Add interface to reserve bad page.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In case of fatal errors, set FED status when interrupt is received. Set
the flag on other devices in the hive before RAS recovery work.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Asad Kamal <asad.kamal@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add ras event id support for ACA.
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add aca deferred error type support
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Each RAS block has different requirement for gpu reset in poison
consumption handling.
Add support for mmhub RAS poison consumption handling.
v2: remove the mmhub poison support for kfd int v10.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add amdgpu ras event id support to better distinguish different
error information sources in dmesg logs.
the following log will be identify by event id:
{event_id} interrupt to inform RAS event
{event_id} ACA logs
{event_id} errors statistic since from current injection/error query
{event_id} errors statistic since from gpu load
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Check amdgpu_ras_mask to fix ineffective ras_mask setting
due to special asic without sram ecc enable but with poison
supported.
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>
For a RAS error that needs a full reset to recover, set the fatal error
status. Clear the status once the device is reset.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Asad Kamal <asad.kamal@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Update boot time errors polling sequence to align with
the latest firmware change.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Frank Min <Frank.Min@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
- move aca init/fini function into ras init/fini to adapt gpu reset
sequence.
- add new function amdgpu_aca_reset()
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The driver unload hangs because the page retirement
kthread cannot be stopped as it is sleeping and waiting
on page retirement event to occur. Add kthread_should_stop()
to the event condition to wake up the kthread when kthread
stop is called during driver unload.
Fixes: 3fdcd0a31d ("drm/amdgpu: Prepare for asynchronous processing of umc page retirement")
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
skip call ras_late_init callback if ras block is not supported.
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use asynchronous polling to handle umc_v12_0 poisoning.
v2:
1. Change function name.
2. Change the debugging information content.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The high three bits of ras features mask indicate socket
id, it should skip to check high three bits of ras features
mask before disable all ras features.
Signed-off-by: Stanley.Yang <Stanley.Yang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
fix array index out of bounds issue for ras_block_string[] array.
Fixes: 30df05fb74 ("drm/amdgpu: Align ras block enum with firmware")
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Move ras capablity check to amdgpu_ras_check_supported.
Driver will query ras capablity through psp interace, or
vbios interface, or specific ip callbacks.
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>
Separate deferred error from UE and CE and log it
individually.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add aca sysfs node support
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v1:
add ACA error query interface
v2:
Add a new helper function to determine whether to use ACA or MCA.
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add ACA bank dump debugfs support
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add ras helper function to query boot time gpu
errors.
v2: use aqua_vanjaram smn addressing pattern
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Le Ma <le.ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Support poison error injection.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Remove "no user action is needed" for correctable and deferred error
to avoid confusion.
Signed-off-by: Candice Li <candice.li@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>
Not needed any more with firmware fixes
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>
Return invalid error code -EINVAL for invalid block id.
Fixes the below:
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c:1183 amdgpu_ras_query_error_status_helper() error: we previously assumed 'info' could be null (see line 1176)
Suggested-by: Hawking Zhang <Hawking.Zhang@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: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fixes the below smatch warnings:
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c:2543 amdgpu_ras_recovery_init() warn: Please consider using kzalloc instead of kmalloc
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c:2830 amdgpu_ras_init() warn: Please consider using kzalloc instead of kmalloc
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>
-----BEGIN PGP SIGNATURE-----
iQFSBAABCAA8FiEEq68RxlopcLEwq+PEeb4+QwBBGIYFAmV2PMQeHHRvcnZhbGRz
QGxpbnV4LWZvdW5kYXRpb24ub3JnAAoJEHm+PkMAQRiGhgQIAKSdjtpkv+IeDwph
umLHxDlCTT8nl+TAhEYuxSdMiHt1i7+zyF0snnHhGthBRnBZBf+PDdkQOmqtjLZj
e591VGY7tdNCAVUwmRRaI82JFz9Pl1k8DXL3f+CE1+MfbpsirAecoIAL0rsvg+4f
ICKSAOBZonL7GT7IGrsbxgqGKCLC+2PcgNT4pADBdjtbC1nmVSLK21dtWr6i4xsn
CT9MxAnej1+EOpreyHhNZUVqfQy0/04x4OY0u/EOXU3PgpfGTrAzkyfwzcd/eit0
x59TZ0owDVXfvvdXUUAC0zG1th2ek5LKNtL+C4rKOoAbq7GZoQfjN5M/Ug+yykwE
aqm6CJ8=
=tZ+R
-----END PGP SIGNATURE-----
Backmerge tag 'v6.7-rc5' into drm-next
Linux 6.7-rc5
Alex requested this for some amdkfd work relying on the symbols exports.
Signed-off-by: Dave Airlie <airlied@redhat.com>
sort error data list to optimize the printing order.
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
enable MCA debug mode on APU device by default.
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Refactor code such that ras block decides the default mca debug mode,
and not swsmu block.
