This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [patch] Adjust AT_EXECFN on explicit loader invocation
- From: OndÅej BÃlka <neleai at seznam dot cz>
- To: Paul Pluzhnikov <ppluzhnikov at google dot com>
- Cc: Adhemerval Zanella <azanella at linux dot vnet dot ibm dot com>, libc-alpha at sourceware dot org
- Date: Mon, 18 Nov 2013 22:48:50 +0100
- Subject: Re: [patch] Adjust AT_EXECFN on explicit loader invocation
- Authentication-results: sourceware.org; auth=none
- References: <ye6qeh8lrcsk dot fsf at elbrus2 dot mtv dot corp dot google dot com> <CALoOobMreT4Ukk44cDRnwt=mDJfqtuamEMm9QHndOh-MuZJoRA at mail dot gmail dot com> <5277EA1D dot 9080305 at linux dot vnet dot ibm dot com> <CALoOobNcSjUKMs4GBsYm7sr5D++JZrAv6ZpehHC=5QYWrPXSdw at mail dot gmail dot com> <20131118161925 dot GA3928 at domone dot podge> <CALoOobPX9d5uP7rn4OykZn6=2Ofvnran06vZL6-Z5r-tDzrsuA at mail dot gmail dot com>
On Mon, Nov 18, 2013 at 01:25:09PM -0800, Paul Pluzhnikov wrote:
> On Mon, Nov 18, 2013 at 8:19 AM, OndÅej BÃlka <neleai@seznam.cz> wrote:
>
> > This is quite confusing when stack grows up, a this part will skip a
> > AT_NULL auxval
>
> I don't believe that's true. AFAIU, there will still be a NULL entry
> between the last envp and the auxv[].
>
Then this makes less sense as AT_NULL is last entry a test would fail
without reading auxval.
Is there any reason not use getauxval(AT_EXECFN) for test?
> However, I don't have any machines that use strack grows up, so I can't test it.
>
> > and what happens if there is no AT_EXECFN?
>
> The test would fail. Are there any kernels on which glibc works, which
> have aux vector, and which do not put AT_EXECFN into it?
>
> Thanks,
> --
> Paul Pluzhnikov