IMU is a new firmware for GFX11.
There are four means by which firmware version can be queried
from the driver: device attributes, vf2pf, debugfs,
and the AMDGPU_INFO_FW_VERSION option in the amdgpu info ioctl.
Add IMU as an option for those four methods.
V2: Added debugfs
Reviewed-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: David Francis <David.Francis@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When using a device based on DCN32/321,
we have an issue where a second
4k@60Hz display does not light up,
and the system becomes unresponsive
for a few minutes. In the debug process,
it was possible to see a hang
in the function dcn20_post_unlock_program_front_end
in this part:
for (j = 0; j < TIMEOUT_FOR_PIPE_ENABLE_MS*1000
&& hubp->funcs->hubp_is_flip_pending(hubp); j++)
mdelay(1);
}
The hubp_is_flip_pending always returns positive
for waiting pending flips which is a symptom of
pipe hang. Additionally, the dmesg log shows
this message after a few minutes:
BUG: soft lockup - CPU#4 stuck for 26s!
...
[ +0.000003] dcn20_post_unlock_program_front_end+0x112/0x340 [amdgpu]
[ +0.000171] dc_commit_state_no_check+0x63d/0xbf0 [amdgpu]
[ +0.000155] ? dc_validate_global_state+0x358/0x3d0 [amdgpu]
[ +0.000154] dc_commit_state+0xe2/0xf0 [amdgpu]
This confirmed the hypothesis that we had a pipe
hanging somewhere. Next, after checking the
ftrace entries, we have the below weird
sequence:
[..]
2) | dcn10_lock_all_pipes [amdgpu]() {
2) 0.120 us | optc1_is_tg_enabled [amdgpu]();
2) | dcn20_pipe_control_lock [amdgpu]() {
2) | dc_dmub_srv_clear_inbox0_ack [amdgpu]() {
2) 0.121 us | amdgpu_dm_dmub_reg_write [amdgpu]();
2) 0.551 us | }
2) | dc_dmub_srv_send_inbox0_cmd [amdgpu]() {
2) 0.110 us | amdgpu_dm_dmub_reg_write [amdgpu]();
2) 0.511 us | }
2) | dc_dmub_srv_wait_for_inbox0_ack [amdgpu]() {
2) 0.110 us | amdgpu_dm_dmub_reg_read [amdgpu]();
2) 0.110 us | amdgpu_dm_dmub_reg_read [amdgpu]();
2) 0.110 us | amdgpu_dm_dmub_reg_read [amdgpu]();
2) 0.110 us | amdgpu_dm_dmub_reg_read [amdgpu]();
2) 0.110 us | amdgpu_dm_dmub_reg_read [amdgpu]();
2) 0.110 us | amdgpu_dm_dmub_reg_read [amdgpu]();
2) 0.110 us | amdgpu_dm_dmub_reg_read [amdgpu]();
[..]
We are not expected to read from dmub register
so many times and for so long. From the trace log,
it was possible to identify that the function
dcn20_pipe_control_lock was triggering the dmub
operation when it was unnecessary and causing
the hang issue. This commit drops the unnecessary
dmub code and, consequently, fixes the second display not
lighting up the issue.
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If the number of pages from the userptr BO differs from the SG BO then the
allocated memory for the SG table doesn't get freed before returning
-EINVAL, which may lead to a memory leak in some error paths. Fix this by
checking the number of pages before allocating memory for the SG table.
Fixes: 264fb4d332 ("drm/amdgpu: Add multi-GPU DMA mapping helpers")
Signed-off-by: Rafael Mendonca <rafaelmendsr@gmail.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>
MMHUB 2.1.x versions don't have ATCL2. Remove accesses to ATCL2 registers.
Since they are non-existing registers, read access will cause a
'Completer Abort' and gets reported when AER is enabled with the below patch.
Tagging with the patch so that this is backported along with it.
v2: squash in uninitialized warning fix (Nathan Chancellor)
Fixes: 8795e182b0 ("PCI/portdrv: Don't disable AER reporting in get_port_device_capability()")
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
[why]
MES response time in sriov may be longer than default value
due to reset or init in other VF. A timeout value specific
to sriov is needed.
[how]
When in sriov, adjust the timeout value to calculated
worst case scenario.
Signed-off-by: Yiqing Yao <yiqing.yao@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY]
0, original pstate X
1, ctx_A_create -> ctx_A->stable_pstate = X
2, ctx_A_set_pstate (Y) -> current pstate is Y (PEAK or STANDARD)
3, ctx_B_create -> ctx_B->stable_pstate = Y
4, ctx_A_destroy -> restore pstate to X
5, ctx_B_destroy -> restore pstate to Y
Above sequence will cause final pstate is wrong (Y), should be original X.
