amdgpu_benchmark_do_move() can exit the loop early if
amdgpu_copy_buffer() or dma_fence_wait() fails.
In the error path, the function jumps to the exit label
without releasing adev->mman.default_entity.lock, which
leaves the mutex held and results in a lock imbalance.
This can block subsequent users of default_entity and
potentially cause deadlocks.
Move the mutex_unlock() to the common exit path so the
lock is released on both success and error returns.
This fixes:
drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c:57 amdgpu_benchmark_do_move()
warn: inconsistent returns '&adev->mman.default_entity.lock'.
v2:
- Drop unrelated initialization of 'r'
- Keep the change focused on the mutex imbalance fix (Pierre).
v3:
- Removed empty line
Fixes: 30f2daedf4 ("drm/amdgpu: add missing lock in amdgpu_benchmark_do_move")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Cc: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
pp_features sorting uses int8_t sort_feature[] to store driver
feature enum indices. On newer ASICs the enum index can exceed 127,
causing signed overflow and silently dropping entries from the output.
Switch the array to int16_t so all enum indices are preserved.
Signed-off-by: Asad Kamal <asad.kamal@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add PMFW info to dmesg log for SMUv13 SOCs. It's helpful as diagnostic
data for some driver load issues.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Asad Kamal <asad.kamal@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The reason the RAP is not granting access to 0x58200 is that
a dedicated RSMU slot would have to be spent for this address range,
and MPASP is close to running out of RSMU slots.
This will help to fix PSP TOC load failure during secureboot.
GFX Driver Need to use indirect access for SMN address regs.
Signed-off-by: sguttula <suresh.guttula@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Now that it is guaranteed there can be no entity if there is no hw ip
block we can remove the open coded protection during CS parsing.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
References: 55414ad5c9 ("drm/amdgpu: error out on entity with no run queue")
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Currently there are two different behaviour modes when userspace tries to
operate on not present HW IP blocks. On a machine without UVD, VCE and VPE
blocks, this can be observed for example like this:
$ sudo ./amd_fuzzing --r cs-wait-fuzzing
...
amd_cs_wait_fuzzing DRM_IOCTL_AMDGPU_CTX r 0
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_GFX r 0
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_COMPUTE r 0
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_DMA r 0
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_UVD r -1
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_VCE r 0
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_UVD_ENC r -1
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_VCN_DEC r 0
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_VCN_ENC r 0
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_VCN_JPEG r 0
amd_cs_wait_fuzzing AMDGPU_WAIT_CS AMD_IP_VPE r 0
We can see that UVD returns an errno (-EINVAL) from the CS_WAIT ioctl,
while VCE and VPE return unexpected successes.
The difference stems from the fact the UVD is a load balancing engine
which retains the context, so with a workaround implemented in
amdgpu_ctx_init_entity(), but which does not account for the fact hardware
block may not be present.
This causes a single NULL scheduler to be passed to
drm_sched_entity_init(), which immediately rejects this with -EINVAL.
The not present VCE and VPE cases on the other hand pass zero schedulers
to drm_sched_entity_init(), which is explicitly allowed and results in
unusable entities.
As the UVD case however shows, call paths can handle the errors, so we can
consolidate this into a single path which will always return -EINVAL if
the HW IP block is not present.
We do this by rejecting it early and not calling drm_sched_entity_init()
when there is no backing hardware.
This also removes the need for the drm_sched_entity_init() to handle the
zero schedulers and NULL scheduler cases, which means that we can follow
up by removing the special casing from the DRM scheduler.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
References: f34e8bb7d6 ("drm/sched: fix null-ptr-deref in init entity")
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
To avoid race condition and avoid UAF cases, implement kref
based queues and protect the below operations using xa lock
a. Getting a queue from xarray
b. Increment/Decrement it's refcount
Every time some one want to access a queue, always get via
amdgpu_userq_get to make sure we have locks in place and get
the object if active.
A userqueue is destroyed on the last refcount is dropped which
typically would be via IOCTL or during fini.
v2: Add the missing drop in one the condition in the signal ioclt [Alex]
v3: remove the queue from the xarray first in the free queue ioctl path
[Christian]
- Pass queue to the amdgpu_userq_put directly.
- make amdgpu_userq_put xa_lock free since we are doing put for each get
only and final put is done via destroy and we remove the queue from xa
with lock.
- use userq_put in fini too so cleanup is done fully.
v4: Use xa_erase directly rather than doing load and erase in free
ioctl. Also remove some of the error logs which could be exploited
by the user to flood the logs [Christian]
Signed-off-by: Sunil Khatri <sunil.khatri@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Replace non-atomic vm->process_info assignment with cmpxchg()
to prevent race when parent/child processes sharing a drm_file
both try to acquire the same VM after fork().
Reviewed-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alysa Liu <Alysa.Liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v1:
The metrics->EnergyAccumulator field has been deprecated on newer pmfw.
v2:
add smu 13.0.0/13.0.7/13.0.10 support.
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
adapt sync info func for pmfw eeprom
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The scratch aperture or gmc private aperture in flat memory contains
57 bits of data on gfx v12.1.0 compared to the 32 bits from previous.
