Re: [PATCH] drm/amdgpu: remove distinction between explicit and implicit sync (v2)

with this. This is for implicit sync, which is required by DRI3. This fix allows removing existing inefficiencies from drivers, so it's a good thing. Marek On Wed., Jun. 10, 2020, 03:56 Chunming Zhou, <mailto:zhou...@amd.com>> wrote: 在 2020/6/10 15:41, Christian König 写道: That's true, bu

Re: [PATCH] drm/amdgpu: remove distinction between explicit and implicit sync (v2)

在 2020/6/10 15:41, Christian König 写道: That's true, but for now we are stuck with the implicit sync for quite a number of use cases. My problem is rather that we already tried this and it backfired immediately. I do remember that it was your patch who introduced the pipeline sync flag

[PATCH] MAINTAINERS: Remove me from amdgpu maintainers

Glad to spend time on kernel driver in past years. I've moved to new focus in umd and couldn't commit enough time to discussions. Signed-off-by: Chunming Zhou --- MAINTAINERS | 1 - 1 file changed, 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 938316092634..4ca508bd4c9e 100644

Re: drm/amdgpu: apply AMDGPU_IB_FLAG_EMIT_MEM_SYNC to compute IBs too

Yes, same question. In fact, PAL cmd stream has itself Relase/Acquire packets. That we use the flag is per your request. -David 在 2020/4/27 22:53, Christian König 写道: Yeah, but is Mesa going to use it? Christian. Am 27.04.20 um 15:54 schrieb Marek Olšák: PAL requested it and they are

Re: [PATCH] drm/ttm: Schedule out if possibe in bo delayed delete worker

We can have both of yours, I think. Even switch to use spin_trylock, I think we are ok to have cond_resched() Xinhui added in this patch. That can give more chance to urgent task to use cpu. -David 在 2020/4/9 22:59, Christian König 写道: Why we break out the loops when there are pending bos

Re: [PATCH] drm/amdgpu: resvert "disable bulk moves for now"

RB on it to go ahead. -David 在 2019/9/12 18:15, Christian König 写道: > This reverts commit a213c2c7e235cfc0e0a161a558f7fdf2fb3a624a. > > The changes to fix this should have landed in 5.1. > > Signed-off-by: Christian König > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 2 -- > 1 file

Re: [PATCH] drm/amdgpu: grab the id mgr lock while accessing passid_mapping

Reviewed-by: Chunming Zhou 在 2019/9/10 16:56, Christian König 写道: > Ping! > > Am 09.09.19 um 13:59 schrieb Christian König: >> Need to make sure that we actually dropping the right fence. >> Could be done with RCU as well, but to complicated for a fix. >> >&g

Re: [PATCH 3/3] drm/amdgpu: remove amdgpu_cs_try_evict

Reviewed-by: Chunming Zhou for series. -David 在 2019/9/3 17:09, Christian König 写道: > Trying to evict things from the current working set doesn't work that > well anymore because of per VM BOs. > > Rely on reserving VRAM for page tables to avoid contention. > > Signed-off-b

Re: [PATCH] drm/amdgpu: fix dma_fence_wait without reference

Reviewed-by: Chunming Zhou 在 2019/8/16 21:21, Christian König 写道: > We need to grab a reference to the fence we wait for. > > Signed-off-by: Christian König > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 27 ++--- > 1 file changed, 15 insertion

Re: [PATCH] drm/amdgpu: fix a potential information leaking bug

he multiplication > 1024*sizeof(*data) as the size parameter for memset() though there is > no risk of integer overflow. > > Signed-off-by: Wang Xiayang Reviewed-by: Chunming Zhou -David > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 2 +- > 1 file changed, 1

Re: Intermittent errors when using amdgpu_job_submit_direct

在 2019/7/10 3:26, Kuehling, Felix 写道: > On 2019-07-09 8:58 a.m., Zhou, David(ChunMing) wrote: >> I've raised it up when Christian make page fault, at that patch, >> amdgpu_job_submit_direct uses exclusive page fault ring for that. >> >> But if you use amdgpu_job_submit_direct for gerneral rings

Re: [PATCH] drm/ttm: Fix the memory delay free issue

It doesn't make sense that freeing BO still uses per-vm resv. I remember when BO is in release list, its resv will be from per-vm resv copy. Could you check it? -David 在 2019/7/10 17:29, Emily Deng 写道: > For vulkan cts allocation test cases, they will create a series of bos, and > then free >

Re: Intermittent errors when using amdgpu_job_submit_direct

I've raised it up when Christian make page fault, at that patch, amdgpu_job_submit_direct uses exclusive page fault ring for that. But if you use amdgpu_job_submit_direct for gerneral rings ocuppied by scheduler, I guess varias bugs will happen. -David 在 2019/7/9 12:53, Kuehling, Felix 写道: >

