:
Squash the two patches into one.
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Shashank Sharma
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_doorbell.h| 5 +++--
drivers/gpu/drm/amd/amdgpu/amdgpu_doorbell_mgr.c| 13 +
.../gpu/drm/amd/amdkfd
.
So that doorbell offset will be aligned based on the doorbell
size.
v2:
- Addressed the review comment from Felix.
v3:
- Adding doorbell_size as parameter to get db absolute offset.
v4:
Squash the two patches into one.
Arvind Yadav (1):
drm/amdkfd: get doorbell's absolute offset based
On 10/4/2023 10:29 PM, Felix Kuehling wrote:
On 2023-10-04 12:16, Arvind Yadav wrote:
This patch is to align the absolute doorbell offset
based on the doorbell's size. So that doorbell offset
will be aligned for both 32 bit and 64 bit.
v2:
- Addressed the review comment from Felix.
v3
Deucher
Signed-off-by: Shashank Sharma
Signed-off-by: Arvind Yadav
---
.../gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 6 +-
drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c | 13 +++--
.../gpu/drm/amd/amdkfd/kfd_process_queue_manager.c | 4 +++-
3 files changed, 19
-by: Shashank Sharma
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_doorbell.h | 5 +++--
drivers/gpu/drm/amd/amdgpu/amdgpu_doorbell_mgr.c | 13 +
2 files changed, 12 insertions(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_doorbell.h
b/drivers
.
So that doorbell offset will be aligned based on the doorbell
size.
v2:
- Addressed the review comment from Felix.
v3:
- Adding doorbell_size as parameter to get db absolute offset.
Arvind Yadav (2):
drm/amdgpu: Adding db_size to get doorbell absolute offset
drm/amdkfd: get doorbell's
This patch is to adjust the absolute doorbell offset
against the doorbell id considering the doorbell
size of 32/64 bit.
v2:
- Addressed the review comment from Felix.
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Shashank Sharma
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd
/64 bit.
v2:
- Addressed the review comment from Felix.
Arvind Yadav (1):
drm/amdkfd: Fix unaligned doorbell absolute offset for gfx8
drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
--
2.34.1
Adding felix.kuehl...@amd.com for review.
Thanks
~Arvind
On 9/27/2023 9:46 PM, Arvind Yadav wrote:
This patch is to adjust the absolute doorbell offset
against the doorbell id considering the doorbell
size of 32/64 bit.
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Shashank Sharma
Adding felix.kuehl...@amd.com for review.
Thanks
~Arvind
On 9/27/2023 9:46 PM, Arvind Yadav wrote:
On older chips, the absolute doorbell offset within
the doorbell page is based on the queue ID.
KFD is using queue ID and doorbell size to get an
absolute doorbell offset in userspace
This patch is to adjust the absolute doorbell offset
against the doorbell id considering the doorbell
size of 32/64 bit.
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Shashank Sharma
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 11
of 32/64 bit.
Arvind Yadav (1):
drm/amdkfd: Fix unaligned doorbell absolute offset for gfx8
drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 11 ++-
1 file changed, 10 insertions(+), 1 deletion(-)
--
2.34.1
On 8/28/2023 9:13 PM, Helen Mae Koike Fornazier wrote:
On Monday, August 28, 2023 09:26 -03, Arvind Yadav wrote:
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
Sharma
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 14 ++
1 file changed, 2 insertions(+), 12 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 2d94f1b63bd6..70777fcfa626 100644
This patch switches the GPU workload mode to/from
compute mode, while submitting compute workload.
v3:
- Addressed the review comment about changing the
function name from *_set() to *_get().
Cc: Christian Koenig
Signed-off-by: Alex Deucher
Reviewed-by: Shashank Sharma
Signed-off-by: Arvind
Sharma
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_job.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
index c3d9d75143f4..c5032762d497 100644
--- a/drivers/gpu/drm/amd/amdgpu
: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c| 2 ++
drivers/gpu/drm/amd/amdgpu/amdgpu_workload.c | 7 +++
drivers/gpu/drm/amd/include/amdgpu_workload.h | 2 ++
3 files changed, 11 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu
.
