On 06/15/2015 05:58 AM, Zeng Linggang wrote: > Hi, > > On Fri, 2015-06-12 at 21:01 +0200, Michael Kerrisk (man-pages) wrote: >> Hi Zeng Linggang, >> >> On 05/27/2015 05:42 PM, Samuel Bronson wrote: >>> On Wed, May 27, 2015 at 6:12 AM, Zeng Linggang <zenglg.jy@xxxxxxxxxxxxxx> wrote: >>>> After research, We think fexecve() is thread-safe. But, there >>>> is not marking of fexecve() in glibc document. >>> >>> Um, fexecve() is just like execve(), only on a file descriptor. (In >>> fact, it appears to be implemented using execve() on >>> /proc/self/fd/$FD.) >>> >>> As such, the following point from execve(2) applies: >>> >>> * All threads other than the calling thread are destroyed during an >>> execve(). Mutexes, condition variables, and other pthreads objects >>> are not preserved. >>> >>> This does not seem to correspond with the usual notion of "thread >>> safe", though I guess it would be technically safe to, say, have >>> several threads racing to run fexecve(). Further text is probably >>> warranted here. >> >> Did you have any thoughts on Samuel's point? > > Thanks for your remind, I almost miss it. > > As we can see, the detail information has been list in 'execve'. I think > it is not necessary to add further text. As long as you think it's okay, I'll go with that. I've applied your patch. Cheers, Michael -- Michael Kerrisk Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/ Linux/UNIX System Programming Training: http://man7.org/training/ -- To unsubscribe from this list: send the line "unsubscribe linux-man" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