Re: [PATCH 1/5] drm/amdgpu: allow direct submission in the VM backends

在 2019/6/28 20:18, Christian König 写道: > This allows us to update page tables directly while in a page fault. > > Signed-off-by: Christian König > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 5 > drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c | 4 +++ >

Re: [PATCH 1/2] drm/amdgpu: fix transform feedback GDS hang on gfx10

please take care of .emit_ib_size member, otherwise it looks ok to me. -David 在 2019/6/20 8:02, Marek Olšák 写道: > From: Marek Olšák > > Signed-off-by: Marek Olšák > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h | 3 ++- > drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 12 ++-- > 2

[PATCH] drm/amdgpu: add DRIVER_SYNCOBJ_TIMELINE to amdgpu

Change-Id: I2b1af1478fbddbb5084b90b3ff85c2eb964bd217 Signed-off-by: Chunming Zhou --- 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

Re: [PATCH 06/10] drm/ttm: fix busy memory to fail other user v10

在 2019/5/23 19:03, Christian König 写道: > [CAUTION: External Email] > > Am 23.05.19 um 12:24 schrieb zhoucm1: >> >> >> On 2019年05月22日 20:59, Christian König wrote: >>> [CAUTION: External Email] >>> >>> BOs on the LRU might be blocked during command submission >>> and cause OOM situations. >>> >>>

[PATCH libdrm 3/7] wrap syncobj timeline query/wait APIs for amdgpu v3

v2: symbos are stored in lexical order. v3: drop export/import and extra query indirection Signed-off-by: Chunming Zhou Acked-by: Christian König --- amdgpu/amdgpu-symbol-check | 2 ++ amdgpu/amdgpu.h| 39 ++ amdgpu/amdgpu_cs.c | 23

[PATCH libdrm 2/7] add timeline wait/query ioctl v2

v2: drop export/import Signed-off-by: Chunming Zhou --- xf86drm.c | 44 xf86drm.h | 6 ++ 2 files changed, 50 insertions(+) diff --git a/xf86drm.c b/xf86drm.c index 2c19376b..17e3d880 100644 --- a/xf86drm.c +++ b/xf86drm.c @@ -4256,3 +4256,47

[PATCH libdrm 7/7] add syncobj timeline tests v3

v2: drop DRM_SYNCOBJ_CREATE_TYPE_TIMELINE, fix timeout calculation, fix some warnings v3: add export/import and cpu signal testing cases Signed-off-by: Chunming Zhou Acked-by: Christian König --- tests/amdgpu/Makefile.am | 3 +- tests/amdgpu/amdgpu_test.c | 11 ++ tests/amdgpu

[PATCH libdrm 6/7] wrap transfer interfaces

Signed-off-by: Chunming Zhou Acked-by: Christian König --- amdgpu/amdgpu.h| 22 ++ amdgpu/amdgpu_cs.c | 16 2 files changed, 38 insertions(+) diff --git a/amdgpu/amdgpu.h b/amdgpu/amdgpu.h index d2480dbe..9d9b0832 100644 --- a/amdgpu/amdgpu.h +++ b

[PATCH libdrm 4/7] add timeline signal/transfer ioctls v2

v2: use one transfer ioctl Signed-off-by: Chunming Zhou --- xf86drm.c | 33 + xf86drm.h | 6 ++ 2 files changed, 39 insertions(+) diff --git a/xf86drm.c b/xf86drm.c index 17e3d880..acd16fab 100644 --- a/xf86drm.c +++ b/xf86drm.c @@ -4257,6 +4257,21

[PATCH libdrm 5/7] expose timeline signal/export/import interfaces v2

v2: adapt to new one transfer ioctl Signed-off-by: Chunming Zhou Acked-by: Christian König --- amdgpu/amdgpu-symbol-check | 3 ++ amdgpu/amdgpu.h| 51 amdgpu/amdgpu_cs.c | 68 ++ 3 files changed, 122

[PATCH libdrm 1/7] addr cs chunk for syncobj timeline

Signed-off-by: Chunming Zhou --- include/drm/amdgpu_drm.h | 9 + 1 file changed, 9 insertions(+) diff --git a/include/drm/amdgpu_drm.h b/include/drm/amdgpu_drm.h index d0701ffc..3d0318e6 100644 --- a/include/drm/amdgpu_drm.h +++ b/include/drm/amdgpu_drm.h @@ -528,6 +528,8 @@ struct

[PATCH 1/2] drm/ttm: fix busy memory to fail other user v6

: fix some missing v5: handle first_bo unlock and bo_get/put v6: abstract unified iterate function, and handle all possible usecase not only pinned bo. Change-Id: I21423fb922f885465f13833c41df1e134364a8e7 Signed-off-by: Chunming Zhou --- drivers/gpu/drm/ttm/ttm_

[PATCH 2/2] drm/amd/display: use ttm_eu_reserve_buffers instead of amdgpu_bo_reserve

add ticket for display bo, so that it can preempt busy bo. Change-Id: I9f031cdcc8267de00e819ae303baa0a52df8ebb9 Signed-off-by: Chunming Zhou --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 22 ++- 1 file changed, 17 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm

Re: [PATCH v5 6/6] drm/amdgpu: Avoid HW reset if guilty job already signaled.

