On 27/04/17 07:04 PM, Christian König wrote:
> Am 27.04.2017 um 10:18 schrieb Michel Dänzer:
>> From: Michel Dänzer
>>
>> This reverts commit cb341a319f7e66f879d69af929c3dadfc1a8f31e.
>>
>> The purpose of the refactor was for amdgpu_crtc_prepare/submit_flip to
>> be used by the DC code, but that's
Change-Id: I19202f1e54ce6a1b8b54aacbc0d42dbba7605662
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 97 ++
.../gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c | 23 +
.../gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h | 1 +
.../gpu/drm/amd
export disablesmcctf to eventmgr.
need to disable temperature alert when s3/s4.
otherwise, when resume back,enable temperature
alert will fail.
Change-Id: I6bbc85b0fc389e442a44ad4d0db3db3342ff5955
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c | 10 ++
d
Change-Id: I7d7f760c5996ddb00a17b34d4b0a3070a965
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/eventmgr/eventsubchains.c | 2 +-
drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c | 5 +
drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.h | 1 +
3 files changed, 7 inse
Reviewed-by: Ken Wang
From: Junwei Zhang
Sent: Friday, April 28, 2017 11:10:46 AM
To: amd-gfx@lists.freedesktop.org
Cc: Deucher, Alexander; Wang, Ken; Zhang, Jerry
Subject: [PATCH] drm/amdgpu: add cu info wave_front_size
missed that for gfx v9 info export
Signe
missed that for gfx v9 info export
Signed-off-by: Junwei Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 503010a..e330009 100644
--- a/drivers/gpu/drm/amd/amdgpu/amd
Reviewed-by: Junwei Zhang
Regards,
Jerry (Junwei Zhang)
Linux Base Graphics
SRDC Software Development
_
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of
> zhoucm1
> Sent: Friday, April 28, 2017 10:59
>
This is drm patch for dedicated vmid.
Kernel patch has pushed to staging, anyone could have a quick review on
this? Jerry? Guess Christian is in dream at current time.
Regards,
David Zhou
On 2017年04月28日 10:38, Chunming Zhou wrote:
Change-Id: I3c1ea377dad8f6d64a992b071a6ec36a1143e416
Signed-of
Nice catch!
Reviewed-by: Junwei Zhang
Regards,
Jerry (Junwei Zhang)
Linux Base Graphics
SRDC Software Development
_
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of
> Chunming Zhou
> Sent: Friday, April
the case could happen when gpu reset:
1. when gpu reset, cs can be continue until sw queue is full, then push job
will wait with holding pd reservation.
2. gpu_reset routine will also need pd reservation to restore page table from
their shadow.
3. cs is waiting for gpu_reset complete, but gpu res
Change-Id: I3c1ea377dad8f6d64a992b071a6ec36a1143e416
Signed-off-by: Chunming Zhou
---
include/drm/amdgpu_drm.h | 22 ++
1 file changed, 22 insertions(+)
diff --git a/include/drm/amdgpu_drm.h b/include/drm/amdgpu_drm.h
index 26d02ba..28cda3a 100644
--- a/include/drm/amdgpu_drm
On 04/27/2017 08:04 PM, Christian König wrote:
Am 27.04.2017 um 12:23 schrieb zhoucm1:
On 2017年04月27日 18:11, Christian König wrote:
Am 27.04.2017 um 11:08 schrieb Zhang, Jerry (Junwei):
On 04/27/2017 04:25 PM, Chunming Zhou wrote:
the case could happen when gpu reset:
1. when gpu reset, cs
This is super simple elimination of else branch and I should
probably even use unlikely in
if (ring->count_dw < count_dw) {
However, amdgpu_ring_write() has similar if condition, but does not
return after DRM_ERROR and it looks suspicious. On error, we still
adding v to ring and keeping c
The changes in the workarounds function use DRM_INFO rather than DRM_DEBUG
because CZs with EDC are often used in embedded environments and any info can
be useful especially in the case of an intermittent problem.
>From e1ce383592c275b58ad95bd80b5479af8c1f9dae Mon Sep 17 00:00:00 2001
From: Davi
Opening the DRM file (/dev/dri/card%d) triggers all sorts of KMD
work to happen which is not useful if the KMD is hung or not working.
Since --top is the only user of the file currently we simply defer
opening it until --top is invoked.
