Am 27.09.22 um 05:37 schrieb Asher Song:
This patch is used to fix following compiling error that occurs in some
old kernels.
error: ‘for’ loop initial declarations are only allowed in C99 mode
for (int i = 0; i < dc->res_pool->res_cap->num_dsc; i++) {
Signed-off-by: Asher Song
Reviewed-b
Am 26.09.22 um 23:40 schrieb Shashank Sharma:
This patch switches the GPU workload mode to/from
compute mode, while submitting compute workload.
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
Feel free to add my acked-by, but Felix should probably take a look as well.
Christian.
Am 26.09.22 um 23:40 schrieb Shashank Sharma:
This patch and switches the GPU workload based profile based
on the workload hint information saved in the workload context.
The workload profile is reset to NONE when the job is done.
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
Re
Am 26.09.22 um 23:40 schrieb Shashank Sharma:
This patch adds new IOCTL flags in amdgpu_context_IOCTL to set
GPU workload profile. These calls will allow a user to switch
to a GPU power profile which might be better suitable to its
workload type. The currently supported workload types are:
"
Am 26.09.22 um 23:40 schrieb Shashank Sharma:
This patch adds new functions which will allow a user to
change the GPU power profile based a GPU workload hint
flag.
Cc: Alex Deucher
Signed-off-by: Shashank Sharma
Alex needs to take a closer look at this stuff, but feel free to add my
acked-b
Am 26.09.22 um 23:40 schrieb Shashank Sharma:
Allow the user to specify a workload hint to the kernel.
We can use these to tweak the dpm heuristics to better match
the workload for improved performance.
V3: Create only set() workload UAPI (Christian)
Signed-off-by: Alex Deucher
Signed-off-by:
Hey Michel,
JIadong is working on exposing high/low priority gfx queues for gfx9 and
older hw generations by using mid command buffer preemption.
I know that you have been working on Gnome Mutter to make use from
userspace for this. Do you have time to run some tests with that?
Thanks,
Chri
Reviewed-by: Guchun Chen
Regards,
Guchun
-Original Message-
From: Song, Asher
Sent: Tuesday, September 27, 2022 11:38 AM
To: Chen, Guchun ; Cui, Flora ; Shi,
Leslie ; Ma, Jun ;
amd-gfx@lists.freedesktop.org
Cc: Song, Asher
Subject: [PATCH] drm/amdgpu: fix a compiling error in old ke
This patch is used to fix following compiling error that occurs in some
old kernels.
error: ‘for’ loop initial declarations are only allowed in C99 mode
for (int i = 0; i < dc->res_pool->res_cap->num_dsc; i++) {
Signed-off-by: Asher Song
---
drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c
[AMD Official Use Only - General]
>I need more time for an in deep review of this, but form the one mile high
>view it looks correct to me now.
>Can we do some pre-commit qa testing with this?
I changed drm test "Command submission Test (GFX)" to send high priority ibs
meanwhile running Manhat
[AMD Official Use Only - General]
Reviewed-by: Likun Gao .
Regards,
Likun
-Original Message-
From: Zhang, Hawking
Sent: Tuesday, September 27, 2022 10:31 AM
To: amd-gfx@lists.freedesktop.org; Gao, Likun ; Deucher,
Alexander
Cc: Zhang, Hawking
Subject: [PATCH] drm/amdgpu: add rlc_sr_
[AMD Official Use Only - General]
Reviewed-by: Likun Gao .
Regards,
Likun
-Original Message-
From: Zhang, Hawking
Sent: Tuesday, September 27, 2022 10:30 AM
To: amd-gfx@lists.freedesktop.org; Gao, Likun ; Deucher,
Alexander
Cc: Zhang, Hawking
Subject: [PATCH] drm/amdgpu: make rlc fw
To allow upload the list via psp
Signed-off-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.c | 8
1 file changed, 8 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.c
index 13675b3aa218..792333206362 100644
--- a/d
To allow kernel change
drm/amdgpu/gfx10: switch to amdgpu_gfx_rlc_init_microcode
backward compatible with old rlc firmware
Signed-off-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/a
[AMD Official Use Only - General]
> -Original Message-
> From: amd-gfx On Behalf Of
> Shashank Sharma
> Sent: Tuesday, September 27, 2022 5:40 AM
> To: amd-gfx@lists.freedesktop.org
> Cc: Deucher, Alexander ; Somalapuram,
> Amaranath ; Koenig, Christian
> ; Sharma, Shashank
>
> Subject
On 2022-09-26 17:35, Lyude Paul wrote:
On Mon, 2022-09-26 at 16:03 +1000, Alistair Popple wrote:
When the module is unloaded or a GPU is unbound from the module it is
possible for device private pages to be left mapped in currently running
processes. This leads to a kernel crash when the pages
This was fixed in initialize_cpsch before, but not in initialize_nocpsch.
