On 2/26/25 11:20 AM, Keith Busch wrote: > From: Keith Busch <kbusch@xxxxxxxxxx> > > Provide an interface for the kernel to leverage the existing > pre-registered buffers that io_uring provides. User space can reference > these later to achieve zero-copy IO. > > User space must register an empty fixed buffer table with io_uring in > order for the kernel to make use of it. Just a suggestion, but might make sense to not use ->release() as a gating whether this is a kernel buffer or not, there's room in the struct anyway with the 'u8 perm' having holes anyway. And if we did that, then we could just have a default release that does the unpin and put rather than needing to check and have branches for the two types of release. Yes the indirect function call isn't free either, like the branches aren't, but I don't think it matters on the release side. At least not enough to care, and it'd help streamline the code a bit and not overload ->release() with meaning "oh this is a kernel buffer". -- Jens Axboe