Signed-off-by: Tom St Denis
---
src/app/top.c | 7 +++
From: Christian König
This way we can finally use some more stats.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 30 --
1 file changed, 12 insertions(+), 18 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
b/drivers/
From: Christian König
Instead of specifying interruptible and no_wait_gpu manually.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 8
drivers/gpu/drm/nouveau/nouveau_bo.c| 6 --
drivers/gpu/drm/radeon/radeon_ttm.c | 8
drivers/gpu/
From: Christian König
Give moving a BO into place an operation context to work with.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 9 ++---
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c| 3 ++-
drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c | 3 ++-
drivers/g
From: Christian König
Instead of the global statistics use the per context bytes moved counter.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 7 +--
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 5 +
2 files changed, 2 insertions(+), 10 deletions(-)
di
From: Christian König
Add some statistics how many bytes we have moved.
Signed-off-by: Christian König
---
drivers/gpu/drm/ttm/ttm_bo.c | 1 +
include/drm/ttm/ttm_bo_api.h | 1 +
2 files changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 0c
From: Christian König
Instead of passing down the parameters manually to every function.
Signed-off-by: Christian König
---
drivers/gpu/drm/ttm/ttm_bo.c | 69 +++-
1 file changed, 30 insertions(+), 39 deletions(-)
diff --git a/drivers/gpu/drm/ttm/ttm_bo
From: Christian König
Extern is the default for function declerations anyway.
Signed-off-by: Christian König
---
include/drm/ttm/ttm_bo_driver.h | 113 +++-
1 file changed, 54 insertions(+), 59 deletions(-)
diff --git a/include/drm/ttm/ttm_bo_driver.h b/inc
From: Christian König
Instead of specifying if sleeping should be interruptible.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 3 ++-
drivers/gpu/drm/ttm/ttm_bo.c | 12 +---
include/drm/ttm/ttm_bo_api.h | 5 ++---
3 files
From: Christian König
Extern is the default for function declerations anyway and this
solves a bunch of 80char per line issues.
Signed-off-by: Christian König
---
include/drm/ttm/ttm_bo_api.h | 135 ++-
1 file changed, 56 insertions(+), 79 deletions(-)
From: Christian König
Instead of passing the parameters manually.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 21 +++--
drivers/gpu/drm/nouveau/nouveau_bo.c| 27 ---
drivers/gpu/drm/qxl/qxl_ttm.c | 9 --
From: Christian König
Not used any more.
Signed-off-by: Christian König
---
drivers/gpu/drm/ttm/ttm_bo.c | 5 ++---
include/drm/ttm/ttm_bo_api.h | 1 -
2 files changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 04cbecc..7033
KFD doesn't use CGS.
Regards,
Felix
-Original Message-
From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of
Deucher, Alexander
Sent: Wednesday, April 26, 2017 12:36 PM
To: 'Christian König'; amd-gfx@lists.freedesktop.org
Subject: RE: [PATCH] drm/amdgpu: remove unus
Am 27.04.2017 um 16:10 schrieb Julien Isorce:
Encountered a dozen of exact same backtraces when mesa's
pb_cache_release_all_buffers is called after that a gpu reset failed.
v2: Remove superfluous error message added in v1.
bug: https://bugs.freedesktop.org/show_bug.cgi?id=96271
Signed-off-by:
Encountered a dozen of exact same backtraces when mesa's
pb_cache_release_all_buffers is called after that a gpu reset failed.
v2: Remove superfluous error message added in v1.
bug: https://bugs.freedesktop.org/show_bug.cgi?id=96271
Signed-off-by: Julien Isorce
---
drivers/gpu/drm/radeon/radeo
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Junwei Zhang
> Sent: Thursday, April 27, 2017 4:31 AM
> To: amd-gfx@lists.freedesktop.org
> Cc: Deucher, Alexander; Wang, Ken; Zhang, Jerry
> Subject: [PATCH 2/2] drm/amdgpu: bump version for
A really good improvement
Reviewed-by: Monk Liu
-Original Message-
From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of
zhoucm1
Sent: Thursday, April 27, 2017 6:24 PM
To: Christian König ; Zhang, Jerry
; amd-gfx@lists.freedesktop.org
Cc: Liu, Monk
Subject: Re: [P
On Thu, Apr 27, 2017 at 9:22 AM, Christian König
wrote:
> From: Christian König
>
> Those functions are all unused and some not even implemented.
>
> v2: keep cgs_get_pci_resource it is used by the ACP driver.
