The GDS and GWS blocks default to allowing all VMIDs to
access all entries. Graphics VMIDs can handle setting
these limits when the driver launches work. However,
compute workloads under HWS control don't go through the
kernel driver. Instead, HWS firmware should set these
limits when a process is put into a VMID slot.
Disable access to these devices by default by turning off
all mask bits (for OA) and setting BASE=SIZE=0 (for GDS
and GWS) for all compute VMIDs. If a process wants to use
these resources, they can request this from the HWS
firmware (when such capabilities are enabled). HWS will
then handle setting the base and limit for the process when
it is assigned to a VMID.
This will also prevent user kernels from getting 'stuck' in
GWS by accident if they write GWS-using code but HWS
firmware is not set up to handle GWS reset. Until HWS is
enabled to handle GWS properly, all GWS accesses will
MEM_VIOL fault the kernel.
v2: Move initialization outside of SRBM mutex
Signed-off-by: Joseph Greathouse <Joseph.Greathouse@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The dpm sensor function already does this for us. This fixes
the freq*_input files with the new SMU implementation.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The register debugfs interface was using the wrong bitmask for vmid
selection for GFX_CNTL.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
As the lock was already held on the entrance to smu_handle_task.
- V2: lock in small granularity
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Under memory pressure, buffer moves between RAM to VRAM can
fail when there is no GTT space available. In those cases
amdgpu_bo_move falls back to ttm_bo_move_memcpy, which seems to
succeed, although it doesn't really support non-contiguous or
invisible VRAM. This manifests as VM faults with corrupted page
table entries in KFD eviction stress tests.
Print some helpful messages when lack of GTT space is causing buffer
moves to fail. Check that source and destination memory regions are
supported by ttm_bo_move_memcpy before taking that fallback.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Not used anymore.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Noticed-by: Dave Airlie <airlied@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Prefetch mode 0 is not supported and can lead to hangs with certain very
specific code patterns. Set a sound prefetch mode for all VMIDs rather
than forcing all shaders to set the prefetch mode at the beginning.
Reduce code duplication a bit while we're at it. Note that the 64-bit
address mode enum and the retry all enum are both 0, so the only
functional change is in the INITIAL_INST_PREFETCH field.
Signed-off-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The legacy navi10 sos binary will not carry on kdb image. the kdb_start_addr
is actually the start address of sys_drv image and shouldn't be sent to psp
bootloader.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: John Clements <john.clements@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When starting a new mm_node, the page_offset becomes 0.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The simple_strtol() function is deprecated. kstrto[l,u32]() is
the correct replacement as it can properly handle overflows.
This patch replaces the deprecated simple_strtol() use introduced recently.
As clk is of type uint32_t, we are safe to use kstrtou32().
It is also safe to return zero on string parsing error,
similar to the case of returning zero if buf is empty in parse_clk().
Fixes: bb5a2bdf36 ("drm/amdgpu: support dpm level modification under virtualization v3")
Signed-off-by: Wang Xiayang <xywang.sjtu@sjtu.edu.cn>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It is annoying to have #warnings that trigger in randconfig
builds like
drivers/gpu/drm/amd/amdgpu/soc15.c:653:3: error: "Enable CONFIG_DRM_AMD_DC for display support on SOC15."
drivers/gpu/drm/amd/amdgpu/nv.c:400:3: error: "Enable CONFIG_DRM_AMD_DC for display support on navi."
Remove these and rely on the users to turn these on.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The perf counter for Vega20 is 108, instead of 104 which it was on all
previous GPUs, so add a check to use the appropriate value.
Signed-off-by: Kent Russell <kent.russell@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The ability to select GFX GRBM me/pipe/queue/vmid was missing from
the gfx10 driver. This patch adds it. Used by the debugfs register
interface to select GFX resources when read/writing registers.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add 5 bits to the offset for SRBM selection to handle VMIDs. Also
update the select_me_pipe_q() callback to also select VMID.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Apply the same setting to SH_MEM_CONFIG and VM_CONTEXT1_CNTL. This
makes the noretry param no longer KFD-specific. On GFX10 I'm not
changing SH_MEM_CONFIG in this commit because GFX10 has different
retry behaviour in the SQ and I don't have a way to test it at the
moment.
Suggested-by: Christian König <Christian.Koenig@amd.com>
CC: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by : Shaoyun.liu < Shaoyun.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Starting from navi10, driver should send Key Database Load command
to bootloader before loading sys_drv and sos
Signed-off-by: John Clements <John.Clements@amd.com>
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Xiaojie Yuan <xiaojie.yuan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The command will be sent to psp bootloader from driver
to ask psp bootloader to exerise tOS, sys_drv and kdb loading
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Xiaojie Yuan <xiaojie.yuan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use the IP discovery table rather than hardcoding the
settings in the driver.
Reviewed-by: Xiaojie Yuan <xiaojie.yuan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
ID of DCE_HWIP from vbios is DMU_HWID,
mismatch cause null pointer crash in navi10 modprobe.
