The kfunc descriptors include argument prototypes generated while calls are
collected. Some argument classifications need program auxiliary state derived
from referenced maps, such as the arena associated with the program.
This avoids a footgun in get_kfunc_arg_type() checks where we do
validation on whether program has prog->aux->arena and it hasn't been
resolved yet.
check_and_resolve_insns() records used maps and populates that state. It must
remain after bpf_check_btf_info(), which applies kernel-side CO-RE relocations,
so that instruction validation and the program tag observe the relocated
instruction stream.
Move only add_kfuncs() after instruction and resource resolution. Subprogram
discovery and validation remain before the full BTF phase because that phase
needs the complete subprogram layout. Add a short comment describing the
resource resolution phase at the call site.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Reviewed-by: Amery Hung <ameryhung@gmail.com>
Link: https://patch.msgid.link/20260808003938.3486067-4-memxor@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
add_subprog_and_kfunc() combines two operations with different ordering
requirements. Subprogram discovery must precede validation of func_info and
line_info, while kfunc descriptors are only needed by the verifier after its
initial program setup is complete.
Split the helper into add_subprogs() and add_kfuncs() so each operation can be
placed according to its actual dependencies. Keep both calls adjacent and in
their existing phase for now, and add short comments describing their roles.
No functional change is intended for valid programs.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Reviewed-by: Amery Hung <ameryhung@gmail.com>
Link: https://patch.msgid.link/20260808003938.3486067-3-memxor@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
BTF processing is split around subprogram discovery. The first phase gets
program BTF and imports func_info because a BTF-tagged exception callback
may not be referenced by any instruction. Subprogram discovery needs this
metadata to find it.
The later phase validates func_info and line_info against the complete
subprogram table and applies CO-RE relocations. This split breaks a real
dependency cycle rather than merely running the same checks early.
Rename bpf_check_btf_info_early() and check_btf_func_early() to preparation
names that reflect this role. Add short call-site comments to make the two
phases and their responsibilities clear.
No functional change is intended.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Reviewed-by: Amery Hung <ameryhung@gmail.com>
Link: https://patch.msgid.link/20260808003938.3486067-2-memxor@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
After the previous commit bpf_is_reg64() is only used in a context
where destination register's property is queried, and only for
instructions for which insn_def_regno() >= 0. Hence, simplify the
function by:
- removing unused parameters;
- removing code paths considering BPF_JMP{,32} instructions;
- streamlining the condition expressions.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20260807-static-zext-v4-6-b6c270013c77@gmail.com
As reported in the thread [1], the verifier's 32-bit operations zero
extension logic is broken. This logic is responsible for correct
semantics of 32-bit operations on s390 architecture.
According to BPF semantics, operation `w1 += 1` is supposed to zero
extend the upper half of the register `r1`. On s390 the JIT relies on
the verifier emitting explicit zero extension before such operations.
The verifier attempts to minimize the amount of zero extensions
inserted by tracking whether upper halves of the 64-bit registers are
ever used. Previously such tracking worked as follows:
- bpf_reg_state->subreg_def field was set by do_check_insn()
for each operation defining lower but not the upper halves
of the register.
- Whenever an operation reading the whole register was verified,
the verifier checked register's subreg_def and set
bpf_insn_aux_data->zext_dst flag as true via a call to
mark_insn_zext() function.
- After the verification was complete, a special pass
bpf_opt_subreg_zext_lo32_rnd_hi32() extended 32-bit operations
with bpf_insn_aux_data->zext_dst set as true by adding
explicit zero extension.
Note that the logic above relies on bpf_reg_state->subreg_def,
which is a property of a current verifier state.
Before the commit [2] two additional steps happened:
- The verifier tracked upper and lower register halves' liveness as
flags REG_LIVE_READ{32,64} in bpf_reg_state->live.
- The function propagate_liveness() called mark_insn_zext()
in order to transfer the knowledge about which registers have
their upper halves alive (and thus might require zero extension).
The commit [2] removed the two steps described above,
hence making possible a situation like below:
- The register's upper half is set and is used on some verification
path P1 and the register happens not to be marked as precise.
- The checkpoint C is created while processing some instruction
between register initialization and usage.
- On some other verification path P2 the register's upper half is not
initialized and that path ends hitting the checkpoint C.
- In such a case the register's initialization on path P2 would lack
zext_dst mark, making it possible for the program to inject
an arbitrary value in the register's upper half.
This commit replaces subreg_def based logic with computing zext_dst
statically, as a part of the bpf_compute_live_registers() analysis:
- The analysis now tracks usage of upper and lower halves of the
registers separately.
- If some instruction defines a 32-bit subregister, but not the whole
register, *and* the upper half of the register is alive after that
instruction, the instruction is marked as zext_dst.
There is one notable drop in precision: whenever a BPF subprogram is
called, all 64 bits of parameter registers are presumed to be used.
The assumption is that such a drop in precision would not inflict
a noticeable performance penalty.
[1] https://lore.kernel.org/bpf/CAGKGUv=sOuqQtA1Ub-5JXfA4FPosJFYKAQE4B79cK+P1erxqtg@mail.gmail.com/
[2] commit 107e169799 ("bpf: disable and remove registers chain based liveness")
Fixes: 107e169799 ("bpf: disable and remove registers chain based liveness")
Reported-by: Min-gyu Kim <gimm78064@gmail.com>
Reported-by: STAR Labs SG <info@starlabs.sg>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/CAGKGUv=sOuqQtA1Ub-5JXfA4FPosJFYKAQE4B79cK+P1erxqtg@mail.gmail.com/
Link: https://lore.kernel.org/bpf/20260807-static-zext-v4-5-b6c270013c77@gmail.com
Extend compute_live_registers() to track upper and lower register
halves' liveness separately. This is mostly straightforward:
- use/def masks are extended to track 2 bits per register;
- compute_insn_live_regs() is updated to properly track these
2 bits according to the instruction semantics.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20260807-static-zext-v4-4-b6c270013c77@gmail.com
The following patches are going to remove bpf_is_reg64() users from
everywhere except fixups.c, and also make it dependent on functions
local to fixups.c. Move the function before hand to simplify the
review. Non functional change.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20260807-static-zext-v4-3-b6c270013c77@gmail.com
bpf_do_misc_fixups() converts the following address space cast
instructions to 32-bit moves:
- cast from address space 1 (user) to address space 0 (kernel)
- cast from address space 0 (kernel) to address space 1 (user)
iff associated arena map has a BPF_F_NO_USER_CONV flag.
Extract a predicate detecting such instructions for use in the
following patches.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20260807-static-zext-v4-2-b6c270013c77@gmail.com
At the moment there are more callsites that want bpf_verbose_insn() to
not print a newline after the instruction, than callsites that want a
newline. Drop '\n' from disasm.c. Non-functional change.
The changes in bpftool are verified by writing a bpf program using a
variety of instructions and comparing `prog dump xlated` output in the
following modes: plain, opcodes, visual, visual opcodes. The output
before and after the changes is identical.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Quentin Monnet <qmo@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20260807-static-zext-v4-1-b6c270013c77@gmail.com
stack_map_get_build_id_offset() introduced a per-CPU irq_work to defer
mmap_read_unlock() from NMI context, and bpf_find_vma() later reused the
same mmap_unlock_work. Both callers only check whether the work is busy
before taking mmap_lock, so a nested caller can reuse the slot before the
first caller queues it. Two read locks may then be acquired while only one
deferred unlock runs, leaking a read lock and blocking exit_mmap().
Reserve the per-CPU slot before mmap_read_trylock(). Use the same wrapper
in stackmap and bpf_find_vma() so both callers release the reservation on
trylock failure. Keep rejecting the slot while the irq_work remains busy.
Release it after the irq_work callback unlocks the mm.
Fixes: eac9153f2b ("bpf/stackmap: Fix deadlock with rq_lock in bpf_get_stack()")
Reported-by: syzbot+cdd6c0925e12b0af60cc@syzkaller.appspotmail.com
Reported-by: sashiko-bot@kernel.org
Signed-off-by: Sanghyun Park <sanghyun.park.cnu@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Closes: https://syzkaller.appspot.com/bug?extid=cdd6c0925e12b0af60cc
Closes: https://lore.kernel.org/r/20260630033745.B80201F000E9@smtp.kernel.org
Link: https://lore.kernel.org/bpf/20260805031425.2157475-2-sanghyun.park.cnu@gmail.com
In mmio_print_mark(), a raw C cast (struct print_entry *)entry is used to
obtain the print_entry pointer.
Use the standard trace_assign_type() macro instead, matching the usage in
mmio_print_rw() and mmio_print_map().
Link: https://patch.msgid.link/178524301013.56416.9116249028160618790.stgit@devnote2
Assisted-by: Antigravity:gemini-3.6-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Syzkaller reported a storage null-ptr-deref issue after replacing prog.
This occurs in the following scenario:
1. prog A, an empty prog, is attached to a cgrp.
2. prog B uses BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE and calls the
bpf_get_local_storage helper.
3. link_update is called to replace prog A with prog B.
The reason is that __cgroup_bpf_replace fails to alloc and assign the
required cgrp storage for the incoming replacement prog. Consequently,
the new prog inherits an uninit storage, leading to null-ptr-deref panic
when kick the new prog.
Fix this by rejecting a link update if new_prog's cgroup storage is
incompatible with link->prog.
Fixes: 0c991ebc8c ("bpf: Implement bpf_prog replacement for an active bpf_cgroup_link")
Signed-off-by: Pu Lehui <pulehui@huawei.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Reviewed-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/20260728132336.2857800-1-pulehui@huaweicloud.com [0]
Link: https://lore.kernel.org/bpf/f87b53c0-8f00-45a6-82db-8242fa9b143f@huaweicloud.com [1]
Link: https://lore.kernel.org/bpf/20260807104403.1013064-1-pulehui@huaweicloud.com
Merge updates related to system sleep handling and runtime PM, a power
capping update, and power management tools updates for 7.3-rc1:
- Rename the PM core module parameter prefix to "pm" and allow the PM
transition (DPM) watchdog to be disabled by default (Tzung-Bi Shih)
- Fix off-by-one in wakelocks number limit check in the system sleep
sysfs interface (Haowen Tu)
- Remove kernel-doc markings from helper descriptions in the core
hibernation code (Adi Nata)
- Use %pe to print error pointer values in the hibernation core (Ronan
Marchal)
- Fix memory leak in snapshot_write_next() error path (Malaya Kumar
Rout)
- Delay allocating and linking the next swap_map_page in the
hibernation image saving code until another image page actually
needs to be recorded (Haesung Kim)
- Stop setting runtime_error on runtime resume callback failures to
allow drivers to recover from resume issues (Praveen Talari)
- Handle PMU registration failure during probe in the intel_rapl_tpmi
driver (Sumeet Pawnikar)
- Avoid optional imports in intel_pstate_tracer unless they are really
needed (Yousef Alhouseen)
- Add generic CPPC performance display to the cpupower utility, build
and call CPPC information on non-AMD processors, make cpupower print
kernel and hardware frequency information, and add libm to cpupower
for generic CPPC view (Jeremy Linton)
- Remove conditional return with no effect from cpupower (Sang-Heon
Jeon)
* pm-sleep:
PM: sleep: Allow disabling DPM watchdog by default
PM: sleep: Rename module parameters prefix to "pm"
PM: hibernate: swap: defer linking the next map page
PM: hibernate: Fix memory leak in snapshot_write_next() error path
PM: hibernate: Use %pe to print error pointer values
PM: hibernate: Remove kernel-doc markings from helper descriptions
PM: sleep: Fix off-by-one in wakelocks number limit check
* pm-runtime:
PM: runtime: Only set runtime_error on suspend callback failures
* pm-powercap:
powercap: intel_rapl_tpmi: Handle PMU registration failure during probe
* pm-tools:
cpupower: remove conditional return with no effect
tools/power: intel_pstate_tracer: avoid optional imports for help
cpupower: Add libm to cpupower for generic CPPC view
cpupower: Print kernel and hardware frequency information
cpupower: Build and call CPPC information on non-AMD processors
cpupower: Add generic CPPC performance display
Merge cpufreq updates for 7.3-rc1:
- Minor fixes and cleanups in assorted cpufreq drivers (Dan Carpenter,
Guru Das Srinagesh, Haoxiang Li, Karl Mehltretter, Sasha Finkelstein,
and Pan Chuang)
- Fix cpufreq table creation and bios_limits() callback in the Rust
bindings (Priya Bala Govindasamy)
- Add IPQ5210 support to qcom-nvmem driver (Varadarajan Narayanan)
- Adjust the .adjust_perf() cpufreq driver callback to allow the
maximum performance value to be passed to drivers and update the
intel_pstate driver to use it (Rafael Wysocki)
- Set policy->cur to the actual requested frequency in the intel_pstate
driver when the performance policy is used (Rafael Wysocki)
- Simplify HWP handling on Broadwell processors in intel_pstate (Rafael
Wysocki)
- Fix setting minimum P-state at init time in intel_pstate (Rafael
Wysocki)
- Consolidate frequency values computation in intel_pstate and clean up
code in that driver (Rafael Wysocki)
- Add missing kernel-doc desciptions for structure and union members in
the amd-pstate driver (David Vernet)
- Handle missing policy in dynamic EPP callbacks in the amd-pstate
driver (EDAMAMEX)
- Introduce EXPORT_SYMBOL_FOR_PSTATE_UT() to export amd-pstate driver
symbols to the amd-pstate-ut subdriver (K Prateek Nayak)
- Add dynamic EPP as an "energy_performance_preference" mode in
amd-pstate, remove the "amd_dynamic_epp" kernel command line option
and the "dynamic_epp" sysfs attribute, and update the dynamic_epp
documentation accordingly (K Prateek Nayak)
- Add unit tests for CPPC Performance Priority and the "dynamic" EPP
mode in the amd-pstate driver (K Prateek Nayak)
- Set min_limit_freq based on bios_min_perf in amd-pstate and remove
the defensive check for bios_min_perf from it (K Prateek Nayak)
- Fix EPP return type and handle errors in amd-pstate during
initialization, toggle auto_sel in active mode on shared memory
systems, and cache the firmware programmed EPP value (Marco Scardovi)
- Skip tests in amd-pstate-ut if the amd-pstate driver is not in active
use (Qianheng Peng)
- Replace sprintf() with sysfs_emit() in sysfs show in the cpufreq
schedutil governor and fix a self-contradictory comment in
sugov_iowait_apply() (Zhongqiu Han)
- Fix the usage example for the sampling_rate tunable of the ondemand
cpufreq governor in admin-guide (wangxiaodong)
* pm-cpufreq: (40 commits)
cpufreq: imx6q: fix out-of-bounds write when probed more than once
cpufreq: imx6q: fix devres accumulation across driver rebind
rust: cpufreq: Fix temporary write in Registration::bios_limit_callback
rust: cpufreq: Add CPUFREQ_TABLE_END as last table entry in TableBuilder::to_table
cpufreq: intel_pstate: Adjust policy->cur in active mode to policy
cpufreq/amd-pstate: Document missing kernel-doc members
cpufreq/amd-pstate-ut: Add unit test for CPPC Performance Priority
cpufreq/amd-pstate-ut: Add unit test for "dynamic" EPP mode
cpufreq/amd-pstate: Reduce the scope of exported symbols
Documentation/amd-pstate: Update dynamic_epp documentation with new behavior
cpufreq/amd-pstate: Remove "amd_dynamic_epp" cmdline and "dynamic_epp" sysfs
cpufreq/amd-pstate: Add dynamic EPP as an "energy_performance_preference" mode
cpufreq/amd-pstate: Extract platform profile to EPP conversion into a helper
cpufreq/amd-pstate: Remove the defensive check for bios_min_perf
cpufreq/amd-pstate: Set min_limit_freq based on bios_min_perf
cpufreq: apple-soc: Calculate frequency as a 64-bit value
kselftest: cpufreq: Backup and restore governor for sptests
selftests/cpufreq: Remove unnecessary sudo from quick_shuffle()
selftests/cpufreq: Remove unused local variables from switch_show_governor()
cpufreq/amd-pstate: handle missing policy in dynamic EPP callbacks
...
Commit 9c63e84db2 ("sched/core: Disable SD_PREFER_SIBLING on asymmetric
CPU capacity domains") removed the SD_PREFER_SIBLING from the domains with
asymmetric capacity. This was done to avoid spreading tasks to sibling
scheduling groups with less capacity, but this does not happen: checks for
capacity in update_sd_pick_busiest(), sched_balance_find_src_group(), and
sched_balance_find_src_rq() prevent migrations from high- to low-capacity
CPUs if the busiest group is not overloaded.
The cluster topology is a notable example: some systems have scheduling
domains spanning CPUs of asymmetric capacity, grouped into two or more
equal-capacity clusters sharing an L2 cache. When CONFIG_SCHED_CLUSTER is
enabled, SD_PREFER_SIBLING is needed in the domain to spread load across
these clusters.
CPUs with spare capacity, big or small, have always helped overloaded
groups. Once the overloading condition disappears, misfit load will still
be used to move high-utilization tasks to bigger CPUs if they have spare
capacity.
Adding the SD_PREFER_SIBLING flag shifts load balancing in shared-LLC
domains from equalizing the number of idle CPUs to equalizing the number
of running tasks. This enables migrations among clusters from newly-idle
load balance, where the outgoing task is already dequeued but the CPU
has not yet transitioned to idle.
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Tested-by: Christian Loehle <christian.loehle@arm.com>
Tested-by: Andrea Righi <arighi@nvidia.com>
Link: https://patch.msgid.link/20260720-rneri-fix-cas-clusters-v6-6-bb500bf4afd4@linux.intel.com
sched_balance_find_src_rq() avoids selecting a runqueue with a single
running task as busiest if doing so results in migrating the task to a
CPU with less than ~5% of extra capacity. It also unintentionally
prevents migrations between CPUs of identical capacity.
When CONFIG_SCHED_CLUSTER is enabled, load should be balanced across
clusters of CPUs with the same capacity. Allowing migration between CPUs
of identical capacity is necessary to meet this goal.
Use get_actual_cpu_capacity() to reflect architectural capacity as well
as diminished capacity due to hardware or cpufreq pressure. Guard this
check with the sched_cluster_active static key so that systems without
cluster topology are unaffected.
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Tested-by: Christian Loehle <christian.loehle@arm.com>
Tested-by: Andrea Righi <arighi@nvidia.com>
Link: https://patch.msgid.link/20260720-rneri-fix-cas-clusters-v6-5-bb500bf4afd4@linux.intel.com
In domains with asymmetric capacity, identifying misfit load in a
scheduling group is not useful when the destination CPU cannot help (i.e.,
its capacity exceeds the group's maximum CPU capacity by less than ~5%). In
such cases, it also prevents load balance among clusters of equal capacity
when CONFIG_SCHED_CLUSTER is enabled. This happens because
update_sd_pick_busiest() skips candidate groups of type misfit_task if the
destination CPU has similar capacity.
Skipping misfit load accounting in this situation allows the group to be
classified as has_spare or fully_busy and lets load balancing proceed. Keep
marking scheduling groups as overloaded when misfit tasks are present. The
sg_overloaded flag propagates to the root domain and allows bigger CPUs in
it to help via newly idle balance.
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Reviewed-by: Chen Yu <yu.c.chen@intel.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Tested-by: Christian Loehle <christian.loehle@arm.com>
Tested-by: Andrea Righi <arighi@nvidia.com>
Link: https://patch.msgid.link/20260720-rneri-fix-cas-clusters-v6-4-bb500bf4afd4@linux.intel.com
update_sd_pick_busiest() may incorrectly select a fully_busy group as the
busiest group when its per-CPU capacity exceeds that of the destination
CPU. This happens because the type of busiest group is initialized to
group_has_spare and allows the fully_busy group to win the type comparison.
update_sd_pick_busiest() should not choose a candidate scheduling group
with at most one runnable task if its per-CPU capacity is greater than that
of the destination CPU. Such a check already exists, but it is done too
late: after the type comparison, preventing a subsequent fully_busy group
of equal per-CPU capacity from being correctly selected.
Move this check to occur before comparing group types.
Fixes: 0b0695f2b3 ("sched/fair: Rework load_balance()")
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Reviewed-by: Chen Yu <yu.c.chen@intel.com>
Reviewed-by: Tim Chen <tim.c.chen@linux.intel.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Tested-by: Christian Loehle <christian.loehle@arm.com>
Tested-by: Andrea Righi <arighi@nvidia.com>
Link: https://patch.msgid.link/20260720-rneri-fix-cas-clusters-v6-3-bb500bf4afd4@linux.intel.com
The argument sg_overloaded of update_sg_lb_stats() is only consumed when
balancing at the root domain. It only makes sense to update it in such a
case. Commit 3229adbe78 ("sched/fair: Do not compute overloaded status
unnecessarily during lb") updated the logic accordingly but missed the case
in which the root domain has the SD_ASYM_CPUCAPACITY flag. Fix this.
Fixes: 3229adbe78 ("sched/fair: Do not compute overloaded status unnecessarily during lb")
Reported-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Tested-by: Christian Loehle <christian.loehle@arm.com>
Tested-by: Andrea Righi <arighi@nvidia.com>
Link: https://patch.msgid.link/20260720-rneri-fix-cas-clusters-v6-2-bb500bf4afd4@linux.intel.com
When picking a busiest CPU with only one running task, the function
sched_balance_find_src_rq() skips candidate CPUs if the destination CPU has
less than ~5% extra capacity. This condition only holds if all the SMT
siblings of a CPU are idle.
SMT siblings share the computing resources of a physical core and this
results in reduced capacity if more than one sibling is busy.
Skipping a CPU as described would prevent the load balancer from pulling
tasks from a scheduling group previously and correctly identified as
group_smt_balance (i.e., one with more than one task running).
Do not skip a candidate CPU of similar capacity if it has busy SMT
siblings.
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Tested-by: Andrea Righi <arighi@nvidia.com>
Link: https://patch.msgid.link/20260720-rneri-fix-cas-clusters-v6-1-bb500bf4afd4@linux.intel.com
find_new_ilb() selects the first idle housekeeping CPU without
considering whether another thread is running on the same physical core.
On an SMT system, the idle load balancer can therefore activate both
siblings even when another housekeeping CPU has an entirely idle core.
On most SMT systems, this is not problematic because the idle load
balancer is a short-lived activity and the transient wakeup of a sibling
has negligible performance impact.
However, this can be particularly costly on NVIDIA Olympus cores used in
Vera. Briefly activating an otherwise idle sibling can reduce the
performance available to the other sibling and this effect does not
necessarily end once the activated sibling becomes idle: after the ILB
finishes and its CPU enters WFI, full single-thread performance is
restored only after the sibling has remained idle for a qualification
interval (10 Ki cycles on the tested Vera system). Repeated short
sibling wakeups can therefore sustain the interference even with little
actual overlap.
Prevent this by preferring an idle housekeeping CPU whose entire SMT
core is idle. Retain the first idle CPU as a fallback when no fully idle
core is available, so NOHZ balancing continues to make forward progress.
Once a partially busy core has been examined, skip its remaining SMT
siblings to avoid repeating the core-idle check on wide SMT systems.
Tests performed using an ad hoc GEMM benchmark running one CPU-intensive
task per SMT core within its CPU affinity mask improved from
approximately 6.2 TFLOP/s to 9.4 TFLOP/s.
Note that this preference may wake a fully idle physical core instead of
using an idle sibling of an active core, potentially increasing ILB
wakeup latency or energy consumption on some architectures. It may also
scan additional CPUs before selecting the one to run the ILB. The
selection falls back to the first idle CPU when no fully idle SMT core
is available. Non-SMT systems continue to select the first idle
housekeeping CPU.
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Mete Durlu <meted@linux.ibm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260804151324.918020-1-arighi@nvidia.com
perf_group_detach() handles leader and sibling detach differently. When the
group leader is detached, all siblings are promoted to singleton events and
their group_leader pointer is reset to themselves. When a sibling is
detached, it is removed from the leader's sibling_list, but its
group_leader pointer is left pointing at the old leader.
That is harmless when the sibling is being closed and freed immediately, as
in the DETACH_DEAD path. It is not safe when the sibling is detached but
kept alive, such as during CPU hotplug with DETACH_GROUP. In that case the
sibling is removed from the context, while its file descriptor can still
keep it alive.
A typical failing sequence is:
- A group contains leader L and sibling S.
- CPU hot-unplug detaches S with DETACH_GROUP, removing it from
L->sibling_list but leaving S->group_leader == L.
- L is later closed and freed.
- A PERF_IOC_FLAG_GROUP ioctl on S follows S->group_leader and
dereferences the freed leader.
This was reproduced by running the perf event fuzzer, CPU hotplug, and a
stress workload concurrently:
Unable to handle kernel paging request at virtual address 006b6b6b6b6b6cdb
CPU: 2 PID: 12489 Comm: perf_fuzzer 6.18.7 PREEMPT
pc : perf_ioctl+0x34c/0xc68
x20: ffffff89a3fa2c70 x8 : 6b6b6b6b6b6b6b6b
Code: 943c4a0e 340047a0 f9404a94 f9411e88 (f940b908)
Call trace:
perf_ioctl+0x34c/0xc68 (P)
__arm64_sys_ioctl+0xa0/0xf4
invoke_syscall+0x58/0xe4
el0_svc_common+0xa8/0xdc
do_el0_svc+0x1c/0x28
el0_svc+0x40/0xc0
el0t_64_sync_handler+0x68/0xdc
el0t_64_sync+0x1c4/0x1c8
The fault happened in perf_ioctl(), where perf_event_for_each() follows
the stale group_leader pointer and perf_event_for_each_child() then
dereferences the freed leader's context.
Fix the use-after-free by promoting the detached sibling to a singleton.
Also fix __event_disable() cgroup accounting and event state change.
Fixes: 8a49542c05 ("perf_events: Fix races in group composition")
Assisted-by: PatchWise:gpt-5.5
Signed-off-by: Aditya Chillara <aditya.chillara@oss.qualcomm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260807-fix-group-leader-uaf-v3-1-b0c2310c9a0d@oss.qualcomm.com
perf_event_remove_on_exec() sets remove-on-exec events to the EXIT state
and detaches their group relationships. The event's file descriptor can
remain open, however, and perf_event_open() currently accepts that event
as a group leader because its early validation rejects only REVOKED and
DEAD events.
A new sibling can consequently be linked to the detached leader. When
the leader is closed, perf_group_detach() observes that its
PERF_ATTACH_GROUP bit is already clear and skips the new sibling. The
sibling then retains a group_leader pointer to the freed event.
Reject group leaders in the EXIT state. Perform the check while holding
the shared context mutex so that an exec in the target task cannot detach
the leader between validation and group attachment.
[peterz: make the earlier test fully consistent]
Fixes: 037a3c43ed ("perf/core: Detach event groups during remove_on_exec")
Assisted-by: Codex:gpt-5.6-sol
Signed-off-by: Kyle Zeng <kylebot@openai.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260806205655.75722-1-kylebot@openai.com
queued_spin_unlock() gates its contended_release trace call behind a
static branch, so a NOP sits on the unlock path even while the
tracepoint is disabled. Removing that requires replacing the unlock
implementation only while contended_release is enabled, which needs a
callback when the tracepoint is toggled.
Convert contended_release to TRACE_EVENT_FN() and add weak no-op
arch_contended_release_trace_reg()/arch_contended_release_trace_unreg()
hooks.
The default hooks are empty, so this is a no-op until an architecture
overrides them.
No functional change intended.
Signed-off-by: Dmitry Ilvokhin <d@ilvokhin.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juergen Gross <jgross@suse.com>
Link: https://patch.msgid.link/1c2fcccfb584c075c02890c484f22c76a1948bf1.1785778551.git.d@ilvokhin.com
Unlike mutex and rw_semaphore, qspinlock has no owner field, so "perf
lock contention --lock-owner" cannot attribute a contended spinlock to
its holder. The waiter-side contention_begin event records that a
spinlock is contended, but not by whom. Firing contended_release in the
holder's context at unlock is the only way to capture the holder of a
contended spinlock.
Combine the contention check, trace call and release in an out-of-line
queued_spin_release_traced() so the compiler need not preserve the lock
pointer in a callee-saved register across the call.
The check in queued_spin_unlock() is paid on every unlock, even while
the tracepoint is disabled: a static-branch NOP on x86_64, and a few
more instructions to manage a stack frame elsewhere. Gate it behind
CONFIG_QUEUED_SPINLOCKS_TRACE_CONTENDED_RELEASE (default n) so nobody
pays for a tracepoint they do not use. Sleeping locks fire
contended_release regardless.
On x86 this generic path is used only with PARAVIRT_SPINLOCKS=n (e.g.
defconfig). PARAVIRT_SPINLOCKS=y kernels keep the paravirt static_call
unlock and are wired up separately.
All below are with the QUEUED_SPINLOCKS_TRACE_CONTENDED_RELEASE option
enabled.
_raw_spin_unlock(), x86_64 defconfig, GCC 11, tracepoint compiled in but
disabled. The unlock is the single 'movb'. The only instruction added to
the executed path is the 2-byte static-branch NOP. The CALL to the
traced helper and the JMP back are emitted out of line and are reached
only once the static branch is patched on:
endbr64 ; 4 bytes
xchg %ax,%ax ; 2 static-branch NOP
; (added)
movb $0x0,(%rdi) ; 3 unlock (single store)
A: decl %gs:__preempt_count ; 7
je B ; 2
jmp __x86_return_thunk ; 5
call queued_spin_release_traced ; 5 out of line, reached
; only when the
; tracepoint is on
jmp A ; 2 (added)
B: call __SCT__preempt_schedule ; 5
jmp __x86_return_thunk ; 5
Baseline is the same stream without the NOP and the out-of-line
CALL/JMP: 31 bytes vs 40 (+9 bytes).
Binary size impact on x86_64, defconfig: +680 bytes (+0.00%), since all
standard configs out-of-line unlock. Architectures with inlined unlock
(s390 (always), csky and loongarch (both when !PREEMPTION)) will see a
bigger increase in binary size.
On the same path (x86_64, PARAVIRT_SPINLOCKS=n) with the tracepoint
disabled, a _raw_spin_unlock()-heavy nginx workload [1] shows no
measurable difference between baseline and patched kernels in
throughput, latency, cycles, instructions, IPC, or L1 instruction-cache
misses (kernel and total): all deltas stay within run-to-run noise.
Unlike x86, on arm64 the frame setup code (STP, MOV and LDP) lands on
the executed path in addition to static-branch NOP. Binary size impact
on arm64, defconfig: +932 bytes (+0.00%).
The _raw_spin_unlock()-heavy nginx workload reflects the larger hot
path: L1 instruction-cache misses rise ~1.4% (kernel and total) and
instruction count ~0.4%, consistent with the per-unlock frame.
cpu_cycles, throughput and latency show no measurable change and are
within run-to-run noise.
Architectures with fully custom qspinlock implementations (e.g.
PowerPC) are not covered by this change.
[1]: https://lore.kernel.org/all/aiphFXe_TPNPxZ_n@shell.ilvokhin.com/
Signed-off-by: Dmitry Ilvokhin <d@ilvokhin.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juergen Gross <jgross@suse.com>
Link: https://patch.msgid.link/0d998e22a0c595f670cfc6725bb683323aced5cb.1785778551.git.d@ilvokhin.com
A task performing a custom private hash resize can remain blocked in
uninterruptible sleep indefinitely. The hung-task detector reports:
INFO: task futex-resizer:314 blocked for more than 10 seconds.
task:futex-resizer state:D stack:14824 pid:314 tgid:312 ppid:311
Call Trace:
__schedule+0x521/0xf30
schedule+0x22/0xa0
futex_hash_allocate+0x3db/0x490
__do_sys_prctl+0x6f5/0xbd0
do_syscall_64+0xf9/0x530
entry_SYSCALL_64_after_hwframe+0x77/0x7f
Kernel panic - not syncing: hung_task: blocked tasks
futex_pivot_pending() allows the resize request to continue when
either no replacement hash is pending (hash_new == NULL) or the current
hash reference count has reached zero.
After the final-reference wake, another futex task can complete the
pivot between the two observations:
T1 T2
futex_hash_allocate()
wait_var_event(mm, ...)
futex_pivot_pending(mm)
hash_new != NULL
futex_hash()
futex_ref_get(old) -> false
futex_pivot_hash(mm)
hash_new = NULL
__futex_pivot_hash(mm, new)
rcu_assign_pointer(hash, new)
fph = rcu_dereference(hash) /* new */
futex_ref_is_dead(fph) -> false
schedule()
The pivot changes the state from hash_new != NULL with a dead current
hash to hash_new == NULL with a live current hash. Because
futex_pivot_pending() reads hash_new and hash without serialization,
the resize task can observe hash_new in the pre-pivot state and hash in
the post-pivot state, causing futex_pivot_pending() to return false even
though the pivot has completed. The task then goes to sleep after the
wakeup has already been consumed.
Serialize state reads in futex_pivot_pending() using futex_mm_phash::lock.
This guarantees that futex_pivot_pending() observes hash_new and hash
atomically, eliminating the race condition.
Fixes: bd54df5ea7 ("futex: Allow to resize the private local hash")
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Yao Kai <yaokai34@huawei.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260804125530.3933754-1-yaokai34@huawei.com
- Fix BPF verifier to preserve full pointer state for commuted
scalar += pointer arithmetic (Yiyang Chen, Eduard Zingerman)
- Fix a use-after-free of request sockets in the BPF TCP
iterator batching (Jose Fernandez)
- Fix a use-after-free of sk_redir in the BPF sockmap send
verdict path (Chengfeng Ye)
- Fix a netns reference imbalance in the BPF conntrack kfuncs
(Chengfeng Ye)
- Fix bpf_get_fsverity_digest() dynptr assumptions and silent
digest truncation (Eric Biggers)
- Fix bpf_tcp_{gen,check}_syncookie to check sk_state before
sk_protocol to make sure it is a full socket (Luxiao Xu)
- Fix rqspinlock to reset the tail when preserving the queue
on deadlock (Kumar Kartikeya Dwivedi)
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
-----BEGIN PGP SIGNATURE-----
iIsEABYKADMWIQTFp0I1jqZrAX+hPRXbK58LschIgwUCanXeCRUcZGFuaWVsQGlv
Z2VhcmJveC5uZXQACgkQ2yufC7HISIMfLQD9EWzi5MVBTcvg0XsHY1GZZBZUpfwo
VCrfPm9vHAVuqQ0A/0D9vWVRf1UEk9ccn+ebVKPuTuydGnDRR0Qovuca4gQF
=jZY8
-----END PGP SIGNATURE-----
Merge tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Pull BPF fixes from Daniel Borkmann:
- Fix BPF verifier to preserve full pointer state for commuted
scalar += pointer arithmetic (Yiyang Chen, Eduard Zingerman)
- Fix a use-after-free of request sockets in the BPF TCP iterator
batching (Jose Fernandez)
- Fix a use-after-free of sk_redir in the BPF sockmap send verdict
path (Chengfeng Ye)
- Fix a netns reference imbalance in the BPF conntrack kfuncs
(Chengfeng Ye)
- Fix bpf_get_fsverity_digest() dynptr assumptions and silent
digest truncation (Eric Biggers)
- Fix bpf_tcp_{gen,check}_syncookie to check sk_state before
sk_protocol to make sure it is a full socket (Luxiao Xu)
- Fix rqspinlock to reset the tail when preserving the queue
on deadlock (Kumar Kartikeya Dwivedi)
* tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
rqspinlock: Reset tail when preserving queue on deadlock
bpf: Check sk_state before sk_protocol in bpf_tcp_*_syncookie
fsverity: Fix silent truncation in bpf_get_fsverity_digest()
fsverity: Fix bpf_get_fsverity_digest() dynptr assumptions
bpf: tcp: Fix use-after-free in bpf_iter_tcp_established_batch()
bpf: Fix netns reference imbalance in conntrack kfuncs
bpf, sockmap: Fix sk_redir use-after-free in send verdict
selftests/bpf: Cover commuted pointer state propagation
bpf: Propagate untrusted pointer state in commuted arithmetic
bpf: Preserve pointer state for commuted arithmetic
bpf: Simplify sanitize_err() signature
A BPF_LOAD_ACQ is not rewritten to a BPF_PROBE_MEM load by the verifier,
unlike a regular BPF_LDX, so the JIT emits a plain load with no exception
table entry and a fault panics the kernel instead of being handled.
Reject the source pointer types that a BPF_LDX would have had that fault
protection applied to, i.e. the ones bpf_convert_ctx_accesses() turns
into BPF_PROBE_MEM: a bare PTR_TO_BTF_ID, PTR_TO_BTF_ID | PTR_UNTRUSTED,
PTR_TO_BTF_ID | MEM_ALLOC | PTR_UNTRUSTED and PTR_TO_MEM | MEM_RDONLY |
PTR_UNTRUSTED.
This is reachable e.g. by loading ->mm out of a trusted task_struct
yields an untrusted pointer to mm_struct, and it is NULL for a kernel
thread:
[...]
SEC("tp_btf/sched_switch")
int BPF_PROG(demo, bool preempt, struct task_struct *prev,
struct task_struct *next)
{
struct mm_struct *mm = next->mm; /* untrusted */
out_ldx = (__u64)mm->pgd; /* BPF_LDX */
out_acq = load_acquire(&mm->pgd); /* BPF_LOAD_ACQ */
return 0;
}
[...]
Both dereference the same pointer, but only the BPF_LDX is protected
(x86-64 JIT, jump targets shown prog-relative):
[...]
; out_ldx = (__u64)mm->pgd;
17: movq $-10485760, %r10
1e: movq %rsi, %r11
21: addq $184, %r11
28: subq %r10, %r11
2b: movabsq $140737498841088, %r10
35: cmpq %r10, %r11
38: ja 0x3e <-- kernel addr?
3a: xorl %edi, %edi <-- no: dst = 0, skip the load
3c: jmp 0x45
3e: movq 184(%rsi), %rdi <-- yes: load + extable entry
[...]
; load_acquire(&mm->pgd)
53: movq %rsi, %rdi
56: movq 184(%rdi), %rax <-- no check, no extable entry
[...]
Note that BPF_PROBE_MEM is not visible in a bpftool xlated dump, as
bpf_insn_prepare_dump() rewrites it back to BPF_MEM.
A PTR_TRUSTED pointer is deliberately not on the list. Such a load is
not converted either, but it does not need to be, since the pointer is
guaranteed live, so load-acquire from it stays allowed.
The check is gated on BPF_LOAD_ACQ so that atomic RMW and store-release
error messages are unchanged; writes (RMW / store-release) to such
pointers are already rejected elsewhere, so only load-acquire needs this.
Fixes: 880442305a ("bpf: Introduce load-acquire and store-release instructions")
Reported-by: STAR Labs SG <info@starlabs.sg>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20260806201047.333389-1-daniel@iogearbox.net
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
These comments describing the page allocator are out of date:
- __alloc_pages() is no longer a public API and has no business being
described outside of mm/.
- The `wait` variable is gone.
It may be out of date for other reasons too but this patch is just
fixing the issues that stood out.
To fix it:
- Instead of referring to a specific function, instead to "the page
allocator"
- Completely drop out-of-date details of that function's internal
behaviour, since they were irrelevant anyway.
Link: https://lore.kernel.org/20260715-spin-trylock-followup-v3-2-fc4d246f705d@google.com
Signed-off-by: Brendan Jackman <jackmanb@google.com>
Suggested-by: Zi Yan <ziy@nvidia.com>
Link: https://lore.kernel.org/all/DJP11T5V7BDW.2FZZZ8R6LOY4I@nvidia.com/
Reviewed-by: Zi Yan <ziy@nvidia.com>
Reviewed-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Acked-by: Tejun Heo <tj@kernel.org>
Cc: David Hildenbrand <david@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Lorenzo Stoakes <ljs@kernel.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Michal Koutný <mkoutny@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Waiman Long <longman@redhat.com>
Cc: Brendan Jackman <brendan.jackman@linux.dev>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Currently, the destruction of the waiter queue is suppressed for
rqspinlock in cases where a deadlock is detected. Deadlock checks happen
relatively frequently (on entry for AA, within 1ms for ABBA), and waiter
threads may not be involved in locking scenarios involving deadlocks.
Thus, it is useful to not flush the queue and let other waiters take a
stab at acquiring the lock after we detect a deadlock and exit.
However, we need to follow the same logic as what we did previously for
the waitq_timeout label: reset the tail, and if we cannot, signal the
next waiter appropriately. In case of deadlocks, this signal would just
mark the MCS node as unlocked, and in case of timeouts, it would signal
RES_TIMEOUT_VAL. The difference thus is in the value propagated, which
decides whether the queue remains active or gets flushed.
Not doing the tail reset, and waiting for the next waiter can lead to
cases where we are the final waiter, and thus no next waiter arrives,
leading to intermittent stalls in this path. Once the next waiter does
join, we will be unblocked. In the theoretical case when the next waiter
never joins, we risk stalling indefinitely.
This can only happen for ABBA deadlocks, since entry into the wait queue
is guarded with AA checks. A precise sequence of executions leading up
to this scenario can be:
CPU 0 holds lock A.
CPU 1 holds lock B.
CPU 2 attempts lock B, becomes the pending waiter for B.
CPU 0 attempts lock B. B has locked+pending bits set, thus CPU 0 queues.
CPU 1 attempts lock A.
CPU 0 detects an ABBA deadlock.
Once deadlock detection happens for CPU 0, it will sit waiting for the
next waiter in the queue to populate node->next, which will experience
delays until such a waiter arrives.
Fix this by adjusting the logic for the check for deadlocks preceding
the waitq_timeout label. It would make sense to consolidate code for
both cases and use 'ret' to distinguish the value being propagated, but
that is left as an exercise for a future refactoring task to avoid diff
noise in this patch.
Fixes: 7bd6e5ce5b ("rqspinlock: Disable queue destruction for deadlocks")
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://patch.msgid.link/20260802021759.1139457-1-memxor@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Add an optional reclaim callback to struct dmem_cgroup_region. When
dmem.max is set below the current usage of a cgroup pool, the new limit
is applied immediately (so that concurrent allocations are throttled
while reclaim is in progress) and then the driver is asked to evict
memory to bring usage back below the limit.
Reclaim is attempted up to a bounded number of times. No error is
returned to userspace if usage remains above the limit after reclaim,
and a pending signal will abort the reclaim loop early. This matches
the behavior of memory.max in the memory cgroup controller.
Also honor O_NONBLOCK so that if that flag is set during the
max value write, no reclaim is initiated. The idea is to avoid
charging the reclaim cost to the writer of the max value.
v2:
- Write max before reclaim is attempted (Maarten)
- Let signals abort the reclaim without error (Maarten)
- If a new max value is written with the O_NONBLOCK flag,
reclaim is not attempted (Maarten)
- Extract region from the pool parameter rather than
passing it explicitly to set_resource_xxx().
v3:
- Use an rw_semaphore (unregister_sem) to protect reclaim callbacks
against concurrent region unregistration: readers (reclaim) hold the
read side; dmem_cgroup_unregister_region() takes the write side to
drain in-flight callbacks before returning. (Sashiko-bot)
v5:
- Rebased on the introduction of struct dmem_cgroup_init.
- Use nonblock=true in reset_all_resource_limits() to avoid sleeping
inside rcu_read_lock() in dmemcs_offline(). (Sashiko-bot)
- Compare usage against the truncated limit value stored in cnt.max,
not the original u64. (Sashiko-bot)
- Use a DMEM_MAX_RECLAIM_RETRIES (16) retry budget instead of 5, matching
the memcg controller's MAX_RECLAIM_RETRIES. Only -ENOSPC (no progress)
counts against the retry budget; other errors terminate the loop
immediately.
v6:
- Fix dmem_cgroup_ops->reclaim docstring: -ENOSPC does not stop reclaim
immediately but is retried up to DMEM_MAX_RECLAIM_RETRIES times; only
other negative errors terminate the loop. (Sashiko-bot)
v7:
- Replace the per-region rw_semaphore with a static SRCU domain
(dmemcg_srcu). SRCU is a better fit than rwsem for this use: it
avoids the per-region lock overhead on every reclaim call, and
synchronize_srcu() at unregister time is a rare operation. (Maarten)
- Trim in-function comments to focus on what rather than how.
Assisted-by: GitHub_Copilot:claude-sonnet-4.6
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Tested-by: Thadeu Lima de Souza Cascardo <cascardo@igalia.com>
Link: https://patch.msgid.link/20260725100036.2372-4-thomas.hellstrom@linux.intel.com
Signed-off-by: Maarten Lankhorst <dev@lankhorst.se>
Replace the bare u64 size argument to dmem_cgroup_register_region() and
drmm_cgroup_register_region() with a const struct dmem_cgroup_init *
pointer. The struct currently carries only the size field, but using a
struct makes the API extensible: future callers can supply additional
initialization parameters without adding more positional arguments.
Update all in-tree callers (amdgpu, xe) to use a compound-literal
initializer.
v5:
- Commit introduced.
Assisted-by: GitHub_Copilot:claude-sonnet-4.6
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Tested-by: Thadeu Lima de Souza Cascardo <cascardo@igalia.com>
Link: https://patch.msgid.link/20260725100036.2372-3-thomas.hellstrom@linux.intel.com
Acked-by: Dave Airlie <airlied@redhat.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Maarten Lankhorst <dev@lankhorst.se>
param_set_charp() stores charp parameters in allocated memory after slab is
available, and releases the previous value when the parameter is updated.
The previous value is released before the replacement allocation succeeds.
If kmalloc_parameter() fails, the setter returns -ENOMEM with the parameter
left as NULL.
Failing zswap's compressor update before zswap is initialized can later
trigger:
BUG: kernel NULL pointer dereference, address: 0000000000000000
RIP: 0010:strcmp+0x10/0x30
Call Trace:
zswap_setup+0x3b1/0x490
zswap_enabled_param_set+0x5b/0xa0
param_attr_store+0x93/0xe0
module_attr_store+0x1c/0x30
kernfs_fop_write_iter+0x116/0x1f0
Allocate and copy the replacement first, then replace the parameter value
only after allocation succeeds.
Fixes: e180a6b775 ("param: fix charp parameters set via sysfs")
Cc: stable@vger.kernel.org
Signed-off-by: Jiacheng Yu <yujiacheng3@huawei.com>
Reviewed-by: Petr Pavlu <petr.pavlu@suse.com>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
In elf_validity_cache_sechdrs, section sizes and offsets are validated,
unless the section type is SHT_NULL or SHT_NOBITS.
Later, elf_validity_cache_secstrings and elf_validity_cache_index_str
access the section name table (.shstrtab) and symbol string table
(.strtab) headers without first ensuring that their types are
SHT_STRTAB. If a section type is SHT_NULL or SHT_NOBITS, sh_offset has
not been validated and may reference out-of-bounds memory when
dereferenced in elf_validity_cache_secstrings or
elf_validity_cache_strtab.
Validate that both string section headers are of type SHT_STRTAB before
caching them.
Cc: stable@vger.kernel.org
Signed-off-by: Thiébaud Weksteen <tweek@google.com>
Reviewed-by: Aaron Tomlin <atomlin@atomlin.com>
Reviewed-by: Petr Pavlu <petr.pavlu@suse.com>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Disabling preemption or local IRQs keeps the current CPU in an RCU
read-side critical section, but in_rcu_cs() does not account for either
state. The verifier therefore rejects safe kptr accesses and invalidates
pointers when another RCU source ends.
Include preemption-disabled and IRQ-disabled state in in_rcu_cs().
Invalidate RCU-protected pointers on RCU unlock, preempt enable, or IRQ
restore only after the final protection ends.
Signed-off-by: Ning Ding <dingning04@gmail.com>
Link: https://lore.kernel.org/bpf/20260805233940.3966981-2-dingning04@gmail.com
[ kkd: Simplify was_in_rcu_cs on spin unlock and adjust the selftest. ]
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Use strcmp() instead of strlen()+memcmp() to compare module names in
kmod_dup_request_lookup(), since all strings are NUL-terminated.
Reviewed-by: Aaron Tomlin <atomlin@atomlin.com>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Use scope-based cleanup helpers for kmod_dup_mutex and kmod_req to shorten
the code and to clarify where the lock is taken in
kmod_dup_request_exists_wait().
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
The kmod dups code preallocates kmod_dup_req before taking kmod_dup_mutex
to avoid allocating memory while holding the lock. This provides little
benefit, since the allocation is fast and can safely be done under the
lock. On the other hand, it leads to unnecessary allocations when the
request turns out to be a duplicate and slightly complicates the code.
Allocate kmod_dup_req only when needed and introduce a helper function
alloc_kmod_req() to initialize the structure.
Reviewed-by: Aaron Tomlin <atomlin@atomlin.com>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
The kmod dups code uses RCU to ensure that a kmod_dup_req instance is freed
only after it is no longer referenced. When releasing an instance, the
kmod_dup_request_delete() function removes the kmod_dup_req from the
dup_kmod_reqs list, waits via synchronize_rcu() and finally frees it.
However, this doesn't work correctly because parallel users referencing the
instance in kmod_dup_request_exists_wait() don't enter an RCU read-side
critical section. This can result in a use-after-free.
The kmod_dup_request_exists_wait() function may need to hold a valid
reference to a kmod_dup_req instance across a blocking wait until the
corresponding modprobe command completes. This makes it unsuitable for RCU.
Fix the issue by changing the lifecycle management of kmod_dup_req to use
reference counting.
Fixes: 8660484ed1 ("module: add debugging auto-load duplicate module support")
Reviewed-by: Aaron Tomlin <atomlin@atomlin.com>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
When kmod_dup_request_announce() announces the completion of
a request_module() call to duplicate waiters, it queues a work item to
invoke kmod_dup_request_complete(), and only that function calls
complete_all().
This adds an arbitrary delay that is unnecessary and provides little
benefit. Call complete_all() directly from kmod_dup_request_announce()
instead.
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Historically, various parameter-handling code kept pointers into
module::args, most notably the charp support. However, in 2009,
commit e180a6b775 ("param: fix charp parameters set via sysfs") changed
charp parameters to kstrdup() the input string as well. As a result,
module::args now mostly wastes memory.
The last users that still pointed into module::args have now been cleaned
up, so remove this data.
Reviewed-by: Aaron Tomlin <atomlin@atomlin.com>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
module_total_size() returns unsigned int but uses a signed int
accumulator. While the result is numerically correct, the type
mismatch is misleading.
Change the accumulator to unsigned int to match the return type.
Signed-off-by: Naveen Kumar Chaudhary <naveen.osdev@gmail.com>
Reviewed-by: Sami Tolvanen <samitolvanen@google.com>
[ppavlu: correct the commit title]
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Both try_add_failed_module() and kmod_dup_request_exists_wait() use
memcpy() with strlen() to copy module names into fixed-size
char[MODULE_NAME_LEN] buffers. Neither performs a bounds check on the
copy. Current callers always pass names originating from
mod->name (itself char[MODULE_NAME_LEN]), so this is not exploitable
today. However both functions accept a plain const char * with no
documented length contract, making them latent buffer overflows if a
future caller passes a longer string.
Replace memcpy() with strscpy() in both sites, which bounds the copy
to MODULE_NAME_LEN and always NUL-terminates.
Signed-off-by: Naveen Kumar Chaudhary <naveen.osdev@gmail.com>
Reviewed-by: Petr Pavlu <petr.pavlu@suse.com>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
The comment wrongly references to /sys/modules/XYZ/parameters/ directory
instead of /sys/module/XYZ/parameters/. Fix it.
Signed-off-by: Zenghui Yu <zenghui.yu@linux.dev>
Reviewed-by: Aaron Tomlin <atomlin@atomlin.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
The symbol lookup code assumes the queried address resides in either
MOD_TEXT or MOD_INIT_TEXT. This breaks for addresses in other module
memory regions (e.g. rodata or data), resulting in incorrect upper
bounds and wrong symbol size.
Select the module memory region the address belongs to instead of
hardcoding text sections. Also initialize the lower bound to the start
of that region, as searching from address 0 is unnecessary.
Cc: stable@vger.kernel.org
Signed-off-by: Stanislaw Gruszka <stf_xl@wp.pl>
Reviewed-by: Petr Pavlu <petr.pavlu@suse.com>
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
The verifier currently keeps a hard-coded list of kfuncs that may be
called while holding a bpf_spin_lock. With KF_SPINLOCK_SAFE available,
retaining this list creates two sources of truth and requires verifier
changes whenever another lock-safe kfunc is added.
