Re: Buildbot failure in Wildebeest Builder on whole buildset

[Arthur Cohen <arthur.cohen@embecosm.com>]

[-- Attachment #1.1.1: Type: text/plain, Size: 1232 bytes --]

Hi Mark,

On 3/7/22 15:42, Mark Wielaard wrote:
> Hi Arthur,
> 
> On Mon, 2022-03-07 at 15:32 +0100, Arthur Cohen wrote:
>> I suspect this might be due to bors, our merging bot on github, and me
>> asking it to merge multiple pull requests this morning (around 4 or 5).
>> Not really sure.
>>
>> But the bug is fixed and properly placed in the history :)
> 
> Aha, I see, you are right. For some reason bors adds commits on top of
> old commits with a merge instead of doing a simple rebase on top of the
> latest commit. Making the git history look like a Christmas tree:
> https://code.wildebeest.org/git/mirror/gccrs/log/
> 
> Your new commits were indeed on top of a 5 day old commit, so the
> buildbot dutifully tested it in context. And that context didn't
> include the fix yet.
> 
> Would it make sense to tell bors to do a rebase instead of a merge
> before pushing to trunk?
> 
> Cheers,
> 
> Mark

I've checked, and I don't think that bors is able to do that. Since it 
maintains a staging queue containing multiple PRs and then merges it 
onto the main branch if all tests pass, I think the christmas tree is 
unavoidable.

I'd be happy to be proved wrong however.

Kindly,

Arthur

[-- Attachment #1.1.2: OpenPGP public key --]
[-- Type: application/pgp-keys, Size: 3195 bytes --]

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 840 bytes --]