+Monk. GPU reset is used widely in SRIOV, so need virtulizatino guy take a look. But out of curious, why guilty job can signal more if the job is already set to guilty? set it wrongly? -David 在 2019/4/18 23:00, Andrey Grodzovsky 写道: > Also reject TDRs if another one already running. > > v2:

Re: [PATCH v5 4/6] drm/sched: Keep s_fence->parent pointer

+Monk to response this patch. 在 2019/4/18 23:00, Andrey Grodzovsky 写道: > For later driver's reference to see if the fence is signaled. > > v2: Move parent fence put to resubmit jobs. > > Signed-off-by: Andrey Grodzovsky > Reviewed-by: Christian König > --- >

Re: [PATCH v5 3/6] drm/scheduler: rework job destruction

Hi Andrey, static void drm_sched_process_job(struct dma_fence *f, struct dma_fence_cb *cb) { ...     spin_lock_irqsave(>job_list_lock, flags);     /* remove job from ring_mirror_list */     list_del_init(_job->node);     spin_unlock_irqrestore(>job_list_lock, flags); [David]

Re: dynamic DMA-buf sharing between devices

> > Am 17.04.19 um 15:52 schrieb Chunming Zhou: >> Thanks Christian, great job. I will verify it this week when I finish my >> current work on hand. >> >> -David >> >> 在 2019/4/17 2:38, Christian König wrote: >>> Hi everybody, >>>

Re: dynamic DMA-buf sharing between devices

Thanks Christian, great job. I will verify it this week when I finish my current work on hand. -David 在 2019/4/17 2:38, Christian König wrote: > Hi everybody, > > core idea in this patch set is that DMA-buf importers can now provide an > optional invalidate callback. Using this callback and

[PATCH libdrm 5/8] add timeline signal/transfer ioctls v2

v2: use one transfer ioctl Signed-off-by: Chunming Zhou --- xf86drm.c | 33 + xf86drm.h | 6 ++ 2 files changed, 39 insertions(+) diff --git a/xf86drm.c b/xf86drm.c index 66e0c985..d57c4218 100644 --- a/xf86drm.c +++ b/xf86drm.c @@ -4280,6 +4280,21

[PATCH libdrm 7/8] wrap transfer interfaces

Signed-off-by: Chunming Zhou --- amdgpu/amdgpu.h| 22 ++ amdgpu/amdgpu_cs.c | 16 2 files changed, 38 insertions(+) diff --git a/amdgpu/amdgpu.h b/amdgpu/amdgpu.h index b5bd3ed9..2350835b 100644 --- a/amdgpu/amdgpu.h +++ b/amdgpu/amdgpu.h @@ -1670,6

[PATCH libdrm 3/8] add timeline wait/query ioctl v2

v2: drop export/import Signed-off-by: Chunming Zhou --- xf86drm.c | 44 xf86drm.h | 6 ++ 2 files changed, 50 insertions(+) diff --git a/xf86drm.c b/xf86drm.c index 18ad7c58..66e0c985 100644 --- a/xf86drm.c +++ b/xf86drm.c @@ -4279,3 +4279,47

[PATCH libdrm 2/8] addr cs chunk for syncobj timeline

Signed-off-by: Chunming Zhou --- include/drm/amdgpu_drm.h | 9 + 1 file changed, 9 insertions(+) diff --git a/include/drm/amdgpu_drm.h b/include/drm/amdgpu_drm.h index e3a97da4..ab53f2e0 100644 --- a/include/drm/amdgpu_drm.h +++ b/include/drm/amdgpu_drm.h @@ -528,6 +528,8 @@ struct

[PATCH libdrm 6/8] expose timeline signal/export/import interfaces v2

v2: adapt to new one transfer ioctl Signed-off-by: Chunming Zhou --- amdgpu/amdgpu-symbol-check | 3 ++ amdgpu/amdgpu.h| 51 amdgpu/amdgpu_cs.c | 68 ++ 3 files changed, 122 insertions(+) diff --git a/amdgpu

[PATCH libdrm 4/8] wrap syncobj timeline query/wait APIs for amdgpu v3

v2: symbos are stored in lexical order. v3: drop export/import and extra query indirection Signed-off-by: Chunming Zhou Signed-off-by: Christian König --- amdgpu/amdgpu-symbol-check | 2 ++ amdgpu/amdgpu.h| 39 ++ amdgpu/amdgpu_cs.c | 23

