From: David Laight
[ Upstream commit 2b97aaf74ed534fb838d09867d09a3ca5d795208 ]
The bodies of __signed_type_use() and __unsigned_type_use() are much the
same size as their names - so put the bodies in the only line that expands
them.
Similarly __signed_type() is defined separately for 64bit and
[Why]
Newer VPE microcode has functionality that will decrease DPM level
only when a workload has run for 2 or more seconds. If VPE is turned
off before this DPM decrease, the SOC can get stuck with a higher
DPM level.
This can happen from amdgpu's ring buffer test because it's a short
quick work
https://github.com/intel-lab-lkp/linux/commits/Tvrtko-Ursulin/drm-ttm-Add-getter-for-some-pool-properties/20251010-052711
base: git://anongit.freedesktop.org/drm/drm drm-next
patch link:
https://lore.kernel.org/r/20251008115314.55438-3-tvrtko.ursulin%40igalia.com
patch subject: [PATCH v3 2/
Add a query for compute queues. Userspace can use this to
query the size of the EOP buffers for compute user queues.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 26 +
include/uapi/drm/amdgpu_drm.h | 8
2 files changed, 34
The CSA and EOP buffers have different alignement requirements.
Hardcode them for now as a bug fix. A proper query will be added in
a subsequent patch.
Fixes: 9e46b8bb0539 ("drm/amdgpu: validate userq buffer virtual address and
size")
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/
These were never used and are duplicated with the
interface that is used. Maybe leftovers from a previous
revision of the patch that added them.
Fixes: 90c448fef312 ("drm/amdgpu: add new AMDGPU_INFO subquery for userq
objects")
Signed-off-by: Alex Deucher
---
include/uapi/drm/amdgpu_drm.h | 21
If active queue buffer is freed, kfd_lookup_process_by_mm return NULL,
means process exited and mm is gone, it is fine to evict queue then
free queue buffer CPU mapping and memory from do_exit.
Only show warning message if process mm is still alive when queue
buffer is freed.
Fixes: b049504e211e
1. Added VF logic to init data exchange region using the offsets from
dynamic(v2) critical regions;
Signed-off-by: Ellen Pan
---
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 88
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h | 1 +
2 files changed, 77 insertions(+), 12 deletio
By design the MES will return an array result that is twice the number
of hung doorbells it can report.
i.e. if up k reported doorbells are supported, then the
second half of the array, also of length k, holds the HQD information
(type/queue/pipe) where queue 1 corresponds to index 0 and k,
queue
On Fri, Oct 10, 2025 at 11:06 AM Lijo Lazar wrote:
>
> For legacy SOCs, discovery binary is sideloaded. Instead of checking for
> binary blob, use a flag to determine if discovery region needs to be
> reserved.
>
> Signed-off-by: Lijo Lazar
Reviewed-by: Alex Deucher
> ---
> drivers/gpu/drm/am
On Fri, Oct 10, 2025 at 11:11 AM Lijo Lazar wrote:
>
> Add amdgpu_discovery_info structure to keep all discovery related
> information.
>
> Signed-off-by: Lijo Lazar
> ---
>
> v2: Remove superfluous discovery_ prefix for members (Alex)
>
> drivers/gpu/drm/amd/amdgpu/amdgpu.h | 10 +-
>
On Fri, Oct 10, 2025 at 1:01 AM Ellen Pan wrote:
>
> 1. Added VF logic to init data exchange region using the offsets from
> dynamic(v2) critical regions;
>
> Signed-off-by: Ellen Pan
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 88
> drivers/gpu/drm/amd/amdgpu/am
For legacy SOCs, discovery binary is sideloaded. Instead of checking for
binary blob, use a flag to determine if discovery region needs to be
reserved.
Signed-off-by: Lijo Lazar
---
drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 10 -
drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h | 1 +
Add amdgpu_discovery_info structure to keep all discovery related
information.
