Due to ADL lookup performed on the begin and end functions
of `for (auto _ : State)`, std::iterator_traits may get
incidentally instantiated. This patch ensures the library
can tolerate that.
* Don't include <sys/resource.h> on Fuchsia.
It doesn't support POSIX resource measurement and timing APIs.
Change-Id: Ifab4bac4296575f042c699db1ce5a4f7c2d82893
* Add BENCHMARK_OS_FUCHSIA for Fuchsia
Change-Id: Ic536f9625e413270285fbfd08471dcb6753ddad1
* Improve State packing: put important members on first cache line.
This patch does a few different things to ensure commonly accessed
data is on the first cache line of the `State` object.
First, it moves the `error_occurred_` member to reside after
the `started_` and `finished_` bools, since there was internal
padding there that was unused.
Second, it moves `batch_leftover_` and `max_iterations` further up
in the struct declaration. These variables are used in the calculation
of `iterations()` which users might call within the loop. Therefore
it's more important they exist on the first cache line.
Finally, this patch turns the bool members into bitfields. Although
this shouldn't have much of an effect currently, because padding is
still needed between the last bool and the first size_t, it should
help in future changes that require more "bool like" members.
* Remove bitfield change for now
* Move bools (and their padding) to end of "first cache line" vars.
I think it makes the most sense to move the padding required
following the group of bools to the end of the variables we want
on the first cache line.
This also means that the `total_iterations_` variable, which is the
most accessed, has the same address as the State object.
* Fix static assertion after moving bools
* Support State::KeepRunningBatch().
State::KeepRunning() can take large amounts of time relative to quick
operations (on the order of 1ns, depending on hardware). For such
sensitive operations, it is recommended to run batches of repeated
operations.
This commit simplifies handling of total_iterations_. Rather than
predecrementing such that total_iterations_ == 1 signals that
KeepRunning() should exit, total_iterations_ == 0 now signals the
intention for the benchmark to exit.
* Create better fast path in State::KeepRunningBatch()
* Replace int parameter with size_t to fix signed mismatch warnings
* Ensure benchmark State has been started even on error.
* Simplify KeepRunningBatch()
* Implement KeepRunning() in terms of KeepRunningBatch().
* Improve codegen by helping the compiler undestand dead code.
* Dummy commit for build bots' benefit.
* Attempt to fix travis timeouts during apt-get.
During some builds, travis fails to update the apt-get indexes.
This causes the build to fail in different ways.
This patch attempts to avoid this issue by manually calling
apt-get update. I'm not sure if it'll work, but it's worth a try.
* Fix missing semicolons in command
The appveyor bot sometimes fails because the time it
outputs is 6 digits long, but the output test regex expects at most
5 digits. This patch increases the size to 6 digits to placate the
test. This should not *really* affect the correctness of the test.
We're propagating extra warning flags to the gtest build, which
can cause it to fail. This patch prevents passing "-Wextra" to
gtest, since the library itself doesn't test with that flag.
* Support State::KeepRunningBatch().
State::KeepRunning() can take large amounts of time relative to quick
operations (on the order of 1ns, depending on hardware). For such
sensitive operations, it is recommended to run batches of repeated
operations.
This commit simplifies handling of total_iterations_. Rather than
predecrementing such that total_iterations_ == 1 signals that
KeepRunning() should exit, total_iterations_ == 0 now signals the
intention for the benchmark to exit.
* Create better fast path in State::KeepRunningBatch()
* Replace int parameter with size_t to fix signed mismatch warnings
* Ensure benchmark State has been started even on error.
* Simplify KeepRunningBatch()
When users satisfy the GTest dependancy by placing a googletest
directory in the project, the targets from GTest and GMock incorrectly
get installed along side this library. We shouldn't be installing
our test dependancies.
This patch forces the options that control installation for googletest
to OFF.
For people who get this library via CMake's AddExternalProject like me.
Would like a long term tutorial from someone who really understands CMake on how to actually link an externalproject's dependencies to another added external project.
* Add support for GTest based unit tests.
As Dominic and I have previously discussed, there is some
need/desire to improve the testing situation in Google Benchmark.
One step to fixing this problem is to make it easier to write
unit tests by adding support for GTest, which is what this patch does.
By default it looks for an installed version of GTest. However the
user can specify -DBENCHMARK_BUILD_EXTERNAL_GTEST=ON to instead
download, build, and use copy of gtest from source. This is
quite useful when Benchmark is being built in non-standard configurations,
such as against libc++ or in 32 bit mode.
