From: David Laight <David.Laight@ACULAB.COM>
To: 'Oleg Nesterov' <oleg@redhat.com>,
Deepa Dinamani <deepa.kernel@gmail.com>,
Al Viro <viro@ZenIV.linux.org.uk>,
"Eric W. Biederman" <ebiederm@xmission.com>,
Linus Torvalds <torvalds@linux-foundation.org>
Cc: "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"akpm@linux-foundation.org" <akpm@linux-foundation.org>,
"arnd@arndb.de" <arnd@arndb.de>,
"dbueso@suse.de" <dbueso@suse.de>,
"axboe@kernel.dk" <axboe@kernel.dk>,
"dave@stgolabs.net" <dave@stgolabs.net>,
"e@80x24.org" <e@80x24.org>,
"jbaron@akamai.com" <jbaron@akamai.com>,
"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
"linux-aio@kvack.org" <linux-aio@kvack.org>,
"omar.kilani@gmail.com" <omar.kilani@gmail.com>,
"tglx@linutronix.de" <tglx@linutronix.de>,
"stable@vger.kernel.org" <stable@vger.kernel.org>
Subject: RE: pselect/etc semantics (Was: [PATCH v2] signal: Adjust error codes according to restore_user_sigmask())
Date: Wed, 29 May 2019 16:54:40 +0000 [thread overview]
Message-ID: <b05cec7f9e8f457281e689576a7a360f@AcuMS.aculab.com> (raw)
In-Reply-To: <20190529161157.GA27659@redhat.com>
From: Oleg Nesterov
> Sent: 29 May 2019 17:12
> Al, Linus, Eric, please help.
>
> The previous discussion was very confusing, we simply can not understand each
> other.
>
> To me everything looks very simple and clear, but perhaps I missed something
> obvious? Please correct me.
>
> I think that the following code is correct
>
> int interrupted = 0;
>
> void sigint_handler(int sig)
> {
> interrupted = 1;
> }
>
> int main(void)
> {
> sigset_t sigint, empty;
>
> sigemptyset(&sigint);
> sigaddset(&sigint, SIGINT);
> sigprocmask(SIG_BLOCK, &sigint, NULL);
>
> signal(SIGINT, sigint_handler);
>
> sigemptyset(&empty); // so pselect() unblocks SIGINT
>
> ret = pselect(..., &empty);
^^^^^ sigint
>
> if (ret >= 0) // sucess or timeout
> assert(!interrupted);
>
> if (interrupted)
> assert(ret == -EINTR);
> }
>
> IOW, if pselect(sigmask) temporary unblocks SIGINT according to sigmask, this
> signal should not be delivered if a ready fd was found or timeout. The signal
> handle should only run if ret == -EINTR.
Personally I think that is wrong.
Given code like the above that has:
while (!interrupted) {
pselect(..., &sigint);
// process available data
}
You want the signal handler to be executed even if one of the fds
always has available data.
Otherwise you can't interrupt a process that is always busy.
One option is to return -EINTR if a signal is pending when the mask
is updated - before even looking at anything else.
Signals that happen later on (eg after a timeout) need not be reported
(until the next time around the loop).
> (pselect() can be interrupted by any other signal which has a handler. In this
> case the handler can be called even if ret >= 0. This is correct, I fail to
> understand why some people think this is wrong, and in any case we simply can't
> avoid this).
You mean any signal that isn't blocked when pselect() is called....
> This was true until 854a6ed56839a ("signal: Add restore_user_sigmask()"),
> now this is broken by the signal_pending() check in restore_user_sigmask().
>
> This patch https://lore.kernel.org/lkml/20190522032144.10995-1-deepa.kernel@gmail.com/
> turns 0 into -EINTR if signal_pending(), but I think we should simply restore
> the old behaviour and simplify the code.
>
> See the compile-tested patch at the end. Of course, the new _xxx() helpers
> should be renamed somehow. fs/aio.c doesn't look right with or without this
> patch, but iiuc this is what it did before 854a6ed56839a.
>
> Let me show the code with the patch applied. I am using epoll_pwait() as an
> example because it looks very simple.
>
>
> static inline void set_restore_sigmask(void)
> {
> // WARN_ON(!TIF_SIGPENDING) was removed by this patch
> current->restore_sigmask = true;
> }
>
> int set_xxx(const sigset_t __user *umask, size_t sigsetsize)
> {
> sigset_t *kmask;
^ no '*' here, add & before uses.