[PATCH libdrm 8/8] add syncobj timeline tests v3

v2: drop DRM_SYNCOBJ_CREATE_TYPE_TIMELINE, fix timeout calculation, fix some warnings v3: add export/import and cpu signal testing cases Signed-off-by: Chunming Zhou Signed-off-by: Christian König --- tests/amdgpu/Makefile.am | 3 +- tests/amdgpu/amdgpu_test.c | 11 ++ tests

[PATCH libdrm 1/8] new syncobj extension v3

v2: drop not implemented IOCTLs and flags v3: add transfer/signal ioctls Signed-off-by: Chunming Zhou Signed-off-by: Christian König --- include/drm/drm.h | 35 +++ 1 file changed, 35 insertions(+) diff --git a/include/drm/drm.h b/include/drm/drm.h index

Re: [PATCH] drm/amdgpu: fix old fence check in amdgpu_fence_emit

kely(rcu_dereference_protected(*ptr, 1))) { >>>> Isn't this line redundant with dma_fence_get_rcu_safe? I think it's >>>> unnecessary. >>>> Otherwise looks ok to me. >>> The key point is lock()+dma_fence_get_rcu_safe(ptr)+unlock() is rather

Re: [PATCH] drm/amdgpu: fix old fence check in amdgpu_fence_emit

在 2019/4/1 21:05, Christian König 写道: > Am 01.04.19 um 04:54 schrieb Zhou, David(ChunMing): >> >>> -Original Message- >>> From: amd-gfx On Behalf Of >>> Christian K?nig >>> Sent: Saturday, March 30, 2019 2:33 AM >>> To: amd-gfx@lists.freedesktop.org >>> Subject: [PATCH] drm/amdgpu: fix

[PATCH libdrm 4/8] wrap syncobj timeline query/wait APIs for amdgpu v3

v2: symbos are stored in lexical order. v3: drop export/import and extra query indirection Signed-off-by: Chunming Zhou Signed-off-by: Christian König --- amdgpu/amdgpu-symbol-check | 2 ++ amdgpu/amdgpu.h| 39 ++ amdgpu/amdgpu_cs.c | 23

[PATCH libdrm 8/8] add syncobj timeline tests v3

v2: drop DRM_SYNCOBJ_CREATE_TYPE_TIMELINE, fix timeout calculation, fix some warnings v3: add export/import and cpu signal testing cases Signed-off-by: Chunming Zhou Signed-off-by: Christian König --- tests/amdgpu/Makefile.am | 3 +- tests/amdgpu/amdgpu_test.c | 12 ++ tests

[PATCH libdrm 5/8] add timeline signal/transfer ioctls v2

v2: use one transfer ioctl Signed-off-by: Chunming Zhou --- xf86drm.c | 33 + xf86drm.h | 6 ++ 2 files changed, 39 insertions(+) diff --git a/xf86drm.c b/xf86drm.c index 66e0c985..d57c4218 100644 --- a/xf86drm.c +++ b/xf86drm.c @@ -4280,6 +4280,21

[PATCH libdrm 7/8] wrap transfer interfaces

Signed-off-by: Chunming Zhou --- amdgpu/amdgpu.h| 22 ++ amdgpu/amdgpu_cs.c | 16 2 files changed, 38 insertions(+) diff --git a/amdgpu/amdgpu.h b/amdgpu/amdgpu.h index b5bd3ed9..2350835b 100644 --- a/amdgpu/amdgpu.h +++ b/amdgpu/amdgpu.h @@ -1670,6

[PATCH libdrm 6/8] expose timeline signal/export/import interfaces v2

v2: adapt to new one transfer ioctl Signed-off-by: Chunming Zhou --- amdgpu/amdgpu-symbol-check | 3 ++ amdgpu/amdgpu.h| 51 amdgpu/amdgpu_cs.c | 68 ++ 3 files changed, 122 insertions(+) diff --git a/amdgpu

[PATCH libdrm 3/8] add timeline wait/query ioctl v2

v2: drop export/import Signed-off-by: Chunming Zhou --- xf86drm.c | 44 xf86drm.h | 6 ++ 2 files changed, 50 insertions(+) diff --git a/xf86drm.c b/xf86drm.c index 18ad7c58..66e0c985 100644 --- a/xf86drm.c +++ b/xf86drm.c @@ -4279,3 +4279,47

[PATCH libdrm 2/8] addr cs chunk for syncobj timeline

Signed-off-by: Chunming Zhou --- include/drm/amdgpu_drm.h | 9 + 1 file changed, 9 insertions(+) diff --git a/include/drm/amdgpu_drm.h b/include/drm/amdgpu_drm.h index be84e43c..bfa04dd8 100644 --- a/include/drm/amdgpu_drm.h +++ b/include/drm/amdgpu_drm.h @@ -523,6 +523,8 @@ struct

