[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

bug#21694: 'clone' syscall binding unreliable

From: Thompson, David
Subject: bug#21694: 'clone' syscall binding unreliable
Date: Fri, 16 Oct 2015 19:12:53 -0400

On Fri, Oct 16, 2015 at 4:39 PM, Ludovic Courtès <address@hidden> wrote:
> I’m reporting the problem and (hopefully) the solution, but I think we’d
> better double-check this.
> The problem: Running the test below in a loop sometimes gets a SIGSEGV
> in the child process (on x86_64, libc 2.22.)
> --8<---------------cut here---------------start------------->8---
> (use-modules (guix build syscalls) (ice-9 match))
> (match (clone (logior CLONE_NEWUSER
>                       CLONE_CHILD_SETTID
>                       CLONE_CHILD_CLEARTID
>                       SIGCHLD))
>   (0
>    (throw 'x))                                    ;XXX: sometimes segfaults
>   (pid
>    (match (waitpid pid)
>      ((_ . status)
>       (pk 'status status)
>       (exit (not (status:term-sig status)))))))
> --8<---------------cut here---------------end--------------->8---
> Looking at (guix build syscalls) though, I see an ABI mismatch between
> our definition and the actual ‘syscall’ C function, and between our
> ‘clone’ definition and the actual C function.
> This leads to the attached patch, which also fixes the above problem for me.
> Could you test this patch?

The patch looks good.  Thanks for catching this!

> Now, there remains the question of CLONE_CHILD_SETTID and
> CLONE_CHILD_CLEARTID.  Since we’re passing NULL for ‘ctid’, I expect
> that these flags have no effect at all.

I added those flags in commit ee78d02 because they solved a real issue
I ran into.  Adding those flags made 'clone' look like a
'primitive-fork' call when examined with strace.

> Conversely, libc uses these flags to update the thread ID in the child
> process (x86_64/arch-fork.h):
> --8<---------------cut here---------------start------------->8---
> #define ARCH_FORK() \
>   INLINE_SYSCALL (clone, 4,                                                   
> \
> \
>                   NULL, &THREAD_SELF->tid)
> --8<---------------cut here---------------end--------------->8---
> This is certainly useful, but we’d have troubles doing it from the FFI…
> It may that this is fine if the process doesn’t use threads.

Right, so here's what 'primitive-fork' does:

child_tidptr=0x7fc5398cea10) = 13247

Here's what 'clone' does:

= 14038

In practice it may not be a problem since most of the time you'd
'exec' after cloning.  Is there any reliable way to get a hold of
whatever THREAD_SELF is?  I wish the libc 'clone' function didn't have
that silly callback and behaved like 'fork', then we could have
avoided these issues altogether.

- Dave

reply via email to

[Prev in Thread] Current Thread [Next in Thread]