By default mca debug mode is set to false.
v2: squash in uninitialized value fix (Alex)
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>
Direct error query mode and firmware error query mode
are supported for now.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Not all platforms support RAS.
Fixes: 73582be11a ("drm/amdgpu: bypass RAS error reset in some conditions")
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
PMFW is responsible for RAS error reset in some conditions, driver can
skip the operation.
v2: add check for ras->in_recovery, it's set earlier than
amdgpu_in_reset.
v3: fix error in gpu reset check.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Enable it by default on APU platform.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Stanley.Yang <Stanley.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
the origin function might return the wrong node.
Fixes: 5b1270beb3 ("drm/amdgpu: add ras_err_info to identify RAS error source")
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Record the debug mode status in RAS.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Enable RAS feature by default for aqua vanjaram on apu
platform.
Signed-off-by: Stanley.Yang <Stanley.Yang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
typo fix.
Fixes: 5b1270beb3 ("drm/amdgpu: add ras_err_info to identify RAS error source")
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Candice Li <candice.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix delete nodes that it has been freed.
Fixes: 5b1270beb3 ("drm/amdgpu: add ras_err_info to identify RAS error source")
Signed-off-by: Stanley.Yang <Stanley.Yang@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Set VCN/JPEG RAS masks to enable software RAS for
VCN and JPEG.
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>
Make the code architecture more simple.
v2: reuse ras_reset_error_count in ras_reset_error_status.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If one of the devices in the hive detects a
fatal error, need to send ras recovery reset
message to PMFW of all devices in the hive.
For that add a flag in hive to indicate that
it's undergoing ras recovery
Signed-off-by: Asad Kamal <asad.kamal@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.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>
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>
It should first check block ras obj whether be set, it should
return 0 directly if block ras obj or hw_ops is not set.
If block doesn't support RAS just return 0 is fine.
Changed from V1:
return 0 directly if block ras obj or hw ops is not set
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>
This patch fixes a memory leak in the amdgpu_ras_feature_enable() function.
The leak occurs when the function sends a command to the firmware to enable
or disable a RAS feature for a GFX block. If the command fails, the kfree()
function is not called to free the info memory.
Fixes: 9f051d6ff1 ("drm/amdgpu: Free ras cmd input buffer properly")
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Cong Liu <liucong2@kylinos.cn>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use an inline function for version check. Gives more flexibility to
handle any format changes.
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>
So driver doesn't generate incorrect message until
the new format is settled down for aqua_vanjaram
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Do not access the pointer for ras input cmd buffer
if it is even not allocated.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Stanley Yang <Stanley.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Disable gfx ras command is needed in some use cases
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>
Mode1 reset needs to recover mp1 in fatal error case
for mp0 v13_0_10.
v2:
Define a macro to wrap psp function calls.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
RAS global isr will only be invoked by hardware
interrupt. Don't need to query ras capability in isr
In addition, amdgpu_ras_interrupt_fatal_error_handler
ensures the isr won't be called from guest linux
side by accident. The RAS cap check in isr that
introduced to fix sriov crash is not needed any more
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>
Treat SDMA/VCN/JPEG as RAS capable IP blocks in poison mode.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Register RAS fatal error interrupt and add handler.
v2: only register NBIO RAS for dGPU platform.
change nbio_v7_9_set_ras_controller_irq_state and nbio_v7_9_set_ras_err_event_athub_irq_state
to dummy functions.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For non-GFX IP blocks, set up ras obj if ras feature
is allowed. For GFX IP blocks, force issue ras
enable_feature command to firmware and only set up ras
obj if ras feature is allowed
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>
For GFX IP that only supports poison consumption, GFX
RAS won't be marked as enabled. i.e., hardware doesn't
support gfx sram ecc. But driver still needs to issue
firmware to enable poison consumption mode for GFX IP.
In such case, check poison mode and treat GFX IP as
RAS capable IP block.
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>
Some IP blocks only support partial ras feature and don't
have ras counter and/or ras error status register at all.
Driver should not create err_count sysfs node for those
IP blocks.
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>
Only disable RAS by default for aqua vanjaram on APU platform.
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>
The VBIOS part number is read both in amdgpu_atom_parse() as well
as in atom_get_vbios_pn() and stored twice in the `struct atom_context`
structure. Remove the first unnecessary read and move the `pr_info`
line from that read into the second.
v2: squash in unused variable removal
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Disable RAS feature by default for aqua vanjaram on APU platform.
Changed from V1:
Splite Disable RAS by default on APU platform into a
separated patch.
Changed from V2:
Avoid to modify global variable amdgpu_ras_enable.
Signed-off-by: Stanley.Yang <Stanley.Yang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Enable RAS for aqua vanjaram.
Changed from V1:
Split the change in amdgpu_ras_asic_supported into a
separated patch.
Changed from V2:
Avoid to modify global variable amdgpu_ras_enable.
