9078011f93
[benchmark] Add autoreleasepool to objc bridging tests
...
We might be returning unowned objects as autorelease'd results.
I consciously do this in the test rather than in the driver because this makes
it easier to run the test in isolation.
2016-03-18 13:07:45 -07:00
4492e589e5
[benchmarks] ARC is not the default like I assumed
...
We are hence leaking ...
2016-03-18 12:01:49 -07:00
50c048d7e9
Add ObjectiveC bridging stub test support
...
This allows for testing the autogenerated bridging stub code.
2016-03-17 13:42:36 -07:00
9d9488078e
[gardening] Fix recently introduced comparisons to true: if (x == true) { … }
2016-03-16 23:51:42 +01:00
c9ef32c4d0
Move the objective-c bridging tests into the unit-tests folder
2016-03-16 15:44:28 -07:00
ff70c7016f
[gardening] Fix formatting of recently introduced header
2016-03-16 23:26:29 +01:00
5ea595a5d4
Add Objective-C bridging tests for NSString, NSArray, NSDictionary, NSSet
...
The tests test conditional and forced bridging from objective-c and bridging to
objective-c.
2016-03-16 15:16:45 -07:00
1edb62dc38
[Python] Make flake8 linting pass without errors/warning (w/ default rules)
2016-03-13 20:19:51 +01:00
ee3e7a58b3
[benchmark] Add versions of the set tests with class element types
2016-03-13 08:58:59 -07:00
191b3c87a1
[benchmark] Add RGBHistogramOfObjects
...
Another dictionary benchmark with objects as key and value types.
2016-03-13 08:45:30 -07:00
548ba33e1d
[benchmark] Add DictionarySwapOfObjects
...
Another dictionary benchmark with objects as key and value types.
2016-03-13 08:28:58 -07:00
4e43f745cd
[benchmark] Add DictionaryRemoveOfObjects
...
Another dictionary benchmark with objects as key and value types.
2016-03-13 08:23:36 -07:00
f6a1926c8c
[benchmark] Add Dictionary3OfObjects
...
Another dictionary benchmark with objects as key and value types.
2016-03-13 08:13:06 -07:00
577b0f2230
[benchmark] Add Dictionary2OfObjects
...
Another dictionary benchmark with objects as key and value types.
2016-03-13 08:03:41 -07:00
d021e8d5c9
[benchmark] Add DictionaryOfObjects benchmark
...
This tests performance of dictionary when both key and value are objects.
(We have fast paths otherwise because we know that the dictionary cannot be
backed by an objective c dictionary)
2016-03-13 08:03:11 -07:00
9c8a01c214
[Python] Remove unused methods get_{data,output,status}()
2016-03-12 20:21:16 +01:00
8bf6df8539
Merge pull request #1615 from practicalswift/remove-global-ignores-add-local-noqa
...
[Python] Replace global linting excludes with local line-level excludes ("noqa")
2016-03-11 14:42:15 +01:00
f89c70135b
Merge pull request #1624 from shahmishal/master
...
[Perf script] Fix output format for compare script
2016-03-10 11:56:55 -08:00
fcbea8bd90
[Perf script] Fix output format for compare script
2016-03-10 11:54:27 -08:00
d5326bfdc4
[Python] Replace global linting excludes with local line-level excludes ("noqa")
...
Replace the project global linting rule excludes (as defined in .pep8) with
fine-grained "# noqa" annotations.
By using noqa annotation the excludes are made on a per line basis instead of
globally.
These annotations are used where we make deliberate deviations from the standard
linting rules.
To lint the Python code in the project:
$ flake8
To install flake8:
$ pip install flake8
See https://flake8.readthedocs.org/en/latest/ for details.
