Jack Xiao [Fri, 20 Mar 2020 04:10:00 +0000 (12:10 +0800)]
drm/amdgpu/gfx10: inherit vmid from mqd
For MES manages vmid assignment, let vmid inherit from mqd instead of
ib packet setting.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Fri, 20 Mar 2020 04:03:08 +0000 (12:03 +0800)]
drm/amdgpu/gfx10: associate mes queue id with fence v2
Associate mes queue id with fence, so that EOP trap handler can look up
which queue has issued the fence.
v2: move mes queue flag to amdgpu_mes_ctx.h
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Fri, 20 Mar 2020 03:37:31 +0000 (11:37 +0800)]
drm/amdgpu/gfx10: use per ctx CSA for de metadata
As MES requires per context preemption, use per context CSA address
for DE metadata to correctly enable context MCBP preemption.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Fri, 20 Mar 2020 03:10:20 +0000 (11:10 +0800)]
drm/amdgpu/gfx10: use per ctx CSA for ce metadata
As MES requires per context preemption, use per context CSA address
for CE metadata to correctly enable context MCBP preemption.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Wed, 1 Jul 2020 05:32:36 +0000 (13:32 +0800)]
drm/amdgpu/gfx10: implement mqd functions of gfx/compute eng v2
Refine the existing gfx/compute mqd functions, and add them
to engine mqd layer.
v2: rebase fix.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Fri, 20 Mar 2020 05:59:48 +0000 (13:59 +0800)]
drm/amdgpu: assign the cpu/gpu address of fence from ring
assign the cpu/gpu address of fence for the normal or mes ring
from ring structure.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Wed, 30 Mar 2022 14:04:38 +0000 (22:04 +0800)]
drm/amdgpu: initialize/finalize the ring for mes queue
Iniailize/finalize the ring for mes queue which submits the command
stream to the mes-managed hardware queue.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Fri, 20 Mar 2020 02:54:45 +0000 (10:54 +0800)]
drm/amdgpu: use ring structure to access rptr/wptr v2
Use ring structure to access the cpu/gpu address of rptr/wptr.
v2: merge gfx10/sdma5/sdma5.2 patches
Signed-off-by: Jack Xiao <Jack.Xiao@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>
Jack Xiao [Thu, 11 Jun 2020 03:27:47 +0000 (11:27 +0800)]
drm/amdgpu: define ring structure to access rptr/wptr/fence
Define ring structure to access the cpu/gpu address of rptr/wptr/fence
instead of dynamic calculation.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Mon, 16 Mar 2020 09:12:23 +0000 (17:12 +0800)]
drm/amdgpu: add mes ctx data in amdgpu_ring
Add mes context data structure in amdgpu_ring.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Thu, 26 Mar 2020 16:38:05 +0000 (00:38 +0800)]
drm/amdgpu: add the per-context meta data v3
The per-context meta data is a per-context data structure associated
with a mes-managed hardware ring, which includes MCBP CSA, ring buffer
and etc.
v2: fix typo
v3: a. use structure instead of typedef
b. move amdgpu_mes_ctx_get_offs_* to amdgpu_ring.h
c. use __aligned to make alignement
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Wed, 1 Jul 2020 04:23:06 +0000 (12:23 +0800)]
drm/amdgpu: add helper function to initialize mqd from ring v4
Add the helper function to initialize mqd from ring configuration.
v2: use if/else pair instead of ?/: pair
v3: use simpler way to judge hqd_active
v4: fix parameters to amdgpu_gfx_is_high_priority_compute_queue
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jack Xiao [Wed, 1 Jul 2020 03:48:52 +0000 (11:48 +0800)]
drm/amdgpu: define MQD abstract layer for hw ip
Define MQD abstract layer for hw ip, for the passing
mqd configuration not only from ring but more sources,
like user queue.
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alice Wong [Mon, 2 May 2022 15:40:18 +0000 (11:40 -0400)]
drm/amdgpu/ucode: Remove firmware load type check in amdgpu_ucode_free_bo
When psp_hw_init failed, it will set the load_type to AMDGPU_FW_LOAD_DIRECT.
During amdgpu_device_ip_fini, amdgpu_ucode_free_bo checks that load_type is
AMDGPU_FW_LOAD_DIRECT and skips deallocating fw_buf causing memory leak.
Remove load_type check in amdgpu_ucode_free_bo.
Signed-off-by: Alice Wong <shiwei.wong@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This issue happens because "pipe_ctx->stream_res.tg"
needs to be initialized first before reading its members.
This commit fixes this issue by properly initializing
the pointer before accessing the target data.
Evan Quan [Wed, 6 Apr 2022 21:44:44 +0000 (17:44 -0400)]
drm/amd/pm: enable SCPM support for SMU
With SCPM enabled, the pptable used will be signed. It cannot
be used directly by driver. To get the raw pptable, we need to
rely on the combo pptable(and its revelant SMU message).
Also, the pptable transferring(to SMU) will be performed by PSP.
Some SMU messages will be not available to driver any more.
