From: KOSAKI Motohiro <kosaki.motohiro@gmail.com>
To: "H. Peter Anvin" <hpa@zytor.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@gmail.com>,
Alexey Dobriyan <adobriyan@gmail.com>,
akpm@linux-foundation.org, viro@zeniv.linux.org.uk,
torvalds@linux-foundation.org, drepper@gmail.com,
linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org
Subject: Re: [PATCH] nextfd(2)
Date: Wed, 04 Apr 2012 09:31:58 -0700 [thread overview]
Message-ID: <4F7C777E.10608@gmail.com> (raw)
In-Reply-To: <4F7A3CC2.1040200@zytor.com>
(4/2/12 4:56 PM), H. Peter Anvin wrote:
> On 04/02/2012 04:17 PM, KOSAKI Motohiro wrote:
>>
>> Sorry for the long delay comment. I realized this thread now. I think
>> /proc no mount case is not good explanation for the worth of this patch. The problem
>> is, we can't use opendir() after fork() if an app has multi threads.
>>
>> SUS clearly say so,
>> http://pubs.opengroup.org/onlinepubs/009695399/functions/fork.html
>>
>> we can only call async-signal-safe functions after fork() when multi threads and
>> opendir() call malloc() internally.
>>
>> As far as I know, OpenJDK has a such fork-readdir-exec code and it can
>> make deadlock
>> when spawnning a new process. Unfortunately Java language perfeter to
>> make a lot of threads rather than other language.
>>
>> This patch can solve such multi threaded case.
>>
>> offtopic, glibc malloc is a slightly clever. It reinitialize its
>> internal lock when fork by using thread_atfork() hook. It mean glibc malloc can be used after
>> fork() and the technique can avoid this issue. But, glibc malloc still has several
>> performance problem and many people prefer to use jemalloc or google malloc instead. Then,
>> they hit an old issue, bah.
>>
>
> OK, so what you're saying here is:
>
> Linux doesn't actually have a problem unless:
> 1. You use the library implementation of opendir/readdir/closedir;
> 2. You use a nonstandard malloc for the platform which doesn't
> correctly set up fork hooks (which I would consider a bug);
Right. but I'm argue "correctly set up" term because SUS/POSIX don't require it.
It is only a workaround of buggy userland in glibc. SUS still says you can't
use opendir and typical userland people don't want ignore SUS as far as possible.
> You can deal with this in one of two ways:
>
> 2. Fix your malloc().
> 1. Use the low level open()/getdents()/close() functions instead of
> opendir()/readdir()/closedir().
Ideally possible. but practically impossible. 2) people don't use a their
own malloc. they only uses open sources alternative malloc. And, I think
you have too narrowing concern. Even though malloc people adds a workaround,
the standard inhibit to use it and people may continue to use more dangerous
RLIM_NOFILE loop. 1) I haven't seen _practical_ userland software uses such
linux internal hacking. Almost all major software can run on multiple OSs.
>> and I've received a request that linux aim fdwalk() several times. Example,
>
> It doesn't sound very hard to implement fdwalk() in terms of
> open/getdents/close without using malloc; since the fdwalk() interface
> lets you use the stack for storage. You can then implement closefrom()
> in terms of fdwalk(). Something like this (untested):
>
> int fdwalk(int (*func)(void *, int), void *cd)
> {
> char buf[4096]; /* ... could be less... */
> const char *p, *q;
> const struct linux_dirent *dp
> int dfd, fd;
> unsigned char c;
> int rv = 0;
> int sz;
>
> dfd = open("/proc/self/fd", O_RDONLY|O_DIRECTORY|O_CLOEXEC);
> if (dfd< 0)
> return -1;
>
> /*** XXX: may want to check for procfs magic here ***/
>
> while ((sz = getdents(dfd, buf, sizeof buf))> 0) {
> p = buf;
>
> while (sz> offsetof(struct linux_dirent, d_name)) {
> dp = (const struct linux_dirent *)p;
>
> if (sz< dp->d_reclen)
> break;
>
> q = dp->d_name;
> p += dp->d_reclen;
> sz -= dp->d_reclen;
>
> fd = 0;
> while (q< p&& (c = *q++)) {
> c -= '0';
> if (c>= 10)
> goto skip;
> fd = fd*10 + c;
> }
>
> if (fd != dfd)
> rv = func(cd, fd);
> skip:
> ;
> }
> }
>
> if (close(dfd))
> return -1;
>
> return rv;
> }
It can. but more ugly. no?