Mark every kfunc currently accepted by kfunc_spin_allowed() with
KF_SPINLOCK_SAFE. This covers the graph, numeric iterator, resource
spin lock, arena, and stream kfuncs.
Remove the obsolete category checks and make kfunc_spin_allowed() rely
solely on the kfunc registration metadata. This preserves the behavior
of existing kfuncs while using the same mechanism for built-in and
module kfuncs.
Signed-off-by: Kaitao Cheng <chengkaitao@kylinos.cn>
Acked-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/20260805153340.34776-3-kaitao.cheng@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Introduce the KF_SPINLOCK_SAFE kfunc metadata flag in BTF so kfuncs may
be explicitly marked as safe to call while holding bpf_spin_lock.
Allow kfuncs defined in kernel modules to be marked with KF_SPINLOCK_SAFE.
Example: BTF_ID_FLAGS(func, $kfunc_name, KF_SPINLOCK_SAFE)
Signed-off-by: Kaitao Cheng <chengkaitao@kylinos.cn>
Acked-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/20260805153340.34776-2-kaitao.cheng@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
DECLARE_EVENT_CLASS emitted __bpf_trace_btf_ids_<call> through
BTF_ID_LIST_GLOBAL, i.e. a global symbol named after the event class.
The class name is not unique across the kernel, so the symbol
multiply-defines whenever two translation units instantiate the same
class.
Switch to the file-local BTF_ID_LIST: the list is reached
only through the event_class_<call>.btf_ids pointer, initialised in the
same unit, so tracefs readers never reference the symbol by name and
resolve_btfids still fills the now-local .BTF_ids entries.
The handcrafted syscall classes are the one cross-unit consumer: give
them their own local BTF_ID_LIST rather than importing the generated
sys_{enter,exit} lists.
Link: https://patch.msgid.link/20260730-b4-fix_btf_tracefs-v2-1-6b66da8dc103@meta.com
Fixes: eadc0725ab ("tracing: Expose tracepoint BTF ids via tracefs")
Reported-by: Mark Brown <broonie@kernel.org>
Closes: https://lore.kernel.org/all/ff58b01c-3f5e-4d55-be82-609d2faaf12e@sirena.org.uk/
Signed-off-by: Mykyta Yatsenko <yatsenko@meta.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
bloom_map_alloc() has two 32-bit-specific problems when the computed
bitmap reaches the U32_MAX fallback case.
First, BITS_TO_BYTES(U32_MAX) is evaluated with 32-bit arithmetic. The
addition performed by DIV_ROUND_UP wraps, so the map allocates only the
fixed-size bloom filter object while keeping bitset_mask == U32_MAX.
Subsequent updates can then write past the allocated object.
Second, fixing only the allocation size is not sufficient. The bloom hash
is a u32, but set_bit() takes a signed long bit number and x86 test_bit()
eventually feeds the index to variable_test_bit(long, ...). On 32-bit
kernels, hashes in [0x80000000, U32_MAX] therefore become negative bit
offsets. x86 bt/bts with a memory operand interpret those offsets relative
to the supplied base, so a map with bitset_mask == U32_MAX can read or
write before bloom->bitset even after allocating the full 512 MiB bitmap.
Keep the U32_MAX fallback, but split each hash into a word pointer and an
in-word bit number before calling test_bit() or set_bit(). The bitops
argument is then always in [0, BITS_PER_LONG - 1], while BIT_WORD(h) still
selects the intended word in the full bitmap.
Compute the bitset size from (u64)bitset_mask + 1 before passing the final
size to bpf_map_area_alloc(). This fixes the original under-allocation and
keeps the allocated storage consistent with the addressable bitset.
Exploitation note: local privilege escalation is possible on a 32-bit x86
kernel using the under-allocation bug from a binary with CAP_BPF.
Fixes: 9330986c03 ("bpf: Add bloom filter map implementation")
Signed-off-by: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/bpf/20260805060228.2703051-1-Jeremy.Jean@oss.cyber.gouv.fr
Assisted-by: Codex:gpt-5
When CONFIG_FUNCTION_ERROR_INJECTION is disabled, a sleepable tracing prog
is allowed to attach to '__x64_'-alike prefix symbols.
It is because the verifier does not verify whether the symbol is a kernel
function or a bpf prog. That said, a sleepable tracing prog is allowed to
attach to a bpf prog target whose name has '__x64_'-alike prefix.
For example, a sleepable fentry prog attaches to a '__x64_sys_nop' XDP
prog, and copies buffer from a user pointer with bpf_copy_from_user()
helper. After attaching the XDP prog to lo interface, the kernel BUG
could be triggered by 'ping -c 1 -W 1 127.0.0.1':
[ 3.460756] BUG: sleeping function called from invalid context at kernel/bpf/trampoline.c:1324
Fix it by disallowing sleepable prog always when its target
btf is not a kernel's btf.
Fixes: 16d9c56606 ("bpf: Always allow sleepable programs on syscalls")
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Viktor Malik <vmalik@redhat.com>
Link: https://lore.kernel.org/bpf/20260805150810.34907-2-leon.hwang@linux.dev
There's no need to modify the trace object bpf_get_stackid_pe, we just
need to pass the needed callchain length in separate argument.
This way we can have callchain pointers const and remove the trace->nr
modification and restoration.
Assisted-by: Codex:GPT-5.5
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-13-jolsa@kernel.org
There's no need to modify the trace object bpf_get_stack_pe, we just
need to pass the needed callchain length in separate argument.
This way we can have callchain pointers const and remove the trace->nr
modification and restoration.
Assisted-by: Codex:GPT-5.5
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-12-jolsa@kernel.org
get_perf_callchain() returns a per-CPU perf_callchain_entry buffer and
releases its recursion slot via put_callchain_entry() before returning,
so nothing keeps the entry reserved while __bpf_get_stack() consumes
it below.
A preemptible BPF program (e.g. a non-sleepable raw tracepoint program
on a PREEMPT kernel, which runs under migrate_disable() but not
preempt_disable()) can be scheduled out between obtaining the entry
and the copy. Another task scheduled on the same CPU then reuses the
same per-CPU buffer and overwrites trace->nr with a larger value.
copy_len is then computed from the inflated trace->nr and can exceed
the caller's buffer, causing an out-of-bounds write in the memcpy()
and in the build_id path.
The rcu_read_lock() taken here alone does not prevent this. It is
only taken on the may_fault path, and under CONFIG_PREEMPT_RCU it does
not disable preemption; it merely keeps perf's callchain buffer array
alive (freed via call_rcu()) and does nothing to stop another task
from reusing the entry.
Disable preemption around obtaining the callchain entry and copying
it into the caller's buffer, so the entry cannot be reused underneath
us and trace->nr stays bounded by max_depth. Build ID resolution may
fault and is therefore deferred until after preemption is re-enabled;
by then the instruction pointers have already been copied into buf,
so it operates only on that private copy. Note, preempt_disable() also
subsumes the buffer-lifetime guarantee the rcu_read_lock() provided,
since a preempt-disabled section is an RCU read-side critical section
for the callchain buffers' call_rcu() reclaim.
Fixes: c195651e56 ("bpf: add bpf_get_stack helper")
Reported-by: Tao Chen <chen.dylane@linux.dev>
Reported-by: STAR Labs SG <info@starlabs.sg>
Signed-off-by: Daniel Borkmann <borkmann@iogearbox.net>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/bpf/20260803210149.296496-11-jolsa@kernel.org
Closes: https://lore.kernel.org/bpf/20260206090653.1336687-1-chen.dylane@linux.dev/
[ changed Fixes: commit ]
Both bpf_get_task_stack and bpf_get_task_stack_sleepable helpers that
use __bpf_get_task_stack have buf defined as ARG_PTR_TO_UNINIT_MEM
argument and we should initialize the buf on every return path.
Adding missing buf memset for __bpf_get_task_stack fail paths. This
provides deterministic buffer contents, which is useful when the buffer
is used directly as a map key.
Fixes: 06ab134ce8 ("bpf: Refcount task stack in bpf_get_task_stack")
Fixes: b992f01e66 ("bpf: Guard against accessing NULL pt_regs in bpf_get_task_stack()")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-10-jolsa@kernel.org
Now with the new callchain_* helper functions we can process trace_in
case directly in bpf_get_stack_pe function and remove it from
__bpf_get_stack which makes things easier for preemption fix in
following change.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-9-jolsa@kernel.org
The new callchain_finalize function calls the build-id retrieval
(if needed) and zeroes the buffer. This makes things easier for
preemption fix in following change.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-8-jolsa@kernel.org
The new callchain_store function stores trace entries buffer into
user supplied buffer. It covers both just-ip and buildid data.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-7-jolsa@kernel.org
Replacing __bpf_get_stackid calls with sequence of following functions:
stackid_fastpath
stackid_new_bucket
stackid_install
This makes code more structured and allows us to easily disable
preemption only in bpf_get_stackid in following changes.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-5-jolsa@kernel.org
The new stackid_fastpath does the fast stack hash and trace check, that
does not need new bucket allocation. It covers both just-ip and buildid
code paths.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-3-jolsa@kernel.org
The new stackid_init function stores all the necessary bits for stackid
trace and it will be used by other functions in following changes.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260803210149.296496-2-jolsa@kernel.org
syzbot is reporting KCOV state corruption on PREEMPT_RT kernels, for the
temporary storage used for saving/restoring remote KCOV state is currently
allocated as the per-CPU area.
On PREEMPT_RT kernels, softirq handlers run as preemptible task threads
(e.g., ksoftirqd). If a softirq context preempts a task running a remote
KCOV session, it safely saves the task's state into the per-CPU area.
However, if that softirq thread is subsequently preempted by a higher-
priority softirq thread on the same CPU, the second softirq will overwrite
the same per-CPU area, permanently destroying the original task's KCOV
state.
Fix this data corruption by moving the temporary storage from the per-CPU
area to the per-thread area. Since each softirq thread now owns its own
task context, nested softirq preemption no longer causes data overwrites.
Note that while the temporary storage is now on a per-thread basis, the
per-CPU kcov_percpu_data.lock must be retained, for we need to ensure that
kcov_remote_start() and kcov_remote_stop() operate atomically without
racing against asynchronous interrupts that manipulate the current task's
KCOV state.
It is likely that GFP_KERNEL allocation by vmalloc_node() in kcov_init()
has already called panic() before returning NULL, for there will be no
OOM-killable userspace processes when __init function of built-in module
runs. But this patch also fixes crashing the kernel when vmalloc_node()
in kcov_init() returned NULL, for kcov_init() left per-CPU irq_area == NULL
but kcov_remote_start() depends on per-CPU irq_area != NULL, resulting in
(1) doing vmalloc() in kcov_remote_start() despite !in_task() context
(2) out-of-array-bounds access if (1) succeeded but
kcov->remote_size < CONFIG_KCOV_IRQ_AREA_SIZE
(3) always leak memory allocated by (1), eventually killing all
OOM-killable userspace processes
problems.
Link: https://lore.kernel.org/43552d09-2ce2-4b19-b0d3-a2d1ab952145@I-love.SAKURA.ne.jp
Reported-by: syzbot+3f51ad7ac3ae57a6fdcc@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=3f51ad7ac3ae57a6fdcc
Reported-by: syzbot+47cf95ca1f9dcca872c8@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=47cf95ca1f9dcca872c8
Reported-by: syzbot+8a173e13208949931dc7@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=8a173e13208949931dc7
Reported-by: syzbot+90984d3713722683112e@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=90984d3713722683112e
Analyzed-by: AI Mode in Google Search (no mail address)
Fixes: 5ff3b30ab5 ("kcov: collect coverage from interrupts")
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reviewed-by: Alexander Potapenko <glider@google.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Clark Williams <williams@redhat.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Marco Elver <elver@google.com>
Cc: Mark Brown <broonie@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
parse() hands nla_strscpy() len as dstsize, and nla_strscpy() copies at
most dstsize - 1 bytes. When the attr payload comes in without a trailing
NUL, srclen == len >= dstsize and the last character of the cpumask string
gets cut off. Register "0-15" and you are silently listening on "0-1",
exit data for the rest never shows up.
The bug only bites when the sender doesn't NUL terminate the payload;
senders that include the NUL were always fine (srclen gets decremented for
the trailing NUL, so srclen < dstsize). Thats probably why this survived
20 years. And the policy is NLA_STRING, not NLA_NUL_STRING, so a payload
without the trailing NUL is legit input here.
Skip the kmalloc/nla_strscpy dance entirely and use nla_strdup(), which
already allocates srclen + 1 and terminates. The nla_len() bounds checks
stay as they were.
Link: https://lore.kernel.org/EC49FE41-7F5F-41E0-A07A-ABEB8ECA514D@grrlz.net
Fixes: f9fd8914c1 ("[PATCH] per-task delay accounting taskstats interface: control exit data through cpumasks")
Signed-off-by: Bradley Morgan <include@grrlz.net>
Reported-by: Oleg Deomi <oleg.deomi@gmail.com>
Closes: https://lore.kernel.org/CAByWkfZ6b1=3H9pwkz-dDQOs9cZaF-HYQ6b9Yb0=Hq2r1Vv_Pw@mail.gmail.com
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Balbir Singh <bsingharora@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Once destroy() returns the stack slot is no longer tracked as iterator
state, so zeroing it is dead work. Make the kfunc a no-op and inline the
call to a single BPF_JA 0 (the fixup can't drop the instruction outright,
so emit a nop; the JITs elide it).
Suggested-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260804134601.2305303-5-puranjay@kernel.org
bpf_iter_num_next() runs on every bpf_for() iteration, so inlining it
drops a call from the loop body. R1 points to the iterator; the returned
pointer to s->cur is R1 itself, since s->cur is first.
s->cur and s->end are int, so the kfunc's s->cur + 1 >= s->end is a
signed 32-bit compare and the inlined code needs no sign extension.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260804134601.2305303-4-puranjay@kernel.org
bpf_for() expands to the bpf_iter_num_{new,next,destroy}() kfuncs, which
the verifier emits as regular calls. They are tiny and only touch the
8-byte on-stack iterator state, so open-code them in bpf_fixup_kfunc_call()
like the other special kfuncs there.
Start with bpf_iter_num_new(): R1 points to the iterator, R2/R3 hold
start/end. The inlined sequence mirrors the kfunc and returns the same
-EINVAL / -E2BIG / 0.
start > end is rejected first, so end - start fits in a u32; range-check
it as u32 on both sides ((u32)(end - start) in the kfunc). A movsx-based
check would emit a cpuv4 instruction that some JITs (x86-32, mips32,
sparc64) decode as a plain move and get wrong.
The emitted instructions are plain BPF, so the interpreter path stays
correct and no jit_required marking is needed.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260804134601.2305303-3-puranjay@kernel.org
The comment on the s->cur + 1 >= s->end check claims the (s64) cast is
needed to avoid overflow when s->cur == s->end == INT_MAX. It isn't:
s->cur + 1 is computed in int and wraps before the cast, so the cast
changes nothing (INT_MAX + 1 compares the same either way).
The wraparound is the point. bpf_iter_num_new() sets s->cur = start - 1,
which wraps to INT_MAX for start == INT_MIN, and the wrapping s->cur + 1
brings it back to start. (s64)s->cur + 1 would instead break iterators
starting at INT_MIN.
Drop the cast and reword the comment. No functional change; the wrap is
well-defined under -fno-strict-overflow.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20260804134601.2305303-2-puranjay@kernel.org
CONFIG_SYSCTL used to make sense as a separate hidden bool before commit
61a47c1ad3 ("sysctl: Remove the sysctl system call"); it was selected
by both CONFIG_SYSCTL_SYSCALL and CONFIG_PROC_SYSCTL.
Today CONFIG_PROC_SYSCTL is the only selector, so the two are always
equal. Kill the hidden bool, rename the PROC_SYSCTL prompt to SYSCTL,
and s/CONFIG_PROC_SYSCTL/CONFIG_SYSCTL/ tree-wide.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Joel Granados <joel.granados@kernel.org>
cad_pid is global, and kill_cad_pid() is only used in the root namespace.
However, due to pid_table_root_permissions(), a non-root user can unshare
pid/user namespaces and modify it from the child namespace. This makes no
sense and is simply wrong.
Move it to kern_reboot_table[] where it logically belongs; this ensures
that only GLOBAL_ROOT_UID can read/modify this sysctl.
Note that this patch doesn't preserve "#ifdef CONFIG_PROC_SYSCTL" around
the "cad_pid"; CONFIG_PROC_SYSCTL selects CONFIG_SYSCTL, so it is always
set when kern_reboot_table[] is compiled.
Cc: stable@vger.kernel.org
Fixes: e054bcbe7e ("sysctl: move cad_pid into kernel/pid.c")
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Alexey Gladkov <legion@kernel.org>
Reviewed-by: Bradley Morgan <include@grrlz.net>
Reviewed-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Signed-off-by: Joel Granados <joel.granados@kernel.org>
Fix kernel-doc warnings in kernel/sysctl.c by adding Returns.
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Joel Granados <joel.granados@kernel.org>
Add colon ":" after argument name where it is missing
Add doc for proc_int_conv and proc_dointvec_conv
Signed-off-by: Joel Granados <joel.granados@kernel.org>
Remove "_minmax" from proc_doulongvec_minmax_conv as it does not enforce
min/max limits but serves as a generic converter for unsigned long
vectors. Update function declaration in sysctl.h, definition in
sysctl.c, and caller in jiffies.c accordingly.
Signed-off-by: Joel Granados <joel.granados@kernel.org>
Make do_proc_vec static and parametrize by proc_vec_type enum which
defines the type being processed and selects which converter is "live".
Signed-ness and size are calculated based on proc_vec_type and
table->data is now walked as raw bytes and advanced by the element size;
the converter still performs the actual typed load/store. Pass converter
as a union to avoid a cast from void*. The public
proc_do{int,uint,ulong}vec_conv() prototypes and all converter
signatures in kernel/, fs/ and the header are therefore unchanged.
Remove do_proc_doulongvec_minmax. proc_doulongvec_minmax_conv uses a
converter callback passed by the caller instead of conversions based on
conv{mul,div}. Create uni and bi-direction converters for milliseconds
to jiffies in proc_doulongvec_ms_jiffies_minmax; which is the only user
of proc_doulongvec_minmax_conv.
Replace do_proc_douintvec{,_w,_r} functions with a call to do_proc_vec.
Disallow vectors for uint by returning -EINVAL when more than one
element is detected.
Signed-off-by: Joel Granados <joel.granados@kernel.org>
Updates all douintvec converter function signatures to include a bool
*negp parameter. This is a preparation commit required to eventually run
all converters under the same function. The negp argument will be
ignored as it is not relevant for the uint type. Note that
do_proc_uint_conv_pipe_maxsz in pipe.c is also modified.
Signed-off-by: Joel Granados <joel.granados@kernel.org>
Move the converter assignment out of do_proc_dointvec into the caller.
Both the test for NULL and the assignment are meant to stay within the
sysctl.c context. This is in preparation of using a typed macro to for
the integer proc vector function.
Signed-off-by: Joel Granados <joel.granados@kernel.org>
check_atomic_load() calls check_load_mem() before atomic_ptr_type_ok().
For a load-acquire that fetches into its own source register (dst_reg ==
src_reg), check_load_mem() overwrites src_reg's type with the type of the
loaded value, so the subsequent atomic_ptr_type_ok() no longer sees the
source pointer and fails to reject the disallowed types (ctx, pkt,
flow_keys, sock).
Since bpf_convert_ctx_accesses() does not rewrite atomic loads, the raw
access to the underlying kernel object is left in place. The destination
type is taken from the ctx access itself, so a load-acquire of the sk
field of struct __sk_buff for example leaves the register typed as
PTR_TO_SOCK_COMMON_OR_NULL, which type_is_sk_pointer() does not match
either, while it actually holds unconverted struct sk_buff bytes. Once
the NULL check has passed this is a type confusion, not just a leak of
kernel data.
Validate src_reg with check_reg_arg() and check the source pointer type
with atomic_ptr_type_ok() before the load again, mirroring
check_atomic_rmw(). Out-of-range register numbers are already rejected
earlier by check_and_resolve_insns() (commit 503d21ef8e ("bpf: Do
register range validation early")), and the only exemption there,
is_stack_arg_ldx(), requires BPF_LDX | BPF_MEM | BPF_DW and thus never
matches a BPF_ATOMIC insn. atomic_ptr_type_ok() can therefore not
dereference register state out of bounds, that is, the out-of-bounds
read addressed by the Fixes commit below does not reappear (as proven
also via selftest).
Fixes: c03bb2fa32 ("bpf: Fix out-of-bounds read in check_atomic_load/store()")
Reported-by: STAR Labs SG <info@starlabs.sg>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260804201917.253491-1-daniel@iogearbox.net
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
This helps to find a common subtree of two resources, which is important
when determining whether it's helpful to evict one resource in favor of
another.
To facilitate this, add a common helper to find the ancestor of two
cgroups using each cgroup's ancestor array.
Tested-by: Thadeu Lima de Souza Cascardo <cascardo@igalia.com>
Reviewed-by: Maarten Lankhorst <dev@lankhorst.se>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Signed-off-by: Natalie Vock <natalie.vock@gmx.de>
Link: https://patch.msgid.link/20260804-dmemcg-aggressive-protect-v8-2-07af96681bf8@gmx.de
Callers can use this feedback to be more aggressive in making space for
allocations of a cgroup if they know it is protected.
These are counterparts to memcg's mem_cgroup_below_{min,low}.
Reviewed-by: Maarten Lankhorst <dev@lankhorst.se>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Signed-off-by: Natalie Vock <natalie.vock@gmx.de>
Link: https://patch.msgid.link/20260804-dmemcg-aggressive-protect-v8-1-07af96681bf8@gmx.de
Be consistent in using vma_start_pgoff() and vma_end_pgoff(), which clearly
indicates which part of the VMA the page offset refers to and aids
greppability.
This is part of a broader series laying the ground to provide a virtual
page offset for MAP_PRIVATE-file backed anon folios.
No functional change intended.
Link: https://lore.kernel.org/20260710-b4-pre-scalable-cow-v2-19-2a5aa403d977@kernel.org
Signed-off-by: Lorenzo Stoakes <ljs@kernel.org>
Acked-by: Marek Szyprowski <m.szyprowski@samsung.com> # for kernel/dma
Reviewed-by: Gregory Price <gourry@gourry.net>
Acked-by: Pedro Falcato <pfalcato@suse.de>
Reviewed-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Cc: Ackerley Tng <ackerleytng@google.com>
Cc: David Hildenbrand (Arm) <david@kernel.org>
Cc: Kai Huang <kai.huang@intel.com>
Cc: SJ Park <sj@kernel.org>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Liam R. Howlett (Oracle) <liam@infradead.org>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The family of vma_interval_tree_() functions manipulate the
address_space (which, of course, is generally referred to as 'mapping')
reverse mapping, but are named the 'VMA' interval tree.
VMAs may be mapped by an anon_vma, an address_space, or both. Therefore
calling the mapping interval tree a 'VMA' interval tree is rather
confusing.
This is also inconsistent with the anon_vma_interval_tree_*() functions
which explicitly reference the rmap object to which they pertain.
Rename the vma_interval_tree_*() functions to mapping_rmap_tree_*() to
correct this.
We will rename the anon rmap functions similarly in a subsequent patch.
No functional change intended.
Link: https://lore.kernel.org/20260710-b4-pre-scalable-cow-v2-8-2a5aa403d977@kernel.org
Signed-off-by: Lorenzo Stoakes <ljs@kernel.org>
Reviewed-by: Gregory Price <gourry@gourry.net>
Reviewed-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Reviewed-by: Zi Yan <ziy@nvidia.com>
Cc: Ackerley Tng <ackerleytng@google.com>
Cc: David Hildenbrand (Arm) <david@kernel.org>
Cc: Kai Huang <kai.huang@intel.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Pedro Falcato <pfalcato@suse.de>
Cc: SJ Park <sj@kernel.org>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Liam R. Howlett (Oracle) <liam@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The file-backed mapping interval tree functions vma_interval_tree_*()
accept a raw rb_root_cached pointer to determine the tree in which they are
operating.
However, in each case, this is always associated with an address_space data
type.
So simply pass a pointer to that instead to simplify the code, and more
clearly differentiate between these operations and those concerning
anonymous mappings.
While we're here, make the generated interval tree functions static as they
do not need to be used externally (any previously existing external users
have now been removed).
We also rename VMA parameters from 'node' to 'vma' as calling this a node
is simply confusing, update the input index types to pgoff_t since they
reference page offsets and rename the parameters to pgoff_start and
pgoff_last.
No functional change intended.
Link: https://lore.kernel.org/20260710-b4-pre-scalable-cow-v2-6-2a5aa403d977@kernel.org
Signed-off-by: Lorenzo Stoakes <ljs@kernel.org>
Reviewed-by: Pedro Falcato <pfalcato@suse.de>
Reviewed-by: Gregory Price <gourry@gourry.net>
Reviewed-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Reviewed-by: Zi Yan <ziy@nvidia.com>
Cc: Ackerley Tng <ackerleytng@google.com>
Cc: David Hildenbrand (Arm) <david@kernel.org>
Cc: Kai Huang <kai.huang@intel.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: SJ Park <sj@kernel.org>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Liam R. Howlett (Oracle) <liam@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
mm/internal.h becomes more and more bloated.
Move declarations related to vmalloc to a new mm/vmalloc.h header.
No functional changes.
Link: https://lore.kernel.org/20260709-internal-h-v2-3-695631425968@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Acked-by: Muchun Song <muchun.song@linux.dev>
Acked-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Acked-by: David Hildenbrand (Arm) <david@kernel.org>
Acked-by: Lorenzo Stoakes <ljs@kernel.org>
Acked-by: Pratyush Yadav <pratyush@kernel.org>
Acked-by: SJ Park <sj@kernel.org>
Cc: Alexander Graf <graf@amazon.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Brendan Jackman <jackmanb@google.com>
Cc: Brendan Jackman <brendan.jackman@linux.dev>
Cc: Dennis Zhou <dennis@kernel.org>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Marco Elver <elver@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Pasha Tatashin <pasha.tatashin@soleen.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: "Uladzislau Rezki (Sony)" <urezki@gmail.com>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "mm: split a couple of headers from internal.h", v2.
mm/internal.h becomes more and more bloated.
Split declarations related to mm_init, memblock, vmalloc and sparse into
new headers.
This patch (of 3):
mm/internal.h becomes more and more bloated.
Move declarations for related to mm/mm_init.c and mm/memblock.c to a new
mm/mm_init.h header.
No functional changes.
[rppt@kernel.org: split stubfs from internal.h to mm_init.h]
Link: https://lore.kernel.org/alJd1BLypyK9Mpaw@kernel.org
Link: https://lore.kernel.org/20260709-internal-h-v2-0-695631425968@kernel.org
Link: https://lore.kernel.org/20260709-internal-h-v2-1-695631425968@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Acked-by: Muchun Song <muchun.song@linux.dev>
Acked-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Acked-by: David Hildenbrand (Arm) <david@kernel.org>
Acked-by: Lorenzo Stoakes <ljs@kernel.org>
Acked-by: Pratyush Yadav <pratyush@kernel.org>
Acked-by: SJ Park <sj@kernel.org>
Cc: Alexander Graf <graf@amazon.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Brendan Jackman <jackmanb@google.com>
Cc: Brendan Jackman <brendan.jackman@linux.dev>
Cc: Dennis Zhou <dennis@kernel.org>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Marco Elver <elver@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Pasha Tatashin <pasha.tatashin@soleen.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: "Uladzislau Rezki (Sony)" <urezki@gmail.com>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
A potential invalid storage access issue can occur after replacing a
cgroup bpf prog.
This occurs in the following scenario:
1. prog1 with storage is attached to a cgroup in multi-attach mode.
2. prog1 is replaced with prog2 using BPF_F_REPLACE in multi-attach
mode, but fails midway (e.g. in bpf_trampoline_link_cgroup_shim or
update_effective_progs).
3. A new prog3 is attached to the cgroup in multi-attach mode.
The reason is that __cgroup_bpf_attach overwrites pl->storage with the
new storage prior to attachment completion. When attachment fails
midway, the cleanup path calls bpf_cgroup_storages_free(new_storage) to
free the newly allocated storage, but fails to restore pl->storage back
to old_storage.
Consequently, the still-active prog1 holds invalid or dangling storage
pointers, leading to an invalid memory access when prog1 executes and
calls bpf_get_local_storage. Additionally, original pl->flags and
cgrp->bpf.flags[atype] are left unrestored.
Fix this by saving old_pl_flags, old_storage, and old_flags prior to the
update, and properly restoring all of them in the cleanup path on error.
Fixes: 7d9c342789 ("bpf: Make cgroup storages shared between programs on the same cgroup")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Pu Lehui <pulehui@huawei.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/20260803013934.4036646-1-pulehui@huaweicloud.com
Commit 8da2a88383 ("rv: Add KUnit tests for some DA/HA monitors")
introduced a division of a 64-bit value by 1000 in the nomiss KUnit
test. This does not compile on 32-bit systems, as standard division of
64-bit values leads to an undefined reference to __udivdi3.
Fix the build on 32-bit systems by using div_u64().
Fixes: 8da2a88383 ("rv: Add KUnit tests for some DA/HA monitors")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202608020311.hYjqOG5k-lkp@intel.com
Reviewed-by: Nam Cao <namcao@linutronix.de>
Link: https://lore.kernel.org/r/20260803150622.322806-1-gmonaco@redhat.com
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
blkdev.h gets included in various places outside the block layer just
for struct blk_plug and related plugging functions.
Split blk_plug into a separate helper to reduce the amount of code
that needs to get rebuilt when blkdev.h changes and to slightly
reduce compile times.
In io_uring this requires pulling in a few other headers explicitly that
previously were implicitly included through blkdev.h.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner (Amutable) <brauner@kernel.org>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://patch.msgid.link/20260804125524.740996-1-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
futex_private_hash_put() drops the reference to fph before evaluating
fph->mm for wake_up_var(). futex_ref_put() enables preemption again before
returning. If that put drops the final reference and the task is preempted,
another task can pivot to the replacement hash and free the old hash after
an RCU grace period. The first task then reads fph->mm from the freed
allocation when it resumes.
KASAN reports a slab-use-after-free in futex_private_hash_put(), with the
read at offset 24 in a freed kmalloc-512 allocation. The allocation and
free stacks point to futex_hash_allocate() and the RCU free path,
respectively.
Load the mm pointer while the fph reference is still held and pass the
saved value to wake_up_var(). wake_up_var() uses the pointer as a waitqueue
key and does not dereference the mm through it.
Fixes: bd54df5ea7 ("futex: Allow to resize the private local hash")
Signed-off-by: Felix Hoffmann <f3lix.dev@gmx.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260731155024.1150011-1-f3lix.dev@gmx.de
In a sleepable BPF program, a spin lock can provide the only RCU protection
for a kptr. The final bpf_spin_unlock() ends that protection, but the
verifier leaves the pointer valid. Another CPU can then free the object
before the pointer is used. A capability-limited runtime PoC triggered a
task_struct use-after-free in __bpf_get_task_stack().
Record whether the program is in an RCU-protected context before releasing
the lock. Invalidate RCU-protected pointers only when the unlock leaves the
final such context. This preserves valid pointers in non-sleepable programs
and inside an explicit RCU read-side section.
Fixes: 5861d1e8db ("bpf: Allow bpf_spin_{lock,unlock} in sleepable progs")
Assisted-by: Codex:gpt-5.6-sol
Assisted-by: ChatGPT:GPT-5.6-Pro
Signed-off-by: Ning Ding <dingning04@gmail.com>
Link: https://lore.kernel.org/bpf/20260803112615.3362122-2-dingning04@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
HugeTLB pages can be preserved memory. So they are never allocated from
scratch. Instead, they are allocated from the memory blocks with no
preserved memory. These areas are detected at runtime on each boot.
But since they are allocated via memblock, they show up as RSRV_KERN,
and blow up the scratch size when scratch scale is in use.
All hugetlb pages are marked RSRV_HUGETLB. Subtract their size from
RSRV_KERN when calculating scratch sizes.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-23-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Motivation
==========
The scratch space is allocated by the first kernel in the KHO chain, and
is reused by all subsequent kernels. The size of the space is either set
via the commandline by the system administrator or by calculating the
amount of memory used by the kernel and adding a multiplier. In either
case, the scratch space is a heuristic and is liable to fill up and fail
allocation if a kernel uses more memory than expected.
In addition, gigantic huge pages (usually 1 GiB) are allocated via
memblock, and in a KHO boot that memory comes from the scratch space. In
hypervisors it is common to dedicate a major part of the system's memory
to gigantic hugepages for VM memory.
If this memory needs to come from scratch space, then scratch needs to
be greater than the memory needed for huge pages, which is impractical.
In addition, hugepages can be preserved memory. Allocating them from
scratch violates the assumption that scratch contains no preserved
memory.
Methodology
===========
Discover areas that don't contain any preserved memory at boot by
walking the preserved memory radix tree. Mark them as scratch to allow
allocations from them. This makes KHO more resilient to memory pressure
and allows supporting huge page preservation.
Since the preserved memory radix tree mixes both physical address and
order into a single key, and does not track table pages, it is difficult
to identify free areas from it directly. Walk the tree and digest it
down into another radix tree. The latter tracks blocks of
KHO_SCRATCH_EXT_BLKSIZE (1 GiB as of now) granularity. Then walk the
digested tree and mark the areas between the present keys as scratch.
Performance
===========
The discovery algorithm traverses the preserved memory radix tree
exactly once. While it does use memory for the digested radix tree,
since the blocks are split by 1 GiB, a single bitmap with 4k pages can
track up to 32 TiB of memory. So there are likely to be very few radix
tree pages used in this tracking. For systems with all physical memory
below 32 TiB, this should result in a total of 6 pages being
used (KHO_TREE_MAX_DEPTH == 6).
An alternate way of achieving this would be to call kho_mem_retrieve()
earlier in boot and mark all the KHO preservations as reserved. But that
can blow up memblock.reserved with a bunch of 4K pages scattered
everywhere, which will reduce performance of subsequent allocations.
Since the free blocks are tracked in chunks of 1 GiB, this won't blow up
memblock.memory as much.
There is no inherent reason for using 1 GiB as the discovered block
size. This can be changed later if needed. Currently, KHO is mainly
targeted for server grade systems with hundreds of gigabytes to
terabytes of memory. So 1 GiB is a reasonable granularity for those
systems. For smaller systems this doesn't work as well, but we can
arrive at a better heuristic when we have concrete use cases.
Practical evaluation
====================
The testing is done on a x86_64 qemu VM running under KVM with 64G
memory and 12 CPUs. The machine pre-allocates 50 1G pages.
Since the performance scales with how busy the radix tree is, tests are
done with 2 preservation patterns: first with two 1M memfds, second with
two 1G memfds, both using 4k pages.
Test case 1 - 1M memfd
~~~~~~~~~~~~~~~~~~~~~~
This test case has two memfds with 1M memory each in 4k pages, plus
other preservations from LUO core and other KHO users.
This is how the radix tree stats look like:
radix_nodes: 0x13
nr_preservations: 0x214
mem_preserved: 0x227000
per order preservations:
order 0: 0x20f
order 1: 0x4
order 4: 0x1
and this is how long it takes to extend the scratch after KHO boot:
KHO: KHO extend time: 47 us
KHO: KHO extend total mem: 0xe6c17b000 (~57G)
Test case 2 - 1G memfd
~~~~~~~~~~~~~~~~~~~~~~
This test case has two memfds with 1G memory each in 4k pages, plus
other preservations from LUO core and other KHO users.
This is how the radix tree stats look like:
radix_nodes: 0x28
nr_preservations: 0x80816
mem_preserved: 0x80829000
per order preservations:
order 0: 0x80811
order 1: 0x4
order 4: 0x1
and this is how long it takes to extend the scratch after KHO boot:
KHO: KHO extend time: 22514 us
KHO: KHO extend total mem: 0xd3f200000 (~52G)
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-19-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Currently the preserved memory radix tree is initialized from
kho_memory_init(), which happens relatively late in MM init. In a coming
patch, the tree will be used from kho_memory_init_early(). Move the tree
initialization there.
Simplify some of the code in kho_mem_retrieve() by getting rid of the
err variable and jumping to err directly.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-16-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
In a future patch, mm init will use kho_scratch_overlap() for deciding
the migrate type of pageblocks it initializes. The earliest user
currently is free_area_init(). kho_scratch_overlap()
relies on kho_scratch pointer being initialized. Introduce
kho_memory_init_early() to do this.
kho_populate() would normally be a good place to do this, but
unfortunately, phys_to_virt() does not work at that point on ARM64. So
we need yet another initialization function.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-15-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Support for discovering memory blocks with no preserved memory will be
added in coming patches. These areas will also be marked as scratch to
allow allocations from them. Memblock will switch to looking through the
scratch array to decide the right migratetype.
Expose kho_scratch_overlap() to KHO users. Since it is now used by
non-debug code, move it out of kexec_handover_debug.c and into
kexec_handover.c. Gate the overlap checks in kho_preserve_folio() and
kho_preserve_pages() by IS_ENABLED(CONFIG_KEXEC_HANDOVER_DEBUG) instead.
Since kexec_handover_debug.c is now empty, delete it.
Add a stub for kho_scratch_overlap() to memblock tests to make sure it
compiles. It will be used in memblock by a coming commit.
No functional changes.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-14-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Move the initialization logic of the radix tree into
kho_radix_init_tree() instead of having users open-code it. Makes the
boundaries cleaner and reduces code duplication when a new user of the
radix tree will be added in a future commit.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-13-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Add kho_radix_destroy_tree() which allows destroying the radix tree and
freeing all its pages.
This is will be used by the upcoming scratch extension mechanism. It
creates a radix tree to track free blocks and then frees them after
telling memblock about them.
Reviewed-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-12-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
The KHO radix tree allocates memory for table pages from the buddy
allocator using get_zeroed_page(). This is not available in early boot
when memblock is still active.
Using the radix tree in early boot is useful for KHO to track metadata
about its memory. One such example is for tracking free blocks for
memory allocation when scratch runs out of space. This feature will be
added in the following commits.
Add kho_radix_{alloc,free}_node() which allocate and free the table
pages. They use slab_is_available() to decide which allocator to use.
While slab_is_available() indicates availability of the slab allocator,
it gets initialized right after buddy so it serves the same practical
purpose.
Reviewed-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-11-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Add an opaque data pointer argument to kho_radix_walk_cb_t. This can be
used by callers to pass extra information to the callback.
Reviewed-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-10-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
The KHO memory preservation radix tree does not mark the table pages
themselves as preserved. This is done to avoid a circular dependency
where preserving a page can lead of allocating other preserved pages.
This means any walker looking for free ranges of memory outside of
scratch areas will ignore the table
Add a table callback that is invoked for each table page. The callback
is given the physical address of the table page.
This is useful for the upcoming mechanism that discovers blocks of
memory with no preserved pages and lets them be used for boot memory.
Another use case is for users of the radix tree other than KHO itself.
The radix tree does not preserve its own pages due to the circular
dependency described above. But external users of the radix tree would
need to preserve and restore their pages for the radix tree to survive
past early boot. They can use this callback to do so.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-9-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
A future commit will add more callbacks for the KHO radix tree. Add a
struct for collecting the callbacks.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-8-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
The memory retrieval logic is spread out across kho_mem_retrieve() and
kho_memory_init(). The incoming scratch area is initialized at
kho_memory_init(), and the error handling is done there too.
Consolidate all this logic into kho_mem_retrieve() to make the code
cleaner.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-7-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
This allows other functions to also use the radix tree.
While at it, add kho_get_mem_map() helper to get the virtual address of
the preserved memory map and use that helper instead of duplicating the
code to get the preserved memory map from the FDT.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-6-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Patch series "taskstats: fix cgroupstats invalid fd handling and add
selftests".
This series fixes an issue where cgroupstats mishandles invalid file
descriptors, and introduces a functional kselftest to prevent regressions.
When an invalid file descriptor is passed via CGROUPSTATS_CMD_ATTR_FD,
cgroupstats_user_cmd() returns 0 instead of an error code. This leads to
two broken behaviors depending on netlink flags:
- Callers without NLM_F_ACK block indefinitely on recv().
- Callers with NLM_F_ACK receive a misleading success ACK (errno == 0)
but no actual statistics payload.
The first patch addresses this by returning -EBADF when the fd cannot be
resolved. The second patch adds a comprehensive kselftest covering both
the valid cgroup v1 query and the invalid fd paths (with and without
NLM_F_ACK) to ensure the fixes work as intended.
This patch (of 2):
cgroupstats_user_cmd() returns 0 without sending a reply or a netlink
error when the fd passed via CGROUPSTATS_CMD_ATTR_FD does not resolve to
an open file in the caller's table. As a result:
- clients that did not set NLM_F_ACK block on recv() indefinitely
waiting for a CGROUPSTATS_CMD_NEW message that is never emitted;
- clients that set NLM_F_ACK receive a misleading "success" ACK (errno
== 0) with no statistics payload.
Return -EBADF instead so the netlink layer propagates the error to
userspace as expected.
Link: https://lore.kernel.org/cover.1783713230.git.cyyzero16@gmail.com
Link: https://lore.kernel.org/b4fd9e288e4a48efebaf41b4ffcdb204b06675c4.1783713230.git.cyyzero16@gmail.com
Signed-off-by: Yiyang Chen <cyyzero16@gmail.com>
Acked-by: Balbir Singh <balbirs@nvidia.com>
Cc: Dr. Thomas Orgis <thomas.orgis@uni-hamburg.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
max_threads is initialized once by fork_init()->set_max_threads(), and
thereafter is mostly read in hot path, such as copy_process().
The fact that it is mostly read and not written to makes it candidates for
__read_mostly declarations.
I have already tested on my machine(arm64,256core,kernel-7.1.3) with
'hackbench -P -l 500 -g 100/500/1000/2000' and the results show that there
is an average of 13% improvement in performance.
before patch:
hackbench -g 100 500 1000 2000
1.094 4.651 13.305 36.322
after patch:
hackbench -g 100 500 1000 2000
0.823 4.487 11.107 32.987
Above data are the average values obtained from multiple tests, and there
was indeed some fluctuation in the data during the tests.
Link: https://lore.kernel.org/20260707092354.465847-1-jansef.jian@hj-micro.com
Signed-off-by: JianChunfu <jansef.jian@hj-micro.com>
Reviewed-by: Lorenzo Stoakes <ljs@kernel.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Kees Cook <kees@kernel.org>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
In has_pending_signals there was a switch/case used for optimizations.
However, today's compilers perform loop unrolling efficiently, thus it is
not needed anymore.
Put i inside the for declaration so we do not risk its escape from the
scope. Moreover, i starts now from 0 and counts up, as it is a more usual
pattern.
Link: https://lore.kernel.org/20260520062849.183621-2-andrea.calabrese@amarulasolutions.com
Signed-off-by: Andrea Calabrese <andrea.calabrese@amarulasolutions.com>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Cc: Adrian Huang <adrianhuang0701@gmail.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Marco Elver <elver@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
send_signal_locked() unconditionally copies siginfo before the namespace
translation to avoid corrupting a shared siginfo.
Not that I think this can actually hurt performance-wise, just it doesn't
look clean to me; the copy is only needed in the unlikely case when the
translation will actually change something.
Defer it to the two cases where si_pid/si_uid are rewritten, and while at
it add #ifdef's just for completeness.
Link: https://lore.kernel.org/akkaAgNfUby5_3nM@redhat.com
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Reviewed-by: Bradley Morgan <include@grrlz.net>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Eric Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
prepare_kill_siginfo(PIDTYPE_TGID) fills si_code = SI_USER and sets
si_pid/si_uid in the sender's namespace. Then send_signal_locked()
translates si_pid/si_uid to the target's namespace.
SEND_SIG_NOINFO exists precisely for the case when si_code == SI_USER
and si_pid/si_uid are the sender's ids; this is exactly what sys_kill()
does via prepare_kill_siginfo(PIDTYPE_TGID). Change sys_kill() to use
it directly.
SEND_SIG_NOINFO produces the same result: si_code = SI_USER, and
__send_signal_locked() computes si_pid/si_uid directly in the target's
namespace. The force computation is also the same: both check if the
sender is visible in the target's pid namespace.
This is just a cleanup and microoptimization (especially with [1]),
this skips the has_si_pid_and_uid() block in send_signal_locked() and
offloads the namespace translation logic to
__send_signal_locked(SEND_SIG_NOINFO) which uses the simpler
computations.
NOTE: As a "side effect" this also fixes the kill(pid < 0, sig) case
where send_signal_locked() rewrites si_pid/si_uid in the shared
siginfo, corrupting it for subsequent recipients. But for other group
senders like __kill_pgrp_info() we still need the fix from Bradley
Morgan [1] who found this problem.
TODO: kill prepare_kill_siginfo() and change other users to use
SEND_SIG_NOINFO too. This needs trivial changes in
__send_signal_locked() and TP_STORE_SIGINFO().
Link: https://lore.kernel.org/aj6btAZqYuv59a8w@redhat.com
Link: https://lore.kernel.org/all/20260622164029.11474-1-include@grrlz.net/ [1]
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Reviewed-by: Bradley Morgan <include@grrlz.net>
Cc: Eric Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
send_signal_locked() rewrites sender ids for the target namespace. Group
sends reuse the same siginfo, so one recipient can affect the next.
Copy the siginfo before changing it.
Link: https://lore.kernel.org/86a8857d58d43ee26a8b365b837fd24830343494.1782159692.git.include@grrlz.net
Fixes: 7a0cf09494 ("signal: Correct namespace fixups of si_pid and si_uid")
Signed-off-by: Bradley Morgan <include@grrlz.net>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Adrian Huang <adrianhuang0701@gmail.com>
Cc: Aleksandr Nogikh <nogikh@google.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Marco Elver <elver@google.com>
Cc: "Masami Hiramatsu (Google)" <mhiramat@kernel.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Fix misspellings of "softlockup" in the watchdog enabled bit definitions
and related comments. Also fix a nearby "successful" typo.
No functional change.
Link: https://lore.kernel.org/20260615174557.1836562-1-edcr1790@gmail.com
Signed-off-by: Matthew Chen <edcr1790@gmail.com>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Inside parse_one(), the core parameter-parsing engine prints the address
of the parameter-set callback function using %p: pr_debug("handling %s
with %p\n", param, params[i].ops->set);
Since the string value of the parameter being parsed (val) is already
available, print the parameter name and its value instead, and avoid
tainting the kernel by exposing a kernel-ptr.
Link: https://lore.kernel.org/20260618190715.3563047-1-jim.cromie@gmail.com
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The "resource sanity check" warning need not mean that there is a
functional issue and in the cases when there isn't one, it is just useless
and confusing noise.
For this reason, downgrade the log level of it from "warn" to "debug".
This reduces log clutter while keeping the diagnostic information
available for debugging purposes if needed.
Link: https://lore.kernel.org/20260523000131.7086-1-sakunix@yahoo.com
Signed-off-by: Manuel Quintero Fonseca <sakunix@yahoo.com>
Cc: Andriy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Hans de Goede <hansg@kernel.org>
Cc: Mika Westeberg <mika.westerberg@linux.intel.com>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
An 'F' entry, and every interpreter a 'B' entry binds, holds a file open
from registration until the entry goes away, pinning the file, its inode,
the mount it came from and that mount's superblock. Nothing bounds how
many of those a user namespace can hold. An entry binds at most
BINFMT_MISC_INTERP_MAX interpreters, but nothing caps the entries.
Charge each binding to the user namespace and uid that makes it against a
new UCOUNT_BINFMT_MISC_INTERPRETERS. Going over budget causes -ENOSPC.
A per-instance cap would suck. Instances are keyed on the user
namespace. So any constant is multiplied by the number of namespaces the
caller creates. Creating those is virtually free. A ucount charges the
namespace and every one of its ancestors. And a namespace can raise only
its own limit. So nesting buys nothing.
The knob is /proc/sys/user/max_binfmt_misc_interpreters. Leave it at the
max_threads/2 default fork_init() gives a new type. No existing
configuration comes close to that.
binfmt_misc is tristate, which makes it the first ucount user that can be
built as a module. Export inc_ucount() and dec_ucount(); without them
CONFIG_BINFMT_MISC=m fails to link. Export them to binfmt_misc alone:
charging a ucount type is not something a module has any business doing
in general, and the list is trivial to extend if a second user shows up.
init_user_ns and init_binfmt_misc are already exported for the same
module.
Link: https://patch.msgid.link/20260803-work-binfmt_misc-interplimit-v1-1-4a2435500bd9@kernel.org
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
When rescue demand on a cpu persistently exceeds the configured bandwidth,
tasks age on that cpu's rescue DSQ until the stall watchdog fires. The
watchdog blames the waiting task's owner, but the misbehaving party is
whoever floods the queue, not whoever happens to time out.
Track each sched's recent rescue consumption per cpu as a decaying average.
Once the oldest waiter on a cpu's rescue DSQ has been queued past a
threshold derived from the rescue knobs (4s at the defaults), the rescue
timer ejects the sub with the highest recent consumption on that cpu with
SCX_EXIT_ERROR_RESCUE. With no recent consumer there is no victim and
nothing is ejected - the generic stall watchdog eventually blames the
waiter's owner instead. Ejections on a cpu are spaced one threshold apart so
the freed bandwidth can drain the backlog before another sub is judged.
The overload check only wins the race against the stall watchdog when the
watchdog timeout clears the threshold, and a single in-budget wait must not
cross the trigger on its own. Warn on a scheduler whose timeout doesn't fit
and on knobs whose funding period exceeds half the threshold.
v2: - Track kill_at in jiffies_64 - on 32-bit, the time_before() grace check
wraps 2^31 ticks after the last ejection and suppresses ejections.
(sashiko AI)
- Track rescue_avg_at in jiffies_64 likewise - the unsigned long decay
delta truncates mod 2^32 on 32-bit and can revive a weeks-old usage
average in the victim pick.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A local DSQ insert lacking the needed caps is diverted to the reject DSQ and
bounced back through ops.enqueue() so the scheduler can re-decide. That
recovery assumes the scheduler has somewhere legal to send the task. When it
doesn't, e.g. when the task's affinity is restricted to cids delegated away,
the task starves until the stall watchdog ejects the scheduler. An exiting
task is worse - it skips ops.enqueue() and the rejection becomes a
self-requeuing cycle that burns the CPU until the watchdog fires.
Add SCX_ENQ_RESCUE, a fallback modifier on local DSQ inserts. When the
insert would be rejected for missing caps, the kernel takes over and runs
the task on the target CPU without consulting the owning scheduler. The
kernel sets the flag itself when enqueueing an exiting task.
Rescue is a last-resort forward-progress backstop with a persistent
disadvantage, not a way around cap enforcement. A per-CPU token bucket
accrues rescue_bandwidth_ppt (default 2%) of CPU time and rescues run one at
a time in arrival order. Each is granted a slice of the rescue_quantum_us
(default 5ms) quantum divided across the waiters, waits at the tail of the
local DSQ claiming no priority, and rejoins its scheduler as a fresh arrival
once the slice is served.
The schedulers keep their normal control over an admitted rescuee and may
preempt or reslice it. Service is measured on CPU time actually received, so
neither shortens the rescue. Prolonged denial escalates - the remaining
slice turns into protected execution (SCX_TASK_PROTECTED) and the rescuee
preempts the current task. Escalation is paced by the same bucket, and
delivered service converges on the configured bandwidth no matter how
aggressively the schedulers dispatch.
Both knobs are root-only and SCX_RESCUE_DISABLE turns rescue off, making
SCX_ENQ_RESCUE inserts reject as usual.
v2: - Add SCX_OPS_OPEN() fix-ups for the new ops fields so cpu-form
schedulers setting them still load on older kernels. (Andrea)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A BPF scheduler can displace any of its tasks at will - cut a running one's
slice with an SCX_ENQ_PREEMPT dispatch, an SCX_KICK_PREEMPT kick or a direct
shortening, and jump a queued one with HEAD insertions. Sometimes the kernel
needs a slice and a DSQ position to stick regardless.
