GFX V11 has GC block as default off IP.
Every time AMDGPU driver sends a request to PMFW
to unload MP1, PMFW will put GC in reset and
power down the voltage.Hence, skipping reset
for APUs with GFX V11 or later to avoid reset
related failures.
Fixes: 34355e6183 ("drm/amdgpu: Fix GFX hang on SteamDeck when amdgpu is reloaded")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Shubhankar Milind Sardeshpande <Shubhankar.MilindSardeshpande@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit d0a8cadffc818f51d05bc234d8da1af228bc59a3)
Cc: stable@vger.kernel.org
For triggering the dpc event with a single device, we still need
to set the in_link_reset flag and the dpc status.
Signed-off-by: Ce Sun <cesun102@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The correct fix is to tie the global xa entry lifetime to the
queue lifetime: insert in amdgpu_userq_create() and erase in
amdgpu_userq_cleanup(), both at the well-defined doorbell_index key,
making the operation O(1) and resolve the fence driver UAF problem
by binding the userq driver fence to per queue.
v2: clean up the local variables initialization. (Christian)
Signed-off-by: Prike Liang <Prike.Liang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch continues the effort to refactor workqueue APIs, which has begun
with the changes introducing new workqueues and a new alloc_workqueue flag:
commit 128ea9f6cc ("workqueue: Add system_percpu_wq and system_dfl_wq")
commit 930c2ea566 ("workqueue: Add new WQ_PERCPU flag")
The point of the refactoring is to eventually alter the default behavior of
workqueues to become unbound by default so that their workload placement is
optimized by the scheduler.
Before that to happen after a careful review and conversion of each individual
case, workqueue users must be converted to the better named new workqueues with
no intended behaviour changes:
system_wq -> system_percpu_wq
system_unbound_wq -> system_dfl_wq
This way the old obsolete workqueues (system_wq, system_unbound_wq) can be
removed in the future.
Suggested-by: Tejun Heo <tj@kernel.org>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch continues the effort to refactor workqueue APIs, which has begun
with the changes introducing new workqueues and a new alloc_workqueue flag:
commit 128ea9f6cc ("workqueue: Add system_percpu_wq and system_dfl_wq")
commit 930c2ea566 ("workqueue: Add new WQ_PERCPU flag")
The point of the refactoring is to eventually alter the default behavior of
workqueues to become unbound by default so that their workload placement is
optimized by the scheduler.
Before that to happen after a careful review and conversion of each individual
case, workqueue users must be converted to the better named new workqueues with
no intended behaviour changes:
system_wq -> system_percpu_wq
system_unbound_wq -> system_dfl_wq
This way the old obsolete workqueues (system_wq, system_unbound_wq) can be
removed in the future.
Suggested-by: Tejun Heo <tj@kernel.org>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In amdgpu_device_fini_hw(), deferred coredump formatting work may still
be pending when hardware and IP components are being torn down. Since
the work may access device registers and memory that will be freed or
powered off, it must be completed before proceeding.
Add a flush_work() call for adev->coredump_work, guarded by
CONFIG_DEV_COREDUMP, to ensure any pending coredump work finishes
before the device enters the early IP fini stage.
This avoids potential use-after-free or accessing hardware resources
that are no longer available.
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Suggested-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Jesse Zhang <jesse.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu_device_get_job_timeout_settings() passes a pointer directly
to the global amdgpu_lockup_timeout[] buffer into strsep().
strsep() destructively replaces delimiter characters with '\0'
in-place.
On multi-GPU systems, this function is called once per device.
When a multi-value setting like "0,0,0,-1" is used, the first
GPU's call transforms the global buffer into "0\00\00\0-1". The
second GPU then sees only "0" (terminated at the first '\0'),
parses a single value, hits the single-value fallthrough
(index == 1), and applies timeout=0 to all rings — causing
immediate false job timeouts.
Fix this by copying into a stack-local array before calling
strsep(), so the global module parameter buffer remains intact
across calls. The buffer is AMDGPU_MAX_TIMEOUT_PARAM_LENGTH
(256) bytes, which is safe for the stack.
v2: wrap commit message to 72 columns, add Assisted-by tag.
v3: use stack array with strscpy() instead of kstrdup()/kfree()
to avoid unnecessary heap allocation (Christian).
This patch was developed with assistance from Claude (claude-opus-4-6).
Assisted-by: Claude:claude-opus-4-6
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Ruijing Dong <ruijing.dong@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Update the way drm_coredump_printer is used based on its documentation
and Xe's code: the main idea is to generate the final version in one go
and then use memcpy to return the chunks requested by the caller of
amdgpu_devcoredump_read.
The generation is moved to a separate worker thread.
This cuts the time to copy the dump from 40s to ~0s on my machine.
---
v3:
- removed adev->coredump_in_progress and instead use work as
the synchronisation mechanism
- use kvfree instead of kfree
---
Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If discovery has failed for any reason (such as no support for a block)
then there is no need to unwind all the IP blocks in fini. In this
condition there can actually be failures during the unwind too.
