No need to have special handling for swSMU supported ASICs.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
No need to have special handling for swSMU supported ASICs.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix header guard and make internal functions static. Fixes the below warnings:
drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu_reset.h:24:9: warning: '__AMDUGPU_RESET_H__' is used as a header guard here, followed by #define of a different macro [-Wheader-guard]
drivers/gpu/drm/amd/amdgpu/aldebaran.c:110:6: warning: no previous prototype for function 'aldebaran_async_reset' [-Wmissing-prototypes]
drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu13/aldebaran_ppt.c:1435:5: warning: no previous prototype for function 'aldebaran_mode2_reset' [-Wmissing-prototypes]
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reported-by: kernel test robot <lkp@intel.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add aldebaran to devices which support recovery
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>
v1: Aldebaran uses reset control to support mode2 reset. The sequences to
reset and restore hardware context are specific to a particular
configuration.
v2: Clear bus mastering before reset.
Fix coding style issues, drop unwanted variables and info log.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Expose PG/CG set states functions for other clients
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v1: Adds a function to load a list of FWs as passed by the caller. This is
needed as only a select need to loaded for some use cases.
v2: Omit unrelated change, remove info log, fix return value when count is 0
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This prefers reset control based handling if it's implemented
for a particular ASIC. If not, it takes the legacy path. It uses
the legacy method of preparing environment (job, scheduler tasks)
and restoring environment.
v2: remove unused variable (Alex)
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v1: Add generic amdgpu_reset_control to handle different types of resets. It
may be added at device, hive or ip level. Each reset control has a list
of handlers associated with it to handle different types of reset. Reset
control is responsible for choosing the right handler given a particular
reset context.
Handler objects may implement a set of functions on how to handle a
particular type of reset.
prepare_env = Prepare environment/software context (not used currently).
prepare_hwcontext = Prepare hardware context for the reset.
perform_reset = Perform the type of reset.
restore_hwcontext = Restore the hw context after reset.
restore_env = Restore the environment after reset (not used currently).
Reset context carries the context of reset, as of now this is based on
the parameters used for current set of resets.
v2: Fix coding style
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
Previous tdr design treats the first job in job_timeout as the bad job.
But sometimes a later bad compute job can block a good gfx job and
cause an unexpected gfx job timeout because gfx and compute ring share
internal GC HW mutually.
[How]
This patch implements an advanced tdr mode.It involves an additinal
synchronous pre-resubmit step(Step0 Resubmit) before normal resubmit
step in order to find the real bad job.
1. At Step0 Resubmit stage, it synchronously submits and pends for the
first job being signaled. If it gets timeout, we identify it as guilty
and do hw reset. After that, we would do the normal resubmit step to
resubmit left jobs.
2. For whole gpu reset(vram lost), do resubmit as the old way.
v2: squash in build fix (Alex)
Signed-off-by: Jack Zhang <Jack.Zhang1@amd.com>
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
BO with ttm_bo_type_sg type can also have tiling_flag and metadata.
So so BO type check for only ttm_bo_type_kernel.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reported-by: Tom StDenis <Tom.StDenis@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tiling flag and metadata are only needed for BOs created by
amdgpu_gem_object_create(), so we can remove those from the
base class.
v2: * squash tiling_flags and metadata relared patches into one
* use BUG_ON for non ttm_bo_type_device type when accessing
tiling_flags and metadata._
v3: *include to_amdgpu_bo_user
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>
Use amdgpu_bo_create_user() for all the BO allocations for
ttm_bo_type_device type.
v2: include amdgpu_amdkfd_alloc_gws() as well it calls amdgpu_bo_create()
for ttm_bo_type_device
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>
Implement a new struct amdgpu_bo_user as subclass of
struct amdgpu_bo and a function to created amdgpu_bo_user
bo with a flag to identify the owner.
v2: amdgpu_bo_to_amdgpu_bo_user -> to_amdgpu_bo_user()
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>
Allow allocating BO structures with different structure size
than struct amdgpu_bo.
v2: Check bo_ptr_size in all amdgpu_bo_create() caller.
