add and_mask since the programming logic of golden setting changed
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Tianci.Yin <tianci.yin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The table grows quickly during debug/development effort when
multiple RAS errors are injected. Allow to avoid this by setting
table header back to empty if needed.
v2: Switch to debugfs entry instead of load time parameter.
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This will allow to reset the table on the fly.
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
this interface is related to specific version of umc, distinguish it
from ras_late_init
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>
move umc ras init from ras module to umc block, generic ras module
should pay less attention to specific ras block.
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>
umc late init is umc specific, it's more suitable to be put in umc block
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>
define sched_policy in case CONFIG_HSA_AMD is not
enabled, with this there is no need to check for CONFIG_HSA_AMD
else where in driver code.
Suggested-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Shirish S <shirish.s@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If CONFIG_HSA_AMD is not set, build fails:
drivers/gpu/drm/amd/amdgpu/amdgpu_device.o: In function `amdgpu_device_ip_early_init':
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:1626: undefined reference to `sched_policy'
Use CONFIG_HSA_AMD to guard this.
Fixes: 1abb680ad371 ("drm/amdgpu: disable gfxoff while use no H/W scheduling policy")
Signed-off-by: Shirish S <shirish.s@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Move the ASIC specific code into a new callback function.
v2: mask the flags for SI and CIK instead of a BUG_ON().
v3: remove last missed BUG_ON().
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Unify how we map the UAPI flags to the PTE hardware flags for a mapping.
Only the MTYPE is actually ASIC dependent, all other flags should be
copied over 1 to 1 and ASIC differences are handled later on.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch is to add asic flag to enable device probe during kfd init.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
While gfxoff is enabled, the mmVM_XXX registers will be 0xfffffff while the GFX
is in "off" state. KFD queue creattion doesn't use ring based method, so it will
trigger a VM fault.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
There is no point retrying page faults in VMID0. Those faults are
always fatal.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-and-Tested-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
As more and more new asics start to reuse the old device IDs before
launch, there is a need to quickly override the existing asic type
corresponding to the reused device ID through a kernel parameter. With
this, engineers no longer need to rely on local hack patches,
facilitating cooperation across teams.
Signed-off-by: Yong Zhao <Yong.Zhao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We need to check if the nbios funcs exist before
checking the individual pointers.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
ras recovery_init should be called after ttm init,
bad page reserve should be put in front of gpu reset since i2c
may be unstable during gpu reset.
add cleanup for recovery_init and recovery_fini
v2: add more comment and print.
remove cancel_work_sync in recovery_init.
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>
save umc error records to ras bad page array
v2: add bad pages before gpu reset
v3: add NULL check for adev->umc.funcs
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
support eeprom records load and save for ras,
move EEPROM records storing to bad page reserving
v2: remove redundant check for con->eh_data
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
change bps type from retired page to eeprom table record, prepare for
saving umc error records to eeprom
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>
Fix typo which messed up the calculation.
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-and-tested-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Restoring clock gating break SMU opeartion afterwards, avoid
this until this further invistigated with SMU.
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-and-tested-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This optimizes out the pci device id usage in KFD and makes the code
more maintainable.
Signed-off-by: Yong Zhao <Yong.Zhao@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
These wptrs must be pinned and GPU accessible when this is called
from hqd_load functions. So they should never fault. This resolves
a circular lock dependency issue involving four locks including the
DQM lock and mmap_sem.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oak Zeng <Oak.Zeng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add support for loading XGMI/RAS FW
Signed-off-by: John Clements <john.clements@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
change r type from bool to int, suitable for both bool and int return
value
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add sw_fini interface of df_funcs.
This interface will remove sysfs file of df_cntr_avail
function.
The old behavior only create sysfs of df_cntr_avail
in sw_init, but never remove it for lack of sw_fini
interface. With this,driver will report create
sysfs fail when it's loaded for the second time.
Signed-off-by: Jack Zhang <Jack.Zhang1@amd.com>
Reviewed-by: Jonathan Kim <Jonathan.Kim@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu_nbio_ras_late_init is used to init nbio specfic
ras debugfs/sysfs node and nbio specific interrupt handler.
It can be shared among nbio 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>
amdgpu_mmhub_ras_late_init is used to init mmhub specfic
ras debugfs/sysfs node and mmhub specific interrupt handler.
It can be shared among mmhub 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>
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>
amdgpu_gfx_ras_late_init is used to init gfx specfic
ras debugfs/sysfs node and gfx specific interrupt handler.
It can be shared among gfx 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>
amdgpu_gmc_ras_late_init is used to init gmc specfic
ras debugfs/sysfs node and gmc specific interrupt handler.
It can be shared among gmc 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>
to prevent access to dangling pointers
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>
In case of RAS error allow user configure auto system
reboot through ras_ctrl.
This is also part of the temproray work around for the RAS
hang problem.
v4: Use latest kernel API for disk sync.
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@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>
Issue 1:
In XGMI case amdgpu_device_lock_adev for other devices in hive
was called to late, after access to their repsective schedulers.
So relocate the lock to the begining of accessing the other devs.
Issue 2:
Using amdgpu_device_ip_need_full_reset to switch the device list from
all devices in hive to the single 'master' device who owns this reset
call is wrong because when stopping schedulers we iterate all the devices
in hive but when restarting we will only reactivate the 'master' device.
Also, in case amdgpu_device_pre_asic_reset conlcudes that full reset IS
needed we then have to stop schedulers for all devices in hive and not
only the 'master' but with amdgpu_device_ip_need_full_reset we
already missed the opprotunity do to so. So just remove this logic and
always stop and start all schedulers for all devices in hive.
Also minor cleanup and print fix.
v4: Minor coding style fix.
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>
Trying to evict things from the current working set doesn't work that
well anymore because of per VM BOs.
Rely on reserving VRAM for page tables to avoid contention.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This hopefully helps reduce the contention for page tables.
v2: adjust maximum reported VRAM size as well
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Make VM updates depend on the moving fence instead of the exclusive one.
Makes it less likely to actually have a dependency.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
gds clearing workaround should only be applied on asics that support gfx ras
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>
