[why]
Updating the cursor enablement register can be a slow operation and accumulates
when high polling rate cursors cause frequent updates asynchronously to the
cursor position.
[how]
Since the cursor enable bit is cached there is no need to update the
enablement register if there is no change to it. This removes the
read-modify-write from the cursor position programming path in HUBP and
DPP, leaving only the register writes.
Reviewed-by: Josip Pavic <josip.pavic@amd.com>
Signed-off-by: Aric Cyr <Aric.Cyr@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
To ensure DML validation receives the correct tiling information,
such as swizzle mode or array mode, based on the active GFX format
[How]
- For new GFX format passed swizzle_mode to DML.
- For legacy GFX format passed array_mode to DML.
- Dynamically determined the appropriate tiling info based on the
active GFX format.
[Description]
This commit ensures that the correct GFX tiling information is passed
to DML. Depending on the active GFX format, the appropriate tiling info
is passed to DML. This change accommodates the different GFX formats
supported by latest platforms, ensuring compatibility and proper
DML validation.
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Signed-off-by: Karthi Kandasamy <karthi.kandasamy@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY]
The `dc_tiling_info` union previously did not have a field to
specify the active GFX format, assuming only one format would
be used per DCN version. from DCN4+, support for switching
between different GFX formats is introduced, requiring a way
to track which format is currently in use.
[HOW]
Updated the `dc_tiling_info` union to include a new field that
explicitly indicates the currently used GFX format.
This allows the system to determine the active GFX format
and take the correct programming path accordingly.
[Description]
The union `dc_tiling_info` has been updated to support multiple
GFX formats by adding a new field for identifying the active format.
This update ensures that the correct programming path is followed
based on the selected format. All references to `dc_tiling_info`
in the codebase have been updated to reflect the new structure.
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Signed-off-by: Karthi Kandasamy <karthi.kandasamy@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[why]
initialize the power state for dc use,
but dc_set_power_state it not called at D3.
It would cause can't recognize last power state
[how]
remove initialize the power state for dc use, it is not necessary.
Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Brandon Syu <Brandon.Syu@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY]
We like to have pretrain for dpia link so that dp and dp tunneling
have aligned behavior. The Main difficult for dpia pretrain is that
encoder can not get corresponded dpia port when link detection
in current implementation.
[HOW]
1. create enable/disable dpia output functions for dcn35 encoder
and have dpia_id and other necessary info as inputs.
2. dcn35 dpia use the new functions to enable/disable output.
3. have a option to enable/disable the change.
Reviewed-by: Wenjing Liu <wenjing.liu@amd.com>
Reviewed-by: Meenakshikumar Somasundaram <meenakshikumar.somasundaram@amd.com>
Signed-off-by: Peichen Huang <PeiChen.Huang@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
No common dsc params found between encoder and decoder is one
of the reason that could prevent dsc from properly enabled.
Dump the params to a specific timing to help locate possible
invalid dsc params in either encoder or decoder side.
Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Fangzhi Zuo <Jerry.Zuo@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
MAX/MIN macros maybe defined already, hence add a guard around them to
prevent errors that complain about redefinition like:
drivers/gpu/drm/amd/amdgpu/../dal-dev/modules/hdcp/hdcp_ddc.c:31: error: "MIN" redefined [-Werror]
31 | #define MIN(a, b) ((a) < (b) ? (a) : (b))
|
In file included from ./include/linux/kernel.h:28,
from ./include/linux/cpumask.h:11,
from ./include/linux/smp.h:13,
from ./include/linux/lockdep.h:14,
from ./include/linux/spinlock.h:63,
from ./include/linux/mmzone.h:8,
from ./include/linux/gfp.h:7,
from ./include/linux/slab.h:16,
from drivers/gpu/drm/amd/amdgpu/../display/dc/os_types.h:37,
from drivers/gpu/drm/amd/amdgpu/../display/modules/inc/mod_hdcp.h:29,
from drivers/gpu/drm/amd/amdgpu/../display/modules/hdcp/hdcp.h:29,
from drivers/gpu/drm/amd/amdgpu/../display/modules/hdcp/hdcp_ddc.c:29:
./include/linux/minmax.h:329: note: this is the location of the previous definition
329 | #define MIN(a,b) __cmp(min,a,b)
|
cc1: all warnings being treated as errors
Reviewed-by: Sun peng Li <sunpeng.li@amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why & How]
We actually have the capability to calculate independent CRC for 2 crc window
at the same time. Extend dm with the capability by having array to
configure/maintain multiple crc windows. Add the flexibility but use 1st CRC
instance only for now. Can change to use the 2nd CRC instance if needed.