This patch documents the newly added v2 branch, which will
be used to stage, test, and receive feedback on upcoming
features, most of which will be breaking changes which can't
be directly applied to master.
This patch primarily changes the BENCHMARK_UNREACHABLE()
implementation under MSVC to use __assume(false) instead
of being a NORETURN function, which ironically caused
unreachable code warnings.
Second, since the NOTHROW function attempt generated the
warnings we meant to avoid, it has been replaced with a dummy
null statement.
* Improve CPU Cache info reporting -- Add Windows support.
This patch does a couple of thing regarding CPU Cache reporting.
First, it adds an implementation on Windows. Second it fixes
the JSONReporter to correctly (and actually) output the CPU
configuration information.
And finally, third, it detects and reports the number of
physical CPU's that share the same cache.
* Refactor System information collection.
This patch refactors the system information collection,
and in particular information about the target CPU. The
motivation is to make it easier to access CPU information,
and easier to add new information as need be.
This patch additionally adds information about the cache
sizes of the CPU.
* Address review comments: Clean up integer types.
This commit cleans up the integer types used in ValueUnion to
follow the Google style guide.
Additionally it adds a BENCHMARK_UNREACHABLE macro to assist
in documenting/catching unreachable code paths.
* Rename ValueUnion accessors.
Define BENCHMARK_OS_NETBSD for NetBSD.
Add detection of cpuinfo_cycles_per_second and cpuinfo_num_cpus.
This code shared detection of these properties with FreeBSD.
* [Tools] A new, more versatile benchmark output compare tool
Sometimes, there is more than one implementation of some functionality.
And the obvious use-case is to benchmark them, which is better?
Currently, there is no easy way to compare the benchmarking results
in that case:
The obvious solution is to have multiple binaries, each one
containing/running one implementation. And each binary must use
exactly the same benchmark family name, which is super bad,
because now the binary name should contain all the info about
benchmark family...
What if i tell you that is not the solution?
What if we could avoid producing one binary per benchmark family,
with the same family name used in each binary,
but instead could keep all the related families in one binary,
with their proper names, AND still be able to compare them?
There are three modes of operation:
1. Just compare two benchmarks, what `compare_bench.py` did:
```
$ ../tools/compare.py benchmarks ./a.out ./a.out
RUNNING: ./a.out --benchmark_out=/tmp/tmprBT5nW
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:16:44
------------------------------------------------------
Benchmark Time CPU Iterations
------------------------------------------------------
BM_memcpy/8 36 ns 36 ns 19101577 211.669MB/s
BM_memcpy/64 76 ns 76 ns 9412571 800.199MB/s
BM_memcpy/512 84 ns 84 ns 8249070 5.64771GB/s
BM_memcpy/1024 116 ns 116 ns 6181763 8.19505GB/s
BM_memcpy/8192 643 ns 643 ns 1062855 11.8636GB/s
BM_copy/8 222 ns 222 ns 3137987 34.3772MB/s
BM_copy/64 1608 ns 1608 ns 432758 37.9501MB/s
BM_copy/512 12589 ns 12589 ns 54806 38.7867MB/s
BM_copy/1024 25169 ns 25169 ns 27713 38.8003MB/s
BM_copy/8192 201165 ns 201112 ns 3486 38.8466MB/s
RUNNING: ./a.out --benchmark_out=/tmp/tmpt1wwG_
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:16:53
------------------------------------------------------
Benchmark Time CPU Iterations
------------------------------------------------------
BM_memcpy/8 36 ns 36 ns 19397903 211.255MB/s
BM_memcpy/64 73 ns 73 ns 9691174 839.635MB/s
BM_memcpy/512 85 ns 85 ns 8312329 5.60101GB/s
BM_memcpy/1024 118 ns 118 ns 6438774 8.11608GB/s
BM_memcpy/8192 656 ns 656 ns 1068644 11.6277GB/s
BM_copy/8 223 ns 223 ns 3146977 34.2338MB/s
BM_copy/64 1611 ns 1611 ns 435340 37.8751MB/s
BM_copy/512 12622 ns 12622 ns 54818 38.6844MB/s
BM_copy/1024 25257 ns 25239 ns 27779 38.6927MB/s
BM_copy/8192 205013 ns 205010 ns 3479 38.108MB/s
Comparing ./a.out to ./a.out
Benchmark Time CPU Time Old Time New CPU Old CPU New
------------------------------------------------------------------------------------------------------