Add SCX_TASK_PROTECTED, guarding both:
- The slice becomes immutable. Every scheduler-reachable write is refused
and counted as SCX_EV_SLICE_DENIED. Higher scheduling classes are
unaffected. PREEMPT|IMMED can't preempt a running protected task and gets
reenqueued.
- A protected task that reached the head of its DSQ keeps it - HEAD
insertions land behind the leading run of protected tasks and reenqueue
sweeps skip them. Only rq-owned DSQs can hold protected tasks, so the walk
runs only for them.
The bit lives in p->scx.flags so that both the refusal and the head walk
read it under the rq lock that protects it.
Protection ends when the slice is consumed, when the task leaves the rq
except for a save/restore on the running task, on a yield, when the
scheduler enters bypass, and when the task leaves scx. The flag is
kernel-internal and not used yet.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
p->scx.slice and p->scx.dsq_vtime writes have no synchronization rules. The
dsq insert kfuncs write both fields synchronously from whatever context
they're called in - a direct dispatch from ops.select_cpu() writes with only
pi_lock held - and, as the kfuncs are safe to call spuriously with the
invalid dispatch discarded later, a scheduler can modify any task's slice by
spuriously calling them. The latter stands in the way of an upcoming patch
which adds kernel-granted slices that the schedulers must not be able to
modify.
Give both fields explicit rules. While the task is running, sleeping or
queued on an rq-owned DSQ, the rq lock protects them - these are the states
where the kernel consumes the slice. While queued on a user DSQ or on the
BPF side, the kernel neither consumes nor decides on the fields and every
writer acts for the BPF scheduler - synchronizing the writers is the
scheduler's responsibility and whichever write lands last wins.
To conform, an insert kfunc no longer writes the fields when called. The
values travel with the dispatch and take effect when the task is inserted. A
discarded dispatch has no side effects. The rq lock rule is asserted at the
slice store.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
SCX_ENQ_IGNORE_CAPS is kernel-internal and marks a placement the kernel
forces. scx_caps_for_enq() waives the enqueue cap for it, but a PREEMPT
insert still picks up the preemption cap requirement from
scx_caps_for_preempt(). Update scx_caps_for_preempt() to take enq_flags and
require nothing when SCX_ENQ_IGNORE_CAPS is set.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The dsq insert kfuncs reject __SCX_ENQ_INTERNAL_MASK bits in
scx_dsq_insert_preamble() instead of scx_vet_enq_flags(). A scheduler can
smuggle internal flags such as SCX_ENQ_CLEAR_OPSS through the dsq move
kfuncs and corrupt the dispatch protocol. Move the rejection into
scx_vet_enq_flags(). The vtime move wrapper OR'd the internal
SCX_ENQ_DSQ_PRIQ bit into enq_flags before the vet; the bit now goes in
inside scx_dsq_move() after the vet.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_bpf_now() couples the valid-or-fresh rq clock read to the current rq.
The read is useful for kernel-internal timing against a specific rq,
including a remotely locked one. Factor it out into __scx_bpf_now().
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
set_task_slice(), task_unlink_from_dsq(), move_local_task_to_local_dsq(),
init_dsq() and dump_line() will be used outside ext.c. Add the scx_ prefix
and declare them in internal.h. The scx_sched_all list will also be used
outside ext.c, drop its static. No functional changes.
v2: Declare scx_sched_all outside the CONFIG_EXT_SUB_SCHED block - the
definition is unconditional. (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The following rescue execution addition gives the function a third possible
destination, making a name that enumerates the outcomes a poor fit. Rename
to the destination-neutral scx_resolve_local_dsq(). No functional changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The KHO radix tree was designed to track preserved pages. So it does not
provide the capability to track any 64-bit key. Instead, it limits the
key width to how much it needs for tracking PFNs and their orders.
Limiting the width reduces the number of levels in the tree.
KHO is not expected to be the only user of the radix tree. With the API
generalized to allow other users, now it is possible to add any key to
the tree.
Check the key width at kho_radix_add_key(), and error out if it exceeds
what the tree can handle. Do this instead of increasing the tree depth
since right now there are no users that need to use wider keys, so this
avoids memory overhead and ABI breakage.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-4-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
The KHO radix tree constants are somewhat hard to understand. The tree
depth essentially comes from the max key width. The max key width comes
from the need to store a 52-bit PFN plus one more bit for the order.
All this is very obscure with the corrent code. The PFN width is defined
as KHO_ORDER_0_LOG2, which makes very little sense to a new reader not
already familiar with what the value means. Then the fact that an extra
bit is needed is hidden in the KHO_TREE_MAX_DEPTH calculation.
Simplify this by removing KHO_ORDER_0_LOG2 and replace it with
KHO_RADIX_KEY_WIDTH. Update the comment to explain why this value is
used. This moves the +1 from KHO_TREE_MAX_DEPTH to KHO_RADIX_KEY_WIDTH,
making things clearer.
Update kho_{encode,decode}_radix_key() to not use KHO_ORDER_0_LOG2.
Instead, refactor the code and comments to make it clearer how the
encoding and decoding is done.
In kho_encode_radix_key(), add a new variable for the shift for physical
address. Use that in calculating where the order bit goes and in
calculating the shifted PFN. Update comments to make this clearer.
In kho_radix_decode_key(), turn order_bit to 0-indexed to simplify the
eventual calculation for order. Touch up comments to make the
computation clearer.
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-3-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
The KHO radix tree is a data structure that can track the presence or
absence of an arbitrary key, with nothing inherently tied to KHO memory
preservation tracking. This was one of the design goals of the radix
tree. This was done to enable it to be re-used by other users of KHO.
Despite that, the radix tree APIs are very closely tied to KHO memory
preservation tracking. Adding a key is done by kho_radix_add_page(),
which encodes it as a page tracking operation and takes in PFN and
order. kho_radix_del_page() does the same. These functions encode the
key internally that goes into the radix tree. kho_radix_walk_tree() does
the same by baking the PFN and order into the callback arguments.
Generalize the APIs by taking the key directly and doing the encoding at
the callers. Rename the functions to kho_radix_add_key() and
kho_radix_del_key(). In practice, this removes a line each from the
functions and moves the encoding function call to the callers.
Similarly, update kho_radix_tree_walk_callback_t to take the key
directly.
Now that key encoding is no longer an inherent part of the radix tree
and can be decided by the user, rename kho_radix_{encode,decode}_key()
to kho_{encode,decode}_radix_key(). This moves them out of the
"kho_radix_" name space into the "kho_" namespace. This emphasizes that
this is KHO's way of encoding the key for its radix tree.
Reviewed-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260801084833.1897543-2-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
* fix a regression caused by allowing coexistence of KHO with deferred
initialization of the memory map.
-----BEGIN PGP SIGNATURE-----
iQEzBAABCgAdFiEEeOVYVaWZL5900a/pOQOGJssO/ZEFAmpweGsACgkQOQOGJssO
/ZHe4gf/b06J+YW9GO7FGQclCf3s30DbL8nNeEkb74lpWU8CAfk2vbo9QyrlYH8+
5wtISj9+SdBwMfY4XFdDWOur1EHD47N52xbT5zL3yp1UEPpXYbggiiKuzwCXK1MZ
lVhyc3XfPmF4eRP5V0Nr7AsN1pAujH83uRrTgg//ZJziOaOtLi7ZWJZa0Oae3UY1
z+/bRIM21DzkL2wD2fitlYxYBmrooYRS6FVr+wv/khBnLxAtcWGM+nfHGI4/i4bK
5oz/9D3+nD/rQBqOT6F+Vf6Dz3zPtJPLe7+B459x24f2yPzJplIt6cDq9nxZKbY2
L7q3npvcsts7iLXVsKSeaHeEfn2l7g==
=xzQP
-----END PGP SIGNATURE-----
Merge tag 'liveupdate-fixes-2026-08-03' of git://git.kernel.org/pub/scm/linux/kernel/git/liveupdate/linux
Pull liveupdate fix from Mike Rapoport:
- fix a regression caused by allowing coexistence of KHO with deferred
initialization of the memory map
* tag 'liveupdate-fixes-2026-08-03' of git://git.kernel.org/pub/scm/linux/kernel/git/liveupdate/linux:
kho: align kho_scratch to MAX_ORDER_NR_PAGES pages
The built-in idle masks are reset with all online CPUs marked idle
before sched_ext is enabled. Busy CPUs can therefore be incorrectly
advertised as idle until their next idle transition.
Initialize the masks empty so that the initial state is conservative.
When bypass is lifted, every CPU is rescheduled and idle-to-idle
re-picks populate the masks with CPUs that are actually idle. Later
idle transitions keep the masks up to date.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Reviewed-by: Kuba Piecuch <jpiecuch@google.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
- More lifecycle fixes for the new sub-scheduler support: a failed enable
could tear down a never-linked sub-scheduler in a way that races the root
scheduler's disable and leads to a use-after-free, tasks that were not on
the ext class could still get the enable callback, and a policy-rejection
path silently rewrote a running task's scheduling policy instead of
aborting the scheduler.
- Scheduler enable/disable could deadlock with cgroup removal and a
concurrent cgroup weight write through kernfs. Fixed by reordering lock
acquisition.
- Sync wakeups could leave the waker CPU incorrectly marked idle in the
built-in idle-CPU tracking.
- A selftest fix for sleeping tasks whose CPU affinity changes before
wakeup.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCam/mog4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGe0NAQC2UrrXeSm97RTCv+5HBfom/8sjPdB2tiAwGiUt
fUUQ+AD9FP30urJaoiALUiL+EGWpLFejwUCXNXlT+0kN2TiFxQM=
=EuWf
-----END PGP SIGNATURE-----
Merge tag 'sched_ext-for-7.2-rc6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext
Pull sched_ext fixes from Tejun Heo:
- More lifecycle fixes for the new sub-scheduler support: a failed
enable could tear down a never-linked sub-scheduler in a way that
races the root scheduler's disable and leads to a use-after-free,
tasks that were not on the ext class could still get the enable
callback, and a policy-rejection path silently rewrote a running
task's scheduling policy instead of aborting the scheduler.
- Scheduler enable/disable could deadlock with cgroup removal and a
concurrent cgroup weight write through kernfs. Fixed by reordering
lock acquisition.
- Sync wakeups could leave the waker CPU incorrectly marked idle in the
built-in idle-CPU tracking.
- A selftest fix for sleeping tasks whose CPU affinity changes before
wakeup.
* tag 'sched_ext-for-7.2-rc6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:
selftests/sched_ext: Handle sleeping task affinity changes in numa test
sched_ext: Mark waker CPU busy when selected in WAKE_SYNC case
sched_ext: Don't enable non-ext tasks in the sub-sched task loops
sched_ext: Skip sub-disable teardown for never-linked sub-schedulers
sched_ext: Take cgroup_lock() first in scx_cgroup_lock()
sched_ext: Reject setting disallow from init_task outside the enable path
- A pressure trigger's poll timer could be re-armed while the last trigger
was being torn down and then fire after the cgroup was freed. Tie the
timer to the cgroup's lifetime and shut it down when the cgroup is freed.
- Writing to a pressure file forked a worker kthread while holding the
cgroup mutex, creating lock dependencies from the mutex to the whole fork
path. A pressure write racing a sched_ext scheduler enable, which blocks
forks before grabbing the mutex, deadlocked. Fork the worker with the
mutex dropped.
- Documentation fix for io.latency behavior on non-rotational devices.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCam/mlw4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGVhRAQCyNBHCHpaY9erKTezenGDK6D+LqbfLWiWuIntB
swiwLQEA6h6Rgob2GDDRHOey7+XwF6PHh6xoh4FhSYboeEZd1wc=
=Re+7
-----END PGP SIGNATURE-----
Merge tag 'cgroup-for-7.2-rc6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup
Pull cgroup fixes from Tejun Heo:
- A pressure trigger's poll timer could be re-armed while the last
trigger was being torn down and then fire after the cgroup was freed.
Tie the timer to the cgroup's lifetime and shut it down when the
cgroup is freed.
- Writing to a pressure file forked a worker kthread while holding the
cgroup mutex, creating lock dependencies from the mutex to the whole
fork path. A pressure write racing a sched_ext scheduler enable,
which blocks forks before grabbing the mutex, deadlocked.
Fork the worker with the mutex dropped.
- Documentation fix for io.latency behavior on non-rotational devices.
* tag 'cgroup-for-7.2-rc6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
Docs/admin-guide/cgroup-v2: document io.latency rotational vs non-rotational behavior
sched/psi: Shut down rtpoll_timer in psi_cgroup_free()
sched/psi: Create the psimon kthread outside of cgroup_mutex
Commit 814cba835e ("bpf, x86: Fix trampoline stack size for 128-bit
arguments") changed the x86 trampoline to compute the number of
registers from arg_size for every argument, which removed the last user
of BTF_FMODEL_STRUCT_ARG. No other architecture or verifier code looks
at the flag, so remove the macro and the code in __get_type_fmodel_flags()
which sets it.
Keep BTF_FMODEL_SIGNED_ARG at BIT(1) rather than renumbering it to
BIT(0), so BIT(0) is available for a future flag.
No functional change.
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Leon Hwang <leon.hwang@linux.dev>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20260803052726.2821447-1-yonghong.song@linux.dev
uretprobes_srcu currently uses normal SRCU, which issues
two smp_mb() per read lock/unlock pair. This overhead is
paid on every uretprobe hit.
Switch to SRCU-fast-updown, which eliminates the per-reader
memory barriers by moving the ordering cost to the
grace-period side (synchronize_rcu() instead of smp_mb()).
This is acceptable because grace periods (uprobe
unregistration) are infrequent compared to reader-side
uretprobe hits.
The updown flavor is required because the SRCU read lock is
taken in prepare_uretprobe() when a return instance is
created and is held until that return instance is finalized.
The traced thread returns to user space in between, so the
lock is inherently released in a different context from
where it was acquired: on the normal return path via
uprobe_handle_trampoline() -> hprobe_finalize(), or from
ri_timer() (expiry) or dup_utask() (fork) via
hprobe_expire(). srcu_down_read_fast() / srcu_up_read_fast()
are designed for this acquire-here / release-elsewhere
pattern and, unlike the same-context srcu_read_lock_fast()
variant, do not carry the lockdep read-side tracking that
would warn on it.
The short, same-context SRCU sections in ri_timer() and
dup_utask() (which guard the uprobe against reuse across the
hprobe_expire() cmpxchg) instead use guard(srcu_fast_updown)
for proper lockdep coverage.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://patch.msgid.link/20260706172744.3920417-3-puranjay@kernel.org
Once binfmt_misc is loaded load_misc_binary() runs for every execve()
on the system since binfmt_misc registers at the head of the formats
list. Every exec therefore performs read_lock() and read_unlock() on
the entries_lock of the relevant binfmt_misc instance, i.e., two
atomic read-modify-writes on a shared cacheline. User namespaces
without their own binfmt_misc mount fall back to an ancestor's
instance so on container-heavy systems every exec on the machine
typically ends up hammering the cacheline of init_binfmt_misc. On
PREEMPT_RT the rwlock additionally turns the handler lookup into a
sleeping lock on the exec fast path.
The lock protects very little. Entries are immutable after publication
except for the Enabled bit which is already toggled locklessly via
set_bit()/clear_bit() and entry lifetime is already handled by the
users refcount via get_binfmt_handler()/put_binfmt_handler(). The read
lock's only remaining job is to make "the entry is still linked" and
"take a reference" atomic with respect to the unlink sites.
Switch the lookup to an RCU walk:
* Lookup walks the entry list under rcu_read_lock() and acquires a
reference via refcount_inc_not_zero(). The refcount can only drop to
zero after an entry has been unlinked so a failed increment means
the walk raced with an unlink. Restarting the search is bounded
because an unlinked entry cannot be found again.
* The unlink sites use hlist_del_init_rcu() which keeps the forward
pointer intact for concurrent walkers and preserves hlist_unhashed()
as the protection against double removal.
* The final put frees the entry via kfree_rcu() as a concurrent walker
may still dereference its flags, magic, mask, and inline strings.
They all live in the entry allocation itself and thus stay valid
until a grace period has elapsed. Closing the interpreter file stays
synchronous. It is only used with a reference already held and all
final puts run in process context.
* Writers remain serialized by the inode lock of the root dentry with
one exception. bm_evict_inode() called from generic_shutdown_super()
during umount unlinks entries without holding it. Keep a spinlock
around the unlink sites instead of relying on superblock lifetime
rules to make that exclusion implicit.
Handler removal semantics are unchanged. An exec that acquired a
reference just before its handler was unregistered already completes
with the removed handler today. The read lock never protected against
that, it only made the window smaller.
With this an exec that matches no binfmt_misc entry, the common case,
no longer writes to any shared cacheline at all.
Link: https://patch.msgid.link/20260710-work-binfmt_misc-locking-v3-5-a162f7cb58d6@kernel.org
Reviewed-by: Jori Koolstra <jkoolstra@xs4all.nl>
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
The upcoming conversion of the handler lookup to RCU walks cannot use
list_del_init(): reinitializing the forward pointer of a removed entry
would make a concurrent lockless walker standing on that entry loop
back onto it indefinitely. The removal paths do rely on
reinitialization though because bm_{entry,status}_write() and
bm_evict_inode() need to detect whether an entry has already been
unlinked.
hlists support exactly this pattern: hlist_del_init_rcu() keeps the
forward pointer of the removed entry intact for concurrent walkers and
only zeroes ->pprev with hlist_unhashed() serving as the linked test.
Convert the entry list to an hlist now while keeping the rwlock so the
subsequent RCU conversion is a pure locking change. hlist_add_head()
inserts at the head just as list_add() did so lookup precedence
between registered handlers is unchanged.
Link: https://patch.msgid.link/20260710-work-binfmt_misc-locking-v3-4-a162f7cb58d6@kernel.org
Reviewed-by: Jori Koolstra <jkoolstra@xs4all.nl>
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
Reported by the Sashiko AI review.
arena_vm_fault() returns VM_FAULT_RETRY when it can't take
arena->spinlock, but it never took mmap_lock. The fault path assumes a
VM_FAULT_RETRY handler already dropped mmap_lock and re-takes it on the
retry, so mmap_lock gets taken twice and can deadlock:
do_user_addr_fault()
{
fault = handle_mm_fault(...); // calls arena_vm_fault()
if (fault & VM_FAULT_RETRY)
goto retry; // re-locks mmap_lock
mmap_read_unlock(mm);
}
Return VM_FAULT_SIGBUS instead, for two reasons:
1. We could keep VM_FAULT_RETRY, but then we'd have to drop the fault
lock first and cap the retry ourselves, the way __folio_lock_or_retry()
does.
2. A failed raw_res_spin_lock_irqsave() already means a possible deadlock
was detected, so retrying just hits the same lock again.
So returning VM_FAULT_RETRY here is overkill.
Fixes: b8467290ed ("bpf: arena: make arena kfuncs any context safe")
Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/20260728060517.95183-1-jiayuan.chen@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Kfunc argument checking re-derives each argument's kfunc_ptr_arg_type
from BTF on every verification of a call in check_kfunc_args(). Now that
get_kfunc_arg_type() is a function of the kfunc's BTF alone, it no
longer inspects register state. The classification can be computed once
when the call is added and cached. This is a step toward describing
kfuncs with a bpf_func_proto and sharing the helper argument-checking
path.
Generate the classification at bpf_add_kfunc_call() time:
- Extend struct bpf_func_proto to be able to describe a kfunc: widen
arg_type[] and the arg_btf_id[]/arg_size[] union from 5 to
MAX_BPF_FUNC_ARGS, since a kfunc may take up to 12 arguments (5 in
registers, 7 on the stack).
- Embed a bpf_func_proto in struct bpf_kfunc_desc, populated by
gen_kfunc_arg_proto() which runs get_kfunc_arg_type() for each
argument and stores the result in proto.arg_type[]. Grow the
descriptor table's descs[] as a flexible array to not waste memory.
- check_kfunc_args() reads the cached classification from meta->fn
The KF_ARG_PTR_TO_CTX classification depends on the resolved program type,
and for BPF_PROG_TYPE_EXT that is the target program's type, which
resolve_prog_type() reads from prog->aux->saved_dst_prog_type. That field
is normally recorded later during verification in check_attach_btf_id(),
after bpf_add_kfunc_call() has run.
Record saved_dst_prog_type and saved_dst_attach_type from dst_prog at
program load time in bpf_prog_load() so the resolved type is available
at add-call time without reordering check_attach_btf_id(). This keeps
e.g. an freplace of an XDP program calling bpf_xdp_metadata_rx_hash()
classifying its struct xdp_md * argument as context.
The classification result is unchanged; it is only computed earlier and
cached.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-19-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Add kfunc scalar argument types, classify them in get_kfunc_arg_type()
along side with pointer arguments and move scalar type verification
into the main switch in check_kfunc_args(). This keeps BTF-based
classification separate from register validation for every argument,
paving the way for generating the kfunc argument prototype at add-call
time. No functional change intended.
KF_ARG_MEM_SIZE and KF_ARG_CONST_MEM_SIZE now are reachable. Therefore,
remove the fallthrough from KF_ARG_PTR_TO_MEM case and adjust the
register indexing.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-18-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Now that get_kfunc_ptr_arg_type() classifies a kfunc pointer argument
from its BTF alone, express a nullable argument by OR-ing PTR_MAYBE_NULL
into the classified type, and resolve a NULL register after
classification instead of before it.
Previously check_kfunc_args() short-circuited a nullable argument passed
a NULL register with a continue placed before get_kfunc_ptr_arg_type(),
so the NULL never reached classification. That kept a register-state
decision (bpf_register_is_null()) ahead of the BTF-based classification.
This mirrors how helper arguments carry PTR_MAYBE_NULL in their
bpf_arg_type and is a step toward describing kfuncs with a bpf_func_proto:
the nullability now travels with the per-argument classification, so it is
captured when the prototype is generated at add-call time.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-17-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
get_kfunc_ptr_arg_type() decided part of a kfunc pointer argument's type
from the caller's register: a PTR_TO_BTF_ID (or reg2btf_ids) register made
the argument KF_ARG_PTR_TO_BTF_ID, otherwise it fell through to a memory
buffer. Folding register state into argument classification prevents
describing a kfunc's arguments from its BTF alone, which is a prerequisite
for generating a helper-like prototype and eventually sharing the argument
checking (check_func_arg()) between helpers and kfuncs.
Classify pointer arguments from BTF only, and resolve them against the
register in check_kfunc_args():
- A pointer to a struct that is not paired with a __sz/__szk size
argument is classified KF_ARG_PTR_TO_BTF_ID and then checked against
the register. A register carrying a BTF ID (PTR_TO_BTF_ID or a
reg2btf_ids type) must be referenced or trusted and is matched against
the expected type. The only relaxation is when the struct is composed
of scalars, the register may be verified as a fixed-size memory buffer
sized from the BTF type; anything else is rejected.
- A pointer paired with a size argument is always a memory buffer and is
never classified as BTF_ID, so the __sz/__szk case no longer detours
through BTF_ID.
The new design now accepts one previously rejected case: passing
PTR_TO_BTF_ID to a pointer to scalar w/o a following __sz/__szk. The
argument will be classified as KF_ARG_PTR_TO_MEM | MEM_FIXED_SIZE. The
PTR_TO_BTF_ID register will go through check_mem_reg() ->
check_helper_mem_access() -> check_ptr_to_btf_access(). For a pointer to
scalar arg, a kernel btf id will be rejected unless explicitly granted
by btf_struct_access(); a program allocated btf id will be allowed.
The referenced-or-trusted check thus moves into the KF_ARG_PTR_TO_BTF_ID
resolution, alongside the type match.
get_kfunc_ptr_arg_type() no longer needs the register, so drop its regs
and reg parameters; it is now a pure function of the kfunc's BTF.
When a register cannot satisfy a BTF_ID argument, report the register type
passed and, when the expected struct has a reg2btf_ids mapping, the
register type that would be accepted, instead of a confusing "socket".
Update the affected selftest messages accordingly.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-16-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
A kfunc memory-pointer argument comes in two flavors: a fixed-size buffer
whose access size is derived from the pointed-to BTF type, and a
variable-size buffer paired with a following __sz/__szk size argument.
Both were represented by separate kfunc_ptr_arg_type values
(KF_ARG_PTR_TO_MEM vs KF_ARG_PTR_TO_MEM_SIZE) with the pointer classified
as the latter when a size argument followed.
Mirror how helpers describe the same distinction: classify both as
KF_ARG_PTR_TO_MEM and OR in MEM_FIXED_SIZE for the fixed-size case, just
as helpers use ARG_PTR_TO_MEM | MEM_FIXED_SIZE. The switches now key on
base_type(kf_arg_type) so the flag rides along, and the KF_ARG_PTR_TO_MEM
handler either resolves the size from BTF (MEM_FIXED_SIZE) or falls
through to the mem/size-pair check, which validates the buffer against the
following size register and skips it. No functional change.
Currently, KF_ARG_MEM_SIZE and KF_ARG_CONST_MEM_SIZE are only reachable
from ARG_PTR_TO_MEM fallthrough. A patch later will merge scalar
checking into the same switch and remove the fallthrough.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-15-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
get_kfunc_ptr_arg_type() returned KF_ARG_PTR_TO_NULL when a nullable
pointer argument was passed a NULL register. This folded a register-state
decision (bpf_register_is_null()) into what is otherwise BTF-based
argument classification, and it short-circuited before the BTF_ID/MEM
resolution.
Drop KF_ARG_PTR_TO_NULL and handle the NULL case in check_kfunc_args()
instead: a nullable argument that is actually NULL is skipped. Note that
it is okay to skip even when it is a mem+size pair because the size
argument check has been moved to the scalar section. The skip is done
before get_kfunc_ptr_arg_type() so that a NULL passed to a nullable
non-scalar-struct argument is not newly rejected by the BTF_ID/MEM
resolution.
No functional change.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-14-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
check_kfunc_args() already makes sure a scalar value is passed to a
scalar kfunc argument. Drop the check in is_kfunc_arg_mem_size() and
is_kfunc_arg_const_mem_size() to further decouple
get_kfunc_ptr_arg_type() from register state (a prerequisite for
generating a helper-like prototype from kfunc's BTF).
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-13-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
To align helper and kfunc pointer to memory argument handling, move
kfunc constant memorry size argument handling to the kfunc scalar
section. In addition, factor out constant scalar argument handling.
The constant size argument (__szk) of a kfunc memory/size pair was
recorded into meta->arg_constant by a dedicated block in the
KF_ARG_PTR_TO_MEM_SIZE case, duplicating the "only one constant
argument" and "must be a known constant" checks already in the generic
scalar argument handling. That block also did an explicit i++ to skip
the size argument.
This also fixes a precision gap: the old dedicated block did not mark
the size register precise, relying on check_mem_size_reg() for that. But
check_mem_size_reg() is skipped when the buffer is a nullable arg passed
as NULL (e.g. bpf_dynptr_slice(_rdwr) with a NULL buffer), so in that
case the __szk value was recorded and used for regs[R0].mem_size without
marking it precise. Routing the size through the scalar path marks it
precise in all cases.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-11-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
ARG_CONST_SIZE does not require a constant: check_mem_size_reg() accepts
any bounded scalar and verifies the memory access against its maximum
(reg_umax). Rename ARG_CONST_SIZE and ARG_CONST_SIZE_OR_ZERO to
ARG_MEM_SIZE and ARG_MEM_SIZE_OR_ZERO to reflect that. ARG_CONST_ALLOC_
SIZE_OR_ZERO, which does require a constant, is left unchanged.
Pure rename, no functional change.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-10-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Fixed-size memory arguments went through two paths: helpers called
check_helper_mem_access() directly, while kfuncs and global subprogs
used check_mem_reg(). Route the helper MEM_FIXED_SIZE case through
check_mem_reg() too so all three share the same check.
This also fixes a bug in the helper path. When passing a NULL to
PTR_MAYBE_NULL | ARG_PTR_TO_FIXED_SIZE_MEM argument, the program would
be falsely rejected by check_helper_mem_access(). This is not
triggerable since there is no such kind of helper. Also, note that
check_reg_type() still make sure NULL cannot be passed to an argument
not marked with PTR_MAYBE_NULL.
It also tightens the poisoned-stack-slot check. check_mem_reg() encoded
"a STACK_POISON slot may be read" as a negative access size for any
PTR_TO_STACK argument, but that is only sound for global subprogs, where
static stack liveness proved the callee body does not read those slots
(2cb27158ad ("bpf: poison dead stack slots")). Since check_mem_reg() is
also used for kfuncs, kfuncs accidentally inherited it and could read a
poisoned (dead, possibly uninitialized) stack slot. Restrict the negative
size to global subprogs (meta == NULL) so kfuncs, like helpers, require
the whole argument initialized.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-9-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Helper ARG_CONST_SIZE and kfunc KF_ARG_PTR_TO_MEM_SIZE memory arguments
already share check_mem_size_reg(), but the kfunc path reached it
through a thin wrapper, check_kfunc_mem_size_reg(). The wrapper existed
only to invoke check_mem_size_reg() twice. Once for BPF_READ and once for
BPF_WRITE because a kfunc mem argument may be both read and written,
whereas a helper argument carries a single access direction.
Let check_mem_size_reg() take a bitmask of access directions (widening
access_type to u32) and perform each requested access, then pass
BPF_READ | BPF_WRITE from the kfunc call site. This removes the
check_kfunc_mem_size_reg() wrapper so helper and kfunc mem+size arguments
run through exactly the same code.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-7-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
bpf_map_lookup_elem() is typed to return PTR_TO_MAP_VALUE for every
map, but for some map kinds the looked up value is actually a
different object: an inner map, a socket or an xsk socket.
Until now this reinterpretation happened once the pointer was
converted from its NULL-able form to a concrete value.
Such reinterpretation logic placement led to mark_ptr_not_null_reg()
being called for a temporary register copy in check_mem_reg() and
check_kfunc_mem_size_reg() (check_mem_size_reg() was buggy because of
not calling it). The temporary copy was necessary to pass
reinterpreted parameters as nullable helper and kfunc arguments.
Avoid this complication by refining map lookup result type right away.
The test case verifier_map_in_map/on_the_inner_map_pointer needs an
update because the verifier now prints a concrete NULL-able type for
the lookup.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-6-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
kfunc now shares the same bpf_call_arg_meta with helpers. Pass kfunc's
own meta to check_mem_reg() and check_kfunc_mem_size() instead of NULL
or a temporary meta on the stack.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-5-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Kfuncs used a single '__map' suffix (KF_ARG_PTR_TO_MAP) for two different
things: a verifier-known map matched by map_uid against a bound
timer/wq/task_work object (bpf_wq_init, bpf_task_work_schedule*), and an
opaque 'struct bpf_map *' used only at runtime (bpf_arena_*), which may be a
map fd or a PTR_TO_BTF_ID struct bpf_map (e.g. a bpf_map iterator's ctx->map).
That combined path only accepted the btf map form due to type confusion. The
'if (!reg->map_ptr)' check reads reg->map_ptr, which aliases reg->btf in the
bpf_reg_state union. A PTR_TO_BTF_ID register always has a non-NULL reg->btf,
so the guard silently passed and validation fell through to
process_kf_arg_ptr_to_btf_id(). It also recorded PTR_TO_BTF_ID info in
meta->map, which would be meaningless.
Split the annotation to avoid such type confusion and to align with
helper:
- '__const_map' -> KF_ARG_CONST_MAP_PTR: verifier-known map, handled by
process_map_ptr_arg() like helper ARG_CONST_MAP_PTR.
- '__map' -> KF_ARG_PTR_TO_BTF_ID: opaque struct bpf_map, validated by
process_kf_arg_ptr_to_btf_id(). A map fd still matches via
reg2btf_ids[CONST_PTR_TO_MAP], so bpf_arena_alloc_pages(&map) keeps
working.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-4-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Both the helper ARG_CONST_MAP_PTR and the kfunc KF_ARG_PTR_TO_MAP
recorded the map pointer in meta->map and, when a map was already
bound by a preceding timer/workqueue/task_work argument, rejected a
mismatching map.
Factor the logic into a single process_map_ptr_arg() used by both
paths. The bound-object name (timer, workqueue, or bpf_task_work) is
derived from the bound map's btf_record, and the register numbers in
the message are computed from the map argument position instead of
being hard-coded.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-3-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Drop process_timer_{helper,kfunc}() since bpf_call_arg_meta is now
shared by helper and kfunc. Call process_timer_func() directly.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260801074633.1595644-2-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
If the SCX_ENABLING -> SCX_ENABLED cmpxchg at the tail of
scx_root_enable_workfn() fails, the function jumps to err_disable
without setting ret. At that point ret still holds the return value
of the last successful __scx_init_task() call, which is 0, so the
err_disable fallback reports the meaningless message:
scx_root_enable() failed (0)
Set ret = -EBUSY, consistent with the other enable-state guards at
the top of the same function, so the fallback always reports a real
errno.
Signed-off-by: Liang Luo <luoliang@kylinos.cn>
Signed-off-by: Tejun Heo <tj@kernel.org>
The kernel-doc comment for sched_ext_ops::sub_cgroup_id uses the old
@cgroup_id name, which no longer matches the struct member. This
produces two kernel-doc warnings:
Warning: struct member sub_cgroup_id not described in sched_ext_ops
Warning: Excess struct member cgroup_id description in sched_ext_ops
Update the @param name to match the actual member.
Signed-off-by: Liang Luo <luoliang@kylinos.cn>
Signed-off-by: Tejun Heo <tj@kernel.org>
The trailing semicolon belongs at the point of use, not in the macro
definition. All uses have been verified to have their own semicolons.
This was found using the following Coccinelle semantic patch:
@r@
identifier i : script:ocaml() { String.lowercase_ascii i = i };
expression e;
@@
*#define i(...) e;
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Signed-off-by: Tejun Heo <tj@kernel.org>
A robust futex unlock stores 0 over the whole futex value - wiping
FUTEX_WAITERS - and wakes a single waiter. That wakeup is a one-shot
notification: the protocol relies on its recipient to either acquire the
futex (and eventually unlock while aware of the remaining contention) or
re-arm FUTEX_WAITERS before sleeping again. If the woken waiter is killed
before it can do either, the kernel must jump in and wake the next task
down the line.
This is a known complication of the futex protocol with a previous
partial fix in commit ca16d5bee5 ("futex: Prevent robust futex exit
race"). Unfortunately, that fix is insufficient.
If a third task re-acquired the futex through the uncontended fast
path in the meantime, the notification is lost: robust exit processing
sees that it is owned by another task and does nothing, while the new
owner sees no FUTEX_WAITERS when it unlocks and wakes nobody.
The remaining waiters sleep forever behind a free futex:
A owns the futex, B and C sleep in FUTEX_WAIT
uval == A | FUTEX_WAITERS
A robust unlock: store 0, FUTEX_WAKE(1) wakes B
uval == 0
D fast path acquire: cmpxchg(0 -> D)
uval == D, no FUTEX_WAITERS
B killed before acting on the wakeup
B exit walk, pending op: owner D != B -> no action
D unlock: no FUTEX_WAITERS -> no wake
C sleeps forever
This is clearly a shortcoming in the implementation, which fails to keep
the FUTEX_WAITERS bit consistent.
Work around this by augmenting the robust list exit processing to also
perform the extra wakeup if the futex word is owned by another thread but
FUTEX_WAITERS is not set.
This does not fix the problem of a non-contended take over/release and free
sequence, which has been discussed for years and has been addressed by
commit 3ca9595d9f ("futex: Add support for unlocking robust futexes") and
subsequent changes, but failed to take the problem described above into
account.
A more complete solution which is based on the in kernel unlock of
contended robust futexes has been discussed in the context of this change
and should show up in mainline sooner than later.
[ tglx: Amend change log slightly and fixup coding style ]
Fixes: ca16d5bee5 ("futex: Prevent robust futex exit race")
Signed-off-by: Keno Fischer <keno@juliahub.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Assisted-by: ClaudeCode:claude-fable-5 tla+
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260730194705.38981-1-keno@juliacomputing.com
- Reset dropped_count in mmio_reset_data()
When mmio_reset_data() is called, it does not reset the dropped_count
so that subsequent runs will have incorrect reporting.
- Add NULL check for mmio_trace_array in logging functions
The functions __trace_mmiotrace_rw() and __trace_mmiotrace_map()
may have the 'tr' variable passed to it as NULL. But they both
dereference it without checking if it is NULL first.
- Check return value of __register_event() in trace_module_add_events()
If __register_event() fails, the call after it (__add_event_to_tracers())
will create a file for it. If the module fails to load and its memory
is freed, the file will still point to it and it will not be removed
as the registering of the event did not complete. Only call
__add_event_to_tracers() if the __register_event() was successful.
- Fix false positive match in regex_match_full()
The regex full matching uses a strncmp() to test against the match
string and the value. It should not match if value is a prefix of
the string to match. Check to make sure the length of the strings
match before comparing.
- Fix reader page read offset for remote buffers
A page swapped in by __rb_get_reader_page_from_remote() retains its
stale read offset, causing subsequent reads to skip events or read
past valid data.
- Fix memory leak of subbuf_ids in rb_allocate_cpu_buffer()
Remote buffers allocate a subbuf_ids array. If the allocator function
fails after it is allocated, it does not free it, resulting in a
memory leak.
-----BEGIN PGP SIGNATURE-----
iIkEABYKADIWIQRRSw7ePDh/lE+zeZMp5XQQmuv6qgUCam1A0RQccm9zdGVkdEBn
b29kbWlzLm9yZwAKCRAp5XQQmuv6qlW/AQDCpddDwAMwN80i3tn5uiqec9JLZT5I
DR6G+0zEF9a5DAD2IxiD3gH/mPVZ3CdiqfElBqjc0Pq1dx414xmzXBv+CA==
=GUpD
-----END PGP SIGNATURE-----
Merge tag 'trace-v7.2-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace
Pull tracing fixes from Steven Rostedt:
- Reset dropped_count in mmio_reset_data()
When mmio_reset_data() is called, it does not reset the dropped_count
so that subsequent runs will have incorrect reporting.
- Add NULL check for mmio_trace_array in logging functions
The functions __trace_mmiotrace_rw() and __trace_mmiotrace_map() may
have the 'tr' variable passed to it as NULL. But they both
dereference it without checking if it is NULL first.
- Check return value of __register_event() in trace_module_add_events()
If __register_event() fails, the __add_event_to_tracers() call after
it will create a file for it. If the module fails to load and its
memory is freed, the file will still point to it and it will not be
removed as the registering of the event did not complete.
Only call __add_event_to_tracers() if the __register_event() was
successful.
- Fix false positive match in regex_match_full()
The regex full matching uses a strncmp() to test against the match
string and the value. It should not match if value is a prefix of the
string to match. Check to make sure the length of the strings match
before comparing.
- Fix reader page read offset for remote buffers
A page swapped in by __rb_get_reader_page_from_remote() retains its
stale read offset, causing subsequent reads to skip events or read
past valid data.
- Fix memory leak of subbuf_ids in rb_allocate_cpu_buffer()
Remote buffers allocate a subbuf_ids array. If the allocator function
fails after it is allocated, it does not free it, resulting in a
memory leak.
* tag 'trace-v7.2-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
ring-buffer: Fix subbuf_ids memory leak in rb_allocate_cpu_buffer() error path
ring-buffer: Fix reader page read offset for remote buffers
tracing/filters: Fix false positive match in regex_match_full()
tracing: Check return value of __register_event() in trace_module_add_events()
tracing/mmiotrace: Add NULL check for mmio_trace_array in logging functions
tracing/mmiotrace: Reset dropped_count in mmio_reset_data()
btf_distill_func_proto() builds the function model used for the
fentry/fexit/fmod_ret/fsession trampolines and struct_ops. It has
accepted a 16-byte __int128 return value since the trampoline was
introduced: __get_type_size() returns the integer's type size, and the
return-type check only rejected ret < 0.
But the BPF trampoline preserves only 8 bytes of the return value (RAX on
x86, i.e. R0). For an attach type that reads the target's return value the
second half (RDX / R3) is neither saved nor restored, so a program
attached to a function returning a 16-byte value corrupts the value seen
by the real caller and itself observes only half of it. struct_ops
trampolines have the same limitation.
This affects the attach types that read the target's return value: fexit,
fmod_ret and fsession (plus the _multi variants of fexit and fsession),
and struct_ops. fentry/fentry_multi run before the target returns and are
unaffected.
Reject a >8 byte return value for these attach types in
bpf_check_attach_target() and bpf_check_attach_btf_id_multi(), and for
struct_ops in bpf_struct_ops_desc_init().
Fixes: fec56f5890 ("bpf: Introduce BPF trampoline")
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Acked-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/20260729050159.2585809-1-yonghong.song@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
In rb_allocate_cpu_buffer(), cpu_buffer->subbuf_ids is allocated using
kcalloc() when buffer->remote is non-NULL. If a subsequent page allocation
fails (e.g., ring_buffer_desc_page() returns NULL or rb_allocate_pages()
fails), execution jumps to fail_free_reader.
While __free(kfree) automatically frees the outer cpu_buffer structure
at scope exit, kfree(cpu_buffer) does not recursively free nested heap
pointers such as cpu_buffer->subbuf_ids, resulting in a memory leak.
Fix this by explicitly freeing cpu_buffer->subbuf_ids in the
fail_free_reader error unwinding path when cpu_buffer->remote is set.
Link: https://patch.msgid.link/178550740672.380917.6067449683620196150.stgit@devnote2
Fixes: 2e67fabd8b ("ring-buffer: Introduce ring-buffer remotes")
Assisted-by: Antigravity:gemini-3.6-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Reviewed-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The untrusted PTR_TO_MEM early return skips pointer offset tracking
because accesses go through probe-read handling. Moving it after full
pointer-state propagation ensures scalar += untrusted_pointer leaves the
destination as PTR_TO_MEM instead of an unrelated scalar.
Fixes: f2362a57ae ("bpf: allow void* cast using bpf_rdonly_cast()")
Signed-off-by: Yiyang Chen <chenyy23@mails.tsinghua.edu.cn>
Tested-by: Daniel Wade <danjwade95@gmail.com>
Link: https://patch.msgid.link/20260729-c3-035-public-bpf-v4-v4-3-8ee297e2346b@mails.tsinghua.edu.cn
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
When scalar += pointer is handled in adjust_ptr_min_max_vals(), the
destination register inherits the pointer state from the source pointer.
Copying only selected fields is fragile because pointer provenance is
tracked by several bpf_reg_state fields.
Use the caller's temporary offset register to preserve the scalar operand
while replacing the destination with the full pointer state. This preserves
the frame number for PTR_TO_STACK registers and keeps parent identity
fields consistent.
Fixes: f4d7e40a5b ("bpf: introduce function calls (verification)")
Signed-off-by: Yiyang Chen <chenyy23@mails.tsinghua.edu.cn>
Tested-by: Daniel Wade <danjwade95@gmail.com>
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Link: https://patch.msgid.link/20260729-c3-035-public-bpf-v4-v4-2-8ee297e2346b@mails.tsinghua.edu.cn
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
The sanitize_err() function is called when:
- ptr += scalar
- scalar += ptr
- scalar += scalar
ALU operations are processed.
This commit drops offset and pointer registers parameters from its
signature to simplify the follow-up changes for 'scalar += ptr' case.
regs[src].type is safe to access, as it is not mutated by the callers.
Signed-off-by: Yiyang Chen <chenyy23@mails.tsinghua.edu.cn>
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Link: https://patch.msgid.link/20260729-c3-035-public-bpf-v4-v4-1-8ee297e2346b@mails.tsinghua.edu.cn
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Validate the functionality of LTL monitors by injecting events in a
controlled environment (KUnit) and expecting reactions, just like it is
done in DA monitors.
Reviewed-by: Nam Cao <namcao@linutronix.de>
Link: https://lore.kernel.org/r/20260723074534.43521-15-gmonaco@redhat.com
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
Some monitors do not only rely on tracepoint arguments but also on the
currently executing task.
This makes it more challenging to mock events in KUnit.
Define wrapper functions around current, the functionality is mocked
only during KUnit, an additional function call is avoided using a static
branch unless any (even unrelated) KUnit test is running.
Rely on a global mock_current variable that is set only by the RV KUnit
tests and cleared on teardown. Unrelated KUnit tests that happen to
trigger RV handlers would see it null and use current.
Reviewed-by: Nam Cao <namcao@linutronix.de>
Reviewed-by: Wen Yang <wen.yang@linux.dev>
Link: https://lore.kernel.org/r/20260723074534.43521-14-gmonaco@redhat.com
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
Validate the functionality of DA monitors by injecting events in a
controlled environment (KUnit) and expecting reactions.
Events handlers are exported directly from the monitor source files
without using system events and with dummy arguments (e.g. no real
tasks). If the provided sequence of events incurs a violation, the test
expects the stub version of rv_react() to be called.
This testing method can validate the entire monitor implementation since
it sits between the monitor and the system (in place of the
tracepoints). All sorts of system and timing events can be emulated
without affecting the running kernel.
Handlers and monitor functions are exported as part of a struct to
simplify the process of running KUnit tests from kernel modules.
Reviewed-by: Nam Cao <namcao@linutronix.de>
Link: https://lore.kernel.org/r/20260723074534.43521-13-gmonaco@redhat.com
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
Export rv_get_task_monitor_slot, rv_put_task_monitor_slot, and rv_react
to GPL modules so they can be accessed by KUnit and future monitors
built as kernel modules.
Reviewed-by: Nam Cao <namcao@linutronix.de>
Link: https://lore.kernel.org/r/20260723074534.43521-12-gmonaco@redhat.com
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
* pm-cpufreq:
cpufreq/amd-pstate: handle missing policy in dynamic EPP callbacks
cpufreq/amd-pstate: Cache the firmware programmed EPP value
cpufreq/amd-pstate: Toggle auto_sel in active mode on shared memory systems
cpufreq/amd-pstate: Fix EPP return type and handle errors during initialization
cpufreq: amd-pstate-ut: Skip tests when amd-pstate driver is not active
cpufreq: schedutil: Replace sprintf() with sysfs_emit() in sysfs show
cpufreq: schedutil: Fix self-contradictory comment in sugov_iowait_apply()
Documentation: admin-guide: cpufreq: fix sampling_rate example command
cpufreq: intel_pstate: Move two functions closer to callers
cpufreq: intel_pstate: Consolidate frequency values computation
cpufreq: intel_pstate: Introduce intel_pstate_update_freq_limits()
cpufreq: intel_pstate: Fix setting minimum P-state at init time
cpufreq: intel_pstate: Rename INTEL_PSTATE_HWP_BROADWELL
cpufreq: intel_pstate: Simplify HWP handling on Broadwell
cpufreq: intel_pstate: Adjust the .adjust_perf() driver callback
cpufreq: intel_pstate: Rearrange checks in hybrid_get_cost()
Align the rv_monitor variable name in LTL to the generic rv_this as it
is already done for DA/HA monitors. This improves consistency and eases
assumptions across model classes.
Reviewed-by: Nam Cao <namcao@linutronix.de>
Link: https://lore.kernel.org/r/20260723074534.43521-2-gmonaco@redhat.com
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
No caller checks the return value from stop_one_cpu_nowait(). All
callers require the callback to run and arrange for the target CPU's
stopper to remain enabled while queuing the work. In particular, commit
f0498d2a54 ("sched: Fix stop_one_cpu_nowait() vs hotplug") added
preemption protection to the scheduler callers so that queuing must
succeed once the target CPU has been observed online.
Therefore, a failure is an unrecoverable violation rather than a condition
individual callers can recover from. Diagnose it with WARN_ON_ONCE() in
stop_one_cpu_nowait(). A check in the common helper covers current and
future callers consistently, while individual checks would duplicate
the same non-recoverable handling at every call site.
Make the function return void because there is no longer a meaningful
result for callers to consume.
On UP, warn if the supplied CPU is not the current CPU because the work
cannot be scheduled in that case.
Signed-off-by: Yury Norov <ynorov@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Bradley Morgan <include@grrlz.net>
Reviewed-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Link: https://patch.msgid.link/20260729022355.325058-1-ynorov@nvidia.com
KCSAN reports a data race between readers_active_check() and a
concurrently executing reader:
BUG: KCSAN: data-race in readers_active_check / percpu_down_write
race at unknown origin, with read to 0xffff9f3eb5bf5f30 of 4 bytes
by task 1271 on cpu 14:
readers_active_check+0x...
percpu_down_write+0x152/0x1f0
value changed: 0xfffffff9 -> 0xfffffff8
readers_active_check() calls per_cpu_sum(*sem->read_count), which
iterates over all CPUs and reads each CPU's per-CPU read_count
variable. Concurrently, a reader on a remote CPU is modifying its own
CPU's read_count via this_cpu_inc() / this_cpu_dec() as it enters and
exits the critical section. These are plain reads and writes to the
same per-CPU storage, hence KCSAN flags a data race.
This race is benign. readers_active_check() is called from the
percpu_down_write() wait loop (rcuwait_wait_event) after sem->block is
already set. At this point:
- New readers must immediately back out (they see block set, decrement
their counter, and wake the writer), so counters can only decrease.
- If the sum catches a reader's increment before its decrement,
readers_active_check() sees a non-zero sum and returns false. The
writer merely iterates the wait loop again -- a harmless retry.
- A false zero (observing sum == 0 while a reader is still active)
cannot happen: per_cpu_sum() reads each CPU's counter, and each
per-CPU int read is atomic on all architectures, so an active
reader's counter is always seen as non-zero.
Annotate the read with data_race() to suppress the KCSAN warning and
document the intentional nature of this unlocked access.
Signed-off-by: Sun Shaojie <sunshaojie@kylinos.cn>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260623104132.505117-1-sunshaojie@kylinos.cn
register_lock_class() can return NULL when the lock class pool is
exhausted, graph_lock() fails, or key validation fails. However,
__lock_set_class() uses the return value directly in pointer arithmetic
without a NULL check:
class = register_lock_class(lock, subclass, 0);
hlock->class_idx = class - lock_classes;
If class is NULL, this computes a wild offset that corrupts
hlock->class_idx. The subsequent reacquire_held_locks() call will
invoke hlock_class() with this corrupted index, leading to a NULL or
out-of-bounds pointer dereference.
Add the missing NULL check, consistent with how __lock_acquire() already
handles this case at the same call site.
Fixes: 64aa348edc ("lockdep: lock_set_subclass - reset a held lock's subclass")
Signed-off-by: Naveen Kumar Chaudhary <naveen.osdev@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Waiman Long <longman@redhat.com>
Reviewed-by: Dmitry Ilvokhin <d@ilvokhin.com>
Link: https://patch.msgid.link/h2kfw43n4527x6mgi2lwpz2rieqnfzgictpv4wr5nyfjkc47co@2r5vz4uz44db
Forking a task that has a pending uretprobe can oops the kernel with a
NULL pointer dereference in the clone() path:
BUG: kernel NULL pointer dereference, address: 0000000000000018
Oops: 0002 [#1] SMP NOPTI
RIP: 0010:hprobe_expire
CR2: 0000000000000018
Call Trace:
uprobe_copy_process
copy_process
kernel_clone
__x64_sys_clone
do_syscall_64
entry_SYSCALL_64_after_hwframe
This was found on real hosts on Meta fleet.
I've got the impression that this is what is happening:
CPU 1 CPU 2 (traced task)
----- -------------------
hit uprobe, prepare_uretprobe():
hprobe LEASED, refcount >= 1
uprobe_unregister()
put_uprobe(): refcount -> 0
fork() -> dup_utask()
hprobe_expire(hprobe, true)
try_get_uprobe() -> NULL
get_uprobe(NULL) <-- Oops
Only take the extra reference when the uprobe is non-NULL; a NULL means
it is gone and is the correct value to return.
Fixes: dd1a756778 ("uprobes: SRCU-protect uretprobe lifetime (with timeout)")
Signed-off-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260729-uprobe-v1-1-61896b87c867@debian.org
SWIOTLB_FORCE has no remaining in-tree users. Forced bouncing is now
controlled through the swiotlb=force command line option via
swiotlb_force_bounce.
Remove the unused flag and simplify the force_bounce initialization.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-24-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Check the return value when converting swiotlb pools between encrypted and
decrypted mappings. If the default pool cannot be decrypted after early
initialization, mark the pool fully used so it cannot satisfy future bounce
allocations.
