glibc-2.35 - Entering slush ABI freeze / Reviewing release blockers.

glibc-2.35 - Entering slush ABI freeze / Reviewing release blockers.

[Carlos O'Donell]

Community,

We are into the slushy ABI freeze. If you need to make ABI changes that are not
already in the "Release Blockers" [1] list please start a discussion about them
here to decide if they are really release blockers.

We have 3 weeks of bug fixing and stabilization before I plan to cut the release.

-- 
Cheers,
Carlos.

[1] https://sourceware.org/glibc/wiki/Release/2.35#Release_blockers.3F

Re: glibc-2.35 - Entering slush ABI freeze / Reviewing release blockers.

[Siddhesh Poyarekar]

On 06/01/2022 20:09, Carlos O'Donell via Libc-alpha wrote:
> Community,
> 
> We are into the slushy ABI freeze. If you need to make ABI changes that are not
> already in the "Release Blockers" [1] list please start a discussion about them
> here to decide if they are really release blockers.
> 
> We have 3 weeks of bug fixing and stabilization before I plan to cut the release.
> 

Could we have the _FORTIFY_SOURCE header fixes[1] block the release? 
The header inconsistencies result in a number of functions (stpcpy, 
pread, etc.) not being fortified in !_GNU_SOURCE cases.

Thanks,
Siddhesh

[1] https://patchwork.sourceware.org/project/glibc/list/?series=6132

Re: glibc-2.35 - Entering slush ABI freeze / Reviewing release blockers.

[Adhemerval Zanella]

On 06/01/2022 11:39, Carlos O'Donell via Libc-alpha wrote:
> Community,
> 
> We are into the slushy ABI freeze. If you need to make ABI changes that are not
> already in the "Release Blockers" [1] list please start a discussion about them
> here to decide if they are really release blockers.
> 
> We have 3 weeks of bug fixing and stabilization before I plan to cut the release.
> 

Thanks for volunteer for 2.35 release manager.

I have added both 'posix: Add terminal control setting support for posix_spawn' [1]
and the 'Multiple rtld-audit fixes' [2]. Former is a request from shell developers
that I tried to include on 2.34 and I don't want to postpone to 2.36.

Latter bumps the LA_CURRENT for all ports (la_symbind for bind-now) and the aarch64
also changes the aarch64 interface (fixing current issues and add a way to support
SVE).

[1] https://patchwork.sourceware.org/project/glibc/patch/20211223173003.1037286-1-adhemerval.zanella@linaro.org/
[2]https://patchwork.sourceware.org/project/glibc/list/?series=6085

Re: glibc-2.35 - Entering slush ABI freeze / Reviewing release blockers.

[Carlos O'Donell]

On 1/6/22 09:47, Siddhesh Poyarekar wrote:
> On 06/01/2022 20:09, Carlos O'Donell via Libc-alpha wrote:
>> Community,
>> 
>> We are into the slushy ABI freeze. If you need to make ABI changes
>> that are not already in the "Release Blockers" [1] list please
>> start a discussion about them here to decide if they are really
>> release blockers.
>> 
>> We have 3 weeks of bug fixing and stabilization before I plan to
>> cut the release.
>> 
> 
> Could we have the _FORTIFY_SOURCE header fixes[1] block the release?
> The header inconsistencies result in a number of functions (stpcpy,
> pread, etc.) not being fortified in !_GNU_SOURCE cases.

Yes.

I *also* want _FORITFY_SOURCE=3 for glibc 2.35.

I want glibc 2.35 to work with gcc 12 to enable more fortification.

With glibc 2.35 and gcc 12 coming out close to eachother I don't want to wait
6 months for the fixes to land in released versions of the toolchain.

-- 
Cheers,
Carlos.

Re: glibc-2.35 - Entering slush ABI freeze / Reviewing release blockers.

[Siddhesh Poyarekar]

On 10/01/2022 21:22, Carlos O'Donell wrote:
> On 1/6/22 09:47, Siddhesh Poyarekar wrote:
>> On 06/01/2022 20:09, Carlos O'Donell via Libc-alpha wrote:
>>> Community,
>>>
>>> We are into the slushy ABI freeze. If you need to make ABI changes
>>> that are not already in the "Release Blockers" [1] list please
>>> start a discussion about them here to decide if they are really
>>> release blockers.
>>>
>>> We have 3 weeks of bug fixing and stabilization before I plan to
>>> cut the release.
>>>
>>
>> Could we have the _FORTIFY_SOURCE header fixes[1] block the release?
>> The header inconsistencies result in a number of functions (stpcpy,
>> pread, etc.) not being fortified in !_GNU_SOURCE cases.
> 
> Yes.
> 
> I *also* want _FORITFY_SOURCE=3 for glibc 2.35.
> 
> I want glibc 2.35 to work with gcc 12 to enable more fortification.
> 
> With glibc 2.35 and gcc 12 coming out close to eachother I don't want to wait
> 6 months for the fixes to land in released versions of the toolchain.

I'm fairly close to home on the gcc patchset, so I can make this happen. 
  The patch I've posted to enable _FORTIFY_SOURCE=3[1] in glibc should 
be unchanged; it failed because the partial patch didn't detect everything.

I'll add it as a desirable since I can always backport if for some 
reason I'm unable to finish the gcc work in time.

Thanks,
Siddhesh

[1] 
https://patchwork.sourceware.org/project/glibc/patch/20211217040753.4176265-1-siddhesh@sourceware.org/