[Bug 111306] gbm creates BO with wrong pitch when dri3_get_modifiers returns modifiers, causing drmModeAddFB2WithModifiers to fail

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=111306

--- Comment #5 from Lionel Landwerlin <lionel.g.landwerlin at linux.intel.com> ---
Meeeeh... 

Looking at this I wonder how anything actually works.

The first issue is that i965 doesn't appear to look at the GBM_BO_USE_SCANOUT
flag in usage to compute the alignment restrictions of its surface.
So if I get this right you should get the same issue without using modifiers.
And we just happen to have things working by chance...

Second I see a fundamental issue with the GBM api.
The API doesn't let you specify a usage together with modifiers, but these are
2 sets of restrictions.
It's perfectly legal to create compressed surfaces which alignment's properties
don't match display requirements.

Fixing the first issue is trivial.
Not that easy with the second one.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-3d-bugs/attachments/20190808/318f1ce7/attachment.html>