[HOW]
When ctx_B create,
if ctx_A touched pstate setting
(not auto, stable_pstate_ctx != NULL),
set ctx_B->stable_pstate the same value as ctx_A saved,
if stable_pstate_ctx == NULL,
fetch current pstate to fill
ctx_B->stable_pstate.
Signed-off-by: Chengming Gui <Jack.Gui@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Commit 8795e182b0 ("PCI/portdrv: Don't disable AER reporting in get_port_device_capability()")
uncovered a bug in amdgpu that required a reordering of the driver
init sequence to avoid accessing a special register on the GPU
before it was properly set up leading to an PCI AER error. This
reordering uncovered a different hw programming ordering dependency
in some APUs where the SDMA doorbells need to be programmed before
the GFX doorbells. To fix this, move the SDMA doorbell programming
back into the soc15 common code, but use the actual doorbell range
values directly rather than the values stored in the ring structure
since those will not be initialized at this point.
This is a partial revert, but with the doorbell assignment
fixed so the proper doorbell index is set before it's used.
Fixes: e3163bc8ff ("drm/amdgpu: move nbio sdma_doorbell_range() into sdma code for vega")
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: skhan@linuxfoundation.org
Cc: stable@vger.kernel.org
The two structs internal to struct panfrost_dump_object_header were
named, but sadly that is incompatible with C++, causing an error: "an
anonymous union may only have public non-static data members".
However nothing refers to struct pan_reg_hdr and struct pan_bomap_hdr
and there's no need to export these definitions, so lets drop them. This
fixes the C++ build error with the minimum change in userspace API.
Reported-by: Adrián Larumbe <adrian.larumbe@collabora.com>
Fixes: 730c2bf4ad ("drm/panfrost: Add support for devcoredump")
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Signed-off-by: Steven Price <steven.price@arm.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221017104602.142992-2-steven.price@arm.com
Commit 35a3b82f1b ("drm/connector: Introduce drmm_connector_init")
introduced the function drmm_connector_init() with a parameter for an
optional ddc pointer to the i2c controller used to access the DDC bus.
However, the underlying call to __drm_connector_init() was always
setting it to NULL instead of passing the ddc argument around.
This resulted in unexpected null pointer dereference on platforms
expecting to get a DDC controller.
Fixes: 35a3b82f1b ("drm/connector: Introduce drmm_connector_init")
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://lore.kernel.org/r/20221019143442.1798964-1-maxime@cerno.tech
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
The xrgb2101010 format conversion test (unlike for other formats) does
an endianness conversion on the results. However, it always converts
TEST_BUF_SIZE 32-bit integers, which results in reading from (and
writing to) more memory than in present in the result buffer. Instead,
use the buffer size, divided by sizeof(u32).
The issue could be reproduced with KASAN:
./tools/testing/kunit/kunit.py run --kunitconfig drivers/gpu/drm/tests \
--kconfig_add CONFIG_KASAN=y --kconfig_add CONFIG_KASAN_VMALLOC=y \
--kconfig_add CONFIG_KASAN_KUNIT_TEST=y \
drm_format_helper_test.*xrgb2101010
Reported-by: Linux Kernel Functional Testing <lkft@linaro.org>
Fixes: 4531143196 ("drm/format-helper: Add KUnit tests for drm_fb_xrgb8888_to_xrgb2101010()")
Signed-off-by: David Gow <davidgow@google.com>
Reviewed-by: Maíra Canal <mairacanal@riseup.net>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Reviewed-by: José Expósito <jose.exposito89@gmail.com>
Signed-off-by: Javier Martinez Canillas <javierm@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221019073239.3779180-1-davidgow@google.com
A user reported a bug on CAPE VERDE system where uvd_v3_1
IP component failed to initialize as there is an issue with
BO move code from one memory to other.
In function amdgpu_mem_visible() called by amdgpu_bo_move(),
when there are no blocks to compare or if we have a single
block then break the loop.
Fixes: 312b4dc11d ("drm/amdgpu: Fix VRAM BO swap issue")
Signed-off-by: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
If mes is not dequeued during fini, mes will be in an uncleaned state
during reload, then mes couldn't receive some commands which leads to
reload failure.
[How]
Perform MES dequeue via MMIO after all the unmap jobs are done by mes
and before kiq fini.
v2: Move the dequeue operation inside kiq_hw_fini.
