linux/tools
David Vernet 6c831c4684 bpf: Treat KF_RELEASE kfuncs as KF_TRUSTED_ARGS
KF_RELEASE kfuncs are not currently treated as having KF_TRUSTED_ARGS,
even though they have a superset of the requirements of KF_TRUSTED_ARGS.
Like KF_TRUSTED_ARGS, KF_RELEASE kfuncs require a 0-offset argument, and
don't allow NULL-able arguments. Unlike KF_TRUSTED_ARGS which require
_either_ an argument with ref_obj_id > 0, _or_ (ref->type &
BPF_REG_TRUSTED_MODIFIERS) (and no unsafe modifiers allowed), KF_RELEASE
only allows for ref_obj_id > 0.  Because KF_RELEASE today doesn't
automatically imply KF_TRUSTED_ARGS, some of these requirements are
enforced in different ways that can make the behavior of the verifier
feel unpredictable. For example, a KF_RELEASE kfunc with a NULL-able
argument will currently fail in the verifier with a message like, "arg#0
is ptr_or_null_ expected ptr_ or socket" rather than "Possibly NULL
pointer passed to trusted arg0". Our intention is the same, but the
semantics are different due to implemenetation details that kfunc authors
and BPF program writers should not need to care about.

Let's make the behavior of the verifier more consistent and intuitive by
having KF_RELEASE kfuncs imply the presence of KF_TRUSTED_ARGS. Our
eventual goal is to have all kfuncs assume KF_TRUSTED_ARGS by default
anyways, so this takes us a step in that direction.

Note that it does not make sense to assume KF_TRUSTED_ARGS for all
KF_ACQUIRE kfuncs. KF_ACQUIRE kfuncs can have looser semantics than
KF_RELEASE, with e.g. KF_RCU | KF_RET_NULL. We may want to have
KF_ACQUIRE imply KF_TRUSTED_ARGS _unless_ KF_RCU is specified, but that
can be left to another patch set, and there are no such subtleties to
address for KF_RELEASE.

Signed-off-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/r/20230325213144.486885-4-void@manifault.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2023-03-25 16:56:22 -07:00
..
accounting tools/accounting/procacct: remove some unused variables 2022-11-18 13:55:09 -08:00
arch bpf-next-for-netdev 2023-03-06 20:36:39 -08:00
bootconfig tools/bootconfig: fix single & used for logical condition 2023-02-22 08:27:35 +09:00
bpf tools/resolve_btfids: Add /libsubcmd to .gitignore 2023-03-06 16:23:01 +01:00
build tools build: Add test echo-cmd 2023-02-03 13:54:22 -03:00
certs
cgroup tools:cgroup:memcg_shrinker remove redundant import 2023-01-18 17:12:59 -08:00
counter
debugging
edid
firewire
firmware
gpio tools: gpio: fix -c option of gpio-event-mon 2023-01-27 14:05:46 +01:00
hv
iio tools/iio/iio_utils:fix memory leak 2023-01-21 17:52:26 +00:00
include bpf: Update the struct_ops of a bpf_link. 2023-03-22 22:53:02 -07:00
io_uring
kvm/kvm_stat tools/kvm_stat: update exit reasons for vmx/svm/aarch64/userspace 2022-11-09 12:26:52 -05:00
laptop
leds
lib libbpf: Use .struct_ops.link section to indicate a struct_ops with a link. 2023-03-22 22:53:02 -07:00
memory-model tools: memory-model: Make plain accesses carry dependencies 2023-01-03 20:47:04 -08:00
mm tools/mm: allow users to provide additional cflags/ldflags 2023-02-02 22:33:24 -08:00
net/ynl tools: net: add __pycache__ to gitignore 2023-02-24 11:55:47 -08:00
objtool - Daniel Verkamp has contributed a memfd series ("mm/memfd: add 2023-02-23 17:09:35 -08:00
pci
pcmcia
perf - Daniel Verkamp has contributed a memfd series ("mm/memfd: add 2023-02-23 17:09:35 -08:00
power platform-drivers-x86 for v6.3-1 2023-02-21 17:32:50 -08:00
rcu
scripts selftests/bpf: Fix cross compilation with CLANG_CROSS_FLAGS 2023-02-22 13:04:55 -08:00
spi
testing bpf: Treat KF_RELEASE kfuncs as KF_TRUSTED_ARGS 2023-03-25 16:56:22 -07:00
thermal
time
tracing tracing updates for 6.3: 2023-02-23 10:20:49 -08:00
usb tools: usb: ffs-aio-example: Fix build error with aarch64-*-gnu-gcc toolchain(s) 2022-11-09 12:37:56 +01:00
verification tools/rv: Remove unneeded semicolon 2023-01-31 19:21:12 -05:00
virtio tools/virtio: enable to build with retpoline 2023-02-20 19:26:59 -05:00
wmi
Makefile