For late initialization, return the `set_memory_decrypted()` failure. For
restricted DMA pools, fail device initialization if the reserved pool
cannot be decrypted.
This prevents swiotlb from using pools whose encryption attributes do not
match their metadata, and avoids returning pages with uncertain encryption
state back to the allocator.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Petr Tesarik <ptesarik@suse.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-23-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
swiotlb_dyn_free() is used after removing a dynamic swiotlb pool from
RCU-protected lists. It can call swiotlb_free_tlb(), which may need to
restore the encryption state of an unencrypted pool with
set_memory_encrypted() before freeing the pages.
RCU callbacks run in atomic context, but set_memory_encrypted() is not
guaranteed to be atomic-safe on all architectures. For example, page
attribute updates may allocate page tables or take sleeping locks.
Use queue_rcu_work() for dynamic pool freeing instead. This keeps the RCU
grace period before freeing a published pool, while running the actual pool
teardown from workqueue context. Use the same helper for the transient-pool
error path, since that path may also be reached from atomic DMA mapping
context.
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Petr Tesarik <ptesarik@suse.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-22-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
ret in dma_direct_alloc() and dma_direct_alloc_pages() holds the returned
CPU mapping, not a generic return value. Rename it to cpu_addr and update
the remaining uses to match.
This makes the allocation paths easier to follow and keeps the local naming
consistent with what the variable actually represents.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Petr Tesarik <ptesarik@suse.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-21-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Make the dma-direct helpers derive the DMA address encoding from
__DMA_ATTR_ALLOC_CC_SHARED instead of implicitly relying on
force_dma_unencrypted() inside phys_to_dma_direct()
Pass an explicit unencrypted/decrypted state into phys_to_dma_direct(),
make the alloc paths return DMA addresses that match the requested buffer
encryption state. Also only call dma_set_decrypted() when
__DMA_ATTR_ALLOC_CC_SHARED is actually set.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Link: https://lore.kernel.org/r/20260717180442.110954-20-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Devices that are DMA non-coherent and require a remap were skipping
dma_set_decrypted(), leaving DMA buffers encrypted even when the device
requires unencrypted access. Move the call after the if (remap) branch
so that both the direct and remapped allocation paths correctly mark the
allocation as decrypted (or fail cleanly) before use.
Fix dma_direct_alloc() and dma_direct_free() to apply set_memory_*() to the
linear-map alias of the backing pages instead of the remapped CPU address.
Also disallow highmem pages for __DMA_ATTR_ALLOC_CC_SHARED, because highmem
buffers do not provide a usable linear-map address.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-19-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Teach dma_direct_map_phys() to select the DMA address encoding based on
DMA_ATTR_CC_SHARED.
Use phys_to_dma_unencrypted() for decrypted mappings and
phys_to_dma_encrypted() otherwise. If a device requires unencrypted DMA
but the source physical address is still encrypted, force the mapping
through swiotlb so the DMA address and backing memory attributes remain
consistent.
Update the arm64, x86, s390 and powerpc secure-guest setup to not use
swiotlb force option
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-18-aneesh.kumar@kernel.org
[mszyprow: rebased onto latest changes in arch/arm64/mm/init.c]
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
dma_direct_map_phys() is too large to benefit from being inlined. Move its
implementation to direct.c and leave the declaration in direct.h.
No functional change in this patch
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-17-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Teach dma_capable() about DMA_ATTR_CC_SHARED so the capability
check can reject encrypted DMA addresses for devices that require
unencrypted/shared DMA.
Also propagate DMA_ATTR_CC_SHARED in swiotlb_map() when the selected
SWIOTLB pool is decrypted so the capability check sees the correct DMA
address attribute.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Petr Tesarik <ptesarik@suse.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-16-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Fold encrypted/decrypted pgprot selection into dma_pgprot() so callers
do not need to adjust the page protection separately.
Update dma_pgprot() to apply pgprot_decrypted() when DMA_ATTR_CC_SHARED or
__DMA_ATTR_ALLOC_CC_SHARED is set and pgprot_encrypted() otherwise Convert
the dma-direct mmap paths to pass DMA_ATTR_CC_SHARED instead of open-coding
force_dma_unencrypted() handling around dma_pgprot().
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-15-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Teach swiotlb to distinguish between encrypted and decrypted bounce
buffer pools, and make allocation and mapping paths select a pool whose
state matches the requested DMA attributes.
Add a cc_shared flag to io_tlb_mem, initialize it for the default and
restricted pools, and propagate __DMA_ATTR_ALLOC_CC_SHARED into swiotlb
pool allocation. Reject swiotlb alloc/map requests when the selected pool
does not match the required encrypted/decrypted state.
Also return DMA addresses with the matching phys_to_dma_{encrypted,
unencrypted} helper so the DMA address encoding stays consistent with the
chosen pool.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-14-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Change swiotlb_tbl_map_single() to take the DMA mapping attributes by
reference and update the direct callers accordingly.
This is a preparatory change for a follow-up patch which updates the
attributes based on the selected swiotlb pool. Keeping the signature change
separate makes the follow-up patch easier to review.
No functional change in this patch.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Petr Tesarik <ptesarik@suse.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-13-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Teach the atomic DMA pool code to distinguish between encrypted and
unencrypted pools, and make pool allocation select the matching pool based
on DMA attributes.
Introduce a dma_gen_pool wrapper that records whether a pool is
unencrypted, initialize that state when the atomic pools are created, and
use it when expanding and resizing the pools. Update dma_alloc_from_pool()
to take attrs and skip pools whose encrypted state does not match
__DMA_ATTR_ALLOC_CC_SHARED. Update dma_free_from_pool() accordingly.
Also pass __DMA_ATTR_ALLOC_CC_SHARED from the swiotlb atomic allocation
path so decrypted swiotlb allocations are taken from the correct atomic
pool.
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-12-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Propagate force_dma_unencrypted() into __DMA_ATTR_ALLOC_CC_SHARED in the
dma-direct allocation path and use the attribute to drive the related
decisions.
This updates dma_direct_alloc(), dma_direct_free(), and
dma_direct_alloc_pages() to fold the forced unencrypted case into attrs.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Petr Tesarik <ptesarik@suse.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-11-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Move swiotlb allocation out of __dma_direct_alloc_pages() and handle it in
dma_direct_alloc() / dma_direct_alloc_pages().
This is needed for follow-up changes that simplify the handling of
memory encryption/decryption based on the DMA attribute flags.
swiotlb backing pages are already mapped decrypted by
swiotlb_update_mem_attributes() and rmem_swiotlb_device_init(), so
dma-direct should not call dma_set_decrypted() on allocation nor
dma_set_encrypted() on free for swiotlb-backed memory.
Update alloc/free paths to detect swiotlb-backed pages and skip
encrypt/decrypt transitions for those paths. Keep the existing highmem
rejection in dma_direct_alloc_pages() for swiotlb allocations.
Only for "restricted-dma-pool", we currently set `for_alloc = true`, while
rmem_swiotlb_device_init() decrypts the whole pool up front. This pool is
typically used together with "shared-dma-pool", where the shared region is
accessed after remap/ioremap and the returned address is suitable for
decrypted memory access. So existing code paths remain valid.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Jiri Pirko <jiri@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Link: https://lore.kernel.org/r/20260717180442.110954-8-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
swiotlb_alloc_tlb() can allocate from the DMA atomic pool when a decrypted
pool is needed from atomic context. With CONFIG_DMA_DIRECT_REMAP, the
atomic pool is backed by remapped virtual addresses, which are not the same
as the direct-map addresses returned by phys_to_virt().
swiotlb_init_io_tlb_pool() currently reconstructs the pool virtual address
from the physical start address. For atomic-pool backed allocations this
stores the wrong address in pool->vaddr. Later, swiotlb_free_tlb() passes
that address to dma_free_from_pool(), which will fail to recognize the
chunk
Pass the virtual address returned by the allocation path into
swiotlb_init_io_tlb_pool(), and store that address in pool->vaddr. This
keeps the pool free path using the same virtual address as the allocator.
Fixes: 79636caad3 ("swiotlb: if swiotlb is full, fall back to a transient memory pool")
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Petr Tesarik <ptesarik@suse.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Link: https://lore.kernel.org/r/20260717180442.110954-6-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
dma_direct_alloc_pages() may satisfy atomic allocations from the coherent
atomic pools. The pool allocation is keyed by the virtual address stored in
the gen_pool, but the pages API returns only the backing struct page.
On architectures with CONFIG_DMA_DIRECT_REMAP, atomic pool chunks are added
to the gen_pool using their remapped virtual address.
dma_direct_free_pages() reconstructs a linear-map address with
page_address(page) and passes that to dma_free_from_pool(). That address
does not match the gen_pool virtual range, so the pool lookup can fail and
the code can fall through to freeing a pool-owned page through the normal
page allocator path.
Add a page-based pool free helper that looks up the owning pool chunk by
physical address, translates it back to the gen_pool virtual address, and
frees that address to the pool. Use it from dma_direct_free_pages() while
keeping the existing virtual-address helper for coherent allocation frees.
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-5-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
atomic_pool_expand() frees the allocated pages from the remove_mapping
error path only when CONFIG_DMA_DIRECT_REMAP is enabled.
When CONFIG_DMA_DIRECT_REMAP is disabled, failures after page allocation,
such as gen_pool_add_virt(), jump to remove_mapping and return without
freeing the pages.
Move __free_pages(page, order) out of the CONFIG_DMA_DIRECT_REMAP block so
that cleanup paths always release the allocation.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Link: https://lore.kernel.org/r/20260717180442.110954-3-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Commit 5b138c534f ("dma-direct: factor out a dma_direct_alloc_from_pool
helper") changed dma_direct_alloc_from_pool() to return the CPU address
from dma_alloc_from_pool(). That fits dma_direct_alloc(), but
dma_direct_alloc_pages() also uses the helper and expects a struct page *.
Fix this by making dma_direct_alloc_from_pool() return the struct page *
again, and pass the CPU address back through an out-parameter for the
dma_direct_alloc() caller.
Fixes: 5b138c534f ("dma-direct: factor out a dma_direct_alloc_from_pool helper")
Cc: stable@vger.kernel.org
Tested-by: Michael Kelley <mhklinux@outlook.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar@kernel.org>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Link: https://lore.kernel.org/r/20260717180442.110954-2-aneesh.kumar@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Replace silent WARN_ON_ONCE with WARN_ONCE that prints the actual error
code from bpf_trampoline_unlink_prog(). This aids debugging of race
conditions during link teardown, while keeping the warning rate limited
to avoid log flooding.
This will be very helpful for speeding up trouble-shooting of some crash
UAF due to bpf_trampoline_unlink_prog failures.
No change to unlink behavior.
Signed-off-by: Xu Xin <xu.xin16@zte.com.cn>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Acked-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/20260729141159128mEJmS_aujBKr-cBu1p_UI@zte.com.cn
In bpf_link_show_fdinfo and bpf_link_get_info_by_fd, link->prog is
accessed without holding any locks. If the prog is concurrently replaced
via bpf_link_update, the old prog can be freed, leading to a potential
UAF issue.
Fix this by accessing link->prog under RCU protection to safely fetch
the pointer and guarantee its lifetime while reading its fields.
Fixes: 0c991ebc8c ("bpf: Implement bpf_prog replacement for an active bpf_cgroup_link")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Pu Lehui <pulehui@huawei.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Reviewed-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/f87b53c0-8f00-45a6-82db-8242fa9b143f@huaweicloud.com [0]
Link: https://lore.kernel.org/bpf/20260728025457.2814876-1-pulehui@huaweicloud.com
In bpf_netns_link_update_prog, the checks for old_prog and prog type
are currently performed locklessly before acquiring netns_bpf_mutex.
This creates a race condition that can lead to a UAF issue.
If two threads concurrently execute BPF_LINK_UPDATE on the same netns
link, the following execution path can trigger a UAF:
CPU0 CPU1
bpf_netns_link_update_prog
if (old_prog && old_prog != link->prog)
return -EPERM;
bpf_netns_link_update_prog
if (old_prog && old_prog != link->prog)
...
old_prog = xchg(&link->prog, new_prog);
bpf_prog_put(old_prog);
if (new_prog->type != link->prog->type) <-- trigger UAF
Fix this by moving the old_prog and prog->type checks inside the
netns_bpf_mutex critical section. Meanwhile, use guard() to simplify
lock management and avoid all the goto jumping.
Fixes: 7f045a49fe ("bpf: Add link-based BPF program attachment to network namespace")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Pu Lehui <pulehui@huawei.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Reviewed-by: Amery Hung <ameryhung@gmail.com>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/f87b53c0-8f00-45a6-82db-8242fa9b143f@huaweicloud.com [0]
Link: https://lore.kernel.org/bpf/20260728023259.2813482-1-pulehui@huaweicloud.com
-----BEGIN PGP SIGNATURE-----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=oqQe
-----END PGP SIGNATURE-----
Merge tag 'audit-pr-20260730' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/audit
Pull audit fixes from Paul Moore:
- Fix potential integer overflows in audit_log_n_string()
Similar to the earlier fix to audit_log_n_hex() that you merged
earlier in July. Expect a cleaner, and generally better fix for these
functions in an upcoming merge window, but this addresses the problem
in a small patch that should be easy for people to backport.
- Fix potential use-after-free in audit_del_rule()
* tag 'audit-pr-20260730' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/audit:
audit: fix potential use-after-free in audit_del_rule()
audit: fix potential integer overflow in audit_log_n_string()
- Allow fast frequency switching in the ACPI CPPC library only when
every supported control used by the driver callback has an address
space already accepted for fast access (Christian Loehle)
- Skip writes to unsupported performance controls in the ACPI CPPC
library (Christian Loehle)
- Update cppc_cpufreq_update_perf_limits() to read policy->min and
policy->max once and, if the lockless snapshot is inconsistent,
reduce the minimum to the observed maximum, along the lines of
cpufreq_driver_resolve_freq() (Christian Loehle)
- Fix a possible memory leak in the powernowk8_cpu_init() error
paths (Abdun Nihaal)
- Loosen the requirement on lowest nonlinear frequency != min freq in
the amd-pstate driver that is too tight for new systems some of
which actually have the lowest nonlinear frequency identical to the
minimum frequency (Mario Limonciello)
- Prevent amd-pstate from loading on unsupported hardware (Rong Zhang)
- Address an initialization race in the schedutil governor when it
runs on multi-CPU cpufreq policies, by making it initialize all
per-CPU structures first and only then publish the per-CPU
utilization update hooks (Zhongqiu Han)
-----BEGIN PGP SIGNATURE-----
iQFGBAABCAAwFiEEcM8Aw/RY0dgsiRUR7l+9nS/U47UFAmprmRgSHHJqd0Byand5
c29ja2kubmV0AAoJEO5fvZ0v1OO1Ro0H/3llnBhbu9nq7QcFmlPqcRh2yqWozPoG
oJTYEc/FEvPTEy65b6YxPFm64heiVTpv2gSZk0NVLnsXmDc3qNfoVFN7BvY8soPV
X4hEZbVi/ZE84l2gmc3t9HjK+j9mxfyWiWj/QDqZvZzf3v0QN8Klo9SvzlwXlnXX
1KdBpPPZYv7H0xIIOiL7K17JbvVTAXIwsV2SidOkm+29xArMYnHnzxqaIoh70kx7
vvRLR0NwubHzPZt66L5TFQCwixa+O71YaQ2l1k6TOzZ5G2BEtriXjAaimAd88iAu
Ogpie91bcERZdmoBIGtBCg9jSrHIqJE8ItugBsZw21BnIBsAKpgpJgw=
=U/2b
-----END PGP SIGNATURE-----
Merge tag 'pm-7.2-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management fixes from Rafael Wysocki:
"These fix issues related to cpufreq, in the ACPI CPPC library and the
generic CPPC cpufreq driver, in the powernow-k8 and amd-pstate
drivers, and in the schedutil governor:
- Allow fast frequency switching in the ACPI CPPC library only when
every supported control used by the driver callback has an address
space already accepted for fast access (Christian Loehle)
- Skip writes to unsupported performance controls in the ACPI CPPC
library (Christian Loehle)
- Update cppc_cpufreq_update_perf_limits() to read policy->min and
policy->max once and, if the lockless snapshot is inconsistent,
reduce the minimum to the observed maximum, along the lines of
cpufreq_driver_resolve_freq() (Christian Loehle)
- Fix a possible memory leak in the powernowk8_cpu_init() error paths
(Abdun Nihaal)
- Loosen the requirement on lowest nonlinear frequency != min freq in
the amd-pstate driver that is too tight for new systems some of
which actually have the lowest nonlinear frequency identical to the
minimum frequency (Mario Limonciello)
- Prevent amd-pstate from loading on unsupported hardware (Rong
Zhang)
- Address an initialization race in the schedutil governor when it
runs on multi-CPU cpufreq policies, by making it initialize all
per-CPU structures first and only then publish the per-CPU
utilization update hooks (Zhongqiu Han)"
* tag 'pm-7.2-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
cpufreq: powernow-k8: Fix possible memory leak in powernowk8_cpu_init()
ACPI: CPPC: Skip writes to unsupported performance controls
cpufreq/amd-pstate: Prevent the driver from loading on unsupported hardware
cpufreq/amd-pstate: Loosen requirement on lowest nonlinear frequency != min freq
cpufreq: schedutil: Publish util hooks only after all sg_cpu are initialized
cpufreq: cppc: Sanitize lockless policy limit snapshots
ACPI: CPPC: Check all controls for fast switching
The comments for call_rcu_tasks() and call_rcu_tasks_rude() referred
to "RCU read-side critical sections", which is ambiguous. specify
"rcu-tasks" and "rcu-tasks rude" respectively to match the actual
flavor being described.
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This 'commit 1612160b91 ("rcu-tasks: Eliminate deadlocks involving
do_exit() and RCU tasks")' has reimplemented the rcu_tasks_postscan(),
and the synchronize_srcu() has been removed, this commit therefore
rename tasks_rcu_exit_srcu_stall_timer to tasks_rcu_exit_stall_timer.
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The rdp->cpu_no_qs.s field is accessed only by the current CPU,
but can be accessed both at task level and from interrupt handlers.
All accesses from interrupts-enabled code must therefore be marked.
This commit therefore converts from __this_cpu_read() to this_cpu_read(),
but only in interrupts-enabled code, that is, the rcu_qs() function.
KCSAN located this issue.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
When CONFIG_KASAN=y and CONFIG_KASAN_STACK=y builds, the
show_rcu_gp_kthreads() exceeds the 1024-byte frame-size limit:
make kernel/rcu/tree.o KCFLAGS="-fstack-usage"
DESCEND objtool
DESCEND bpf/resolve_btfids
INSTALL libsubcmd_headers
CC kernel/rcu/tree.o
In file included from kernel/rcu/tree.c:4998:
kernel/rcu/tree_stall.h: In function 'show_rcu_gp_kthreads':
kernel/rcu/tree_stall.h:994:1: warning: the frame size of 1656 bytes is larger than 1024 bytes [-Wframe-larger-than=]
grep show_rcu kernel/rcu/tree.su
tree_nocb.h:1622:13:show_rcu_nocb_state 896 dynamic,bounded
tree_stall.h:933:6:show_rcu_gp_kthreads 1784 dynamic,bounded
tree_stall.h:1102:13:sysrq_show_rcu 16 static
Wrap the pr_info() into two noinline_for_stack helpers function:
show_rcu_state() print rcu_state status, and show_rcu_node()
print single rcu_node status.
After apply this change:
grep show_rcu kernel/rcu/tree.su
tree_stall.h:955:22:show_rcu_node 696 dynamic,bounded
tree_stall.h:930:22:show_rcu_state 872 dynamic,bounded
tree_nocb.h:1622:13:show_rcu_nocb_state 896 dynamic,bounded
tree_stall.h:972:6:show_rcu_gp_kthreads 544 static
tree_stall.h:1113:13:sysrq_show_rcu 16 static
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
rcu_all_qs() and rcu_note_context_switch() read/clear the per-CPU
->rcu_urgent_qs and ->rcu_need_heavy_qs flags with plain raw_cpu_read()
and this_cpu_write(), while the RCU core clears them with WRITE_ONCE() in
rcu_disable_urgency_upon_qs(). KCSAN flags the resulting same-CPU race:
BUG: KCSAN: data-race in rcu_all_qs / rcu_disable_urgency_upon_qs
It is benign -- the flags are advisory and rcu_all_qs() re-reads
->rcu_urgent_qs with smp_load_acquire() before acting on it -- but these
are the last unmarked accesses to the two flags; every other access
already uses READ_ONCE()/WRITE_ONCE()/smp_*. Mark them to match. No
functional change.
Reproduced on a PREEMPT_NONE, CONFIG_KCSAN_INTERRUPT_WATCHER=y kernel with
a pthreads program whose threads (two per CPU) loop reading a large file:
for (;;) {
int fd = open("/proc/kallsyms", O_RDONLY);
while (read(fd, buf, sizeof(buf)) > 0)
;
close(fd);
}
The read()s drive cond_resched() -> rcu_all_qs() while the busy CPUs keep
the grace period urgent, so the RCU core clears the flags concurrently.
Fixes: 2dba13f0b6 ("rcu: Switch urgent quiescent-state requests to rcu_data structure")
Signed-off-by: Itai Handler <itai.handler@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The rcu_check_gp_start_stall() works entirely on rnp parameter,
and never uses it's rdp parameter. this commit therefore drop it,
updating both callers and the declaration.
No functional change.
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The rdp->cpu_no_qs.b.norm field is accessed only by the current CPU,
but can be accessed both at task level and from interrupt handlers.
All accesses from interrupts-enabled code must therefore be marked.
This commit therefore converts from __this_cpu_read() to this_cpu_read(),
from __this_cpu_write() to this_cpu_write(), and plain C-language accesses
to READ_ONCE() and WRITE_ONCE(), but only in interrupts-enabled code.
KCSAN located this issue.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Commit 14a8570564 ("sched/deadline: Use revised wakeup rule for
dl_server") applies the revised wakeup rule to any server, as a result
servers that are not running (dl_defer_running == 0) and start with a
deadline overflow get enqueued and can boost tasks as if they were
running, invalidating the defer rule and the documented state model.
Apply the revised wakeup rule only for deferrable servers that are
marked as running.
Fixes: 14a8570564 ("sched/deadline: Use revised wakeup rule for dl_server")
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Tested-by: Andrea Righi <arighi@nvidia.com>
Link: https://patch.msgid.link/20260522125833.264145-1-gmonaco@redhat.com
A page swapped in by __rb_get_reader_page_from_remote() retains its
stale read offset, causing subsequent reads to skip events or read
past valid data. Fix it.
Link: https://patch.msgid.link/20260729133609.4022734-1-vdonnefort@google.com
Fixes: fbd1743ecb ("ring-buffer: Add non-consuming read for ring-buffer remotes")
Signed-off-by: Vincent Donnefort <vdonnefort@google.com>
Reviewed-by: Keir Fraser <keirf@google.com>
Tested-by: Keir Fraser <keirf@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
`audit_del_rule()` destroys `e->rule.exe` via `audit_remove_mark_rule()`
before unlinking the rule from RCU-visible filter lists and waiting for a
grace period. Concurrent readers in `audit_filter()` and
`audit_filter_rules()` still dereference `e->rule.exe`, while the fsnotify
mark can be freed on an independent lifetime path. This creates a
use-after-free window during rule deletion.
Fix this by unlinking the rule from the RCU-visible lists and invoking
`synchronize_rcu()` before calling `audit_remove_mark_rule()` (and other
rule removal helpers). This ensures that all existing RCU readers have
exited the critical section before any underlying resources are destroyed.
Cc: stable@vger.kernel.org
Fixes: 34d99af52a ("audit: implement audit by executable")
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: Codex:gpt-5.4
Signed-off-by: Luxiao Xu <rakukuip@gmail.com>
Signed-off-by: Ren Wei <enjou1224z@gmail.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
audit_log_n_string() computes new_len as "slen + 3" (enclosing quotes
plus the NUL terminator) and stores it into an int, while slen is a
size_t. For a sufficiently large slen the addition can overflow and/or
the result be truncated when assigned to the int new_len, so the
"new_len > avail" check can be bypassed and the subsequent
memcpy(ptr, string, slen) can write past the skb tail.
This is the same class of bug that was fixed for the hex sibling in
commit 65dfde57d1 ("audit: fix potential integer overflow in
audit_log_n_hex()"); both helpers are reached through
audit_log_n_untrustedstring() with the same length source.
Make new_len a size_t and use check_add_overflow() to catch the
overflow, mirroring the audit_log_n_hex() fix. No functional change for
the in-tree callers, which all pass bounded lengths.
Cc: stable@vger.kernel.org
Fixes: 168b717395 ("AUDIT: Clean up logging of untrusted strings")
Signed-off-by: Zhan Xusheng <zhanxusheng@xiaomi.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
regex_match_full() calls strncmp(str, r->pattern, len) where len is the
target field buffer size. When len is smaller than r->len (the filter
pattern length), strncmp() checks only len bytes of r->pattern against
str. If those len bytes match, strncmp() returns 0, resulting in a
false-positive match where a shorter string in a fixed-size field
matches a longer filter pattern.
For example, a 4-byte static string field containing "abcd" matched the
filter pattern "abcdefgh" because strncmp("abcd", "abcdefgh", 4)
returned 0. In this case, @len does NOT include '\0' because it is
fixed-size array.
Fix this by returning 0 (no match) early when len < r->len.
Fixes: 1889d20922 ("tracing/filters: Provide basic regex support")
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/178528488779.124250.5571741156199253769.stgit@devnote2
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
trace_module_add_events() ignores the return value of __register_event()
and unconditionally calls __add_event_to_tracers() for each event.
If __register_event() fails (for example, if event_init() fails), the
trace_event_call is not added to ftrace_events list, but
__add_event_to_tracers() still creates a trace_event_file pointing to it.
If module loading subsequently fails and module memory is freed, tracing
state retains a stale trace_event_call pointer in trace_event_file,
leading to a use-after-free when tracefs or tracing subsystem operations
are later executed.
Fix this by checking the return value of __register_event() and only
calling __add_event_to_tracers() if event registration succeeded.
Fixes: ae63b31e4d ("tracing: Separate out trace events from global variables")
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/178528487878.124250.14170824576025743236.stgit@devnote2
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
mmio_trace_rw() and mmio_trace_mapping() retrieve mmio_trace_array into
tr and pass it to __trace_mmiotrace_rw() and __trace_mmiotrace_map().
If these functions are invoked while mmio_trace_array is NULL (e.g. before
initialization or after disabled), accessing tr->array_buffer.buffer will
result in a NULL pointer dereference crash.
Fix this by adding an explicit NULL check for tr at the beginning of
__trace_mmiotrace_rw() and __trace_mmiotrace_map().
Link: https://patch.msgid.link/178524300062.56416.8362487250709962380.stgit@devnote2
Fixes: f984b51e07 ("ftrace: add mmiotrace plugin")
Assisted-by: Antigravity:gemini-3.6-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
mmio_reset_data() is called during tracer initialization, reset, and
start. While it resets overrun_detected and prev_overruns, it neglects
to reset dropped_count. Consequently, dropped event counts from prior
tracing sessions persist in dropped_count and corrupt overrun reports
in subsequent runs.
Fix this by explicitly calling atomic_set(&dropped_count, 0) in
mmio_reset_data().
Link: https://patch.msgid.link/178524299122.56416.16277704230639425172.stgit@devnote2
Fixes: 173ed24ee2 ("mmiotrace: count events lost due to not recording")
Assisted-by: Antigravity:gemini-3.6-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
In register_fprobe(), get_ips_from_filter() resolves target function
addresses and increments module reference counts via try_module_get() for
symbols in kernel modules. If get_ips_from_filter() fails on the second
pass and returns an error, register_fprobe() returned directly without
releasing module references acquired up to that point.
Fix this by ensuring the cleanup loop executing module_put() runs even when
get_ips_from_filter() returns a negative error.
Link: https://lore.kernel.org/all/178528125360.101985.4144133640239273153.stgit@devnote2/
Fixes: d24fa977ee ("tracing: fprobe: Fix to lock module while registering fprobe")
Assisted-by: Antigravity:gemini-3.6-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
The futex address must be naturally aligned and this is checked via
"address % size" where `address' is the supplied address and `size' is
the expected size of futex. It is guaranteed that `size' is power of two
but the compiler does not see it and creates here a `div' operation
(x86, arm, gcc-15).
We can take advantage of the pow2 property and rewrite it as
"address & (size-1)".
As per testing, the command
|perf bench futex hash -f 1 -b 16384 -t 1 -r 30
improved from
| [thread 0] futex: 0x5619f931f740 [ 7001583 ops/sec ]
to
| [thread 0] futex: 0x55da173e5740 [ 7376137 ops/sec ]
or by 5.3%
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260701161736.xYYizA0e@linutronix.de
Runtime constify the read-only after init data __futex_shift(shift_32),
__futex_mask(mask_32), and __futex_queues(ptr) used in __futex_hash()
hot path to avoid referencing global variable.
This also allows __futex_queues to be allocated dynamically to
"nr_node_ids" slots instead of reserving config dependent MAX_NUMNODES
(1 << CONFIG_NODES_SHIFT) worth of slots upfront.
Runtime constants are initialized before their first access and
runtime_const_init() provides necessary barrier to ensure subsequent
accesses are not reordered against their initialization.
No functional changes intended.
perf bench futex on a 3rd Gen EPYC (2 x 64C/128T):
+----------------+-----------+-----------+-----------+--------------+
| Benchmark | Kernel 1 | Kernel 2 | Unit | % Improvement|
| | (avg/5) | (avg/5) | | (K2 vs K1) |
+----------------+-----------+-----------+-----------+--------------+
| Wake-parallel | 0.01614 | 0.00456 | ms | +71.75% |
| Requeue | 0.26394 | 0.24644 | ms | +6.63% |
| Lock-pi | 34.0 | 57.2 | ops/sec | +68.24% |
+----------------+-----------+-----------+-----------+--------------+
Performance testing on a 144-thread Intel(R) Xeon(R) CPU E7-8890 v3 (4 NUMA nodes):
+-------------------------------------------------------------+
| perf bench futex hash -b 0 |
+----------------------+------------+------------+------------+
| Configuration | As-is | Patched | Delta |
+----------------------+------------+------------+------------+
| 1 thread, 1 futex | 6,449,632 | 6,532,004 | +1.28% |
| 144 threads, 1024 fx | 2,111,486 | 2,139,685 | +1.34% |
+----------------------+------------+------------+------------+i
[ prateek: Dynamically allocate __futex_queues, mark the global data
__ro_after_init since they are constified after futex_init(). ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reported-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> # MAX_NUMNODES bloat
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Charlie Jenkins <thecharlesjenkins@gmail.com>
Tested-by: Charlie Jenkins <thecharlesjenkins@gmail.com>
Link: https://patch.msgid.link/20260227161841.GH606826@noisy.programming.kicks-ass.net
Link: https://patch.msgid.link/20260728052540.4728-9-kprateek.nayak@amd.com
kho_preserved_memory_reserve() calculates the size of a preservation by
doing 1 << (order + PAGE_SHIFT). Since the '1' is a 32-bit integer, it
can only be shifted by 31. That is, it will only work for preservations
up to 2 GiB. Larger preservations will trigger undefined behaviour.
While preservations larger than 2 GiB can't be obtained via folios
currently, they can be obtained via kho_preserve_pages().
For example, memblock reserve_mem uses kho_preserve_pages().
Reservations larger than 2 GiB are valid and will trigger this bug if
properly aligned.
Fix it by using 1UL for shifting.
Fixes: fc33e4b44b ("kexec: enable KHO support for memory preservation")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260727150240.889555-1-pratyush@kernel.org
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
The nodemasks in these structures may come from a variety of sources,
including tasks and cpusets - and should never be modified by any code
when being passed around inside another context.
Link: https://lore.kernel.org/20260609002919.3967782-1-gourry@gourry.net
Signed-off-by: Gregory Price <gourry@gourry.net>
Reviewed-by: Lorenzo Stoakes <ljs@kernel.org>
Acked-by: David Hildenbrand (Arm) <david@kernel.org>
Reviewed-by: Barry Song <baohua@kernel.org>
Acked-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Tested-by: SeongJae Park <sj@kernel.org>
Acked-by: SeongJae Park <sj@kernel.org>
Acked-by: Waiman Long <longman@redhat.com>
Acked-by: Zi Yan <ziy@nvidia.com>
Cc: Axel Rasmussen <axelrasmussen@google.com>
Cc: Baoquan He <baoquan.he@linux.dev>
Cc: Brendan Jackman <jackmanb@google.com>
Cc: Chris Li <chrisl@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Kairui Song <kasong@tencent.com>
Cc: Kemeng Shi <shikemeng@huaweicloud.com>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Michal Koutný <mkoutny@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Shakeel Butt <shakeel.butt@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Wei Xu <weixugc@google.com>
Cc: Yuanchu Xie <yuanchu@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The BUG_ON(axp->pid_count >= AUDIT_AUX_PIDS) check in
audit_signal_info_syscall() will panic the kernel if the signal target
pid array overflows, which is too severe for this situation.
Replace it with a WARN_ON_ONCE() and return of -EINVAL, instead.
Signed-off-by: Ricardo Robaina <rrobaina@redhat.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
The BUG_ON(!mutex_is_locked(&audit_filter_mutex)) check in
audit_add_to_parent() will panic the kernel if the mutex is not held,
which is too severe for this situation.
Replace it with lockdep_assert_held(), instead.
Signed-off-by: Ricardo Robaina <rrobaina@redhat.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
enable_trace_fprobe() sets the file link or the TP_FLAG_PROFILE flag and
then registers each trace_fprobe in the probe list. If
__register_trace_fprobe() fails partway through, the function returns
immediately without unregistering the trace_fprobes it already registered
or undoing the file link / flag it set, leaving the event half-enabled and
leaking the registered fprobe(s).
enable_trace_kprobe() already handles this with a rollback path. Do the
same for fprobe: on failure, unregister all probes and clear the file link
or profile flag.
Link: https://lore.kernel.org/all/20260724064208.480030-1-raushan.jhon@gmail.com/
Fixes: 334e5519c3 ("tracing/probes: Add fprobe events for tracing function entry and exit.")
Cc: stable@vger.kernel.org
Signed-off-by: Raushan Patel <raushan.jhon@gmail.com>
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
traceprobe_expand_meta_args() parses $argN with simple_strtoul() and
calls sprint_nth_btf_arg(n - 1, ...). For $arg0, n is 0 so the index is
-1. Because ctx->nr_params is signed, the "idx >= nr_params" guard in
sprint_nth_btf_arg() does not catch the negative index, and
ctx->params[-1].name_off is read out of bounds.
The normal per-argument path (parse_probe_vars()) already rejects
$arg0 via its argument-number check, but meta-argument expansion runs
before per-argument parsing and substitutes the value first, bypassing
that check.
Reject $arg0 explicitly during expansion.
Link: https://lore.kernel.org/all/20260724054435.146279-1-raushan.jhon@gmail.com/
Fixes: 18b1e870a4 ("tracing/probes: Add $arg* meta argument for all function args")
Cc: stable@vger.kernel.org
Signed-off-by: Raushan Patel <raushan.jhon@gmail.com>
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
MAX_COMMON_HEAD_LEN (63) was used to allocate a temporary buffer for
formatting command heads in trace_kprobe_match_command_head() and
trace_uprobe_match_command_head(). However, the buffer size is too
short for some longer symbols. Especially, with rust code, the symbol
can be mangled and become very long.
Refactor trace_kprobe_match_command_head() to perform direct string
comparisons using strcmp() and strncmp(), eliminating the need for a
temporary buffer and removing the MAX_COMMON_HEAD_LEN string length
restriction on probe symbol names.
For trace_uprobe_match_command_head(), since tu->filename is already
matched via strncmp(), use a fixed 64-byte stack buffer solely for
formatting offset and ref_ctr_offset (which requires at most 39 bytes).
With all users converted, remove the MAX_COMMON_HEAD_LEN definition from
trace_probe.h.
Link: https://lore.kernel.org/all/178521361102.34226.9650586522488974115.stgit@devnote2/
Reported-by: Zhan Xusheng <zhanxusheng1024@gmail.com>
Link: https://lore.kernel.org/all/20260724023317.624074-1-zhanxusheng@xiaomi.com/
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
In preparation for removing the strlcat API[1],
replace the string concatenation logic with a struct seq_buf,
which tracks the current position and the remaining space internally.
Use seq_buf_str() to NUL-terminate before passing to early_enable_events().
Link: https://github.com/KSPP/linux/issues/370 [1]
Link: https://patch.msgid.link/20260713045249.69942-1-woradorn.laon@gmail.com
Signed-off-by: Woradorn Laodhanadhaworn <woradorn.laon@gmail.com>
[ Moved placement of #include <linux/seq_buf.h> ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Replace strcpy() with strscpy() in __trace_find_cmdline() for
consistency with the existing strscpy() call in the same function,
and to avoid potential buffer overflow as flagged by the Kernel
Self Protection Project.
Link: https://patch.msgid.link/20260705173648.5418-1-posheng.lin.tw@gmail.com
Signed-off-by: Po-Sheng Lin <posheng.lin.tw@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Currently on boot up and when modules are loaded, the trace event
infrastructure will examine the TP_printk's of every event looking to see
if it dereferences pointers on the ring buffer via printk formats like
"%pB" and such. What it doesn't do is check if the arguments themselves
do a dereference from a pointer.
This was brought with a fix[1] to the fsl_edma event that had in the
arguments of the TP_printk(): "__entry->edma->membase"
The __entry->edma is a pointer saved in the ring buffer. The dereference
from TP_printk() happens when the user reads the "trace" file which can be
seconds, minutes, hours, days, weeks, or even months later! There is no
guarantee that the __entry->edma pointer will still be pointing to what it
was when it was recorded, and could crash the kernel when a user reads the
event.
Add logic to the test_event_printk() that also checks for this case and
warn if the event dereferences a pointer from the ring buffer.
[1] https://lore.kernel.org/all/20260630200022.1826420-1-martin@kaiser.cx/
Link: https://patch.msgid.link/20260630184836.74d477b6@gandalf.local.home
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Reviewed-by: Martin Kaiser <martin@kaiser.cx>
Reviewed-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
preemptirq_delay_test accepts cpu_affinity as a module parameter and,
when it is non-negative, writes that CPU directly into a temporary
cpumask from the worker thread. Values outside nr_cpu_ids can set a bit
outside the allocated cpumask before the test reports a normal affinity
error.
Validate the requested CPU in preemptirq_delay_run() before setting it
in the temporary cpumask. Invalid affinity requests are reported by
the test thread and skipped before cpumask_set_cpu() can touch an
out-of-range bit.
Link: https://patch.msgid.link/20260628131021.2208632.6a5c6c959813.preemptirq-delay-test-invalid-cpu-affinity@trailofbits.com
Assisted-by: Codex:gpt-5.5-cyber-preview
Signed-off-by: Samuel Moelius <sam.moelius@trailofbits.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
A single closing curly bracket should be put into a trace sequence buffer.
Thus use the corresponding function “trace_seq_putc”.
The source code was transformed by using the Coccinelle software.
Link: https://patch.msgid.link/d215fa89-9a62-4067-86ec-833290f35c80@web.de
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
A single string should be put into a sequence within a loop.
Thus use the corresponding function “seq_puts” for one selected call.
The source code was transformed by using the Coccinelle software.
Link: https://patch.msgid.link/1cf327f0-49a6-477f-a06f-2b22a167db24@web.de
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Single characters should be put into a sequence. Thus use the corresponding
function “seq_putc” for selected calls.
The source code was transformed by using the Coccinelle software.
Link: https://patch.msgid.link/6bcaa4da-05c6-4097-90f5-3969f8a1dfbc@web.de
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
expr_str() allocates a fixed MAX_FILTER_STR_VAL buffer and then builds
expression names with a series of raw strcat() appends. Nested operands,
constants, field flags, and generated field names can push the rendered
string past that fixed limit before the name is attached to the hist
field.
Build expression strings with seq_buf and return -E2BIG when the
rendered name would exceed MAX_FILTER_STR_VAL. This keeps the existing
tracing-side limit while replacing the raw append logic with bounded
construction.
Link: https://patch.msgid.link/20260611055945.22348-4-pengpeng@iscas.ac.cn
Signed-off-by: Pengpeng Hou <pengpeng@iscas.ac.cn>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
expr_str() currently reports all failure cases as NULL, so callers cannot
distinguish invalid recursion depth from allocation failure or later
string construction errors.
Return ERR_PTR()-encoded errors from expr_str() and make parse_unary()
and parse_expr() propagate them. Clear expr->name before destroying the
hist field so the error pointer is not freed as a string.
Link: https://patch.msgid.link/20260611055945.22348-3-pengpeng@iscas.ac.cn
Signed-off-by: Pengpeng Hou <pengpeng@iscas.ac.cn>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
expr_str() allocates a temporary expression buffer and manually frees it
on some error paths.
Convert the buffer to __free(kfree) and return it with return_ptr() on
success. This keeps ownership handling separate from the later ERR_PTR()
conversion and string-bound change.
Link: https://patch.msgid.link/20260611055945.22348-2-pengpeng@iscas.ac.cn
Signed-off-by: Pengpeng Hou <pengpeng@iscas.ac.cn>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The HIST_FIELD_FL_CONST path uses the fixed "u64" type string.
Point hist_field->type directly to the string literal, matching the
HIST_FIELD_FL_HITCOUNT path. The release path already uses kfree_const(),
so no duplication is needed.
Link: https://patch.msgid.link/20260527023450.2137639-1-pengyu@kylinos.cn
Signed-off-by: Yu Peng <pengyu@kylinos.cn>
Acked-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Add events/<sys>/<event>/btf_ids, a per-template file that exposes
the BTF ids resolve_btfids fills in for each tracepoint:
btf_obj_id BTF object owning the ids below
raw_btf_id FUNC_PROTO of __bpf_trace_<call> (named args), consumed
by raw_tp / tp_btf BPF programs
tp_btf_id trace_event_raw_<call> ring-buffer record, consumed by
classic BPF_PROG_TYPE_TRACEPOINT programs
DECLARE_EVENT_CLASS now emits a 2-entry BTF_ID_LIST (FUNC __bpf_trace_*
and STRUCT trace_event_raw_*) and stores the pointer in
trace_event_class.
Per-syscall events under syscalls/ share the handcrafted classes
event_class_syscall_{enter,exit} instead of going through
DECLARE_EVENT_CLASS. Wire those classes to the BTF id lists
generated for sys_enter / sys_exit so all ~700 per-syscall
events expose the shared dispatcher prototype and record.
The per-syscall events do not own their own tracepoint
(they share sys_enter/sys_exit), so raw_btf_id is reported as 0
on those events; the meaningful raw_btf_id is exposed on
raw_syscalls/sys_{enter,exit}/btf_ids where raw_tp / tp_btf
programs can actually attach.
Link: https://patch.msgid.link/20260518-generic_tracepoint-v2-2-b755a5cf67bb@meta.com
Signed-off-by: Mykyta Yatsenko <yatsenko@meta.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Drop the static qualifier and add prototypes to <linux/btf.h> so the
tracing core can look up module BTF and translate ids stored by
resolve_btfids (which are local to a module's split BTF) into the
runtime ids used by the kernel.
Used by the upcoming events/<sys>/<event>/btf_ids tracefs interface.
Link: https://patch.msgid.link/20260518-generic_tracepoint-v2-1-b755a5cf67bb@meta.com
Signed-off-by: Mykyta Yatsenko <yatsenko@meta.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
- new driver mm81x for an S1G device
- new driver nxpwifi for NXP devices
(mostly forked off from mwifiex)
- ath12k: much kernel infrastructure integration work
- brcmfmac: DPP support, some Cypress part update
- nl80211: per-link statistics support
-----BEGIN PGP SIGNATURE-----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=eCJ1
-----END PGP SIGNATURE-----
Merge tag 'wireless-2026-07-26' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next
Johannes Berg says:
====================
wireless-next-2026-07-26
Mostly driver changes this time:
- new driver mm81x for an S1G device
- new driver nxpwifi for NXP devices
(mostly forked off from mwifiex)
- ath12k: much kernel infrastructure integration work
- brcmfmac: DPP support, some Cypress part update
- nl80211: per-link statistics support
====================
Link: https://patch.msgid.link/20260726105205.942922-60-johannes@sipsolutions.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
There is no particular reason to keep bpf_res_spin_lock() disabled in
tracing programs, since it is safe against reentrancy and deadlocks.
Remove the restriction for tracing programs covered by the predicate
is_tracing_prog_type().
This is a prerequisite before the definition of is_tracing_prog_type()
is updated to include raw_tp, fentry, fexit, and fmod_ret. Existing
tracing programs will be updated to use bpf_res_spin_lock() instead when
it is available.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260719113551.1294284-2-memxor@gmail.com
scx_link_sched() carries each failure out of the locked section through
err_msg and ret because scx_error() used to take scx_sched_lock and couldn't
be called under it. That restriction is gone, so report each failure at the
site it's detected and return directly. The scx_error() here claims the exit
on the sched being linked, which has no descendants yet, and the locked
propagation walk is deferred, so nothing reacquires scx_sched_lock inline.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_hardlockup() defers the abort to an irq_work because exit claiming used
to take scx_sched_lock and couldn't run from NMI. The deferral is now
unnecessary - claiming is NMI-safe and asserting ->aborting is exactly what
breaks the live-locks that hard-lock CPUs. Call handle_lockup() directly and
drop the irq_work. This also makes the self-detected case recoverable: the
perf watchdog fires on the hard-locked CPU itself, where a queued irq_work
never runs with IRQs off.
Also fix the return value: %true used to be returned whenever sched_ext was
loaded, suppressing the kernel's hardlockup report even when the abort was
refused. Return %true only when this call initiated the abort.
Fixes: bd2d76455b ("sched_ext: Defer scx_hardlockup() out of NMI")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The per-cpu kick lists are protected by IRQ masking which doesn't stop NMIs,
so scx_bpf_kick_cpu() from NMI silently drops the kick after a one-time
warning. A dropped kick can leave a CPU idle when the scheduler believes it
was woken, which is a correctness problem for the scheduler even if the
kernel is fine. Now that scx_error() works from NMI, abort the scheduler
instead so that the bug is surfaced deterministically. The warned_nmi_kick
tracking is no longer needed.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The bstr exit kfuncs format the message into a shared static buffer under a
raw spinlock before initiating the exit. The lock can't be taken from NMI
and needlessly serializes all bstr exits system-wide.
Now that exit claiming is lock-free, reverse the order: claim the exit first
and format directly into the exit_info message buffer which the claim winner
owns exclusively. The new scx_exit_bstr() implements the sequence, replacing
scx_bstr_format(), and the shared buffer and lock are deleted; the formatter
itself is what bpf_trace_printk() already runs from NMI. scx_prog_sched()
callers were relying on the lock for RCU protection, which is now provided
explicitly.
A malformed format no longer changes or fails the requested operation:
scx_bpf_exit_bstr() keeps its graceful exit kind and scx_bpf_sub_kill_bstr()
still kills the child, with a fallback message carrying the formatting
errno, while the sched that supplied the bad format is aborted for its bug.
Before this and the previous patch, an "any" category kfunc called from NMI
context could trigger scx_error() and deadlock - e.g. a tracing prog
attached to a function running in NMI calling scx_bpf_dsq_peek() on a
non-existent DSQ would try to grab scx_sched_lock, which may be held by the
interrupted CPU. This and the previous patch fix the deadlock: scx_error()
and the bstr exit kfuncs, and thus scx_bpf_error() and scx_bpf_exit(), are
now safe to call from any context including NMI.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_claim_exit() claims descendants' exits by walking the subtree under
scx_sched_lock, making exit claiming, and thus scx_error(), unusable from
NMI and from under scx_sched_lock. However, kfuncs raising errors can run
from NMI-attached BPF progs, the hardlockup handler runs in NMI, and
scx_link_sched() wants to report failures under the lock.
The walk does two things with different urgencies: ->aborting must be
asserted synchronously to break IRQs-off dispatch-path live-locks, while the
descendants' exit_kind claims can happen later. Split them: sweep ->aborting
locklessly under RCU to unwedge the system and defer the locked
SCX_EXIT_PARENT walk to a new irq_work, both of which are NMI-safe.
The sweep stores each node's ->aborting and then reads its children list
while scx_link_sched() inserts and then checks the parent's ->aborting, the
two sides paired by full barriers - one side always sees the other. A link
that sees ->aborting undoes its insert and fails. As the undo's
list_del_rcu() leaves ->sibling non-empty, list_empty() can no longer
identify a never-linked sched during teardown - add sch->linked instead.
trace_sched_ext_exit can now fire from NMI and is called after the
->aborting stores so that its callbacks don't hold up live-lock recovery.
The exit backtrace is skipped for NMI exits as stack_trace_save()'s
NMI-safety is arch-dependent and undocumented.
v2: Move trace_sched_ext_exit() after the ->aborting stores (Andrea).
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
New HW no longer implements a separate class for latency-sensitive QPs
and advertises this by a new cap bit.
Signed-off-by: Leon Romanovsky <leon@kernel.org>
For when a device is not DMA capable, the max mapping size would be 0, so
make dma_max_mapping_size() reflect that.
Signed-off-by: John Garry <john.g.garry@oracle.com>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://patch.msgid.link/20260629085310.2298552-2-john.g.garry@oracle.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Unlike local reenqueues, cap rejections have no repeat limit. A
malfunctioning scheduler can keep re-inserting a task to a cid it lacks caps
on, cycling the task through reject and reenqueue. This was assumed safe
because a task that never runs trips the stall watchdog. However, the
reenqueue irq_work re-arms itself and outranks the timer vector, blocking
everything else on the CPU including stall detection and recovery, until the
NMI hardlockup detector fires.
Local reenqueues already have a repeat cap, SCX_REENQ_LOCAL_MAX_REPEAT,
which needs generalizing to cover all reenqueues. It also has an attribution
problem. Counted per-cpu on root, it tears down the whole hierarchy even
when a sub-scheduler caused the repeated reenqueues.
Generalize by bounding every reenqueue with one per-task counter. reenq_cnt
is bumped in scx_do_enqueue_task() on each SCX_ENQ_REENQ, the single path
every reenqueue producer passes through, and cleared in clr_task_runnable()
when the task is picked to run and in scx_disable_task() when it leaves the
scheduler's control. Past SCX_REENQ_MAX_REPEAT the task's owning scheduler
is ejected with a new SCX_EXIT_ERROR_REENQ and the task is left stranded to
be picked up during sched exit.
The SCX_EV_REENQ_LOCAL_REPEAT event becomes SCX_EV_REENQ_REPEAT, counting
repeat reenqueues from all sources.
v2: Count SCX_EV_REENQ_REPEAT only when a reenqueue leads to another
reenqueue, not on every reenqueue.
v3: - Also clear reenq_cnt in scx_disable_task() so that the count doesn't
carry over to the next owner across sched class switches, scheduler
replacement or sub-scheduler rehoming (Andrea Righi).
- Update the stale SCX_EV_REENQ_LOCAL_REPEAT references in sched-ext.rst
(Andrea Righi).
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
- Move rb_desc->nr_page_va before updating dynamic array
The rb_descr->page_va is a dynamic array counted by nr_page_va. But the
updating of the page_va[] is done before the nr_page_va is incremented
causing a build with CONFIG_UBSAN_BOUNDS to flag it as an overflow.
Move the increment of the counted by value before the array element is
updated.
- Propagate errors from remote event bulk updates
The return value of trace_remote_enable_event() was not being checked by
remote_events_dir_enable_write() where it would silently fail. Have it
check the return value and propagate that back up to user space.
- Fix resource leak on mmiotrace trace_pipe close
The mmiotrace tracer was created in 2008 before the trace_pipe had a close
callback to allow tracers to do clean up from trace_pipe open. The
trace_pipe close cleanup callback was added in 2009 but the mmiotrace
tracer was not updated. It had a hack to do the cleanup in the read call,
where it may leak if user space did not read the entire buffer.
Add a callback to mmiotrace trace_pipe close do to the cleanup properly.