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>
Add VCN3 IB parsing to figure out to which instance we can send the
stream for decode.
v2: remove VCN instance limit as well, fix amdgpu_cs_find_mapping,
check supported formats instead of unsupported.
v3: fix typo and error handling
v4: make sure the message BO is CPU accessible
v5: fix addr calculation once more
v6: only check message buffers
v7: fix constant and use defines
v8: fix create msg calculation
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Sonny Jiang <sonny.jiang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The VCN3 instances can do both decode as well as encode.
Share the scheduler load balancing score and remove fixing encode to
only the second instance.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-and-Tested-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Allow separate ring to share the same scheduler score.
No functional change.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-and-Tested-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
A previous fix I did left a rather complicated loop in
amdgpu_securedisplay_debugfs_write() for what could be expressed in a
simple sprintf, as Rasmus pointed out.
This drops the leading 0x for each byte, but is otherwise
much nicer.
Suggested-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
On initializing the framebuffer, call drm_any_plane_has_format to do a
check if the modifier is supported. drm_any_plane_has_format calls
dm_plane_format_mod_supported which is extended to validate that the
modifier is on the list of the plane's supported modifiers.
The bug was caught using igt-gpu-tools test: kms_addfb_basic.addfb25-bad-modifier
Tested on ChromeOS Zork by turning on the display, running an overlay
test, and running a YT video.
=== Changes from v1 ===
Explicitly handle DRM_FORMAT_MOD_INVALID modifier.
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Mark Yacoub <markyacoub@chromium.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix the following coccicheck warning:
drivers/gpu//drm/amd/amdgpu/amdgpu_ras.c:434:9-17: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_xgmi.c:220:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_xgmi.c:249:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/df_v3_6.c:208:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_psp.c:2973:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_vram_mgr.c:75:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_vram_mgr.c:112:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_vram_mgr.c:58:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_vram_mgr.c:93:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_vram_mgr.c:125:9-17: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_gtt_mgr.c:52:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_gtt_mgr.c:71:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_device.c:140:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_device.c:164:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_device.c:186:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_device.c:208:8-16: WARNING:
use scnprintf or sprintf
drivers/gpu//drm/amd/amdgpu/amdgpu_atombios.c:1916:8-16: WARNING:
use scnprintf or sprintf
Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
That is unused for quite some time now.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use positive logic to check for RAS
support. Rename the function to actually indicate
what it is testing for. Essentially, make the
function a predicate with the correct name.
Cc: Stanley Yang <Stanley.Yang@amd.com>
Cc: Alexander Deucher <Alexander.Deucher@amd.com>
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Setting amdgpu.noretry=1 as default for Arcturus.
Signed-off-by: Philip Cox <Philip.Cox@amd.com>
Reviewed-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
updated host to send boot config to psp to enable GECC for sienna cichlid
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 interface support for setting boot config
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>
[what]
currently driver recover vram after full access, which may hit
a corner case that meanwhile another whole gpu reset may be
triggered by another VF, which will cause vram recover fail
then fail the whole device reset.
[how]
move the recover vram into full access. So another bad VF will
not disturb the recover sequence for this vf.
Signed-off-by: Horace Chen <horace.chen@amd.com>
Reviewed by: Monk.Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use other APIs which are with the same functionality but much
more clean.
V2: drop mediate unneeded interface
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>
Also drop unnecessary header file and declarations.
