Jens Axboe

Closing this one as we do return if there's data to be read, just not how much. The latter would be a trivial addition however.

Closing this one as it's a mix of a bunch of things, some of which are available at this point. Feel free to open a new issue, but please keep...

Yeah I don't think we should, there's no good reasons to do that.

Let me just kill this O_ATOMIC stuff first, that never ended up upstream.

> Let me just kill this O_ATOMIC stuff first, that never ended up upstream. Done

I've already explained what needs to be done to support this. If someone wants to do the work, then I'd be happy to review it. But I personally don't have...

Something like this should do it: ``` diff --git a/src/include/liburing.h b/src/include/liburing.h index 70c177431faf..28eba2e57535 100644 --- a/src/include/liburing.h +++ b/src/include/liburing.h @@ -1382,6 +1382,14 @@ IOURINGINLINE void io_uring_buf_ring_advance(struct io_uring_buf_ring *br, io_uring_smp_store_release(&br->tail, new_tail); }...

Added, with documentation update.

Not sure why it's so interesting to post on HN, honestly most of the commentary there is vitriol and not very useful. What are we trying to accomplish? For the...

Here's one from this week: https://lore.kernel.org/io-uring/[email protected]/