- Fix a possible NULL pointer dereference in the mmiotrace tracer
If the mmio_pipe_open() fails to find a PCI device, it will set the
hiter->dev pointer to NULL. The read function will blindly dereference
that pointer. Fix the read call to check to see if that pointer is
populated before dereferencing it.
- Fix union collision of module and refcnt for dynamic events
In 'struct trace_event_call', the 'module' pointer and the 'refcnt' atomic
variable share the same memory space in a union. The filter on module
logic only checked if the 'module' was set to determine if the event
belonged to the module. As dynamic events are always builtin, it doesn't
need the 'module' field of the structure and used a refcount. But the
module filtering logic would then mistaken these dynamic events as a
module and call module_name(event->module) on it.
Add a check to see if the event is a dynamic event and if so, do not check
it for being part of the given module.
- Reset the top level buffer in selftests before running instances
The ftracetest selftest initializes each instance before executing the
tests. But it does not reset the top level buffer. Dynamic events are only
added and removed by the top level so any left over dynamic events will
not be removed by the reset in the instances.
Left over dynamic events can cause the tests to incorrectly fail. Reset
the top level buffer before running the instances.
- Make the context_switch counter 64 bit
The code to read user space for a system call trace event or for a
trace_marker will disable migration, enable preemption, read user space
into a per CPU buffer, disable preemption and enable migration again.
It checks if the per CPU context switch counter to see if it changed, and
if it did not, it would know that the per CPU buffer was not touched by
another task.
But the save counter was 32 bit and it would compare it to the 64 bit
context_switch variable. A long running system could have the
context_switch variable greater that 1<<32 in which case the compare will
always fail. The compare will promote the 32 bit int saved value to 64 bit
and compare it to the full 64 bit counter. Since the top 32 bits of the
saved value was zero, it would never match.
- Fix a use-after-free of the event_enable trigger
The event_enable trigger allows for enabling one event when another event
is triggered. When the trigger is removed, it must go through a
synchronization phase to make sure it is not triggered again. The trigger
itself is delayed by the "bulk delay" logic that was recently added.
But the code that frees the event_enable data used to rely on the trigger
code to do the synchronization. Now that the code uses the call RCU
functions (and a workqueue), that delay no longer is there.
Add a callback private_data_free() function that allows triggers to clean
up data after the synchronization phase has completed.
- Move the module_ref counter into the delay callback
Since an event of the event_enable trigger can enable an event for a
module, it ups the module ref count for that event's module. This prevents
the event from trying to enable an event that no longer exists and cause a
use-after-free bug.
The ref counter was set back down when the trigger was removed but not
after thy synchronization phase. This could lead to the module data being
accessed after module was unloaded.
Move the module ref decrement into the private_data_free() callback of the
event_enable trigger.
- Add mutex to protect parser in ftrace filtering
The set_ftrace_filter file uses a parsing descriptor that is allocated at
open and modified by writes. If multiple threads were to write to the
descriptor at the same time, it can corrupt the parser.
Add a mutex around the modifications of the parser descriptor.
- Fix possible corruption in perf syscall tracing
The perf system call trace events can now read user space. To do so, the
reads of user space enable preemption and disables it again. During this
time that preemption is enabled, the task can migrate. The perf event list
head is assigned via a per CPU pointer. It is done before the user space
part is called. If the user space reading migrates the task to another
CPU, then the head pointer is no longer valid.
Re-assign the head pointer after the reading of user space to keep it
using the correct data.
-----BEGIN PGP SIGNATURE-----
iIoEABYKADIWIQRRSw7ePDh/lE+zeZMp5XQQmuv6qgUCamYDGhQccm9zdGVkdEBn
b29kbWlzLm9yZwAKCRAp5XQQmuv6qkhYAP9G5wDTVQQzitl900iWp9rvQ2Qm5UWN
JDnK2HO1elmj0AD8CRiHBI5W3O2yUmoO4bOFZ9YFXz+DqJ1jwkDs5FfqjAU=
=EMD9
-----END PGP SIGNATURE-----
Merge tag 'trace-v7.2-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace
Pull tracing fixes from Steven Rostedt:
- Move rb_desc->nr_page_va before updating dynamic array
The rb_descr->page_va is a dynamic array counted by nr_page_va. But
the updating of the page_va[] is done before the nr_page_va is
incremented causing a build with CONFIG_UBSAN_BOUNDS to flag it as an
overflow.
Move the increment of the counted by value before the array element
is updated.
- Propagate errors from remote event bulk updates
The return value of trace_remote_enable_event() was not being checked
by remote_events_dir_enable_write() where it would silently fail.
Have it check the return value and propagate that back up to user
space.
- Fix resource leak on mmiotrace trace_pipe close
The mmiotrace tracer was created in 2008 before the trace_pipe had a
close callback to allow tracers to do clean up from trace_pipe open.
The trace_pipe close cleanup callback was added in 2009 but the
mmiotrace tracer was not updated. It had a hack to do the cleanup in
the read call, where it may leak if user space did not read the
entire buffer.
Add a callback to mmiotrace trace_pipe close do to the cleanup
properly.
- Fix a possible NULL pointer dereference in the mmiotrace tracer
If the mmio_pipe_open() fails to find a PCI device, it will set the
hiter->dev pointer to NULL. The read function will blindly
dereference that pointer. Fix the read call to check to see if that
pointer is populated before dereferencing it.
- Fix union collision of module and refcnt for dynamic events
In 'struct trace_event_call', the 'module' pointer and the 'refcnt'
atomic variable share the same memory space in a union. The filter on
module logic only checked if the 'module' was set to determine if the
event belonged to the module. As dynamic events are always builtin,
it doesn't need the 'module' field of the structure and used a
refcount. But the module filtering logic would then mistaken these
dynamic events as a module and call module_name(event->module) on it.
Add a check to see if the event is a dynamic event and if so, do not
check it for being part of the given module.
- Reset the top level buffer in selftests before running instances
The ftracetest selftest initializes each instance before executing
the tests. But it does not reset the top level buffer. Dynamic events
are only added and removed by the top level so any left over dynamic
events will not be removed by the reset in the instances.
Left over dynamic events can cause the tests to incorrectly fail.
Reset the top level buffer before running the instances.
- Make the context_switch counter 64 bit
The code to read user space for a system call trace event or for a
trace_marker will disable migration, enable preemption, read user
space into a per CPU buffer, disable preemption and enable migration
again. It checks if the per CPU context switch counter to see if it
changed, and if it did not, it would know that the per CPU buffer was
not touched by another task.
But the save counter was 32 bit and it would compare it to the 64 bit
context_switch variable. A long running system could have the
context_switch variable greater that 1<<32 in which case the compare
will always fail. The compare will promote the 32 bit int saved value
to 64 bit and compare it to the full 64 bit counter. Since the top 32
bits of the saved value was zero, it would never match.
- Fix a use-after-free of the event_enable trigger
The event_enable trigger allows for enabling one event when another
event is triggered. When the trigger is removed, it must go through a
synchronization phase to make sure it is not triggered again. The
trigger itself is delayed by the "bulk delay" logic that was recently
added. But the code that frees the event_enable data used to rely on
the trigger code to do the synchronization. Now that the code uses
the call RCU functions (and a workqueue), that delay no longer is
there.
Add a callback private_data_free() function that allows triggers to
clean up data after the synchronization phase has completed.
- Move the module_ref counter into the delay callback
Since an event of the event_enable trigger can enable an event for a
module, it ups the module ref count for that event's module. This
prevents the event from trying to enable an event that no longer
exists and cause a use-after-free bug.
The ref counter was set back down when the trigger was removed but
not after thy synchronization phase. This could lead to the module
data being accessed after module was unloaded.
Move the module ref decrement into the private_data_free() callback
of the event_enable trigger.
- Add mutex to protect parser in ftrace filtering
The set_ftrace_filter file uses a parsing descriptor that is
allocated at open and modified by writes. If multiple threads were to
write to the descriptor at the same time, it can corrupt the parser.
Add a mutex around the modifications of the parser descriptor.
- Fix possible corruption in perf syscall tracing
The perf system call trace events can now read user space. To do so,
the reads of user space enable preemption and disables it again.
During this time that preemption is enabled, the task can migrate.
The perf event list head is assigned via a per CPU pointer. It is
done before the user space part is called. If the user space reading
migrates the task to another CPU, then the head pointer is no longer
valid.
Re-assign the head pointer after the reading of user space to keep it
using the correct data.
* tag 'trace-v7.2-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
tracing: perf: Fix stale head for perf syscall tracing
ftrace: Add global mutex to serialize trace_parser access
tracing: Delay module ref count for "enable_event" trigger
tracing: Fix use-after-free freeing trigger private data
tracing: Fix context switch counter truncation
selftests/ftrace: Reset triggers at top level before instance loop
tracing: Fix union collision of module and refcnt for dynamic events
tracing: Fix mmiotrace possible NULL dereferencing of hiter->dev
tracing: Fix resource leak on mmiotrace trace_pipe close
tracing: Propagate errors from remote event bulk updates
tracing/remotes: Fix page_va[] access before counter update in trace_remote_alloc_buffer()
RDMA device names are unique only within a network namespace, but an
RDMA cgroup can account resources for devices from multiple namespaces.
Duplicate names therefore make cgroup output ambiguous and can cause
limit writes to select the wrong device.
Use the system-wide RDMA device index to distinguish duplicate names
while preserving the existing UAPI for unique names. Reject ambiguous
name-only writes with -ENOTUNIQ and expose a complete device view to
administrators.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Link: https://patch.msgid.link/20260716132316.1495242-9-jiri@resnulli.us
Signed-off-by: Leon Romanovsky <leon@kernel.org>
dsq->first_task is __rcu for the lockless scx_bpf_dsq_peek(). The task
removal path compares it against the departing task with a plain load, which
sparse flags. The comparison runs under the dsq lock and only tests
identity, so rcu_access_pointer() is the fit.
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_root is __rcu and naked accesses were left as transitional markers for
the multi-scheduler transition, to be converted to accesses through the
associated scheduler instances. Most accesses have since been converted to
resolve the sched from the program or task at hand. The remaining naked
sites divide into ones that semantically always want the root sched, which
this patch resolves, and one that is left to a later patch.
The resolved sites:
- The SCX_OPS_TID_TO_TASK validation and the ecaps sync kick already hold a
sched whose ancestors[] pins the root as entry 0 with plain pointers
stable for the sched's lifetime. Reach the root through the sched at hand.
- The dispatch entry, class switch, idle notification and fork init paths
only execute while the scheduler is live and scx_root never changes inside
the live window, so no update can race them. Add scx_root_protected_live()
which documents that invariant and resolves with a plain load.
- The hotplug path, including the ecaps reseeds, runs with the hotplug lock
held, which excludes the scx_root writers. Add scx_root_protected(), which
accepts either the hotplug lock or scx_enable_mutex.
- Is-root tests use a zero level instead of comparing against the global.
touch_core_sched_dispatch() stays naked, to be resolved by a later patch.
Signed-off-by: Tejun Heo <tj@kernel.org>
cgrp->scx_sched is __rcu and published with rcu_assign_pointer() but every
reader loads it with a plain access, so sparse flags all of them. The reads
are lock-protected: enable/disable paths rewrite the field under all of
scx_enable_mutex, scx_fork_rwsem and cgroup_mutex, and cgroup creation
inherits the parent's sched under cgroup_mutex before the new cgroup is
reachable, so holding any one of the three locks makes the read stable.
Add scx_cgroup_sched() which states the protection with
rcu_dereference_check() and convert the readers. No functional changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
While booting with KHO, the following crash was observed:
BUG: unable to handle page fault for address: ff19164fffff8328
RIP: 0010:__free_one_page+0x1a1/0x6b0
Call Trace:
<TASK>
[<ffffffff913208bf>] free_one_page+0xaf/0x240
[<ffffffff93973288>] deferred_free_pages+0xa8/0xd0
[<ffffffff93971b4f>] deferred_init_memmap_chunk+0x10f/0x1b0
[<ffffffff9396e265>] padata_mt_helper+0x65/0xa0
[<ffffffff90fac402>] process_scheduled_works+0x202/0x410
[<ffffffff90fae739>] worker_thread+0x1f9/0x2d0
[<ffffffff90fb62fd>] kthread+0x27d/0x2f0
[<ffffffff90fae540>] ? __pfx_worker_thread+0x10/0x10
[<ffffffff90fb6080>] ? __pfx_kthread+0x10/0x10
[<ffffffff90efdc55>] ret_from_fork+0x145/0x280
[<ffffffff90fb6080>] ? __pfx_kthread+0x10/0x10
[<ffffffff90e2e46a>] ret_from_fork_asm+0x1a/0x30
</TASK>
deferred_init_memmap_chunk() interleaves initialization of struct pages
with freeing them. This works fine without KHO because free regions
will never be buddy neighbors. However, with KHO, free memory will be split
into (free && scratch) and (free && !scratch), that can be buddy neighbors.
KHO scratch is aligned to CMA_MIN_ALIGNMENT_PAGES pages but buddy looks
at the neighborhood of MAX_ORDER_NR_PAGES pages. These values are
configurable but CMA_MIN_ALIGNMENT_PAGES is always less or equal to
MAX_ORDER_NR_PAGES. In the crashing configuration they were set as
follows:
CMA_MIN_ALIGNMENT_PAGES = 1 << 9
MAX_ORDER_NR_PAGES = 1 << 10
So while freeing one chunk, buddy accessed uninitialized struct pages
from another chunk, tried to merge the blocks and crashed.
To fix this, let's just align KHO scratch to MAX_ORDER_NR_PAGES pages.
Fixes: c6073743d0 ("kho: make preserved pages compatible with deferred struct page init")
Signed-off-by: Michal Clapinski <mclapinski@google.com>
Link: https://patch.msgid.link/20260717134028.2880508-1-mclapinski@google.com
[rppt: massaged the changelog]
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
LUO keeps track of successful retrieve attempts on an FLB. It does so
to avoid multiple retrievals of the same FLB. Multiple retrievals cause
problems because once the FLB is retrieved, the serialized data
structures are likely freed and the FLB is likely in a very different
state from what the code expects.
All this works well when retrieve succeeds. When it fails,
luo_flb_retrieve_one() returns the error immediately, without ever
storing anywhere that a retrieve was attempted or what its error code
was. If the user attempts to retrieve another file registered with the
same FLB, LUO will attempt to call the FLB's retrieve() callback again.
The retry is problematic for much of the same reasons listed above. The
FLB is likely in a very different state than what the retrieve logic
normally expects (e.g. some KHO pages may have already been restored and
freed).
There is no sane way of attempting the retrieve again. Remember the
error retrieve returned and directly return it on a retry.
This is done by changing the retrieved bool to a retrieve_status
integer. A value of 0 means retrieve was never attempted, a positive
value means it succeeded, and a negative value means it failed and the
error code is the value.
This is similar to commit f85b1c6af5 ("liveupdate: luo_file: remember
retrieve() status") which did the same for LUO files.
Fixes: cab056f2aa ("liveupdate: luo_flb: introduce File-Lifecycle-Bound global state")
Assisted-by: Gemini:gemini-3-pro-preview
Signed-off-by: David Matlack <dmatlack@google.com>
Reviewed-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Link: https://patch.msgid.link/20260528174140.1921129-3-dmatlack@google.com
Signed-off-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Increment the outgoing FLB refcount in liveupdate_flb_get_outgoing() so
that the FLB structure cannot be freed while the caller is actively
using it. Add an additional liveupdate_flb_put_outgoing() function so
the caller can explicitly indicate when it is done using the outgoing
FLB.
During a Live Update, the kernel may need to fetch the outgoing FLB
outside of the scope of a file handler's preserve() and unpreserve()
callbacks. In that situation there is no way for the caller to protect
itself against the outgoing FLB from being freed while it is using it.
Incrementing the reference count in liveupdate_flb_get_outgoing()
ensures it cannot be freed.
This change also aligns the outgoing FLB lifecycle management with the
incoming FLB, since the latter uses the same get/put semantics.
Fixes: cab056f2aa ("liveupdate: luo_flb: introduce File-Lifecycle-Bound global state")
Assisted-by: Gemini:gemini-3-pro-preview
Signed-off-by: David Matlack <dmatlack@google.com>
Reviewed-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Link: https://patch.msgid.link/20260528174140.1921129-2-dmatlack@google.com
Signed-off-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
The UAPI documents liveupdate_session_finish::reserved as requiring zero,
but luo_session_finish() currently ignores it and finishes the session.
Accepting nonzero values prevents the field from being safely repurposed
by a future extension.
Reject nonzero reserved values before changing session state, matching
LIVEUPDATE_SESSION_GET_NAME.
Fixes: 16cec0d265 ("liveupdate: luo_session: add ioctls for file preservation")
Assisted-by: Codex:gpt-5.6-sol
Reviewed-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Signed-off-by: Jackie Liu <liuyun01@kylinos.cn>
Link: https://patch.msgid.link/20260716012607.22020-2-liu.yun@linux.dev
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
In ftrace, the trace_parser structure is allocated and initialized when
a trace file is opened, and is subsequently used across write and release
handlers to parse user input.
The affected handler paths and their specific functions are:
- Open paths: ftrace_regex_open(), ftrace_graph_open()
- Write paths: ftrace_regex_write(), ftrace_graph_write()
- Release paths: ftrace_regex_release(), ftrace_graph_release()
If userspace opens a trace file descriptor and shares it across multiple
threads, concurrent write calls will race on the parser's internal state,
specifically the 'idx', 'cont', and 'buffer' fields, leading to corrupted
input or undefined behavior.
Fix this by adding a global mutex, parser_lock, to serialize all access
to trace_parser across write and release paths, preventing concurrent
corruption of parser state.
Fixes: e704eff3ff ("ftrace: Have set_graph_function handle multiple functions in one write")
Fixes: 689fd8b65d ("tracing: trace parser support for function and graph")
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260725024721.1983675-1-wutengda@huaweicloud.com
Signed-off-by: Tengda Wu <wutengda@huaweicloud.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
- Fix tcp_bpf_sendmsg() error path mistaking a concurrently-freed
sk_psock->cork for the local temporary message and freeing it
again. (Chengfeng Ye)
- Reject passing scalar NULL to nonnull arg of a global subprog.
Previously the verifier did not account for the cases directly
passing scalars to a global subprog, e.g.: 'global_func(0);'
would pass even if 'global_func' argument was marked nonnull.
(Amery Hung)
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRINyQBqoQUC24dy5htleuBPyPTXgUCamPsBwAKCRBtleuBPyPT
Xo/uAP0YR6/9em1W/sZnss6Wkwfwd9NdmmlrJqSlt7aq+A+W7AD/V5HbigM4pJ76
CZFkKvnzhhU3dSXm/dBT1/GGJbL4pwo=
=uzKM
-----END PGP SIGNATURE-----
Merge tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Pull bpf fixes from Eduard Zingerman:
- Fix tcp_bpf_sendmsg() error path mistaking a concurrently-freed
sk_psock->cork for the local temporary message and freeing it again
(Chengfeng Ye)
- Reject passing scalar NULL to nonnull arg of a global subprog.
Previously the verifier did not account for the cases directly
passing scalars to a global subprog, e.g.: 'global_func(0);' would
pass even if 'global_func' argument was marked nonnull (Amery Hung)
* tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
bpf, sockmap: Fix cork use-after-free in tcp_bpf_sendmsg()
selftests/bpf: Test passing scalar NULL to nonnull global subprog
bpf: Reject passing scalar NULL to nonnull arg of a global subprog
Add 'P' flag to the diagnostic line to indicate whether any per-cpu
rtpcp's lazy_timer is pending. this helps diagnose stalls where
rcu-task callbacks are queued but the kthread stay sleep because
the lazy_timer has not yet fired and no grace period has started.
The output is as follows:
[ 31.319540][ T77] call_rcu_tasks() has failed boot-time tests.
[ 31.320205][ T77] rcu_tasks: RTGS_WAIT_CBS(11) since 7518 g:4 i:0 kCuUP l:150000
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit updates the comment "We can't create the thread unless
interrupts are enabled." from 'commit 4929c913bd ("rcu: Make
call_rcu_tasks() tolerate first call with irqs disabled")' to be more
clear and also to cover deferred wakeup and to take into account the fact
that kthread creation has been move to core_initcall() time by 'commit
c63eb17ff0 ("rcu: Create call_rcu_tasks() kthread at boot time")'.
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Although the rdp->rcu_cpu_has_work field is accessed only by the
corresponding CPU, it can be accessed by both interrupt handlers via
invoke_rcu_core_kthread() and at task level via rcu_cpu_kthread().
This means that we need this_cpu_read() rather than __this_cpu_read(),
this_cpu_write() rather than __this_cpu_write(), and READ_ONCE()
rather than plain C-language loads. The exception is the boot-time
rcu_spawn_core_kthreads(), which cannot race with kthreads that have
not yet been spawned.
This commit therefore makes it so.
KCSAN located this issue.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Ever since the kvfree_rcu() tracing moved out of the callback-enqueue
path, rcutree_enqueue() no longer looks at the callback function
pointer: By the time it is invoked, __call_rcu_common() has already
stored the function into rhp->func, and the enqueue path only adds
the rcu_head to the segmented callback list and emits tracepoints
that do not take the function pointer.
Nevertheless, the function pointer is still threaded through
call_rcu_core(), call_rcu_nocb(), and rcutree_enqueue(), forcing
each level to carry a dead argument.
Remove the parameter from all three functions, from the no-CBs stub,
and from the corresponding declarations. Anything needing the
callback function can still get it from rhp->func.
No functional change.
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The ->expedited_need_qs counter was part of the old expedited
grace-period machinery that spun waiting for each CPU to check in.
The current implementation instead tracks holdout CPUs via the
rcu_node structures' ->expmask fields and waits on the ->exp_wq[]
wait queues, so nothing reads or writes ->expedited_need_qs any
longer.
Remove the field to avoid wasting space in rcu_state and to keep
readers of the expedited code from searching for nonexistent users.
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Currently raw_cpu_write() is used to clear the ->rcu_need_heavy_qs
field of the per-CPU rcu_data structure. However, on x86 this is a
normal assignment, which does not play well with concurrent accesses.
This commit therefore upgrades the uses of raw_cpu_write() to its
concurrency-safe counterpart WRITE_ONCE() of an rdp pointer obtained
from this_cpu_ptr(&rcu_data).
KCSAN located this issue.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Currently __this_cpu_read() and __this_cpu_write() is used to access the
->cpu_no_qs.b.exp field of the per-CPU rcu_data structure. However,
this can fail when the accesses can happen in interrupt handlers, as
recently started being exercised by rcutorture. This commit therefore
upgrades the uses of __this_cpu_read() and __this_cpu_write() to their
interrupt-safe counterparts this_cpu_read() and this_cpu_write().
KCSAN located this issue.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Add nexp and exp_interval parameters to rcuscale that spawn kthreads
running synchronize_rcu_expedited() in a loop. This generates concurrent
expedited GP load while the normal writers measure GP or callback
latency.
When combined with gp_async=1 (which uses call_rcu() for writers), this
tests how effectively callbacks benefit from expedited grace periods.
With RCU callback expedited GP tracking, the async callbacks should
complete faster because they piggyback on the expedited GPs rather than
waiting for normal GPs.
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Even when rcu_pending() triggers rcu_core(), the normal callback
advancement path through note_gp_changes() -> __note_gp_changes() bails
out when rdp->gp_seq == rnp->gp_seq (no normal GP change). Since
expedited GPs do not update rnp->gp_seq, rcu_advance_cbs() is never
called and callbacks remain stuck in RCU_WAIT_TAIL.
Add a direct callback advancement block in rcu_core() that checks for GP
completion via rcu_segcblist_nextgp() combined with
poll_state_synchronize_rcu_full(). When detected, trylock rnp and call
rcu_advance_cbs() to move completed callbacks to RCU_DONE_TAIL. Wake the
GP kthread if rcu_advance_cbs() requests a new grace period.
Uses trylock to avoid adding contention on rnp->lock. If the lock is
contended, callbacks will be advanced on the next tick.
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
rcu_pending() decides whether rcu_core() should run on the current CPU's
timer tick. It does not account for expedited grace periods: after an
expedited GP completes, a non-offloaded CPU's callbacks remain in
RCU_WAIT_TAIL (not yet advanced to RCU_DONE_TAIL) and rcu_core() is
never invoked to advance them.
Detect that case via rcu_segcblist_nextgp() combined with a new
memory-ordering-free poll variant,
poll_state_synchronize_rcu_full_unordered(). This keeps rcu_pending()
cheap: it runs on every tick that has pending callbacks, so it must
not pay for the two memory barriers in
poll_state_synchronize_rcu_full(). The check is only a hint to run
rcu_core(); the ordered re-check and the actual callback advancement
happen there.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
When an expedited grace period completes, rcu_exp_wait_wake() wakes
waiters on rnp->exp_wq[] but does not notify the NOCB rcuog kthreads. An
rcuog kthread that is waiting for a grace period sleeps on the leaf
rcu_node's ->nocb_gp_wq[] with a wait condition based on the grace-period
state, so without a wakeup, callbacks on offloaded CPUs that could
benefit from the expedited GP wait until the rcuog kthread wakes for some
other reason (e.g. the next normal GP or a timer).
Make the rcuog grace-period wait honour expedited GPs and wake it when
one completes:
- nocb_gp_wait() now records the grace period to wait for as a struct
rcu_gp_seq (both normal and expedited), tracks the earliest pending
normal and expedited sequence across the group, and releases the wait
via poll_state_synchronize_rcu_full() so it wakes for whichever
completes first. ->nocb_gp_seq is widened to struct rcu_gp_seq
accordingly.
- rcu_exp_wait_wake() calls the new rcu_nocb_exp_cleanup() on leaf
nodes, which wakes both ->nocb_gp_wq[0] and ->nocb_gp_wq[1] (the
expedited sequence does not share parity with the normal ->gp_seq the
waiter indexed with). Both this path and rcu_nocb_gp_cleanup() use
the shared rcu_nocb_cleanup_wake() helper, which checks swait_active()
first; the smp_mb() in rcu_gp_cleanup()/rcu_exp_wait_wake() orders the
grace-period state update before that check.
A stub rcu_nocb_exp_cleanup() is provided for CONFIG_RCU_NOCB_CPU=n.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Update documentation comments throughout the RCU callback infrastructure
to reflect the transition from a single grace-period sequence number to
the full struct rcu_gp_seq that tracks both normal and expedited grace
periods.
The ->gp_seq[] array documentation in rcu_segcblist.h is updated to
describe dual (normal and expedited) GP tracking. The
rcu_segcblist_advance(), rcu_segcblist_accelerate(), and
rcu_advance_cbs() comments are updated to refer to the struct rcu_gp_seq
state (gsp) instead of the old bare grace-period sequence number (seq).
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Currently, RCU callbacks only track normal grace-period sequence
numbers. This means callbacks must wait for normal grace periods to
complete even when expedited grace periods have already elapsed.
Use the full struct rcu_gp_seq (which tracks both the normal and
expedited grace-period sequences) throughout the callback
infrastructure.
rcu_segcblist_advance() now checks both normal and expedited GP
completion via poll_state_synchronize_rcu_full(), and becomes
parameterless since it reads the grace-period state internally.
rcu_segcblist_accelerate() stores the full state (both sequences)
instead of just the normal one. rcu_accelerate_cbs() and
rcu_accelerate_cbs_unlocked() use get_state_synchronize_rcu_full() to
capture both sequences, and the NOCB advance checks use
poll_state_synchronize_rcu_full() instead of comparing only the normal
sequence.
srcu_segcblist_advance() becomes a standalone implementation because it
compares SRCU sequences directly and cannot use
poll_state_synchronize_rcu_full(), which reads RCU-specific globals.
srcu_segcblist_accelerate() sets the ->exp field to
RCU_GET_STATE_NOT_TRACKED so that poll_state_synchronize_rcu_full()
compares only ->norm and ignores ->exp.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Triggers are now delayed from freeing, but can still be triggered until
after the RCU grace period has ended. The freeing of the enable_event data
is put into the private_data_free() callback, but the put of the module
refcount is done immediately.
It is possible that if a module is removed that has an event that would
enable (or disable) it is still active, it can read the data of the module
after it is removed causing a use-after-free bug.
Move the trace_event_put_ref() that releases the module into the delayed
callback so that the module can not be removed until any reference to its
events are finished.
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260724132415.1b5005db@gandalf.local.home
Reported-by: Sashiko <sashiko-bot@kernel.org>
Link: https://sashiko.dev/#/patchset/20260724030523.19081-1-devnexen%40gmail.com
Fixes: 61d445af0a ("tracing: Add bulk garbage collection of freeing event_trigger_data")
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Commit 61d445af0a ("tracing: Add bulk garbage collection of freeing
event_trigger_data") moved the kfree() of event_trigger_data to a kthread
that runs tracepoint_synchronize_unregister() before freeing. That removed
the synchronization the trigger .free callbacks used to get implicitly and
inline from trigger_data_free().
event_hist_trigger_free(), event_hist_trigger_named_free() and
event_enable_trigger_free() free their satellite data (hist_data, cmd_ops,
enable_data) right after trigger_data_free() returns. With the
synchronization now deferred to the kthread, a concurrent tracepoint
handler can still reach that data through the list_del_rcu()'d trigger,
causing a use-after-free.
The histogram teardown must stay synchronous: remove_hist_vars() and
unregister_field_var_hists() have to detach a synthetic event from the
histogram before the trigger-removal write returns, otherwise a following
command races in and the synthetic-event removal fails with -EBUSY, as the
trigger-synthetic-eprobe.tc selftest catches. Make those callbacks wait
with the correct barrier - tracepoint_synchronize_unregister(), matching
the free kthread - before freeing.
The enable trigger has no such synchronous requirement, and a blocking
synchronize there would re-serialize the path that commit deliberately
deferred. Give it an optional private_data_free() callback that the free
kthread runs after its grace period, and free enable_data from there.
Link: https://patch.msgid.link/20260724030523.19081-1-devnexen@gmail.com
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Suggested-by: Steven Rostedt <rostedt@goodmis.org>
Fixes: 61d445af0a ("tracing: Add bulk garbage collection of freeing event_trigger_data")
Signed-off-by: David Carlier <devnexen@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
scx_bpf_cidperf_set() reaches cpufreq with no cap check, so any cid-form
sub-sched can steer the frequency of any cid in its view, including ones it
holds nothing on.
Gate it behind a new SCX_CAP_PERF rather than SCX_CAP_BASE: hardware control
is a separate axis from queue access - a parent may well delegate scheduling
on a cid without handing over its frequency. PERF neither implies nor is
implied by the other caps. The check runs under the target rq's lock, which
ecaps updates are also folded under, so it is authoritative - a write can
never land after a revoke has taken effect. Denials are counted in
SCX_EV_SUB_CIDPERF_DENIED.
The operation is synchronous and the outcome is reported to the caller:
scx_bpf_cidperf_set() now returns 0 or -errno, -EACCES on denial. The
cid-form interface is still under initial development, so the signature is
changed in place without versioning.
scx_qmap grants PERF alongside its existing cid grants so the cpuperf demo
keeps working in sub-scheds.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Factor the cpuperf target write out of scx_bpf_cpuperf_set() into
scx_cpuperf_set() which takes the acting sched and returns 0 or -errno, and
flatten the nested validation into early returns. No functional change.
Prep for gating the write behind a cap and reporting the outcome from the
cid-form kfunc.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
kick_one_cpu() silently skips a kick when the kicking sub-sched lacks
SCX_CAP_BASE on the target cid, as does kick_one_cpu_if_idle() for idle
kicks. The skips are sound with the same logic as the reenq gate but are
invisible today, unlike the preempt degradation counted in
SCX_EV_SUB_PREEMPT_DENIED. Count them in a new SCX_EV_SUB_KICK_DENIED event
so every cap denial is observable.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_bpf_dsq_reenq() with an SCX_DSQ_LOCAL_ON target schedules deferred reenq
work on the cid's cpu, raising an IPI when the target rq isn't the locked
one. Nothing checks caps along the way, so a sub-sched holding no cap at all
on a cid can force its cpu to take IPIs and rq lock cycles at will. The
analogous scx_bpf_kick_cid() path gates delivery on SCX_CAP_BASE in
kick_one_cpu() to prevent exactly this.
Apply the same rule at the reenq scheduling point: if the calling sched
lacks SCX_CAP_BASE on the target cid, drop the reenq and count it in the new
SCX_EV_SUB_REENQ_DENIED event. The check is lockless, which is fine: a reenq
slipping through right after a revoke is harmless, and a wrong denial can't
happen - if the caller has seen its ownership of the cpu, the check sees it
too.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The trampoline could be corrupted by the blindly
'tr->flags = BPF_TRAMP_F_TAIL_CALL_CTX' in verifier.
1. A fexit attached to a tail_call_reachable prog. 'tr->flags' became
'BPF_TRAMP_F_CALL_ORIG | BPF_TRAMP_F_TAIL_CALL_CTX'. And, the
trampoline would poke the target prog's nop insn using jmp insn instead
of call insn.
2. Another fexit loaded with the same tail_call_reachable prog target.
'tr->flags' became 'BPF_TRAMP_F_TAIL_CALL_CTX'.
3. Close the first fexit link. Due to no BPF_TRAMP_F_CALL_ORIG in
'tr->flags', the trampoline will fail to restore the prog's nop insn
using call insn.
[ 3.410719] WARNING: kernel/bpf/syscall.c:3551 at bpf_tracing_link_release+0x53/0x60, CPU#1: test_progs/98
...
[ 3.428793] bpf_link_free+0x58/0x130
[ 3.429293] bpf_link_release+0x23/0x30
Fix the warning by updating 'tr->flags' with '|=' and lock.
Fixes: 2b5dcb31a1 ("bpf, x64: Fix tailcall infinite loop")
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
Reviewed-by: Pu Lehui <pulehui@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20260722151909.69142-2-leon.hwang@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
BPF subprograms use the bpf_callback_t ABI, but task work invokes the
callback through a three-argument function pointer. This trips kCFI.
Store and invoke the callback as bpf_callback_t.
Fixes: 38aa7003e3 ("bpf: task work scheduling kfuncs")
Signed-off-by: Mykyta Yatsenko <yatsenko@meta.com>
Link: https://lore.kernel.org/bpf/20260724-task_work_cfi-v1-1-2616691781ed@meta.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
trace_user_fault_read() samples nr_context_switches_cpu() before enabling
preemption and retries the user copy if the counter changes. The helper
returns unsigned long long because rq->nr_switches is u64, but the saved
value is unsigned int.
Once a CPU has performed 2^32 context switches, assigning the counter to
cnt discards its upper bits. The comparison after the copy promotes cnt
back to unsigned long long, but the lost bits remain zero, so it reports a
change even when the task was never scheduled out. Every retry then fails
the same way until the 100-try guard warns and the user copy is abandoned.
This affects long-running systems and workloads with high context-switch
rates. A CPU switching 1,000 times per second takes about 50 days.
Store the sampled count in unsigned long long so the full value is
preserved.
Cc: stable@vger.kernel.org
Fixes: 64cf7d058a ("tracing: Have trace_marker use per-cpu data to read user space")
Link: https://patch.msgid.link/20260717173252.3431565-1-usama.arif@linux.dev
Reported-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Usama Arif <usama.arif@linux.dev>
Acked-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Reviewed-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
In 'struct trace_event_call', the 'module' pointer and the 'refcnt'
atomic variable share the same memory space in a union. For dynamic
events, the union member is 'refcnt', which acts as an active
reference counter.
When a dynamic event (such as kprobe, uprobe, fprobe, eprobe, or
wprobe) has a non-zero reference count (e.g. due to active event
triggers or perf attachments), its 'call->module' evaluates to a
small non-zero integer instead of NULL.
When filtering or setting events for a specific module (e.g., writing
':mod:<module>' to 'set_event'), the code in
'__ftrace_set_clr_event_nolock()' and 'update_event_fields()' reads
'call->module' directly without checking whether the event is dynamic.
This causes the kernel to treat the small integer (refcnt) as a
'struct module' pointer, leading to a NULL/invalid pointer dereference
(Oops) when dereferencing the module name.
Fix this by ensuring that the 'TRACE_EVENT_FL_DYNAMIC' flag is checked
before treating 'call->module' as a valid pointer in these code paths.
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/178425670947.84440.11344393611899824907.stgit@devnote2
Fixes: 4c86bc531e ("tracing: Add :mod: command to enabled module events")
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The flags passed to the pick-idle kfuncs are values from the
scx_pick_idle_cpu_flags enum, whose members are prefixed
SCX_PICK_IDLE_ (SCX_PICK_IDLE_CORE, SCX_PICK_IDLE_IN_NODE).
Three kernel-doc comments in idle.c erroneously used
%SCX_PICK_IDLE_CPU_* which does not correspond to any defined flag
name, while the adjacent scx_bpf_pick_idle_cpu_node() correctly
documents %SCX_PICK_IDLE_*.
Fix the three occurrences to use the correct SCX_PICK_IDLE_* prefix.
Signed-off-by: Liang Luo <luoliang@kylinos.cn>
Signed-off-by: Tejun Heo <tj@kernel.org>
Add a wakeup monitor to detect a lower-priority task waking up a
higher-priority task.
The rtapp/sleep monitor already detects this. However, that monitor
triggers an error in the context of the wakee task and user only gets
the stacktrace of that task. It is also extremely useful to get the
stacktrace of the waker task, which this monitor offers. In other
words, this monitor complements the rtapp/sleep monitor.
Signed-off-by: Nam Cao <namcao@linutronix.de>
Reviewed-by: Gabriele Monaco <gmonaco@redhat.com>
Link: https://lore.kernel.org/r/ba5658fa13e49ada466b84a2c211f233037180b5.1781852967.git.namcao@linutronix.de
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
The rtapp/sleep monitor's primary purpose is detecting common mistakes
with user-space real-time design. Monitoring real-time issues with
kernel threads is a bonus.
However, accomodating kernel threads complicates the monitor due to
the edge cases which is seen by the monitor as lower-priority task
waking higher-priority task:
- kthread_stop() wakes up the task in order to stop it.
- The rcu thread and migration thread can be woken by any task.
- The ktimerd thread is woken near the end of irq_exit_rcu(), where
the preempt counter is "broken" and falsely says this is task
context. This requires the monitor to use the hardirq_context flag
instead of the preempt counter.
Beside complicating the monitor, the final case also requires enabling
CONFIG_TRACE_IRQFLAGS (so that "hardirq_context" can be used). This
adds overhead to the kernel even when the monitor is not active. This
may be an obstacle to enabling this monitor in distros' kernels.
Furthermore, kernel threads usually are started before the monitor is
enabled. Consequently, the threads' states (i.o.w. the monitor's
atomic propositions for the threads) are not fully known to the
monitor. As a result, the kernel threads mostly cannot be monitored.
Overall, the downsides of accomodating kernel threads outweights the
benefits. Thus, exclude kernel threads to simplify the monitor.
Signed-off-by: Nam Cao <namcao@linutronix.de>
Reviewed-by: Gabriele Monaco <gmonaco@redhat.com>
Link: https://lore.kernel.org/r/eec2ca5224bcdacc45b8e1eb2f0e68109e1cae7a.1781852967.git.namcao@linutronix.de
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
CLOCK_REALTIME is the only clock that often is misused in real-time
applications. The other clocks either are safe for real-time uses
(CLOCK_TAI, CLOCK_MONOTONIC, CLOCK_BOOTTIME) or are unlikely to be misused
(CLOCK_AUX, CLOCK_PROCESS_CPUTIME_ID).
Update the monitor to only warn about CLOCK_REALTIME.
While at it, update the out-of-sync documentation.
Signed-off-by: Nam Cao <namcao@linutronix.de>
Reviewed-by: Gabriele Monaco <gmonaco@redhat.com>
Link: https://lore.kernel.org/r/c7ceb5c6263ee8f43a2676acae669cf486b0d903.1781852967.git.namcao@linutronix.de
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
The rtapp/sleep monitor detects real-time tasks which go to sleep in an
real-time-unsafe manner. If this happen, the monitor triggers a trace event
in the sched_wakeup tracepoint's handler.
However, the invoking context of that trace event is not the most
informative, because of the stack trace of that event is the wakeup's code
path which is not very helpful:
74.669317: rv:error_sleep: condvar[254]: violation detected
ltl_validate+0x345 ([kernel.kallsyms])
handle_sched_wakeup+0x34 ([kernel.kallsyms])
ttwu_do_activate+0xff ([kernel.kallsyms])
sched_ttwu_pending+0x104 ([kernel.kallsyms])
__flush_smp_call_function_queue+0x15b ([kernel.kallsyms])
__sysvec_call_function_single+0x18 ([kernel.kallsyms])
sysvec_call_function_single+0x66 ([kernel.kallsyms])
asm_sysvec_call_function_single+0x1a ([kernel.kallsyms])
pv_native_safe_halt+0xf ([kernel.kallsyms])
default_idle+0x9 ([kernel.kallsyms])
default_idle_call+0x33 ([kernel.kallsyms])
do_idle+0x234 ([kernel.kallsyms])
cpu_startup_entry+0x24 ([kernel.kallsyms])
start_secondary+0xf8 ([kernel.kallsyms])
common_startup_64+0x13e ([kernel.kallsyms])
What would be much more valuable is the stack trace of the task itself.
Instead of using the sched_wakeup tracepoint, use the sched_exit
tracepoint. This makes the event happen in the task's context, making
the stack trace far more informative for user:
rv:error_sleep: condvar[254]: violation detected
ltl_validate+0x345 ([kernel.kallsyms])
handle_sched_exit+0x39 ([kernel.kallsyms])
__schedule+0x80f ([kernel.kallsyms])
schedule+0x22 ([kernel.kallsyms])
futex_do_wait+0x33 ([kernel.kallsyms])
__futex_wait+0x8c ([kernel.kallsyms])
futex_wait+0x73 ([kernel.kallsyms])
do_futex+0xc6 ([kernel.kallsyms])
__x64_sys_futex+0x121 ([kernel.kallsyms])
do_syscall_64+0xf3 ([kernel.kallsyms])
entry_SYSCALL_64_after_hwframe+0x77 ([kernel.kallsyms])
__futex_abstimed_wait_common64+0xc6 (inlined)
__futex_abstimed_wait_common+0xc6 (/usr/lib/x86_64-linux-gnu/libc.so.6)
Signed-off-by: Nam Cao <namcao@linutronix.de>
Reviewed-by: Gabriele Monaco <gmonaco@redhat.com>
Link: https://lore.kernel.org/r/d97b4b5c476e5792b6875ec9bbf8dc214f999516.1781852967.git.namcao@linutronix.de
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
A global subprogram argument tagged __arg_nonnull is set up as a
non-nullable PTR_TO_MEM. However the verifier does not check against a
scalar NULL, leading to real NULL pointer dereference. Reject it as
well.
Fixes: 94e1c70a34 ("bpf: support 'arg:xxx' btf_decl_tag-based hints for global subprog args")
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://patch.msgid.link/20260723221815.367797-1-ameryhung@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Add missing function parameter descriptions and use the correct
function name in kernel-doc comments to avoid kernel-doc warnings:
Warning: kernel/sched/ext/ext.c:2692 function parameter 'sch' not described in 'finish_dispatch'
Warning: kernel/sched/ext/ext.c:5309 function parameter 'stalled_mask' not described in 'scx_rcu_cpu_stall'
Warning: kernel/sched/ext/ext.c:5405 function parameter 'cpu' not described in 'scx_hardlockup'
Warning: kernel/sched/ext/ext.c:8470 expecting prototype for scx_bpf_dsq_insert(). Prototype was for scx_bpf_dsq_insert___v2() instead
Warning: kernel/sched/ext/ext.c:8784 expecting prototype for scx_bpf_dsq_move_to_local(). Prototype was for scx_bpf_dsq_move_to_local___v2() instead
Warning: kernel/sched/ext/ext.c:9498 expecting prototype for scx_bpf_reenqueue_local(). Prototype was for scx_bpf_reenqueue_local___v2() instead
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
For the smp_mb() after kthread_run() in rcu_spawn_tasks_kthread_generic()
from these commits:
'commit eacd6f04a1 ("rcu-tasks: Move Tasks RCU to its own file")'
'commit 84a8f446ff ("rcu: Defer rcu_tasks_kthread() creation till
first call_rcu_tasks()")'
the memory order pairing as follows:
rcu_spawn_tasks_kthread()
->t = kthread_run(rcu_tasks_kthread, ...);
->smp_mb(); /* Ensure others see full kthread. */
->WRITE_ONCE(rcu_tasks_kthread_ptr, t);
call_rcu_tasks()
->if (READ_ONCE(rcu_tasks_kthread_ptr))
->wake_up(&rcu_tasks_cbs_wq)
->try_to_wake_up()
lock pi_lock
->smp_mb__after_spinlock()
//see full kthread
Because the 'commit d119357d07 ("rcu-tasks: Treat only synchronous
grace periods urgently")' moved the kthread_ptr assignment into the
rcu_tasks_kthread() function, the following memory order pairings are
sufficient:
The runq's raw_spinlock/unlock(or smp_mb__after_spinlock()) from
wake_up_process() in kthread_run() and __schedule() provides memory order
barrier when the kthread is first scheduled, this ensures the kthread's
func observes all of the kthread's initialization.
The kthread's smp_store_release(&rtp->kthread_ptr, ...) in
rcu_tasks_kthread() and smp_load_acquire(&rtp->kthread_ptr)
in call_rcu_tasks_generic() compose release/acquire pairing, the
cumulativity of smp_store_release() propagates visibility of the kthread's
initialization through the scheduler chain.
This commit therefore remove smp_mb() in rcu_spawn_tasks_kthread_generic().
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds a dump RCU task status function to rcu tasks tests,
used to obtain more information to help debug when the RCU tasks
boot-time tests failed.
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The RCU tasks trace has been reimplemented by 'commit c27cea4416
("rcu: Re-implement RCU Tasks Trace in terms of SRCU-fast")', the
rcu_tasks structure's->n_ipis_fails is no longer used, this commit
therefore remove it.
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Paul McKenney noted that a softirq (or irq_work) handler arming for a
deferred QS can fire and find rcu_preempt_depth() > 0 -- the task is
still inside its outer reader, so rcu_preempt_need_deferred_qs() bails
without reporting the QS. At that point the queued mechanism has been
consumed but ->defer_qs_pending stays in DEFER_QS_PENDING.
In the meantime, the only remaining path back to a quiescent state on
this CPU may be a local_irq_disable()/_enable() pair that does not
call preempt_check_resched() (it is just `sti`/`cli`). patch 6's
unconditional set_need_resched_current() makes need_resched true, but
without an irq_work being raised the next outer rcu_read_unlock_special()
hits the P-gate at the arming code:
if (rdp->defer_qs_pending != DEFER_QS_PENDING) {
rdp->defer_qs_pending = DEFER_QS_PENDING;
irq_work_queue_on(...); // <-- skipped
}
so no irq_work is queued for the hardirq-exit preempt_schedule_irq()
path either. The deferred QS now waits until the next timer tick (or
similar preempt-safe boundary), needlessly extending expedited grace
period latency.
Clear ->defer_qs_pending in the bail-out path of rcu_preempt_deferred_qs()
when rcu_preempt_depth() > 0. The recursion guard semantics introduced
by commit b41642c877 ("rcu: Fix rcu_read_unlock() deadloop due to IRQ
work").
The clear is also safe against fresh recursion at this exact program
point: rcu_preempt_depth() > 0 guarantees we are still inside an outer
reader, so any inner rcu_read_unlock() from tracing infrastructure
brings nesting back to outer (>0), never to 0. The slow path of
rcu_read_unlock_special() is structurally unreachable under that
condition, so no recursive raise_softirq_irqoff()/irq_work_queue_on()
can be triggered by the clear. Essentially, the mechanism will work to
prevent the following recursion which Xiongfeng had previously reported:
irq_exit() -> __irq_exit_rcu()
-> tick_irq_exit() -> tick_nohz_irq_exit() -> tick_nohz_stop_sched_tick()
-> trace_tick_stop() // BPF prog hooked here
-> rcu_read_unlock_special()
-> irq_work_queue_on(&rdp->defer_qs_iw, rdp->cpu) // self-IPI re-enters irq_exit
Reported-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Currently rdp->defer_qs_pending transitions from DEFER_QS_PENDING to
DEFER_QS_IDLE at two sites: rcu_preempt_deferred_qs_irqrestore() and
rcu_preempt_deferred_qs_handler() (depth>0 reset). Both write the
IDLE value directly.
Introduce a single inline helper rcu_defer_qs_clear() in tree.h and
route both sites through it. This becomes the single
PENDING->IDLE transition point for upcoming work.
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
RCU stall warnings currently print task states as raw hexadecimal
values, requiring developers to manually decode them.
Use task_state_to_char() so that stall warnings show the same symbolic
task-state representation used elsewhere in the kernel.
For example:
->state=0x402 becomes ->state=I
->state=0x0 becomes ->state=R
->state=0x2 becomes ->state=D
This improves readability while preserving the underlying diagnostic
information.
Suggested-by: Zqiang <qiang.zhang@linux.dev>
Co-developed-by: Wang Lian <lianux.mm@gmail.com>
Signed-off-by: Wang Lian <lianux.mm@gmail.com>
Signed-off-by: Kunwu Chan <kunwu.chan@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
SRCU and Tasks RCU do not track expedited grace periods. When their
callback state is checked via poll_state_synchronize_rcu_full(), the
uninitialized or zeroed exp field could cause false-positive
completion detection.
This commit adds an RCU_GET_STATE_NOT_TRACKED sentinel value (0x2) that
these subsystems can place into exp to indicate that expedited GP
tracking is not applicable. The expedited sequence check in
poll_state_synchronize_rcu_full() is guarded to skip entries marked with
this sentinel.
This is needed to allow rcu_segcblist_advance() and rcu_accelerate_cbs()
to work with both normal and expedited grace periods via
get_state_synchronize_rcu_full() and poll_state_synchronize_rcu_full().
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Change the type of the per-segment ->gp_seq[] array in struct
rcu_segcblist from unsigned long to struct rcu_gp_seq. This prepares the
callback tracking infrastructure to record both normal and expedited
grace periods per segment.
The rcu_segcblist_nextgp(), rcu_segcblist_advance(), and
rcu_segcblist_accelerate() helpers now take a struct rcu_gp_seq * instead
of an unsigned long, and all callers use the .norm field for comparisons
and assignments. The SRCU and Tasks RCU wrappers construct a struct
rcu_gp_seq with only .norm set and forward to the core helpers.
No functional change: only the .norm field is used.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit extracts the tail-pointer cleanup and segment compaction
logic from rcu_segcblist_advance() into a new static helper function,
rcu_segcblist_advance_compact(). This shared logic will be reused by the
upcoming srcu_segcblist_advance() standalone implementation, which
cannot call the core rcu_segcblist_advance() because that function will
use RCU-specific globals.
No functional change.
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Add srcu_segcblist_advance() and srcu_segcblist_accelerate() wrappers
that forward to the core rcu_segcblist_advance() and
rcu_segcblist_accelerate() functions, and switch all SRCU (srcutree.c)
and Tasks RCU (tasks.h) callers to use these wrappers.
This isolates SRCU and Tasks RCU from upcoming changes to the core
advance/accelerate functions, which will switch to struct
rcu_gp_seq for dual normal/expedited GP tracking. Because SRCU and
Tasks RCU use only normal GP sequences, their wrappers will maintain the
existing unsigned long interface.
No functional change.
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The polled grace-period state structure rcu_gp_oldstate holds a snapshot
of the normal (and, on SMP, expedited) grace-period sequence numbers.
Upcoming changes store this structure in the callback segment list, where
the "oldstate" name reads poorly: there it represents the grace period a
segment is waiting on and is also compared against the current
grace-period state.
Rename struct rcu_gp_oldstate to the more neutral struct rcu_gp_seq, and
shorten its members rgos_norm and rgos_exp to norm and exp. Local
variables and parameters of this type are renamed from rgosp/rgos to
gsp/gs accordingly.