Factor sdma bitmap initialization into a helper function to apply the
correct implementation in both cases without duplicating it.
v2: Added a range check
Reported-by: Ellis Michael
Signed-off-by: Felix Kuehling
---
.../
Hi August, thanks for the log.
It seems the eDP panel does not fully satisfy the amdgpu requirements
for enabling PSR SU, but we're enabling it anyways.
I suspect it may be due to the "DP_FORCE_PSRSU_CAPABILITY" bit being set
in it's DPCD registers, I'm checking with some devs to see if that i
On 2022-09-26 15:40, Philip Yang wrote:
Unified memory usage with xnack off is tracked to avoid oversubscribe
system memory. When switching xnack mode from off to on, subsequent
free ranges allocated with xnack off will not unreserve memory when
xnack is on, cause memory accounting unbalanced.
On Mon, Sep 26, 2022 at 4:21 PM Bokun Zhang wrote:
>
> - Under SRIOV, we need to send REQ_GPU_FINI to the hypervisor
> during the suspend time. Furthermore, we cannot request a
> mode 1 reset under SRIOV as VF. Therefore, we will skip it
> as it is called in suspend_noirq() function.
>
> - I
This patch switches the GPU workload mode to/from
compute mode, while submitting compute workload.
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
---
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 14 +++---
1 file changed, 11 insertions(+), 3 deletions(-)
diff --git a/driver
This patch and switches the GPU workload based profile based
on the workload hint information saved in the workload context.
The workload profile is reset to NONE when the job is done.
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
---
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
This patch adds new IOCTL flags in amdgpu_context_IOCTL to set
GPU workload profile. These calls will allow a user to switch
to a GPU power profile which might be better suitable to its
workload type. The currently supported workload types are:
"None": Default workload profile
"3D": Workloa
Allow the user to specify a workload hint to the kernel.
We can use these to tweak the dpm heuristics to better match
the workload for improved performance.
V3: Create only set() workload UAPI (Christian)
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
---
include/uapi/drm/amdgpu_dr
This patch adds new functions which will allow a user to
change the GPU power profile based a GPU workload hint
flag.
Cc: Alex Deucher
Signed-off-by: Shashank Sharma
---
drivers/gpu/drm/amd/amdgpu/Makefile | 2 +-
.../gpu/drm/amd/amdgpu/amdgpu_ctx_workload.c | 97 +++
AMDGPU SOCs supports dynamic workload based power profiles, which can
provide fine-tuned performance for a particular type of workload.
This patch series adds an interface to set/reset these power profiles
based on the workload type hints. A user can set a hint of workload
type being submistted to
On Mon, 2022-09-26 at 16:03 +1000, Alistair Popple wrote:
> When the module is unloaded or a GPU is unbound from the module it is
> possible for device private pages to be left mapped in currently running
> processes. This leads to a kernel crash when the pages are either freed
> or accessed from t
On Mon, 2022-09-26 at 16:03 +1000, Alistair Popple wrote:
> nouveau_dmem_fault_copy_one() is used during handling of CPU faults via
> the migrate_to_ram() callback and is used to copy data from GPU to CPU
> memory. It is currently specific to fault handling, however a future
> patch implementing ev
- Under SRIOV, we need to send REQ_GPU_FINI to the hypervisor
during the suspend time. Furthermore, we cannot request a
mode 1 reset under SRIOV as VF. Therefore, we will skip it
as it is called in suspend_noirq() function.
- In the resume code path, we need to send REQ_GPU_INIT to the
hyp
Unified memory usage with xnack off is tracked to avoid oversubscribe
system memory. When switching xnack mode from off to on, subsequent
free ranges allocated with xnack off will not unreserve memory when
xnack is on, cause memory accounting unbalanced.