Reset num_ip_blocks to zero during failure path and skip the unnecessary
cleanup path.
Suggested-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
I found a few more paths that cleanup fails due to a NULL version pointer
on unsupported hardware.
Add NULL checks as applicable.
Fixes: 39fc2bc4da ("drm/amdgpu: Protect GPU register accesses in powergated state in some paths")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When GPU initialization fails due to an unsupported HW block
IP blocks may have a NULL version pointer. During cleanup in
amdgpu_device_fini_hw, the code calls amdgpu_device_set_pg_state and
amdgpu_device_set_cg_state which iterate over all IP blocks and access
adev->ip_blocks[i].version without NULL checks, leading to a kernel
NULL pointer dereference.
Add NULL checks for adev->ip_blocks[i].version in both
amdgpu_device_set_cg_state and amdgpu_device_set_pg_state to prevent
dereferencing NULL pointers during GPU teardown when initialization has
failed.
Fixes: 39fc2bc4da ("drm/amdgpu: Protect GPU register accesses in powergated state in some paths")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If the device has not recovered after slot reset is called, it goes to
out label for error handling. There it could make decision based on
uninitialized hive pointer and could result in accessing an uninitialized
list.
Initialize the list and hive properly so that it handles the error
situation and also releases the reset domain lock which is acquired
during error_detected callback.
Fixes: 732c6cefc1 ("drm/amdgpu: Replace tmp_adev with hive in amdgpu_pci_slot_reset")
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Ce Sun <cesun102@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Now that DC supports external DP bridge encoders,
it has reached feature parity with the legacy non-DC display
driver on CIK APUs: Kaveri, Kabini, Mullins.
Use the DC display driver by default on SI APUs, unless it is
explicitly disabled using the amdgpu.dc=0 module parameter.
DC brings proper support for DP/HDMI audio, DP MST, VRR,
10-bit colors, some HDR features, atomic modesetting, etc.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We only want to stop the work queues, not mess with the
fences, etc.
v2: add the job back to the pending list.
v3: return the proper job status so scheduler adds the
job back to the pending list
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This was done entirely with mindless brute force, using
git grep -l '\<k[vmz]*alloc_objs*(.*, GFP_KERNEL)' |
xargs sed -i 's/\(alloc_objs*(.*\), GFP_KERNEL)/\1)/'
to convert the new alloc_obj() users that had a simple GFP_KERNEL
argument to just drop that argument.
Note that due to the extreme simplicity of the scripting, any slightly
more complex cases spread over multiple lines would not be triggered:
they definitely exist, but this covers the vast bulk of the cases, and
the resulting diff is also then easier to check automatically.
For the same reason the 'flex' versions will be done as a separate
conversion.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is the result of running the Coccinelle script from
scripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to
avoid scalar types (which need careful case-by-case checking), and
instead replace kmalloc-family calls that allocate struct or union
object instances:
Single allocations: kmalloc(sizeof(TYPE), ...)
are replaced with: kmalloc_obj(TYPE, ...)
Array allocations: kmalloc_array(COUNT, sizeof(TYPE), ...)
are replaced with: kmalloc_objs(TYPE, COUNT, ...)
Flex array allocations: kmalloc(struct_size(PTR, FAM, COUNT), ...)
are replaced with: kmalloc_flex(*PTR, FAM, COUNT, ...)
(where TYPE may also be *VAR)
The resulting allocations no longer return "void *", instead returning
"TYPE *".
Signed-off-by: Kees Cook <kees@kernel.org>
xgmi support atomics between links. Set them to true. This only set for
GFX12 onwards to avoid regression on older generations
v2: Use correct xgmi flag that indicates CPU connection
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Send unload command to smu during modprobe -r amdgpu for smu 13/14.
1. This can fix the high voltage/temperatue issue after driver is unloaded.
2. Reloading driver could fail but with the debug port based mode1 reset
during driver is reloaded, it is good and safe.
Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The commit 6a23e7b433 ("drm/amd: Clean up kfd node on surprise
disconnect") introduced early KFD cleanup when drm_dev_is_unplugged()
returns true. However, this causes hangs during normal module unload
(rmmod amdgpu).
The issue occurs because drm_dev_unplug() is called in amdgpu_pci_remove()
for all removal scenarios, not just surprise disconnects. This was done
intentionally in commit 39934d3ed5 ("Revert "drm/amdgpu: TA unload
messages are not actually sent to psp when amdgpu is uninstalled"") to
fix IGT PCI software unplug test failures. As a result,
drm_dev_is_unplugged() returns true even during normal module unload,
triggering the early KFD cleanup inappropriately.
The correct check should distinguish between:
- Actual surprise disconnect (eGPU unplugged): pci_dev_is_disconnected()
returns true
- Normal module unload (rmmod): pci_dev_is_disconnected() returns false
Replace drm_dev_is_unplugged() with pci_dev_is_disconnected() to ensure
the early cleanup only happens during true hardware disconnect events.