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>
gcc warns about an sprintf() that uses the same buffer as source
and destination, which is undefined behavior in C99:
drivers/gpu/drm/amd/amdgpu/amdgpu_securedisplay.c: In function 'amdgpu_securedisplay_debugfs_write':
drivers/gpu/drm/amd/amdgpu/amdgpu_securedisplay.c:141:6: error: 'sprintf' argument 3 overlaps destination object 'i2c_output' [-Werror=restrict]
141 | sprintf(i2c_output, "%s 0x%X", i2c_output,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
142 | securedisplay_cmd->securedisplay_out_message.send_roi_crc.i2c_buf[i]);
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/amd/amdgpu/amdgpu_securedisplay.c:97:7: note: destination object referenced by 'restrict'-qualified argument 1 was declared here
97 | char i2c_output[256];
| ^~~~~~~~~~
Rewrite it to remember the current offset into the buffer instead.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
clang points out that the %hu format string does not match the type
of the variables here:
drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c:263:7: warning: format specifies type 'unsigned short' but the argument has type 'unsigned int' [-Wformat]
version_major, version_minor);
^~~~~~~~~~~~~
include/drm/drm_print.h:498:19: note: expanded from macro 'DRM_ERROR'
__drm_err(fmt, ##__VA_ARGS__)
~~~ ^~~~~~~~~~~
Change it to a regular %u, the same way a previous patch did for
another instance of the same warning.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Tom Rix <trix@redhat.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu_hdp.h has been included at line 91, so remove
the duplicate include.
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Wan Jiabing <wanjiabing@vivo.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
To make size is 4 byte aligned. Use &~0x3ULL instead of &3ULL.
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix size comparison in the resource cursor.
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>
The RLC was put into a wrong state on runtime suspend. Thus the RLC
autoload will fail on the succeeding runtime resume. By adding an
intermediate PPSMC_MSG_PrepareMp1ForUnload(some GC hard reset involved,
designed for PnP), we can bring RLC back into the desired state.
V2: integrate INTERRUPTS_ENABLED flag clearing into current
mp1 state set routines
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Enable clockgating for VCN and JPEG blocks on aldebaran
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
s/traing/training/
...Plus the entire sentence construction for better readability.
Signed-off-by: Bhaskar Chowdhury <unixbhaskar@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If userptr pages have been pinned but not bounded,
they remain uncleared.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Daniel Gomez <daniel@qtec.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
GFX is in gfxoff mode during s0ix so we shouldn't need to
actually tear anything down and restore it.
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We handle it properly within the CG/PG functions directly
now.
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Not needed as the device is in gfxoff state so the CG/PG state
is handled just like it would be for gfxoff during runtime gfxoff.
This should also prevent delays on resume.
Reworked from Pratik's original patch (Alex)
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Pratik Vishwakarma <Pratik.Vishwakarma@amd.com>
Provide and explanation as to why we skip GFX and PSP for
S0ix. GFX goes into gfxoff, same as runtime, so no need
to tear down and re-init. PSP is part of the always on
state, so no need to touch it.
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The SMU expects CGPG to be enabled when entering S0ix.
with this we can re-enable SMU suspend.
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This really needs to be done to properly tear down
the device. SMC, PSP, and GFX are still problematic,
need to dig deeper into what aspect of them that is
problematic.
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
No functional change.
v2: use correct dev
v3: rework
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Move the non-DC specific code into the DCE IP blocks similar
to how we handle DC. This cleans up the common suspend
and resume pathes.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Vram is system memory, so no need to evict.
v2: use PM_EVENT messages
v3: use correct dev
v4: use driver flags
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Set flags at the top level pmops callbacks to track
state. This cleans up the current set of flags and
properly handles S4 on S0ix capable systems.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
During system hibernation suspend still need un-gate gfx CG/PG firstly to handle HW
status check before HW resource destory.
Signed-off-by: Prike Liang <Prike.Liang@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>
There's no need to keep vgaswitcheroo around for HG
systems. They don't use muxes and their power control
is handled via ACPI.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Once the device has runtime suspended, we don't need to power it
back up again for system suspend. Likewise for resume, we don't
to power up the device again on resume only to power it back off
again via runtime pm because it's still idle.
v2: add DPM_FLAG_SMART_PREPARE as well
Reviewed-by: Evan Quan <evan.quan@amd.com>
Acked-by: Rajneesh Bhardwaj <rajneesh.bhardwaj@amd.com> (v1)
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
as per:
https://www.kernel.org/doc/html/latest/driver-api/pm/devices.html
The prepare callback is required to support the DPM_FLAG_SMART_SUSPEND
driver flag. This allows runtime pm to auto complete when the
system goes into suspend avoiding a wake up on suspend and on resume.