- Added *_clear_all function to clear all the power profile.
- scheduling delay work to clear the power profile when refcount
becomes zero.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_workload.c | 118
the review comment.
- Changing the function name from *_set() to *_get().
- Now setting a power profile when refcount is zero.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_workload.c | 59
Deucher
Reviewed-by: Shashank Sharma
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/Makefile | 2 +-
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 3 ++
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c| 4 ++
drivers/gpu/drm/amd/amdgpu/amdgpu_workload.c | 44
.
v3:
- Addressed all review comment.
- Changed the function name from *_set() to *_get().
- Now clearing all the profile in work handler.
- Added *_clear_all function to clear all the power profile.
Arvind Yadav (7):
drm/amdgpu: Added init/fini functions for workload
drm/amdgpu: Add new function
On 8/22/2023 6:16 PM, Lazar, Lijo wrote:
On 8/22/2023 5:41 PM, Yadav, Arvind wrote:
Hi Lijo,
The *_set function will set the GPU power profile and the *_put
function will schedule the
smu_delayed_work task after 100ms delay. This smu_delayed_work task
will clear a GPU
power profile
On 8/22/2023 6:24 PM, Lazar, Lijo wrote:
On 8/22/2023 5:52 PM, Yadav, Arvind wrote:
On 8/22/2023 12:01 PM, Lazar, Lijo wrote:
On 8/21/2023 12:17 PM, Arvind Yadav wrote:
This patch adds a suspend function that will clear the GPU
power profile before going into suspend state.
v2:
- Add
On 8/22/2023 11:55 AM, Lazar, Lijo wrote:
On 8/21/2023 12:17 PM, Arvind Yadav wrote:
This patch adds a function which will change the GPU
power profile based on a submitted job. This can optimize
the power performance when the workload is on.
v2:
- Splitting workload_profile_set
On 8/22/2023 12:01 PM, Lazar, Lijo wrote:
On 8/21/2023 12:17 PM, Arvind Yadav wrote:
This patch adds a suspend function that will clear the GPU
power profile before going into suspend state.
v2:
- Add the new suspend function based on review comment.
Cc: Shashank Sharma
Cc: Christian
<100ms>
workload
_|``|
Jobs (COMPUTE) |```|__|```|___||___||___
Jobs (VIDEO) ___|```|___||___||___||___
On 8/22/2023 10:21 AM, Lazar, Lijo wrote:
On 8/21/2023 12:17 PM, Arvind
On 8/21/2023 11:40 PM, Alex Deucher wrote:
On Mon, Aug 21, 2023 at 1:54 PM Yadav, Arvind wrote:
On 8/21/2023 9:52 PM, Alex Deucher wrote:
On Mon, Aug 21, 2023 at 2:55 AM Arvind Yadav wrote:
This patch adds a function which will change the GPU
power profile based on a submitted job
On 8/21/2023 11:36 PM, Alex Deucher wrote:
On Mon, Aug 21, 2023 at 2:55 AM Arvind Yadav wrote:
This patch adds a function which will change the GPU
power profile based on a submitted job. This can optimize
the power performance when the workload is on.
v2:
- Splitting workload_profile_set
On 8/21/2023 9:52 PM, Alex Deucher wrote:
On Mon, Aug 21, 2023 at 2:55 AM Arvind Yadav wrote:
This patch adds a function which will change the GPU
power profile based on a submitted job. This can optimize
the power performance when the workload is on.
v2:
- Splitting workload_profile_set
On 8/21/2023 7:09 PM, Shashank Sharma wrote:
On 21/08/2023 08:47, Arvind Yadav wrote:
This patch adds a function which will clear the GPU
power profile after job finished.
This is how it works:
- schedular will set the GPU power profile based on ring_type.
