In amdgpu_vkms_conn_get_modes(), the return value of drm_cvt_mode()
is assigned to mode, which will lead to a NULL pointer dereference
on failure of drm_cvt_mode(). Add a check to avoid null pointer
dereference.
Signed-off-by: Ma Ke <make_ruc2021@163.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
introduced "ras_err_info" to better identify a RAS ERROR source.
NOTE:
For legacy chips, keep the original RAS error print format.
v1:
RAS errors may come from different dies during a RAS error query,
therefore, need a new data structure to identify the source of RAS ERROR.
v2:
- use new data structure 'amdgpu_smuio_mcm_config_info' instead of
ras_err_id (in v1 patch)
- refine ras error dump function name
- refine ras error dump log format
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(No effect outside the ras_mgr data structure)
Since a new member was added to the ras_err_data data structure,
it becomes unreasonable for the ras_mgr instance to contain this data,
because ras mgr only uses the 2 member information of ue_count/ce_count in err_data.
This patch changes the code err_data into built-in structure members,
making the code directly compatible.
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This is needed to correctly handle BOs imported into compute VM from gfx.
Both kfd and gfx should use same bo_va and set bo_va->ref_count correctly
when map the Bos into same VM, otherwise we may trigger kernel general
protection when iterate mappings over bo_va's valids or invalids list.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Xiaogang Chen <Xiaogang.Chen@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ramesh Errabolu <Ramesh.Errabolu@amd.com>
Tested-by: Xiaogang Chen <Xiaogang.Chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Expose ras table version & schema info to sysfs
v2: Updated schema to get poison support info
from ras context, removed asic specific checks
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>
PSP OS updates the version information in register. On APUs with PSPv13,
PSP OS will already be loaded with SBIOS. Hence use the version register
instead of using information in driver binary header.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu_uvd_suspend() allocates memory and copies objects into that
allocated memory. This fails under memory pressure. Instead move
majority of this code into a prepare step when swap can still be
allocated.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If any IP blocks allocate memory during their hw_fini() sequence
this can cause the suspend to fail under memory pressure. Introduce
a new phase that IP blocks can use to allocate memory before suspend
starts so that it can potentially be evicted into swap instead.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Linux PM core has a prepare() callback run before suspend.
If the system is under high memory pressure, the resources may need
to be evicted into swap instead. If the storage backing for swap
is offlined during the suspend() step then such a call may fail.
So move this step into prepare() to move evict majority of
resources and update all non-pmops callers to call the same callback.
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2362
Reviewed-by: Christian König <christian.koenig@amd.com>
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>
Add CG support for GFX/MC/HDP/ATHUB/IH/BIF.
Add PG support for GFX.
Signed-off-by: Li Ma <li.ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
IP discovery region has increased to > 8K on some SOCs.Maximum reserve
size is upto 12K, but not used. For now increase to 10K.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Return -EINVAL when MMSCH init fail which can be handle by function
amdgpu_device_reset_sriov correctly.
Signed-off-by: Lin.Cao <lincao12@amd.com>
Reviewed-by: Jingwen Chen <Jingwen.Chen2@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Remove IB end boundary requirement,
VPE has no such limitions, use existing
amdgpu_ring_generic_pad_ib() instead.
This is missed in rebase.
Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When MES is oversubscribed it may not frequently check for new
command submissions from driver if the scheduling load is high.
Response latency as high as 5 seconds has been observed.
Enable a flag which adds a check for new commands between
scheduling quantums.
Signed-off-by: Jay Cornwall <jay.cornwall@amd.com>
Cc: Alexandru Tudor <alexandru.tudor@amd.com>
Reviewed-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
SI hardware does not have doorbells at all, however currently the code
will try to do the allocation and thus fail, makes SI AMDGPU not usable.
Fix this failure by skipping doorbells allocation when doorbells count
is zero.
Fixes: 54c30d2a8d ("drm/amdgpu: create kernel doorbell pages")
Reviewed-by: Shashank Sharma <shashank.sharma@amd.com>
Signed-off-by: Icenowy Zheng <uwu@icenowy.me>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
SI hardware does not have doorbells at all, however currently the code
will try to do the allocation and thus fail, makes SI AMDGPU not usable.
Fix this failure by skipping doorbells allocation when doorbells count
is zero.
Fixes: 54c30d2a8d ("drm/amdgpu: create kernel doorbell pages")
Reviewed-by: Shashank Sharma <shashank.sharma@amd.com>
Signed-off-by: Icenowy Zheng <uwu@icenowy.me>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Here, Adding db_size in byte to find the doorbell's
absolute offset for both 32-bit and 64-bit doorbell sizes.
So that doorbell offset will be aligned based on the doorbell
size.
v2:
- Addressed the review comment from Felix.
v3:
- Adding doorbell_size as parameter to get db absolute offset.
v4:
Squash the two patches into one.
Cc: Christian Koenig <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Shashank Sharma <shashank.sharma@amd.com>
Signed-off-by: Arvind Yadav <Arvind.Yadav@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add hub->ctx_distance when read CONTEXT1_CNTL, align w/
write back operation.
v2: fix coding style errors reported by checkpatch.pl (Christian)
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Lang Yu <lang.yu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The amdgpu_ras_get_context may return NULL if device
not support ras feature, so add check before using.
Signed-off-by: Stanley.Yang <Stanley.Yang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add support to read Manufacturer Name and FRU File Id fields. Also add
sysfs device attributes for external usage.
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>
