diff options
| author | David Turner <dturner@twopensource.com> | 2014-06-18 05:54:42 (GMT) |
|---|---|---|
| committer | Junio C Hamano <gitster@pobox.com> | 2014-06-18 17:57:18 (GMT) |
| commit | 745224e04a03e4544c58d5d38d3c54f67100f8eb (patch) | |
| tree | 615a8a79eaedbb737408f3b3c2ff8f0940f92684 /t/valgrind | |
| parent | dde8a902c774268628015705a7767f61af18b865 (diff) | |
| download | git-745224e04a03e4544c58d5d38d3c54f67100f8eb.zip git-745224e04a03e4544c58d5d38d3c54f67100f8eb.tar.gz git-745224e04a03e4544c58d5d38d3c54f67100f8eb.tar.bz2 | |
refs.c: SSE2 optimizations for check_refname_component
Optimize check_refname_component using SSE2 on x86_64.
git rev-parse HEAD is a good test-case for this, since it does almost
nothing except parse refs. For one particular repo with about 60k
refs, almost all packed, the timings are:
Look up table: 29 ms
SSE2: 23 ms
This cuts about 20% off of the runtime.
Ondřej Bílka <neleai@seznam.cz> suggested an SSE2 approach to the
substring searches, which netted a speed boost over the SSE4.2 code I
had initially written.
Signed-off-by: David Turner <dturner@twitter.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Diffstat (limited to 't/valgrind')
| -rw-r--r-- | t/valgrind/default.supp | 8 |
1 files changed, 8 insertions, 0 deletions
diff --git a/t/valgrind/default.supp b/t/valgrind/default.supp index 0a6724f..332ab1a 100644 --- a/t/valgrind/default.supp +++ b/t/valgrind/default.supp @@ -49,3 +49,11 @@ Memcheck:Addr4 fun:copy_ref } +{ + ignore-sse-check_refname_format + Memcheck:Addr8 + fun:check_refname_format + fun:cmd_check_ref_format + fun:handle_builtin + fun:main +} |