This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH - ping] Don't unlink /dev/null on darwin


"Peter O'Gorman" <peter@pogma.com> writes:
> Peter O'Gorman wrote:
>> I'll look into this then, might be hard to get the $$ through to the
>> shell cleanly though.
>
> Turned out to be fairly easy, I verified this works by putting in a
> sleep 120; and doing an ls:
> peter% ls vis_te*
> vis_temp_file25933.o
> (and later)
> peter% ls vis_te*
> tcsh: ls: No match.
> peter% ls vis_te*
> vis_temp_file26359.o
> (and later)
> peter% ls vis_te*
> tcsh: ls: No match.
>
> Patch attached, is this one okay?

Yes, this one is OK.

zw

> +	* mklibgcc.in (vis_hide): Don't use -o /dev/null on the test. On
> +	at least darwin -o /dev/null can result in as unlinking /dev/null.
> +	Use a temp file, vis_temp_file$$.o instead.

Don't provide ChangeLog entries as patches - by the time they get
around to being applied, they won't match.  Just plain text is fine.

Also, we generally don't put the rationale in the ChangeLog, I would
have written

        * mklibgcc.in (vis_hide): Use a temporary object file, not
        -o /dev/null.

zw


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]