>
> Signed-off-by: Christian König
Reviewed-by: Alex Deucher
> ---
> drivers/gpu/d
Am 26.04.2017 um 18:36 schrieb Deucher, Alexander:
-Original Message-
From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
Of Christian König
Sent: Wednesday, April 26, 2017 9:14 AM
To: amd-gfx@lists.freedesktop.org
Subject: [PATCH] drm/amdgpu: remove unused and mostly u
From: Christian König
Those functions are all unused and some not even implemented.
v2: keep cgs_get_pci_resource it is used by the ACP driver.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c | 198 ---
drivers/gpu/drm/amd/include/cgs_common.h |
Reviewed-by: Christian König for patch #3 and #5.
Am 27.04.2017 um 15:09 schrieb Chunming Zhou:
The current kernel implementation, which grabs the idle VMID from pool when
emitting the job may:
The back-to-back submission from one process could use different VMID.
The submission to
The current kernel implementation, which grabs the idle VMID from pool when
emitting the job may:
The back-to-back submission from one process could use different VMID.
The submission to different queues from single process could use different
VMID
It works well in most case but cannot
It will be used for reserving vmid.
Change-Id: Ib7169ea999690c8e82d0dcbccdd2d97760c0270a
Signed-off-by: Chunming Zhou
Reviewed-by: Junwei Zhang
Reviewed-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 1 +
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 18 ++
driv
v2: rename dedicated_vmid to reserved_vmid
Change-Id: Id728e20366c8a1ae90d4e901dc80e136e2a613bb
Signed-off-by: Chunming Zhou
Reviewed-by: Junwei Zhang
Reviewed-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 17 -
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 2 ++
add reserve/unreserve vmid funtions.
v3:
only reserve vmid from gfxhub
v4:
fix racy condition
Change-Id: I5f80dc39dc9d44660a96a2b710b0dbb4d3b9039d
Signed-off-by: Chunming Zhou
Reviewed-by: Junwei Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 62 +++---
1 file ch
v2: move #define to amdgpu_vm.h
v3: move reserved vmid counter to id_manager,
and increase counter before allocating vmid
v4: rename to reserved_vmid_num
Change-Id: Ie5958cf6dbdc1c8278e61d9158483472d6f5c6e3
Signed-off-by: Chunming Zhou
Reviewed-by: Junwei Zhang
Reviewed-by: Christian König
---
Change-Id: I1065e0430ed44f7ee6c29214b72e35a7343ea02b
Signed-off-by: Chunming Zhou
Reviewed-by: Junwei Zhang
Reviewed-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
b/d
v2: move sync waiting only when flush needs
v3: fix racy
v4: peek fence instead of get fence, and fix potential context starved.
Change-Id: I64da2701c9fdcf986afb90ba1492a78d5bef1b6c
Signed-off-by: Chunming Zhou
Reviewed-by: Junwei Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 79 ++
Am 27.04.2017 um 12:58 schrieb Julien Isorce:
But always print an error.
Encountered a dozen of exact same backtraces when mesa's
pb_cache_release_all_buffers is called after that a gpu reset failed.
An other approach would be to check rdev->vm_manager.enabled instead
of rdev->accel_working in
Am 27.04.2017 um 12:26 schrieb zhoucm1:
On 2017年04月27日 17:51, Christian König wrote:
Patch #1, #2, #4 and #6 are Reviewed-by: Christian König
.
Patch #3:
+/* Select the first entry VMID */
+idle = list_first_entry(&id_mgr->ids_lru, struct amdgpu_vm_id,
list);
+list_del_init(&i
Am 27.04.2017 um 12:23 schrieb zhoucm1:
On 2017年04月27日 18:11, Christian König wrote:
Am 27.04.2017 um 11:08 schrieb Zhang, Jerry (Junwei):
On 04/27/2017 04:25 PM, Chunming Zhou wrote:
the case could happen when gpu reset:
1. when gpu reset, cs can be continue until sw queue is full, then
pu
Am 27.04.2017 um 12:57 schrieb Julien Isorce:
Like done in radeon_cs_ioctl.
In particular it avoids mesa to call map/unmap:
radeon_create_bo
ioctl(DRM_RADEON_GEM_CREATE) -> ok
ioctl(DRM_RADEON_GEM_VA-MAP)
radeon_destroy_bo
ioctl(DRM_RADEON_GEM_VA-UNMAP)
Encountered also cas
But always print an error.
Encountered a dozen of exact same backtraces when mesa's
pb_cache_release_all_buffers is called after that a gpu reset failed.
An other approach would be to check rdev->vm_manager.enabled instead
of rdev->accel_working in the other function radeon_gem_object_close.
But
Like done in radeon_cs_ioctl.