Apply this for hybrid gfx and BOCO systems where d3cold is
provided by the ACPI platform.
v2: check if device is runtime suspended in prepare.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
KIQ ring is being operated by kfd as well as amdgpu.
KFD is using kiq lock, we should the same from amdgpu side
as well.
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 33cf440d59.
And it will enable mode-2 gpu reset for vangogh,
it asks PSP firmware version is 00.1A.00.0F or newer.
Signed-off-by: Xiaojian Du <Xiaojian.Du@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When recovery thread has begun GPU reset, there should be not other
threads to access hardware, otherwise system randomly hang.
v2 (chk): rewritten from scratch, use trylock and lockdep instead of
hand wiring the logic.
v3: add in_irq check
v4: change to check in_task
Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
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>
amd-drm-next-5.13-2021-03-23:
amdgpu:
- Debugfs cleanup
- Various cleanups and spelling fixes
- Flexible array cleanups
- Initial AMD Freesync HDMI
- Display fixes
- 10bpc dithering improvements
- Display ASSR support
- Clean up and unify powerplay and swsmu interfaces
- Vangogh fixes
- Add SMU gfx busy queues for RV/PCO
- PCIE DPM fixes
- S0ix fixes
- GPU metrics data fixes
- DCN secure display support
- Backlight type override
- Add initial support for Aldebaran
- RAS fixes
- Prime fixes for A+A systems
- Reset fixes
- Initial resource cursor support
- Drop legacy IO BAR requirements
- Various power fixes
amdkfd:
- MMU notifier fixes
- APU fixes
radeon:
- Debugfs cleanups
- Flexible array cleanups
UAPI:
- amdgpu: Add a new INFO ioctl interface to query video capabilities
rather than hardcoding them in userspace. This allows us to provide
fine grained asic capabilities (e.g., if a particular part is
bandwidth limited, we can limit the capabilities). Proposed userspace:
https://gitlab.freedesktop.org/leoliu/drm/-/commits/info_video_capshttps://gitlab.freedesktop.org/leoliu/mesa/-/commits/info_video_caps
- amdkfd: bump the driver version. There was a problem with reporting
some RAS features on older versions of the driver. Proposed userspace:
7cdd63475c
Danvet: A bunch of conflicts all over, but it seems to compile ... I
did put the call to dc_allow_idle_optimizations() on a single line
since it looked a bit too jarring to be left alone.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210324040147.1990338-1-alexander.deucher@amd.com
Instead of having a global lock for potentially less contention.
Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/424010/
The HW is not in production yet. Driver support is still in development.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
As noted during the review this approach doesn't make sense at all.
We should not apply any limitation on the VRAM applications can use inside the kernel.
If an application or end user wants to reserve a certain amount of VRAM for bad pages handling we should do this in the upper layer.
This reverts commit f89b881c81.
Signed-off-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 are seeing VM page faults with this. Revert the change until the bugs
are fixed.
This reverts commit 94ae8dc557.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm_gem_object_put() should be paired with drm_gem_object_lookup().
All gem objs are saved in fb->base.obj[]. Need put the old first before
assign a new obj.
Trigger VRAM leak by running command below
$ service gdm restart
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We set the same variable a few lines above. Drop the duplicate
setting.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since this is a "revert of a revert", the end effect is that freesync
video is back to its original state, the way it was before the first
revert.
Signed-off-by: Nikola Cornij <nikola.cornij@amd.com>
Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
SMU internal might need to check this pending_reset setting to decide the reset method
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This is to fix the case where it only enable the light SMU
on normal device init. This feature actually need to be enabled after ASIC
been reset as well.
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Some ASICs do not have GOP driver to copy vbios image into
VFCT table. And it will go to next check.
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It was leftover from radeon where it was required for some
specific old hardware. It hasn't been required for ages
and the driver already falls back to MMIO when legacy IO
is not available. Legacy IO also seems to be problematic on
on some thunderbolt devices. Drop it.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Nicholas Johnson <nicholas.johnson-opensource@outlook.com.au>
Interrupts on are non-reentrant on linux. This is just an ancient
leftover from radeon where irq processing was kicked of from different
places.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
memalloc_nofs_save/restore are no longer sufficient to prevent recursive
lock warnings when holding locks that can be taken in MMU notifiers. Use
memalloc_noreclaim_save/restore instead.