Signed-off-by: YuBiao Wang <YuBiao.Wang@amd.com>
Reviewed-by: Jack Xiao <Jack.Xiao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When booting a kernel compiled with CONFIG_CFI_CLANG on a machine with
an RX 6700 XT, there is a CFI failure in kfd_destroy_mqd_cp():
[ 12.894543] CFI failure at kfd_destroy_mqd_cp+0x2a/0x40 [amdgpu] (target: hqd_destroy_v10_3+0x0/0x260 [amdgpu]; expected type: 0x8594d794)
Clang's kernel Control Flow Integrity (kCFI) makes sure that all
indirect call targets have a type that exactly matches the function
pointer prototype. In this case, hqd_destroy()'s third parameter,
reset_type, should have a type of 'uint32_t' but every implementation of
this callback has a third parameter type of 'enum kfd_preempt_type'.
Update the function pointer prototype to match reality so that there is
no more CFI violation.
Link: https://github.com/ClangBuiltLinux/linux/issues/1738
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Building 32-bit images may fail with the following error.
drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_util_32.c:
In function ‘dml32_UseMinimumDCFCLK’:
drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_util_32.c:3142:1:
error: the frame size of 1096 bytes is larger than 1024 bytes
This is seen when building i386:allmodconfig with any of the following
compilers.
gcc (Debian 12.2.0-3) 12.2.0
gcc (Ubuntu 9.4.0-1ubuntu1~20.04.1) 9.4.0
The problem is not seen if the compiler supports GCC_PLUGIN_LATENT_ENTROPY
because in that case CONFIG_FRAME_WARN is already set to 2048 even for
32-bit builds.
dml32_UseMinimumDCFCLK() was introduced with commit dda4fb85e4
("drm/amd/display: DML changes for DCN32/321"). It declares a large
number of local variables. Increase the frame size for the affected
file to 2048, similar to other files in the same directory, to enable
32-bit build tests with affected compilers.
Fixes: dda4fb85e4 ("drm/amd/display: DML changes for DCN32/321")
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Reported-by: Łukasz Bartosik <ukaszb@google.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
In SRIOV multi-vf, dpm is always disabled, and pm_attr_list won't
be initialized. There will be a NULL pointer call trace after
removing the dpm check condition in amdgpu_pm_sysfs_fini.
BUG: kernel NULL pointer dereference, address: 0000000000000000
RIP: 0010:amdgpu_device_attr_remove_groups+0x20/0x90 [amdgpu]
Call Trace:
<TASK>
amdgpu_pm_sysfs_fini+0x2f/0x40 [amdgpu]
amdgpu_device_fini_hw+0xdf/0x290 [amdgpu]
[How]
List pm_attr_list should be initialized when dpm is disabled.
Fixes: a6ad27cec5 ("drm/amd/pm: Remove redundant check condition")
Signed-off-by: ZhenGuo Yin <zhenguo.yin@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
temporarily disable thermal alert on smu_v13_0_10 due to kfd test fail.
will enable it again after confirming the thermal hardware setting.
Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 16fb4dca95.
Unfortunately, that commit causes fan monitors can't be read and written
properly.
Fixes: 16fb4dca95 ("drm/amdgpu: getting fan speed pwm for vega10 properly")
Signed-off-by: Asher Song <Asher.Song@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
- refactor mode2 on v11.0.7 to align with aldebaran
- comment out using mode2 reset as default for now, will introduce
another controller to replace previous reset_level_mask
v2: squash in unused variable removal (Alex)
Signed-off-by: Victor Zhao <Victor.Zhao@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit dac6b80818.
This commit reverted the AMDGPU_SKIP_MODE2_RESET as it conflicts with
the original design of reset handler. Will redesign it.
Fixes: dac6b80818 ("drm/amdgpu: let mode2 reset fallback to default when failure")
Signed-off-by: Victor Zhao <Victor.Zhao@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 5bd8d53f6f.
This commit breaks the reset logic for aldebaran, revert it for now.
Will move the mask inside the reset handler.
Fixes: 5bd8d53f6f ("drm/amdgpu: add debugfs amdgpu_reset_level")
Signed-off-by: Victor Zhao <Victor.Zhao@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For asic with VF MMIO access protection avoid using CPU for VM table updates.
CPU pagetable updates have issues with HDP flush as VF MMIO access protection
blocks write to mmBIF_BX_DEV0_EPF0_VF0_HDP_MEM_COHERENCY_FLUSH_CNTL register
during sriov runtime.
v3: introduce virtualization capability flag AMDGPU_VF_MMIO_ACCESS_PROTECT
which indicates that VF MMIO write access is not allowed in sriov runtime
Signed-off-by: Danijel Slivka <danijel.slivka@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