When switching xnack mode from on to off, n
[AMD Official Use Only - General]
> -Original Message-
> From: amd-gfx On Behalf Of Felix
> Kuehling
> Sent: Wednesday, September 21, 2022 6:30 PM
> To: amd-gfx@lists.freedesktop.org
> Cc: Ellis Michael
> Subject: [PATCH] drm/amdkfd: Fix UBSAN shift-out-of-bounds warning
>
> Caution: Th
On Mon, Sep 26, 2022 at 04:03:06PM +1000, Alistair Popple wrote:
> Since 27674ef6c73f ("mm: remove the extra ZONE_DEVICE struct page
> refcount") device private pages have no longer had an extra reference
> count when the page is in use. However before handing them back to the
> owning device drive
On 9/26/2022 5:19 PM, Christian König wrote:
Am 26.09.22 um 17:14 schrieb Sharma, Shashank:
Hello Christian,
On 9/26/2022 5:10 PM, Christian König wrote:
Am 26.09.22 um 17:02 schrieb Shashank Sharma:
Allow the user to specify a workload hint to the kernel.
We can use these to tweak the dp
Am 26.09.22 um 17:14 schrieb Sharma, Shashank:
Hello Christian,
On 9/26/2022 5:10 PM, Christian König wrote:
Am 26.09.22 um 17:02 schrieb Shashank Sharma:
Allow the user to specify a workload hint to the kernel.
We can use these to tweak the dpm heuristics to better match
the workload for imp
Hello Christian,
On 9/26/2022 5:10 PM, Christian König wrote:
Am 26.09.22 um 17:02 schrieb Shashank Sharma:
Allow the user to specify a workload hint to the kernel.
We can use these to tweak the dpm heuristics to better match
the workload for improved performance.
Signed-off-by: Alex Deucher
Am 26.09.22 um 17:02 schrieb Shashank Sharma:
Allow the user to specify a workload hint to the kernel.
We can use these to tweak the dpm heuristics to better match
the workload for improved performance.
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
---
include/uapi/drm/amdgpu_dr
Am 26.09.22 um 17:04 schrieb Arunpravin Paneer Selvam:
On 9/26/2022 12:01 PM, Christian König wrote:
Am 26.09.22 um 07:25 schrieb Arunpravin Paneer Selvam:
DRM buddy manager allocates the contiguous memory requests in
a single block or multiple blocks. So for the ttm move operation
(incase of
On 9/26/2022 12:01 PM, Christian König wrote:
Am 26.09.22 um 07:25 schrieb Arunpravin Paneer Selvam:
DRM buddy manager allocates the contiguous memory requests in
a single block or multiple blocks. So for the ttm move operation
(incase of low vram memory) we should consider all the blocks to
This patch switches the GPU workload mode to/from
compute mode, while submitting compute workload.
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
---
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 14 +++---
1 file changed, 11 insertions(+), 3 deletions(-)
diff --git a/driver
Allow the user to specify a workload hint to the kernel.
We can use these to tweak the dpm heuristics to better match
the workload for improved performance.
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
---
include/uapi/drm/amdgpu_drm.h | 18 ++
1 file changed, 18 i
This patch and switches the GPU workload based profile based
on the workload hint information saved in the workload context.
The workload profile is reset to NONE when the job is done.
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
---
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
This patch adds new functions which will allow a user to
change the GPU power profile based a GPU workload hint
flag.
Cc: Alex Deucher
Signed-off-by: Shashank Sharma
---
drivers/gpu/drm/amd/amdgpu/Makefile | 2 +-
.../gpu/drm/amd/amdgpu/amdgpu_ctx_workload.c | 97 +++
This patch adds new IOCTL flags in amdgpu_context_IOCTL to get
and set GPU workload profile. These calls will allow a user to
switch to a GPU power profile which might be better suitable to
its workload type. The currently supported workload types are:
"None": Default workload profile
"3D":
AMDGPU SOCs supports dynamic workload based power profiles, which can
provide fine-tuned performance for a particular type of workload.
This patch series adds an interface to set/reset these power profiles
based on the workload type hints. A user can set a hint of workload
type being submistted to
Am 26.09.22 um 11:22 schrieb Asher Song:
The patch fix following compiling error:
error: ‘for’ loop initial declarations are only allowed in C99 mode
for (int i = 0; i < dc->res_pool->res_cap->num_dsc; i++) {
Newer kernels are compiled in C99 mode, so this in now legal.
