0 should be right driver return value, 0x1 is the right firmware
return value. So switch to 0 at last.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Xiaojian Du <xiaojian.du@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
PCIE MMIO bar needs to be restored firstly after the reset event
triggers. So it's unable to access the registers to wait for response
from SMU. Becasue the value of mmMP1_SMN_C2PMSG_90 is invalid at that
moment.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Enable GPU reset when we encounter a hang.
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
We need to save and restore PCI config space.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Vangogh will use mode2 reset, so plumb it through the nv
soc driver.
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
GPU reset is handled via SMU similar to previous APUs.
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
[Why]
When host trigger a whole gpu reset, guest will keep
waiting till host finish reset. But there's a work
queue in guest exchanging data between vf&pf which need
to access frame buffer. During whole gpu reset, frame
buffer is not accessable, and this causes the call trace.
[How]
After vf get reset notification from pf, stop data exchange.
Signed-off-by: Jingwen Chen <Jingwen.Chen2@amd.com>
Signed-off-by: Jack Zhang <Jack.Zhang1@amd.com>
Reviewed-by: Monk Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
VHG based APU will support feature mask checking.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
VanGogh supports feature mask checking which exposed by smu firmware. It
has to initial at first, otherwise, all SMU_FEATURE_xxx masks are
invalid.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The get_allowed_feature_mask is superfluous on vangogh.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The user prefers to know the real response value from C2PMSG 90 register
which is written by firmware not -EIO.
v2: return C2PMSG 90 value
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
VCN/JPEG PG won't be a feature mask bit which exposed by vangogh smu
firmware. So remove it.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
KASAN reported a slab-out-of-bounds read of size 1 in
kdf_create_vcrat_image_cpu().
This occurs when, for example, when on an x86_64 with a single NUMA node
because kfd_fill_iolink_info_for_cpu() is a no-op, but afterwards the
sub_type_hdr->length, which is out-of-bounds, is read and multiplied by
entries. Fortunately, entries is 0 in this case so the overall
crat_table->length is still correct.
Check if there were any entries before de-referencing sub_type_hdr which
may be pointing to out-of-bounds memory.
Fixes: b7b6c38529 ("drm/amdkfd: Calculate CPU VCRAT size dynamically (v2)")
Suggested-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Jeremy Cline <jcline@redhat.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For multiple VF, after engine hang,as host driver will first
encounter FLR, so has no meanning to set compute to 60s.
v2:
Refine the patch and comment
Signed-off-by: Emily.Deng <Emily.Deng@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
commit a861736dae ("drm/amd/display: Fixed Intermittent blue screen on OLED panel")
causes power regression for many users. It seems that this change causes
the MCLK to get forced high; this creates a regression for many users
since their devices were not able to drop to a low state after this
change. For this reason, this reverts commit
a861736dae.
Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1407
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Naveed Ashfaq <Naveed.Ashfaq@amd.com>
Cc: Hersen Wu <hersenxs.wu@amd.com>
Cc: Roman Li <roman.li@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[why]
driver has sent PMFW VBIOSSMC_MSG_SetDisplayIdleOptimizations while SMU
still processing a previous VBIOSSMC_MSG_SetHardMinDcfclkByFreq message.
[how]
same as RN, change the time out to2s.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Charlene Liu <Charlene.Liu@amd.com>
Reviewed-by: Chris Park <Chris.Park@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
The initial purpose of dcn10 pipe split is to support some high
bandwidth mode which requires dispclk greater than max dispclk. By
initial bring up power measurement data, it showed power consumption is
less with pipe split for dcn block. This could be reason for enable pipe
split by default. By battery life measurement of some Chromebooks,
result shows battery life is longer with pipe split disabled.
