[OE-core] [PATCH 4/6] tcmode-default: Bump gcc,glibc,gdb

[Bruce Ashfield]

On 2016-05-15 5:25 AM, Khem Raj wrote:
> On Sun, May 15, 2016 at 12:03 AM, Richard Purdie
> <richard.purdie at linuxfoundation.org> wrote:
>> On Fri, 2016-05-13 at 09:46 -0700, Khem Raj wrote:
>>> gcc 6.1, glibc 2.24, gdb 7.11
>>>
>>> Signed-off-by: Khem Raj <raj.khem at gmail.com>
>>> ---
>>>  meta/conf/distro/include/tcmode-default.inc | 6 +++---
>>>  1 file changed, 3 insertions(+), 3 deletions(-)
>>
>> I'll merge most of this series however before we can move to gcc6 by
>> default there are some remaining issues:
>>
>> https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb/b
>> uilds/788/steps/BuildImages/logs/stdio
>>
>> qt4-x11-free fails to compile
>
> This can be fixed easily. I will see if I can cook up one.
>
>> linux-yocto 4.1 also fails
>
> I am sure if we upgrade to latest point release of 4.1 then most of
> fixes may be available.
> but there are more than 1 fixes that would require backport

Two 4.1 -stable updates came out just recently. I already have them
queued here. I'll send them along today .. will have a scan to see if
the gcc6 changes were picked up by the maintainer.

>
>>
>> (on qemux86-64, qemux86, qemuppc, qemumips, qemuarm)
>>
>> https://autobuilder.yoctoproject.org/main/builders/nightly-ppc/builds/7
>> 78/steps/BuildImages_1/logs/stdio
>>
>> linux-yocto 4.4 fails on ppc
>>
>> https://autobuilder.yoctoproject.org/main/builders/nightly-non-gpl3/bui
>> lds/768/steps/BuildImages/logs/stdio
>
> This patch https://patchwork.ozlabs.org/patch/614548/ is required and
> I gave pointers for that to Bruce
> and was hoping that with
> http://git.openembedded.org/openembedded-core/commit/?id=f1c75b93a4e11425e595c5ce043fbb0276a41931
> the above fix is bundled in but Bruce might be able to conform.
>

Confirmed that the fix is in that commit. In linux-yocto:

Author: Khem Raj <raj.khem at gmail.com>
Date:   Mon Apr 25 05:47:01 2016 +0000

     powerpc/ptrace: Fix out of bounds array access warning

     gcc-6 correctly warns about a out of bounds access

     arch/powerpc/kernel/ptrace.c:407:24: warning: index 32 denotes an 
offset greater than size of 'u64[32][1] {aka long long unsigned 
int[32][1]}' [-Warray-bounds]
             offsetof(struct thread_fp_state, fpr[32][0]));
                             ^

     check the end of array instead of beginning of next element to fix this

     Signed-off-by: Khem Raj <raj.khem at gmail.com>
     Cc: Kees Cook <keescook at chromium.org>
     Cc: Michael Ellerman <mpe at ellerman.id.au>
     Cc: Segher Boessenkool <segher at kernel.crashing.org>
     Signed-off-by: Bruce Ashfield <bruce.ashfield at windriver.com>

All the branches have that commit now, and it is covered by that SRCREV
bump. So there's something else lurking for ppc.

Bruce

>
>>
>> non-GPLv3 on elfutils and diffutils
>
> Yes, now, I wonder if we can backport fixes from newer versions and
> still call it GPL-2 without getting
> consent from patch authors. This is going to be a bit twisted.
>
>>
>> (There may be other failures too, the build hasn't completed yet)
>>
>> We'll need to come up with a plan to address these issues (possibly
>> upgrades for some of the kernels?) before we can change the defaults.
>>
>> Cheers,
>>
>> Richard
>>
>>