This commit adds the cleaner shader microcode for GFX10.3.0 GPUs. The
cleaner shader is a piece of GPU code that is used to clear or
initialize certain GPU resources, such as Local Data Share (LDS), Vector
General Purpose Registers (VGPRs), and Scalar General Purpose Registers
(SGPRs).
Clearing these resources is important for ensuring data isolation
between different workloads running on the GPU. Without the cleaner
shader, residual data from a previous workload could potentially be
accessed by a subsequent workload, leading to data leaks and incorrect
computation results.
The cleaner shader microcode is represented as an array of 32-bit words
(`gfx_10_3_0_cleaner_shader_hex`). This array is the binary
representation of the cleaner shader code, which is written in a
low-level GPU instruction set.
When the cleaner shader feature is enabled, the AMDGPU driver loads this
array into a specific location in the GPU memory. The GPU then reads
this memory location to fetch and execute the cleaner shader
instructions.
The cleaner shader is executed automatically by the GPU at the end of
each workload, before the next workload starts. This ensures that all
GPU resources are in a clean state before the start of each workload.
This addition is part of the cleaner shader feature implementation. The
cleaner shader feature helps resource utilization by cleaning up GPU
resources after they are used. It also enhances security and reliability
by preventing data leaks between workloads.
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: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It ensures that appropriate error codes are returned when an error
condition is detected
Fixes the below;
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c:2849 amdgpu_ras_add_bad_pages() warn: missing error code here? 'amdgpu_umc_pages_in_a_row()' failed.
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c:2884 amdgpu_ras_add_bad_pages() warn: missing error code here? 'amdgpu_ras_mca2pa()' failed.
v2: s/-EIO/-EINVAL, retained the use of -EINVAL from
amdgpu_umc_pages_in_a_row & and amdgpu_ras_mca2pa_by_idx, when the
RAS context is not initialized or the convert_ras_err_addr function is
unavailable. (Thomas)
V3: Returning 0 as the absence of eh_data is acceptable. (Tao)
Fixes: a8d133e625 ("drm/amdgpu: parse legacy RAS bad page mixed with new data in various NPS modes")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Cc: YiPeng Chai <yipeng.chai@amd.com>
Cc: Tao Zhou <tao.zhou1@amd.com>
Cc: Hawking Zhang <Hawking.Zhang@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: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Users might switch to ROCM build does not include MEC SJT FW and driver
needs to consider this case.w
Signed-off-by: yfeng1 <yfeng1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Before, every time fdinfo is queried we try to lock all the BOs in the
VM and calculate memory usage from scratch. This works okay if the
fdinfo is rarely read and the VMs don't have a ton of BOs. If either of
these conditions is not true, we get a massive performance hit.
In this new revision, we track the BOs as they change states. This way
when the fdinfo is queried we only need to take the status lock and copy
out the usage stats with minimal impact to the runtime performance. With
this new approach however, we would no longer be able to track active
buffers.
Signed-off-by: Yunxiang Li <Yunxiang.Li@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241219151411.1150-6-Yunxiang.Li@amd.com
Signed-off-by: Christian König <christian.koenig@amd.com>
If IH primary ring and KFD ih fifo overflows, we may miss CP, SDMA
interrupts and cause application soft hang. Show warning message with
ring name if overflow happens.
Add function to get ih ring name to avoid duplicating it. To keep
warning message consistent between GPU generations, change all
*_ih.c except ASICs older than Vega which has only one ih ring.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
After GPU page fault, there are lots of page fault interrupts generated
at short period even with CAM filter enabled because the fault address
is different. Each page fault copy to KFD ih fifo to send event to user
space by KFD interrupt worker, this could cause KFD ih fifo overflow
while other processes generate events at same time.
KFD process is aborted after GPU page fault, we only need one GPU page
fault interrupt sent to KFD ih fifo to send memory exception event to
user space.
Incease KFD ih fifo size to 2 times of IH primary ring size, to handle
the burst events case.
This patch handle the gfx v9 path, cover retry on/off and CAM filter
on/off cases.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This partially reverts commit 194eb174cb.
This commit introduced a new state variable into adev without even
remotely worrying about CPU barriers.
Since we already have the amdgpu_in_reset() function exactly for this
use case partially revert that.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The sdma context empty interrupt is dropped in amdgpu_irq_dispatch
as unregistered interrupt src_id 243, this interrupt accounts to 1/3 of
total interrupts and causes IH primary ring overflow when running
stressful benchmark application. Disable this interrupt has no side
effect found.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The use of 1 << i in scheduler mask calculations can result in an
unintentional integer overflow due to the expression being
evaluated as a 32-bit signed integer.
This patch replaces 1 << i with 1ULL << i to ensure the operation
is performed as a 64-bit unsigned integer, preventing overflow
Discovered in coverity scan, CID 1636393, 1636175, 1636007, 1635853
Fixes: c5c63d9cb5 ("drm/amdgpu: add amdgpu_gfx_sched_mask and amdgpu_compute_sched_mask debugfs")
Signed-off-by: Karol Przybylski <karprzy7@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Driver has different ways to fetch VBIOS. If one of the methods doesn't
find an authentic one, it will show misleading info messages eventhough
a subsequent method finds a valid VBIOS. Keep the message level at debug
and add device context.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since 2320c9e6a7 ("drm/sched: memset() 'job' in drm_sched_job_init()")
accessing job->base.sched can produce unexpected results as the initialisation
of (*job)->base.sched done in amdgpu_job_alloc is overwritten by the
memset.
This commit fixes an issue when a CS would fail validation and would
be rejected after job->num_ibs is incremented. In this case,
amdgpu_ib_free(ring->adev, ...) will be called, which would crash the
machine because the ring value is bogus.
To fix this, pass a NULL pointer to amdgpu_ib_free(): we can do this
because the device is actually not used in this function.
The next commit will remove the ring argument completely.
Fixes: 2320c9e6a7 ("drm/sched: memset() 'job' in drm_sched_job_init()")
Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add support for the drm_panic module, which displays a pretty user
friendly message on the screen when a Linux kernel panic occurs.
Adapt Lu Yao's code to use common helpers derived from
Jocelyn's patch. This extends the non-DC code to enable
access to non-CPU accessible VRAM and adds support for
other DCE versions.
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Lu Yao <yaolu@kylinos.cn>
Cc: Jocelyn Falempe <jfalempe@redhat.com>
Cc: Harry Wentland <harry.wentland@amd.com>
- The previous patch only considered the case for baremetal
and is not applicable for SRIOV code path. We also need to
init fw_share for SRIOV VF
Fixes: 928cd772e1 ("drm/amdgpu/vcn: reset fw_shared when VCPU buffers corrupted on vcn v4.0.3")
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Bokun Zhang <bokun.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
VF device sets the RAS flag when mailbox data can't be read properly.
There is no conclusive way to tell if the real source is RAS error.
Therefore VF schedules a KFD based reset which doesn't set RAS source.
SKip checking RAS source for any VF scheduled recovery.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reported-by: Vojislav Tomasevic <vojislav.tomasevic@amd.com>
Reviewed-by: Yiqing Yao <yiqing.yao@amd.com>
Tested-by: Yiqing Yao <yiqing.yao@amd.com>
Fixes: e1ee2111ca ("drm/amdgpu: Prefer RAS recovery for scheduler hang")
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Extracts the resume sequence for per sdma instance from sdma_v7_0_gfx_resume.
This function can be used in start or restart scenarios of specific instances.
Signed-off-by: Jesse Zhang <jesse.zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
