search for: vhost_set_vmap_dirti

On Thu, Mar 07, 2019 at 10:45:57AM +0800, Jason Wang wrote: > > On 2019/3/7 ??12:31, Michael S. Tsirkin wrote: > > > +static void vhost_set_vmap_dirty(struct vhost_vmap *used) > > > +{ > > > + int i; > > > + > > > + for (i = 0; i < used->npages; i++) > > > + set_page_dirty_lock(used->pages[i]); > > This seems to rely on

On Thu, Mar 07, 2019 at 10:45:57AM +0800, Jason Wang wrote: > > On 2019/3/7 ??12:31, Michael S. Tsirkin wrote: > > > +static void vhost_set_vmap_dirty(struct vhost_vmap *used) > > > +{ > > > + int i; > > > + > > > + for (i = 0; i < used->npages; i++) > > > + set_page_dirty_lock(used->pages[i]); > > This seems to rely on

On Wed, Mar 06, 2019 at 02:18:12AM -0500, Jason Wang wrote: > It was noticed that the copy_user() friends that was used to access > virtqueue metdata tends to be very expensive for dataplane > implementation like vhost since it involves lots of software checks, > speculation barrier, hardware feature toggling (e.g SMAP). The > extra cost will be more obvious when transferring small

On Wed, Mar 06, 2019 at 02:18:12AM -0500, Jason Wang wrote: > It was noticed that the copy_user() friends that was used to access > virtqueue metdata tends to be very expensive for dataplane > implementation like vhost since it involves lots of software checks, > speculation barrier, hardware feature toggling (e.g SMAP). The > extra cost will be more obvious when transferring small

On 2019/3/7 ??12:31, Michael S. Tsirkin wrote: >> +static void vhost_set_vmap_dirty(struct vhost_vmap *used) >> +{ >> + int i; >> + >> + for (i = 0; i < used->npages; i++) >> + set_page_dirty_lock(used->pages[i]); > This seems to rely on page lock to mark page dirty. > > Could it happen that page writeback will check the > page, find it

It was noticed that the copy_user() friends that was used to access virtqueue metdata tends to be very expensive for dataplane implementation like vhost since it involves lots of software checks, speculation barrier, hardware feature toggling (e.g SMAP). The extra cost will be more obvious when transferring small packets since the time spent on metadata accessing become more significant. This

On Thu, Mar 07, 2019 at 10:34:39AM -0500, Michael S. Tsirkin wrote: > On Thu, Mar 07, 2019 at 10:45:57AM +0800, Jason Wang wrote: > > > > On 2019/3/7 ??12:31, Michael S. Tsirkin wrote: > > > > +static void vhost_set_vmap_dirty(struct vhost_vmap *used) > > > > +{ > > > > + int i; > > > > + > > > > + for (i = 0; i <

This series tries to access virtqueue metadata through kernel virtual address instead of copy_user() friends since they had too much overheads like checks, spec barriers or even hardware feature toggling. This is done through setup kernel address through vmap() and resigter MMU notifier for invalidation. Test shows about 24% improvement on TX PPS. TCP_STREAM doesn't see obvious improvement.

This series tries to access virtqueue metadata through kernel virtual address instead of copy_user() friends since they had too much overheads like checks, spec barriers or even hardware feature toggling. This is done through setup kernel address through vmap() and resigter MMU notifier for invalidation. Test shows about 24% improvement on TX PPS. TCP_STREAM doesn't see obvious improvement.