- Schedular will clear the GPU
On 8/21/2023 7:24 PM, Shashank Sharma wrote:
On 21/08/2023 15:35, Yadav, Arvind wrote:
On 8/21/2023 6:36 PM, Shashank Sharma wrote:
Hey Arvind,
On 21/08/2023 08:47, Arvind Yadav wrote:
The'struct amdgpu_smu_workload' initialization/cleanup
functions is added by this patch.
v2
On 8/21/2023 7:16 PM, Shashank Sharma wrote:
On 21/08/2023 08:47, Arvind Yadav wrote:
This patch is to switch the GPU workload profile based
on the submitted job. The workload profile is reset to
default when the job is done.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
On 8/21/2023 7:13 PM, Shashank Sharma wrote:
On 21/08/2023 08:47, Arvind Yadav wrote:
This patch adds a suspend function that will clear the GPU
power profile before going into suspend state.
v2:
- Add the new suspend function based on review comment.
Cc: Shashank Sharma
Cc: Christian
On 8/21/2023 6:36 PM, Shashank Sharma wrote:
Hey Arvind,
On 21/08/2023 08:47, Arvind Yadav wrote:
The'struct amdgpu_smu_workload' initialization/cleanup
functions is added by this patch.
v2:
- Splitting big patch into separate patches.
- Added new fini function.
Cc: Shashank Sharma
Cc
Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 14 ++
1 file changed, 2 insertions(+), 12 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 2d94f1b63bd6..70777fcfa626 100644
--- a/drivers
This patch switches the GPU workload mode to/from
compute mode, while submitting compute workload.
Cc: Christian Koenig
Signed-off-by: Alex Deucher
Signed-off-by: Shashank Sharma
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 8 +---
1 file changed, 5
This patch is to switch the GPU workload profile based
on the submitted job. The workload profile is reset to
default when the job is done.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_job.c | 5 +
1 file
This patch adds a suspend function that will clear the GPU
power profile before going into suspend state.
v2:
- Add the new suspend function based on review comment.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu
function
will cancel this work and set the GPU power profile based on
preferences.
v2:
- Splitting workload_profile_set and workload_profile_put
into two separate patches.
- Addressed review comment.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_workload.c | 56 +++
drivers/gpu/drm/amd/include/amdgpu_workload.h | 3 +
2 files changed, 59 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_workload.c
b
The'struct amdgpu_smu_workload' initialization/cleanup
functions is added by this patch.
v2:
- Splitting big patch into separate patches.
- Added new fini function.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/Makefile
.
Arvind Yadav (7):
drm/amdgpu: Added init/fini functions for workload
drm/amdgpu: Add new function to set GPU power profile
drm/amdgpu: Add new function to put GPU power profile
drm/amdgpu: Add suspend function to clear the GPU power profile.
drm/amdgpu: Switch on/off GPU workload profile
On 8/14/2023 8:28 PM, Shashank Sharma wrote:
Hey Arvind,
On 14/08/2023 09:34, Arvind Yadav wrote:
This patch adds a function which will allow to
change the GPU power profile based on a submitted job.
This can optimize the power performance when the
workload is on.
Cc: Shashank Sharma
Cc
On 8/14/2023 9:35 PM, Shashank Sharma wrote:
Ah, Thanks for pointing that out Alex.
@Arvind, please refer to the patch
(https://patchwork.freedesktop.org/patch/504854/?series=109060=4)
in previous series of SMU workload hints with UAPI (here:
https://patchwork.freedesktop.org/series/109060
On 8/14/2023 8:03 PM, Alex Deucher wrote:
On Mon, Aug 14, 2023 at 3:35 AM Arvind Yadav wrote:
This patch adds a function which will allow to
change the GPU power profile based on a submitted job.
This can optimize the power performance when the
workload is on.
A few minor comments inline
On 8/14/2023 5:35 PM, Christian König wrote:
Am 14.08.23 um 09:34 schrieb Arvind Yadav:
This patch adds a function which will allow to
change the GPU power profile based on a submitted job.
This can optimize the power performance when the
workload is on.
Cc: Shashank Sharma
Cc: Christian
-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 14 ++
1 file changed, 2 insertions(+), 12 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 2d94f1b63bd6..70777fcfa626 100644
--- a/drivers/gpu/drm/amd
This patch is to switch the GPU workload profile based
on the submitted job. The workload profile is reset to
default when the job is done.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/amdgpu_job.c | 5 +
1 file
This patch adds a function which will allow to
change the GPU power profile based on a submitted job.
This can optimize the power performance when the
workload is on.