[PATCH libdrm 1/8] new syncobj extension v3

v2: drop not implemented IOCTLs and flags v3: add transfer/signal ioctls Signed-off-by: Chunming Zhou Signed-off-by: Christian König --- include/drm/drm.h | 35 +++ 1 file changed, 35 insertions(+) diff --git a/include/drm/drm.h b/include/drm/drm.h index

[PATCH 6/9] drm/amdgpu: add timeline support in amdgpu CS v3

syncobj wait/signal operation is appending in command submission. v2: separate to two kinds in/out_deps functions v3: fix checking for timeline syncobj Signed-off-by: Chunming Zhou Cc: Tobias Hector Cc: Jason Ekstrand Cc: Dave Airlie Cc: Chris Wilson Cc: Lionel Landwerlin Reviewed

[PATCH 7/9] drm/syncobj: add transition iotcls between binary and timeline v2

we need to import/export timeline point. v2: unify to one transfer ioctl Signed-off-by: Chunming Zhou Cc: Lionel Landwerlin Reviewed-by: Lionel Landwerlin --- drivers/gpu/drm/drm_internal.h | 2 + drivers/gpu/drm/drm_ioctl.c| 2 + drivers/gpu/drm/drm_syncobj.c | 74

[PATCH 5/9] drm/syncobj: use the timeline point in drm_syncobj_find_fence v4

From: Christian König Implement finding the right timeline point in drm_syncobj_find_fence. v2: return -EINVAL when the point is not submitted yet. v3: fix reference counting bug, add flags handling as well v4: add timeout for find fence Signed-off-by: Christian König Cc: Lionel Landwerlin

[PATCH 8/9] drm/syncobj: add timeline signal ioctl for syncobj v5

Signed-off-by: Chunming Zhou Cc: Tobias Hector Cc: Jason Ekstrand Cc: Dave Airlie Cc: Chris Wilson Cc: Lionel Landwerlin Reviewed-by: Lionel Landwerlin --- drivers/gpu/drm/drm_internal.h | 2 + drivers/gpu/drm/drm_ioctl.c| 2 + drivers/gpu/drm/drm_syncobj.c | 73

[PATCH 9/9] drm/amdgpu: update version for timeline syncobj support in amdgpu

Signed-off-by: Chunming Zhou Reviewed-by: Lionel Landwerlin --- 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 8a0732088640..4d8db87048d3

[PATCH 4/9] drm/syncobj: add timeline payload query ioctl v6

user mode can query timeline payload. v2: check return value of copy_to_user v3: handle querying entry by entry v4: rebase on new chain container, simplify interface v5: query last signaled timeline point, not last point. v6: add unorder point check Signed-off-by: Chunming Zhou Cc: Tobias Hector

[PATCH 3/9] drm/syncobj: add support for timeline point wait v8

for xxx_WAIT_AVAILABLE v6: rebase and rework on new container v7: drop _WAIT_COMPLETED, it is the default anyway v8: correctly handle garbage collected fences Signed-off-by: Chunming Zhou Signed-off-by: Christian König Cc: Tobias Hector Cc: Jason Ekstrand Cc: Dave Airlie Cc: Chris Wilson Cc: Lionel

[PATCH 2/9] drm/syncobj: add new drm_syncobj_add_point interface v4

From: Christian König Use the dma_fence_chain object to create a timeline of fence objects instead of just replacing the existing fence. v2: rebase and cleanup v3: fix garbage collection parameters v4: add unorder point check, print a warn calltrace Signed-off-by: Christian König Cc: Lionel

[PATCH 0/9] *** timeline syncobj support ***

new dma_fence_chain container v7 drm/syncobj: add new drm_syncobj_add_point interface v4 drm/syncobj: use the timeline point in drm_syncobj_find_fence v4 Chunming Zhou (6): drm/syncobj: add support for timeline point wait v8 drm/syncobj: add timeline payload query ioctl v6 drm/amdgpu: add

[PATCH 1/9] dma-buf: add new dma_fence_chain container v7

From: Christian König Lockless container implementation similar to a dma_fence_array, but with only two elements per node and automatic garbage collection. v2: properly document dma_fence_chain_for_each, add dma_fence_chain_find_seqno, drop prev reference during garbage collection if it's

Re: [PATCH 8/9] drm/syncobj: add timeline signal ioctl for syncobj v5

在 2019/3/28 20:53, Lionel Landwerlin 写道: > On 25/03/2019 08:32, Chunming Zhou wrote: >> v2: individually allocate chain array, since chain node is free >> independently. >> v3: all existing points must be already signaled before cpu perform >> signal operation, >

