| Age | Commit message (Collapse) | Author |
|---|
|
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
|
|
* add an integer revision field
* add an extra information string field
* zlib benchmark: new revision[1] is 2, so that old results are obviously old,
and now the zlib version string will be stored in extra information.
[1] as of 6a8e19a14305079b03e45eeb0580a45104f300dd
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
While the shell isn't rewritten to take advantage of GTK+3 only, let's
just ignore these deprecation warnings.
|
|
As discussed in
https://github.com/lpereira/hardinfo/issues/345
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
For generated reports, the benchmark results are now sorted.
Also, only the (10) nearest results are included to avoid
output clutter. Perhaps the number could be set in a
hardinfo parameter.
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
Build the configuration file path using g_get_user_config_dir() instead of g_get_home_dir().
|
|
If you just need the hardware report, this makes it so much
faster.
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
If (end - start) / threads left a remainder, benchmark_parallel_for()
would start an extra thread for the leftover elements. Now, it has the
last thread process the few extra elements.
Added a note:
benchmark_parallel_for(): element [start] included, but [end] is excluded.
callback(): expected to processes elements [start] through [end] inclusive.
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
The idea here is to make adding/removing benchmark tests easier by
having benchmark.c be the framework and all the functioned needed
for individual benchmarks live in benches.c.
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
New benchmark types:
* benchmark_crunch_for(): call function repeatedly for a number
of seconds; result is number of completions.
* benchmark_parallel(): one call for each available thread
up to n_threads; result is sum of return values.
Tweaks:
* Store return values from callback via
benchmark_parallel_for_dispatcher()
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
For running benchmarks from the command line, add a new param
to specify the output format:
-g "conf" gives a line suitable for benchmark.conf
-g "shell" gives the complete "moreinfo" shell data
-g "short" (or nothing) gives the bench_value string as before
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
Benchmark results store actual number of threads used by benchmark
when it was run. Previously, results assumed all available threads
were used.
Examples:
* CPU Fib only uses one
* FPU FFT uses 4, 2, or 1
* N-Queens uses 10, 5, 2, or 1
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
|
|
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
* More information can be saved in benchmark.conf, while still being
compatible with older versions.
* Selecting a result give additional information about the result
and the machine the produced it.
|
|
Current CPU configurations aren't properly represented in Hardinfo.
For SMT, each hardware thread is still reported as a CPU. Clusters
with different CPU clock rates are not reported. It is common for
ARM to pair a cluster of fast cores with a cluster of slower, but
more power-efficient cores. These changes attempt to address this.
The getProcessorFrequency method now returns the processor's max
frequency for all its cores. While the new
getProcessorFrequencyDesc lists each unique frequency with a prefix
Nx with the count of cores at that freqency.
Benchmark results have been reformated to use the de-prefixed
getProcessorName and getProcessorFrequencyDesc.
As an example from benchmark.conf:
4x AMD Phenom(tm) II X4 940 Processor | 800 MHz
becomes:
AMD Phenom(tm) II X4 940 Processor | 4x 800 MHz
Which, I think, makes much more sense, as it works well with
this kind of thing:
Qualcomm Snapdragon 691 | 4x 1400 MHz + 4x 800 MHz
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
* More easily translated.
* The strings are more likely to be re-usable.
* A small change to one, or to a shell param, does not lose all
of them.
Signed-off-by: Burt P <pburt0@gmail.com>
|
|
|
|
|
|
There was compile warning in benchmark.c:
Scanning dependencies of target benchmark
[ 48%] Building C object CMakeFiles/benchmark.dir/modules/benchmark.c.o
/home/maxpayne/hardinfo/modules/benchmark.c: In function ‘get_benchmark_results’:
/home/maxpayne/hardinfo/modules/benchmark.c:565:37: warning: format ‘%d’ expects argument of type ‘int’, but argument 5 has type ‘long unsigned int’ [-Wformat=]
gchar *result = g_strdup_printf("[param]\n"
^
[ 50%] Building C object CMakeFiles/benchmark.dir/modules/benchmark/blowfish.c.o
|
|
|
|
|
|
|
|
gcc was (correctly) assuming that benchmark_parallel_for() returned
an integer, which was being to a double and messing up results. Added
the missing prototype.
Also, populate bench_results array with sane values on module init.
Some other minor cleanups in benchmark code.
|
|
|
|
|
|
|
|
|