[How]
Disable pipe split by default. Pipe split could be still enabled when
required dispclk is greater than max dispclk.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Hersen Wu <hersenxs.wu@amd.com>
Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Roman Li <Roman.Li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[why]
Required for DSC MST
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Nikola Cornij <nikola.cornij@amd.com>
Reviewed-by: Zhan Liu <Zhan.Liu@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
- Reverts "drm/amd/display: Revert HUBP blank behaviour for now"
- Hubp blank will fail if the pipe is locked (this is the case on
linux), so add a check to make sure pipe isn't locked, if it is then
defer the blank to post_unlock.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY]
DCN30 has a bug where blanking HUBP blocks pstate allow unless
HUBP_DISABLE is toggled afterwards.
[HOW]
Create a HW sequence for blanking HUBP.
1. Wait for enter VBLANK
2. Set HUBP_BLANK
3. Make sure HUBP_IN_BLANK = 1
4. Toggle HUBP_DISABLE on and off to perform soft reset
All existing calls to hubp->funcs->set_blank should be replaced with
this new sequence.
In wait_for_mpcc_disconnect, only blank the pipe being disconnected, and
leave all other pipes unmodified.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
A call to BREAK_TO_DEBUGGER is triggered when powering down memory
[How]
Remove the check for GAMCOR_MEM_PWR_STATE when powering off the memory
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Jacky Liao <ziyu.liao@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
Seamless boot stream has hw resource assigned, already. 'add' is
actually rebuild the assignment.
[How]
Swap seamless boot stream to pipe 0 (if needed) to ensure pipe_ctx
matches
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Raymond Yang <rayyang@amd.com>
Reviewed-by: Martin Leung <Martin.Leung@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[why]
When OS reboots, and panel is turned off, T12 may not be maintained.
T12 is defined as the interval between VDDC off (occurs at shutdown) and
the next VDDC on (occurs when eDP is POST-ed)
[how]
DC already tracks panel power off time. Add a DC interface which DM can
call during shutdown. Ideally this should be as late as possible during
the shutdown sequence so the extra delay is minimal.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Jun Lei <jun.lei@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
1. the corrected timing to write DPCD 300h AMD signature is before
link training.
2. and VBIOS will writes correctted AMD signature by after AGESA 1.0.0,
so driver doesn't need to write DPCD 300h again to switch config.
3. there are some OLED panel will clean backlight level to 0 once
receive DPCD 300h, so we will see flicker issue at that time.
[How]
read DPCD 300h signature before write it to avoid reprogram again, if
AMD signature was writed corrected by VBIOS, driver will not reprogram
it again.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Yu-ting Shen <Yu-ting.Shen@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
When switching single pipe to split pipe, the bandwidth check is just
for first pipe. The 2nd pipe with abnormal(or zero) dpp clock when pipe
ready and unlock leads the garbage on display.
[How]
Removed external increasing dpp clock check, the internal function
already loops all of pipes to check whether update dpp clock.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Chiawen Huang <chiawen.huang@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY]
We want to make enabling test pattern a part of the stream update code
path. This change is the first step towards that goal.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY]
DCN10 uses plane visibility to determine when to unblank HUBP; there is
no reason DCN20+ should not do the same.
[HOW]
In addition to changing the check in HWSEQ, we must change
is_pipe_tree_visible so that it checks ODM pipe topologies as well as
MPC. Since we're now checking both ODM and MPC topologies, the helper
function names have been changed to reference "parent" and "child"
instead of "top" and "bottom".
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
Driver apply the monitor patch into global debug flag. When apply
monitor patch, fec feature will always disable.
[How]
Separate fec debug flag and monitor patch
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Lewis Huang <Lewis.Huang@amd.com>
Reviewed-by: Wenjing Liu <Wenjing.Liu@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY]
The stack variable "val" is potentially unpopulate it, so initialize it
with the value 0xf (indicating an invalid mux)
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Reviewed-by: Martin Leung <Martin.Leung@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
In dc_link_dp_set_test_pattern, we assume all pipes have a stream, which
can cause null pointer dereference.