>
> if (!umask)
> return 0;
> if (sigsetsize != sizeof(sigset_t))
> return -EINVAL;
> if (copy_from_user(kmask, umask, sizeof(sigset_t)))
> return -EFAULT;
>
> // we can safely modify ->saved_sigmask/restore_sigmask, they has no meaning
> // until the syscall returns.
> set_restore_sigmask();
> current->saved_sigmask = current->blocked;
> set_current_blocked(kmask);
>
> return 0;
> }
>
>
> void update_xxx(bool interrupted)
> {
> // the main reason for this helper is WARN_ON(!TIF_SIGPENDING) which was "moved"
> // from set_restore_sigmask() above.
> if (interrupted)
> WARN_ON(!test_thread_flag(TIF_SIGPENDING));
> else
> restore_saved_sigmask();
> }
I looked at the code earlier, but failed to find the code that actually
delivers the signals.
It may be 'racy' with update_xxx() regardless of whether that is
looking for -EINTR or just a pending signal.
I assume that TIF_SIGPENGING is used to (not) short-circuit the
system call return path so that signals get delivered.
So that it is important that update_xxx() calls restore_saved_sigmask()
if there is no signal pending.
(Although a signal can happen after the test - which can/will be ignored
until the signal is enabled again.)
restore_saved_sigmask() must itself be able to set TIF_SIGPENDING
(the inner sigmask could be more restrictive!).
If restore_saved_sigmask() isn't called here, the syscall return
path must do it after calling all the handlers and after clearing
TIF_SIGPENDING, and then call unmasked handlers again.
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)
next prev parent reply other threads:[~2019-05-29 16:54 UTC|newest]
Thread overview: 95+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-05-22 3:21 [PATCH v2] signal: Adjust error codes according to restore_user_sigmask() Deepa Dinamani
2019-05-22 15:05 ` Oleg Nesterov
2019-05-22 15:55 ` Deepa Dinamani
2019-05-22 16:14 ` Oleg Nesterov
2019-05-22 16:33 ` Deepa Dinamani
2019-05-23 9:03 ` David Laight
2019-05-23 14:59 ` Oleg Nesterov
2019-05-23 16:18 ` David Laight
2019-05-23 16:36 ` Oleg Nesterov
2019-05-23 16:56 ` David Laight
2019-05-23 18:06 ` Deepa Dinamani
2019-05-23 20:41 ` Deepa Dinamani
2019-05-23 21:06 ` Deepa Dinamani
2019-05-24 9:58 ` David Laight
2019-05-24 14:10 ` Oleg Nesterov
2019-05-24 15:16 ` Deepa Dinamani
2019-05-24 16:33 ` Oleg Nesterov
2019-05-24 17:01 ` Deepa Dinamani
2019-05-27 15:04 ` Oleg Nesterov
2019-05-28 20:47 ` Deepa Dinamani
2019-05-29 16:57 ` Oleg Nesterov
2019-05-29 18:42 ` Deepa Dinamani
2019-05-28 9:02 ` David Laight
2019-05-28 9:12 ` David Laight
2019-05-28 11:37 ` Deepa Dinamani
2019-05-28 12:04 ` David Laight
2019-05-24 14:19 ` Oleg Nesterov
2019-05-24 14:29 ` Deepa Dinamani
2019-05-24 14:51 ` Oleg Nesterov
2019-05-24 13:29 ` Oleg Nesterov
2019-05-24 14:59 ` David Laight
2019-05-24 15:09 ` David Laight
2019-05-24 15:46 ` Oleg Nesterov
2019-05-24 15:44 ` Oleg Nesterov
2019-05-24 16:40 ` David Laight
2019-05-23 14:33 ` Oleg Nesterov
2019-05-22 22:18 ` Chris Down
2019-05-22 22:52 ` Deepa Dinamani
2019-05-29 16:11 ` pselect/etc semantics (Was: [PATCH v2] signal: Adjust error codes according to restore_user_sigmask()) Oleg Nesterov
2019-05-29 16:54 ` David Laight [this message]
2019-05-29 18:50 ` Eric Wong
