Re: [PATCH] aio_run_iocb should always retry

> Note that there's an existing problem with failed cancellations in the
> code that should be fixed (and I'm sure you'll clean it up).  Patch
> attached (and soon to become irrelevant, I think).

Yeah, thanks for bringing this up.  I agree that sys_io_cancel()  
shouldn't have had an effect if it returns an error.  Though this  
patch is still racey in today's implementation, it'll certainly be  
the right thing once we wrap some serialization around it.

Along those lines, I don't think we should lose events if the  
copy_to_user() fails.  I think I'll pin the pages containing the  
event result before doing any work so we have a chance to return  
errors if the pointer is bogus.

- z

--
To unsubscribe, send a message with 'unsubscribe linux-aio' in
the body to [email protected].  For more info on Linux AIO,
see: http://www.kvack.org/aio/
Don't email: <a href=mailto:"[email protected]">[email protected]</a>