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Xiaojie Yuan <xiaojie.yuan@amd.com>
Signed-off-by: tiancyin <tianci.yin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
With mm_nodes larger than 4GB, byte_count in amdgpu_fill_buffer would
overflow.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We need to serialize access to the psp ring if there are multiple
callers at runtime.
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
New parameters added for navi lack documentation.
Reviewed-by: Xiaojie Yuan <xiaojie.yuan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This RELEASE_MEM use has the Release semantic, which means we should write
back but not invalidate. Invalidations only make sense with the Acquire
semantic (ACQUIRE_MEM), or when RELEASE_MEM is used to do the combined
Acquire-Release semantic, which is a barrier, not a fence.
The undesirable side effect of doing invalidations for the Release semantic
is that it invalidates caches while shaders are running, because the Release
can execute in the middle of the next IB.
UMDs should use ACQUIRE_MEM at the beginning of IBs. Doing cache
invalidations for a fence (like in this case) doesn't do anything
for correctness.
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Split it between navi10 and newer and everything before
navi10.
Reviewed-by: Xiaojie Yuan <xiaojie.yuan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Need to add appropriate ifdef.
Acked-by: Leo Li <sunpeng.li@amd.com>
Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When CONFIG_PERF_EVENTS is disabled, we cannot compile the pmu
portion of the amdgpu driver:
drivers/gpu/drm/amd/amdgpu/amdgpu_pmu.c:48:38: error: no member named 'hw' in 'struct perf_event'
struct hw_perf_event *hwc = &event->hw;
~~~~~ ^
drivers/gpu/drm/amd/amdgpu/amdgpu_pmu.c:51:13: error: no member named 'attr' in 'struct perf_event'
if (event->attr.type != event->pmu->type)
~~~~~ ^
...
Use conditional compilation for this file.
Fixes: 9c7c85f7ea ("drm/amdgpu: add pmu counters")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
clang warns:
In file included from drivers/gpu/drm/amd/amdgpu/nv.c:53:
drivers/gpu/drm/amd/amdgpu/../amdgpu/mes_v10_1.h:24:9: warning:
'__MES_V10_1_H__' is used as a header guard here, followed by #define of
a different macro [-Wheader-guard]
#ifndef __MES_V10_1_H__
^~~~~~~~~~~~~~~
drivers/gpu/drm/amd/amdgpu/../amdgpu/mes_v10_1.h:25:9: note:
'__MES_v10_1_H__' is defined here; did you mean '__MES_V10_1_H__'?
#define __MES_v10_1_H__
^~~~~~~~~~~~~~~
__MES_V10_1_H__
1 warning generated.
Capitalize the V.
Fixes: 886f82aa7a ("drm/amdgpu/mes10.1: add ip block mes10.1 (v2)")
Link: https://github.com/ClangBuiltLinux/linux/issues/582
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
An upcoming change in the hmm_range_register API requires passing in
a pointer to an hmm_mirror instead of mm_struct. To access the
hmm_mirror we need pass bo instead of ttm to amdgpu_ttm_tt_get_user_pages
because mirror is part of amdgpu_mn structure, which is accessible from bo.
v2: fix building without CONFIG_HMM_MIRROR (Arnd)
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add mode1 (by psp) reset for navi asic.
Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Perform a ras_suspend to disable ras on all IPs to workaround
some ROCm stability issue.
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Acked-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add baco reset support for smu11.
it can help gpu do asic reset when gpu recovery.
Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
kmemdup is introduced to duplicate a region of memory in a neat way.
Rather than kmalloc/kzalloc + memcpy, which the programmer needs to
write the size twice (sometimes lead to mistakes), kmemdup improves
readability, leads to smaller code and also reduce the chances of mistakes.
Suggestion to use kmemdup rather than using kmalloc/kzalloc + memcpy.
Reviewed-by: Christian König <Christian.Koenig@amd.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Signed-off-by: Fuqian Huang <huangfq.daxian@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Restore the code that resets mem->invalid. Othewise so mapping
userptrs after they got an MMU notifiers would always be skipped.
This also avoids unnecessarily calling get_user_pages on BOs that
have not been invalidated since the last try.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We don't want eviction fences to trigger when waiting for page table
updates to complete during restore. In theory there shouldn't be any
unsignaled eviction fences in the PD reservation object, but I'm
seeing them in instrumented code for reasons not fully understood.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since amdgpu has always requested PCIE atomics, kfd don't
need duplicated PCIE atomics enablement. Referring to amdgpu
request result is enough.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
GPU atomics operation depends on PCIE atomics support.
Always enable PCIE atomics ops support in case that
it hasn't been enabled.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The new field in amdgpu device is used to record whether the
system has PCIE atomics support. The field can be exposed to
UMD or kfd whether PCIE atomics have supported.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
MGPU fan boost feature should not be enabled until all the
devices from the same hive are all back from reset.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add the gfx10 equivalent of the gfx9 code.
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v2: update emit_ib_size
(though it's still wrong because it was wrong before)
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Make mem_busy_percent sysfs interface invisible on Vega10.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Set default value of this kernel parameter to 9000
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The buf swap field doesn't exist on RB1.
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The athub pg features enabling should be indicated by pg_flags.
Reported-by: Lijo Lazar <Lijo.Lazar@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
