public inbox for libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Florian Weimer <fweimer@redhat.com>
To: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Cc: libc-alpha@sourceware.org
Subject: Re: [PATCH 1/3] linux: Move flockfile/_IO_flockfile into libc
Date: Thu, 06 May 2021 22:34:41 +0200	[thread overview]
Message-ID: <87k0obd2ji.fsf@oldenburg.str.redhat.com> (raw)
In-Reply-To: <20210506200614.2928007-1-adhemerval.zanella@linaro.org> (Adhemerval Zanella's message of "Thu, 6 May 2021 17:06:12 -0300")

* Adhemerval Zanella:

> diff --git a/stdio-common/flockfile.c b/stdio-common/flockfile.c
> index 0bc856a465..c838dcd538 100644
> --- a/stdio-common/flockfile.c
> +++ b/stdio-common/flockfile.c
> @@ -17,13 +17,18 @@
>     <https://www.gnu.org/licenses/>.  */
>  
>  #include <stdio.h>
> +#include <stdio-lock.h>
> +#include <sys/single_threaded.h>
>  
>  void
>  __flockfile (FILE *stream)
>  {
> +  bool multithread = __libc_single_threaded == 0;
> +  if (multithread)
> +    {
> +      stream->_flags2 |= _IO_FLAGS2_NEED_LOCK;
> +      _IO_lock_lock (*stream->_lock);
> +    }
>  }

I'm afraid this is not correct because a thread can become multithreaded
between the flockfile/funlockfile calls.  As a result, the lock state
would be wrong at funlockfile time.

In order to optimize locks in such cases, it is necessary to perform the
same actions as the locking/unlocking operations, but without atomic
instructions.

The stream->_flags2 update has a data race with other _flags2 updates
(e.g., in __malloc_stats), but that's a preexisting bug that should be
fixed separately (there might be other affected places).

Thanks,
Florian


  parent reply	other threads:[~2021-05-06 20:34 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-06 20:06 Adhemerval Zanella
2021-05-06 20:06 ` [PATCH 2/3] linux: Move ftrylockfile/_IO_ftrylockfile " Adhemerval Zanella
2021-05-06 20:06 ` [PATCH 3/3] linux: Move funlockfile/_IO_funlockfile " Adhemerval Zanella
2021-05-06 20:34 ` Florian Weimer [this message]
2021-05-06 21:01   ` [PATCH 1/3] linux: Move flockfile/_IO_flockfile " Adhemerval Zanella
2021-05-07  6:37     ` Florian Weimer

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=87k0obd2ji.fsf@oldenburg.str.redhat.com \
    --to=fweimer@redhat.com \
    --cc=adhemerval.zanella@linaro.org \
    --cc=libc-alpha@sourceware.org \
    /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 read-only IMAP folder(s) and NNTP newsgroup(s).