On 2021-12-02 at 14:54:21 +0530, Tejas Upadhyay wrote:
> Replace filling the GGTT entirely with scratch pages to avoid invalid
> accesses from VT-d overfetch of scanout by only surrounding scanout vma
> with guard pages. This eliminates the 100+ms delay in resume where we
> have to repopulate the G
Replace filling the GGTT entirely with scratch pages to avoid invalid
accesses from VT-d overfetch of scanout by only surrounding scanout vma
with guard pages. This eliminates the 100+ms delay in resume where we
have to repopulate the GGTT with scratch.
This should also help in avoiding slow suspe
Replace filling the GGTT entirely with scratch pages to avoid invalid
accesses from VT-d overfetch of scanout by only surrounding scanout vma
with guard pages. This eliminates the 100+ms delay in resume where we
have to repopulate the GGTT with scratch.
This should also help in avoiding slow suspe