We should probably
The patch fix following compiling error:
error: ‘for’ loop initial declarations are only allowed in C99 mode
for (int i = 0; i < dc->res_pool->res_cap->num_dsc; i++) {
Signed-off-by: Asher Song
---
drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c | 4 ++--
1 file changed, 2 insertions(+), 2
I was thinking about that as well, yes.
Might be a good idea to just change the alignment check in
amdgpu_bo_create():
/* Memory should be aligned at least to a page size. */
page_align = ALIGN(bp->byte_align, PAGE_SIZE) >>
PAGE_SHIFT;
Something like ALIGN(b
Device drivers can use the migrate_vma family of functions to migrate
existing private anonymous mappings to device private pages. These pages
are backed by memory on the device with drivers being responsible for
copying data to and from device memory.
Device private pages are freed via the pgmap-
This series aims to fix a number of page reference counting issues in drivers
dealing with device private ZONE_DEVICE pages. These result in use-after-free
type bugs, either from accessing a struct page which no longer exists because it
has been removed or accessing fields within the struct page wh
Signed-off-by: Alistair Popple
---
lib/test_hmm.c | 119 +-
lib/test_hmm_uapi.h| 1 +-
tools/testing/selftests/vm/hmm-tests.c | 49 +++-
3 files changed, 148 insertions(+), 21 deletions(-)
diff --git a/lib/test_hmm.c
migrate_device_coherent_page() reuses the existing migrate_vma family of
functions to migrate a specific page without providing a valid mapping
or vma. This looks a bit odd because it means we are calling
migrate_vma_*() without setting a valid vma, however it was considered
acceptable at the time
On Sun, Sep 11, 2022 at 11:28:43PM -0600, Jim Cromie wrote:
> hi Greg, Dan, Jason, DRM-folk,
>
> heres follow-up to V6:
> rebased on driver-core/driver-core-next for -v6 applied bits (thanks)
> rework drm_debug_enabled{_raw,_instrumented,} per Dan.
>
> It excludes:
> nouveau parts (immature
amdgpu_bo_kmap() returns error when fails to map buffer object. Add the
error check and propagate the error.
Signed-off-by: Li Zhong
---
v2: revise the compile error
drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c | 5 -
drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c | 5 -
2 files
On Thu, Sep 22, 2022 at 8:04 PM Lazar, Lijo wrote:
>
>
>
> On 9/23/2022 1:36 AM, Li Zhong wrote:
> > Check the return value of amdgpu_kv_smc_bapm_enable() and log the error
> > when it fails.
> >
> > Signed-off-by: Li Zhong
> > ---
> > drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c | 5 -
> >
To whoom it may belong: Deepest sorrow to inform you that my husband Helmut
passed by on July 21. This access will be cleared in a view days. Irene
Ursprüngliche Nachricht Von: Christian König
Datum: 01.09.22 08:50 (GMT+01:00) An: Dmitry
Osipenko , David Airlie , Gerd
Hoffma
25.08.2022 17:36, Hans de Goede пишет:
> Before this commit when we want userspace to use the acpi_video backlight
> device we register both the GPU's native backlight device and acpi_video's
> firmware acpi_video# backlight device. This relies on userspace preferring
> firmware type backlight devi
When the CPU tries to access a device private page the migrate_to_ram()
callback associated with the pgmap for the page is called. However no
reference is taken on the faulting page. Therefore a concurrent
migration of the device private page can free the page and possibly the
underlying pgmap. Thi
Hi Leo,
On 2022-09-23 20:41, Leo Li wrote:
Hi August,
Can you provide a dmesg log with drm.debug=0x16 enabled in kernel cmdline?
Log is available here: https://files.augustwikerfors.se/dmesg.2022-09-23.txt
This is what I did during that log:
1. Boot the system
2. Type into the password field
When the module is unloaded or a GPU is unbound from the module it is
possible for device private pages to be left mapped in currently running
processes. This leads to a kernel crash when the pages are either freed
or accessed from the CPU because the GPU and associated data structures
and callback
nouveau_dmem_fault_copy_one() is used during handling of CPU faults via
the migrate_to_ram() callback and is used to copy data from GPU to CPU
memory. It is currently specific to fault handling, however a future
patch implementing eviction of data during teardown needs similar
functionality.
Refac
Since 27674ef6c73f ("mm: remove the extra ZONE_DEVICE struct page
refcount") device private pages have no longer had an extra reference
count when the page is in use. However before handing them back to the
owning device driver we add an extra reference count such that free
pages have a reference c
61 matches
Mail list logo
