From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/13197 Path: news.gmane.org!.POSTED!not-for-mail From: Rich Felker Newsgroups: gmane.linux.lib.musl.general Subject: Re: [PATCH] fexecve: implement in terms of execveat when it exists Date: Tue, 4 Sep 2018 19:30:28 -0400 Message-ID: <20180904233028.GF1878@brightrain.aerifal.cx> References: <20180902171124.GW1878@brightrain.aerifal.cx> Reply-To: musl@lists.openwall.com NNTP-Posting-Host: blaine.gmane.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: blaine.gmane.org 1536103718 19669 195.159.176.226 (4 Sep 2018 23:28:38 GMT) X-Complaints-To: usenet@blaine.gmane.org NNTP-Posting-Date: Tue, 4 Sep 2018 23:28:38 +0000 (UTC) User-Agent: Mutt/1.5.21 (2010-09-15) To: musl@lists.openwall.com Original-X-From: musl-return-13213-gllmg-musl=m.gmane.org@lists.openwall.com Wed Sep 05 01:28:34 2018 Return-path: Envelope-to: gllmg-musl@m.gmane.org Original-Received: from mother.openwall.net ([195.42.179.200]) by blaine.gmane.org with smtp (Exim 4.84_2) (envelope-from ) id 1fxKkM-00051b-LQ for gllmg-musl@m.gmane.org; Wed, 05 Sep 2018 01:28:34 +0200 Original-Received: (qmail 21634 invoked by uid 550); 4 Sep 2018 23:30:42 -0000 Mailing-List: contact musl-help@lists.openwall.com; run by ezmlm Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-ID: Original-Received: (qmail 21612 invoked from network); 4 Sep 2018 23:30:41 -0000 Content-Disposition: inline In-Reply-To: Original-Sender: Rich Felker Xref: news.gmane.org gmane.linux.lib.musl.general:13197 Archived-At: On Sun, Sep 02, 2018 at 01:42:26PM -0400, Joseph C. Sible wrote: > On Sun, Sep 2, 2018 at 1:11 PM Rich Felker wrote: > > > > On Sun, Sep 02, 2018 at 12:04:55AM -0400, Joseph Sible wrote: > > > This lets fexecve work even when /proc isn't mounted. > > > --- > > > src/process/fexecve.c | 7 +++++++ > > > 1 file changed, 7 insertions(+) > > > > > > diff --git a/src/process/fexecve.c b/src/process/fexecve.c > > > index 6507b42..905487e 100644 > > > --- a/src/process/fexecve.c > > > +++ b/src/process/fexecve.c > > > @@ -1,13 +1,20 @@ > > > +#define _GNU_SOURCE > > > #include > > > #include > > > +#include > > > +#include "syscall.h" > > > > > > void __procfdname(char *, unsigned); > > > > > > int fexecve(int fd, char *const argv[], char *const envp[]) > > > { > > > +#ifdef SYS_execveat > > > + return syscall(SYS_execveat, fd, "", argv, envp, AT_EMPTY_PATH); > > > +#else > > > char buf[15 + 3*sizeof(int)]; > > > __procfdname(buf, fd); > > > execve(buf, argv, envp); > > > if (errno == ENOENT) errno = EBADF; > > > return -1; > > > +#endif > > > } > > > -- > > > 2.7.4 > > > > This breaks programs running on any kernel older than 3.19. > > > > Instead it needs to be something like > > > > int r = __syscall(SYS_execveat, fd, "", argv, envp, AT_EMPTY_PATH); > > if (r!=-ENOSYS) return __syscall_ret(r); > > ... > > > > with no #ifdef. #ifdef SYS_anything is only valid in musl when the > > existence of the syscall is arch-specific. The defines come from musl > > itself, so trying to use it for something version-specific does not > > make sense; it would be unconditionally true or false. > > > > Rich > > Good catch. Updated patch below. > -- >8 -- > Subject: [PATCH v2] fexecve: implement in terms of execveat when it exists > > This lets fexecve work even when /proc isn't mounted. > --- > src/process/fexecve.c | 5 +++++ > 1 file changed, 5 insertions(+) > > diff --git a/src/process/fexecve.c b/src/process/fexecve.c > index 6507b42..5cac05e 100644 > --- a/src/process/fexecve.c > +++ b/src/process/fexecve.c > @@ -1,10 +1,15 @@ > +#define _GNU_SOURCE > #include > #include > +#include > +#include "syscall.h" > > void __procfdname(char *, unsigned); > > int fexecve(int fd, char *const argv[], char *const envp[]) > { > + int r = __syscall(SYS_execveat, fd, "", argv, envp, AT_EMPTY_PATH); > + if(r != -ENOSYS) return __syscall_ret(r); > char buf[15 + 3*sizeof(int)]; > __procfdname(buf, fd); > execve(buf, argv, envp); > -- > 2.7.4 Thanks, merging. In the future please send patches as attachments (preferably type text/plain if it's easy to pick) unless you're sure they'll apply correctly with git-am from the body. Rich