glibc/benchtests
Siddhesh Poyarekar 9298ecba15 Accept output arguments to benchmark functions
This patch adds the ability to accept output arguments to functions
being benchmarked, by nesting the argument type in <> in the args
directive.  It includes the sincos implementation as an example, where
the function would have the following args directive:

  ## args: double:<double *>:<double *>

This simply adds a definition for a static variable whose pointer gets
passed into the function, so it's not yet possible to pass something
more complicated like a pre-allocated string or array.  That would be
a good feature to add if a function needs it.

The values in the input file will map only to the input arguments.  So
if I had a directive like this for a function foo:

  ## args: int:<int *>:int:<int *>

and I have a value list like this:

1, 2
3, 4
5, 6

then the function calls generated would be:

foo (1, &out1, 2, &out2);
foo (3, &out1, 4, &out2);
foo (5, &out1, 6, &out2);
2013-12-05 10:12:59 +05:30
..
acos-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
acosh-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
asin-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
asinh-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
atan-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
atanh-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
bench-bcopy.c Copy over string performance tests into benchtests 2013-06-11 15:08:13 +05:30
bench-bzero.c Copy over string performance tests into benchtests 2013-06-11 15:08:13 +05:30
bench-memccpy.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-memchr.c benchtests: Add memrchr benchmark 2013-09-06 09:24:52 -03:00
bench-memcmp.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-memcpy.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-memmem.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-memmove.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-mempcpy.c Copy over string performance tests into benchtests 2013-06-11 15:08:13 +05:30
bench-memrchr.c benchtests: Add memrchr benchmark 2013-09-06 09:24:52 -03:00
bench-memset.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-modf.c Prevent optimizing out of benchmark function call 2013-05-17 19:10:33 +05:30
bench-rawmemchr.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-skeleton.c benchtests: Rename argument to TIMING_INIT macro. 2013-09-11 15:18:20 +01:00
bench-stpcpy_chk.c Port remaining string benchmarks 2013-06-11 20:51:55 +05:30
bench-stpcpy.c Copy over string performance tests into benchtests 2013-06-11 15:08:13 +05:30
bench-stpncpy.c Copy over string performance tests into benchtests 2013-06-11 15:08:13 +05:30
bench-strcasecmp.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strcasestr.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strcat.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strchr.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strchrnul.c Copy over string performance tests into benchtests 2013-06-11 15:08:13 +05:30
bench-strcmp.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strcpy_chk.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strcpy.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strcspn.c Copy over string performance tests into benchtests 2013-06-11 15:08:13 +05:30
bench-string.h Remove duplicate ifunc benchtests. 2013-11-26 12:48:33 +01:00
bench-strlen.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strncasecmp.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strncat.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strncmp.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strncpy.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strnlen.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strpbrk.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strrchr.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strsep.c benchtests: Add strsep benchmark 2013-11-18 06:49:44 -06:00
bench-strspn.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strstr.c benchtests: Switch string benchmarks to use bench-timing.h. 2013-09-04 15:40:12 +01:00
bench-strtod.c benchtests: Add strtod benchmark 2013-11-11 11:24:07 -02:00
bench-timing.h 2013-11-13 Steve Ellcey <sellcey@mips.com> 2013-11-13 08:48:25 -08:00
cos-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
cosh-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
exp-inputs New inputs for exp 2013-10-28 16:35:08 +05:30
log-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
Makefile Benchmark test for sqrt function. 2013-12-02 09:37:18 -08:00
pow-inputs Benchmark inputs for pow 2013-10-28 16:36:46 +05:30
README Accept output arguments to benchmark functions 2013-12-05 10:12:59 +05:30
rint-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
sin-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
sincos-inputs Accept output arguments to benchmark functions 2013-12-05 10:12:59 +05:30
sinh-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
sqrt-inputs Benchmark test for sqrt function. 2013-12-02 09:37:18 -08:00
tan-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30
tanh-inputs Add more directives to benchmark input files 2013-10-07 11:51:25 +05:30

Using the glibc microbenchmark suite
====================================

The glibc microbenchmark suite automatically generates code for specified
functions, builds and calls them repeatedly for given inputs to give some
basic performance properties of the function.

Running the benchmark:
=====================

The benchmark can be executed by invoking make as follows:

  $ make bench

This runs each function for 10 seconds and appends its output to
benchtests/bench.out.  To ensure that the tests are rebuilt, one could run:

  $ make bench-clean

The duration of each test can be configured setting the BENCH_DURATION variable
in the call to make.  One should run `make bench-clean' before changing
BENCH_DURATION.

  $ make BENCH_DURATION=1 bench

The benchmark suite does function call measurements using architecture-specific
high precision timing instructions whenever available.  When such support is
not available, it uses clock_gettime (CLOCK_PROCESS_CPUTIME_ID).  One can force
the benchmark to use clock_gettime by invoking make as follows:

  $ make USE_CLOCK_GETTIME=1 bench

Again, one must run `make bench-clean' before changing the measurement method.

Adding a function to benchtests:
===============================

If the name of the function is `foo', then the following procedure should allow
one to add `foo' to the bench tests:

- Append the function name to the bench variable in the Makefile.

- Make a file called `foo-inputs` to provide the definition and input for the
  function.  The file should have some directives telling the parser script
  about the function and then one input per line.  Directives are lines that
  have a special meaning for the parser and they begin with two hashes '##'.
  The following directives are recognized:

  - args: This should be assigned a colon separated list of types of the input
    arguments.  This directive may be skipped if the function does not take any
    inputs.  One may identify output arguments by nesting them in <>.  The
    generator will create variables to get outputs from the calling function.
  - ret: This should be assigned the type that the function returns.  This
    directive may be skipped if the function does not return a value.
  - includes: This should be assigned a comma-separated list of headers that
    need to be included to provide declarations for the function and types it
    may need (specifically, this includes using "#include <header>").
  - include-sources: This should be assigned a comma-separated list of source
    files that need to be included to provide definitions of global variables
    and functions (specifically, this includes using "#include "source").
  - name: See following section for instructions on how to use this directive.

  Lines beginning with a single hash '#' are treated as comments.  See
  pow-inputs for an example of an input file.

Multiple execution units per function:
=====================================

Some functions have distinct performance characteristics for different input
domains and it may be necessary to measure those separately.  For example, some
math functions perform computations at different levels of precision (64-bit vs
240-bit vs 768-bit) and mixing them does not give a very useful picture of the
performance of these functions.  One could separate inputs for these domains in
the same file by using the `name' directive that looks something like this:

  ##name: 240bit

See the pow-inputs file for an example of what such a partitioned input file
would look like.

Benchmark Sets:
==============

In addition to standard benchmarking of functions, one may also generate
custom outputs for a set of functions.  This is currently used by string
function benchmarks where the aim is to compare performance between
implementations at various alignments and for various sizes.

To add a benchset for `foo':

- Add `foo' to the benchset variable.
- Write your bench-foo.c that prints out the measurements to stdout.
- On execution, a bench-foo.out is created in $(objpfx) with the contents of
  stdout.