In particular it avoids mesa to call map/unmap:
radeon_create_bo
ioctl(DRM_RADEON_GEM_CREATE) -> ok
ioctl(DRM_RADEON_GEM_VA-MAP)
radeon_destroy_bo
ioctl(DRM_RADEON_GEM_VA-UNMAP)
Encountered also cases where the vm_manager succeeded to be enabled after
On 2017年04月27日 17:51, Christian König wrote:
Patch #1, #2, #4 and #6 are Reviewed-by: Christian König
.
Patch #3:
+/* Select the first entry VMID */
+idle = list_first_entry(&id_mgr->ids_lru, struct amdgpu_vm_id,
list);
+list_del_init(&idle->list);
+vm->reserved_vmid[vmhub]
On 2017年04月27日 18:11, Christian König wrote:
Am 27.04.2017 um 11:08 schrieb Zhang, Jerry (Junwei):
On 04/27/2017 04:25 PM, Chunming Zhou wrote:
the case could happen when gpu reset:
1. when gpu reset, cs can be continue until sw queue is full, then
push job will wait with holding pd reservat
Am 27.04.2017 um 11:08 schrieb Zhang, Jerry (Junwei):
On 04/27/2017 04:25 PM, Chunming Zhou wrote:
the case could happen when gpu reset:
1. when gpu reset, cs can be continue until sw queue is full, then
push job will wait with holding pd reservation.
2. gpu_reset routine will also need pd rese
Am 27.04.2017 um 10:18 schrieb Michel Dänzer:
From: Michel Dänzer
This reverts commit cb341a319f7e66f879d69af929c3dadfc1a8f31e.
The purpose of the refactor was for amdgpu_crtc_prepare/submit_flip to
be used by the DC code, but that's no longer the case.
Signed-off-by: Michel Dänzer
---
dri
Am 27.04.2017 um 09:18 schrieb Trigger Huang:
Support for MCBP/Virtualization in combination with chained IBs is
formal released on firmware feature version #46. So enable it
according to firmware feature version, otherwise, world switch will
hang.
Signed-off-by: Trigger Huang
Reviewed-by: Ch
Patch #1, #2, #4 and #6 are Reviewed-by: Christian König
.
Patch #3:
+ /* Select the first entry VMID */
+ idle = list_first_entry(&id_mgr->ids_lru, struct amdgpu_vm_id, list);
+ list_del_init(&idle->list);
+ vm->reserved_vmid[vmhub] = idle;
+ mutex_unlock(&id_mgr-
On 27/04/17 04:39 PM, Michel Dänzer wrote:
> On 27/04/17 04:19 AM, Alex Xie wrote:
>> Hi,
>>
>> I knew this is part of ioctl. And I intended to fix this interruptible
>> waiting in an ioctl.
>
> It's by design, nothing that needs fixing in the case of an ioctl.
>
>
>> 1. The wait is short. There
Am 27.04.2017 um 06:42 schrieb zhoucm1:
On 2017年04月27日 10:52, Zhang, Jerry (Junwei) wrote:
On 04/26/2017 07:10 PM, Chunming Zhou wrote:
v2: move sync waiting only when flush needs
Change-Id: I64da2701c9fdcf986afb90ba1492a78d5bef1b6c
Signed-off-by: Chunming Zhou
---
drivers/gpu/drm/amd/amd
On 27/04/17 06:05 PM, Christian König wrote:
>
>> 2. In this function and caller functions, the error handling for such
>> interrupt is complicated and risky. When the waiting is interrupted by
>> a signal, the callers of this function need to handle this interrupt
>> error. I traced the calling s
Am 27.04.2017 um 04:24 schrieb Zhang, Jerry (Junwei):
On 04/27/2017 10:13 AM, zhoucm1 wrote:
On 2017年04月26日 20:26, Christian König wrote:
Am 26.04.2017 um 13:10 schrieb Chunming Zhou:
add reserve/unreserve vmid funtions.
v3:
only reserve vmid from gfxhub
Change-Id: I5f80dc39dc9d44660a96a2b7
On 27/04/17 05:15 AM, Felix Kuehling wrote:
>
> Hi Michel,
>
> You said in an earlier email that it doesn't have to be convenient. With
> that in mind, I went back to a minimalistic solution that doesn't need
> any additional Kconfig options and only uses two module parameters (one
> in radeon, o
1. The wait is short. There is not much benefit by interruptible
waiting. The BO is a frame buffer BO. Are there many competitors to
lock this BO? If not, the wait is very short. This is my main reason
to change.