Fixes: f920e413ff ("mm: track mmu notifiers in fs_reclaim_acquire/release")
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>
cause:
It is necessary to send ras disable command to ras-ta during gfx
block ras later init, because the ras capability is disable read
from vbios for vega20 gaming, but the ras context is released
during ras init process, this will cause send ras disable command
to ras-to failed.
how:
Delay releasing ras context, the ras context
will be released after gfx block later init done.
Changed from V1:
move release_ras_context into ras_resume
Changed from V2:
check BIT(UMC) is more reasonable before access eeprom table
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>
arcturus and sienna_cichlid share the same version
of umc_info interface (umc_info v33). arcturus uses
umc_config to indicate ECC capability, while
sienna_cichlid uses umc_config1 to indicate ECC
capability. driver needs to check either umc_config
or umc_config1 to decide ECC capability for ASICs
that use umc_info v33 interface.
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>
Separate the drm_mm_node walking from the actual handling.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Oak Zeng <Oak.Zeng@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Separate the drm_mm_node walking from the actual handling.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Oak Zeng <Oak.Zeng@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Separate the drm_mm_node walking from the actual handling.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Oak Zeng <Oak.Zeng@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Separate the drm_mm_node walking from the actual handling.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Oak Zeng <Oak.Zeng@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Separate the drm_mm_node walking from the actual handling.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Oak Zeng <Oak.Zeng@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Separate the drm_mm_node walking from the actual handling.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Oak Zeng <Oak.Zeng@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Separate the drm_mm_node walking from the actual handling.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Oak Zeng <Oak.Zeng@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Allows to walk over the drm_mm nodes in a TTM resource object.
v2: squash in fix from Felix
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Oak Zeng <Oak.Zeng@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cleanup pdb0 bo before bo_fini gets called
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
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>
1. make function mmhub_v1_7_setup_vm_pt_regs static
2. indent a if statement
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Rajneesh Bhardwaj <rajneesh.bhardwaj@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
ioremap_cache is not supported on some architecture
such as s390. Put the codes into a #ifdef to fix
some compile error reported by test robot.
v2: squash in non-x86 fix
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reported-by: Kernel test robot <lkp@intel.com>
Reviewed-by: Christian Konig <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 73bf5cad26.
Fixed in newer firmware
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
gfx_v8_0_parse_sq_irq() is using in_task() to distinguish if it is
invoked from a workqueue worker or directly from the interrupt handler.
The usage of in_interrupt() in drivers is phased out and Linus clearly
requested that code which changes behaviour depending on context should
either be separated or the context be conveyed in an argument passed by the
caller, which usually knows the context.
gfx_v8_0_parse_sq_irq() is invoked directly either from a worker or from
the interrupt service routine. The worker is only bypassed if the worker
is already busy.
Add an argument `from_wq' to gfx_v8_0_parse_sq_irq() which is true if
invoked from the worker.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
gfx_v9_0_get_gpu_clock_counter() acquires a mutex_t lock and is the only
caller of gfx_v9_0_kiq_read_clock().
If it safe to acquire a mutex_t then gfx_v9_0_get_gpu_clock_counter() is
always invoked from preemptible context.
Remove in_interrupt() because it superfluous as it will always return
false.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The usage of in_interrupt() in gmc_v*_process_interrupt() is intended to
use a different code path if invoked from the interrupt handler vs
invoked from the workqueue.
The usage of in_interrupt() in drivers is phased out and Linus clearly
requested that code which changes behaviour depending on context should
either be separated or the context be conveyed in an argument passed by the
caller, which usually knows the context.
gmc_v*_process_interrupt() is invoked via the ->process() callback
from amdgpu_ih_process() which in turn is invoked either from
amdgpu_irq_handler() (the interrupt handler) or from
amdgpu_irq_handle_*() which is a workqueue.
amdgpu_irq::ih is always processed from the interrupt handler, the other
three struct amdgpu_ih_ring members are processed from a workqueue.