Cc: Shashank Sharma
Cc: Christian Koenig
Cc: Alex Deucher
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/amd/amdgpu/Makefile
on submitted job. This can optimize the power
performance when the particular workload is on.
Arvind Yadav (3):
drm/amdgpu: Add new api to switch on/off power profile mode
drm/amdgpu: Switch on/off GPU workload profile
Revert "drm/amd/amdgpu: switch on/off vcn power profile mode"
drive
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/scheduler/sched_main.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/scheduler/sched_main.c
b/drivers/gpu/drm/scheduler/sched_main.c
index 820c0c5544e1..ea7bfa99d6c9 100644
--- a/drivers/gpu/drm/scheduler/
Here's enabling software signaling on fence for syncobj.
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/drm_syncobj.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c
index 0c2be8360525..2ab7a1a9eeb4 100644
--- a/drivers/gpu
amdgpu]
amdgpu_ih_process+0x86/0x100 [amdgpu]
amdgpu_irq_handler+0x24/0x60 [amdgpu]
__handle_irq_event_percpu+0x4b/0x190
handle_irq_event_percpu+0x15/0x50
handle_irq_event+0x39/0x60
handle_edge_irq+0xaf/0x210
__common_interrupt+0x6e/0x110
common_interrupt+0xc1/0xe0
Signed-off-by:
On 10/17/2022 8:20 PM, Christian König wrote:
Am 17.10.22 um 16:30 schrieb Arvind Yadav:
-This is purely a timing issue. Here, sometimes Job free
is happening before the job is done.
To fix this issue moving 'dma_fence_cb' callback from
job(struct drm_sched_job) to scheduler fence (struct
NOPTI
CPU: 2 PID: 0 Comm: swapper/2 Not tainted 6.0.0-rc2-custom #1
Arvind : [dma_fence_default_wait _START] timeout = -1
Hardware name: AMD Dibbler/Dibbler, BIOS RDB1107CC 09/26/2018
RIP: 0010:drm_sched_job_done.isra.0+0x11/0x140 [gpu_sched]
Code: 8b fe ff ff be 03 00 00 00 e8 7b da b7 e3 e9 d4 fe
ROR("fence add callback failed (%d)\n", r);
}
bool drm_sched_fence_remove_parent_cb(struct drm_sched_fence *s_fence) {
return dma_fence_remove_callback(s_fence->parent,
_fence->cb);
}
Thanks
~Arvind
Regards,
Christian.
Am 12.10.
: supervisor read access in kernel mode
#PF: error_code(0x) - not-present page
PGD 0 P4D 0
Oops: [#1] PREEMPT SMP NOPTI
CPU: 2 PID: 0 Comm: swapper/2 Not tainted 6.0.0-rc2-custom #1
Arvind : [dma_fence_default_wait _START] timeout = -1
Hardware name: AMD Dibbler/Dibbler, BIOS RDB1107CC 09/26
On 9/30/2022 4:56 PM, Christian König wrote:
Am 30.09.22 um 10:48 schrieb Arvind Yadav:
BUG: kernel NULL pointer dereference, address: 0088
#PF: supervisor read access in kernel mode
#PF: error_code(0x) - not-present page
PGD 0 P4D 0
Oops: [#1] PREEMPT SMP NOPTI
BUG: kernel NULL pointer dereference, address: 0088
#PF: supervisor read access in kernel mode
#PF: error_code(0x) - not-present page
PGD 0 P4D 0
Oops: [#1] PREEMPT SMP NOPTI
CPU: 2 PID: 0 Comm: swapper/2 Not tainted 6.0.0-rc2-custom #1
Arvind : [dma_fence_default_wait
NULL check is added for s_fence->parent.
Signed-off-by: Arvind Yadav
---
changes in v1 : subject 'drm/sched:' was missing.
drivers/gpu/drm/scheduler/sched_main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/scheduler/sched_main.c
b/drivers/gpu/
NULL check is added for s_fence->parent.
