Jens Axboe

There aren't really any post-prep side -EINVAL possible for a linked timeout, hence why I was suspecting it's something that the prep side doesn't like (like, for example, an invalid...

It will not block, io_uring is actually async, aio is just not.

```RQF_NOWAIT``` _is_ the mechanism. If something is blocking on other IO off the submission path, then that piece is buggy. In this case that looks like XFS is. I can...

XFS needs something like the below. Not really tested, may be daemons lurking... But it's certainly losing the NOWAIT flag, which is the bug here. ``` diff --git a/fs/iomap/direct-io.c b/fs/iomap/direct-io.c...

> This is 6.5, but I believe the behavior is the same all the back to 5.x and forward to the tip. Could very well be. > Agreed, though fs/iomap/direct-io.c...

Don't think this is a liburing issue, none of those two headers are liburing.

I realize it comes up when compiling that example, just saying that it isn't doing anything special or including odd headers. Does it compile if you remove linux/ipv6.h from that...

Yep agree, we do need a specific target that just builds the library parts, not examples or tests. Something ala: ``` diff --git a/Makefile b/Makefile index d6d86355950e..e57a65be17c2 100644 --- a/Makefile...

Please include some actual details on what you are doing, what devices are involved, what kernel, etc. As it stands, this question is rather useless.

Please try a more recent and supported kernel, the ancient ones don't have proper -EAGAIN handling on the storage side (which I'm guessing is what you are doing, you didn't...