[PATCH 2/2] Get rid of online_cpus() when determining grep threads num
- From
Victor Leschuk <vleschuk@gmail.com>
- Date
- Dec 15, 2015, 15:31 UTC
- Message-ID
- <1450193500-22468-3-git-send-email-vleschuk@accesssoftek.com>
- In-Reply-To
- <1450193500-22468-1-git-send-email-vleschuk@accesssoftek.com>
Number of threads now doesn't depend on online_cpus(), e.g. if specific number is not configured GREP_NUM_THREADS_DEFAULT (8) threads will be used even on 1-core CPU.
Reason: multithreading can improve performance even on single core machines as IO is also a major factor here. Using multiple threads can significantly boost grep performance when working on slow filesystems (or repo isn't cached) or through network (for example repo is located on NFS).
Signed-off-by: Victor Leschuk <vleschuk@accesssoftek.com> --- builtin/grep.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/builtin/grep.c b/builtin/grep.c index e9aebab..1315905 100644 --- a/builtin/grep.c +++ b/builtin/grep.c @@ -827,7 +827,7 @@ int cmd_grep(int argc, const char **argv, const char *prefix) } #ifndef NO_PTHREADS - if (list.nr || cached || online_cpus() == 1 || show_in_pager) { + if (list.nr || cached || show_in_pager) { /* Can not multi-thread object lookup */ num_threads = 0; }
-- 2.6.3.369.g3e7f205.dirty