summaryrefslogtreecommitdiff
path: root/docs/libc_assembler.md
diff options
context:
space:
mode:
authorChristopher Ferris <cferris@google.com>2018-08-06 12:18:32 -0700
committerChristopher Ferris <cferris@google.com>2018-08-06 12:18:32 -0700
commita473be239570817b065dc7acbc6968c3ad402133 (patch)
treee341a54d393bff19fa792d3a3b6eab0a185a2a78 /docs/libc_assembler.md
parent782cd01b098f86bef787955d5264e694d0e7e807 (diff)
Fix libc assembler documentation.
Test: NA Change-Id: I71485a3a44a56f2ff7ed590aec171464b57f56f6
Diffstat (limited to 'docs/libc_assembler.md')
-rw-r--r--docs/libc_assembler.md24
1 files changed, 20 insertions, 4 deletions
diff --git a/docs/libc_assembler.md b/docs/libc_assembler.md
index 151f2653b..44c00366d 100644
--- a/docs/libc_assembler.md
+++ b/docs/libc_assembler.md
@@ -18,15 +18,15 @@ there is no major difference in performance on each.
Benchmark 64 bit memcmp:
- /data/benchmarktest64/bionic-benchmarks/bionic-benchmarks --bionic_xml=string.xml memcmp
+ /data/benchmarktest64/bionic-benchmarks/bionic-benchmarks --bionic_xml=string.xml --benchmark_filter=memcmp
Benchmark 32 bit memcmp:
- /data/benchmarktest/bionic-benchmarks/bionic-benchmarks --bionic_xml=string.xml memcmp
+ /data/benchmarktest/bionic-benchmarks/bionic-benchmarks --bionic_xml=string.xml --benchmark_filter=memcmp
Locking to a specific cpu:
- /data/benchmarktest/bionic-benchmarks/bionic-benchmarks --bionic_cpu=2 --bionic_xml=string.xml memcmp
+ /data/benchmarktest/bionic-benchmarks/bionic-benchmarks --bionic_cpu=2 --bionic_xml=string.xml --benchmark_filter=memcmp
## Performance
The bionic benchmarks are used to verify the performance of changes to
@@ -56,7 +56,7 @@ Here is an example of how the benchmark should be executed. For this
command to work, you need to change directory to one of the above
directories.
- bionic-benchmarks --bionic_xml=suites/string.xml memcmp
+ bionic-benchmarks --bionic_xml=string.xml --benchmark_filter=memcmp
The last argument is the name of the one function that you want to
benchmark.
@@ -102,6 +102,22 @@ of issues, you can run a subset of the tests again. At the very least, it's
always a good idea to rerun the suite a couple of times to verify that
there isn't a high variation in the numbers.
+If you want to verify a single benchmark result, you can run a single test
+using a command like this:
+
+ bionic-benchmarks --bionic_xml=string.xml --benchmark_filter=BM_string_memcmp/1/1/0
+
+Where the argument to the filter argument is the name of the benchmark from
+the output. Sometimes this filter can still match multiple benchmarks, to
+guarantee that you only run the single benchmark, you can execute the benchmark
+like so:
+
+ bionic-benchmarks --bionic_xml=string.xml --benchmark_filter=BM_string_memcmp/1/1/0$
+
+NOTE: It is assumed that these commands are executed in adb as the shell user
+on device. If you are trying to run this using adb directly from a host
+machine, you might need to escape the special shell characters such as **$**.
+
## Testing
Run the bionic tests to verify that the new routines are valid. However,