Question on backporting. [Was Re: [PATCH v2 2/2] powerpc: Add HWCAP3/HWCAP4 data to TCB for Power Architecture.]

Question on backporting. [Was Re: [PATCH v2 2/2] powerpc: Add HWCAP3/HWCAP4 data to TCB for Power Architecture.]

[Peter Bergner]

On 3/19/24 6:50 PM, Peter Bergner wrote:
> On 12/13/23 5:47 PM, Peter Bergner wrote:
>> On 12/5/23 6:38 AM, Manjunath Matti wrote:
>>> This patch adds a new feature for powerpc.  In order to get faster
>>> access to the HWCAP3/HWCAP4 masks, similar to HWCAP/HWCAP2 (i.e. for
>>> implementing __builtin_cpu_supports() in GCC) without the overhead of
>>> reading them from the auxiliary vector, we now reserve space for them
>>> in the TCB.
>>>
>>> This is an ABI change for GLIBC 2.39.
>>>
>>> Suggested-by: Peter Bergner <bergner@linux.ibm.com>
>>> ---
>>>  elf/dl-diagnostics.c                          |  2 +
>>>  elf/dl-support.c                              |  2 +
>>>  elf/elf.h                                     |  4 ++
>>>  sysdeps/generic/ldsodefs.h                    |  2 +
>>>  sysdeps/powerpc/dl-procinfo.c                 |  6 ++-
>>>  sysdeps/powerpc/dl-procinfo.h                 | 52 +++++++++++++------
>>>  sysdeps/powerpc/hwcapinfo.c                   | 11 ++--
>>>  sysdeps/unix/sysv/linux/dl-parse_auxv.h       |  2 +
>>>  sysdeps/unix/sysv/linux/dl-sysdep.c           |  2 +
>>>  .../unix/sysv/linux/powerpc/cpu-features.c    |  2 +
>>>  .../unix/sysv/linux/powerpc/cpu-features.h    |  2 +
>>>  sysdeps/unix/sysv/linux/powerpc/libc-start.c  |  6 +++
>>>  12 files changed, 74 insertions(+), 19 deletions(-)
>>
>> LGTM.  The only catch is that this patch is dependent on the associated
>> kernel patch that adds the AT_HWCAP3 and AT_HWCAP4 getting upstream so
>> glibc can rely on their values.  I'm still working on upstreaming that.
> 
> The kernel patch this is dependent on has finally reached Linus's tree,
> so I have pushed this patch.

As someone new to the Power maintainership role, can someone help explain
the process/rules for backporting changes into the release branches?
The above patch which was reviewed and approved before the 2.39 release
had to wait for the Linux kernel patch which it was dependent on, before
being pushed, thus missing the 2.39 release (patch 1/2 did make it).

As port maintainer do I have discretion to decide whether this patch
can be backported to the release branches or do I need consensus that
backporting it is ok?

I'll mention I would also like to backport Amrita's recent Power11 patch
as well to the 2.39 release branch.

Peter

Re: Question on backporting. [Was Re: [PATCH v2 2/2] powerpc: Add HWCAP3/HWCAP4 data to TCB for Power Architecture.]

[Tulio Magno Quites Machado Filho]

Peter Bergner <bergner@linux.ibm.com> writes:

> As someone new to the Power maintainership role, can someone help explain
> the process/rules for backporting changes into the release branches?

I like the explanation from this section of the wiki:
https://sourceware.org/glibc/wiki/Release/#General_policy

> As port maintainer do I have discretion to decide whether this patch
> can be backported to the release branches or do I need consensus that
> backporting it is ok?

In my interpretation, consensus is still important, but your input is
also very important.
With that said I believe both patches fall into the following:

    ... any patch on master that doesn't change ABI or API is immediately
    suitable for backporting to a stable branch.

This is what I'd do, assuming there are no time constraints:

1. Submit both patches to libc-stable.
2. If nobody objects, push the changes to the stable branch.

-- 
Tulio Magno

Re: Question on backporting. [Was Re: [PATCH v2 2/2] powerpc: Add HWCAP3/HWCAP4 data to TCB for Power Architecture.]

[Adhemerval Zanella Netto]

On 20/03/24 16:08, Tulio Magno Quites Machado Filho wrote:
> Peter Bergner <bergner@linux.ibm.com> writes:
> 
>> As someone new to the Power maintainership role, can someone help explain
>> the process/rules for backporting changes into the release branches?
> 
> I like the explanation from this section of the wiki:
> https://sourceware.org/glibc/wiki/Release/#General_policy
> 
>> As port maintainer do I have discretion to decide whether this patch
>> can be backported to the release branches or do I need consensus that
>> backporting it is ok?
> 
> In my interpretation, consensus is still important, but your input is
> also very important.
> With that said I believe both patches fall into the following:
> 
>     ... any patch on master that doesn't change ABI or API is immediately
>     suitable for backporting to a stable branch.
> 
> This is what I'd do, assuming there are no time constraints:
> 
> 1. Submit both patches to libc-stable.
> 2. If nobody objects, push the changes to the stable branch.
> 

Both patch should be ok to backport, since the trickiest that added
__tcb_parse_hwcap_and_convert_at_platform was done on 2.39.