Signed-off-by: Evan Quan <evan.quan@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Evan Quan [Thu, 31 Mar 2022 22:10:35 +0000 (18:10 -0400)]
drm/amd/pm: move bootup values retrieving to ->sw_init
Firsrt of all, the operations involved is to interact with
VBIOS. They are fully supported at ->sw_init phase.
Secondly, the new mechanism to upload pptable to SMU is
introduced. With the new mechanism, the pptable transferring
has to be via PSP. That requires the pptable ucode(and necessary
bootupp values retrieving) must be ready before ->hw_init phase of PSP.
Signed-off-by: Evan Quan <evan.quan@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alice Wong [Thu, 28 Apr 2022 01:03:54 +0000 (21:03 -0400)]
drm/amdgpu/psp: deallocate memory when psp_load_fw failed
psp_load_fw failure would cause memory leak for psp tmr and psp ring
because psp_hw_init is not called as psp block is not fully initialized.
Clean up psp tmr and psp ring when psp_load_fw fail by calling
psp_free_shared_bufs and psp_ring_destroy.
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alice Wong <shiwei.wong@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Elena Sakhnovitch [Fri, 11 Mar 2022 19:03:09 +0000 (14:03 -0500)]
drm/amd/pm: Disable fan control if not supported
On Sienna Cichild, not all platforms use PMFW based fan control
(ex: fanless systems). On such ASICs fan control by PMFW will be
disabled in PPTable. Disable hwmon knobs for fan control also as
it is not possible to report or control fan speed on such platforms
through driver.
v3: FeaturesToRun casted as uint64_t
Signed-off-by: Elena Sakhnovitch <elena.sakhnovitch@amd.com> Reviewed-by: Lijo Lazar <lijo.lazar@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Thu, 21 Apr 2022 05:21:52 +0000 (01:21 -0400)]
drm/amdgpu/psp: move PSP memory alloc from hw_init to sw_init
Memory allocations should be done in sw_init. hw_init should
just be hardware programming needed to initialize the IP block.
This is how most other IP blocks work. Move the GPU memory
allocations from psp hw_init to psp sw_init and move the memory
free to sw_fini. This also fixes a potential GPU memory leak
if psp hw_init fails.
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Fri, 29 Apr 2022 00:17:38 +0000 (20:17 -0400)]
drm/amdgpu: Free user pages if kvmalloc_array fails
To cleanup the BOs of bo_list which have got user pages.
Signed-off-by: Philip Yang <Philip.Yang@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/smu: Increace dpm level count only for smu v13.0.2
Only V13.0.2 on SMU v13 will get 0 based max level from fw and
increment by one, other ASIC will not need for this.
V2: replace the asic_type check with ip versioning check.
Harry Wentland [Tue, 19 Apr 2022 17:03:12 +0000 (13:03 -0400)]
drm/amd/display: Avoid reading audio pattern past AUDIO_CHANNELS_COUNT
A faulty receiver might report an erroneous channel count. We
should guard against reading beyond AUDIO_CHANNELS_COUNT as
that would overflow the dpcd_pattern_period array.
Signed-off-by: Harry Wentland <harry.wentland@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Wed, 27 Apr 2022 22:50:49 +0000 (18:50 -0400)]
drm/amdgpu: Free user pages if amdgpu_cs_parser_bos failed
Otherwise userspace resubmit the BOs again will trigger kernel WARNING
and fail the command submission.
Signed-off-by: Philip Yang <Philip.Yang@amd.com> Tested-by: Robert Święcki <robert@swiecki.net> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Candice Li [Wed, 27 Apr 2022 10:02:45 +0000 (18:02 +0800)]
drm/amdgpu: Fix build warning for TA debugfs interface
Remove the redundant codes to fix build warning
when CONFIG_DEBUG_FS is disabled.
Reported-by: Randy Dunlap <rdunlap@infradead.org> Signed-off-by: Candice Li <candice.li@amd.com> Reviewed-by: Yang Wang <kevinyang.wang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Changed from v1:
Switch WREG32/RREG32_PCIE to use indirect reg access
helper for sco15 and onwards
Acked-by: Christian König <christian.koenig@amd.com> 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>
Stanley.Yang [Wed, 4 Aug 2021 07:43:17 +0000 (15:43 +0800)]
drm/amdgpu: add new write field for soc21
add new write field macro to handle soc21
registers with reg prefix
Acked-by: Christian König <christian.koenig@amd.com> 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>
Hawking Zhang [Mon, 28 Feb 2022 10:35:02 +0000 (18:35 +0800)]
drm/amdgpu: switch to atomfirmware_asic_init
Some initial settings now are not available from
the atom data table. The assumption that !ps[0]
|| !ps[1] in amdgpu_atom_asic_init is not valid.
In addition, driver needs to strictly follow
atomfirmware structure (asic_init_parameters) to
initialize parameters used to execute asic_init
function, otherwise, the execution of asic_init
would fail.
This shall be applicable to all soc15 adapters,but
let make the transition on soc21 first.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
projects / users / jedix / linux-maple.git / log