On Tue, Nov 10, 2015 at 8:28 AM, Victor Leschuk <vleschuk@xxxxxxxxx> wrote:
> "git grep" can now be configured (or told from the command line)
> how many threads to use when searching in the working tree files.
>
> Signed-off-by: Victor Leschuk <vleschuk@xxxxxxxxxxxxxxxx>
> ---
As an aid for reviewers, please use this space (below the "---" line)
to describe what changed since the previous version. It's also helpful
if you can provide a link to the previous version in the mailing list
archive.
> Documentation/config.txt | 7 +++++
> Documentation/git-grep.txt | 15 ++++++++++
> builtin/grep.c | 50 +++++++++++++++++++++++-----------
> contrib/completion/git-completion.bash | 1 +
> 4 files changed, 57 insertions(+), 16 deletions(-)
>
> diff --git a/Documentation/config.txt b/Documentation/config.txt
> @@ -1447,6 +1447,13 @@ grep.extendedRegexp::
> option is ignored when the 'grep.patternType' option is set to a value
> other than 'default'.
>
> +grep.threads::
> + Number of grep worker threads, use it to tune up performance on
> + your machines. Leave it unset (or set to 0) for default behavior,
> + which for now is using 8 threads for all systems.
> + Default behavior can be changed in future versions
> + to better suite hardware and circumstances.
s/suite/suit/ (here and elsewhere)
Was the conclusion of discussion that there should be some explanation
here that this is more about tuning for I/O rather than CPU, or did I
misunderstand?
> gpg.program::
> Use this custom program instead of "gpg" found on $PATH when
> making or verifying a PGP signature. The program must support the
> diff --git a/Documentation/git-grep.txt b/Documentation/git-grep.txt
> index 4a44d6d..91027b6 100644
> --- a/Documentation/git-grep.txt
> +++ b/Documentation/git-grep.txt
> diff --git a/builtin/grep.c b/builtin/grep.c
> @@ -832,14 +846,18 @@ int cmd_grep(int argc, const char **argv, const char *prefix)
> }
>
> #ifndef NO_PTHREADS
> - if (list.nr || cached || online_cpus() == 1)
> - use_threads = 0;
> + if (list.nr || cached)
> + num_threads = 0; /* Can not multi-thread object lookup */
> + else if (num_threads == 0)
> + num_threads = GREP_NUM_THREADS_DEFAULT; /* User didn't specify value, or just wants default behavior */
> + else if (num_threads < 0)
> + die("Invalid number of threads specified (%d)", num_threads);
The original code consulted online_cpus(), but the new code does not.
This is a sufficiently significant change that it deserves mention in
the commit message. In fact, it's really a distinct change which might
deserve being done in its own preparatory patch (with an explanation
something along the lines of "even single-core machines may benefit
from threading when the bottleneck is I/O").
> #else
> - use_threads = 0;
> + num_threads = 0;
> #endif
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html