Reviewed-by: HaoPing Liu <haoping.liu@amd.com>
Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why & How]
Since now we can set multiple crc windows for secure display, add a new input
parameter for dc_stream_get_crc to indicate to fetch crc from which crc
engine.
Reviewed-by: HaoPing Liu <haoping.liu@amd.com>
Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why & How]
Have to support multiple CRC windows setting to dmub. Add new dmub forward
functions for supporting/forwarding multiple crc windows setting to dmub.
Reviewed-by: HaoPing Liu <haoping.liu@amd.com>
Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
For DP HBlank expansion/reduction, the HBlank parameters of the original
EDID timing needs to be notified to the sink in order for the timing to
be reduced back to the original HBlank size.
[How]
Add parameter in dc_crtc_timing to track the increased HBlank.
Reviewed-by: Michael Strauss <michael.strauss@amd.com>
Signed-off-by: George Shen <george.shen@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
Some features, such as HBlank expansion/reduction, needs to know how
much HBlank is required to support basic audio.
[How]
Add interface to link to calculate required HBlank size for a given
link + timing combination to support basic audio (i.e. 2-channel 48KHz).
Reviewed-by: Wenjing Liu <wenjing.liu@amd.com>
Signed-off-by: George Shen <george.shen@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
DPCD register RECEIVE_PORT0_CAP contains HBlank expansion/reduction
capabilities of a DP device. These capabilities are required to enable
HBlank expansion/reduction logic.
[How]
Read raw RECEIVE_PORT0_CAP register values and store parsed fields.
Reviewed-by: Wenjing Liu <wenjing.liu@amd.com>
Signed-off-by: George Shen <george.shen@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY & HOW]
Driver disable will deallocate framebuffer to reset IPS state, this will cause IPS start with
INIT state to blindly power gate ONO region to break power sequence. All the gating blocks
should be powered up when releasing hw to ensure all the power optimizations are identical to pre-OS.
Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Duncan Ma <duncan.ma@amd.com>
Signed-off-by: Yihan Zhu <Yihan.Zhu@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Commit 2563391e57 ("drm/amd/display: DML2.1 resynchronization") blew
away the compiler warning fix from commit 2fde4fdddc
("drm/amd/display: Avoid -Wenum-float-conversion in
add_margin_and_round_to_dfs_grainularity()"), causing the warning to
reappear.
drivers/gpu/drm/amd/amdgpu/../display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.c:183:58: error: arithmetic between enumeration type 'enum dentist_divider_range' and floating-point type 'double' [-Werror,-Wenum-float-conversion]
183 | divider = (unsigned int)(DFS_DIVIDER_RANGE_SCALE_FACTOR * (vco_freq_khz / clock_khz));
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~
1 error generated.
Apply the fix again to resolve the warning.
Re-apply again after commit be4e350931 ("drm/amd/display: DML21 Reintegration For Various Fixes")
This should be making its way back to the original DML trees this time. (Alex)
Fixes: be4e350931 ("drm/amd/display: DML21 Reintegration For Various Fixes")
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3841
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY & HOW]
commit 7fb363c575 ("drm/amd/display: Let drm_crtc_vblank_on/off manage interrupts")
lets drm_crtc_vblank_* to manage interrupts in amdgpu_dm_crtc_set_vblank,
and amdgpu_irq_get/put do not need to be called here. Part of that
patch got lost somehow, so fix it up.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Before, every time fdinfo is queried we try to lock all the BOs in the
VM and calculate memory usage from scratch. This works okay if the
fdinfo is rarely read and the VMs don't have a ton of BOs. If either of
these conditions is not true, we get a massive performance hit.