Cc: stable@vger.kernel.org
Reported-by: Cal Peake <cp@absolutedigital.net>
Closes: https://lore.kernel.org/all/b0c22deb-c0fa-3343-33cf-fd9a77d7db99@absolutedigital.net/
Fixes: 6a23e7b433 ("drm/amd: Clean up kfd node on surprise disconnect")
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add a full memory barrier after clearing no_hw_access in
amdgpu_device_mode1_reset() so subsequent PCI state restore
access cannot observe stale state on other CPUs.
Fixes: 7edb503fe4 ("drm/amd/pm: Disable MMIO access during SMU Mode 1 reset")
Signed-off-by: Perry Yuan <perry.yuan@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When an eGPU is unplugged the KFD topology should also be destroyed
for that GPU. This never happens because the fini_sw callbacks never
get to run. Run them manually before calling amdgpu_device_ip_fini_early()
when a device has already been disconnected.
This location is intentionally chosen to make sure that the kfd locking
refcount doesn't get incremented unintentionally.
Cc: kent.russell@amd.com
Closes: https://community.frame.work/t/amd-egpu-on-linux/8691/33
Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
Reviewed-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 6a23e7b433)
Cc: stable@vger.kernel.org
This patch allows kfd driver function correctly when AMD gpu devices got
unplug/replug at run time.
When an AMD gpu device got unplug kfd driver gracefully terminates existing
kfd processes after stops all queues by sending SIGBUS to user process. After
that user space can still use remaining AMD gpu devices. When all AMD gpu
devices at system got removed kfd driver will not response new requests.
Unplugged AMD gpu devices can be re-plugged. kfd driver will use added devices
to function as usual.
The purpose of this patch is having kfd driver behavior as expected during and
after AMD gpu devices unplug/replug at run time.
Signed-off-by: Xiaogang Chen <Xiaogang.Chen@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When an eGPU is unplugged the KFD topology should also be destroyed
for that GPU. This never happens because the fini_sw callbacks never
get to run. Run them manually before calling amdgpu_device_ip_fini_early()
when a device has already been disconnected.
This location is intentionally chosen to make sure that the kfd locking
refcount doesn't get incremented unintentionally.
Cc: kent.russell@amd.com
Closes: https://community.frame.work/t/amd-egpu-on-linux/8691/33
Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
Reviewed-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Skipping power ungate exposed some scenarios that will fail
like below:
```
amdgpu: Register(0) [regVPEC_QUEUE_RESET_REQ] failed to reach value 0x00000000 != 0x00000001n
amdgpu 0000:c1:00.0: amdgpu: VPE queue reset failed
...
amdgpu: [drm] *ERROR* wait_for_completion_timeout timeout!
```
The underlying s2idle issue that prompted this commit is going to
be fixed in BIOS.
This reverts commit 2a6c826cfe.
This was lost in the 6.19 merge so reapply it.
Fixes: 2a6c826cfe ("drm/amd: Skip power ungate during suspend for VPE")
Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reported-by: Konstantin <answer2019@yandex.ru>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=220812
Reported-by: Matthew Schwartz <matthew.schwartz@linux.dev>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 3925683515)
During Mode 1 reset, the ASIC undergoes a reset cycle and becomes
temporarily inaccessible via PCIe. Any attempt to access MMIO registers
during this window (e.g., from interrupt handlers or other driver threads)
can result in uncompleted PCIe transactions, leading to NMI panics or
system hangs.
To prevent this, set the `no_hw_access` flag to true immediately after
triggering the reset. This signals other driver components to skip
register accesses while the device is offline.
A memory barrier `smp_mb()` is added to ensure the flag update is
globally visible to all cores before the driver enters the sleep/wait
state.
Signed-off-by: Perry Yuan <perry.yuan@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 7edb503fe4)
During Mode 1 reset, the ASIC undergoes a reset cycle and becomes
temporarily inaccessible via PCIe. Any attempt to access MMIO registers
during this window (e.g., from interrupt handlers or other driver threads)
can result in uncompleted PCIe transactions, leading to NMI panics or
system hangs.
To prevent this, set the `no_hw_access` flag to true immediately after
triggering the reset. This signals other driver components to skip
register accesses while the device is offline.
A memory barrier `smp_mb()` is added to ensure the flag update is
globally visible to all cores before the driver enters the sleep/wait
state.
Signed-off-by: Perry Yuan <perry.yuan@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We need to check if the fence was signaled without an
error as the per queue resets may have signalled the fence
while attempting to reset the queue.
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add a uma/ directory containing two sysfs files as interfaces to
inspect or change UMA carveout size. These files are:
- uma/carveout_options: a read-only file listing all the available
UMA allocation options and their index.
- uma/carveout: a file that is both readable and writable. On read,
it shows the index of the current setting. Writing a valid index
into this file allows users to change the UMA carveout size to that
option on the next boot.
Co-developed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Yo-Jung Leo Lin (AMD) <Leo.Lin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The drm_*() macros include the device which is helpful for debugging
issues in multi-GPU systems.
Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
