Hi Jesse,
yes, I know that's a well known bug.
The Intel guys have already narrowed it down to a missing
dma_fence_enable_signaling() in the syncobj code path.
I strongly suggest to work together with them to find where that needs
to be added instead.
Regards,
Christian.
Am 30.06.22 um 17:26 schrieb Zhang, Jesse(Jie):
[AMD Official Use Only - General]
Hi Christian,
If we remove the following patch, the "syncobj timeline test" can pass.
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=721255b52700b320c4ae2e23d57f7d9ad1db50b9
The following log is provided by AMD CQE team. They run the amdgpu_test tool on
ubuntu22 (kernel version 5.15.0-39)
Suite: SYNCOBJ TIMELINE Tests
Test: syncobj timeline test ...FAILED
1. sources/drm/tests/amdgpu/syncobj_tests.c:299 -
CU_ASSERT_EQUAL(payload,18)
2. sources/drm/tests/amdgpu/syncobj_tests.c:309 -
CU_ASSERT_EQUAL(payload,20)
You can get more detail information by the attachment.
So we need fix this issue. And if you have any better solution to solve the
issue, please let me know.
Thanks
Jesse
-----Original Message-----
From: Koenig, Christian <[email protected]>
Sent: Wednesday, 29 June 2022 5:12 pm
To: Zhang, Jesse(Jie) <[email protected]>; [email protected];
[email protected]
Cc: Mukunda, Vijendar <[email protected]>; Hiregoudar, Basavaraj <[email protected]>; Dommati, Sunil-kumar
<[email protected]>; Pandey, Ajit Kumar <[email protected]>; Nirmoy Das <[email protected]>; Maarten Lankhorst
<[email protected]>; Maxime Ripard <[email protected]>; Thomas Zimmermann <[email protected]>; David Airlie
<[email protected]>; Daniel Vetter <[email protected]>; Sumit Semwal <[email protected]>; open list:DRM DRIVERS
<[email protected]>; open list <[email protected]>; open list:DMA BUFFER SHARING FRAMEWORK <[email protected]>;
moderated list:DMA BUFFER SHARING FRAMEWORK <[email protected]>
Subject: Re: [PATCH v1] Fix: SYNCOBJ TIMELINE Test failed.
Am 29.06.22 um 08:02 schrieb jie1zhan:
The issue cause by the commit :
721255b527(drm/syncobj: flatten dma_fence_chains on transfer).
Because it use the point of dma_fence incorrectly
Correct the point of dma_fence by fence array
Well that patch is just utterly nonsense as far as I can see.
Signed-off-by: jie1zhan <[email protected]>
Reviewed-by: Christian König <[email protected]>
Reviewed-by: Nirmoy Das <[email protected]>
I have strong doubts that Nirmoy has reviewed this and I certainly haven't
reviewed it.
Christian.
---
drivers/gpu/drm/drm_syncobj.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/drm_syncobj.c
b/drivers/gpu/drm/drm_syncobj.c index 7e48dcd1bee4..d5db818f1c76
100644
--- a/drivers/gpu/drm/drm_syncobj.c
+++ b/drivers/gpu/drm/drm_syncobj.c
@@ -887,7 +887,7 @@ static int drm_syncobj_flatten_chain(struct dma_fence **f)
goto free_fences;
dma_fence_put(*f);
- *f = &array->base;
+ *f = array->fences[0];
return 0;
free_fences:
