[RFT] Glibc 2.23 and binutils 2.26

[RFT] Glibc 2.23 and binutils 2.26

[Khem Raj]

[-- Attachment #1: Type: text/plain, Size: 531 bytes --]

Hi all

upstream glibc and binutils release branches have been cut out and are being made ready for next release
I have put together update patchset for both of them

here are the branches

https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23

FYI These are still using autorev to ensure that we test tip of release branches

Please give them a shot in your environments and report any issues you encounter.

Thanks for help

-Khem




[-- Attachment #2: Message signed with OpenPGP using GPGMail --]
[-- Type: application/pgp-signature, Size: 211 bytes --]

Re: [oe] [RFT] Glibc 2.23 and binutils 2.26

[akuster808]

On 01/17/2016 03:54 AM, Khem Raj wrote:
> Hi all
> 
> upstream glibc and binutils release branches have been cut out and are being made ready for next release
> I have put together update patchset for both of them
> 
> here are the branches
> 
> https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
> https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23

Several Security fixes will be included in glibc-2.23 update.

https://bugzilla.yoctoproject.org/show_bug.cgi?id=8980


CVE-2015-8776 - Passing out of range data to strftime() causes a segfault

https://sourceware.org/bugzilla/show_bug.cgi?id=18985


CVE-2015-8777 - LD_POINTER_GUARD is not ignored for privileged binaries
https://sourceware.org/bugzilla/show_bug.cgi?id=18928

CVE-2015-8778 - hcreate((size_t)-1) should fail with ENOMEM
https://sourceware.org/bugzilla/show_bug.cgi?id=18240

CVE-2014-9761 - nan function unbounded stack allocation
https://sourceware.org/bugzilla/show_bug.cgi?id=16962

CVE-2015-8779 - catopen() Multiple unbounded stack allocations
https://sourceware.org/bugzilla/show_bug.cgi?id=17905

> 
> FYI These are still using autorev to ensure that we test tip of release branches
> 
> Please give them a shot in your environments and report any issues you encounter.

BTW, Mips64 Octeon3 works fine. Aarch64 qemu boot tested.

- Armin
> 
> Thanks for help
> 
> -Khem
> 
> 
> 
> 
> 

Re: [oe] [RFT] Glibc 2.23 and binutils 2.26

[ChenQi]

[-- Attachment #1: Type: text/plain, Size: 1038 bytes --]

On 01/17/2016 07:54 PM, Khem Raj wrote:
> Hi all
>
> upstream glibc and binutils release branches have been cut out and are being made ready for next release
> I have put together update patchset for both of them
>
> here are the branches
>
> https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
> https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23
>
> FYI These are still using autorev to ensure that we test tip of release branches
>
> Please give them a shot in your environments and report any issues you encounter.
>
> Thanks for help
>
> -Khem
>
>
>
>
>

0001-Generate-relocatable-SDKs.patch is included in SRC_URI twice, which 
causes the following error when building binutils-crosssdk-XXX.

ERROR: Command Error: exit status: 1  Output:
Patch 0001-Generate-relocatable-SDKs.patch is already applied; check 
your series file
ERROR: Function failed: patch_do_patch

binutils-2.26.inc and binutils-crosssdk_2.26.bb both has this patch in 
its SRC_RUI.

Regards,
Chen Qi

[-- Attachment #2: Type: text/html, Size: 1779 bytes --]

Re: [oe] [RFT] Glibc 2.23 and binutils 2.26

[Khem Raj]

[-- Attachment #1: Type: text/plain, Size: 1463 bytes --]


> On Jan 26, 2016, at 9:11 PM, ChenQi <Qi.Chen@windriver.com> wrote:
> 
> On 01/17/2016 07:54 PM, Khem Raj wrote:
>> Hi all
>> 
>> upstream glibc and binutils release branches have been cut out and are being made ready for next release
>> I have put together update patchset for both of them
>> 
>> here are the branches
>> 
>> https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
>> https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23
>> 
>> FYI These are still using autorev to ensure that we test tip of release branches
>> 
>> Please give them a shot in your environments and report any issues you encounter.
>> 
>> Thanks for help
>> 
>> -Khem
>> 
>> 
>> 
>> 
>> 
> 
> 0001-Generate-relocatable-SDKs.patch is included in SRC_URI twice, which causes the following error when building binutils-crosssdk-XXX.
> 
> ERROR: Command Error: exit status: 1  Output:
> Patch 0001-Generate-relocatable-SDKs.patch is already applied; check your series file
> ERROR: Function failed: patch_do_patch
> 
> binutils-2.26.inc and binutils-crosssdk_2.26.bb both has this patch in its SRC_RUI.
> 

Thanks for report. fixed and squashed into original commit in same branch as above.

> Regards,
> Chen Qi
> --
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-devel


[-- Attachment #2: Message signed with OpenPGP using GPGMail --]
[-- Type: application/pgp-signature, Size: 211 bytes --]

Re: [RFT] Glibc 2.23 and binutils 2.26

[Alexander Kanavin]

On 01/17/2016 01:54 PM, Khem Raj wrote:
> upstream glibc and binutils release branches have been cut out and are being made ready for next release
> I have put together update patchset for both of them
>
> here are the branches
>
> https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
> https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23

Hello Khem,

when are you planning to submit these?


Alex

Re: [RFT] Glibc 2.23 and binutils 2.26

[Khem Raj]

On Fri, Jan 29, 2016 at 6:14 AM, Alexander Kanavin
<alexander.kanavin@linux.intel.com> wrote:
> On 01/17/2016 01:54 PM, Khem Raj wrote:
>>
>> upstream glibc and binutils release branches have been cut out and are
>> being made ready for next release
>> I have put together update patchset for both of them
>>
>> here are the branches
>>
>> https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
>> https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23
>
>
> Hello Khem,
>
> when are you planning to submit these?

binutils exposes a problem for ppc as a result kernel fails to build.
The fix is to drop
http://git.openembedded.org/openembedded-core/tree/meta/recipes-devtools/gcc/gcc-5.3/0030-Enable-SPE-AltiVec-generation-on-powepc-linux-target.patch

that patch makes generic ppc configs to be specific to fsl ppc.
but we need context as to why this was done

glibc is still not released.

>
>
> Alex
>
>
> --
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core

Re: [RFT] Glibc 2.23 and binutils 2.26

[Richard Purdie]

On Fri, 2016-01-29 at 06:44 -0800, Khem Raj wrote:
> On Fri, Jan 29, 2016 at 6:14 AM, Alexander Kanavin
> <alexander.kanavin@linux.intel.com> wrote:
> > On 01/17/2016 01:54 PM, Khem Raj wrote:
> > > 
> > > upstream glibc and binutils release branches have been cut out
> > > and are
> > > being made ready for next release
> > > I have put together update patchset for both of them
> > > 
> > > here are the branches
> > > 
> > > https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
> > > https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23
> > 
> > 
> > Hello Khem,
> > 
> > when are you planning to submit these?
> 
> binutils exposes a problem for ppc as a result kernel fails to build.
> The fix is to drop
> http://git.openembedded.org/openembedded-core/tree/meta/recipes-devto
> ols/gcc/gcc-5.3/0030-Enable-SPE-AltiVec-generation-on-powepc-linux
> -target.patch
> 
> that patch makes generic ppc configs to be specific to fsl ppc.
> but we need context as to why this was done
> 
> glibc is still not released.

I ran the binutils upgrade in -next on the autobuilder. It looked
mostly good but with poky-lsb we did see:

https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb/b
uilds/635/steps/BuildImages/logs/stdio

https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb/b
uilds/635/steps/BuildImages_1/logs/stdio

Cheers,

Richard

Re: [RFT] Glibc 2.23 and binutils 2.26

[Khem Raj]

On Sun, Jan 31, 2016 at 12:32 AM, Richard Purdie
<richard.purdie@linuxfoundation.org> wrote:
> On Fri, 2016-01-29 at 06:44 -0800, Khem Raj wrote:
>> On Fri, Jan 29, 2016 at 6:14 AM, Alexander Kanavin
>> <alexander.kanavin@linux.intel.com> wrote:
>> > On 01/17/2016 01:54 PM, Khem Raj wrote:
>> > >
>> > > upstream glibc and binutils release branches have been cut out
>> > > and are
>> > > being made ready for next release
>> > > I have put together update patchset for both of them
>> > >
>> > > here are the branches
>> > >
>> > > https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
>> > > https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23
>> >
>> >
>> > Hello Khem,
>> >
>> > when are you planning to submit these?
>>
>> binutils exposes a problem for ppc as a result kernel fails to build.
>> The fix is to drop
>> http://git.openembedded.org/openembedded-core/tree/meta/recipes-devto
>> ols/gcc/gcc-5.3/0030-Enable-SPE-AltiVec-generation-on-powepc-linux
>> -target.patch
>>
>> that patch makes generic ppc configs to be specific to fsl ppc.
>> but we need context as to why this was done
>>
>> glibc is still not released.
>
> I ran the binutils upgrade in -next on the autobuilder. It looked
> mostly good but with poky-lsb we did see:
>
> https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb/b
> uilds/635/steps/BuildImages/logs/stdio
>
> https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb/b
> uilds/635/steps/BuildImages_1/logs/stdio

it looks like -fpie is the factor here, I will try to reproduce it here locally
may be we can get away by using system zlib instead of internal one

>
> Cheers,
>
> Richard
>

Re: [RFT] Glibc 2.23 and binutils 2.26

[Richard Purdie]

On Sun, 2016-01-31 at 08:32 +0000, Richard Purdie wrote:
> On Fri, 2016-01-29 at 06:44 -0800, Khem Raj wrote:
> > On Fri, Jan 29, 2016 at 6:14 AM, Alexander Kanavin
> > <alexander.kanavin@linux.intel.com> wrote:
> > > On 01/17/2016 01:54 PM, Khem Raj wrote:
> > > > 
> > > > upstream glibc and binutils release branches have been cut out
> > > > and are
> > > > being made ready for next release
> > > > I have put together update patchset for both of them
> > > > 
> > > > here are the branches
> > > > 
> > > > https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.
> > > > 26
> > > > https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23
> > > 
> > > 
> > > Hello Khem,
> > > 
> > > when are you planning to submit these?
> > 
> > binutils exposes a problem for ppc as a result kernel fails to
> > build.
> > The fix is to drop
> > http://git.openembedded.org/openembedded-core/tree/meta/recipes-dev
> > to
> > ols/gcc/gcc-5.3/0030-Enable-SPE-AltiVec-generation-on-powepc-linux
> > -target.patch
> > 
> > that patch makes generic ppc configs to be specific to fsl ppc.
> > but we need context as to why this was done
> > 
> > glibc is still not released.
> 
> I ran the binutils upgrade in -next on the autobuilder. It looked
> mostly good but with poky-lsb we did see:
> 
> https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb
> /b
> uilds/635/steps/BuildImages/logs/stdio
> 
> https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb
> /b
> uilds/635/steps/BuildImages_1/logs/stdio

Good news is I fixed the binutils build issues. 

There was a window of available time on the autobuilder so I tested
glibc 2.23 which failed in ltp with:

https://autobuilder.yoctoproject.org/main/builders/nightly-arm/builds/634/steps/BuildImages/logs/stdio

bdflush01.c:84:25: fatal error: sys/kdaemon.h: No such file or directory

That made the builds rather red with many failures but was the only
error I could spot.

Cheers,

Richard

Re: [RFT] Glibc 2.23 and binutils 2.26

[Khem Raj]

On Mon, Feb 1, 2016 at 1:12 AM, Richard Purdie
<richard.purdie@linuxfoundation.org> wrote:
> On Sun, 2016-01-31 at 08:32 +0000, Richard Purdie wrote:
>> On Fri, 2016-01-29 at 06:44 -0800, Khem Raj wrote:
>> > On Fri, Jan 29, 2016 at 6:14 AM, Alexander Kanavin
>> > <alexander.kanavin@linux.intel.com> wrote:
>> > > On 01/17/2016 01:54 PM, Khem Raj wrote:
>> > > >
>> > > > upstream glibc and binutils release branches have been cut out
>> > > > and are
>> > > > being made ready for next release
>> > > > I have put together update patchset for both of them
>> > > >
>> > > > here are the branches
>> > > >
>> > > > https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.
>> > > > 26
>> > > > https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23
>> > >
>> > >
>> > > Hello Khem,
>> > >
>> > > when are you planning to submit these?
>> >
>> > binutils exposes a problem for ppc as a result kernel fails to
>> > build.
>> > The fix is to drop
>> > http://git.openembedded.org/openembedded-core/tree/meta/recipes-dev
>> > to
>> > ols/gcc/gcc-5.3/0030-Enable-SPE-AltiVec-generation-on-powepc-linux
>> > -target.patch
>> >
>> > that patch makes generic ppc configs to be specific to fsl ppc.
>> > but we need context as to why this was done
>> >
>> > glibc is still not released.
>>
>> I ran the binutils upgrade in -next on the autobuilder. It looked
>> mostly good but with poky-lsb we did see:
>>
>> https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb
>> /b
>> uilds/635/steps/BuildImages/logs/stdio
>>
>> https://autobuilder.yoctoproject.org/main/builders/nightly-x86-64-lsb
>> /b
>> uilds/635/steps/BuildImages_1/logs/stdio
>
> Good news is I fixed the binutils build issues.
>
> There was a window of available time on the autobuilder so I tested
> glibc 2.23 which failed in ltp with:
>
> https://autobuilder.yoctoproject.org/main/builders/nightly-arm/builds/634/steps/BuildImages/logs/stdio
>
> bdflush01.c:84:25: fatal error: sys/kdaemon.h: No such file or directory
>
> That made the builds rather red with many failures but was the only
> error I could spot.

yes this is fixed and has been upstreamed as well :) see
https://github.com/linux-test-project/ltp/commit/739bcd5619250b67f7d4eb28eb3ac68ac05db0f1

locally I am carrying a bigger set of fixes ( for musl ) so I did not
propose it yet.
I was expecting that we will upgrade ltp to latest release (20160126)
and get the above fix directly from upstream
before we upgrade to new glibc. If no one is upgrading ltp, I will put
an upgrade patch in my tree.


>
> Cheers,
>
> Richard
>
>