The problem is that all those waits can block the MM subsystem from
reclaiming me
Am 26.04.2017 um 22:15 schrieb Felix Kuehling:
On 17-04-25 02:28 AM, Michel Dänzer wrote:
On 22/04/17 02:05 AM, Felix Kuehling wrote:
__setup doesn't work in modules.
Right. We could build something like
drivers/video/fbdev/core/fb_cmdline.c:video_setup() into the kernel to
handle this, but it
On 04/27/2017 04:25 PM, Chunming Zhou wrote:
the case could happen when gpu reset:
1. when gpu reset, cs can be continue until sw queue is full, then push job
will wait with holding pd reservation.
2. gpu_reset routine will also need pd reservation to restore page table from
their shadow.
3. cs
v2: 64-bit aligned for gpu info
Signed-off-by: Ken Wang
Signed-off-by: Junwei Zhang
Reviewed-by: Alex Deucher
Reviewed-by: Qiang Yu
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 3 +++
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 11 +++
drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 3 +++
Signed-off-by: Junwei Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index ead00d7..857c3be 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dr
the case could happen when gpu reset:
1. when gpu reset, cs can be continue until sw queue is full, then push job
will wait with holding pd reservation.
2. gpu_reset routine will also need pd reservation to restore page table from
their shadow.
3. cs is waiting for gpu_reset complete, but gpu res
From: Michel Dänzer
This reverts commit cb341a319f7e66f879d69af929c3dadfc1a8f31e.
The purpose of the refactor was for amdgpu_crtc_prepare/submit_flip to
be used by the DC code, but that's no longer the case.
Signed-off-by: Michel Dänzer
---
drivers/gpu/drm/amd/amdgpu/amdgpu_display.c | 147 ++
pls ignore this one, will send another new one.
On 2017年04月27日 15:20, Chunming Zhou wrote:
the case could happen when gpu reset:
1. when gpu reset, cs can be continue until sw queue is full, then push job
will wait with holding pd reservation.
2. gpu_reset routine will also need pd reservation
It looks fine for me now.
Reviewed-by: Junwei Zhang
Jerry
On 04/27/2017 01:00 PM, Chunming Zhou wrote:
The current kernel implementation, which grabs the idle VMID from pool when
emitting the job may:
The back-to-back submission from one process could use different VMID.
The submi
On 04/27/2017 01:54 PM, Zhang, Jerry (Junwei) wrote:
On 04/27/2017 01:00 PM, Chunming Zhou wrote:
v2: move sync waiting only when flush needs
v3: fix racy
Change-Id: I64da2701c9fdcf986afb90ba1492a78d5bef1b6c
Signed-off-by: Chunming Zhou
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 72 +++
Reviewed-by: Xiangliang Yu
Thanks!
Xiangliang Yu
> -Original Message-
> From: Trigger Huang [mailto:trigger.hu...@amd.com]
> Sent: Thursday, April 27, 2017 3:18 PM
> To: amd-gfx@lists.freedesktop.org
> Cc: Liu, Monk ; Yu, Xiangliang
> ; Huang, Trigger
> Subject: [PATCH] drm/amdgpu: Ena
On 27/04/17 04:19 AM, Alex Xie wrote:
> Hi,
>
> I knew this is part of ioctl. And I intended to fix this interruptible
> waiting in an ioctl.
It's by design, nothing that needs fixing in the case of an ioctl.
> 1. The wait is short. There is not much benefit by interruptible
> waiting. The BO
Support for MCBP/Virtualization in combination with chained IBs is
formal released on firmware feature version #46. So enable it
according to firmware feature version, otherwise, world switch will
hang.
Signed-off-by: Trigger Huang
---
drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 17 +++--
the case could happen when gpu reset:
1. when gpu reset, cs can be continue until sw queue is full, then push job
will wait with holding pd reservation.
2. gpu_reset routine will also need pd reservation to restore page table from
their shadow.
3. cs is waiting for gpu_reset complete, but gpu res
Reviewed-by: Monk Liu
-Original Message-
From: Trigger Huang [mailto:trigger.hu...@amd.com]
Sent: Thursday, April 27, 2017 3:18 PM
To: amd-gfx@lists.freedesktop.org
Cc: Liu, Monk ; Yu, Xiangliang ;
Huang, Trigger
Subject: [PATCH] drm/amdgpu: Enable chained IB MCBP support
Support for
On 27/04/17 03:45 PM, Gerd Hoffmann wrote:
> Hi,
>
>>> That is done using the RADEON_TILING_SWAP_{16,32}BIT flag mentioned in
>>> another thread?
>>
>> Right.
>>
>>
>>> What about dumb bos? You've mentioned the swap flag isn't used for
>>> those. Which implies they are in little endian byte or
75 matches
Mail list logo