[PATCH 8/9] drm/syncobj: add timeline signal ioctl for syncobj v5

Signed-off-by: Chunming Zhou Cc: Tobias Hector Cc: Jason Ekstrand Cc: Dave Airlie Cc: Chris Wilson Cc: Lionel Landwerlin --- drivers/gpu/drm/drm_internal.h | 2 + drivers/gpu/drm/drm_ioctl.c| 2 + drivers/gpu/drm/drm_syncobj.c | 73 ++ include/uapi/drm

[PATCH 9/9] drm/amdgpu: update version for timeline syncobj support in amdgpu

Signed-off-by: Chunming Zhou --- 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 8a0732088640..4d8db87048d3 100644 --- a/drivers/gpu/drm/amd

[PATCH 2/9] drm/syncobj: add new drm_syncobj_add_point interface v4

From: Christian König Use the dma_fence_chain object to create a timeline of fence objects instead of just replacing the existing fence. v2: rebase and cleanup v3: fix garbage collection parameters v4: add unorder point check, print a warn calltrace Signed-off-by: Christian König Cc: Lionel

[PATCH 7/9] drm/syncobj: add transition iotcls between binary and timeline v2

we need to import/export timeline point. v2: unify to one transfer ioctl Signed-off-by: Chunming Zhou Cc: Lionel Landwerlin --- drivers/gpu/drm/drm_internal.h | 2 + drivers/gpu/drm/drm_ioctl.c| 2 + drivers/gpu/drm/drm_syncobj.c | 74 ++ include/uapi

[PATCH 6/9] drm/amdgpu: add timeline support in amdgpu CS v3

syncobj wait/signal operation is appending in command submission. v2: separate to two kinds in/out_deps functions v3: fix checking for timeline syncobj Signed-off-by: Chunming Zhou Cc: Tobias Hector Cc: Jason Ekstrand Cc: Dave Airlie Cc: Chris Wilson Cc: Lionel Landwerlin --- drivers/gpu

[PATCH 5/9] drm/syncobj: use the timeline point in drm_syncobj_find_fence v4

From: Christian König Implement finding the right timeline point in drm_syncobj_find_fence. v2: return -EINVAL when the point is not submitted yet. v3: fix reference counting bug, add flags handling as well v4: add timeout for find fence Signed-off-by: Christian König Cc: Lionel Landwerlin

[PATCH 1/9] dma-buf: add new dma_fence_chain container v7

From: Christian König Lockless container implementation similar to a dma_fence_array, but with only two elements per node and automatic garbage collection. v2: properly document dma_fence_chain_for_each, add dma_fence_chain_find_seqno, drop prev reference during garbage collection if it's

[PATCH 4/9] drm/syncobj: add timeline payload query ioctl v6

user mode can query timeline payload. v2: check return value of copy_to_user v3: handle querying entry by entry v4: rebase on new chain container, simplify interface v5: query last signaled timeline point, not last point. v6: add unorder point check Signed-off-by: Chunming Zhou Cc: Tobias Hector

[PATCH 3/9] drm/syncobj: add support for timeline point wait v8

for xxx_WAIT_AVAILABLE v6: rebase and rework on new container v7: drop _WAIT_COMPLETED, it is the default anyway v8: correctly handle garbage collected fences Signed-off-by: Chunming Zhou Signed-off-by: Christian König Cc: Tobias Hector Cc: Jason Ekstrand Cc: Dave Airlie Cc: Chris Wilson Cc: Lionel

[PATCH 7/9] drm/syncobj: add transition iotcls between binary and timeline v2

we need to import/export timeline point. v2: unify to one transfer ioctl Signed-off-by: Chunming Zhou Cc: Lionel Landwerlin --- drivers/gpu/drm/drm_internal.h | 2 + drivers/gpu/drm/drm_ioctl.c| 2 + drivers/gpu/drm/drm_syncobj.c | 74 ++ include/uapi

[PATCH 6/9] drm/amdgpu: add timeline support in amdgpu CS v3

syncobj wait/signal operation is appending in command submission. v2: separate to two kinds in/out_deps functions v3: fix checking for timeline syncobj Signed-off-by: Chunming Zhou Cc: Tobias Hector Cc: Jason Ekstrand Cc: Dave Airlie Cc: Chris Wilson Cc: Lionel Landwerlin --- drivers/gpu

[PATCH 9/9] drm/amdgpu: update version for timeline syncobj support in amdgpu

Signed-off-by: Chunming Zhou --- 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 8a0732088640..4d8db87048d3 100644 --- a/drivers/gpu/drm/amd

[PATCH 3/9] drm/syncobj: add support for timeline point wait v8

for xxx_WAIT_AVAILABLE v6: rebase and rework on new container v7: drop _WAIT_COMPLETED, it is the default anyway v8: correctly handle garbage collected fences Signed-off-by: Chunming Zhou Signed-off-by: Christian König Cc: Tobias Hector Cc: Jason Ekstrand Cc: Dave Airlie Cc: Chris Wilson Cc: Lionel