While at it, provide a single definition of the structure in rcupdate.h
rather than separate Tiny-RCU and Tree-RCU definitions, and give it the
->exp field unconditionally. Tiny RCU does not track expedited grace
periods and leaves ->exp unused, but a single definition that always has
->exp lets the shared callback code in rcu_segcblist.c reference it
without CONFIG_SMP guards, including on !SMP builds.
No functional change.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The trace_rcu_this_gp() wrapper forwards only the rcu_node structure's
fields and the requested grace-period sequence number to the
rcu_future_grace_period tracepoint. Its rcu_data pointer parameter
has no users, but every one of the ten call sites must nevertheless
come up with an rcu_data pointer to pass in.
Remove the parameter and update all callers. This also allows
rcu_future_gp_cleanup() to drop the local rcu_data pointer that
existed solely to feed this trace call.
No functional change.
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Each scftorture_invoker() thread decrements n_started, which is
initialized to the number of threads, and is then supposed to wait
until all of its siblings have also checked in before starting the
test proper. However, the wait loop is guarded by
!atomic_dec_return(&n_started), which is true only for the final
thread to arrive, and by then n_started is already zero, so the
final thread does not wait either. The side-effect (possibly positive)
is that no thread ever waits and the start-synchronization barrier is
dead code, with early threads beginning to hammer smp_call_function*()
while later threads are still being spawned.
Invert the test so that every thread other than the last spins until
n_started reaches zero, making the threads start testing together as
intended. The existing torture_must_stop() check in the wait loop
continues to bound the wait during shutdown.
We can also drop the spinning entirely if the intent is to leave it as
dead code, however for the current intent, this patches fixes the code.
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
scf_torture_stats_print() aggregates each invoker thread's counters
into a local scf_statistics structure before printing, but the
n_single_rpc_ofl field is missing from the aggregation loop. As a
result, the "single_rpc_ofl" value printed in the statistics line is
always zero, even when smp_call_function_single() invocations for the
RPC test have failed due to offline CPUs and been counted by the
invoker threads.
Add the missing accumulation so that the printed value reflects the
actual counts.
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The mmiotrace tracer was added May 12th 2008. At that time, resources
created in pipe_open() could not be freed because there was not
pipe_close function pointer of the tracer. The pipe_close function pointer
was added in December 7th, 2009, but the mmiotrace tracer was not updated.
mmio_pipe_open() allocates a header_iter and takes a pci_dev reference
when trace_pipe is opened. mmio_close() frees them, but it was only
wired to the tracer's .close callback.
tracing_release_pipe() invokes .pipe_close, not .close, when the
trace_pipe file is released. As a result, closing trace_pipe with the
mmiotrace tracer active leaked the header_iter allocation and left a
stale pci_dev reference.
Set .pipe_close to mmio_close, matching how function_graph wires both
callbacks to the same handler.
Note, if the trace_pipe is read to completion, it will clean up the
resources, but if one were to run:
# head -n 1 /sys/kernel/tracing/trace_pipe
VERSION 20070824
Over and over again, it would trigger a massive leak.
Cc: stable@vger.kernel.org
Fixes: c521efd170 ("tracing: Add pipe_close interface)
Link: https://patch.msgid.link/20260715143604.14481-1-gaikwad.dcg@gmail.com
Signed-off-by: deepakraog <gaikwad.dcg@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
remote_events_dir_enable_write() ignores the return value from
trace_remote_enable_event(). If a remote rejects an event state change,
the write therefore reports success even though the affected event remains
in its previous state.
Keep trying all events, but retain and return the first error. This matches
__ftrace_set_clr_event_nolock(), which permits partial updates while
notifying userspace when an operation fails.
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260715074455.3897-1-liu.yun@linux.dev
Fixes: 775cb093bc ("tracing: Add events/ root files to trace remotes")
Assisted-by: Codex:gpt-5.6-sol
Signed-off-by: Jackie Liu <liuyun01@kylinos.cn>
Reviewed-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Introduce the "dpm_watchdog_enabled" module parameter to allow the DPM
watchdog to be enabled or disabled at boot time and runtime.
Additionally, introduce the CONFIG_DPM_WATCHDOG_ENABLED Kconfig option
to set the default value of the module parameter at compile time.
The primary motivation for this configurability resolves around Android
GKI (Generic Kernel Image). We want to enable CONFIG_DPM_WATCHDOG in
the GKI so the feature is available. However, because the GKI is shared
across many different devices, we don't want to inadvertently affect
devices that are unaware of this feature. This provides a way to
compile it in, but keep it disabled by default for those devices via the
kernel command line or module parameters.
To maintain backward compatibility, CONFIG_DPM_WATCHDOG_ENABLED relies
on `default y`. Previously, the DPM watchdog was always active if
CONFIG_DPM_WATCHDOG was set. Defaulting this new option to 'y' ensures
that the behavior remains unchanged for existing users and defconfigs
when they upgrade.
Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org>
Link: https://patch.msgid.link/20260720030821.2780257-3-tzungbi@kernel.org
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Delay allocating and linking the next swap_map_page until another image
page actually needs to be recorded.
The previous code linked and wrote a new swap map page as soon as the
current one became full. When the image size was an exact multiple of
MAP_PAGE_ENTRIES, that left an empty final map page that existed only to
terminate the on-disk chain.
Instead, keep a full map page in memory and only allocate the next map page
when the next image page arrives. This preserves the resume chain while
avoiding an unnecessary swap slot allocation and write for the empty
swap map page.
Signed-off-by: Haesung Kim <mattkim513@gmail.com>
Link: https://patch.msgid.link/20260714072627.3165744-1-mattkim513@gmail.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When memory_bm_create() succeeds for copy_bm but fails for zero_bm,
the function returns without freeing the resources allocated for
copy_bm. This results in a memory leak that includes radix tree nodes,
zone structures, and page lists.
Fix this by calling memory_bm_free() to release copy_bm's resources
before returning the error code when zero_bm allocation fails.
Fixes: 005e8dddd4 ("PM: hibernate: don't store zero pages in the image file")
Signed-off-by: Malaya Kumar Rout <malayarout91@gmail.com>
Acked-by: Brian Geffon <bgeffon@google.com>
Link: https://patch.msgid.link/20260711145246.8625-1-malayarout91@gmail.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The slot array already tracks allocation and task_monitor_count
duplicates that state. On an invalid second release, the old code
warns but still decrements the counter, corrupting later allocations.
Use the slot array as the sole source of truth. Return after warning
about an unused slot, and return -EBUSY when no slot is free.
Reviewed-by: Gabriele Monaco <gmonaco@redhat.com>
Signed-off-by: Li Qiang <liqiang01@kylinos.cn>
Link: https://lore.kernel.org/r/20260715015825.1413822-1-liqiang01@kylinos.cn
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
* Avoid running unit tests without amd-pstate
* Fixes for EPP on shared memory systems
* Fixes for dynamic EPP callbacks
* Avoid loading on guests
* Allow lowest nonlinear == minimum freq
-----BEGIN PGP SIGNATURE-----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=rxGv
-----END PGP SIGNATURE-----
Merge tag 'amd-pstate-v7.3-2026-07-22' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/superm1/linux
Merge amd-pstate content for 7.3 (07/22/26) from Mario Limonciello:
"* Avoid running unit tests without amd-pstate
* Fixes for EPP on shared memory systems
* Fixes for dynamic EPP callbacks
* Avoid loading on guests
* Allow lowest nonlinear == minimum freq"
* tag 'amd-pstate-v7.3-2026-07-22' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/superm1/linux: (923 commits)
cpufreq/amd-pstate: handle missing policy in dynamic EPP callbacks
cpufreq/amd-pstate: Cache the firmware programmed EPP value
cpufreq/amd-pstate: Toggle auto_sel in active mode on shared memory systems
cpufreq/amd-pstate: Fix EPP return type and handle errors during initialization
cpufreq: amd-pstate-ut: Skip tests when amd-pstate driver is not active
cpufreq/amd-pstate: Prevent the driver from loading on unsupported hardware
cpufreq/amd-pstate: Loosen requirement on lowest nonlinear frequency != min freq
Linux 7.2-rc4
Revert "drm/amd/display: Restore 5s vbl offdelay for NV3x+ DGPUs"
drm/amd/display: check GRPH_FLIP status before sending event
drm/amd/display: consolidate DCN vblank/flip handling onto vupdate_no_lock
drm/amd: Create a device link between APU display and XHCI devices
drm/amd/display: wire DCN42B mcache programming callback
drm/amd/display: set new_stream to NULL after release
drm/amd/display: Force PWM backlight on Lenovo Legion 5 15ARH05
drm/amdkfd: free MQD managers on DQM init failures
drm/amdgpu/ttm: Consider concurrent VM flushes for buffer entities
drm/amd/pm/smu7: Fix AC/DC switch notification
drm/amdgpu: Disable PCIe dynamic speed switching on Ryzen Pinnacle Ridge
drm/amdgpu: always emit the job vm fence
...
page_va[] is annotated __counted_by(nr_page_va), so nr_page_va must
cover an index before that element is accessed. The allocation loop
writes page_va[id] while nr_page_va is still id and increments it only
afterwards, so every write is one element past the declared count.
The store is out of bounds with respect to the annotation: a build with
CONFIG_UBSAN_BOUNDS on a toolchain that honours __counted_by
(clang >= 20.1, gcc >= 15.1) flags it as an array-index overflow.
Increment nr_page_va before writing the element it now covers. A failed
allocation then leaves the slot counted but NULL; the error path frees
it with free_page(0), which is a no-op.
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260713072823.2668323-1-fuad.tabba@linux.dev
Fixes: 96e43537af ("tracing: Introduce trace remotes")
Signed-off-by: Fuad Tabba <fuad.tabba@linux.dev>
Reviewed-by: Vincent Donnefort <vdonnefort@google.com>
Tested-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
These callbacks are sysfs show paths. Use sysfs_emit() and
cpumask_pr_args() to emit the masks.
This prepares for removing cpumap_print_to_pagebuf().
Link: https://lore.kernel.org/all/akANJ-AT7nHpRMq-@yury/
Acked-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Yury Norov <ynorov@nvidia.com>
Commit b0473dcd4b ("smp: Improve smp_call_function_single()
CSD-lock diagnostics") changed smp_call_function_single() so that,
when CSD lock debugging is enabled, async !wait calls use the
destination CPU csd_data. That improves diagnostics, but it also removes
the single-writer property that made the old csd_lock() safe: multiple
CPUs can now prepare the same destination CPU CSD concurrently.
csd_lock() currently waits for CSD_FLAG_LOCK to clear and then sets the
bit with a non-atomic read-modify-write. Two senders can both see an
unlocked CSD, set the bit, overwrite the callback fields, and enqueue
the same llist node. Re-adding a node that is already the queue head can
make node->next point to itself, leaving the target CPU stuck walking
call_single_queue. Later synchronous work, such as a TLB shootdown, can
then remain queued and trigger soft-lockup warnings or panics.
Keep the single csd_lock() implementation, but when CSD lock debugging is
enabled, acquire CSD_FLAG_LOCK with try_cmpxchg_acquire(). This makes the
destination CPU CSD a real atomic lock in the only configuration where it
can be shared by multiple remote senders, while preserving the existing
non-debug fast path.
Fixes: b0473dcd4b ("smp: Improve smp_call_function_single() CSD-lock diagnostics")
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Link: https://patch.msgid.link/20260716004539.13983-2-paulmck@kernel.org
Commit b0473dcd4b ("smp: Improve smp_call_function_single()
CSD-lock diagnostics") made smp_call_function_single() use the destination
CPU's csd_data when CSD lock debugging is enabled. That lets the debug code
associate a stuck CSD lock with the target CPU, but it also means the CPU
argument is used in per_cpu_ptr() before generic_exec_single() has a chance
to validate it.
This becomes unsafe when smp_call_function_any() cannot find an online CPU
in the supplied mask. In that case the selected CPU can be nr_cpu_ids, and
the !wait path calls get_single_csd_data(cpu) before generic_exec_single()
returns -ENXIO. With csdlock_debug_enabled set, that indexes the per-CPU
offset array with an invalid CPU number.
Use the destination CPU's csd_data only when the CPU number is within
nr_cpu_ids. For invalid CPU numbers, fall back to the local CPU's csd_data
and let generic_exec_single() perform the existing validation and return
-ENXIO.
Fixes: b0473dcd4b ("smp: Improve smp_call_function_single() CSD-lock diagnostics")
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Acked-by: Muchun Song <muchun.song@linux.dev>
Link: https://patch.msgid.link/20260716004539.13983-1-paulmck@kernel.org
* Fix validation of LIVEUPDATE_SESSION_GET_NAME ioctl argument caused by a
wrong resolution of a merge conflict during the last merge window
-----BEGIN PGP SIGNATURE-----
iQEzBAABCgAdFiEEeOVYVaWZL5900a/pOQOGJssO/ZEFAmpg7E4ACgkQOQOGJssO
/ZHdXQf+KPPnhNyDqHgHtnsp/sd4wQ5vGVjfmAwtX7nxpxPenGSwhZ8yKShClMNM
2s0KDQmorqKK1j3MMy+xzI+MUl7uHMnHB4cpOsESrILzfqWUgo5VmEU4beCCYXXs
1JD4eKjAs+7F59g0HE75pwmqyXzrD2nq29xWlVxQCOxsTkGe+j28B/uqw0MI8CTd
66LZzyFRT/VBjyaZYSU03KN5heJsFsckYaoPCxK4bZqaiFi65bCyd/YZ7t6JgqDl
PLx1fwjFpmCz6HArFfw3EnwVzkMoZ/qep4h+h+WMH3jiXUUKFHuJWpHYoXYN1C4r
xuQm3Rpq3HmwR/8VHtwLTJSeNXtfBg==
=KyUS
-----END PGP SIGNATURE-----
Merge tag 'liveupdate-fixes-2026-07-22' of git://git.kernel.org/pub/scm/linux/kernel/git/liveupdate/linux
Pull liveupdate fix from Mike Rapoport:
- Fix validation of LIVEUPDATE_SESSION_GET_NAME ioctl argument caused
by a wrong resolution of a merge conflict during the last merge
window
* tag 'liveupdate-fixes-2026-07-22' of git://git.kernel.org/pub/scm/linux/kernel/git/liveupdate/linux:
liveupdate: fix GET_NAME ioctl argument validation
The nesting depth check and the cgroup online check in
scx_sub_enable_workfn() reach err_disable without setting ret, so
the fallback error added by commit db4e9defd2 ("sched_ext: Record
an error on errno-only sub-enable failure") reports
"scx_sub_enable() failed (0)".
This is currently harmless because both paths record their own
scx_error() first and the first error wins, but it leaves the
fallback broken for these paths. Set -EINVAL and -ENODEV there
so the fallback always reports a real errno.
v2: The validate_ops() path from v1 is already fixed in for-7.3
(sub.c already has ret = scx_validate_ops()), so only the two
remaining paths are addressed.
Signed-off-by: Cui Jian <cjian720@163.com>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
SCX's built-in idle CPU tracking is imperfect and can be out-of-sync
with the actual idle state of CPUs, especially immediately after
enabling SCX due to scx_idle_enable() marking all online CPUs idle.
scx_select_cpu_dfl() skips marking the selected CPU as busy if
the selected CPU is the waker CPU in the SCX_WAKE_SYNC case.
If the waker CPU was marked idle by SCX, it will still be marked idle
after CPU selection and potentially even after switching to the wakee.
In the allowed_cpus selftest, this can manifest as the test failing with
the following message in dmesg:
allowed_cpus.bpf.c:21: CPU 0 should be marked as busy
This patch explicitly marks the waker CPU as busy. With this patch,
the test failure no longer reproduces. There are still some pretty
unlikely races that could make the test fail (e.g. pick_task_idle()
marking the selected CPU idle between selection and validation), but
these can't be fixed easily.
Signed-off-by: Kuba Piecuch <jpiecuch@google.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
There has been quite a bit of activity across the ath drivers.
Significant changes in ath12k include:
Align with new Qualcomm generic Peripheral Authentication Service (PAS).
Ongoing infrastructure changes to support the QCC2072 platform.
Ongoing infrastructure changes to support the IPQ5332 platform.
Enhance datapath statistics.
Tuning of datapath parameters.
In addition, an assortment of cleanups and minor bug fixes across ath6kl,
ath10k, ath11k, ath12k, and carl9170.
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ/mtSHzPUi16IfDEksFbugiYzLewUCamDp7AAKCRAsFbugiYzL
ezyJAQC6D29OEEnel7C3H4gPsLdF8oI3pszt9MHFAhKpqCh/7AEAzzuvfNVfewXU
WeP+bkfi+fLPSChX4UxWsw8xx4vVHgM=
=A8cC
-----END PGP SIGNATURE-----
Merge tag 'ath-next-20260722' of git://git.kernel.org/pub/scm/linux/kernel/git/ath/ath
Jeff Johnson says:
==================
ath.git patches for v7.3 (PR #1)
There has been quite a bit of activity across the ath drivers.
Significant changes in ath12k include:
Align with new Qualcomm generic Peripheral Authentication Service (PAS).
Ongoing infrastructure changes to support the QCC2072 platform.
Ongoing infrastructure changes to support the IPQ5332 platform.
Enhance datapath statistics.
Tuning of datapath parameters.
In addition, an assortment of cleanups and minor bug fixes across ath6kl,
ath10k, ath11k, ath12k, and carl9170.
==================
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
arch_bpf_trampoline_size() allocates a buffer to get actual size required
for a trampoline.
This buffer must be in the module address space because
__arch_prepare_bpf_trampoline() calculates rel32 offsets relatively to
that buffer.
In preparation for enabling ROX mode for EXECMEM_BPF make sure that the
allocated memory is writable.
Add bpf_jit_alloc_exec_rw() wrapper for execmem_alloc_rw() and use it for
buffer allocation in arch_bpf_trampoline_size().
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Acked-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/bpf/20260716-execmem-x86-rox-bpf-v0-v3-4-4e76158c01c5@kernel.org
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
execmem_alloc() can return ROX memory that is already filled with
architecture defined trapping instructions.
In preparation for enabling this mode for BPF on x86, make sure that there
is no redundant management of the ROX memory.
There is no need to fill allocated memory with trapping instructions, to
request permissions reset on free and to set ROX permissions as this all
is handled by execmem_alloc().
Add bpf_jit_mem_is_rox() wrapper for execmem_is_rox(), use it to check if
execmem_alloc() returns ROX memory and skip the redundant steps in that
case.
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Acked-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/bpf/20260716-execmem-x86-rox-bpf-v0-v3-3-4e76158c01c5@kernel.org
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
bpf_jit_alloc_exec() and bpf_jit_free_exec() are wrappers for the
corresponding execmem APIs.
Architectures define the properties of the memory range needed by BPF in
their initialization of execmem and don't need to override neither of
them.
Drop the __weak qualifier from bpf_jit_alloc_exec() and
bpf_jit_free_exec().
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Acked-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/bpf/20260716-execmem-x86-rox-bpf-v0-v3-2-4e76158c01c5@kernel.org
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
bpf_dispatcher->rw_image is a temporary writable buffer that
arch_prepare_bpf_dispatcher() fills and then copies into
bpf_dispatcher->image using bpf_arch_text_copy().
The rel32 offsets emitted by emit_bpf_dispatcher() are calculated against
->image, so ->rw_image does not need to live in the module address range.
Allocate ->rw_image with vzalloc() to avoid permissions dance when
EXECMEM_BPF will be backed by ROX caches.
Using vzalloc() rather than vmalloc() ensures that the memory that
bpf_dispatcher_update() unconditionally copies into the executable buffer
is zeroed, which is not ideal but still better than random memory returned
by the existing bpf_jit_alloc_exec() or plain vmalloc().
Switching from bpf_jit_alloc_exec() to vzalloc() also saves a bit of
space in the more scarce module address space.
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Acked-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/bpf/20260716-execmem-x86-rox-bpf-v0-v3-1-4e76158c01c5@kernel.org
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Use %pe format specifier instead of %ld with PTR_ERR() to print
error pointers as a symbolic error name (e.g. -ENOMEM) instead
of a raw integer value.
Signed-off-by: Ronan Marchal <ronanmarchal29@gmail.com>
[ rjw: Subject rewrite ]
Link: https://patch.msgid.link/20260615191832.75923-1-ronanmarchal29@gmail.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Use sysfs_emit() instead of sprintf() in rate_limit_us_show().
sysfs_emit() is the preferred API for sysfs output as it provides
PAGE_SIZE bounds checking and ensures proper sysfs formatting.
No functional change intended.
Signed-off-by: Zhongqiu Han <zhongqiu.han@oss.qualcomm.com>
Link: https://patch.msgid.link/20260716131546.1159644-1-zhongqiu.han@oss.qualcomm.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Commit 16a03c71bb ("cpufreq: schedutil: Merge initialization code of
sg_cpu in single loop") merged the per-CPU initialization and the
utilization-hook registration into a single loop in sugov_start().
For a shared cpufreq policy this re-introduces the race originally fixed
by commit ab2f7cf141 ("cpufreq: schedutil: Fix sugov_start() versus
sugov_update_shared() race").
The scheduler's util path reaches the hook under RCU-sched and never takes
policy->rwsem, so the rwsem held across sugov_start() cannot serialize the
two. Once the first CPU's hook is published, sugov_update_shared() may run
and, via sugov_next_freq_shared(), read/write each sibling sugov_cpu
(iowait_boost, util, bw_min, ...) concurrently with the memset() still
initializing them, with no lock common to both sides: the update side holds
sg_policy->update_lock while the init side holds only policy->rwsem, which
the scheduler's util path never takes.
The walk only accesses scalar members, never a pointer like ->sg_policy,
so it does not crash today; it merely uses stale (or zero on first start)
values that skew the frequency selection and tracepoints. It is still a
genuine data race, and a latent crash once any pointer member is
dereferenced there.
Restore the two-phase approach: initialize all per-CPU structures first,
and only then publish the per-CPU utilization update hooks.
Fixes: 16a03c71bb ("cpufreq: schedutil: Merge initialization code of sg_cpu in single loop")
Cc: stable@vger.kernel.org
Signed-off-by: Zhongqiu Han <zhongqiu.han@oss.qualcomm.com>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Link: https://patch.msgid.link/20260716115159.848403-1-zhongqiu.han@oss.qualcomm.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Several helpers in snapshot.c are introduced with kernel-doc (/**)
comment blocks but do not describe their parameters with @param
tags.This emits warnings when building
with extra warnings enabled (make W=1), for example:
kernel/power/snapshot.c:469: warning: Function parameter or member
'zone' not described in 'add_rtree_block'
kernel/power/snapshot.c:469: warning: Function parameter or member
'gfp_mask' not described in 'add_rtree_block'
kernel/power/snapshot.c:469: warning: Function parameter or member
'safe_needed' not described in 'add_rtree_block'
kernel/power/snapshot.c:469: warning: Function parameter or member
'ca' not described in 'add_rtree_block'
These are file-local implementation details, not part of the
exported kernel API documented under Documentation/. Replace the
kernel-doc markers with plain block comments for the affected
functions.
Properly documented symbols such as
alloc_rtree_node(), snapshot_read_next() and snapshot_write_next()
remain unchanged.
Signed-off-by: Adi Nata <adinata.softwareengineer@gmail.com>
[ rjw: Subject rewrite ]
Link: https://patch.msgid.link/20260609231626.38839-1-adinata.softwareengineer@gmail.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The cid tables are visible to the cid kfuncs while being modified: the
first enable publishes the global pointers before filling them,
ops.init_cids() overrides rewrite them in place, and re-enables rebuild
them in place. A racing TRACING or SYSCALL program can read unfilled
entries, including uninitialized memory in the kmalloc'd tables, or torn
topo updates.
Tie the tables' lifetimes to the root sched instead: each root enable
builds a fresh set privately and publishes the per-table __rcu globals once
the layout is final, and root disable unpublishes and RCU-frees the set. A
non-NULL global is now always a fully built table which stays valid for the
reader's RCU read section, and lookups stay two loads. Kfuncs treat NULL as
no-mapping, also after the scheduler exits instead of reporting the stale
last mapping.
The cid kfuncs are available whether the root scheduler is cid-form or
cpu-form, the latter to allow gradual migration to cids. Every root
therefore builds and publishes a default mapping.
Every reader must either be gated on scheduler liveness or NULL-check
inside an RCU read section. Fix the two kfuncs that were neither:
scx_bpf_this_cid() read the table with no RCU or preemption protection and
scx_bpf_task_cid() relied on KF_RCU, which doesn't put a sleepable program
in an RCU read section. The hotplug callbacks are instead serialized by
retiring the tables inside the cpus_read_lock() section that clears
scx_root.
v2: Document why every root builds the tables (desc + cid.c comment).
Reported-by: Andrea Righi <arighi@nvidia.com>
Closes: https://lore.kernel.org/r/al3tLtPZZkFjMveK@gpd4
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
select_task_rq_scx() falls into the default path when the scheduler has no
ops.select_cpu or is bypassing. There it calls scx_select_cpu_dfl() and
direct-dispatches to the picked CPU's local DSQ.
While bypassing, neither does anything: the enqueue path routes the task to
a bypass DSQ before consulting the direct-dispatch target, so the direct
dispatch never happens, and the CPU pick at most shifts which CPU's bypass
DSQ receives the task. Worse, when the scheduler does its own idle tracking,
the built-in idle cpumasks the pick consults are not even updated, so it
doesn't work anyway.
Return prev_cpu without the default selection while bypassing and let the
bypass enqueue place the task.
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
check_rq_for_timeouts() blames a runnable stall on the task's owner. Under
a sub-scheduler hierarchy the stalled task can be sitting on a DSQ that a
different scheduler has to drain, e.g. an ancestor's bypass DSQ while the
owner is bypassing. The drainer then escapes blame while the owner is
exited, and when the owner's exit is already claimed, nothing actionable is
reported at all.
Blame the DSQ's owning scheduler instead. The local DSQ is consumed by the
cpu itself and keeps blame on the owner. Detection keeps the owner's timeout
and single-scheduler behavior is unchanged.
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
BPF_REFCOUNT is not marked as a unique field, while it should be. Fix
this oversight.
Fixes: d54730b50b ("bpf: Introduce opaque bpf_refcount struct and add btf_record plumbing")
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://patch.msgid.link/20260719153634.2908692-4-memxor@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
btf_find_struct_field() initializes a fresh seen mask for every recursive
descent. Unique special fields in different levels of the same aggregate
therefore do not see one another. The duplicate fields can reach
btf_parse_fields(), where they trigger an invariant WARN_ON_ONCE(). A
crafted user BTF can consequently trigger the warning before map creation
checks capabilities.
Initialize the seen mask once in btf_find_field() and pass the same pointer
through struct, datasec, and nested-struct walks. This gives the entire field
traversal one shared uniqueness state.
Fixes: 64e8ee8148 ("bpf: look into the types of the fields of a struct type recursively.")
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://patch.msgid.link/20260719153634.2908692-3-memxor@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Sashiko pointed out correctly that the case statement for
BPF_RES_SPIN_LOCK incorrectly checks offset for BPF_SPIN_LOCK.
Fix it by checking res_spin_lock_off instead.
Fixes: 0de2046137 ("bpf: Implement verifier support for rqspinlock")
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://patch.msgid.link/20260719153634.2908692-2-memxor@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Queue and stack pop/peek helpers accept an uninitialized output buffer
because the verifier expects the helper to initialize it. The empty-map
error path clears the buffer, but a failed lock acquisition returns
-EBUSY without writing it.
Clear the output before returning -EBUSY so BPF programs cannot observe
uninitialized stack contents after a failed helper call.
Fixes: a34a9f1a19 ("bpf: Avoid deadlock when using queue and stack maps from NMI")
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/20260719125419.1782196-1-memxor@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Accept only one "region value" pair entry for the dmem.max, dmem.min,
dmem.low files.
This changes the UAPI that otherwise accepted multiple lines for setting
multiple entries in one write. No existing user is known to rely on
writing multiple regions in a single write.
Processing multiple regions in dmemcg_limit_write() could quietly change
first limits before failing on a later one and returning an error to the
writer, with no indication some changes occurred.
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Eric Chanudet <echanude@redhat.com>
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Reviewed-by: Thadeu Lima de Souza Cascardo <cascardo@igalia.com>
Tested-by: Thadeu Lima de Souza Cascardo <cascardo@igalia.com>
Reviewed-by: Natalie Vock <natalie.vock@gmx.de>
Link: https://patch.msgid.link/20260608-cgroup-dmem-write-single-region-v2-1-b0cd6c4ccf1b@redhat.com
Signed-off-by: Maarten Lankhorst <dev@lankhorst.se>
kprobe_perf_func() calls memset(&entry[1], 0, dsize) prior to calling
store_trace_args(). store_trace_args() populates the entry buffer and
handles dynamic data fields.
Remove this redundant memset call to align with kretprobe_perf_func() and
other probe perf functions.
Link: https://lore.kernel.org/all/178454429309.296567.17379087988750549247.stgit@devnote2/
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
fentry_perf_func() calls memset(&entry[1], 0, dsize) prior to calling
store_trace_args(). store_trace_args() populates the entry buffer and
handles dynamic data fields.
Furthermore, passing dsize (the dynamic data byte length) to memset at
&entry[1] (the start of fixed trace arguments) is inaccurate as it zeroes
from the fixed args area rather than the dynamic data region.
Remove this redundant memset call to align with fexit_perf_func() and other
probe perf functions.
Link: https://lore.kernel.org/all/178454428366.296567.16465331441301815980.stgit@devnote2/
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
trace_fprobe_match_command_head() copies trace_fprobe_symbol(tf) into a
local buffer 'buf' of size MAX_COMMON_HEAD_LEN + 1 using snprintf before
comparing with argv[0].
Since trace_fprobe_symbol(tf) already returns a null-terminated string,
comparing it directly with argv[0] via strcmp() avoids stack buffer usage
and potential symbol truncation at MAX_COMMON_HEAD_LEN.
Link: https://lore.kernel.org/all/178454427449.296567.12336315661120939938.stgit@devnote2/
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
In store_trace_entry_data(), edata is cast to unsigned long for pointer
offset arithmetic before being cast back to unsigned long *. Cast edata
to u8 * instead.
Link: https://lore.kernel.org/all/178454425622.296567.16952341966130280432.stgit@devnote2/
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
In trace_probe_compare_arg_type(), prior to entering the comparison loop,
a->nr_args and b->nr_args are checked for equality. Since the loop
condition is i < a->nr_args, i is guaranteed to be less than b->nr_args
inside the loop.
Remove the redundant (b->nr_args <= i) check.
Link: https://lore.kernel.org/all/178454423769.296567.6694636865644203423.stgit@devnote2/
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
eprobe_dyn_event_match() checks if the target event system in argv[0]
matches ep->event_system using strncmp(ep->event_system, argv[0], len).
However, if ep->event_system is longer than len (e.g. "eprobes" vs
"ep/event"), strncmp() still returns 0 because the first len characters
match.
Check that ep->event_system[len] is '\0' to ensure exact system name
matching.
Link: https://lore.kernel.org/all/178454235856.290363.14872590900774231133.stgit@devnote2/
Fixes: 7d5fda1c84 ("tracing: Fix event probe removal from dynamic events")
Cc: stable@vger.kernel.org
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
In __set_print_fmt(), LEN_OR_ZERO is defined as (len ? len - pos : 0).
If len is non-zero but smaller than pos, len - pos evaluates to a negative
integer. When passed as a size argument to snprintf(), this negative value
is cast to a large unsigned size_t, bypassing buffer size limits.
Ensure len > pos before subtracting to avoid integer underflow.
Link: https://lore.kernel.org/all/178454234934.290363.15247317871499514139.stgit@devnote2/
Fixes: 5bf652aaf4 ("tracing/probes: Integrate duplicate set_print_fmt()")
Cc: stable@vger.kernel.org
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
If trace_probe_log.argc is 0 in __trace_probe_log_err(), the loop
constructing the command string will not execute and p will remain equal to
command. Writing to *(p - 1) will cause an out-of-bounds access before
command. This should not happen, but better to be treated.
Reject if trace_probe_log.argc is 0.
Link: https://lore.kernel.org/all/178454233992.290363.18323091580600697731.stgit@devnote2/
Fixes: ab105a4fb8 ("tracing: Use tracing error_log with probe events")
Cc: stable@vger.kernel.org
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
In trace_probe_match_command_args(), a stack buffer buf[MAX_ARGSTR_LEN + 1]
(256 bytes) is used to format "<name>=<comm>". However, since name can
be up to 32 bytes (MAX_ARG_NAME_LEN) and comm up to 255 bytes
(MAX_ARGSTR_LEN), the formatted string can exceed 256 bytes and get
truncated by snprintf(), causing spurious argument matching failures.
Instead of formatting into a temporary buffer on stack, compare the
argument name, the '=' delimiter, and the comm expression directly.
Link: https://lore.kernel.org/all/178454233010.290363.10428767141343428804.stgit@devnote2/
Fixes: eb5bf81330 ("tracing/kprobe: Add per-probe delete from event")
Cc: stable@vger.kernel.org
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Reread the syscall number from pt_regs and stop returning the eventually
modified syscall number.
That moves the reread to the end of syscall_trace_enter() and prepares for
moving it to the call site.
No functional change.
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Michal Suchánek <msuchanek@suse.de>
Reviewed-by: Jinjie Ruan <ruanjinjie@huawei.com>
Link: https://patch.msgid.link/20260712141346.639115923@kernel.org
Move it out of line and let it reread the syscall number on it's own. That
makes the low level entry code denser and allows to move the reread to the
call site of syscall_trace_enter() once the tracer is fixed up.
To prevent the compiler from putting audit_context() out of line and
thereby breaking dead code elimination, mark audit_context()
__always_inline.
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Michal Suchánek <msuchanek@suse.de>
Reviewed-by: Jinjie Ruan <ruanjinjie@huawei.com>
Link: https://patch.msgid.link/20260712141346.576865340@kernel.org
If a watcher deliberately disables interrupts (either by user choice, or
because we're dealing with a scoped reordered access) to avoid detecting
any data races in interrupts, NMIs are still able to fire.
When we set up a watchpoint on a scoped reordered access, we disabled
interrupts because the same CPU cannot observe reordering of its own
accesses. To ensure we observe no false positives from NMIs, disable
access checking for interrupt contexts as well.
Fixes: 69562e4983 ("kcsan: Add core support for a subset of weak memory modeling")
Signed-off-by: Marco Elver <elver@google.com>
Running kernel-doc on timer_migration.h reports:
Warning: kernel/time/timer_migration.h:105 bad line:
The empty line separating the @available paragraph from the @idle
member documentation in the struct tmigr_cpu kernel-doc block lacks
the " *" line prefix that kernel-doc requires on every line inside a
block. The header is not scanned by the build-time kernel-doc checks,
so the warning only shows up when kernel-doc is invoked on the file
directly.
Add the missing prefix. The empty line was introduced when the
@available documentation was expanded by commit 45a13ba52c
("timers/migration: Update stale @online doc to @available").
No functional change.
Signed-off-by: Babanpreet Singh <bbnpreetsingh@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Assisted-by: Claude:claude-sonnet-5
Link: https://patch.msgid.link/20260717035947.7-1-bbnpreetsingh@gmail.com
Hybrid automata monitors's clock variables have two different
representations:
- The invariant representation, which is the timestamp when the invariant
expires
- The guard representation, which is the timestamp when the clock is last
reset
This dual representation makes the logic quite difficult to follow (well,
at least for me). It also complicates the monitors and the generation tool,
as it requires conversion back and forth between the representation.
Simplify by using the clock variables for a single purpose: storing the
time stamp since the clock is last reset.
This also allows simplifying rvgen, which will be done in a follow-up
commit.
Reviewed-by: Gabriele Monaco <gmonaco@redhat.com>
Signed-off-by: Nam Cao <namcao@linutronix.de>
Link: https://lore.kernel.org/r/c0f600dcbf3d8b487c944406851a39146f4d91fa.1781847583.git.namcao@linutronix.de
Signed-off-by: Gabriele Monaco <gmonaco@redhat.com>
EXPORT_SYMBOL_GPL(msi_domain_free_irqs_all) was mistakenly placed after
msi_domain_free_irqs_range() instead of after its intended function
msi_domain_free_irqs_all().
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Reviewed-by: Radu Rendec <radu@rendec.net>
Link: https://patch.msgid.link/20260717143157.1718-1-lirongqing@baidu.com
Two unrelated things go by "cgroup" in the cid form. Sub-schedulers attach
to cgroups, and the cgroup_*() ops deliver cpu controller events. While the
ops names suggest cgroup2 hierarchy, they actually operate on the cpu
controller.
Rename them to cpuctl_* in struct sched_ext_ops_cid, which has no users
outside scx_qmap yet. The cpu form is deployed ABI and keeps the old names.
The layout is unchanged and the kernel keeps calling through the cpu-form
union view.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Sub-schedulers don't get cgroups yet: every task_group is inited on the root
sched and the routing added by the previous patches always resolves to it.
Add the handover: an enabling sub-scheduler takes over the cgroups in its
subtree and a disabling one returns them to its parent.
scx_cgroup_claim_subtree() runs while the sub enables, after the subtree's
cgrp->scx_sched's are set and before any task is claimed. It inits each
subtree task_group on the sub, exits it from the parent and updates
tg->scx.sched. A failed ops.cgroup_init() unwinds the sub-side inits and
aborts the enable with the parent untouched.
Disabling reverses it with scx_cgroup_return_subtree(): exit each cgroup
from the sub, then re-init it on the parent with the current tg->scx.*
values, resyncing weight and bandwidth changes made while the sub had it.
When a re-init fails, the parent is failed and the remaining task_groups
still transfer uninited and get no cgroup ops - the same punting done for
tasks. The dying parent's own disable moves them onward.
The handover walks include dying but not yet offlined task_groups, the same
as root's bulk walks: a removed cgroup keeps hosting scheduling events until
its dying tasks finish their final context switches, and its
ops.cgroup_exit() must follow the last of them. tg on/offlining is excluded
through cgroup_lock(), so either ordering against an rmdir of a subtree
cgroup delivers balanced init/exit pairs.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
With sub-schedulers claiming cgroup subtrees, cgroup ops must be delivered
to each task_group's sched rather than always to root. Add tg->scx.sched to
track which sched initialized the task_group. It is set and cleared together
with SCX_TG_INITED.
Deliver the ops accordingly:
- ops.cgroup_exit() goes to the sched whose ops.cgroup_init() it pairs with.
- ops.cgroup_prep_move/move/cancel_move() go to the task's sched, and only
for moves that don't re-home the task. A re-homing move is reported
through the ops.exit_task/init_task() pair instead. The cgroups passed to
the move ops can be outside the sched's inited set as the cpu controller
can be coarser than the sub-scheduler topology.
- Knobs of a cgroup belong to the parent, so ops.set_weight/idle/bandwidth()
go to the parent task_group's sched.
All task_groups currently resolve to the root sched, so no behavior changes
until sub-schedulers start claiming cgroups.
While at it, scx_cgroup_init() is restructured so both paths share the
recording.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A task's sched (p->scx.sched) must match its cgroup's owner
(cgrp->scx_sched). cgroup migration breaks the invariant:
scx_cgroup_move_task() only fires root's ops.cgroup_move() and never
re-homes the task, leading to wrong-sched scheduling and, once the stale
sched is freed, a use-after-free.
Hook into the new cgroup task migration events and re-home each task whose
destination cgroup is owned by a different sched. The events map naturally
to the transfer: MIGRATING runs the fallible init for the destination sched,
letting it reject the migration the same way ops.cgroup_prep_move() can,
MIGRATED does the re-home, which can't fail, and CANCELED undoes the init
when the migration falls through.
Pre-commit, the task's task_group still reflects the source, so
__scx_init_task() grows an explicit cgroup argument for the migration path
to hand ops.init_task() the destination cgroup.
Signed-off-by: Tejun Heo <tj@kernel.org>
Closes: https://lore.kernel.org/r/alnxrsexEe_nQwqL@gpd4
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_cgroup_enabled is in the CONFIG_EXT_GROUP_SCHED block. The upcoming
cgroup migration re-homing needs the gate outside the block. Move the
definition and flag flips outside CONFIG_EXT_GROUP_SCHED. No functional
changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Factor out scx_rehome_task() and scx_punt_task() from the sub-disable
re-home loop and scx_fail_parent(). The upcoming cgroup migration re-homing
also needs scx_rehome_task(). No functional changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A subsystem can attach to the cgroup hierarchy itself, independent of which
controllers are enabled where - BPF hooks already behave this way and
sched_ext sub-schedulers do too. Controller callbacks can't track task
migrations for them: sched_ext must re-home a task whose migration crosses a
sub-scheduler boundary, but the cpu controller's attach callbacks fire only
when the task_group changes and miss moves whenever the controller topology
is coarser than the sub-scheduler topology.
Add cgroup_task_notifier with per-task migration events mirroring the
can_attach/attach/cancel_attach phases so that a consumer which prepares
per-task state can also veto a migration: CGROUP_TASK_MIGRATING fires
pre-commit, CGROUP_TASK_MIGRATED post-commit and
CGROUP_TASK_MIGRATE_CANCELED unwinds a failed migration. Only migrations
that change a task's dfl cgroup are reported.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The BPF_MOV64_PERCPU_REG insn requires JIT to emit native code to for
'dst_reg = src_reg + <percpu_base_off>'.
However, the interpreter ignores the 'off' at its ALU64_MOV_X label.
The 'off' indicates the insn is BPF_MOV64_PERCPU_REG insn. Then, when
the interpreter loads memory from the register, it will hit a page
fault.
[ 2.545572] BUG: unable to handle page fault for address: ffffffffacaaf034
[ 2.546485] #PF: supervisor read access in kernel mode
[ 2.547167] #PF: error_code(0x0000) - not-present page
[ 2.547850] PGD 134e63067 P4D 134e63067 PUD 134e64063 PMD 10021c063 PTE 800ffffeca550062
[ 2.548912] Oops: Oops: 0000 [#1] SMP PTI
Set jit_required as true in order to disallow interpreter fallback in
core.c::__bpf_prog_select_runtime(), if any BPF_ADDR_PERCPU insn is
patched to the prog.
BTW, rename the helper bpf_map_supports_cpu_flags() to
bpf_map_is_percpu_map().
Fixes: 7bdbf74463 ("bpf: add special internal-only MOV instruction to resolve per-CPU addrs")
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/20260715141122.15783-4-leon.hwang@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
The interpreter does not recognize the BPF_JMP|BPF_JA|BPF_X insn, which
is used for insn_array map. Thereafter, it would hit the BUG_ON() in
___bpf_prog_run() at run time.
[ 2.563726] BPF interpreter: unknown opcode 0d (imm: 0x0)
[ 2.564557] ------------[ cut here ]------------
[ 2.565206] kernel BUG at kernel/bpf/core.c:2349!
[ 2.565882] Oops: invalid opcode: 0000 [#1] SMP PTI
Set jit_required as true when insn_array map is used in the prog in
order to disallow interpreter fallback for gotox insn in
core.c::__bpf_prog_select_runtime().
Fixes: 493d9e0d60 ("bpf, x86: add support for indirect jumps")
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/20260715141122.15783-3-leon.hwang@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Since the interpreter does not support the arena-related insns,
interpreter fallback should not be allowed for these insns in
core.c::__bpf_prog_select_runtime().
Currently, when the interpreter executes the arena ST/LDX/STX insns,
it would hit the BUG_ON() in ___bpf_prog_run() at run time.
[ 2.579196] BPF interpreter: unknown opcode a2 (imm: 0x0)
[ 2.579998] ------------[ cut here ]------------
[ 2.580652] kernel BUG at kernel/bpf/core.c:2349!
[ 2.581314] Oops: invalid opcode: 0000 [#1] SMP PTI
Set jit_required as true when arena map is used in the prog to disallow
interpreter fallback for arena-related insns.
Fixes: 6082b6c328 ("bpf: Recognize addr_space_cast instruction in the verifier.")
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/bpf/20260715141122.15783-2-leon.hwang@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
bpf_arena_free_pages() accepts scalar arena addresses. The runtime
masks the address to the low 32 bits and reconstructs a full user
address from the arena base before returning the range to the arena
free tree.
When the scalar value is below the low 32 bits of the arena base,
full_uaddr falls below user_vm_start. The existing upper-end clipping
then turns this into an out-of-range free-tree offset. A later
allocation can reuse that offset and return an address below the arena
mapping.
Reject such frees before computing the clipped range.
Fixes: 317460317a ("bpf: Introduce bpf_arena.")
Signed-off-by: Yiyang Chen <chenyy23@mails.tsinghua.edu.cn>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/20260717-c10-031-public-bpf-next-v2-b4-v2-1-54b555443a7c@mails.tsinghua.edu.cn
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
arena is the only map type whose map_mem_usage() still returns 0, so
"bpftool map show" and fdinfo always showed 0 memlock for an arena no
matter how many pages it had.
Count the pages that are actually mapped into the arena: bump a counter in
apply_range_set_cb() when a page goes in and drop it in
apply_range_clear_cb() when a page goes out, both under the arena spinlock.
map_mem_usage() then just returns nr_pages << PAGE_SHIFT.
Only real data pages are counted, not the scratch page.
Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/20260717114117.350851-3-jiayuan.chen@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Replace the scratch_page field in the pte-callback data with the arena
pointer; later patches use other arena fields from these callbacks. No
functional change.
Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/20260717114117.350851-2-jiayuan.chen@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Pull to receive:
477869bfaf ("sched_ext: Reject setting disallow from init_task outside the enable path")
5f8b69642d ("sched_ext: Take cgroup_lock() first in scx_cgroup_lock()")
8c13364db9 ("sched_ext: Skip sub-disable teardown for never-linked sub-schedulers")
5cdc928598 ("sched_ext: Don't enable non-ext tasks in the sub-sched task loops")
as dependencies for the upcoming cgroup migration patchset and to
resolve the conflicts with the ext.c/sub.c split on for-7.3.
5f8b69642d comments scx_cgroup_lock() which for-7.3 exported for
sub.c. Resolved by keeping the exported version with the comment.
8c13364db9 and 5cdc928598 patch the pre-split sub-sched enable and
disable paths in ext.c which for-7.3 moved to sub.c. Resolved by
applying the never-linked teardown skip and the class gates to sub.c.
Signed-off-by: Tejun Heo <tj@kernel.org>
Root enable and scx_post_fork() enable a task only if it's on the ext class.
Tasks on other classes, possible under an SCX_OPS_SWITCH_PARTIAL root, are
left READY and enabled by switching_to_scx() when they switch over. The sub
enable-commit pass and the sub-disable re-home loop enable unconditionally,
so a fair-class READY task in the subtree becomes ENABLED while not on
sched_ext. A later switch to SCHED_EXT then trips the task state validation
WARN (ENABLED with the previous state not READY) and calls ops.enable() a
second time.
Gate scx_enable_task() on the task's class in both loops.
Fixes: 337ec00b1d ("sched_ext: Implement cgroup sub-sched enabling and disabling")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A sub-scheduler enable can fail before scx_link_sched() links the sched into
the hierarchy, e.g. when the parent is already being disabled, and cleanup
still runs the full scx_sub_disable().
That is racy against root disable: drain_descendants() is the only ordering
between a sub's disable-time task walk and root disable's all-task teardown,
and an unlinked sub is invisible to it. Root's teardown can thus run between
the never-linked sub's drain and its walk, exiting every task to no
scheduler.
The walk then trips the membership WARN and re-homes the exited tasks onto
the dying hierarchy, a use-after-free.
Skip the cgroup ownership reset and the task walk if @sch was never linked,
indicated by the empty ->sibling as unlinking only happens later in the same
function. The membership WARN remains valid: a linked sub is always waited
on by an ancestor's drain.
Fixes: 337ec00b1d ("sched_ext: Implement cgroup sub-sched enabling and disabling")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_cgroup_lock() write-locks scx_cgroup_ops_rwsem and then takes
cgroup_lock(), which can deadlock through kernfs:
scx enable/disable cgroup rmdir cpu.weight write
------------------ ------------ ----------------
cgroup_lock()
percpu_down_write(rwsem)
cgroup_lock()
kernfs_get_active()
percpu_down_read(rwsem)
kernfs_drain()
The enable path waits for the rmdir to release cgroup_mutex. The rmdir,
deactivating the cpu controller's files, waits in kernfs_drain() for the
write's active reference. The write, in scx_group_set_weight(), waits for
the rwsem behind the pending writer.
Take cgroup_lock() first. The set_* paths take no cgroup locks inside the
read side, so a pending write-lock then only waits for read sections that
always run to completion, and no dependency from the rwsem back to
cgroup_mutex remains.
Fixes: a5bd6ba30b ("sched_ext: Use cgroup_lock/unlock() to synchronize against cgroup operations")
Cc: stable@vger.kernel.org # v6.18+
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The p->scx.disallow revert assumes the root enable path, where the switching
loop reads the reverted policy right afterwards and leaves the task off SCX.
The sub-scheduler disable path also reaches it when re-initializing the
returned tasks on a root parent. Nothing reads the policy there: the task is
enabled on root anyway and keeps running on the ext class with a silently
rewritten policy.
Kill the sched instead, matching the fork and non-root branches, and update
the disallow documentation, which equated !fork with the load path and
pointed at a stale debugfs path for nr_rejected.
Fixes: 337ec00b1d ("sched_ext: Implement cgroup sub-sched enabling and disabling")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
check_kfunc_call() reads meta.func_name when bpf_fetch_kfunc_arg_meta()
returns -EACCES, but that error can come from fetch_kfunc_meta() (e.g.
fd_array_get_btf() rejecting BTF binding for a signed program) before
meta is memset(), leaving it uninitialized and risking a garbage deref
in verbose().
Move the memset() to the start of bpf_fetch_kfunc_arg_meta() so meta is
zeroed on every error return. The intended "not allowed" -EACCES path
still sets func_name first, so its message is unchanged.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://patch.msgid.link/20260715172127.2416388-3-ameryhung@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
For a bloom filter, the value argument of bpf_map_peek_elem() is always
an input. Therefore, the verifier should not allow passing uninitialized
stack memory to it to avoid information leak.
bpf_map_peek_elem() tags its value argument ARG_PTR_TO_MAP_VALUE |
MEM_UNINIT, telling the verifier the callee fills the buffer. This holds
for queue/stack maps, but not for a bloom filter, which reads the buffer
as an input to test set membership and never writes it.
As a result, a program can pass an uninitialized stack buffer to
bpf_map_peek_elem() on a bloom filter. The verifier accepts it and marks
the buffer initialized on return, letting the program read back leftover
kernel stack memory. Bloom maps require CAP_BPF to create, so this is a
CAP_BPF-gated stack infoleak that bypasses the boundary CAP_BPF is meant
to enforce (arbitrary kernel reads are gated behind CAP_PERFMON).
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://patch.msgid.link/20260715172127.2416388-2-ameryhung@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
- Fix a UAF in socket clone early bailout paths (Matt Bobrowski)
- Reject unhashed UDP sockets on sockmap update to prevent refcount leaks
(Michal Luczaj)
- Account for receive queue data in FIONREAD on sockmap sockets without a
verdict program (Mattia Meleleo)
- Reject negative constant offsets for verifier buffer pointers (Sun Jian)
- Fix for tracing of kfuncs with implicit arguments (Ihor Solodrai)
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQR0/74CM8h5+1E9pgtHLTd7Y1QvgwUCalp3/AAKCRBHLTd7Y1Qv
g9J/AQCJzidAmStp1i074sZC7sC1uq+aYIvwycpMTp6hVgyP+AEAw7oKcnGkPGXs
jeEDctaLtxBnPMXGAR5UaeqdNqu9ywQ=
=zQgn
-----END PGP SIGNATURE-----
Merge tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Pull bpf fixes from Kumar Kartikeya Dwivedi:
- Fix a UAF in socket clone early bailout paths (Matt Bobrowski)
- Reject unhashed UDP sockets on sockmap update to prevent refcount
leaks (Michal Luczaj)
- Account for receive queue data in FIONREAD on sockmap sockets without
a verdict program (Mattia Meleleo)
- Reject negative constant offsets for verifier buffer pointers (Sun
Jian)
- Fix for tracing of kfuncs with implicit arguments (Ihor Solodrai)
* tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
selftests/bpf: Cover tracing implicit kfunc args
bpf: Fix tracing of kfuncs with implicit args
selftests/bpf: Cover negative buffer pointer offsets
bpf: Reject negative const offsets for buffer pointers
selftests/bpf: Test FIONREAD on a sockmap socket without a verdict program
bpf, sockmap: Account for receive queue in FIONREAD without a verdict program
selftests/bpf: Fail unbound UDP on sockmap update
selftests/bpf: Adapt sockmap update error handling
bpf, sockmap: Reject unhashed UDP sockets on sockmap update
selftests/bpf: Ensure UDP sockets are bound
bpf: Fix UAF in sock clone early bailouts
The function try_enable_preferred_console() currently has the
non-obvious side effect of returning success for consoles that are
already pre-enabled. This obscures the logic flow during console
registration.