[How]
Add a null pointer check before accessing stream.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In the legacy ringbuffer submission, we still had an open-coded version
of intel_engine_stop_cs() with one additional verification step. Transfer
that verification to intel_engine_stop_cs() itself, and call it.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210113204709.15020-1-chris@chris-wilson.co.uk
Since we are system_highpri_wq, we expected the heartbeat to be
scheduled promptly. However, we see delays of over 10ms upsetting our
assertions. Accept this as inevitable and bump the minimum error
threshold to 20ms (from 6 jiffies).
<6> [616.784749] rcs0: Heartbeat delay: 3570us [2802, 9188]
<6> [616.807790] bcs0: Heartbeat delay: 2111us [745, 4372]
<6> [616.853776] vcs0: Heartbeat delay: 6485us [2424, 11637]
<3> [616.859296] vcs0: Heartbeat delay was 6485us, expected less than 6000us
<3> [616.860901] i915/intel_heartbeat_live_selftests: live_heartbeat_fast failed with error -22
v2: More context from CI.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210113163115.5740-2-chris@chris-wilson.co.uk
Devfreq framework supports 2 modes for monitoring devices.
Use delayed timer as default instead of deferrable timer
in order to monitor the GPU status regardless of CPU idle.
Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
Reviewed-by: Steven Price <steven.price@arm.com>
Signed-off-by: Steven Price <steven.price@arm.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210105164111.30122-1-lukasz.luba@arm.com
The field is only relevant for legacy DRM drivers. Its only non-legacy
user in the DRM core is in drm_file.c. This code is now protected by
CONFIG_DRM_LEGACY. Radeon, the only driver that used the field, has been
changed to maintain it's own copy.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210112081035.6882-7-tzimmermann@suse.de
CONFIG_DRM_VM gets selected by CONFIG_DRM_LEGACY, but nothing else. So
remove it and build drm_vm.o as part of CONFIG_DRM_LEGACY.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210112081035.6882-5-tzimmermann@suse.de
The function is declared in drm_cache.h. I also removed the curly
braces from the for loop to adhere to kernel coding style.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210112081035.6882-3-tzimmermann@suse.de
The AGP wrapper functions serve no purpose. They used to handle
builds that have CONFIG_AGP unset. But their callers are all in
drm_agpsupport.c, which only gets build with CONFIG_AGP.
v2:
* clarify CONFIG_AGP in commit description (Daniel)
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210112081035.6882-2-tzimmermann@suse.de
Moving vc4's mmap code from vc4_mmap() into a GEM object function
allows for the use drm_gem_mmap() and drm_gem_prime_mmap(). The content
of vc4_drm_fpos can then be generated by DEFINE_DRM_GEM_FOPS().
The actual mmap implementation is just a check if the BO is a validated
shader plus the default CMA mmap code.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Maxime Ripard <mripard@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210108140808.25775-4-tzimmermann@suse.de
Initialize all required entries from guc_set_default_submission, instead
of calling the execlists function. The previously inherited setup has
been copied over from the execlist code and simplified by removing the
execlists submission-specific parts.
v2: move setting of relative_mmio flag to engine_setup_common (Chris)
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: John Harrison <john.c.harrison@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> #v1
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20210113021236.8164-5-daniele.ceraolospurio@intel.com
Instead of starting the engine in execlists submission mode and then
switching to GuC, start directly in GuC submission mode. The initial
setup functions have been copied over from the execlists code
and simplified by removing the execlists submission-specific parts.
v2: remove unneeded unexpected starting state check (Chris)
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: John Harrison <john.c.harrison@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20210113021236.8164-4-daniele.ceraolospurio@intel.com
GuC owns the execlists state and the context IDs used for submission, so
the status of the ports and the CSB entries are not something we control
or can decode from the i915 side, therefore we can avoid dumping it. A
follow-up patch will also stop setting the csb pointers when using GuC
submission.
GuC dumps all the required events in the GuC logs when verbosity is set
high enough.
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: John Harrison <John.C.Harrison@Intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20210113021236.8164-3-daniele.ceraolospurio@intel.com
Huang Rui