Replace the in_interrupt() check with a comparison against adev->irq.ih.
A similar check is already done to check if the ih pointer is from
ih_soft.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
There is a spelling mistake in a drm debug message. Fix it.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In gpu reset, XGMI TA unload will cause gpu hang.
Skip it on A+A aldebaran.
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
SMU introduce the new interface to enable light Secondary Bus Reset mode, driver
enable it on passthrough + XGMI configuration
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Read eeprom through SMU doesn't works stable on XGMI reset during test.
skip it for now
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This recent change introduce SDMA interrupt info printing with irq->process function.
These functions do not require a set function to enable/disable the irq
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add IH function to allow caller to wait until ring entries are processed
until the checkpoint write pointer.
This will be primarily used by HMM to drain pending page fault interrupts
before memory unmap to prevent HMM from handling stale interrupts.
Suggested-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
FB BO should not be ttm_bo_type_kernel type and
amdgpufb_create_pinned_object() pins the FB BO anyway.
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In passthrough configuration, hypervisior will trigger the SBR(Secondary bus reset) to the devices
without sync to each other. This could cause device hang since for XGMI configuration, all the devices
within the hive need to be reset at a limit time slot. This serial of patches try to solve this issue
by co-operate with new SMU which will only do minimum house keeping to response the SBR request but don't
do the real reset job and leave it to driver. Driver need to do the whole sw init and minimum HW init
to bring up the SMU and trigger the reset(possibly BACO) on all the ASICs at the same time
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Acked-by: Andrey Grodzovsky andrey.grodzovsky@amd.com
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The gmc.xgmi.head list originally is designed for device list in the XGMI hive. Mix use it
for reset purpose will prevent the reset function to adjust XGMI device list which is required
in next change
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Reviewed-by: Andrey Grodzovsky andrey.grodzovsky@amd.com
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The MQD might not be initialized duirng first init period if the device need to be reset
druing probe. Driver need to proper init them in gpu recovery period
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu driver may be in reset state during init which will not initialize the kfd,
driver need to initialize the KFD after reset by check the flag
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit d86fd724e5.
Disable PSP RAP L0 self test until to RAP feature ready.
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>
To initialize the framebuffer, call drm_gem_fb_init_with_funcs which
verifies that the BO size can fit the FB size by calculating the minimum
expected size of each plane.
The bug was caught using igt-gpu-tools test: kms_addfb_basic.too-high
and kms_addfb_basic.bo-too-small
Tested on ChromeOS Zork by turning on the display and running a YT
video.
=== Changes from v1 ===
1. Added new line under declarations.
2. Use C style comment.
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: "Christian König" <christian.koenig@amd.com>
Cc: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Mark Yacoub <markyacoub@chromium.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Some configurations don't have FB BAR enabled. Avoid reading ROM image
from FB BAR region in such cases.
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>
This patch sets 'AMDGPU_GEM_CREATE_CPU_GTT_USWC' as input
parameter flag, during object creation of an imported DMA
buffer.
In absence of this flag:
1. Function amdgpu_display_supported_domains() doesn't add
AMDGPU_GEM_DOMAIN_GTT as supported domain.
2. Due to which, Function amdgpu_display_user_framebuffer_create()
refuses to create framebuffer for imported DMA buffers.
3. Due to which, AddFB() IOCTL fails.
4. Due to which, amdgpu_present_check_flip() check fails in DDX
5. Due to which DDX driver doesn't allow flips (goes to blitting)
6. Due to which setting Freesync/VRR property fails for PRIME buffers.
So, this patch finally enables Freesync with PRIME buffer offloading.
v2 (chk): instead of just checking the flag we copy it over if the
exporter is an amdgpu device as well.
Signed-off-by: Shashank Sharma <shashank.sharma@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Variable 'bp' seems to be unused residue from previous
logic, and is not required anymore.
