Le 12/06/2019 à 11:23, Paul Mackerras a écrit :
On Wed, Jun 12, 2019 at 09:42:52AM +0200, Christophe Leroy wrote:
Le 12/06/2019 à 09:22, Michael Neuling a écrit :
In commit c1fe190c0672 ("powerpc: Add force enable of DAWR on P9
option") I screwed up some assembler and corrupted a pointer in
r3. This resulted in crashes like the below from Cédric:
Iaw Documentation/process/submitting-patches.rst:
Describe your changes in imperative mood, e.g. "make xyzzy do frotz"
instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy
to do frotz", as if you are giving orders to the codebase to change
its behaviour.
So you could rephrase as follows for instance:
Commit XXXX ("") screwed up some assembler ....
That advice in submitting-patches.rst is certainly appropriate when
talking about the actual change that the patch makes. However, it is
also appropriate to give descriptive background material that helps
the reader to understand why the change is necessary -- in this case,
where and how the bug was introduced. So I'm going to support Mikey
as regards his first few paragraphs.
Does it really matter knowing that it is Mikey who screwed up the
assembler ? For me what's important is to know which commit introduced
the error, not who made the error, isn't it ?
Christophe
I agree that the last paragraph that says "This fixes the bug by ..."
could be reworded as "Fix the bug by ...".
Paul.