[PATCH 8/9] drm/syncobj: add timeline signal ioctl for syncobj v4

v2: individually allocate chain array, since chain node is free independently. v3: all existing points must be already signaled before cpu perform signal operation, so add check condition for that. v4: remove v3 change and add checking to prevent out-of-order Signed-off-by: Chunming Zhou Cc

[PATCH 5/9] drm/syncobj: use the timeline point in drm_syncobj_find_fence v4

From: Christian König Implement finding the right timeline point in drm_syncobj_find_fence. v2: return -EINVAL when the point is not submitted yet. v3: fix reference counting bug, add flags handling as well v4: add timeout for find fence Signed-off-by: Christian König Cc: Lionel Landwerlin

[PATCH 4/9] drm/syncobj: add timeline payload query ioctl v6

user mode can query timeline payload. v2: check return value of copy_to_user v3: handle querying entry by entry v4: rebase on new chain container, simplify interface v5: query last signaled timeline point, not last point. v6: add unorder point check Signed-off-by: Chunming Zhou Cc: Tobias Hector

[PATCH 1/9] dma-buf: add new dma_fence_chain container v6

From: Christian König Lockless container implementation similar to a dma_fence_array, but with only two elements per node and automatic garbage collection. v2: properly document dma_fence_chain_for_each, add dma_fence_chain_find_seqno, drop prev reference during garbage collection if it's

[PATCH 2/9] drm/syncobj: add new drm_syncobj_add_point interface v4

From: Christian König Use the dma_fence_chain object to create a timeline of fence objects instead of just replacing the existing fence. v2: rebase and cleanup v3: fix garbage collection parameters v4: add unorder point check, print a warn calltrace Signed-off-by: Christian König Cc: Lionel

[PATCH 8/9] drm/syncobj: add timeline signal ioctl for syncobj v3

v2: individually allocate chain array, since chain node is free independently. v3: all existing points must be already signaled before cpu perform signal operation, so add check condition for that. Signed-off-by: Chunming Zhou --- drivers/gpu/drm/drm_internal.h | 2 + drivers/gpu/drm

[PATCH 6/9] drm/amdgpu: add timeline support in amdgpu CS v3

syncobj wait/signal operation is appending in command submission. v2: separate to two kinds in/out_deps functions v3: fix checking for timeline syncobj Signed-off-by: Chunming Zhou Cc: Daniel Rakos Cc: Jason Ekstrand Cc: Bas Nieuwenhuizen Cc: Dave Airlie Cc: Christian König Cc: Chris Wilson

[PATCH 5/9] drm/syncobj: use the timeline point in drm_syncobj_find_fence v3

From: Christian König Implement finding the right timeline point in drm_syncobj_find_fence. v2: return -EINVAL when the point is not submitted yet. v3: fix reference counting bug, add flags handling as well v4: add timeout for find fence Signed-off-by: Christian König ---

[PATCH 9/9] drm/amdgpu: update version for timeline syncobj support in amdgpu

Signed-off-by: Chunming Zhou --- 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 8a0732088640..4d8db87048d3 100644 --- a/drivers/gpu/drm/amd

[PATCH 4/9] drm/syncobj: add timeline payload query ioctl v6

user mode can query timeline payload. v2: check return value of copy_to_user v3: handle querying entry by entry v4: rebase on new chain container, simplify interface v5: query last signaled timeline point, not last point. v6: add unorder point check Signed-off-by: Chunming Zhou Cc: Daniel Rakos

[PATCH 7/9] drm/syncobj: add transition iotcls between binary and timeline v2

we need to import/export timeline point. v2: unify to one transfer ioctl Signed-off-by: Chunming Zhou --- drivers/gpu/drm/drm_internal.h | 2 + drivers/gpu/drm/drm_ioctl.c| 2 + drivers/gpu/drm/drm_syncobj.c | 74 ++ include/uapi/drm/drm.h | 10

[PATCH 3/9] drm/syncobj: add support for timeline point wait v8

for xxx_WAIT_AVAILABLE v6: rebase and rework on new container v7: drop _WAIT_COMPLETED, it is the default anyway v8: correctly handle garbage collected fences Signed-off-by: Chunming Zhou Signed-off-by: Christian König Cc: Daniel Rakos Cc: Jason Ekstrand Cc: Bas Nieuwenhuizen Cc: Dave Airlie Cc: Chris Wilson

[PATCH 2/9] drm/syncobj: add new drm_syncobj_add_point interface v4

From: Christian König Use the dma_fence_chain object to create a timeline of fence objects instead of just replacing the existing fence. v2: rebase and cleanup v3: fix garbage collection parameters v4: add unorder point check, print a warn calltrace Signed-off-by: Christian König ---