Cc: Koenig Christian <christian.koenig@amd.com>
Cc: Deucher Alexander <alexander.deucher@amd.com>
Signed-off-by: Shashank Sharma <shashank.sharma@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This temporarily reverts freesync video patches since it causes regression with
eDP displays. This patch is a squashed revert of the following patches:
6f59f229f8 ("drm/amd/display: Skip modeset for front porch change")
d10cd527f5 ("drm/amd/display: Add freesync video modes based on preferred modes")
0eb1af2e82 ("drm/amd/display: Add module parameter for freesync video mode")
Signed-off-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Reviewed-by: Anson Jacob <anson.jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Aldebaran uses more than 4M runtime TMR. The current
hard coded 4M TMR is not big enough for Aldebaran.
Increase it to 8M.
v2: Only do 8M size for ALDEBARAN (Hawking)
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The flag is only applied on fine-grained memory.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Oak Zeng <Oak.Zeng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Page tables in vram mapping to cpu is changed from uncached to
cached in A+A, the snoop bit in VM_CONTEXTx_PAGE_TABLE_BASE_ADDR/
PDE0s/PDE1s/PDE2s/PTE.TFs has to be set so gpuvm walker snoop
page table data out of CPU cache.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
New A+A HW supports cached vram mapped to cpu.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When connected to a host via xGMI, system fatal errors may trigger
warm reset, driver has no change to query edc status before reset.
Therefore in this case, driver should harvest previous error loging
registers during boot, instead of only resetting them.
v2:
1. IP's ras_manager object is created when its ras feature is enabled,
so change to query edc status after amdgpu_ras_late_init called
2. change to enable watchdog timer after finishing gfx edc init
Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reivewed-by: Hawking Zhang <hawking.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This is needed for best machine learning performance. XNACK can still
be enabled per-process if needed.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Tested-by: Alex Sierra <alex.sierra@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add PSP RAP L0 check when RAP TA is loaded.
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>
RAP TA is an optional firmware. if it doesn’t exist,
the driver should bypass psp_rap_invoke() function.
1. bypass psp_rap_invoke() when RAP TA is not loaded.
2. add new parameter (status) to query RAP TA status.
(the status value is different with psp_ta_invoke(),
3. fix the 'rap_status' MThread critical problem.
(used without lock)
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>
When there is no graphics support, KFD can use more of the VMIDs. Graphics
VMIDs are only used for video decoding/encoding and post processing. With
two VCE engines, there is no reason to reserve more than 2 VMIDs for that.
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>
SQ's watchdog timer monitors forward progress, a mask of which waves
caused the watchdog timeout is recorded into ras status registers and
then trigger a system fatal error event.
v2:
1. change *query_timeout_status to *query_sq_timeout_status.
2. move query_sq_timeout_status into amdgpu_ras_do_recovery.
3. add module parameters to enable/disable fatal error event and modify
the watchdog timer.
v3:
1. remove unused parameters of *enable_watchdog_timer
Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The bank number of both VML2 and ATCL2 are changed to 8, so refine
related codes to avoid defining long name arrays.
Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add edc counter/status reset and query functions for gfx block of
aldebaran.
v2: change to clear edc counter explicitly
aldebaran hardware will not clear edc counter after driver reading them,
so driver should clear them explicitly.
Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add GC power brake feature support for Aldebaran.
v2: squash in fixes (Alex)
Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The golden setting was changed recently. update to
the latest one
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
golden settings that should be applied
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Those registers should be programmed as one-time initialization
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Initialization of TRAP_DATA0/1 is still required for the debugger to detect
new waves on Aldebaran. Also, per-vmid global trap enablement may be
required outside of debugger scope so move to init phase.
v2: just add the gfx 9.4.2 changes (Alex)
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>
Create dedicated Aldebaran kfd2kgd callbacks to prepare
for new per-vmid register instructions for debug trap
setting functions and sending host traps.
v2: rebase (Alex)
Signed-off-by: Jonathan Kim <Jonathan.Kim@amd.com>
Reviewed-by: Oak Zeng <Oak.Zeng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This is to keep wavefront context for debug purpose
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Match existing asics.
v2: rebase (Alex)
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Le Ma <Le.Ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
With a recent gart page table re-construction, the gart page
table is now 2-level for some ASICs: PDB0->PTB.
