This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc 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] PowerPC - Fix build with linux 3.7 headers


On Fri, Nov 23, 2012 at 10:33 AM, Carlos O'Donell
<carlos@systemhalted.org> wrote:
> On Fri, Nov 23, 2012 at 10:16 AM, Josh Boyer <jwboyer@gmail.com> wrote:
>> On Fri, Nov 23, 2012 at 8:20 AM, Carlos O'Donell
>> <carlos@systemhalted.org> wrote:
>>> On Fri, Nov 23, 2012 at 8:02 AM, Josh Boyer <jwboyer@gmail.com> wrote:
>>>> On Thu, Nov 22, 2012 at 11:07 PM, Adam Conrad <adconrad@0c3.net> wrote:
>>>>> The following simple patch fixes the build against linux 3.7 headers,
>>>>> where _ASM_POWERPC_ELF_H was renamed to _UAPI_ASM_POWERPC_ELF_H.
>>>>>
>>>>> Already tested and working in the current Ubuntu development series.
>>>>>
>>>>> 2012-11-22  Adam Conrad  <adconrad@0c3.net>
>>>>>
>>>>>        * sysdeps/unix/sysv/linux/powerpc/sys/procfs.h: Support
>>>>>        _UAPI_ASM_POWERPC_ELF_H in the Linux 3.7 headers.
>>>>>
>>>>> ---
>>>>
>>>> I wouldn't commit this.  It is arguably a bug in the kernel headers and
>>>> there is already a patch sent to fix it.  See here:
>>>>
>>>> http://article.gmane.org/gmane.linux.kernel.cross-arch/15847
>>>>
>>>> I believe Andrew Morton already queued this up.
>>>>
>>>> I have no authority in glibc at all, but I'd NAK this for now if I did.
>>>
>>> Josh,
>>>
>>> No true, you reviewed the patch, and made a comment. That counts.
>>
>> Heh, OK.
>>
>>> Adam,
>>>
>>> Has this bug made it into a release kernel that glibc could possibly
>>> be built with?
>>
>> It's in the 3.7-rcX kernels, and still present today in Linus' latest
>> tree but hopefully the final 3.7 kernel has it fixed.  The UAPI rework
>> is a 3.7 "feature" so it isn't present in older kernels.
>
> Thanks, in that case I think we need to watch the bug and fix it in
> glibc if a kernel makes it out with the header defect.
>
> Unless someone else has a stronger argument.

I think that's fine, though I really hope the kernel gets fixed soon.
As noted in the kernel patch, glibc is looking for specific header
guard defines in various kernel headers.  If the kernel ships with the
UAPI prefix in the installed headers, a more in-depth audit of which
kernel header guards are looked at is probably needed.

josh


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