[PATCH 1/9] dma-buf: add new dma_fence_chain container v5

From: Christian König Lockless container implementation similar to a dma_fence_array, but with only two elements per node and automatic garbage collection. v2: properly document dma_fence_chain_for_each, add dma_fence_chain_find_seqno, drop prev reference during garbage collection if it's

Re: [PATCH 1/3] drm/amdgpu: remove non-sense NULL ptr check

The series is Reviewed-by: Chunming Zhou 在 2019/3/8 22:31, Christian König 写道: > It's a bug having a dead pointer in the IDR, silently returning > is the worst we can do. > > Signed-off-by: Christian König > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 10 -- >

[PATCH 9/9] drm/amdgpu: update version for timeline syncobj support in amdgpu

Signed-off-by: Chunming Zhou --- 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 8a0732088640..4d8db87048d3 100644 --- a/drivers/gpu/drm/amd

[PATCH 6/9] drm/amdgpu: add timeline support in amdgpu CS v3

syncobj wait/signal operation is appending in command submission. v2: separate to two kinds in/out_deps functions v3: fix checking for timeline syncobj Signed-off-by: Chunming Zhou Cc: Daniel Rakos Cc: Jason Ekstrand Cc: Bas Nieuwenhuizen Cc: Dave Airlie Cc: Christian König Cc: Chris Wilson

[PATCH 8/9] drm/syncobj: add timeline signal ioctl for syncobj v3

v2: individually allocate chain array, since chain node is free independently. v3: all existing points must be already signaled before cpu perform signal operation, so add check condition for that. Signed-off-by: Chunming Zhou --- drivers/gpu/drm/drm_internal.h | 2 + drivers/gpu/drm

[PATCH 4/9] drm/syncobj: add timeline payload query ioctl v5

user mode can query timeline payload. v2: check return value of copy_to_user v3: handle querying entry by entry v4: rebase on new chain container, simplify interface v5: query last signaled timeline point, not last point. Signed-off-by: Chunming Zhou Cc: Daniel Rakos Cc: Jason Ekstrand Cc: Bas

[PATCH 5/9] drm/syncobj: use the timeline point in drm_syncobj_find_fence v3

From: Christian König Implement finding the right timeline point in drm_syncobj_find_fence. v2: return -EINVAL when the point is not submitted yet. v3: fix reference counting bug, add flags handling as well Signed-off-by: Christian König --- drivers/gpu/drm/drm_syncobj.c | 43

[PATCH 3/9] drm/syncobj: add support for timeline point wait v8

for xxx_WAIT_AVAILABLE v6: rebase and rework on new container v7: drop _WAIT_COMPLETED, it is the default anyway v8: correctly handle garbage collected fences Signed-off-by: Chunming Zhou Signed-off-by: Christian König Cc: Daniel Rakos Cc: Jason Ekstrand Cc: Bas Nieuwenhuizen Cc: Dave Airlie Cc: Chris Wilson

[PATCH 2/9] drm/syncobj: add new drm_syncobj_add_point interface v3

From: Christian König Use the dma_fence_chain object to create a timeline of fence objects instead of just replacing the existing fence. v2: rebase and cleanup v3: fix garbage collection parameters Signed-off-by: Christian König --- drivers/gpu/drm/drm_syncobj.c | 37

[PATCH 1/9] dma-buf: add new dma_fence_chain container v5

From: Christian König Lockless container implementation similar to a dma_fence_array, but with only two elements per node and automatic garbage collection. v2: properly document dma_fence_chain_for_each, add dma_fence_chain_find_seqno, drop prev reference during garbage collection if it's

[PATCH] drm/amdgpu: enable bo priority setting from user space

Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c| 13 + drivers/gpu/drm/amd/amdgpu/amdgpu_gem.h| 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 3 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_object.h

Re: [PATCH] drm/amdgpu: force to use CPU_ACCESS hint optimization

t;>> ; amd-gfx@lists.freedesktop.org >>> Subject: Re: [PATCH] drm/amdgpu: force to use CPU_ACCESS hint >>> optimization >>> >>> Am 06.03.19 um 12:52 schrieb Chunming Zhou: >>>> As we know, visible vram can be placed to invisible when no cpu >>

[PATCH] drm/amdgpu: force to use CPU_ACCESS hint optimization

As we know, visible vram can be placed to invisible when no cpu access. Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 8 +++- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu

Re: [PATCH] drm/amdgpu: csa_vaddr should not larger than AMDGPU_GMC_HOLE_START

在 2019/1/18 17:11, Christian König 写道: Hi Monk, You see that for UMD, it can use 0 to HOLE_START Let me say it once more: The UMD nor anybody else CAN'T use 0 to HOLE_START, that region is reserved for the ATC hardware! We unfortunately didn't knew that initially and also didn't used the ATC,

  1   2   3   4   5   6   7   >