Re: [PATCH 09/27] drm/etnaviv: don't flush workqueue in etnaviv_gpu_wait_obj_inactive

2017-12-04 Thread Russell King - ARM Linux
On Fri, Dec 01, 2017 at 11:36:06AM +0100, Lucas Stach wrote: > There is no need to synchronize with oustanding retire jobs if the object > has gone idle. Retire jobs only ever change the object state from active to > idle, not the other way around. > > The IOVA put race is uncritical, as the

Re: [PATCH 09/27] drm/etnaviv: don't flush workqueue in etnaviv_gpu_wait_obj_inactive

2017-12-01 Thread Lucas Stach
Hi Russell, Am Freitag, den 01.12.2017, 16:59 + schrieb Russell King - ARM Linux: > On Fri, Dec 01, 2017 at 11:36:06AM +0100, Lucas Stach wrote: > > There is no need to synchronize with oustanding retire jobs if the object > > has gone idle. Retire jobs only ever change the object state from

Re: [PATCH 09/27] drm/etnaviv: don't flush workqueue in etnaviv_gpu_wait_obj_inactive

2017-12-01 Thread Philipp Zabel
On Fri, 2017-12-01 at 11:36 +0100, Lucas Stach wrote: > There is no need to synchronize with oustanding retire jobs if the object > has gone idle. Retire jobs only ever change the object state from active to > idle, not the other way around. > > The IOVA put race is uncritical, as the GEM_WAIT

[PATCH 09/27] drm/etnaviv: don't flush workqueue in etnaviv_gpu_wait_obj_inactive

2017-12-01 Thread Lucas Stach
There is no need to synchronize with oustanding retire jobs if the object has gone idle. Retire jobs only ever change the object state from active to idle, not the other way around. The IOVA put race is uncritical, as the GEM_WAIT ioctl itself is holding a reference to the GEM object, so the