Source-Changes-D archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: CVS commit: src/crypto/external/bsd/openssl



Thanks for pointing this out! I did just that.

christos

> On Apr 16, 2025, at 5:54 PM, Taylor R Campbell <riastradh%netbsd.org@localhost> wrote:
> 
>> Module Name:    src
>> Committed By:   christos
>> Date:           Wed Apr 16 15:23:17 UTC 2025
>> 
>> Log Message:
>> Resolve conflicts between 3.0.15 and 3.0.16:
>> - Reverted 3.0.16 changes to idea, rc2, and rc5 to retain binary compatibility
>>  The patches to undo the reversion for the next import (which will require
>>  a version bump) are in patches.
>> - Restored test for ppc machdep.fpu_present
>>  We should probably contact upstream for them to put it back.
> 
> This makes no sense.  There were no upstream changes to idea, rc2, or
> rc5 between 3.0.15 and 3.0.16, as you can see by doing:
> 
> cvs rdiff -pu -r openssl-3-0-15 -r openssl-3-0-16 src/crypto/external/bsd/openssl/dist
> 
> So there should have been no changes when you merged the upstream
> update into HEAD.
> 
> It appears that, instead of doing a three-way merge as in
> <https://mewburn.net/luke/papers/3rdparty-and-cvs.html>, you just
> ditched all our local changes and replaced them by the OpenSSL update.
> 
> And then you selectively reapplied some of them, like the ppc
> machdep.fpu_present test, which was a local change all along, not
> something that upstream removed.  Next time, instead of just
> reapplying this change, you should take it as a giant red flag that
> you aren't actually doing a three-way merge.
> 
> I think you need to redo this from the beginning by reverting the
> 3.0.16 update and using `cvs co -j openssl-3-0-15 -j openssl-3-0-16'
> properly; there are lots of important local changes that you summarily
> discarded this way.

Attachment: signature.asc
Description: Message signed with OpenPGP



Home | Main Index | Thread Index | Old Index