On 10/01/18 22:40, Johannes Schindelin wrote:
> Hi,
>
> On Wed, 10 Jan 2018, Jonathan Nieder wrote:
>
>> Phillip Wood wrote:
>>
>>> From: Phillip Wood <phillip.wood@xxxxxxxxxxxxx>
>>>
>>> If the commit message does not need to be edited then create the
>>> commit without forking 'git commit'. Taking the best time of ten runs
>>> with a warm cache this reduces the time taken to cherry-pick 10
>>> commits by 27% (from 282ms to 204ms), and the time taken by 'git
>>> rebase --continue' to pick 10 commits by 45% (from 386ms to 212ms) on
>>> my computer running linux. Some of greater saving for rebase is
>>> because it no longer wastes time creating the commit summary just to
>>> throw it away.
>>
>> Neat! Dmitry Torokhov (cc-ed) noticed[1]
Thanks for reporting and bisecting this Dmitry. When I was preparing
this series I checked to see if it needed to run the 'pre-commit' hook
but missed the 'prepare-commit-msg' hook.
> that this causes the
>> prepare-commit-msg hook not to be invoked, which I think is
>> unintentional. Should we check for such a hook and take the slowpath
>> when it is present?
>
> We could also easily recreate the functionality:
>
> if (find_hook("pre-commit")) {
> struct argv_array hook_env = ARGV_ARRAY_INIT;
>
> argv_array_pushf(&hook_env, "GIT_INDEX_FILE=%s",
> get_index_file());
> argv_array_push(&hook_env, "GIT_EDITOR=:");
> ret = run_hook_le(hook_env.argv, "pre-commit", NULL);
> argv_array_clear(&hook_env);
> }
Thanks Johannes, though it needs to run the 'prepare-commit-msg' hook,
the current code in master only runs the 'pre-commit' hook when we edit
the message. I'll send a patch with a test.
Best Wishes
Phillip
> (This assumes that the in-process try_to_commit() is only called if the
> commit message is not to be edited interactively, which is currently the
> case.)
>
> Ciao,
> Dscho
>