Signed-off-by: Lijo Lazar
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 8 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 7 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c| 4 +-
drivers/gpu/drm/amd
On 10. 10. 25 14:28, Christian König wrote:
David any objections that I take this patch and make the necessary
modifications?
Sure, please go ahead.
Thanks,
David
People are pinging me about the problem.
Regards,
Christian.
On 09.10.25 17:04, Christian König wrote:
FYI
On 09.10.25 09
On Wed, 2025-10-08 at 09:53 +0100, Tvrtko Ursulin wrote:
> This time round we explore the rate of submitted job queue processing
> with multiple identical parallel clients.
>
> Example test output:
>
> 3 clients:
> t cycle: min avg max : ...
> + 0ms
From: Wang Jiang
The audio detection process in the Radeon driver is as follows:
radeon_dvi_detect/radeon_dp_detect -> radeon_audio_detect ->
radeon_audio_enable -> radeon_audio_component_notify ->
radeon_audio_component_get_eld
When HDMI is unplugged, radeon_dvi_detect is triggered.
At this po
[AMD Official Use Only - AMD Internal Distribution Only]
>-Original Message-
>From: Pan, Ellen
>Sent: Friday, October 10, 2025 10:13 AM
>To: amd-gfx@lists.freedesktop.org
>Cc: Deucher, Alexander ; Koenig, Christian
>; Lazar, Lijo ; Chan, Hing
>Pong ; Pan, Ellen
>Subject: [PATCH 4/6] drm/
Add a DECLARE_DYNDBG_CLASSMAP stub/do-nothing defn, so that callers
will compile properly with CONFIG_DYNAMIC_DEBUG=n
Signed-off-by: Jim Cromie
---
include/linux/dynamic_debug.h | 39 ++-
1 file changed, 20 insertions(+), 19 deletions(-)
diff --git a/include/linu
struct ddebug_class_param keeps a ref to the state-storage of the
param; make both class-types use the same unsigned long storage type.
ISTM this is simpler and safer; it avoids an irrelevant difference,
and if 2 users somehow get class-type mixed up (or refer to the wrong
union member), at least
When writing queries to >control, flags are parsed 1st, since they are
the only required field, and they require specific compositions. So
if the flags draw an error (on those specifics), then keyword errors
aren't reported. This can be mildly confusing/annoying, so explain it
instead.
cc: linux
echo 1000 > /sys/module/test_dynamic_debug/parameters/do_prints
This allows its use as a scriptable load generator, to generate
dynamic-prefix-emits for flag combinations vs undecorated messages.
This will make it easy to assess the cost of the prefixing.
Reading the ./do_prints node also prints
Since commit
85f7f6c0edb8 ("dynamic_debug: process multiple debug-queries on a line")
Multi-query commands have been allowed:
modprobe drm dyndbg="class DRM_UT_CORE +p; class DRM_UT_KMS +p"
modprobe drm dyndbg=<
[ 203.902703] dyndbg: query parse failed
[ 203.902871] dyndbg: processed 2 quer
This new test-fn runs 3 module/submodule modprobe scenarios, variously
using both the generic dyndbg= modprobe arg, and the
test-module's classmap-params to manipulate the test-mod*'s pr_debugs.
In all cases, the current flag-settings are counted and tested vs
expectations.
The 3rd scenario recapi
dynamic-debug has several __sections, each with ,
num_, and it iterates over these with a 2-index for-loop.
These loops are fiddly with the 2 names.
We have only 2 such loops now, but are getting more soon; lets
embed/abstract the fiddlyness in the for_subvec() macro, and avoid
repeating it going
recompose struct _ddebug_info, inserting proper sub-structs.
The struct currently has 2 pairs of ptr,length fields distinguished by
name. Several for-loops operate on these field pairs, soon many more
will be added.
Looping over these blocks by respective field-pairs is repetitive and
fiddly, di
26 matches
Mail list logo