2019-05-30 9:34 ` David Laight
2019-05-30 13:04 ` pselect/etc semantics Eric W. Biederman
2019-05-29 16:56 ` pselect/etc semantics (Was: [PATCH v2] signal: Adjust error codes according to restore_user_sigmask()) Deepa Dinamani
2019-05-29 18:26 ` Deepa Dinamani
2019-05-29 22:32 ` Arnd Bergmann
2019-05-30 1:54 ` pselect/etc semantics Eric W. Biederman
2019-05-30 18:28 ` Arnd Bergmann
2019-05-30 14:40 ` pselect/etc semantics (Was: [PATCH v2] signal: Adjust error codes according to restore_user_sigmask()) Oleg Nesterov
2019-05-30 18:37 ` Arnd Bergmann
2019-05-30 13:01 ` pselect/etc semantics Eric W. Biederman
2019-05-30 15:18 ` David Laight
2019-05-30 16:13 ` Oleg Nesterov
2019-05-30 15:38 ` Eric W. Biederman
2019-05-30 15:48 ` Deepa Dinamani
2019-05-30 16:59 ` Deepa Dinamani
2019-05-30 16:08 ` Oleg Nesterov
2019-05-30 17:20 ` Eric W. Biederman
2019-05-30 16:22 ` David Laight
2019-05-30 15:57 ` Oleg Nesterov
2019-05-30 21:03 ` Eric Wong
2019-06-04 13:41 ` [PATCH] signal: remove the wrong signal_pending() check in restore_user_sigmask() Oleg Nesterov
2019-06-04 15:31 ` Eric W. Biederman
2019-06-04 15:57 ` David Laight
2019-06-04 16:37 ` Arnd Bergmann
2019-06-04 18:14 ` Deepa Dinamani
2019-06-04 18:35 ` Eric Wong
2019-06-04 21:26 ` Linus Torvalds
2019-06-04 22:24 ` Eric Wong
2019-06-04 23:51 ` Eric W. Biederman
2019-06-05 9:04 ` Oleg Nesterov
2019-06-05 8:56 ` Oleg Nesterov
2019-06-05 9:02 ` David Laight
2019-06-05 9:25 ` Oleg Nesterov
2019-06-05 9:58 ` David Laight
2019-06-05 15:58 ` [PATCH -mm 0/1] signal: simplify set_user_sigmask/restore_user_sigmask Oleg Nesterov
2019-06-05 15:58 ` [PATCH -mm 1/1] " Oleg Nesterov
2019-06-06 0:14 ` kbuild test robot
2019-06-06 1:06 ` kbuild test robot
2019-06-06 7:25 ` Oleg Nesterov
2019-06-06 7:30 ` Sedat Dilek
2019-06-05 17:24 ` [PATCH -mm 0/1] " Linus Torvalds
2019-06-06 9:05 ` David Laight
2019-06-06 11:05 ` Oleg Nesterov
2019-06-06 11:29 ` David Laight
2019-06-06 12:41 ` Oleg Nesterov
2019-06-06 13:23 ` David Laight
2019-06-06 10:22 ` Oleg Nesterov
2019-06-06 11:32 ` [PATCH -mm V2 1/1] " Oleg Nesterov
2019-06-06 14:08 ` [PATCH 0/2] select: simplify the usage of restore_saved_sigmask_unless() Oleg Nesterov
2019-06-06 14:08 ` [PATCH 1/2] select: change do_poll() to return -ERESTARTNOHAND rather than -EINTR Oleg Nesterov
2019-06-07 18:05 ` Linus Torvalds
2019-06-06 14:09 ` [PATCH 2/2] select: shift restore_saved_sigmask_unless() into poll_select_copy_remaining() Oleg Nesterov
2019-06-07 21:39 ` [RFC PATCH 0/5]: Removing saved_sigmask Eric W. Biederman
2019-06-11 18:58 ` Oleg Nesterov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=b05cec7f9e8f457281e689576a7a360f@AcuMS.aculab.com \
--to=david.laight@aculab.com \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=axboe@kernel.dk \
--cc=dave@stgolabs.net \
--cc=dbueso@suse.de \
--cc=deepa.kernel@gmail.com \
--cc=e@80x24.org \
--cc=ebiederm@xmission.com \
--cc=jbaron@akamai.com \
--cc=linux-aio@kvack.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=oleg@redhat.com \
--cc=omar.kilani@gmail.com \
--cc=stable@vger.kernel.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=viro@ZenIV.linux.org.uk \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).