Re: [PATCH 1/13 ver 3] rs6000, Remove __builtin_vsx_cmple* builtins

["Kewen.Lin" <linkw@linux.ibm.com>]

Hi Carl,

on 2024/6/6 06:25, Carl Love wrote:
> Kewen:
> 
> On 6/3/24 23:00, Kewen.Lin wrote:
>> Hi Carl,
>>
>> on 2024/5/29 23:52, Carl Love wrote:
>>> This patch was approved in the previous series.  There are no changes to this patch.  Reposting for completeness. 
>> I guess you can just push the approved ones, as there is no dependency
>> between any two of them?  It can help to reduce the size of this series.
> 
> The patches do touch some similar files so they are not completely independent from a patch standpoint.  Functionally they are all independent.
> 
> I tried applying the approved patches only to the current mainline tree.  The approved patches were: 1,3,5 (with tweak), 6, 8, 9, 10, 12.  Patch 5 requires a little rebasing due to a little fuzz in the lines.  Not a big deal.  Patch 8 also doesn't apply cleanly with git.  The patch command gets a little confused when I tried to use it, so I had to manually "recreate" the patch.  The changes are straight forward so that is fairly easy.  The rest of the patches applied cleanly with git. I am guessing there will be some rebasing needed for the non-approved patches to apply them after the approved patches.

IMHO, you can first reorder the patches in your WIP branch (like the approved ones go first) with git rebase -i,
then rebase with the latest trunk, there may be some conflicts but I'd expect there are not many.

> 
> The main reason that I reposted everything was that the patch numbers changed and I wanted it to be fairly clear what was going on.

OK, if you push them, you can also specify the commit hashes for the pushed ones in cover letter.

> 
> I toyed with the idea of committing the 8 approved patches and then working on the additional 5 but I think that is hard as I would have to manually adjust the patch numbers to keep them lined up with version 3 or version 4 has a new numbering patches 1 to 5 (i.e. remapping of version 3 patch numbers).  Either way I think it would be hard/confusing. 
> 
> Given that separating out the approved and non-approved patches causes some re-basing issues, it is probably best to just update the 5 patches, posting them as version 4 and not re-post the whole series. I will just note in the header patch 0/13 the patches that have already been approved.  I hope that is ok?

Sure, I'm totally fine if you prefer this way. :)

BR,
Kewen