SDMA edc counter registers were added in gfx edc counters
array. When querying gfx error counter in that array, there
is no way to differentiate sdma instance number for different
asic and then results to NULL pointer access when trying to
read sdma register base address for instances greater
than 2 on Vega20.
In addition, this also results to wrong gfx error counters
since it actually added sdma edc counters.
Therefore, sdma edc counter registers should be separated
from gfx edc counter regsiter array and only get initialized
when driver tries to enable sdma ras.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm_sched_entity_init() takes drm gpu scheduler list instead of
drm_sched_rq list. This makes conversion of drm_sched_rq list
to drm gpu scheduler list unnecessary
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
VKexample test hang during Occlusion/SDMA/Varia runs.
Clear XNACK_WATERMK in reg SDMA0_UTCL1_WATERMK to fix this issue.
Signed-off-by: chen gong <curry.gong@amd.com>
Reviewed-by: Aaron Liu <aaron.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Find that ring sdma0 test failed if turn on SDMA powergating after do
mode2 reset.
Perhaps the mode2 reset does not reset the SDMA PG state, SDMA is
already powered up so there is no need to ask the SMU to power it up
again. So I skip this function for a moment.
Signed-off-by: chen gong <curry.gong@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
change struct ras_err_data *err_data to void *err_data, align with
umc code and the callback's declaration in each ras block could
pay no attention to the structure type
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The sdma_v4 should be ungated when the IP resume back,
otherwise it will hang up and resume time out error.
Signed-off-by: Prike Liang <Prike.Liang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu_sdma_ras_late_init is used to init sdma specfic
ras debugfs/sysfs node and sdma specific interrupt handler.
It can be shared among sdma generations
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Problem:
Under certain conditions, when some IP bocks take a RAS error,
we can get into a situation where a GPU reset is not possible
due to issues in RAS in SMU/PSP.
Temporary fix until proper solution in PSP/SMU is ready:
When uncorrectable error happens the DF will unconditionally
broadcast error event packets to all its clients/slave upon
receiving fatal error event and freeze all its outbound queues,
err_event_athub interrupt will be triggered.
In such case and we use this interrupt
to issue GPU reset. THe GPU reset code is modified for such case to avoid HW
reset, only stops schedulers, deatches all in progress and not yet scheduled
job's fences, set error code on them and signals.
Also reject any new incoming job submissions from user space.
All this is done to notify the applications of the problem.
v2:
Extract amdgpu_amdkfd_pre/post_reset from amdgpu_device_lock/unlock_adev
Move amdgpu_job_stop_all_jobs_on_sched to amdgpu_job.c
Remove print param from amdgpu_ras_query_error_count
v3:
Update based on prevoius bug fixing patch to properly call amdgpu_amdkfd_pre_reset
for other XGMI hive memebers.
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
ce can also trigger interrupt, and even both ce and ue error can be
found in one ras query, distinguishing between ce and ue in interrupt
handler is uncessary.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Suggested-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Currenly the error check on variable instance is always false because
it is a uint32_t type and this is never less than zero. Fix this by
making it an int type.
Addresses-Coverity: ("Unsigned compared against 0")
Fixes: 7d0e6329df ("drm/amdgpu: update more sdma instances irq support")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
we can simplify all those unnecessary function under
SRIOV for vega10 since:
1) PSP L1 policy is by force enabled in SRIOV
2) original logic always set all flags which make itself
a dummy step
besides,
1) the ih_doorbell_range set should also be skipped
for VEGA10 SRIOV.
2) the gfx_common registers should also be skipped
for VEGA10 SRIOV.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Emily Deng <Emily.Deng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The more 6 sdma instances work fine now with DF fix in vbios:
* mmDF_PIE_AON_MiscClientsEnable(0x1c728)=0x3fe(DF_ALL_INSTANCE)
[9:4]MmhubsEnable=3f (change from 0)
Signed-off-by: Le Ma <le.ma@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Another 6 sdma instances do not work at present. Disable them to unblock KFD
for silicon bringup as a workaround
Signed-off-by: Le Ma <le.ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The bit RSVD_ENG0 to RSVD_ENG5 in GPU_HDP_FLUSH_REQ/GPU_HDP_FLUSH_DONE
can be leveraged for sdma instance 2~7 to poll register/memory.
Signed-off-by: Le Ma <le.ma@amd.com>
Acked-by: Snow Zhang < Snow.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
On Arcturus, sdma instance 5~7 is connected to the second mmhub. The vmhub type
in amdgpu_ring_funcs is constant, so we create an individual amdgpu_ring_funcs
with different vmhub type(AMDGPU_MMHUB_1) for these sdma instances.
Signed-off-by: Le Ma <le.ma@amd.com>
Acked-by: Snow Zhang < Snow.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This change is needed for Arcturus which has 8 sdma instances.
The CG/PG part is not covered for now.
Signed-off-by: Le Ma <le.ma@amd.com>
Acked-by: Snow Zhang < Snow.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