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/scheduler/sched_main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/scheduler/sched_main.c
b/drivers/gpu/drm/scheduler/sched_main.c
index 4f2395d1a791..6684d8846
On 9/30/2022 12:02 AM, Christian König wrote:
Am 29.09.22 um 20:30 schrieb Yadav, Arvind:
On 9/29/2022 11:48 PM, Christian König wrote:
Am 27.09.22 um 19:24 schrieb Arvind Yadav:
Fence signaling must be enabled to make sure that
the dma_fence_is_signaled_locked() function ever returns true
On 9/29/2022 11:48 PM, Christian König wrote:
Am 27.09.22 um 19:24 schrieb Arvind Yadav:
Fence signaling must be enabled to make sure that
the dma_fence_is_signaled_locked() function ever returns true.
Since drivers and implementations sometimes mess this up,
this ensures correct behaviour
Remove the extra signaled bit status check because
it is returning early when the fence is already signaled and
__dma_fence_enable_signaling is checking the status of signaled
bit again.
Signed-off-by: Arvind Yadav
---
drivers/dma-buf/dma-fence.c | 5 -
1 file changed, 5 deletions(-)
diff
Here's enabling software signaling on fence for sw_sync.
Signed-off-by: Arvind Yadav
---
drivers/dma-buf/sw_sync.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/dma-buf/sw_sync.c b/drivers/dma-buf/sw_sync.c
index 348b3a9170fa..d2a52ceac14e 100644
--- a/drivers/dma-buf/sw_sync.c
bugs resulting in not
signaled fences much more obvious.
Signed-off-by: Arvind Yadav
---
include/linux/dma-fence.h | 5 +
1 file changed, 5 insertions(+)
diff --git a/include/linux/dma-fence.h b/include/linux/dma-fence.h
index 775cdc0b4f24..5156dc6be0a6 100644
--- a/include/linux/dma
make any implementation bugs resulting in not
signaled fences much more obvious.
Arvind Yadav (3):
[PATCH 1/3] dma-buf: Remove the signaled bit status check
[PATCH 2/3] dma-buf: Enable signaling on fence for sw_sync
[PATCH 3/3] dma-buf: Check status of enable-signaling bit on debug
drivers
the
sync_file and drm_syncobj tests I would expect to have problems with
this extra check.
Yes, IGT test cases are failing .
~Arvind
Thanks,
Christian.
Am 14.09.22 um 18:43 schrieb Arvind Yadav:
Fence signaling must be enabled to make sure that
the dma_fence_is_signaled() function ever returns
Using the parent fence instead of the finished fence
to get the job status. This change is to avoid GPU
scheduler timeout error which can cause GPU reset.
Signed-off-by: Arvind Yadav
Reviewed-by: Andrey Grodzovsky
---
changes in v1,v2 - Enable signaling for finished fence in sche_main
resulting in not
signaled fences much more obvious.
Signed-off-by: Arvind Yadav
---
Changes in v1,v2 :
1- Addressing Christian's comment to replace
CONFIG_DEBUG_WW_MUTEX_SLOWPATH instead of CONFIG_DEBUG_FS.
2- As per Christian's comment moving this patch at last so
The version of this patch
Here's enabling software signaling on fence for selftest.
Signed-off-by: Arvind Yadav
Reviewed-by: Christian König
---
Changes in v1 :
1- Addressing Christian's comment to remove unnecessary callback.
2- Replacing CONFIG_DEBUG_WW_MUTEX_SLOWPATH instead of CONFIG_DEBUG_FS.
3- The version
dma_fence_wait() should always enable signaling even
when the fence is already signaled.
Signed-off-by: Arvind Yadav
---
Changes in v1..v3: This new patch was not part of previous series.
---
drivers/dma-buf/dma-fence.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git
Here's setting software signaling bit for the stub fence
which is always signaled. If this fence signaling bit is
not set then the AMD GPU scheduler will cause a GPU reset
due to a GPU scheduler cleanup activity timeout.
Signed-off-by: Arvind Yadav
Reviewed-by: Christian König
---
Changes
Remove the signaled bit status check because it is returning
early when the fence is already signaled and
__dma_fence_enable_signaling is checking the status of signaled
bit again.
Signed-off-by: Arvind Yadav
Reviewed-by: Christian König
---
Changes in v1, v2: This new patch was not part
resulting in not
signaled fences much more obvious.
