Re: [PATCH 2/8] aio_poll_wake(): don't set ->woken if we ignore the wakeup

So, what should we do? On 2019/3/7 10:18, Al Viro wrote: > On Thu, Mar 07, 2019 at 12:03:10AM +, Al Viro wrote: >> From: Al Viro >> >> In case of early wakeups, aio_poll() assumes that aio_poll_complete() >> has either already happened or is imminent. In that case we do not >> want to put io

Re: [PATCH 3/8] aio_poll(): sanitize the logics after vfs_poll(), get rid of leak on error

+ if (async && !apt.error) --->may be this should be if (!async && !apt.error) ? On 2019/3/7 8:03, Al Viro wrote: > From: Al Viro > > We want iocb_put() happening on errors, to balance the extra reference > we'd taken. As it is, we end up with a leak. The rules should be > * erro