Eric Blake wrote:
> we also need to
> tweak the O_NOINHERIT flag of an existing fd without having to use dup in
> cloexec.c (I'm hoping the windows functions GetHandleInformation and
> SetHandleInformation work here).
I don't see how to make it work without a lot of dup() calls and without
assumin
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to Bruno Haible on 8/20/2009 2:14 AM:
> This is becoming confusing: Why should the 'fopen' test suddenly test
> fopen_safer,
> not fopen??
>
> I would find it better to introduce a new test module fopen-safer-tests, that
> shares
> the bul
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to Bruno Haible on 8/20/2009 2:14 AM:
> Hi Eric,
>
>> I went ahead and implemented popen_safer.
>
> Well done!
>
>> +/* Get the original definition of popen. It might be defined as a macro.
>> */
>
> This is not needed. I'm not aware o
Hi Eric,
> I went ahead and implemented popen_safer.
Well done!
> +/* Get the original definition of popen. It might be defined as a macro. */
This is not needed. I'm not aware of any platform that defines popen as a
macro:
$ LC_ALL=C fgrep popen */usr/include/stdio.h
aix-4.3.2/usr/include/st
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to Eric Blake on 8/19/2009 11:06 AM:
> I went ahead and implemented popen_safer.
And this patch also includes a first use of O_CLOEXEC; I'm debating about
adding more O_CLOEXEC support throughout gnulib, a piece at a time.
cloexec support ha
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to Eric Blake on 8/17/2009 5:14 PM:
> Even though Bruno's pipe module is generally a better
> interface than popen, I discovered yet another popen
> bug in cygwin today[1] (I fixed a cygwin popen bug in
> August 2006, only to have it regress
Even though Bruno's pipe module is generally a better
interface than popen, I discovered yet another popen
bug in cygwin today[1] (I fixed a cygwin popen bug in
August 2006, only to have it regress in December of
that year when cgf rewrote the implementation to be
faster). So even though m4 no lo