next prev parent reply other threads:[~2012-04-04 16:31 UTC|newest]
Thread overview: 70+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-04-01 12:57 [PATCH] nextfd(2) Alexey Dobriyan
2012-04-01 13:58 ` Konstantin Khlebnikov
2012-04-01 21:30 ` Alexey Dobriyan
2012-04-02 0:09 ` Alan Cox
2012-04-02 8:38 ` Konstantin Khlebnikov
2012-04-02 9:26 ` Cyrill Gorcunov
2012-04-01 15:43 ` Eric Dumazet
2012-04-01 21:31 ` Alexey Dobriyan
2012-04-01 21:36 ` Alan Cox
2012-04-01 17:20 ` Linus Torvalds
2012-04-01 18:28 ` Valentin Nechayev
2012-04-01 21:33 ` Alexey Dobriyan
2012-04-01 19:21 ` Arnd Bergmann
2012-04-01 21:35 ` Alexey Dobriyan
2012-04-01 22:05 ` H. Peter Anvin
2012-04-04 12:13 ` Arnd Bergmann
2012-04-01 22:03 ` H. Peter Anvin
2012-04-01 22:13 ` H. Peter Anvin
2012-04-02 0:08 ` Alan Cox
2012-04-30 9:58 ` Valentin Nechayev
2012-04-02 1:19 ` Kyle Moffett
2012-04-02 1:37 ` H. Peter Anvin
2012-04-02 11:37 ` Ulrich Drepper
2012-04-06 9:54 ` Alexey Dobriyan
2012-04-06 15:27 ` Colin Walters
2012-04-06 16:14 ` H. Peter Anvin
2012-04-06 20:16 ` Alexey Dobriyan
2012-04-06 20:33 ` H. Peter Anvin
2012-04-06 21:02 ` H. Peter Anvin
2012-04-12 10:54 ` Alexey Dobriyan
2012-04-12 11:11 ` Alan Cox
2012-04-12 13:35 ` Alexey Dobriyan
2012-04-12 13:51 ` H. Peter Anvin
2012-04-12 19:21 ` Alexey Dobriyan
2012-04-12 14:09 ` Eric Dumazet
2012-04-06 16:23 ` H. Peter Anvin
2012-04-07 21:21 ` Ben Pfaff
2012-04-11 0:12 ` KOSAKI Motohiro
2012-04-11 0:09 ` KOSAKI Motohiro
2012-04-11 17:58 ` H. Peter Anvin
2012-04-11 18:04 ` Linus Torvalds
2012-04-11 18:11 ` H. Peter Anvin
2012-04-11 19:46 ` KOSAKI Motohiro
2012-04-11 19:49 ` H. Peter Anvin
2012-04-11 20:23 ` KOSAKI Motohiro
2012-04-11 20:32 ` H. Peter Anvin
2012-04-17 18:12 ` KOSAKI Motohiro
2012-04-11 18:00 ` H. Peter Anvin
2012-04-11 19:20 ` KOSAKI Motohiro
2012-04-11 19:22 ` H. Peter Anvin
2012-04-11 19:26 ` KOSAKI Motohiro
2012-04-11 19:28 ` H. Peter Anvin
2012-04-11 19:31 ` KOSAKI Motohiro
2012-04-11 19:32 ` H. Peter Anvin
2012-04-02 23:17 ` KOSAKI Motohiro
2012-04-02 23:56 ` H. Peter Anvin
2012-04-04 11:51 ` Ulrich Drepper
2012-04-04 16:38 ` KOSAKI Motohiro
2012-04-04 16:43 ` Ulrich Drepper
2012-04-04 17:07 ` KOSAKI Motohiro
2012-04-04 17:49 ` Ulrich Drepper
2012-04-04 18:08 ` KOSAKI Motohiro
2012-04-04 16:31 ` KOSAKI Motohiro [this message]
2012-04-04 17:10 ` Colin Walters
2012-04-04 17:25 ` Colin Walters
2012-04-04 23:35 ` KOSAKI Motohiro
2012-04-04 18:44 ` H. Peter Anvin
2012-04-03 19:21 ` Colin Walters
2012-04-04 3:01 ` Al Viro
2012-04-04 17:10 ` KOSAKI Motohiro
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=4F7C777E.10608@gmail.com \
--to=kosaki.motohiro@gmail.com \
--cc=adobriyan@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=drepper@gmail.com \
--cc=hpa@zytor.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--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).