Move the check for pre-enabled consoles directly into the top-level
register_console(). This change makes the handling of pre-enabled
consoles explicit and easier to follow.
Furthermore, this separation lays the groundwork for future cleanups
where try_enable_preferred_console() can be restricted to cases where
an entry actually exists in the console_cmdline[] array.
Also it fixes a possible out-of-bound access when the console_cmdline[]
array is full and no console matched. In fact, the check of
c->user_specified did not make much sense. The new console either
matched and was handled in the for-cycle. Or it did not match
and then *c pointed to an unused entry.
Possible behavior change:
try_enable_preferred_console() will newly be called also with
@user_specified parameter set to "false" when it failed with the "true"
variant. But it looks like the right way to do. It will allow to call
newcon->setup() when the console was preferred by some platform
specific code.
Reported-by: Naveen Kumar Chaudhary <naveen.osdev@gmail.com> # out-of-bound access
Closes: https://lore.kernel.org/r/7sq4tr2nmlz32tvkf6vpsghv6exvqfghsrlvywjcqihzsqqbf7@bspclmti5xg4
Reviewed-by: John Ogness <john.ogness@linutronix.de>
Link: https://patch.msgid.link/20260604101459.393162-2-pmladek@suse.com
Signed-off-by: Petr Mladek <pmladek@suse.com>
The SWIOTLB bounce buffer pool size is hardcoded at 64 MB via
IO_TLB_DEFAULT_SIZE with no compile-time knob to adjust it. On
memory-constrained embedded or mobile platforms equipped with a
hardware IOMMU (e.g., ARM SMMU) covering most DMA-capable devices,
reserving 64 MB at boot is unnecessarily wasteful — the SWIOTLB is
only exercised for devices that bypass the IOMMU or have restricted
DMA address ranges.
Introduce CONFIG_SWIOTLB_DEFAULT_SIZE_MB, an integer Kconfig option
(range 1–64 MB, default 64) that allows platforms to set a smaller
compile-time default. IO_TLB_DEFAULT_SIZE is updated to derive from
this value when CONFIG_SWIOTLB is enabled, preserving the existing
64 MB default when the option is not configured.
The runtime "swiotlb=<nslabs>" kernel parameter override remains
fully supported and takes precedence over the compile-time default.
Signed-off-by: Jagadeesh Pagadala <jpagadal@qti.qualcomm.com>
Signed-off-by: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>
Reviewed-by: Michael Kelley <mhklinux@outlook.com>
Link: https://lore.kernel.org/r/20260702-swiotlb-v2-1-9205f3ba5408@oss.qualcomm.com
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Some JIT compilers, such as x86_64, rely on a register to pass the TCC.
When subprograms of synchronous callback invoke tailcall, C helpers
invoking bpf callback clobber this register, and the corrupted TCC may
bypass the TCC limit, leading to infinite tailcall.
Fix this by rejecting tailcall inside all subprogs of sync callback.
This also cleanly consolidates the existing async and exception callback
checks into a single unified `is_cb` check.
Reported-by: Sashiko <sashiko-bot@kernel.org>
Reported-by: Björn Töpel <bjorn@kernel.org>
Signed-off-by: Pu Lehui <pulehui@huawei.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://patch.msgid.link/20260716120157.835937-3-pulehui@huaweicloud.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Currently in check_max_stack_depth_subprog, when the verifier enters a
new callee branch, the local tail_call_reachable is not properly
synchronized with the callee's state.
Consider a main prog branching into multiple subprogs:
subprog0 -> tailcall
main <
subprog1 -> subprog2
When the verifier finishes checking subprog0 and backtracks to main
prog, the local tail_call_reachable state is left as true. As it
proceeds to subprog1, this uncleared state leaks into the new branch,
falsely marking subprog1 and subprog2 as tailcall reachable.
Fix this by explicitly syncing tail_call_reachable with the callee's
has_tail_call state on entry. The caller's state is safely preserved and
restored via the existing backtracking logic.
Fixes: ebf7d1f508 ("bpf, x64: rework pro/epilogue and tailcall handling in JIT")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Pu Lehui <pulehui@huawei.com>
Link: https://patch.msgid.link/20260716120157.835937-2-pulehui@huaweicloud.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
A kfunc marked with KF_IMPLICIT_ARGS flag takes implicit arguments
(such as bpf_prog_aux) that the verifier injects at load time.
resolve_btfids strips those from the kfunc's BTF-visible prototype and
keeps the real kernel ABI in a counterpart _impl prototype [1].
fentry/fexit/fmod_ret/fsession programs may attach to the BPF kernel
functions, including those with implicit args. However
bpf_check_attach_target() and bpf_check_attach_btf_id_multi() extract
the struct btf_func_model from the wrong BTF prototype of the
kfunc. The btf_func_model is later read to construct the trampoline,
which then causes the injected implicit argument to be clobbered and
the kfunc dereferencing garbage.
Add btf_attach_func_proto() to resolve the real ABI prototype of the
kfunc the way the call site does: by looking up the _impl prototype
for a KF_IMPLICIT_ARGS kfunc. Use it at both attach-target model
construction sites.
To enable this, make two supporting changes:
* pass bpf_verifier_log instead of bpf_verifier_env to
find_kfunc_impl_proto(), so it can be reused from the attach path
* add btf_kfunc_check_flag() to test a flag across all of a kfunc's
hook sets, because a program attaching to a kfunc is not in the
kfunc's call-set
KF_IMPLICIT_ARGS must be consistent across the sets, so
btf_kfunc_check_flag() returns -EINVAL on inconsistency.
btf_kfunc_check_flag() reads the kfunc's flags from the target's
kfunc_set_tab. For a module BTF that table is stable only after the
module is live, so take a module reference around the read, mirroring
how the kfunc call path gates the same lookup with btf_try_get_module().
The remaining call sites of btf_distill_func_proto() are safe as
is. The BPF_TRACE_ITER case distills a registered iterator's
prototype, and bpf_struct_ops_desc_init() distills the
function-pointer members of a struct_ops type. Neither is a kfunc, and
so can't have implicit arguments.
[1] https://lore.kernel.org/all/20260120222638.3976562-1-ihor.solodrai@linux.dev/
Fixes: 64e1360524 ("bpf: Verifier support for KF_IMPLICIT_ARGS")
Reported-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Ihor Solodrai <ihor.solodrai@linux.dev>
Link: https://github.com/sched-ext/scx/issues/3687#issuecomment-4906694106
Link: https://patch.msgid.link/20260713235223.1639022-2-ihor.solodrai@linux.dev
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
With cgroup v2 migration of a multithreaded process having threads
in different cgroups of a threaded subtree, it is possible that
cpuset_can_attach() can be called with tasks that are not migrating with
respect to cpuset if cpuset controller is not enabled in some of the
subtree cgroups. IOW, the old cpuset can be the same as the new one. This
can cause problem when we need to track the set of old cpusets and the
new cpusets in singly linked lists as a cpuset cannot be in both lists.
As reported by Tejun, the following is an example threaded subtree with
partial cpuset delegation that can cause this issue to show up.
P (+cpuset)
|- R (cpuset) <- destination
| `- C (no cpuset) -> effective cpuset == R
`- W (cpuset)
Group leader in R, thread_a in C, thread_b in W; migrate the whole
process into R (echo $PID > R/cgroup.procs). thread_a moves C->R:
its cgroup changes so compare_css_sets() keeps it in the taskset, but
its cpuset css is unchanged (C inherits R's), so task_cs() == cs ==
R. cpuset is in ss_mask because thread_b (W->R) changed. can_attach()
then tags R as a source (thread_a) and the destination (thread_b):
Handle this special case by skipping tasks that are not migrating in
cpuset_can_attach() and avoid calling cpuset_can_attach_check() in this
case. By doing so, the destination cpuset will not be put into source
cpuset linked list.
As the source cpuset cannot be easily determined in cpuset_attach(),
unnecessary work can be performed if a task is not actually
migrating. However, no harm will be done except wasting some
CPU cycles. If it happens that none of the tasks is migrating,
attach_ctx.old_cs will be NULL and task iteration won't be needed.
Reported-by: Tejun Heo <tj@kernel.org>
Closes: https://lore.kernel.org/lkml/e254af713b5345aec3d086771ecf1e71@kernel.org
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The only case where the cgroup_taskset structure requires task migration
to multiple cpusets is when enabling a cpuset controller in cgroup v2
where the newly created child cpusets inherits the same effective CPUs
and memory nodes from the parent. In that case, task migration can happen
directly with no update to tasks' CPU and memory nodes assignment and no
further work needed from the cpuset side except updating nr_deadline_tasks
when DL tasks are involved and setting old_mems_allowed in the child
cpusets.
Do that by tracking all the destination cpusets with a new dst_cs_head
singly linked list. The reset_migrate_dl_data() function is integrated
into clear_attach_data() so that it can be used for both source and
destination cpusets.
A warning will be printed if there are multiple destination cpusets but
it is not on default hierarchy or when the CPUs or memory nodes change.
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The smp_call*() functions handle their required preemption and CPU
pinning internally. The explicit preempt_disable() in
scftorture_invoke_one() is therefore no longer required for correctness.
Keeping the outer preempt_disable() would also prevent scftorture from
exercising the narrowed internal preemption-disabled regions during IPI
dispatch.
Removing the outer preemption protection can expose a CPU hotplug race in
the test validation when use_cpus_read_lock is false. For multicast
operations, SCF_PRIM_MANY or SCF_PRIM_ALL, if only one CPU is online,
smp_call_function_many() correctly skips sending IPIs and leaves scfc_out
false. Without preemption disabled, a CPU hotplug thread can preempt the
test thread, bring a second CPU online and increment num_online_cpus().
When the test thread resumes, the validation check can observe
num_online_cpus() > 1 and falsely trigger the memory-ordering warning,
leaking the scfcp structure.
Remove the preempt_disable() and preempt_enable() pairs around the
smp_call*() invocations in scftorture_invoke_one(). Restrict the
num_online_cpus() > 1 validation to the use_cpus_read_lock=true case,
where the CPU count is stable during the evaluation.
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://patch.msgid.link/20260709122933.4021501-10-zhouchuyi@bytedance.com
smp_call_function_many_cond() handles the preemption and CPU pinning
requirements internally. on_each_cpu_cond_mask() only builds the call
flags and forwards the request to that helper.
Remove the outer preempt_disable() and preempt_enable() pair from
on_each_cpu_cond_mask().
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Muchun Song <muchun.song@linux.dev>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://patch.msgid.link/20260709122933.4021501-9-zhouchuyi@bytedance.com
smp_call_function_many_cond() handles the preemption and CPU pinning
requirements internally. smp_call_function() only forwards the request to
that helper for cpu_online_mask and does not access CPU-local state on
its own.
Remove the outer preempt_disable() and preempt_enable() pair from
smp_call_function().
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Muchun Song <muchun.song@linux.dev>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://patch.msgid.link/20260709122933.4021501-8-zhouchuyi@bytedance.com
smp_call_function_many_cond() still has to keep the caller pinned to the
current CPU while the remote IPI request is built and dispatched. This
protects the queueing state and CPU-hotplug boundary that are required
before the synchronous wait starts:
- It protects the current CPU's per-CPU scratch cpumask,
cfd->cpumask_ipi. Another task running on the same CPU could otherwise
enter smp_call_function_many_cond() and reuse that scratch cpumask
before the current caller has finished building and sending the IPI
request.
- It provides the CPU-hotplug exclusion required by the CSD queueing
side. New CSDs must not be queued after smpcfd_dying_cpu() has flushed
the outgoing CPU's callback queue. Keeping preemption disabled until
all required CSDs have been queued and the corresponding IPIs have
been sent prevents CPU offline from crossing that boundary in the
middle of the queueing operation.
The CSD acquisition side also relies on that caller-side CPU pinning.
csd_lock() waits for CSD_FLAG_LOCK to clear and then marks the CSD busy
with a regular store, so another task on the same CPU must not be
allowed to acquire and reinitialize the same per-CPU CSD concurrently.
After the callbacks have been queued and the IPIs have been sent, the
caller only performs the final csd_lock_wait() completion wait. If it is
preempted there, another task running on the original CPU may enter
smp_call_function_many_cond(), but any attempt to reuse the same per-CPU
CSD will block in csd_lock() until the previous callback clears
CSD_FLAG_LOCK. The final csd_lock_wait() does not acquire or reinitialize
the CSD, so it does not need the same caller-side preemption-disabled
protection.
The wait mask is task-local, so it cannot be overwritten by another task
on the original CPU. The per-CPU CSD storage also remains allocated
across CPU offline, so csd_lock_wait() can safely dereference it even if
the target CPU is offlined after the caller is unpinned.
With those requirements satisfied, enable preemption before the
synchronous csd_lock_wait() loop. This makes the potentially long wait
preemptible and migratable while keeping the CPU-pinned section around
the remote CPU selection and IPI dispatch.
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://patch.msgid.link/20260709122933.4021501-7-zhouchuyi@bytedance.com
smp_call_function_many_cond() uses per-CPU CSD objects when queueing
callbacks to remote CPUs, and the wait path later dereferences those CSDs
from csd_lock_wait().
Making the wait path preemptible allows the initiating task to be
preempted or migrated before it waits for completion. A target CPU can be
offlined in that window. If smpcfd_dead_cpu() frees the target CPU's
per-CPU CSD storage, csd_lock_wait() can later dereference freed memory.
One way to protect the CSD storage is to free it via RCU or after a
synchronization step in the CPU offline path, but that would add
unnecessary complexity and can delay CPU shutdown.
Allocate the per-CPU CSD storage the first time a CPU comes up and keep
it allocated when the CPU is offlined. This allows csd_lock_wait() to
access the CSD even when the target CPU is offlined after preemption is
re-enabled and before the wait is invoked.
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: Muchun Song <muchun.song@linux.dev>
Link: https://patch.msgid.link/20260709122933.4021501-6-zhouchuyi@bytedance.com
smp_call_function_many_cond() uses the per-CPU cfd->cpumask as the list
of remote CPUs to wait for. That is safe while the caller remains pinned
to the current CPU for the whole operation, because another task cannot
run on the same CPU and reuse the per-CPU mask.
The synchronous wait is the long-latency part of the operation. To make
that wait preemptible, the mask iterated by csd_lock_wait() must remain
stable even if the task is preempted or migrates. If the wait used the
per-CPU cfd->cpumask after dropping CPU pinning, another task scheduled
on the original CPU could enter smp_call_function_many_cond() and
overwrite the mask while the first task is still iterating it.
Give each task private IPI cpumask storage and use it as the wait mask in
smp_call_function_many_cond(). Other cpumask storage choices do not fit
this use case:
- Per-CPU storage is the state that becomes unsafe once the wait is
made preemptible. After the caller drops CPU pinning, another task
scheduled on the original CPU can enter smp_call_function_many_cond()
and reuse the same per-CPU mask.
- Stack storage is not suitable for large NR_CPUS or
CONFIG_CPUMASK_OFFSTACK=y configurations. The wait mask needs to
scale with cpumask_size(), and putting that storage on the stack is
not acceptable on large systems.
- Allocating the mask inside smp_call_function_many_cond() would put an
allocation and a failure path in the generic IPI path. A sleeping
allocation is not suitable because callers have historically only
provided a preempt-disabled context, not a sleepable one. GFP_ATOMIC
would avoid sleeping, but a failure fallback would make the latency
improvement opportunistic instead of guaranteed.
The users are not limited to a small, pre-identifiable class of tasks. On
x86, ordinary tasks can reach this path through TLB flushes during exit,
unmap and reclaim, so allocating the mask only for a known subset of
tasks is not straightforward.
The memory cost is explicit: one word is added to task_struct. When
cpumask_size() fits in that word, the mask is stored inline and no
separate allocation is needed. Larger systems allocate cpumask_size() per
task; on x86-64 NR_CPUS=8192 this is 1 KiB per task. For context,
x86 already carries several KiB of per-task architecture and FPU state,
depending on the enabled features and configuration. That does not make
the extra cpumask free, but it puts the large-NR_CPUS case in
perspective.
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://patch.msgid.link/20260709122933.4021501-5-zhouchuyi@bytedance.com
smp_call_function_any() disables preemption across the entire operation:
selecting a target CPU, enqueueing the IPI, and synchronously waiting for
the remote CPU. smp_call_function_single() already re-enables preemption
before the synchronous csd_lock_wait(), so callers of
smp_call_function_any() should benefit from the same shorter
preemption-disabled section.
Simply removing get_cpu() and put_cpu() from smp_call_function_any()
would leave the preemption disablement entirely to
smp_call_function_single(). That opens a preemption window between
selecting the remote CPU, for example via sched_numa_find_nth_cpu(), and
dispatching the IPI in smp_call_function_single(). If the selected CPU is
fully offlined in that window, smp_call_function_single() fails its
cpu_online() check and returns -ENXIO to the caller, violating the
guarantee that smp_call_function_any() executes on any online CPU in the
mask.
Move the remote CPU selection into a common
__smp_call_function_single() helper. Keep the target CPU selection and
IPI dispatch within the same preemption-disabled region, while still
allowing the wait path to use the shorter preemption-disabled section
provided by smp_call_function_single().
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://patch.msgid.link/20260709122933.4021501-4-zhouchuyi@bytedance.com
smp_call_function_single() disables preemption while it validates the
target CPU, prepares the call single data, queues the callback and sends
the IPI.
For the !wait case, preemption protects the per-CPU csd_data from
concurrent modification by another task on the same CPU. For the wait
case, the CSD is stack allocated and no other task can reuse it. CPU
pinning is still required until the callback has been queued and the IPI
has been sent, to ensure that the target CPU cannot be offlined after the
online check but before dispatch.
After generic_exec_single() has queued the callback, the synchronous
csd_lock_wait() invocation at the end of the execution does not require
the caller to remain pinned to the current CPU.
Enable preemption before csd_lock_wait() to shorten the
preemption-disabled section.
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Muchun Song <muchun.song@linux.dev>
Reviewed-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://patch.msgid.link/20260709122933.4021501-3-zhouchuyi@bytedance.com
The CSD lock wait debugging code in __csd_lock_wait() must run with
preemption disabled. The smp function call mechanisms which invoke it
currently keep preemption disabled across the wait, so the debugging code
inherits that guarantee from its callers.
Keeping preemption disabled across the whole smp function call operation
can induce large scheduling latencies. Shortening the caller-side
preemption-disabled region would invoke __csd_lock_wait() with preemption
enabled.
Prepare for that by disabling preemption explicitly around the CSD lock
wait debugging code in __csd_lock_wait().
Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: Muchun Song <muchun.song@linux.dev>
Link: https://patch.msgid.link/20260709122933.4021501-2-zhouchuyi@bytedance.com
LIVEUPDATE_SESSION_GET_NAME was developed in the liveupdate/next branch
while the session type validation change was carried in liveupdate-fixes.
When the conflict between the two branches was resolved, the GET_NAME
operation descriptor picked up the structure and last member from
RETRIEVE_FD.
This makes both its known size and minimum size 16 bytes rather than 72.
A zero-initialized request still succeeds because luo_session_get_name()
writes the full name before luo_ucmd_respond() copies the full GET_NAME
response to userspace. However, copy_struct_from_user() treats the
output-only name field as unknown trailing data and rejects the request
with -E2BIG if any byte in that field is nonzero.
Use the GET_NAME structure and its name field in the descriptor.
Link: https://lore.kernel.org/all/ahWlYXNjGUbkKoHy@sirena.org.uk/
Assisted-by: Codex:gpt-5.6-sol
Reviewed-by: Pratyush Yadav (Google) <pratyush@kernel.org>
Signed-off-by: Jackie Liu <liuyun01@kylinos.cn>
Link: https://patch.msgid.link/20260716012607.22020-1-liu.yun@linux.dev
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
With CONFIG_EXT_SUB_SCHED=y but no sub-scheduler attached - the common case
- hot paths still pay for sub-sched bookkeeping. Gate it behind
__scx_has_subs, a static key counting live sub-schedulers, so that a
root-only system stops paying.
Most conversions are simple skip-if-no-sub tests. scx_idle_notify() is
special - it's a hierarchy walk, so give it a fast path which notifies the
root directly using the same tests as the walk. A pending
SCX_RQ_SUB_IDLE_RENOTIFY can be ignored as no sub can be owed one and the
caller clears the flag either way.
Suggested-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
rq->scx.sub_dispatch_prev is sub-sched-only but was left unconditional. Move
it into the CONFIG_EXT_SUB_SCHED block next to ecaps_to_sync and gate its
updates.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_dispatch_sched() is common dispatch machinery and looks out of place in
sub.h, but it needs scx_cpu_arg() from cid.h and can't move into internal.h
without creating a circular include. Add inlines.h on top of internal.h and
cid.h, and move the function there. The function was sub.h's only cid.h
user, so drop that include. Pure code move, no functional change.
v2: Host the function in a new inlines.h instead of at internal.h's tail,
which formed a circular include with cid.h. Drop sub.h's now-unused
cid.h include. (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_discard_ecaps_to_sync() waited for balance_one() to consume a dying
sched's queued ecaps sync, polling with resched_cpu() + msleep(). The wait
is unbounded - the ext dl_server forces picks through sustained fair or RT
load only while ext tasks are queued, so an ext-idle cpu monopolized by a
higher class can stall the teardown indefinitely.
Remove the node directly instead: take all queued nodes, drop the dying
sched's and resplice the rest. Consumption runs under the rq lock and batch
nodes read as on-list throughout, so the producer-side dedup stays correct.
A node that an in-flight scx_process_sync_ecaps() batch holds across a
dispatch-induced rq unlock still needs a wait, but one bounded by that batch
completing rather than by a future balance.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
This is needed by rust/helpers/srcu.c which now adds
rust_helper_srcu_readers_active() as a wrapper around the SRCU helper
for Rust callers.
To achive this:
1- Move the srcu_readers_active() implementation from
"kernel/rcu/srcutree.c" to "include/linux/srcutree.h".
2- Implement a matching srcu_readers_active() in
"include/linux/srcutiny.h" and use it on the existing open-coded
WARN_ON() check in cleanup_srcu_struct().
Signed-off-by: Onur Özkan <work@onurozkan.dev>
Reviewed-by: Gary Guo <gary@garyguo.net>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Boqun Feng <boqun@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Restructure the SRCU initialization functions so it always follows
one direction:
init_srcu_struct() -> __init_srcu_struct() -> lockdep or generic
This uses the same wrapper style as mutex. It avoids the old confusing
style where init_srcu_struct() and __init_srcu_struct() called each
other in different configs. It also helps Rust side to have simpler
helper for SRCU initialization.
Signed-off-by: Onur Özkan <work@onurozkan.dev>
Reviewed-by: Gary Guo <gary@garyguo.net>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Boqun Feng <boqun@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit abstracts the open-coded dumping of reader segments in the
rcu_torture_cleanup() function into a new rcu_torture_dump_read_segs()
function. This abstraction will allow reader segments to be dumped for
other purposes.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Now that Tasks Trace RCU is implemented in terms of SRCU, it no longer
has any particular need for the IRQ_WORK Kconfig option. This commit
therefore removes the "select IRQ_WORK" from the TASKS_TRACE_RCU Kconfig
option.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The kerneldoc of sugov_iowait_apply() says the IO boost value is increased
in sugov_iowait_apply() and, in the same sentence, that it is decreased by
the same function. That is self-contradictory, and the first part is wrong:
sugov_iowait_apply() only decreases the boost.
The boost is actually increased in sugov_iowait_boost(). Fix the comment to
name sugov_iowait_boost() as the place where the boost is increased, so it
matches the code.
No functional change.
Fixes: fd7d5287fd ("cpufreq: schedutil: Cleanup and document iowait boost")
Signed-off-by: Zhongqiu Han <zhongqiu.han@oss.qualcomm.com>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Link: https://patch.msgid.link/20260703092433.4080165-1-zhongqiu.han@oss.qualcomm.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
In some cases, the processor may not actually stick to the "desired"
performance level programmed through the driver's .adjust_perf()
callback and may go above it, which may not be desirable (for instance,
there may be a UCLAMP_MAX limit set for the task currently running on
the given CPU which should be respected).
Address that by adjusting the .adjust_perf() callback to take an
additional argument, max_perf, representing the maximum allowed
performance level of the CPU and update the intel_pstate driver to
take that argument into account as appropriate.
Accordingly, adjust cpufreq_driver_adjust_perf() and the other existing
user of .adjust_perf(), which is the amd-pstate driver (but the behavior
of that driver is not changed).
While at it, also update the cpufreq_driver_adjust_perf()
documentation to reflect this change and some previous code
changes that have not been taken into account in it.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Zhongqiu Han <zhongqiu.han@oss.qualcomm.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Link: https://patch.msgid.link/6277654.lOV4Wx5bFT@rafael.j.wysocki
[ rjw: Adjusted Rust function formatting ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
psi_schedule_rtpoll_work() is called locklessly from the scheduler hotpath
and can race psi_trigger_destroy() taking down the last rtpoll trigger under
rtpoll_trigger_lock:
psi_schedule_rtpoll_work() psi_trigger_destroy()
rcu_read_lock();
task = rcu_dereference(rtpoll_task);
rcu_assign_pointer(rtpoll_task, NULL);
timer_delete(&rtpoll_timer);
mod_timer(&rtpoll_timer, ...);
rcu_read_unlock();
synchronize_rcu();
kthread_stop(task_to_destroy);
The group can then be freed with the re-armed timer still pending, and
poll_timer_fn() runs on freed memory.
461daba06b ("psi: eliminate kthread_worker from psi trigger scheduling
mechanism") deleted the timer synchronously after the synchronize_rcu(),
which prevented this but raced trigger creation instead: the deletion could
cancel the timer that a new trigger set armed during the grace period and,
as creation also reinitialized the timer at the time, corrupt it.
8f91efd870 ("psi: Fix race between psi_trigger_create/destroy") moved the
initialization into group_init() and the deletion into the locked section,
trading the creation races for the window above.
Neither placement in the destruction path works. A pending timer firing
while the group is alive is harmless though. poll_timer_fn() just wakes the
rtpoll waitqueue and doesn't re-arm itself. Bind the timer to the group's
lifetime instead and shut it down in psi_cgroup_free(). Nothing can arm it
by then. timer_shutdown_sync() because the timer is never armed again.
Fixes: 8f91efd870 ("psi: Fix race between psi_trigger_create/destroy")
Cc: stable@vger.kernel.org # v5.10+
Reported-by: Sashiko AI <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/all/20260711000434.36C4A1F000E9@smtp.kernel.org/
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Tested-by: Matt Fleming <mfleming@cloudflare.com>
Acked-by: Suren Baghdasaryan <surenb@google.com>
a5b98009f1 ("sched/psi: fix race between file release and pressure write")
made pressure_write() hold cgroup_mutex across psi_trigger_create(), which
forks the psimon kthread for the first rtpoll trigger. As kthread creation
depends on the whole fork path, the commit inadvertently created a lot of
unwanted locking dependencies from cgroup_mutex.
sched_ext got hit by one: its enable path blocks forks and then grabs
cgroup_mutex, so a pressure write racing a scheduler enable deadlocks, with
every other fork piling up behind.
Fix it by splitting trigger creation so that the worker is forked with
cgroup_mutex dropped and the kernfs active reference left broken. The latter
matters because rmdir and cgroup.pressure writes drain active references
under cgroup_mutex. Publishing the trigger last keeps error reporting
synchronous and preserves the of->priv lifetime rules.
The trigger registered in the first stage pins the group's rtpoll machinery
across the unlocked window, leaving only creation races to resolve. The
catch-up poll on installation covers scheduling attempts dropped while there
was no worker.
v2: Retagged sched/psi (was cgroup).
Fixes: a5b98009f1 ("sched/psi: fix race between file release and pressure write")
Cc: stable@vger.kernel.org
Cc: Edward Adam Davis <eadavis@qq.com>
Cc: Chen Ridong <chenridong@huaweicloud.com>
Reported-by: Matt Fleming <mfleming@cloudflare.com>
Closes: https://lore.kernel.org/all/20260710100441.2653477-1-matt@readmodwrite.com/
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Tested-by: Matt Fleming <mfleming@cloudflare.com>
Acked-by: Suren Baghdasaryan <surenb@google.com>
The verifier rejects variable offsets for PTR_TO_TP_BUFFER and PTR_TO_BUF
accesses, but it currently accepts a constant negative offset produced by
pointer arithmetic.
Commit 022ac07508 ("bpf: use reg->var_off instead of reg->off for
pointers") moved constant pointer offsets from reg->off to reg->var_off.
However, __check_buffer_access() continued to check only the instruction
offset. An access with reg->var_off equal to -8 and an instruction offset
of zero therefore passes verification.
For writable raw tracepoints, the access end is also calculated from the
unsigned reg->var_off.value. An eight-byte access starting at -8 wraps
the calculated end to zero, allowing the program to load and attach
without increasing max_tp_access.
After ensuring that reg->var_off is constant, calculate the effective
access start using signed arithmetic and reject it when it is negative.
Use the validated start to calculate the access end for both
PTR_TO_TP_BUFFER and PTR_TO_BUF.
Fixes: 022ac07508 ("bpf: use reg->var_off instead of reg->off for pointers")
Signed-off-by: Sun Jian <sun.jian.kdev@gmail.com>
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Cc: stable@vger.kernel.org # 5.2.0
Link: https://patch.msgid.link/20260714093846.18159-2-sun.jian.kdev@gmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Helper and kfunc argument checking carried two separate meta structs: the
verifier-local struct bpf_call_arg_meta and bpf_kfunc_call_arg_meta.
Merge them into a single struct bpf_call_arg_meta. This is groundwork for
sharing argument checking between helpers and kfuncs.
While merging, drop the btf_id field from the helper meta since it is
never used.
No functional change.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260715064047.1793790-8-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
meta->pkt_access is only ever a copy of fn->pkt_access, assigned once in
check_helper_call() and read back in may_access_direct_pkt_data(). Have
may_access_direct_pkt_data() take the bpf_func_proto and read
fn->pkt_access directly, and drop the meta field along with its
assignment.
The only non-NULL caller, check_func_arg(), already has fn in scope; the
remaining callers pass NULL and are unaffected.
No functional change intended.
Suggested-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Link: https://lore.kernel.org/bpf/20260715064047.1793790-7-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
The constant "size of the PTR_TO_MEM returned in R0" argument is handled by
both helpers (ARG_CONST_ALLOC_SIZE_OR_ZERO) and kfuncs (__rdonly_buf_size /
__rdwr_buf_size), each with its own meta field (meta->mem_size,
meta->r0_size) and duplicated validation. Add struct arg_alloc_mem_desc
and a shared process_const_alloc_mem_size(), and replace both fields with
meta->arg_alloc_mem.
The desc records presence with a 'found' flag instead of using a non-zero
size as the sentinel. This also fixes a pre-existing bug on the kfunc
return path: "no size argument" was tested as r0_size == 0, so an explicit
__rdonly_buf_size/__rdwr_buf_size of 0 was treated as absent and fell
through to btf_resolve_size(), giving R0 the size of the pointed-to return
type instead of 0. With 'found', an explicit zero size is honored and
btf_resolve_size() is used only when no size argument was passed.
The size is stored in a u32, matching regs[R0].mem_size. The U32_MAX
check now apply to both helper and kfunc through
process_const_alloc_mem_size().
Fold bpf_session_cookie return size assignment into current kfunc return
size resolution path.
Note that verifier saves kfunc return size through r0_size instead of
mem_size. The later has no active readers so remove it.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260715064047.1793790-5-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
check_func_arg() only ever handles register arguments (the caller
loops over the first MAX_BPF_FUNC_REG_ARGS arguments), so a single
argno_t built from the register number identifies the argument for
every callee. Remove the duplicated argno_from_reg() calls to
simplify check_func_arg().
'regno' is still kept for the few places that need the raw register
number directly (register reads, verbose R%d messages) and for
referring to the neighbouring size/memory argument in the
ARG_CONST_SIZE{,_OR_ZERO} cases.
No functional change intended.
Suggested-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Link: https://lore.kernel.org/bpf/20260715064047.1793790-4-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
To prepare for unifying the helper and kfunc call_arg_meta, group the
scattered MEM_UNINIT ("raw") memory argument fields (raw_mode, regno and
access_size) into a new struct arg_raw_mem_desc. The intention is to
make it clear about when these are set and used instead of fields with
overly generic names.
Identify the raw argument once, up front, in check_raw_mode_ok() (like
check_proto_release_reg() does for release_regno), recording its regno.
check_stack_range_initialized() now recognizes the raw buffer by matching
that regno, so the separate raw_mode flag is no longer needed, and the
per-argument "meta->raw_mode = arg_type & MEM_UNINIT" assignments in
check_func_arg() go away with it.
A raw memory argument can be tagged either ARG_PTR_TO_MEM | MEM_UNINIT or
ARG_PTR_TO_MAP_VALUE | MEM_UNINIT (the output buffer of bpf_map_pop_elem()
and bpf_map_peek_elem()). Either may be passed as a PTR_TO_STACK, which
reaches check_stack_range_initialized() through check_helper_mem_access(),
so both must be treated as raw. Extend arg_type_is_raw_mem() to match the
map value case as well; otherwise check_raw_mode_ok() would not record the
regno for it and an uninitialized stack buffer passed to those helpers
would be wrongly rejected for programs without CAP_PERFMON.
No functional change intended. This patch does not enable raw_mode
memory access for kfunc (i.e., uninit stack will not be allowed to be
passed to kfunc for unprivileged programs). Existing kfuncs with arguments
tagged with __uninit are either priviledged or dynptr kfuncs, which take
another path to make sure the access is checked by check_mem_access().
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260715064047.1793790-3-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
For a MEM_UNINIT ("raw mode") helper argument,
check_stack_range_initialized() open-coded a scan that rejected any
STACK_DYNPTR slot in the range with "potential write to dynptr". This
duplicated, and was stricter than, the handling that runs when the
buffer is actually marked initialized. check_helper_call() later replays
the write byte by byte via check_mem_access(), which goes through
destroy_if_dynptr_stack_slot(), which rejects overwritting a referenced
dynptr. Therefore drop the redundant scan and rely on check_mem_access().
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260715064047.1793790-2-ameryhung@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
luo_session_alloc() calls INIT_LIST_HEAD(&session->file_set.files_list)
followed immediately by luo_file_set_init(), which also performs
INIT_LIST_HEAD() on the same list head. Remove the duplicate call.
Signed-off-by: Hongfu Li <lihongfu@kylinos.cn>
Reviewed-by: Pratyush Yadav <pratyush@kernel.org>
Link: https://patch.msgid.link/20260714074056.95335-1-hongfu.li@linux.dev
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
To avoid potential stack overflows on limited kernel stacks, convert
parse_probe_arg() from a recursive function into a loop-based
implementation with a simple local state stack.
Since recursion is eliminated using a loop with a fixed-size
stack in the context, this restricts the dereference nesting
depth. The maximum nesting depth of dereferences is now restricted
to the same limit as typecasts (TRACEPROBE_MAX_NESTED_LEVEL, which
is 8) and reports the same TOO_MANY_NESTED error. Update ftrace
selftests to reflect this restriction and simplify the checks.
Note that this change slightly alters the behavior of nested
dereferencing in fetcharg. Previously, dereferencing without BTF
allowed for up to 14 levels of nesting, whereas dereferencing
with BTF was limited to 3 levels. With this change, the nesting
depth is now limited to 8 levels in both cases.
Link: https://lore.kernel.org/all/178399141396.27810.5390060618628718661.stgit@devnote2/
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
To support BTF argument parsing (such as accessing fields within nested
structures via typecasting), the maximum argument string length needs
to be extended. Extend MAX_ARGSTR_LEN from 63 to 255.
Since MAX_ARGSTR_LEN was previously reused to format command heads in
trace_*probe_match_command_head() functions, introduce a dedicated
MAX_COMMON_HEAD_LEN (63) macro for matching command heads and switch
these functions to use the new macro.
Link: https://lore.kernel.org/all/178399140457.27810.11387684872148824707.stgit@devnote2/
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Decompose parse_probe_vars() by extracting retval, stack, current task
struct, and function argument parsing logic into dedicated static
helper functions (parse_probe_var_retval, parse_probe_var_stack,
parse_probe_var_current, and parse_probe_var_arg). This simplifies
parse_probe_vars() and improves its readability.
Link: https://lore.kernel.org/all/178399137612.27810.14869178899276413372.stgit@devnote2/
Assisted-by: Antigravity:gemini-3.5-flash
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
fp pointer and unsigned long have the same size on all relevant
architectures that build Linux. Furthermore this struct is only used in
architectures that do not set ARCH_DEFINE_ENCODE_FPROBE_HEADER which is
set only for 64bit architectures (apart from LoongArch).
Both fields are aligned on these architectures so the struct with
__packed and without it are the same.
Remove the __packed as it is unnecessary.
[Masami: Removed Fixes tag because this is not fixing any problem.]
Link: https://lore.kernel.org/all/20260428-topic-fprobe-packed-v7-1-v1-1-9abc9b866b4c@baylibre.com/
Signed-off-by: Markus Schneider-Pargmann (The Capable Hub) <msp@baylibre.com>
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
When tracing the kernel local variables, sometimes we need to get the
CPU local variables. To access it, current simple dereference is not
enough.
Thus, introduce a special this_cpu_read() dereference to access per-cpu
variable for the current CPU (accessing other CPU variable may race with
updates on other CPUs). Also this_cpu_ptr() is for accessing per-cpu
pointer.
Those are working as same as the kernel percpu macro.
Link: https://lore.kernel.org/all/178271367680.1176915.4711734074448973989.stgit@devnote2/
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Since we can use the BTF to cast value to a structure pointer type,
it is useful to introduce "$current" special variable support to
fetcharg.
User can define a fetcharg to access current task_struct properties
using BTF info. e.g.
$current->cpus_ptr
Link: https://lore.kernel.org/all/178271366709.1176915.15320906169981578568.stgit@devnote2/
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Add a field specifier option for the typecast. This works like
container_of() macro.
(STRUCT[,FIELD[.FIELD2...]])VAR
This is equivalent to :
container_of(VAR, struct STRUCT, FIELD[.FIELD2...])
For example:
echo "f tick_nohz_handler next_tick=(tick_sched,sched_timer)timer->next_tick" >> dynamic_events
This will trace tick_nohz_handler() with its tick_sched::next_tick which
is converted from @timer by contianer_of(tick, struct tick_sched, sched_timer).
So, if you enabkle both fprobes:tick_nohz_handler__entry and
timer:hrtimer_expire_entry events, we will see something like:
<idle>-0 [002] d.h1. 3778.087272: hrtimer_expire_entry: hrtimer=00000000d63db328 f
unction=tick_nohz_handler now=3777450051040
<idle>-0 [002] d.h1. 3778.087281: tick_nohz_handler__entry: (tick_nohz_handler+0x4
/0x140) next_tick=3777450000000
Link: https://lore.kernel.org/all/178271365745.1176915.725923927180862257.stgit@devnote2/
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
This allows type casting to various fetchargs without parentheses
by recursively calling parse_probe_arg on the target when type
casting is used.
For example, this allows the following expressions:
- (STRUCT)%REG->FIELD
- (STRUCT)$stackN->FIELD
- (STRUCT)@SYM->FIELD
Note that @SYM+/-OFFSET with typecast needs parentheses like:
- (STRUCT)(@SYM-8)->FIELD
Link: https://lore.kernel.org/all/178271364804.1176915.17600229399446264813.stgit@devnote2/
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
When we hit an open parenthesis right after typecast closing
parenthesis, it means we have nested typecast. This allows us to
typecast a generic data member in a structure to a pointer to
another structure.
For example, to cast a DATA_MEMBER of VAR structure to STRUCT pointer
and get MEMBER value.
(STRUCT)(VAR->DATA_MEMBER)->MEMBER
Also, we can nest typecast.
(STRUCT1)((STRUCT2)$ARG->FIELD2)->FIELD1
Currently the max nest level is limited to 3.
This also allows user to use typecasting for registers or stacks on
kprobe events. e.g.
(STRUCT)(%ax)->MEMBER
(STRUCT)($stack0)->MEMBER
Link: https://lore.kernel.org/all/178271363855.1176915.16793301788257446529.stgit@devnote2/
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Support BTF typecast feature on other probe events, but only if it is
kernel function entry or return, and must use function parameter name
or $retval. This means you can do:
(STRUCT)PARAM->MEMBER
Note: you can not use other variables like $stackN, %reg etc. That
needs nesting support.
To support other probe events, we just need to use last_struct type
when we find a function parameter in parse_btf_arg().
This also updates <tracefs>/README file to show struct typecast.
Link: https://lore.kernel.org/all/178271362928.1176915.12235759508786922490.stgit@devnote2/
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
A cid-form scheduler can grant caps to and revoke them from its child
sub-schedulers but has no way to tear one down. Add scx_bpf_sub_kill() to
evict a direct child with a printf-style reason that reaches the child's
scx_exit_info. No exit code is taken because the child is a separate
scheduler whose exit-code semantics the parent cannot know. The child and
its subtree are disabled through the usual async path under a new exit kind,
SCX_EXIT_PARENT_KILL.
The bstr formatting infrastructure in ext.c is exposed through internal.h
with scx_ prefixes so the kfunc, which lives in sub.c, can format the
reason.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_process_sync_ecaps() consumes ecaps syncs while the sched is bypassing
without delivering ops.sub_ecaps_updated(), leaving reported_ecaps stale.
Nothing re-queued a sync when bypass lifted, so a cid whose caps never
change again would never be notified. Attach-time initial grants hit this
every time: they are consumed during the enable bypass window, so a sched
never learned its initial effective caps through the callback.
Re-queue a sync for every (sched, cpu) with an undelivered delta at the
per-cpu bypass exit in scx_bypass(), next to the idle renotify catch-up. The
next balance on the cpu then delivers the pending delta with proper dispatch
context.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
__scx_update_idle() notified only the root scheduler. A sub-scheduler that
holds a cid needs that cid's idle state to place and kick on it.
Deliver ops.update_idle() to every scheduler that holds SCX_CAP_BASE on the
transitioning cid. The root holds every cap, so a real transition always
reaches it.
Real transitions are not enough on their own. A cid that is already idle
when a sub-sched gains baseline access produces no transition, so the new
holder would never learn it is idle. The ecaps sync arms a re-notify on the
gain, and the next idle pick delivers ops.update_idle() to just that sched,
leaving holders that already track the cpu untouched. A matching loss of
baseline access drops any pending re-notify.
Bypass suppresses ops.update_idle() too, so a cpu that goes idle during a
bypass window and stays idle yields no transition to re-deliver on
un-bypass. Arm the same re-notify for every sched leaving bypass. The acute
case is a child granted cids during its own ops.sub_attach(). The grant
lands while the child is bypassed and the notify walk skips it, so on
un-bypass it holds cids it never saw go idle. The root is owed the same and
is armed through a separate per-rq flag, which keeps this working when
sub-schedulers are compiled out.
v2: Gate the idle catch-up in pick_task_idle() to avoid a double ops.update_idle(). (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_local_or_reject_dsq() authorizes a local-DSQ insert against the caps
of the scheduler doing the insert. On the consume/dispatch paths that is
the scheduler running balance_one(), passed down through
scx_consume_dispatch_q() and move_local_task_to_local_dsq(), so the check
is correct.
The remote-move path loses it. move_remote_task_to_local_dsq()
re-activates @p on the destination rq through enqueue_task_scx(), which
reconstructs the scheduler from the task, i.e. @p's owner. When an
ancestor places a descendant's task - e.g. draining a bypassed
sub-scheduler - the owner is a sub-scheduler of the placer, so
authorizing against the owner checks a narrower cap set and can
spuriously reject a task the placer is entitled to run.
Carry the placing scheduler across the activate_task() boundary the same
way enq_flags already are, via a per-rq field set only for the duration
of the re-activation, and have scx_local_or_reject_dsq() authorize
against it. The placer's caps are a superset of the owner's, so this
admits what the placer may run and keeps rejecting what it may not.
v2: Document @sch in move_remote_task_to_local_dsq()'s kerneldoc. (Andrea)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A kick forces a scheduling event on the target cpu, and a preemption also
evicts the running task. Gate both on caps. Any kick requires baseline
access on the cid, and preempting a task the sub-sched does not own -
whether by a SCX_ENQ_PREEMPT insert or a SCX_KICK_PREEMPT kick - requires
the new SCX_CAP_PREEMPT. Gating either alone would leave a hole - the
weakest cap authorizing preempting kicks, or plain kicks disturbing cpus the
kicker has no access to.
Preempting the sched's own subtree is always allowed, and the cap extends
the right to any task on the cid. PREEMPT implies ENQ, and so ENQ_IMMED.
A preempting insert tests the running task under the target rq lock and is
rejected and reenqueued unless the victim is in the inserter's subtree or it
holds PREEMPT. A migration-disabled task is admitted regardless, but with
SCX_ENQ_PREEMPT stripped.
Kicks are enforced on the delivery path, where the effective caps can be
read coherently under the target rq's lock. A kick from a sub-sched lacking
SCX_CAP_BASE on the cid is dropped, and a SCX_KICK_PREEMPT kick without
PREEMPT for a task outside the kicker's subtree degrades to a plain
reschedule.
Unlike the enqueue caps, PREEMPT is checked only at the instant of the
insert or kick, never as a standing property of a queued task.
v2: Clear SCX_ENQ_PREEMPT on the offline and migration_pending force-admits.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Add SCX_CAP_ENQ, which gates inserting tasks onto a cid's local DSQ. Unlike
IMMED enqueue, plain enqueues can pile up, so ENQ is the stronger cap and
implies ENQ_IMMED. Losing ENQ also triggers the reenq scan. The scan tests
each queued task and the running task against the cap each needs via
scx_caps_for_task(), so an ENQ-only loss reenqueues plain tasks, evicting a
running one, while IMMED tasks, which need only ENQ_IMMED, stay put.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A task's slice grants it cpu occupancy - how long it holds its cpu. In a
sub-scheduler hierarchy cpu access is delegated through revocable
capabilities, so a task's occupancy must follow them. Only its own scheduler
sets its slice, and extending the slice is allowed only while that scheduler
holds baseline cpu access (SCX_CAP_BASE) on the cpu. Otherwise a scheduler
could keep occupying a cpu it has been denied simply by handing out long
slices.
The cap check reads effective caps, which are coherent only under the task's
rq lock, and the kernel decrements the slice under that lock as the task
runs, so a running task's slice can be changed only there while a queued
task's can be set directly. Make scx_bpf_task_set_slice() apply the slice
under the rq lock. Synchronously when the caller already holds it, otherwise
by stashing it in the new p->scx.slice_oob, tagged with the scheduler's id
so a request that outlived a reassignment is dropped. Whether the caller
holds @p's current rq lock is tested with p->scx.runnable_cpu.
Revocation is enforced through the same grant. When a cpu's effective caps
lose SCX_CAP_BASE, the cap-revoke reenq scan also checks the running task
and zeroes its slice to evict it. The scan runs as a balance callback after
the pick, so this catches both the task that was running when the revoke
landed and a capless task the pick just promoted off the local DSQ. The
paths that keep a task on its cpu - holding on to the last runnable task in
balance, the ENQ_LAST reinsertion and the slice refill on pick - skip tasks
lacking baseline access. A migration-disabled task is exempt, mirroring its
capless admission on insert.
v4: Test rq ownership with p->scx.runnable_cpu, closing a remote-wakeup TOCTOU. (sashiko AI)
v3: Keep a pending out-of-band slice request across refill and preserve. (sashiko AI)
v2: Only write slice directly when @p is queued on the held rq. (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Add p->scx.runnable_cpu, the cpu @p is runnable on, or -1 when it is not.
It is stamped as @p joins the runnable_list (set_task_runnable()) and
cleared as it leaves (clr_task_runnable()), both under the rq lock.
task_cpu() can't answer "is @p on this rq" reliably: a remote wakeup changes
it under @p's pi_lock alone, without the source rq lock, so it can read as
the locked rq while @p is really elsewhere. runnable_cpu changes only under
the rq lock, so a caller holding an rq lock can compare against it to know
whether that is @p's current rq.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A later change makes set_task_slice() also drop a pending out-of-band slice
request, so the BPF-triggered writes to p->scx.slice need to funnel through
one place. Introduce set_task_slice() and route those writes through it.
update_curr_scx() decrements curr->scx.slice directly for accounting and is
left alone. No functional change - the helper only assigns p->scx.slice.
v2: Reword the set_task_slice comment to "BPF-triggered writes". (Andrea)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Neither a scx_sched pointer nor its cgroup id uniquely identifies a
scheduler instance. A freed sched's memory can be reallocated, and a cgroup
can detach one sched and attach another. Add a monotonic, never-reused u64
id. A later patch compares it to drop a slice request that outlived a change
of a task's owning scheduler.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Replace the __SCX_CAP_DUMMY placeholder with SCX_CAP_ENQ_IMMED, which gates
inserting IMMED tasks onto a cid's local DSQ. An IMMED enqueue is guaranteed
to either get its task running on the cpu at once or hand it back to the
scheduler, so IMMED work can never pile up on the cpu's queue and a cpu can
be shared across sub-scheds through IMMED access without any of them
swamping it.
That makes ENQ_IMMED the natural baseline, the minimal cap to make any use
of a cpu. SCX_CAP_BASE aliases it so gates on basic cpu access can state the
intention instead of naming ENQ_IMMED.
Enforcement covers inserts and queued tasks. An insert without the cap is
diverted to the reject DSQ, and queued tasks are reenqueued when the cap is
lost. scx_bpf_sub_dispatch() skips a child that lacks the cap on the cpu, as
its inserts would only be rejected. Vacating the running task on cap loss
lands in a later patch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A SAVE/RESTORE requeue re-inserts a running task in place and is immediately
followed by set_next_task_scx(). It is not a real scheduling event: the task
is already admitted to its cid and must return to the local DSQ
unconditionally.
scx_caps_for_enq() maps an enqueue to the cap its local-DSQ insert requires.
Add SCX_ENQ_IGNORE_CAPS, set it on the RESTORE-in-place branch of
enqueue_task_scx(), and have scx_caps_for_enq() require no caps for it, so
the cid admission gate never diverts an in-place restore to the reject DSQ.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
When a sub-scheduler dispatches a task to a CPU it lacks the required
capability on, the task must be rejected rather than allowed to run.
Add the machinery for that. Each rq gets a reject DSQ, a kernel-internal
holding queue that is never run and that the BPF scheduler cannot reach. An
insert that must be refused is diverted there instead of the local DSQ, and
a deferred requeue then hands the parked tasks back to the BPF scheduler to
re-decide. A cap revoke extends this to already-queued tasks. When the
revoke reaches the cpu's effective caps, the cpu scans its local DSQ and
reenqueues the tasks that no longer qualify.
A migration-disabled task must run on its cpu, so a capless one is admitted
anyway and counted in the new SCX_EV_SUB_FORCED_ADMIT event.
This is preparation for the actual sub-sched cap enforcement. The divert is
wired but inert here.
v2: Admit offline-rq and migration_pending inserts to local, not reject. (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The local DSQ is synchronized by the containing rq lock rather than its own
dsq->lock. A later patch adds a second such DSQ. In preparation, factor the
"rq owns the lock" test into dsq_is_rq_owned() and rename
local_dsq_post_enq() to rq_owned_post_enq(), taking @rq explicitly.
No behavior change.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A sub-scheduler that gains or loses effective caps on a cpu may want to act
on it right away - e.g. place or preempt on a newly usable cpu. The existing
ops.sub_caps_updated() doesn't fit as it is delivered asynchronously to
scheduling operations and can arrive before the per-cpu effective caps go
live.
