On Fri, 10 Jan 2020 at 13:21, Andreas Beckmann <[email protected]> wrote: > > Control: tag -1 upstream > Control: forwarded -1 > https://devtalk.nvidia.com/default/topic/1068045/linux/5-4-kernel-breaks-prime-synchronization-/ > > On 10/01/2020 10.37, Raphaël Hertzog wrote: > > We had multiple reports of nvidia users having refresh issues ever since > > they switched to Linux 5.4. > > > > I believe they are affected by the loss of "PRIME synchronization" as > > explained in > > https://devtalk.nvidia.com/default/topic/1068045/linux/5-4-kernel-breaks-prime-synchronization-/ > > Luca, can you reproduce something like that? > > > There's a patch for 440.44 in that thread but I don't know of any upstream > > fix. Direct link to patch: https://gitlab.com/snippets/1927096 > > That patch looks like a crude hack. I have no clue what its intention > is. The only thing I understood for sure is that it abuses the > NV_DRM_DRIVER_HAS_GEM_PRIME_RES_OBJ condition, probably by turning it > into "true". It's also not backwards compatible. > I'm not going to apply it in the current shape. :-( > > > Andreas
Hi, I can reproduce, but I agree with you on the review of the patch, it doesn't look very clean. Given it only affects 5.4 it might be fine to wait for an upstream fix?
