Vega20 arbitrates pstate at hive level and not device level. Last peer to
remote buffer unmap could drop P-State while another process is still
remote buffer mapped.
With this fix, P-States still needs to be disabled for now as SMU bug
was discovered on synchronous P2P transfers. This should be fixed in the
next FW update.
Signed-off-by: Jonathan Kim <Jonathan.Kim@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We were registering the aux device in the MST late_register
rather than the regular one.
v2: handle eDP as well
Fixes: 405a1f9090 ("drm/amdgpu/display: split dp connector registration (v4)")
Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1100
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Initialize thermal controller fields in the PowerPlay table for Hawaii
GPUs, so that fan speeds are reported.
Signed-off-by: Sandeep Raghuraman <sandy.8925@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 3fded222f4
This is work around for vcn1 only. Currently vcn1 has separate
begin_use and idle work handle.
Signed-off-by: James Zhu <James.Zhu@amd.com>
Tested-by: changzhu <Changfeng.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Update the list with supported Arcturus chips, but disable for now until
final list is confirmed.
Ideally we can poll atombios for FRU support, instead of maintaining
this list of chips, but this will enable serial number reading for
supported ASICs for the time-being.
Signed-off-by: Kent Russell <kent.russell@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When the amdgpu in the suspend/resume loop need notify the dpm disabled,
otherwise the smu table will be uninitialize and result in resume failed.
Signed-off-by: Prike Liang <Prike.Liang@amd.com>
Tested-by: Mengbing Wang <Mengbing.Wang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fixes a minor typo in the file.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rajneesh Bhardwaj <rajneesh.bhardwaj@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
reduce cmd submission to smu by caching version info
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: John Clements <john.clements@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit c12b84d6e0.
The original patch causes a RAS event and subsequent kernel hard-hang
when running the KFDMemoryTest.PtraceAccessInvisibleVram on VG20 and
Arcturus
dmesg output at hang time:
[drm] RAS event of type ERREVENT_ATHUB_INTERRUPT detected!
amdgpu 0000:67:00.0: GPU reset begin!
Evicting PASID 0x8000 queues
Started evicting pasid 0x8000
qcm fence wait loop timeout expired
The cp might be in an unrecoverable state due to an unsuccessful queues preemption
Failed to evict process queues
Failed to suspend process 0x8000
Finished evicting pasid 0x8000
Started restoring pasid 0x8000
Finished restoring pasid 0x8000
[drm] UVD VCPU state may lost due to RAS ERREVENT_ATHUB_INTERRUPT
amdgpu: [powerplay] Failed to send message 0x26, response 0x0
amdgpu: [powerplay] Failed to set soft min gfxclk !
amdgpu: [powerplay] Failed to upload DPM Bootup Levels!
amdgpu: [powerplay] Failed to send message 0x7, response 0x0
amdgpu: [powerplay] [DisableAllSMUFeatures] Failed to disable all smu features!
amdgpu: [powerplay] [DisableDpmTasks] Failed to disable all smu features!
amdgpu: [powerplay] [PowerOffAsic] Failed to disable DPM!
[drm:amdgpu_device_ip_suspend_phase2 [amdgpu]] *ERROR* suspend of IP block <powerplay> failed -5
Signed-off-by: Kent Russell <kent.russell@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Set MP1 state to prepare for unload before reloading SMU FW
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: John Clements <john.clements@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Added dedicated function to check if particular fw should be skipped from loading.
Added dedicated function for SMU FW loading via PSP
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: John Clements <john.clements@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
To fit the latest PMFW.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
On the ASIC powered down(in baco or system suspend),
the dpm_enabled will be set as false. Then all access
(e.g. df state setting issued on RAS error event) to
SMU will be blocked.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
As data transfer may starts immediately after i2c eeprom init
completed. Thus i2c eeprom should be initialized after SMU
ready. And i2c data transfer should be prohibited when SMU
down. That is the i2c eeprom fini sequence needs to be
updated also.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
To fit the latest SMC firmware 42.53 and eliminate the
warning on driver loading.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It can support different NV1x ASIC better. And this can guard
no member got missing.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The system reboot failed as some IP blocks enter power gate before perform
hw resource destory. Meanwhile use unify interface to set device CGPG to ungate
state can simplify the amdgpu poweroff or reset ungate guard.
Fixes: 487eca11a3 ("drm/amdgpu: fix gfx hang during suspend with video playback (v2)")
Signed-off-by: Prike Liang <Prike.Liang@amd.com>
Tested-by: Mengbing Wang <Mengbing.Wang@amd.com>
Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The variable dp_ref_clk_khz is being initialized with a value that is
never read and it is being updated later with a new value. The
initialization is redundant and can be removed.
