diff options
| author | Mark Salyzyn <salyzyn@google.com> | 2016-12-28 10:30:57 -0800 |
|---|---|---|
| committer | Mark Salyzyn <salyzyn@google.com> | 2017-01-20 17:53:27 +0000 |
| commit | 142b43d4ab47f8f98748a9d560934a1f8f90c898 (patch) | |
| tree | 674e8a41ddce4873a233f54de88d5a168fd156bf /liblog | |
| parent | 8dbab358f01ed0957ff4d5b1c505981dfe93d5de (diff) | |
| download | core-142b43d4ab47f8f98748a9d560934a1f8f90c898.tar.gz core-142b43d4ab47f8f98748a9d560934a1f8f90c898.tar.bz2 core-142b43d4ab47f8f98748a9d560934a1f8f90c898.zip | |
liblog: BM_log_latency improvement
This reduces BM_log_latency from ~300ns to ~100ns. Because, well, we
read the clock earlier and that is what BM_log_latency tries to
measure. The one major improvement in the switch from kernel logger
to user space logger was that we picked up the timestamp in the
context of the caller before doing anything else, but alas changes
over time neglected this fact and placed isloggable checking ahead
of this important fact.
Test: liblog_benchmarks, check results
Change-Id: I4bc9fc3cf8b1659e88417d967b1d0f3743f9e456
Diffstat (limited to 'liblog')
| -rw-r--r-- | liblog/logger_write.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/liblog/logger_write.c b/liblog/logger_write.c index f19c3ab27..1a2d50612 100644 --- a/liblog/logger_write.c +++ b/liblog/logger_write.c @@ -262,6 +262,8 @@ static int __write_to_log_daemon(log_id_t log_id, struct iovec *vec, size_t nr) } #if defined(__ANDROID__) + clock_gettime(android_log_clockid(), &ts); + if (log_id == LOG_ID_SECURITY) { if (vec[0].iov_len < 4) { return -EINVAL; @@ -351,8 +353,6 @@ static int __write_to_log_daemon(log_id_t log_id, struct iovec *vec, size_t nr) return -EPERM; } } - - clock_gettime(android_log_clockid(), &ts); #else /* simulate clock_gettime(CLOCK_REALTIME, &ts); */ { |