On Wed, 19 Oct 2016 18:00:16 +0200, Alexander Bluhm wrote:
> On Wed, Oct 19, 2016 at 09:34:16AM -0600, Theo de Raadt wrote:
> > Inside strerror_r, I'm unsure why there is a save_errno dance.
>
> That is from the time when we had national language support.
>
> ok?
OK millert@
- todd
On Wed, Oct 19, 2016 at 09:52:13AM -0600, Todd C. Miller wrote:
> On Wed, 19 Oct 2016 09:11:36 -0600, "Todd C. Miller" wrote:
>
> > Currently, syslog_r() avoids using strerror() since it is not
> > reentrant. We have a reentrant strerror_r() so let's use it.
>
> Since strerror_r() always fills i
On Wed, Oct 19, 2016 at 09:34:16AM -0600, Theo de Raadt wrote:
> Inside strerror_r, I'm unsure why there is a save_errno dance.
That is from the time when we had national language support.
ok?
bluhm
Index: lib/libc/string/strerror_r.c
On Wed, 19 Oct 2016 09:11:36 -0600, "Todd C. Miller" wrote:
> Currently, syslog_r() avoids using strerror() since it is not
> reentrant. We have a reentrant strerror_r() so let's use it.
Since strerror_r() always fills in the buffer there is no need to
check the return value. If saved_errno is
On Wed, 19 Oct 2016 09:34:16 -0600, "Theo de Raadt" wrote:
> Inside strerror_r, I'm unsure why there is a save_errno dance.
I don't see anything called by strerror_r() that would touch errno
so we could get rid of the save_errno dance and just do:
if (ret_errno)
errno = r
On Wed, 19 Oct 2016 17:29:18 +0200, Mark Kettenis wrote:
> Perhaps add a comment explicitly stating that OpenBSD's strerror_r()
> *is* reentrant?
I thought that was implicit in the name.
- todd
> > Currently, syslog_r() avoids using strerror() since it is not
> > reentrant. We have a reentrant strerror_r() so let's use it.
> >
> > OK?
>
> Perhaps add a comment explicitly stating that OpenBSD's strerror_r()
> *is* reentrant?
Also, signal-handler safe.
We've done something a little cra
> From: "Todd C. Miller"
> Date: Wed, 19 Oct 2016 09:11:36 -0600
>
> Currently, syslog_r() avoids using strerror() since it is not
> reentrant. We have a reentrant strerror_r() so let's use it.
>
> OK?
Perhaps add a comment explicitly stating that OpenBSD's strerror_r()
*is* reentrant?
> Inde
Currently, syslog_r() avoids using strerror() since it is not
reentrant. We have a reentrant strerror_r() so let's use it.
OK?
- todd
Index: lib/libc/gen/syslog_r.c
===
RCS file: /cvs/src/lib/libc/gen/syslog_r.c,v
retrieving revis
