Date: Sun, 2 Jun 2013 23:22:09 +0200 From: Jilles Tjoelker <jilles@stack.nl> To: Pawel Jakub Dawidek <pjd@FreeBSD.org> Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r251197 - head/lib/libc/sys Message-ID: <20130602212209.GA44549@stack.nl> In-Reply-To: <20130531233109.GD1387@garage.freebsd.pl> References: <201305312131.r4VLVcd5074132@svn.freebsd.org> <20130531233109.GD1387@garage.freebsd.pl>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jun 01, 2013 at 01:31:10AM +0200, Pawel Jakub Dawidek wrote: > On Fri, May 31, 2013 at 09:31:38PM +0000, Jilles Tjoelker wrote: > > Author: jilles > > Date: Fri May 31 21:31:38 2013 > > New Revision: 251197 > > URL: http://svnweb.freebsd.org/changeset/base/251197 > > Log: > > sigaction(2): *at system calls are async-signal safe. > Isn't chflags(2) family safe as well? There are many more non-POSIX functions that are async-signal safe. For example: * bindat()/connectat() (like bind()/connect()) * eaccess() (like faccessat()) * pipe2() (like pipe()) * accept4() (like accept()) * wait4() (like waitpid()) * futimesat() (like utimes()) * strsep() (like strtok_r()) We may not want to document them all as async-signal safe because it may be necessary later to implement some as async-signal unsafe library functions. However, it would be useful to document some of them as async-signal safe. -- Jilles Tjoelker
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20130602212209.GA44549>