BM_memcpy/8 +0.0020 +0.0020 36 36 36 36
BM_memcpy/64 -0.0468 -0.0470 76 73 76 73
BM_memcpy/512 +0.0081 +0.0083 84 85 84 85
BM_memcpy/1024 +0.0098 +0.0097 116 118 116 118
BM_memcpy/8192 +0.0200 +0.0203 643 656 643 656
BM_copy/8 +0.0046 +0.0042 222 223 222 223
BM_copy/64 +0.0020 +0.0020 1608 1611 1608 1611
BM_copy/512 +0.0027 +0.0026 12589 12622 12589 12622
BM_copy/1024 +0.0035 +0.0028 25169 25257 25169 25239
BM_copy/8192 +0.0191 +0.0194 201165 205013 201112 205010
```
2. Compare two different filters of one benchmark:
(for simplicity, the benchmark is executed twice)
```
$ ../tools/compare.py filters ./a.out BM_memcpy BM_copy
RUNNING: ./a.out --benchmark_filter=BM_memcpy --benchmark_out=/tmp/tmpBWKk0k
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:37:28
------------------------------------------------------
Benchmark Time CPU Iterations
------------------------------------------------------
BM_memcpy/8 36 ns 36 ns 17891491 211.215MB/s
BM_memcpy/64 74 ns 74 ns 9400999 825.646MB/s
BM_memcpy/512 87 ns 87 ns 8027453 5.46126GB/s
BM_memcpy/1024 111 ns 111 ns 6116853 8.5648GB/s
BM_memcpy/8192 657 ns 656 ns 1064679 11.6247GB/s
RUNNING: ./a.out --benchmark_filter=BM_copy --benchmark_out=/tmp/tmpAvWcOM
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:37:33
----------------------------------------------------
Benchmark Time CPU Iterations
----------------------------------------------------
BM_copy/8 227 ns 227 ns 3038700 33.6264MB/s
BM_copy/64 1640 ns 1640 ns 426893 37.2154MB/s
BM_copy/512 12804 ns 12801 ns 55417 38.1444MB/s
BM_copy/1024 25409 ns 25407 ns 27516 38.4365MB/s
BM_copy/8192 202986 ns 202990 ns 3454 38.4871MB/s
Comparing BM_memcpy to BM_copy (from ./a.out)
Benchmark Time CPU Time Old Time New CPU Old CPU New
--------------------------------------------------------------------------------------------------------------------
[BM_memcpy vs. BM_copy]/8 +5.2829 +5.2812 36 227 36 227
[BM_memcpy vs. BM_copy]/64 +21.1719 +21.1856 74 1640 74 1640
[BM_memcpy vs. BM_copy]/512 +145.6487 +145.6097 87 12804 87 12801
[BM_memcpy vs. BM_copy]/1024 +227.1860 +227.1776 111 25409 111 25407
[BM_memcpy vs. BM_copy]/8192 +308.1664 +308.2898 657 202986 656 202990
```
3. Compare filter one from benchmark one to filter two from benchmark two:
(for simplicity, the benchmark is executed twice)
```
$ ../tools/compare.py benchmarksfiltered ./a.out BM_memcpy ./a.out BM_copy
RUNNING: ./a.out --benchmark_filter=BM_memcpy --benchmark_out=/tmp/tmp_FvbYg
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:38:27
------------------------------------------------------
Benchmark Time CPU Iterations
------------------------------------------------------
BM_memcpy/8 37 ns 37 ns 18953482 204.118MB/s
BM_memcpy/64 74 ns 74 ns 9206578 828.245MB/s
BM_memcpy/512 91 ns 91 ns 8086195 5.25476GB/s
BM_memcpy/1024 120 ns 120 ns 5804513 7.95662GB/s
BM_memcpy/8192 664 ns 664 ns 1028363 11.4948GB/s
RUNNING: ./a.out --benchmark_filter=BM_copy --benchmark_out=/tmp/tmpDfL5iE
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:38:32
----------------------------------------------------
Benchmark Time CPU Iterations
----------------------------------------------------
BM_copy/8 230 ns 230 ns 2985909 33.1161MB/s
BM_copy/64 1654 ns 1653 ns 419408 36.9137MB/s
BM_copy/512 13122 ns 13120 ns 53403 37.2156MB/s
BM_copy/1024 26679 ns 26666 ns 26575 36.6218MB/s
BM_copy/8192 215068 ns 215053 ns 3221 36.3283MB/s
Comparing BM_memcpy (from ./a.out) to BM_copy (from ./a.out)
Benchmark Time CPU Time Old Time New CPU Old CPU New
--------------------------------------------------------------------------------------------------------------------
[BM_memcpy vs. BM_copy]/8 +5.1649 +5.1637 37 230 37 230
[BM_memcpy vs. BM_copy]/64 +21.4352 +21.4374 74 1654 74 1653
[BM_memcpy vs. BM_copy]/512 +143.6022 +143.5865 91 13122 91 13120
[BM_memcpy vs. BM_copy]/1024 +221.5903 +221.4790 120 26679 120 26666
[BM_memcpy vs. BM_copy]/8192 +322.9059 +323.0096 664 215068 664 215053
```
* [Docs] Document tools/compare.py
* [docs] Document how the change is calculated
When stopping a timer, the current time is subtracted
from the start time. However, when the times are identical,
or sufficiently close together, the subtraction can result
in a negative number.