Addresses-Coverity: ("Unused value")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix the following gcc warning:
drivers/gpu/drm/radeon/ci_dpm.c:82:36: warning: ‘defaults_saturn_pro’
defined but not used [-Wunused-const-variable=]
static const struct ci_pt_defaults defaults_saturn_pro =
^~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/radeon/ci_dpm.c:68:36: warning: ‘defaults_bonaire_pro’
defined but not used [-Wunused-const-variable=]
static const struct ci_pt_defaults defaults_bonaire_pro =
^~~~~~~~~~~~~~~~~~~~
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix the following gcc warning:
drivers/gpu/drm/radeon/si_dpm.c:255:33: warning: ‘dte_data_tahiti_le’
defined but not used [-Wunused-const-variable=]
static const struct si_dte_data dte_data_tahiti_le =
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This code is dead, let's remove it.
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Let format prefixes take care of printing the module name
through pr_fmt and dev_fmt definitions.
Signed-off-by: Aurabindo Pillai <mail@aurabindo.in>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu uses lots of pr_* calls for printing error messages.
With this prefix, errors shall be more obvious to the end
use regarding its origin, and may help debugging.
Prefix format:
[xxx.xxxxx] amdgpu: ...
Signed-off-by: Aurabindo Pillai <mail@aurabindo.in>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix the following gcc warning:
drivers/gpu/drm/amd/amdgpu/../display/dc/dce80/dce80_hw_sequencer.c:43:46:
warning: ‘reg_offsets’ defined but not used [-Wunused-const-variable=]
static const struct dce80_hw_seq_reg_offsets reg_offsets[] = {
^~~~~~~~~~~
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Set up a GPU scheduler based on the ring flag rather
than the ring type.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We don't want a GPU scheduler for this ring.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This allows IPs to flag whether a specific ring requires
a GPU scheduler or not. E.g., sometimes instances of an
IP are asymmetric and have different capabilities.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Get SMC fw size before backdoor loading instead of giving an
certain value, as it may different for different ASIC.
Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Vram lost counter is wrongly increased by two during baco reset.
V2: assumed vram lost for mode1 reset on all ASICs
Signed-off-by: Evan Quan <evan.quan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Prefix RAS message printing in GFX IP with PCI device info,
which assists the debug in multiple GPU case.
Signed-off-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If there is no GPU hang, user still can access
debugfs through kiq.
Signed-off-by: Yintian Tao <yttao@amd.com>
Reviewed-by: Monk Liu <Monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Prefix ras related kernel message logging with PCI
device info by replacing DRM_INFO/WARN/ERROR with
dev_info/warn/err. This can clearly tell user about
GPU device information where ras is. And add some
other ras message printing to make it more clear
and friendly as well.
Suggested-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Uncorrectable error count printing is missed when issuing UMC
UE injection. When going to the error count log function in GPU
recover work thread, there is no chance to get correct error count
value by last error injection and print, because the error status
register is automatically cleared after reading in UMC ecc irq
callback. So add such message printing in UMC ecc irq cb to be
consistent with other RAS error interrupt cases.
Signed-off-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
If reading dpcd happens ahead of hw initialization, then aconnector is NULL
at this point. This is expected, so there is no need to output an error (which will
spam dmesg.log)
[How]
Change type of message from "error" to "DC_LOG_DC".
Signed-off-by: Zhan Liu <zhan.liu@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Acked-by: Zhan Liu <zhan.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Under bare metal, there is no more else to take
care of the GPU register access through MMIO.
Under Virtualization, to access GPU register is
implemented through KIQ during run-time due to
world-switch.
Therefore, under SR-IOV user can only access
debugfs to r/w GPU registers when meets all
three conditions below.
- amdgpu_gpu_recovery=0
- TDR happened
- in_gpu_reset=0
v2: merge amdgpu_virt_can_access_debugfs() into
amdgpu_virt_enable_access_debugfs()
v3: drop ret variable in amdgpu_virt_enable_access_debugfs()
and directly return result
Signed-off-by: Yintian Tao <yttao@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This sequence is recommended by PMFW team for the baco reset
with PMFW reloaded. And it seems able to address the random
failure seen on Arcturus.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: John Clements <John.Clements@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Execution will only reach here if the asserted condition is true.
Hence there is no need for the additional check.
Signed-off-by: Aurabindo Pillai <mail@aurabindo.in>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
fixes unused variable warning.
Reported-by: Eric Biggers <ebiggers@kernel.org>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Mikita Lipski <mikita.lipski@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Make the fw_write_wait default case true since presumably all new
gfx9 asics will have updated firmware. That is using unique WAIT_REG_MEM
packet with opration=1.
Signed-off-by: Aaron Liu <aaron.liu@amd.com>
Tested-by: Aaron Liu <aaron.liu@amd.com>
Tested-by: Yuxian Dai <Yuxian.Dai@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add indirect access support to registers outside of
mmio bar.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
all the register access through kiq is redirected
to amdgpu_kiq_rreg/amdgpu_kiq_wreg
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
those are not needed anymore
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
not needed anymore
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
all the mmCUR_CONTROL instances are in mmr range and
can be accessd directly by using RREG32/WREG32
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