In the case of 2-level gart page table, the page_table_base
of vmid0 should point to PDB0 instead of PTB.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oak Zeng <Oak.Zeng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
More accurate words are used to address a
code review feedback
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For the new 2-level GART table, the last PDE0 points
to PTB. Since PTB is in vram and right now we are
runing under s=0 mode (vram is treated as FB carveout),
so the s bit of this PDE0 should be set to 0.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Aldebaran can share the same initializing shader code witn
arcturus.
Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
With the 2-level gart page table, vram is squeezed into gart aperture
and FB aperture is disabled. Therefore all VRAM virtual addresses are
in the GART aperture. However currently PSP requires TMR addresses
in FB aperture. So we need some design change at PSP FW level to support
this 2-level gart table driver change. Right now this PSP FW support
doesn't exist. To workaround this issue temporarily, FB aperture is
added back and the gart aperture address is converted back to FB aperture
for this PSP TMR address.
Will revert it after we get a fix from PSP FW.
v2: squash in tmr fix for other asics (Kevin)
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Set up HW for 2-level vmid0 page table: 1. Set up
PAGE_TABLE_START/END registers. Currently only plan
to do 2-level page table for ALDEBARAN, so only gfxhub1.0
and mmhub1.7 is changed. 2. Set page table base register.
For 2-level page table, the page table base should point
to PDB0. 3. Disable AGP and FB aperture as they are not
used.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If use gart for FB translation, allocate and fill
PDB0.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add functions to allocate PDB0, map it for CPU access,
and fill it.
Those functions are only used for 2-level vmid0 page
table construction
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If use gart for FB translation, we will squeeze vram into
sysvm aperture. This requires 2 level gart table. Add
page table depth and page table block size parameters
to gmc. This is prepare work to 2-level gart table
construction
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If use GART for FB translation, place both vram and gart to sysvm
aperture. AGP aperture is not set up in this case because it
is not used
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Modify the comment to reflect the fact that, if
use GART for vram address translation for vmid0,
[vram_start, vram_end] will be placed inside SYSVM
aperture, together with GART.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In amdgpu_gmc_gart_location function, gart_size is adjusted
by a smu_prv_buffer_size. This logic shouldn't belong to
this function. Move the logic to the mc_init functions
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
On A+A platform, CPU write page directory and page table in cached
mode. So it is necessary for page table walker to snoop CPU cache.
This setting is necessary for page walker to snoop page directory
and page table data out of CPU cache.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Acked-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
On A+A platform, vram can be mapped as WB. Not necessarily
to always map vram as WC on such platform.
Calling function arch_io_reserve_memtype_wc will mark the
whole vram region as WC. So don't call it for A+A platform.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The psp supplies the link type in the upper 2 bits of the psp xgmi node
information num_hops field. With a new link type, Aldebaran has these
bits set to a non-zero value (1 = xGMI3) so the KFD topology will report
the incorrect IO link weights without proper masking.
The actual number of hops is located in the 3 least significant bits of
this field so mask if off accordingly before passing it to the KFD.
Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Amber Lin <amber.lin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
By default this timestamp is 32 bit counter. It gets
overflowed in around 10 minutes.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If xnack is on, VM retry fault interrupt send to IH ring1, and ring1
will be full quickly. IH cannot receive other interrupts, this causes
deadlock if migrating buffer using sdma and waiting for sdma done while
handling retry fault.
Remove VMC from IH storm client, enable ring1 write pointer overflow,
then IH will drop retry fault interrupts and be able to receive other
interrupts while driver is handling retry fault.
IH ring1 write pointer doesn't writeback to memory by IH, and ring1
write pointer recorded by self-irq is not updated, so always read
the latest ring1 write pointer from register.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
With the current kfd memory accounting scheme, kfd applications
can use up to 15/16 of total system memory. For system which
has small total system memory size it leaves small system memory
for OS. For example, if the system has totally 16GB of system
memory, this scheme leave OS and non-kfd applications only 1GB
of system memory. In many cases, this leads to OOM killer.
This patch changed the KFD system memory accounting scheme.
15/16 of free system memory when kfd driver load. This deduct
the system memory that OS already use.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Suggested-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