Add new helper kfd_init_apertures_v12 for gfx version >= v12.1.0 which
supports 57-bit VA space.
v2:
- update pdd->scratch_limit (Yu, Lang)
- update fixes tag (Felix Kuehling)
- add helper kfd_init_apertures_v12
Fixes: db1882b3ff ("drm/amdkfd: Update LDS, Scratch base for 57bit address")
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Lang Yu <lang.yu@amd.com>
Acked-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
read bad page data from pmfw eeprom when retirement
is triggered, use timestamp read from eeprom
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add read func for pmfw eeprom, and adapt address converting
for bad pages loaded from pmfw eeprom
v2: change label 'Out' to 'out'
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add a new IPID parse interface for umc, so we can
implement it for each ASIC, and so we can call it
in other blocks
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
There is a statement that has a ;; at the end, remove the extraneous ;
and remove extra tabs in the code block.
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Reviewed-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use get_smn_base interface to get IP die instance's base offset in
aqua_vanjaram. encode_ext_smn_addressing is not used.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Clang warns that comparing a __u16 value against 65536 is always false.
num_syncobj_handles is defined as __u16 in both the userq signal and
wait ioctl argument structs, so it can never exceed 65535. The checks
against AMDGPU_USERQ_MAX_HANDLES are therefore redundant and trigger
-Wtautological-constant-out-of-range-compare.
Fixes: Clang -Wtautological-constant-out-of-range-compare in userq
signal/wait ioctls
Fixes: d8e760b799 ("drm/amdgpu: update type for num_syncobj_handles in drm_amdgpu_userq_signal")
Fixes: c561d23204 ("drm/amdgpu: update type for num_syncobj_handles in drm_amdgpu_userq_wait")
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add wrapper funcs for pmfw eeprom interface to make them
easier to be called
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add flag to indicate if pmfw eeprom is supported or
not, and initialize it
v2: change copyright from 2025 to 2026
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add smu interfaces and its data structures for
pmfw eeprom in uniras
v2: add 'const' to smu messages array, and specify
index for each member when initializing.
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add amdgpu_smu_ras_feature_is_enabled to query one feature
is supported or not
v2: change default return value from -EOPNOTSUPP to 0
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add pmfw eeprom related messages into smu_v13_0_6_ras_send_msg
v2: add sriov check before sending smu commands
Signed-off-by: Gangliang Xie <ganglxie@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Make eeprom data and its counter consistent.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Handle RAS eeprom record when UMC_CHANNEL_IDX_V2 is set.
v2: get UMC_CHANNEL_IDX_V2 flag before the clear of it.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
update the type for num_syncobj_handles from __u32 to _u16 with
required padding.
This breaks the UAPI for big-endian platforms but this is deliberate
and harmless since userqueues is still a beta feature. It is enabled
via module parameter and need the right fw support to work.
Signed-off-by: Sunil Khatri <sunil.khatri@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
update the type for num_syncobj_handles from __u64 to _u16 with
required padding.
This breaks the UAPI for big-endian platforms but this is deliberate
and harmless since userqueues is still a beta feature. It is enabled
via module parameter and need the right fw support to work.
Signed-off-by: Sunil Khatri <sunil.khatri@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
queue id always remain a positive value and should
be of type unsigned.
With this we also dont need to typecast the id to other
types specially in xarray functions.
Signed-off-by: Sunil Khatri <sunil.khatri@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This version brings along the following updates:
- Prevent integer overflow when mhz to khz
- Remove always-false branches
- Remove redundant initializers
- Silence unused variable warning
- Initialize replay_state to PR_STATE_INVALID
- Fallback to boot snapshot for dispclk
- Skip cursor cache reset if hubp powergating is disabled
Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Taimur Hassan <Syed.Hassan@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Tested-by: Dan Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since the PAGE_SIZE is 8KB on sparc64, the size of
structure amdsriov_ras_telemetry will exceed 64KB,
so use absolute value to fix the buffer size.
Fixes the issue:
drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:522:2: error: static
assertion failed due to requirement 'sizeof(struct
amdsriov_ras_telemetry) <= 64 << 10': amdsriov_ras_telemetry must be 64 KB
| sizeof(struct amdsriov_ras_telemetry) <=
AMD_SRIOV_MSG_RAS_TELEMETRY_SIZE_KB_V1 << 10,
drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:522:40: note:
expression evaluates to '115616 <= 65536'
| sizeof(struct amdsriov_ras_telemetry) <=
AMD_SRIOV_MSG_RAS_TELEMETRY_SIZE_KB_V1 << 10,
Fixes: cb48a6b2b6 ("drm/amd/ras: use dedicated memory as vf ras command buffer")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202602261700.rVOLIw4l-lkp@intel.com/
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHAT]
Cast to long long before multiplication to prevent overflow
when converting mhz to khz by multiplying by 1000.
This is reported as INTEGER_OVERFLOW errors by Coverity.
Reviewed-by: Roman Li <roman.li@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Tested-by: Dan Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHAT]
program_prealpha_dealpha and hpo_frl_stream_enc_acquired are always
false and all branches depending on them will never be taken.
This is reported as DEADCODE errors by Coverity.
Reviewed-by: Roman Li <roman.li@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Tested-by: Dan Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHAT]
Remove unnecessary default value assignments for variables that
are unconditionally assigned before use.
Linux kernel code style prefers no assignments during initialization
when variables are assigned unconditionally as they can obscures
the actual data flow. In addition, compilers will be able to catch them
if variables are used without being updated later in all conditions.
This is reported as UNUSED_VALUE errors by Coverity.
Reviewed-by: Roman Li <roman.li@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Tested-by: Dan Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