Arvind Yadav (6):
[PATCH v4 1/6] dma-buf: Remove the signaled bit status check
[PATCH v4 2/6] dma-buf: set signaling bit for the stub fence
[PATCH v4 3/6] dma-buf: Enable signaling on fence for selftests
[PATCH v4 4/6] dma-buf: dma_fence_wait
use signaling
enabled by the calling of dma_fence_add_callback() for parent fence. As
a result, dma_fence_is_signaled() will return the correct fence status
and reset worker thread can be cancelled in drm_sched_get_cleanup_job().
~arvind
Andrey
On 2022-09-09 13:08, Arvind Yadav wrote:
Us
resulting in not
signaled fences much more obvious.
Signed-off-by: Arvind Yadav
---
Changes in v1,v2 :
1- Addressing Christian's comment to replace
CONFIG_DEBUG_WW_MUTEX_SLOWPATH instead of CONFIG_DEBUG_FS.
2- As per Christian's comment moving this patch at last so
The version of this patch is also
Using the parent fence instead of the finished fence
to get the job status. This change is to avoid GPU
scheduler timeout error which can cause GPU reset.
Signed-off-by: Arvind Yadav
---
changes in v1,v2 - Enable signaling for finished fence in sche_main()
is removed
---
drivers/gpu/drm
Here's enabling software signaling on fence because
amdgpu_ctx_add_fence() is checking the status of fence
and emits warning.
Signed-off-by: Arvind Yadav
---
Changes in v1, v2: This new patch was not part of previous series.
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 2 ++
1 file changed
Here's enabling software signaling on fence for selftest.
Signed-off-by: Arvind Yadav
---
Changes in v1 :
1- Addressing Christian's comment to remove unnecessary callback.
2- Replacing CONFIG_DEBUG_WW_MUTEX_SLOWPATH instead of CONFIG_DEBUG_FS.
3- The version of this patch is also changed
Here's setting software signaling bit for the stub fence
which is always signaled. If this fence signaling bit is
not set then the AMD GPU scheduler will cause a GPU reset
due to a GPU scheduler cleanup activity timeout.
Signed-off-by: Arvind Yadav
---
Changes in v1 :
1- Addressing Christian's
Remove the signaled bit status check because it is returning
early when the fence is already signaled and
__dma_fence_enable_signaling is checking the status of signaled
bit again.
Signed-off-by: Arvind Yadav
---
Changes in v1, v2: This new patch was not part of previous series.
---
drivers
resulting in not
signaled fences much more obvious.
Arvind Yadav (6):
[PATCH v3 1/6] dma-buf: Remove the signaled bit status check
[PATCH v3 2/6] dma-buf: set signaling bit for the stub fence
[PATCH v3 3/6] dma-buf: Enable signaling on fence for selftests
[PATCH v3 4/6] drm/amdgpu: Enable
On 9/6/2022 12:39 PM, Christian König wrote:
Am 05.09.22 um 18:35 schrieb Arvind Yadav:
Here's on debug enabling software signaling for the stub fence
which is always signaled. This fence should enable software
signaling otherwise the AMD GPU scheduler will cause a GPU reset
due to a GPU
bit status to confirm that enable_signaling
is enabled.
Signed-off-by: Arvind Yadav
---
Changes in v1 :
1- Addressing Christian's comment to replace
CONFIG_DEBUG_WW_MUTEX_SLOWPATH instead of CONFIG_DEBUG_FS.
2- As per Christian's comment moving this patch at last so
The version of this patch
Here's on debug enabling software signaling for selftest.
Signed-off-by: Arvind Yadav
---
Changes in v1 :
1- Addressing Christian's comment to remove unnecessary callback.
2- Replacing CONFIG_DEBUG_WW_MUTEX_SLOWPATH instead of CONFIG_DEBUG_FS.
3- The version of this patch is also changed
Here's on debug enabling software signaling for the stub fence
which is always signaled. This fence should enable software
signaling otherwise the AMD GPU scheduler will cause a GPU reset
due to a GPU scheduler cleanup activity timeout.
