On 10/3/22 1:40 PM, Jens Axboe wrote: > Hi Linus, > > On top of the block and io_uring branches, here are a set of updates for > the passthrough support that was merged in the 6.0 kernel. With these > changes, passthrough NVMe support over io_uring now performs at the same > level as block device O_DIRECT, and in many cases 6-8% better. This pull > request contains: > > - Add support for fixed buffers for passthrough (Anuj, Kanchan) > > - Enable batched allocations and freeing on passthrough, similarly to > what we support on the normal storage path (me) > > Please pull! Geert noticed that there was an issue if io_uring wasn't configured, there's been a patch added for that. I'll send a v2 of this pull request. -- Jens Axboe