In this new revision, we track the BOs as they change states. This way
when the fdinfo is queried we only need to take the status lock and copy
out the usage stats with minimal impact to the runtime performance. With
this new approach however, we would no longer be able to track active
buffers.
Signed-off-by: Yunxiang Li <Yunxiang.Li@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241219151411.1150-6-Yunxiang.Li@amd.com
Signed-off-by: Christian König <christian.koenig@amd.com>
If IH primary ring and KFD ih fifo overflows, we may miss CP, SDMA
interrupts and cause application soft hang. Show warning message with
ring name if overflow happens.
Add function to get ih ring name to avoid duplicating it. To keep
warning message consistent between GPU generations, change all
*_ih.c except ASICs older than Vega which has only one ih ring.
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>
If event slot is not signaled, kfd_signal_event_interrupt goes to slow
path to scan all event slots to find the signaled event, this is needed
for old ASICs that don't have the event ID or the event IDs are
incorrect in the IH payload.
There is case that GPU signal the same event twice, then driver process
the first event interrupt, set_event and event slot is auto-reset, then
for the second event interrupt, KFD goes to slow path as event is not
signaled, just drop the second event interrupt because the application
only need wakeup once.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For CPX mode, each KFD node has interrupt worker to process ih_fifo to
send events to user space. Currently all interrupt workers of same adev
queue to same CPU, all workers execution are actually serialized and
this cause KFD ih_fifo overflow when CPU usage is high.
Use per-GPU unbounded highpri queue with number of workers equals to
number of partitions, let queue_work select the next CPU round robin
among the local CPUs of same NUMA.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
After GPU page fault, there are lots of page fault interrupts generated
at short period even with CAM filter enabled because the fault address
is different. Each page fault copy to KFD ih fifo to send event to user
space by KFD interrupt worker, this could cause KFD ih fifo overflow
while other processes generate events at same time.
KFD process is aborted after GPU page fault, we only need one GPU page
fault interrupt sent to KFD ih fifo to send memory exception event to
user space.
Incease KFD ih fifo size to 2 times of IH primary ring size, to handle
the burst events case.
This patch handle the gfx v9 path, cover retry on/off and CAM filter
on/off cases.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
To handle 40000 to 80000 interrupts per second running CPX mode with 4
streams/queues per KFD node, KFD interrupt handler becomes the
performance bottleneck.
Remove the kfifo_out memcpy overhead by accessing ih_fifo data in-place
and updating rptr with kfifo_skip_count.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This partially reverts commit 194eb174cb.
This commit introduced a new state variable into adev without even
remotely worrying about CPU barriers.
Since we already have the amdgpu_in_reset() function exactly for this
use case partially revert that.
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>
The sdma context empty interrupt is dropped in amdgpu_irq_dispatch
as unregistered interrupt src_id 243, this interrupt accounts to 1/3 of
total interrupts and causes IH primary ring overflow when running
stressful benchmark application. Disable this interrupt has no side
effect found.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The use of 1 << i in scheduler mask calculations can result in an
unintentional integer overflow due to the expression being
evaluated as a 32-bit signed integer.
This patch replaces 1 << i with 1ULL << i to ensure the operation
is performed as a 64-bit unsigned integer, preventing overflow
Discovered in coverity scan, CID 1636393, 1636175, 1636007, 1635853
Fixes: c5c63d9cb5 ("drm/amdgpu: add amdgpu_gfx_sched_mask and amdgpu_compute_sched_mask debugfs")
Signed-off-by: Karol Przybylski <karprzy7@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