Signed-off-by: Stanley.Yang <Stanley.Yang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The address parameter of GFX RAS injection isn't related to XGMI node
number, keep it unchanged.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Candice Li <candice.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Do not compare injection address with mc_vram_size
if mc_vram_size is zero.
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>
Using "is_app_apu" to identify device in the native
APU mode or carveout mode.
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>
The fixed commit listed in the Fixes tag below, introduced a bug in
amdgpu_ras.c::amdgpu_reserve_page_direct(), in that when introducing the new
amdgpu_umc_fill_error_record() and internally in that new function the physical
address (argument "uint64_t retired_page"--wrong name) is right-shifted by
AMDGPU_GPU_PAGE_SHIFT. Thus, in amdgpu_reserve_page_direct() when we pass
"address" to that new function, we should NOT right-shift it, since this
results, erroneously, in the page address to be 0 for first
2^(2*AMDGPU_GPU_PAGE_SHIFT) memory addresses.
This commit fixes this bug.
Cc: Tao Zhou <tao.zhou1@amd.com>
Cc: Hawking Zhang <Hawking.Zhang@amd.com>
Cc: Alex Deucher <Alexander.Deucher@amd.com>
Fixes: 400013b268 ("drm/amdgpu: add umc_fill_error_record to make code more simple")
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Link: https://lore.kernel.org/r/20230610113536.10621-1-luben.tuikov@amd.com
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Report the number of records stored in the RAS EEPROM table in debugfs.
This can be used by user-space to calculate the capacity of the RAS EEPROM
table since "bad_page_cnt_threshold" is also reported in the same place in
debugfs.
See commit 7fb6407145 ("drm/amdgpu: Add bad_page_cnt_threshold to debugfs").
ras_num_recs can already be inferred by dumping the RAS EEPROM table, also in
the same debugfs location, see commit reference c65b0805e7 (drm/amdgpu:
RAS EEPROM table is now in debugfs, 2021-04-08). This commit makes it an
integer value easily shown in a single file.
Cc: Alex Deucher <Alexander.Deucher@amd.com>
Cc: Hawking Zhang <Hawking.Zhang@amd.com>
Cc: Tao Zhou <tao.zhou1@amd.com>
Cc: Stanley Yang <Stanley.Yang@amd.com>
Cc: John Clements <john.clements@amd.com>
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Link: https://lore.kernel.org/r/20230603051043.211548-1-luben.tuikov@amd.com
Acked-by: Alex Deucher <Alexander.Deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add ras info to EEPROM table, app can analyse device ECC
status without GPU driver through EEPROM table ras info.
Signed-off-by: Stanley.Yang <Stanley.Yang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The mask is only needed to be set when RAS block instance number is
more than 1 and invalid bits should be also masked out.
We only check valid bits for GFX and SDMA block for now, and will
add check for other RAS blocks in the future.
v2: move the check under injection operation since the mask is only
used by RAS error inject.
v3: add valid bits handling for SDMA.
v4: print message if the mask is adjusted.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Stanley.Yang <Stanley.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
So GFX RAS injection could use default function if it doesn't define its
own injection interface.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Stanley.Yang <Stanley.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
User can specify injected instances by the mask. For backward
compatibility, the mask value is incorporated into sub block index
without interface change of RAS TA.
User uses logical mask and driver should convert it to physical value
before sending it to RAS TA.
v2: update parameter name.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Stanley.Yang <Stanley.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It turns out STATUS_VALID_FLAG needs to be checked
ahead of any other fields. ADDRESS_VALID_FLAG and
ERR_INFO_VALID_FLAG only manages ADDRESS and ERR_INFO
field respectively. driver should continue poll
ERR CNT field even ERR_INFO_VALD_FLAG is not set.
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>
Persistent edc harvesting is supported in APP APU
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>
Add common helper to reset ras error status. It
applies to IP blocks that follow the new ras error
logging register design, and need to write 0 to
reset the error status. For IP blocks that don't
support the new design, please still implement ip
specific helper.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add common helper to query ras error status and
log error information, including memory block id
and erorr count. The helpers are applicable to IP
blocks that follow the new ras error logging design.
For IP blocks that don't support the new design,
please still implement ip specific helper to query
ras error.
v2: optimize struct amdgpu_ras_err_status_reg_entry
and the implementaion in helper (Lijo/Tao)
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Some ASICs support fatal error event but do not
support pcie_bif ras.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
XGMI RAS should be according to the gmc xgmi physical nodes number,
XGMI RAS should not be enabled if xgmi num_physical_nodes is zero.
Signed-off-by: Stanley.Yang <Stanley.Yang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
GPU will stop working once fatal error is detected.
it will inform driver to do reset to recover from
the fatal error.
v2: squash in logic fix (Srinivasan)
v3: squash in logic fix (Dan)
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Candice Li <candice.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu_ras_register_ras_block should always be invoked
by ras_sw_init, where driver needs to check ras caps
at ip level, instead of asic level.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Stanley Yang <Stanley.Yang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