Signed-off-by: Arvind Yadav
---
Changes in v1 :
1
Here's enabling software signaling for finished fence.
Signed-off-by: Arvind Yadav
---
Changes in v1 :
1- Addressing Christian's comment to remove CONFIG_DEBUG_FS check from
this patch.
2- The version of this patch is also changed and previously
it was [PATCH 2/4]
---
drivers/gpu/drm
calling dma_fence_is_signaled(). To avoid this scenario
on debug kernel, check the DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT bit
status before checking the MA_FENCE_FLAG_SIGNALED_BIT bit status
to confirm that software signaling is enabled.
Arvind Yadav (4):
[PATCH v2 1/4] drm/sched: Enable signaling
On 9/5/2022 7:16 PM, Yadav, Arvind wrote:
On 9/5/2022 4:55 PM, Christian König wrote:
Am 05.09.22 um 12:56 schrieb Arvind Yadav:
Here's on debug adding an enable_signaling callback for finished
fences and enabling software signaling for finished fence.
Signed-off-by: Arvind Yadav
On 9/5/2022 4:56 PM, Christian König wrote:
Am 05.09.22 um 12:56 schrieb Arvind Yadav:
Here's on debug adding an enable_signaling callback for the stub
fences and enabling software signaling for the stub fence which
is always signaled. This fence should enable software signaling
otherwise
On 9/5/2022 4:55 PM, Christian König wrote:
Am 05.09.22 um 12:56 schrieb Arvind Yadav:
Here's on debug adding an enable_signaling callback for finished
fences and enabling software signaling for finished fence.
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/scheduler/sched_fence.c | 12
On 9/5/2022 4:51 PM, Christian König wrote:
Am 05.09.22 um 12:56 schrieb Arvind Yadav:
The core DMA-buf framework needs to enable signaling
before the fence is signaled. The core DMA-buf framework
can forget to enable signaling before the fence is signaled.
To avoid this scenario on the debug
Here's on debug adding an enable_signaling callback for fences
and enabling software signaling for selftest.
Signed-off-by: Arvind Yadav
---
drivers/dma-buf/st-dma-fence-chain.c | 17 +
drivers/dma-buf/st-dma-fence-unwrap.c | 54 +++
drivers/dma-buf/st-dma
.
Signed-off-by: Arvind Yadav
---
drivers/dma-buf/dma-fence.c | 17 +
1 file changed, 17 insertions(+)
diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c
index 066400ed8841..0a67af945ef8 100644
--- a/drivers/dma-buf/dma-fence.c
+++ b/drivers/dma-buf/dma-fence.c
Here's on debug adding an enable_signaling callback for finished
fences and enabling software signaling for finished fence.
Signed-off-by: Arvind Yadav
---
drivers/gpu/drm/scheduler/sched_fence.c | 12
drivers/gpu/drm/scheduler/sched_main.c | 4 +++-
2 files changed, 15
bit status to confirm that enable_signaling
is enabled.
Signed-off-by: Arvind Yadav
---
include/linux/dma-fence.h | 5 +
1 file changed, 5 insertions(+)
diff --git a/include/linux/dma-fence.h b/include/linux/dma-fence.h
index 775cdc0b4f24..60c0e935c0b5 100644
--- a/include/linux/dma-fence.h
calling dma_fence_is_signaled(). To avoid this scenario
on debug kernel, check the DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT bit
status before checking the MA_FENCE_FLAG_SIGNALED_BIT bit status
to confirm that software signaling is enabled.
Arvind Yadav (4):
dma-buf: Check status of enable-signaling bit
If core DMA-buf framework forgets to call dma_fence_enable_signaling()
before calling the dma_fence_is_signaled(). To handle this scenario on
debug kernel the DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT needs to be checked
before checking the actual signaling status.
Signed-off-by: Arvind Yadav
-by: Arvind Yadav
---
drivers/dma-buf/dma-fence-array.c | 6 ++
include/linux/dma-fence-array.h | 2 ++
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/dma-buf/dma-fence-array.c
b/drivers/dma-buf/dma-fence-array.c
index 5c8a7084577b..3ebb6c5fa287 100644
--- a/drivers
1 - 100 of 174 matches
Mail list logo