Add ops.sub_ecaps_updated(cid, before, after), a cid-form callback fired
from scx_process_sync_ecaps() when a sub-sched's effective caps on a cid
change. It runs in dispatch context so the sched can insert, kick or preempt
on the cid directly. @before is the caps as of the last delivery.
Cpu hotplug rides the same machinery. Going down zeroes each sched's ecaps
on the cpu's cid, with queued syncs discarded at consumption while the cpu
is inactive. Coming back up queues a sync for every sched. reported_ecaps is
kept across the down/up cycle, so the resync fires the callback only if
ownership actually changed while the cpu was down.
v2: Compute cid below the active-cpu guard; discard queued syncs on !cpu_active(). (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Checking a sched's caps on a cid would need to test several cap bits against
caps[] to account for implied caps. Also, caps[] modifications aren't
synchronized against scheduling operations on each cpu, which can lead to
awkward race conditions.
Collect them per cpu instead. caps[] under pshard->lock stays the target
configuration. scx_sched_pcpu->ecaps is added, the transposed effective
copy: the set of cap bits the sched holds on that cpu which can be accessed
with a single read. It is stable under the rq lock. It can also be read
locklessly with READ_ONCE().
Grant and revoke only mutate caps[]. They queue a sync request on the target
cpu's rq->scx.ecaps_to_sync and kick it, and the cpu recomputes the queued
scheds' ecaps from caps[] in balance_one() under its own rq lock. A dying
sched runs the sync directly to retire its queued request before freeing. As
held references can defer the freeing past the enclosing root scheduler's
lifetime, root enable discards leftover sync requests before going live.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Wire up ops_cid.sub_caps_updated() to notify sub-scheds of cap changes.
Three constraints shape the design:
1. Static memory. Deliveries use a fixed-size buffer, both for runtime
efficiency and so notifications can't be lost under memory pressure.
2. High-frequency updates. Grant/revoke can mutate caps in bursts, and the
notifier path must absorb that without amplifying it.
3. Recursive grant/revoke from the callback. A child receiving a
notification can call grant/revoke on its own children, which can
cascade recursively down its subtree.
(1) and (2) lead to coalescing into a fixed payload. Each delivery carries a
single (cmask, caps) pair covering every change since the previous one.
Direction (set vs cleared) isn't encoded as it doesn't fit in the fixed-size
summary. The callback queries scx_bpf_sub_caps() for current state. Only one
delivery is in flight per shard. Further changes fold into the same buffer
and ship as the next callback, so a shard's callbacks fire in order.
(3) leads to deferred delivery. Events accumulate during grant/revoke and
are delivered after the shard lock is released.
v2:
- Request a private stack for ops.sub_caps_updated(). (sashiko AI)
- Build cmask_arena_out via scx_cmask_ref, not by re-reading its header.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Caps are per-cid permissions parents delegate to direct children via
scx_bpf_sub_grant() / scx_bpf_sub_revoke(). A child's cap set is always a
subset of its parent's. Sub-scheds check their caps locally, and cross-sched
communication is needed only when the delegation set itself changes.
Caps will be used to implement sub-sched scheduling on the enqueue path.
Picking a cid for a task at a leaf depends on which cids the leaf is allowed
to use, and resolving that programmatically on every enqueue would mean a
cross-sched round-trip call chain, possibly retrying if the request can't be
granted as-is. The dispatch path is different - it runs as top-down
recursion via scx_bpf_sub_dispatch().
Locking is per shard. cid space is split into shards, and each sub-sched has
its own pshard->lock for each shard. Operations are broken up on shard
boundaries. Different shards never contend. Shards are expected to be
topology-aligned and likely to serve as the locality unit when cids are
allocated to schedulers, so per-shard lock granularity scales naturally with
the allocation pattern.
This patch adds the framework with a single dummy cap. Real caps land in
later patches.
The enable path is reordered for pshards. scx_arena_pool_init() moves ahead
of scx_link_sched() so the pshards are allocated before the sched becomes
reachable - scx_alloc_pshards() skips allocation when the arena pool isn't
initialized.
- scx_bpf_sub_grant(): Per-cid all-or-nothing grant to direct child.
- scx_bpf_sub_revoke(): Clear caps on @cmask across @child and its subtree.
- scx_bpf_sub_caps(): Lockless snapshot of caps on a cid range.
/sys/kernel/sched_ext/SCHED/caps shows the caps each scheduler currently
holds.
v4: Move the pshard[] full build/publish and the err_disable scx_error() recording to earlier patches. (sashiko AI)
v3: Build pshard[] fully before publishing it, read it with READ_ONCE. (sashiko AI)
v2: Validate ops before scx_link_sched() publishes the sub. (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A scheduler's BPF programs can outlive it. A timer it armed or a tracing
program it loaded can fire after ops.exit() has run, before the programs are
unloaded, and scx_prog_sched() still resolves the program to its scheduler
through ops->priv. Harmless while kfuncs touch only lifetime-stable state,
but a hazard once a kfunc reads global state a newly loaded scheduler can
change underneath it.
Add scx_sched->dead, set right after ops.exit() and drained with
synchronize_rcu(). It follows exit() rather than preceding it so exit()'s own
kfunc calls still resolve to @sch. scx_prog_sched() returns NULL for a dead
scheduler, so every kfunc's existing !sch bail rejects it at one choke
point.
v2: Check dead in the CONFIG_EXT_SUB_SCHED=n scx_prog_sched() too. (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Factor the sibling/ancestor portion of scx_next_descendant_pre() out as
scx_skip_subtree_pre(), a pre-order walk primitive that skips @pos's
subtree, and call it from scx_next_descendant_pre(). Same locking rules as
the existing primitive.
Used in a follow-up to fast-skip subtrees that have nothing to do during a
descendant walk.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Future kfuncs need to walk descendants without scx_sched_lock. Make the
walker RCU-safe so that they can. A sub-sched's fields are initialized
before it is linked, so a walk that observes a linked node also observes its
setup. In-place changes after linking carry their own ordering.
Switch the children/sibling list ops to RCU and expand the descendant walker
to accept rcu_read_lock as a valid read-side context. Walkers that mutate
keep scx_sched_lock.
A sub-sched can be linked while an ancestor is bypassing, after the bypass
walk that propagates the depth has passed its parent. Bypass state is a
per-cpu flag plus a depth count and can't be established atomically at link
time, so refuse to link under a bypassing ancestor. Take scx_bypass_lock
across linking to check the parent's bypass state coherently.
v3: Reject linking under a bypassing ancestor instead of inheriting bypass_depth. (sashiko AI)
v2: Inherit bypass_depth before publishing @sch on the RCU sibling list.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_call_op_set_cpumask() builds a per-cpu cmask in the set_cmask scratch,
which lives in BPF-writable arena. A scheduler can corrupt the scratch's
inline header (base, nr_cids, alloc_words) from another cpu, so sizing and
indexing the write from it risks an out-of-bounds write.
Drive the build from kernel-known geometry instead.
scx_cmask_ref_init_kern() imposes base and nr_cids rather than reading them,
and scx_cmask_ref_from_cpumask() fills the scratch from the ref. Neither
reads the header back.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
kfuncs taking struct scx_cmask * from BPF arena memory have two problems.
The pointer can be any value the BPF prog hands in, and the header (@base,
@nr_cids, @alloc_words) can be mutated by the prog concurrently with kernel
access.
Add scx_cmask_ref, a validated handle. _init() normalizes the input pointer
into the arena's kern_vm range via scx_arena_to_kaddr() and snapshots the
header, rejecting a range outside the machine or a nr_cids whose words
exceed the declared @alloc_words. Downstream sizing uses the snapshot, not
the live header. _shard() reads slices while _or() and _copy() write back,
all bounded by the snapshot. No callers yet.
struct scx_cmask's bits[] carried __counted_by(alloc_words), so
UBSAN_BOUNDS and FORTIFY_SOURCE bound accesses to the array. That bound is
read from @alloc_words at the access. For an arena cmask @alloc_words is
BPF-writable. A prog that sets it larger than the real allocation makes the
check pass on a genuine overrun, so the annotation catches nothing, and it
only runs under those debug configs. Drop it - _init() validates
@alloc_words explicitly, and kernel-owned cmasks set it themselves.
v2: Validate @alloc_words in _init(), drop __counted_by. (Andrea, sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Add struct scx_pshard and sch->pshard[] indexed by shard_idx, each entry
allocated on its shard's NUMA node from scx_shard_node[si]. The struct
starts empty (one dummy field). Follow-up patches will grow it as
shard-local state lands. Only cid-type schedulers with an arena pool get
pshards.
Allocation happens after ops.init_cids() returns so any
scx_bpf_cid_override() it issues has finalized scx_nr_cid_shards and
scx_shard_node[]. sch->nr_pshards records the array size for the async RCU
free path, which may run after a later scheduler's scx_cid_init() has
rewritten the global.
v3: Build and publish pshard[] fully-formed here rather than a later patch.
v2: Free the partially-allocated pshard array on alloc failure. (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Split kobject_init_and_add() in scx_alloc_and_add_sched(): only
kobject_init() runs there. A new scx_sched_sysfs_add() helper does
kobject_add() (and creates sub_kset when the scheduler implements
ops.sub_attach), called by both enable workfns once @sch is linked and its
sysfs-visible state is initialized. Prep so a future caps attribute can rely
on @sch being fully built by the time it's sysfs-visible. Add early enough
that a stall later in enable still leaves sysfs inspectable.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
An overridden cid mapping invalidates the auto-generated shard layout, so
the override call has to provide both. Extend scx_bpf_cid_override() with a
shard_start[] array that lists the first cid of each shard (starting at 0,
strictly increasing, last shard implicitly extends to num_possible_cpus()).
A scheduler that wants only custom shards with the auto-generated cid
mapping can read the current mapping and pass it back unchanged.
Overridden shards can span NUMA nodes, so scx_shard_node[] is rebuilt by
majority count: each shard is assigned to the node that owns the most cpus
in it.
v2: Snapshot the caller's cpu_to_cid/shard_start arrays before validating. (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Sub-sched operations need a scalable locking / work domain smaller than the
whole cid space. Carve the cid space into topology-respecting shards: each
shard is a contiguous cid range that stays within one LLC, and LLCs larger
than the per-shard cap (default 24 cids, configurable via
ops.cid_shard_size) split into enough shards to fit. A hard cap of
SCX_CID_SHARD_MAX_CPUS prevents pathological sizes under custom
configurations.
No-topo cids pack into their own shards so every cid has a shard assignment.
Also build scx_cid_shard_ranges[] for O(1) shard-to-cid-range lookup and
scx_shard_node[] so callers can size or place work by NUMA without walking
cids. Auto-built shards inherit their LLC's node. No-topo shards carry
NUMA_NO_NODE.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A cid-form scheduler that calls scx_bpf_cid_override() to install a custom
cid layout can only do so from ops.init(). Enable-path setup that depends on
the cid layout thus has to run after ops.init(), and ops.init() itself can't
use anything derived from the final layout, which turned out to be too
restrictive.
Add an ops.init_cids() callback dedicated to finalizing the cid layout. It
runs before the rest of the enable-path setup, so the final layout is in
effect for everything that follows including ops.init(), which now runs
after the arena pool and cmask scratch allocations.
scx_bpf_cid_override() is restricted to ops.init_cids() at load time. It
sits in a kfunc set gated by SCX_KF_ALLOW_INIT_CIDS, a flag set only on the
init_cids op, so the verifier rejects a call from any other context. The
runtime root-only check is dropped as ops.init_cids() only runs during root
enable.
The qmap demo moves its override into a dedicated qmap_init_cids() and,
while at it, introduces an enum for the cid override modes instead of
hard-coded integers.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The kick machinery kept its targets in rq->scx shared by every sched on the
cpu. A preempt kick carried no record of which scheduler requested it.
A later patch needs preempt kicks scoped to the requesting scheduler so a
sub-scheduler can preempt only tasks in its own subtree. Move the kick masks
into the per-sched per-cpu scx_sched_pcpu and have scx_kick_cpu() link the
sched onto a per-cpu list (rq->scx.sched_pcpus_to_kick). The cpu's single
kick irq_work walks that list and kicks each sched's targets on its behalf,
so a kick stays attributed to its scheduler.
The SCX_KICK_WAIT sync set (cpus_to_sync, the kick_sync snapshot and the
balance-callback trigger) stays in rq->scx: the waiter is the cpu, not the
scheduler, and its only writers, the kick irq_work and the wait balance
callback, are cpu-local.
On disable, free_kick_syncs() flushes each cpu's pending kick irq_work
before clearing @ksyncs, so a late kick unlinks its to_kick_node instead of
early-returning on a NULL @ksyncs and leaving the node linked at free.
v3: Flush the kick irq_work in free_kick_syncs() before clearing @ksyncs. (sashiko AI)
v2: Warn once per sched on scx_bpf_kick_cpu() from NMI.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
A per-task op must be dispatched on the scheduler that owns the task.
SCX_CALL_OP_TASK() and its _RET twin take @sch explicitly, and a caller that
passes the wrong scheduler would silently run the op on it. Add a
WARN_ON_ONCE() that @sch matches the task's owner so such a mismatch is
caught rather than hidden.
Two sites legitimately target a scheduler other than the task's owner:
cgroup_move() runs on the root sched, and scx_sub_init_cancel_task() fires
exit_task() on a task not yet associated with @sch. Both switch to the inner
__SCX_CALL_OP_TASK(), which dispatches on the explicit @sch without the
assert.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Improve readability of the sg_dma_len assignment in the P2PDMA cases by
removing duplicated code, which was a direct result of the d0d08f4bd7
("dma-direct: Fix missing sg_dma_len assignment in P2PDMA bus mappings")
fix. No functional change.
Suggested-by: Leon Romanovsky <leon@kernel.org>
Link: https://lore.kernel.org/all/20260604071856.GA245424@unreal/
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Reviewed-by: Leon Romanovsky <leon@kernel.org>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Link: https://lore.kernel.org/r/20260713083022.1110993-1-m.szyprowski@samsung.com
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
- A cpuset that never set its memory nodes could divide by zero when a
task's mempolicy rebinds on CPU hotplug. Rebind against the effective
nodes, which are always populated.
- Documentation fixes for memory.stat, io.stat, and the misc and v1 RDMA
controllers.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCalVWkg4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGet9AQD3/qE93E+PctxXa+rmRfSSjIzqqv9NUZa4THjs
YOUqSAEApVSLPkhg3xHF61q9l+FuocLLY378Uf6LYGWet1SghAQ=
=1Gp+
-----END PGP SIGNATURE-----
Merge tag 'cgroup-for-7.2-rc3-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup
Pull cgroup fixes from Tejun Heo:
- A cpuset that never set its memory nodes could divide by zero when a
task's mempolicy rebinds on CPU hotplug. Rebind against the effective
nodes, which are always populated
- Documentation fixes for memory.stat, io.stat, and the misc and v1
RDMA controllers
* tag 'cgroup-for-7.2-rc3-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
Docs/admin-guide/cgroup-v2: note blkcg_debug_stats gates io.latency stats
Docs/admin-guide/cgroup-v1: document rdma.peak, rdma.events and rdma.events.local
Docs/admin-guide/cgroup-v2: drop stale misc interface file count
cgroup/cpuset: rebind mm mempolicy to effective_mems, not mems_allowed
Docs/admin-guide/cgroup-v2: fix memory.stat doc details
- Lifecycle fixes for the new sub-scheduler support: two use-after-frees
and an enable-failure path that left a half-initialized sub-scheduler
linked.
- Two dispatch-path locking bugs: a spurious scheduler abort from a
migration race, and a lockdep splat from stale runqueue-lock tracking.
- Callback and task-state fixes: stale scheduler-owned state on a task
leaving SCX, a weight callback running after disable, and a bogus
warning on core-scheduling forced idle.
- On nohz_full, finite-slice tasks could miss the tick that expires their
slice. Enable it when such a task is picked, with a selftest.
- Smaller fixes: userspace CPU-mask helpers, ratelimited deprecation
warnings, docs and a sparse annotation.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCalVWgw4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGTmUAP0QHX7Ae/g7yMPEB4slURpnSj/wE7hRCI5jTmay
2iIEJgD/RAhpwUUAp4Abozt0mjQMWh9UyEVjOxPCVNKaNh1XfwU=
=EBCm
-----END PGP SIGNATURE-----
Merge tag 'sched_ext-for-7.2-rc3-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext
Pull sched_ext fixes from Tejun Heo:
- Lifecycle fixes for the new sub-scheduler support: two
use-after-frees and an enable-failure path that left a
half-initialized sub-scheduler linked.
- Two dispatch-path locking bugs: a spurious scheduler abort from a
migration race, and a lockdep splat from stale runqueue-lock
tracking.
- Callback and task-state fixes: stale scheduler-owned state on a task
leaving SCX, a weight callback running after disable, and a bogus
warning on core-scheduling forced idle.
- On nohz_full, finite-slice tasks could miss the tick that expires
their slice. Enable it when such a task is picked, with a selftest.
- Smaller fixes: userspace CPU-mask helpers, ratelimited deprecation
warnings, docs and a sparse annotation.
* tag 'sched_ext-for-7.2-rc3-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:
sched_ext: Skip ops.set_weight() for disabled tasks
tools/sched_ext: scx - Fix cmask_subset(), cmask_equal() and cmask_weight()
sched_ext: Fix premature ops->priv publication in scx_alloc_and_add_sched()
sched_ext: Record an error on errno-only sub-enable failure
selftests/sched_ext: Verify nohz_full tick behavior
sched_ext: Enable tick for finite slices on nohz_full
sched_ext: Preserve rq tracking across local DSQ dispatch
sched_ext: Documentation: Fix ops table header reference
sched_ext: Don't warn on core-sched forced idle in put_prev_task_scx()
sched_ext: Pin parent scx_sched across a child sub-scheduler's lifetime
sched_ext: Annotate ksyncs with __rcu in alloc/free_kick_syncs()
sched_ext: Check remote rq eligibility under task's rq lock
sched_ext: Reset dsq_vtime and slice when a task leaves SCX
sched_ext: Avoid flooding the log with deprecation warnings
Update DMA reserved memory pool allocation log messages to display
sizes in KiB instead of MiB. Using MiB caused allocations less than
1 MiB to be logged as 0 MiB due to integer truncation. KiB provides
better precision for smaller memory regions specified in the Device Tree.
Signed-off-by: Vova Sharaienko <sharaienko@google.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Link: https://lore.kernel.org/r/20260629223759.2637162-1-sharaienko@google.com
restore_dm_crypt_keys_to_thread_keyring() gets a reference to the user
keyring before restoring the saved dm-crypt keys.
The same keyring reference is then passed to add_key_to_keyring() for each
saved key, but add_key_to_keyring() drops that reference on every call.
This is only balanced when exactly one key is restored. With multiple
keys, the keyring reference is dropped too many times and may trigger a
refcount underflow or use-after-free.
When more than five keys are restored, a refcount underflow/use-after-free
warning can be triggered.
The early error paths after lookup_user_key() also return without dropping
the keyring reference.
Keep ownership of the keyring reference in
restore_dm_crypt_keys_to_thread_keyring(), drop it once on all exit paths,
and make add_key_to_keyring() only use the reference without consuming it.
Fixes: 62f17d9df6 ("crash_dump: retrieve dm crypt keys in kdump kernel")
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Reviewed-and-tested-by: Coiby Xu <Coiby.Xu@gmail.com>
Acked-by: Baoquan He <baoquan.he@linux.dev>
Reviewed-by: Bradley Morgan <include@grrlz.net>
Link: https://patch.msgid.link/20260704112509.3717884-1-lgs201920130244@gmail.com
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
- Free field in error path of synthetic event parse
In __create_synth_event() the field was allocated but was not freed in the
error path.
- Fix ring_buffer_event_length() on 8 byte aligned architectures
On architectures with CONFIG_HAVE_64BIT_ALIGNED_ACCESS set to y,
the ring_buffer_event_length() may return the wrong size. This
is because archs with that config set will always use the "big
event meta header" as that is 8 bytes keeping the payload 8 bytes
aligned, even when a 4 byte header could hold the size of the
event.
But ring_buffer_event_length() doesn't take this into account and only
subtracts 4 bytes for the meta header in the length when it should have
subtracted 8 bytes.
- Have osnoise wait for a full rcu synchronization on unregister
osnoise_unregister_instance() used to call synchronize_rcu() before
freeing its copy of the instance but was switched to kfree_rcu().
The osniose tracer has code that traverses the instances that it
uses, and inst is just a pointer to that instance. By using kfree_rcu()
instead of synchronize_rcu(), the instance that the inst pointer is
pointing to can be freed while the osnoise code is still referencing it.
That is, a rmdir on an instance first unregisters the tracer. When the
unregister finishes, the rmdir expects that the tracer is finished with
the instance that it is using. By putting back the synchronize_rcu()
in osnoise_unregister_instance() the unregistering of osnoise will now
return when all the users of the instance have finished.
- Remove an unused setting of "ret" in tracing_set_tracer()
- Fix ring_buffer_read_page() copying events
The commit that changed ring_buffer_read_page() to show dropped events
from the buffer itself, split the "commit" variable between the commit
value (with flags) and "size" that holds the size of the sub-buffer.
A cut and paste error changed the test of the reading from checking the
size of the buffer to the size of the event causing reads to only read one
event at a time.
- Make tracepoint_printk a static variable
When the tracing sysctl knobs were move from sysctl.c to trace.c, the
variable tracepoint_printk no longer needed to be global. Make it static.
- Fix some typos
- Fix NULL pointer dereference in func_set_flag()
The flags update of the function tracer first checks if the value of the
flag is the same and exits if they are, and then it checks if the current
tracer is the function tracer and exits if it isn't. The problem is that
these checks need to be in a reversed order, as if the tracer isn't the
function tracer, then the flag being checked may not exist. Reverse the
order of these checks.
- Fix ufs core trace events to not dereference a pointer in TP_printk()
The TP_printk() part of the TRACE_EVENT() macro is called when the user
reads the "trace" file. This can be seconds, minutes, hours, days, weeks,
and even months after the data was recorded into the ring buffer. Thus,
saving a pointer to an object into the ring buffer and then dereferencing
it from TP_printk() can cause harm as the object the pointer is pointing
to may no longer exist.
Fix all the trace events in ufs core to save the device name in the ring
buffer instead of dereferencing the device descriptor from TP_printk().
- Prevent out-of-bound reads in glob matching of trace events
The filter logic of events allows simple glob logic to add wild cards to
filter on strings. But some events have fields that may not have a
terminating 'nul' character. This may cause the glob matching to go beyond
the string. Change the logic to always pass in the length of the field
that is being matched.
- Add no-rcu-check version of trace_##event##_enabled()
The trace_##event##_enabled() usually wraps trace events to do extra work
that is only needed when the trace event is enabled. But this can hide
events that are placed in locations where RCU is not watching, and can
make lockdep not see these bugs when the event is not enabled.
The trace_##event##_enabled() was updated to always test to make sure RCU
is watching to catch locations that may call events without RCU being
active.
This caused a false positive for the irq_disabled() and related events. As
that use trace_irq_disabled_enabled() to force RCU to be watching when the
event is enabled via the ct_irq_enter() function, calls the event, and
then calls ct_irq_exit() to put RCU back to its original state.
The trace_irq_disabled_enabled() should not trigger a warning when RCU is
not watching because the code within its block handles the case properly.
Make a __trace_##event##_enabled() version for this event to use that
doesn't check RCU is watching as it handles the case when it isn't.
- Fix use-after-free in user_event_mm_dup()
When the enabler is removed from the link list, it is freed immediately.
But it is protected via RCU and needs to be freed after an RCU
grace period. Use queue_rcu_work() so that the event_mutex can also
be taken as user_event_put() takes the mutex on the last reference
is released.
- Free type string in error path of parse_synth_field()
There's an error path in parse_synth_field() where the allocated type
string is not freed.
- Add selftest that tests deferred event teardown
- Fix leak in error path of trace_remote_alloc_buffer()
If page allocation fails, the desc->nr_cpus is not incremented for the
current CPU and the allocations done for it are not freed.
- Fix allocation length in trace_remote_alloc_buffer()
The logic to calculate the struct_len was doing a double count and setting
the value too large. Calculate the size upfront to fix the error and
simplify the logic.
- Fix sparse CPU masks in ring_buffer_desc()
If there are sparse CPUs (gaps in the numbering), the ring_buffer_desc()
will fail as it tests the CPU number against the number of CPUs that are
used.
-----BEGIN PGP SIGNATURE-----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=+//w
-----END PGP SIGNATURE-----
Merge tag 'trace-v7.2-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace
Pull tracing fixes from Steven Rostedt:
- Free field in error path of synthetic event parse
In __create_synth_event() the field was allocated but was not freed
in the error path
- Fix ring_buffer_event_length() on 8 byte aligned architectures
On architectures with CONFIG_HAVE_64BIT_ALIGNED_ACCESS set to y, the
ring_buffer_event_length() may return the wrong size. This is because
archs with that config set will always use the "big event meta
header" as that is 8 bytes keeping the payload 8 bytes aligned, even
when a 4 byte header could hold the size of the event
But ring_buffer_event_length() doesn't take this into account and
only subtracts 4 bytes for the meta header in the length when it
should have subtracted 8 bytes
- Have osnoise wait for a full rcu synchronization on unregister
osnoise_unregister_instance() used to call synchronize_rcu() before
freeing its copy of the instance but was switched to kfree_rcu(). The
osniose tracer has code that traverses the instances that it uses,
and inst is just a pointer to that instance. By using kfree_rcu()
instead of synchronize_rcu(), the instance that the inst pointer is
pointing to can be freed while the osnoise code is still referencing
it
That is, a rmdir on an instance first unregisters the tracer. When
the unregister finishes, the rmdir expects that the tracer is
finished with the instance that it is using. By putting back the
synchronize_rcu() in osnoise_unregister_instance() the unregistering
of osnoise will now return when all the users of the instance have
finished
- Remove an unused setting of "ret" in tracing_set_tracer()
- Fix ring_buffer_read_page() copying events
The commit that changed ring_buffer_read_page() to show dropped
events from the buffer itself, split the "commit" variable between
the commit value (with flags) and "size" that holds the size of the
sub-buffer. A cut and paste error changed the test of the reading
from checking the size of the buffer to the size of the event causing
reads to only read one event at a time
- Make tracepoint_printk a static variable
When the tracing sysctl knobs were move from sysctl.c to trace.c, the
variable tracepoint_printk no longer needed to be global. Make it
static
- Fix some typos
- Fix NULL pointer dereference in func_set_flag()
The flags update of the function tracer first checks if the value of
the flag is the same and exits if they are, and then it checks if the
current tracer is the function tracer and exits if it isn't. The
problem is that these checks need to be in a reversed order, as if
the tracer isn't the function tracer, then the flag being checked may
not exist. Reverse the order of these checks
- Fix ufs core trace events to not dereference a pointer in TP_printk()
The TP_printk() part of the TRACE_EVENT() macro is called when the
user reads the "trace" file. This can be seconds, minutes, hours,
days, weeks, and even months after the data was recorded into the
ring buffer. Thus, saving a pointer to an object into the ring buffer
and then dereferencing it from TP_printk() can cause harm as the
object the pointer is pointing to may no longer exist
Fix all the trace events in ufs core to save the device name in the
ring buffer instead of dereferencing the device descriptor from
TP_printk()
- Prevent out-of-bound reads in glob matching of trace events
The filter logic of events allows simple glob logic to add wild cards
to filter on strings. But some events have fields that may not have a
terminating 'nul' character. This may cause the glob matching to go
beyond the string. Change the logic to always pass in the length of
the field that is being matched
- Add no-rcu-check version of trace_##event##_enabled()
The trace_##event##_enabled() usually wraps trace events to do extra
work that is only needed when the trace event is enabled. But this
can hide events that are placed in locations where RCU is not
watching, and can make lockdep not see these bugs when the event is
not enabled
The trace_##event##_enabled() was updated to always test to make sure
RCU is watching to catch locations that may call events without RCU
being active
This caused a false positive for the irq_disabled() and related
events. As that use trace_irq_disabled_enabled() to force RCU to be
watching when the event is enabled via the ct_irq_enter() function,
calls the event, and then calls ct_irq_exit() to put RCU back to its
original state
The trace_irq_disabled_enabled() should not trigger a warning when
RCU is not watching because the code within its block handles the
case properly. Make a __trace_##event##_enabled() version for this
event to use that doesn't check RCU is watching as it handles the
case when it isn't
- Fix use-after-free in user_event_mm_dup()
When the enabler is removed from the link list, it is freed
immediately. But it is protected via RCU and needs to be freed after
an RCU grace period. Use queue_rcu_work() so that the event_mutex can
also be taken as user_event_put() takes the mutex on the last
reference is released
- Free type string in error path of parse_synth_field()
There's an error path in parse_synth_field() where the allocated type
string is not freed
- Add selftest that tests deferred event teardown
- Fix leak in error path of trace_remote_alloc_buffer()
If page allocation fails, the desc->nr_cpus is not incremented for
the current CPU and the allocations done for it are not freed
- Fix allocation length in trace_remote_alloc_buffer()
The logic to calculate the struct_len was doing a double count and
setting the value too large. Calculate the size upfront to fix the
error and simplify the logic
- Fix sparse CPU masks in ring_buffer_desc()
If there are sparse CPUs (gaps in the numbering), the
ring_buffer_desc() will fail as it tests the CPU number against the
number of CPUs that are used
* tag 'trace-v7.2-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
ring-buffer: Allow sparse CPU masks in ring_buffer_desc()
tracing/remotes: Fix struct_len in trace_remote_alloc_buffer()
tracing/remotes: Fix leak in trace_remote_alloc_buffer() error path
selftests/user_events: Wait for deferred event teardown after unregister
tracing/synthetic: Free type string on error path
tracing/user_events: Fix use-after-free in user_event_mm_dup()
tracing: Add a no-rcu-check version of trace_##event##_enabled()
tracing: Prevent out-of-bounds read in glob matching
ufs: core: tracing: Do not dereference pointers in TP_printk()
tracing: Fix NULL pointer dereference in func_set_flag()
samples: ftrace: Fix typos in benchmark comment
tracing: Make tracepoint_printk static as not exported
ring-buffer: Fix ring_buffer_read_page() copying only one event per page
tracing: Remove unused ret assignment in tracing_set_tracer()
tracing/osnoise: Call synchronize_rcu() when unregistering
ring-buffer: Fix event length with forced 8-byte alignment
tracing/synthetic: Free pending field on error path
The return value of __secure_computing() currently uses 0 to indicate
that a system call should be allowed, and -1 to indicate that it should
be blocked/killed. This 0/-1 pattern is non-intuitive for a security
check function and makes the control flow at the call sites less readable.
Furthermore, any potential future changes to these return values would
require a high-risk, error-prone audit of all its users across different
architectures.
Sanitize this logic by converting the return type of __secure_computing()
to a proper boolean, where 'true' explicitly means 'allow' and 'false'
means 'fail/deny'.
Update all the two dozen or so call sites across the tree to align with
this new boolean semantic. No functional changes are intended, as the
callers still return -1 to the lower-level assembly entry code upon
seccomp denial.
Rename the function to __seccomp_permit_syscall() so that the purpose is
entirely clear.
[ tglx: Rename the function ]
Suggested-by: Thomas Gleixner <tglx@kernel.org>
Suggested-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Mukesh Kumar Chaurasiya (IBM) <mkchauras@gmail.com>
Reviewed-by: Mukesh Kumar Chaurasiya (IBM) <mkchauras@gmail.com>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Link: https://patch.msgid.link/20260707190254.230735780@kernel.org
Since tracing_multi link does not set ftrace_managed, it would fail to
release the tracing_multi link when attaching tracing_multi link and
then attaching fentry link.
[ 3.714215] WARNING: kernel/bpf/trampoline.c:1727 at bpf_trampoline_multi_detach+0x20b/0x240, CPU#1: test_progs/97
...
[ 3.733170] bpf_tracing_multi_link_release+0x14/0x30
[ 3.733890] bpf_link_free+0x58/0x130
[ 3.734414] bpf_link_release+0x23/0x30
Fix it by setting 'ftrace_managed = true' in register_fentry_multi().
Fixes: aef4dfa790 ("bpf: Add bpf_trampoline_multi_attach/detach functions")
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20260711124822.29406-2-leon.hwang@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
bpf_cpumask_populate() writes to its destination with bitmap_copy(), but
the destination is typed as struct cpumask *. That allows the verifier to
accept borrowed cpumask pointers returned by read-only kfuncs, such as
scx_bpf_get_online_cpumask(), as a writable destination.
Make the destination a struct bpf_cpumask * so populate follows the same
ownership rule as the other mutating cpumask kfuncs. Query kfuncs continue
to accept const struct cpumask * inputs.
Fixes: 950ad93df2 ("bpf: add kfunc for populating cpumask bits")
Signed-off-by: Nicholas Dudar <main.kalliope@gmail.com>
Acked-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/20260709182800.2037938-2-main.kalliope@gmail.com
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
unearthed other races in the area (Thomas Gleixner)
Signed-off-by: Ingo Molnar <mingo@kernel.org>
-----BEGIN PGP SIGNATURE-----
iQJFBAABCgAvFiEEBpT5eoXrXCwVQwEKEnMQ0APhK1gFAmpSJ0gRHG1pbmdvQGtl
cm5lbC5vcmcACgkQEnMQ0APhK1hyKA//W5KqLrRgXjH9cS7tF6PLo9NhbZZn5FhH
jxThPvJosNyCj8UEVD+iCRjkqkNxiY9mzx3px2ZVlJg8K35ez2PeZ/zx0wvuSWlM
sD7k4S3JJQ7vAE4CkYvsf91TxCWqiOMw+NLyo7MgDhwsOdgPgUAAKlCdCyoFzRK7
rAV7FQWMQag24donwPiesk8bnU4pWrooq8BoY7XVCy6aHDwwPVxrs1yPyUEikIXq
AT2HiYFOu9ok6rSlJs+A2JQnmke6M6aIJ5dmUGR7gUDRAqlXesmM6jlwklzK0a4O
j6lpE05ouIhZpYzQVTDxvf8cytp0+rutTX8w0lPRlqxYkaeBVEpktxfBAoqFCsoe
GpK8hK0PM57W29zhnTygQib53QQAk9QBQetp0NwUjLSOu3ouhCAOC3wuoctJ4fKY
oNSKGPBhHwJN5932JuxlEll/xbo+bFyHfP87tqn1SA0bgKMSNet5q7+wYE+1gcTp
9LHQnoMpjyRBjuOupnvxqeOVbN3qlNdY6MINFwxBZuKG7UwmYpTH7l72fuUM/eNr
b4UHKgwv3vuQX/BfBoT1mLnzWQTltrfEywBCXHsEHUQ7KFhdRHEbnh3zVDZCT5S3
5N/RS8aFcryD/0/DhxWEhO2DBTgjDDvmOKplIPYFtikDI1dkjZ9dZ0BCI42u6wMH
by25mw2rF2s=
=2I60
-----END PGP SIGNATURE-----
Merge tag 'timers-urgent-2026-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer fix from Ingo Molnar:
- Fix a subtle posix-cpu-timers vs. exec() race, which
unearthed other races in the area (Thomas Gleixner)
* tag 'timers-urgent-2026-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
posix-cpu-timers: Prevent UAF caused by non-leader exec() race
-----BEGIN PGP SIGNATURE-----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=wY2k
-----END PGP SIGNATURE-----
Merge tag 'audit-pr-20260710' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/audit
Pull audit fixes from Paul Moore:
"Two relatively small audit patches to fix potential data races with
the main audit backlog queue as well as possible integer overflows
when logging data as hex strings"
* tag 'audit-pr-20260710' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/audit:
audit: fix potential integer overflow in audit_log_n_hex()
audit: Fix data races of skb_queue_len() readers on audit_queue
No user currently relies on sparse CPU masks, but the descriptor logic already
supports them via linear fallback. Remove the arbitrary limitation.
Link: https://patch.msgid.link/20260709160017.1729517-4-vdonnefort@google.com
Fixes: 2e67fabd8b ("ring-buffer: Introduce ring-buffer remotes")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Pre-calculate desc->struct_len up-front in trace_remote_alloc_buffer()
with trace_buffer_desc_size() to fix double-counting.
While at it, use the accessor __first_ring_buffer_desc().
Link: https://patch.msgid.link/20260709160017.1729517-3-vdonnefort@google.com
Fixes: 96e43537af ("tracing: Introduce trace remotes")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
If page allocation fails in trace_remote_alloc_buffer(), desc->nr_cpus
is not yet incremented for the current CPU. As a consequence, on error,
half-allocated rb_desc will not be freed in trace_remote_free_buffer().
Increment desc->nr_cpus as soon as the first allocation for the current
CPU has succeeded.
Link: https://patch.msgid.link/20260709160017.1729517-2-vdonnefort@google.com
Fixes: 96e43537af ("tracing: Introduce trace remotes")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When switching a task's sched_class away from sched_ext, we get the
following sequence of events in __sched_setscheduler():
sched_change_begin()
switched_from_scx()
scx_disable_task(p)
ops.disable(p)
__setscheduler_params()
set_load_weight()
reweight_task_scx(p)
ops.set_weight(p)
p->sched_class = next_class;
sched_change_end()
...
Notably, ops.set_weight() is called _after_ ops.disable().
This violates the expected semantics of the callbacks, the expectation
being that ops.disable() can only be followed by ops.exit_task() or
ops.enable().
Skipping the weight adjustment for disabled tasks should be harmless
since the weight will be recalculated in scx_enable_task() if the task
ever rejoins SCX.
Fixes: 637b068282 ("sched: Fold sched_class::switch{ing,ed}_{to,from}() into the change pattern")
Cc: stable@vger.kernel.org # v6.19+
Signed-off-by: Kuba Piecuch <jpiecuch@google.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
sched_ext schedulers have state in BPF programs and kernel. scx_dump
provides kernel state and BPF program state on error, but this is static
in what it can provide.
Add a sched_ext_exit tracepoint in scx_claim_exit() so that BPF programs
can dynamically inspect scheduler specific state at the moment of exit.
Pass the exiting scx_sched so attached programs can read its state, and,
since exits propagate through a hierarchy of sub-schedulers, identify
which scheduler each event belongs to.
Signed-off-by: Pat Somaru <patso@likewhatevs.io>
Signed-off-by: Tejun Heo <tj@kernel.org>
map_range() reads rb->aux_pages[], rb->aux_nr_pages and rb->aux_pgoff via
perf_mmap_to_page() while holding only event->mmap_mutex. Those fields are
serialized by rb->aux_mutex, and mmap_mutex is per event.
Thus, two events sharing one rb via PERF_EVENT_IOC_SET_OUTPUT can race
rb_alloc_aux() with map_range(), leading to a page-UAF scenario as follows:
CPU 0 CPU 1
===== =====
rb_alloc_aux() map_range()
[1]: allocate rb->aux_pages[0]
[2]: rb->aux_nr_pages++
[3]: perf_mmap_to_page()
returns rb->aux_pages[0]
[4]: map it as VM_PFNMAP
[5]: rb->aux_pgoff = 1
munmap the page
[6]: free rb->aux_pages[0]
Pages mapped as VM_PFNMAP have no refcount protection, so CPU 1 holds a
mapping to a freed physical frame.
Fix this by taking rb->aux_mutex across the page walk in map_range().
Fixes: b709eb872e ("perf: map pages in advance")
Signed-off-by: Lee Jia Jie <jiajie.lee@starlabs.sg>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: stable@vger.kernel.org
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Now that nothing inflates tick_length beyond tick_length_base (the
adjtime path was converted to use time_offset in the previous commit),
the two fields are always equal.
Remove tick_length_base and keep tick_length as the single field.
Remove the per-second reset and the delta update in
ntp_update_frequency() since there is no separate base to track.
No functional change intended.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Assisted-by: Kiro:claude-opus-4.6-1m
Link: https://patch.msgid.link/20260621220051.1030462-8-dwmw2@infradead.org
time_offset (the exponential PLL phase slew) and time_adjust (the
linear adjtime() slew) can be asked to move the clock in opposite
directions. second_overflow() folds only their *net* into the per-tick
skew_delta, so the cancelling overlap would never be drained from
either tracker by the per-tick code — and if they cancel exactly,
skew_delta is zero and neither converges at all.
Arguably we could just let one of them entirely cancel out the other
immediately, but that would be a change in userspace-visible behaviour.
Instead, preserve the existing behaviour by calculating the "conflict"
portion between the opposing skew each second, and transferring that
amount directly from one tracker to the other.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Assisted-by: Kiro:claude-opus-4.8
Link: https://patch.msgid.link/20260621220051.1030462-7-dwmw2@infradead.org
The legacy adjtime() slew (ADJ_OFFSET_SINGLESHOT) was the last user of
tick_length != tick_length_base: it slewed the clock by inflating
tick_length directly, which delivered the correction imprecisely (e.g.
delivering only 4997.5µs when asked for a 5ms skew).
Deliver it accurately through the same per-tick mechanism that is now
used for time_offset, allowing it to contribute to skew_delta and thus
drive the delivery through ntp_error and mult selection.
To allow for accurate accounting, store the sub-microsecond part of
time_adjust is separately, while keeping time_adjust in microseconds
as that's the external API.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Assisted-by: Kiro:claude-opus-4.8
Link: https://patch.msgid.link/20260621220051.1030462-6-dwmw2@infradead.org
Currently, the phase offset of time_offset and time_adjust is delivered
by adjusting tick_length in second_overflow(), and immediately draining
time_offset/time_adjust by the amount that the tick_length adjustment is
*estimated* to cause. This is fairly approximate, in part because it is
not always correct to assume that precisely NTP_INTERVAL_FREQ ticks will
occur between one call to second_overflow() and the next. It could also
over and under-run in the final second of delivery.
Instead of inflating tick_length, transfer the intended skew directly
into ntp_error each tick to achieve the desired rate.
In second_overflow(), calculate skew_delta which is the per-tick slew
rate, in the same units as time_offset: (ns << NTP_SCALE_SHIFT) / HZ.
In logarithmic_accumulation(), drain up to 'skew_delta' time units from
time_offset into ntp_error to drive the overall effective rate. The new
ntp_drain_skew() function returns the amount which is actually 'claimed'
by time_offset (and in a future patch, time_adjust). Any overrun which
is delivered by the changed 'mult' (as described below) but not claimed
by ntp_drain_skew() will remain in ntp_error to be corrected away in
subsequent ticks.
Simply transferring the precise amount from time_offset to ntp_error
would be sufficent to make the time *eventually* converge, however the
skew delivered is limited by the choice of { mult, mult+1 } each tick
and thus the convergence would be extremely slow.
In theory we could inflate ntp_err_mult with the magnitude of ntp_error
in the general case — but that would cause overcorrection in a tickless
kernel. Instead, in timekeeping_adjust(), take skew_delta into account
when calculating 'mult', such that the available {mult, mult+1} choices
bracket the overall effective rate *including* the skew, to avoid the
delta just building up in ntp_error.
The effect is that the inflated 'mult' causes ntp_error to grow because
xtime_interval is (e.g.) longer than the true tick_length. But then the
same delta is removed again as it's drained from time_offset.
This gives behaviour equivalent to the old tick_length += delta approach
but with exact per-tick accounting of the time_offset actually imparted
to the clock, and no overrun.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Assisted-by: Kiro:claude-opus-4.8
Link: https://patch.msgid.link/20260621220051.1030462-5-dwmw2@infradead.org
cycle_interval is an integer number of counter cycles per NTP interval,
so the real time it represents differs from the nominal
NTP_INTERVAL_LENGTH by up to half a counter period. For coarse
clocksources this is significant: the 3.579545 MHz ACPI PM timer at
HZ=1000 rounds 3579.545 cycles up to 3580, making each tick 1.000127 ms
(+127 PPM).
Commit a386b5af8e ("time: Compensate for rounding on odd-frequency
clocksources") introduced xtime_remainder to compensate for exactly
this, citing the same 127 PPM ACPI PM example. The compensation is
correct and necessary, but it was applied inside the timekeeping
accumulation in timekeeping.c: subtracted in the mult computation in
timekeeping_adjust() and folded into the ntp_error update in
logarithmic_accumulation(). That keeps the base rate correct and leaves
NTP its full symmetric +/-MAXFREQ range rather than +373/-627 PPM, but
the NTP code in ntp.c never sees it: tick_length is computed without the
correction, so ntp.c's notion of how long a tick is disagrees with the
rate timekeeping actually produces.
Make the offset an explicit part of the NTP tick_length instead. Add
ntp_data::cs_tick_adj, a fixed per-second addend that
ntp_update_frequency() includes alongside ntp_tick_adj and time_freq.
tk_setup_internals() computes it from the difference between the real
cycle_interval duration and the nominal interval, stores it in the
timekeeper, and hands it to NTP through a new argument to ntp_clear() --
which already recomputes the frequency and is invoked after every
clocksource (re)configuration. timekeeping_init() now uses TK_UPDATE_ALL
for this; clearing NTP there is otherwise redundant since ntp_init() has
just initialised it.
ntp.c now computes the true tick rate, giving a single source of truth.
Like ntp_tick_adj, cs_tick_adj stays internal to the kernel: userspace
still sees the nominal 1.000000 ms tick via adjtimex and is unaware of
the addends. timekeeping_adjust() and logarithmic_accumulation() use
ntp_tick / xtime_interval directly, and xtime_remainder is removed.
The base-rate arithmetic is unchanged: ntp_tick becomes
xtime_interval << ntp_error_shift, so the mult division yields the same
base mult and the ntp_error accumulation still nets to zero per tick.
Beyond the cleanup of treating all the tick_length contributions
(nominal interval, ntp_tick_adj, cs_tick_adj, time_freq) consistently
as addends in one place, it also prepares for feed-forward discipline:
a future timekeeping_set_reference() will set tick_length to track an
absolute external reference such as a vmclock, and that path needs
ntp.c to own a tick_length that already reflects the clocksource
quantisation, with no hidden correction applied elsewhere.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Assisted-by: Kiro:claude-opus-4.8
Acked-by: John Stultz <jstultz@google.com>
Link: https://patch.msgid.link/20260621220051.1030462-4-dwmw2@infradead.org
consume_remote_task() drops this_rq before unlinking a remote task from
its DSQ and locking src_rq. When called from ops.dispatch(),
scx_locked_rq() keeps pointing to this_rq across the lock dance. The
subsequent switch_rq_lock(src_rq, this_rq) cannot update the tracking
because its guard does not match, leaving the tracking stale while
src_rq is held.
Keep this_rq locked until the task has been unlinked and the DSQ lock
released, then use switch_rq_lock() to switch directly to src_rq. Use
the same helper to restore this_rq after losing the dequeue race. The
successful path already switches back through
move_remote_task_to_local_dsq(), so scx_locked_rq() now follows the
actually held rq throughout the consume path.
Suggested-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/all/455e701bca66bdecde530d225f4dba0a@kernel.org
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_bpf_kick_cid() returned an error code, but the value conveys nothing
actionable and no caller consumes it. The kick is asynchronous, so a
successful return only means it was queued. An invalid @cid is already
reported through scx_error() by scx_cid_to_cpu(), and a missing scheduler
leaves nothing to kick.
Make scx_bpf_kick_cid() return void to match scx_bpf_kick_cpu(). The
cid-form kfuncs are not in practical use yet, so the ABI change is safe.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Direct writes to p->scx.slice and p->scx.dsq_vtime bypass
scx_bpf_task_set_slice/dsq_vtime() and the authority checks they carry.
Those checks exist for sub-schedulers, which attach only through the
cid-form struct_ops, so the direct writes only need to be closed there.
Give sched_ext_ops_cid its own verifier ops that reject the two fields.
cid-form is a new interface with no legacy users, so there is no
compatibility to keep. The cpu-form keeps direct writes, and the deprecation
warning they carried is dropped.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_rq.extra_enq_flags carries scx-specific enqueue flags across the
activate_task() boundary during a cross-rq task move in
move_remote_task_to_local_dsq(). Rename it to remote_activate_enq_flags to
name that role, and fix the stale comment that referenced the old
move_task_to_local_dsq() name.
No functional change.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The cid-form ops overlay their cpu-form siblings at the same struct slot.
Ops whose signature matches the sibling are invoked through the cpu-form
call sites unchanged, but set_cmask() takes an arena cmask address rather
than a cpumask, so scx_call_op_set_cpumask() calls ops_cid.set_cmask()
directly and hand-rolled the kf_tasks[] and locked_rq bracket that
SCX_CALL_OP_TASK() provides. The hand-rolled bracket reset locked_rq to
NULL on exit instead of restoring the saved value, so a nested call would
clobber the outer op's locked-rq tracking.
Parameterize the dispatch macros by the ops-table member and add
SCX_CALL_CID_OP_TASK(), which routes through sch->ops_cid. Convert
scx_call_op_set_cpumask() to it and drop the hand-rolled bracket. The only
behavioral change is that locked_rq is now saved and restored like every
other op call site.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The cmask ops can operate on BPF-arena cmasks which BPF programs may read
and write concurrently. The _RACY op variants existed to make such lockless
reads explicit but this turned out to be too restrictive. Mark the word
accesses in all the two-cmask ops with READ_ONCE/WRITE_ONCE instead and drop
the _RACY variants.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Pull to receive:
db4e9defd2 ("sched_ext: Record an error on errno-only sub-enable failure")
49b3378a75 ("sched_ext: Fix premature ops->priv publication in scx_alloc_and_add_sched()")
e6979d05c6 ("tools/sched_ext: scx - Fix cmask_subset(), cmask_equal() and cmask_weight()")
for further sub-sched changes and to resolve the conflicts with the
sub-sched updates on for-7.3.
db4e9defd2 adds scx_error() to the sub-enable err_disable sink which
for-7.3 moved from ext.c into sub.c. Resolved by applying the fix to
scx_sub_enable_workfn() in sub.c.
49b3378a75 drops RCU_INIT_POINTER() from an scx_alloc_and_add_sched()
unwind label whose body changed with for-7.3's stall_cpus addition.
Resolved by dropping the line from the updated unwind.
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_alloc_and_add_sched() publishes @sch through ops->priv before allocating
the cgroup path. If that allocation fails, the unwind path clears ops->priv
and frees @sch immediately. scx_prog_sched() callers can dereference
ops->priv from RCU context the moment it is set, so freeing without a grace
period can use-after-free a concurrent kfunc caller.
Move the publication below the cgroup path allocation so that every failure
path after publication frees @sch through kobject_put(), whose release path
defers the freeing by a grace period.
Fixes: 105dcd005b ("sched_ext: Introduce scx_prog_sched()")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_sub_enable_workfn() has several failure paths that only return an errno
(e.g. -ENOMEM from an allocation) and jump to err_disable without calling
scx_error(). scx_flush_disable_work() runs the disable, and thus ops.exit(),
only when an error has been recorded, so an errno-only failure leaves the
half-initialized sub-scheduler linked.
Record an error at the err_disable sink so every errno-only failure runs the
disable path.
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The comment for scx_bpf_dsq_insert() references "@dsp_id" in the
description body, but the parameter is named "@dsq_id" in both the
parameter list and the function signature.
Signed-off-by: Liang Luo <luoliang@kylinos.cn>
Acked-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
check_kfunc_args() detects a kfunc argument named rdonly_buf_size or
rdwr_buf_size and stores reg->var_off.value into meta->r0_size, a u64,
and does not bound it. check_kfunc_call() later copies that value into
the returned register's mem_size field:
meta->r0_size = reg->var_off.value;
...
regs[BPF_REG_0].mem_size = meta.r0_size;
regs[BPF_REG_0].mem_size is u32. A constant whose upper 32 bits are set
gets truncated instead of causing a load-time rejection, so the verifier
records a PTR_TO_MEM register with an approximately 4 GiB mem_size for
whatever allocation the kfunc returned. A later access check against
that register uses the truncated, wrong bound.
Reject rdonly_buf_size/rdwr_buf_size values that exceed U32_MAX at the
point meta->r0_size is set.
Fixes: eb1f7f71c1 ("bpf/verifier: allow kfunc to return an allocated mem")
Signed-off-by: Nicholas Dudar <main.kalliope@gmail.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260709155837.1879230-2-main.kalliope@gmail.com
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