For some reason MinGW is the only platform where this problem
manifests. I suspect it's due to MinGW specific behavior in either
the CPU timing code, floating point model, or printf formatting.
Either way, the fix for MinGW should be correct across all platforms.
Describe how to use the cpupower command to disable CPU frequency scaling.
Document this, since there are other ways that don't see to have the same
effect. See #325
* Fix BM_SetInsert example
Move declaration of `std::set<int> data` outside the timing loop, so that the
destructor is not timed.
* Speed up BM_SetInsert test
Since the time taken to ConstructRandomSet() is so large compared to the time
to insert one element, but only the latter is used to determine number of
iterations, this benchmark now takes an extremely long time to run in
benchmark_test.
Speed it up two ways:
- Increase the Ranges() parameters
- Cache ConstructRandomSet() result (it's not random anyway), and do only
O(N) copy every iteration
* Fix same issue in BM_MapLookup test
* Make BM_SetInsert test consistent with README
- Use the same Ranges everywhere, but increase the 2nd range
- Change order of Args() calls in README to more closely match the result of Ranges
- Don't cache ConstructRandomSet, since it doesn't make sense in README
- Get a smaller optimization inside it, by givint a hint to insert()
Recently the library added a new ranged-for variant of the KeepRunning
loop that is much faster. For this reason it should be preferred in all
new code.
Because a library, its documentation, and its tests should all embody
the best practices of using the library, this patch changes all but a
few usages of KeepRunning() into for (auto _ : state).
The remaining usages in the tests and documentation persist only
to document and test behavior that is different between the two formulations.
Also note that because the range-for loop requires C++11, the KeepRunning
variant has not been deprecated at this time.
This patch improves the performance of the KeepRunning loop in two ways:
(A) it removes the dependency on the max_iterations variable, preventing
it from being loaded every iteration.
(B) it loops to zero, instead of to an upper bound. This allows a single
decrement instruction to be used instead of a arithmetic op followed by a
comparison.
* Add C++11 Ranged For loop alternative to KeepRunning
As pointed out by @astrelni and @dominichamon, the KeepRunning
loop requires a bunch of memory loads and stores every iterations,
which affects the measurements.
The main reason for these additional loads and stores is that the
State object is passed in by reference, making its contents externally
visible memory, and the compiler doesn't know it hasn't been changed
by non-visible code.
It's also possible the large size of the State struct is hindering
optimizations.
This patch allows the `State` object to be iterated over using
a range-based for loop. Example:
void BM_Foo(benchmark::State& state) {
for (auto _ : state) {
[...]
}
}
This formulation is much more efficient, because the variable counting
the loop index is stored in the iterator produced by `State::begin()`,
which itself is stored in function-local memory and therefore not accessible
by code outside of the function. Therefore the compiler knows the iterator
hasn't been changed every iteration.
This initial patch and idea was from Alex Strelnikov.
* Fix null pointer initialization in C++03
* Always use inline asm DoNotOptimize with clang.
clang-cl masquerades as MSVC but not GCC, so it was using the
MSVC-compatible definitions of DoNotOptimize and ClobberMemory.
Presumably, it's better in general to use the targeted assembly for
this functionality (the codegen is different), but the specific issue
is that clang-cl deprecates the usage of _ReadWriteBarrier, and this
gets rid of that warning.
* triggering another AppVeyor run