To enable checking of the PEP-8 naming conventions, install the optional
extension pep8-naming:
$ pip install pep8-naming
To enable checking of blind "except:" statements, install the optional
extension flake8-blind-except:
$ pip install flake8-blind-except
To enable checking of import statement order, install the optional
extension flake8-import-order:
$ pip install flake8-import-order
2016-03-10 16:22:48 +01:00
a1156e1854
[Python] Fix recently introduced linting (flake8) regression.
2016-03-10 13:51:17 +01:00
42b7e3f8d6
Merge pull request #1598 from practicalswift/fix-80-column-violations-in-python-code
...
[Python] Fix 80 column violations
2016-03-10 10:35:25 +01:00
02006f20bc
Merge remote-tracking branch 'origin/master' into swift-3-api-guidelines
2016-03-09 16:05:03 -08:00
0796eaad1f
[Python] Fix 80-column violations
2016-03-09 23:52:11 +01:00
046f0636b2
Fix compare_perf_tests.py to support python 3
2016-03-09 14:02:16 -08:00
9c998d7ce8
[leaks-runner] Get more accurate results by subtracting two differing iteration counts.
...
Previously, we had some white listing lists for allocations that were assigned
into globals during a benchmarks running. Now we instead run two different
iterations with the second running an additional time. Then we subtract the
counts.
This enables me to get rid of the whitelist.
2016-03-09 12:56:15 -08:00
ccc8791581
[leaks-checker] Add support for filtering tests and changing the number of samples/iters run. This will make it easy to bisect using the Benchmark_LeaksRunner script.
2016-03-09 10:28:00 -08:00
e9ac8c876a
Add a performance unit test for stack promotion
2016-03-09 07:30:00 -08:00
e232456bf0
[leaks-runner] Print out the RC delta with the results.
2016-03-09 00:37:14 -08:00
30b66ea036
Merge pull request #1584 from practicalswift/python-3-compatible-print
...
[Python] Use Py3k compatible print operator: print "foo" → print("foo")
2016-03-09 08:00:49 +01:00
0645b30caf
Fix deadlock in leaks runner script.
2016-03-08 22:18:36 -08:00
c9000af795
Merge pull request #1526 from practicalswift/fix-pep8-violations-ii
...
[Python] Fix five classes of PEP-8 violations (E101/E111/E128/E302/W191)
2016-03-08 23:55:46 -05:00
0fd0c48648
[Python] Use Py3k compatible print operator: print "foo" → print("foo")
2016-03-08 23:10:52 +01:00
1fae0d1325
Merge remote-tracking branch 'origin/master' into swift-3-api-guidelines
2016-03-08 12:48:48 -08:00
e2de5c2202
[Python] Fix blind except: statements
...
Make sure all Python code in the repo specifies which exceptions to
catch when using `except:`.
Regressions can be catched using `flake8-blind-except` going forward.
2016-03-08 11:08:28 +01:00
265835fdfc
[Python] Use consistent import ordering for Python code
...
Ordering used:
1.) standard library imports
2.) third party imports
3.) local package imports
Each group is individually alphabetized.
2016-03-07 23:25:16 +01:00
183da818df
[Python] Fix five classes of PEP-8 violations (E101/E111/E128/E302/W191)
...
* E101: indentation contains mixed spaces and tabs
* E111: indentation is not a multiple of four
* E128: continuation line under-indented for visual indent
* E302: expected 2 blank lines, found 1
* W191: indentation contains tabs
2016-03-07 22:36:23 +01:00
7fe6916bf6
Merge remote-tracking branch 'origin/master' into swift-3-api-guidelines
2016-03-07 12:10:47 -08:00
5b2800225d
[gardening] Fix recently introduced typos
...
Fixed:
* perfomance → performance
* requirments → requirements
* satisified → satisfied
* template template → template
2016-03-04 11:10:50 +01:00
d197cc5022
Merge pull request #1531 from practicalswift/fix-headers
...
[gardening] Fix recently introduced file headers. Also: "–" → "--".
2016-03-03 16:18:10 -08:00
47888558be
[perftests] Fix thinko due to my brains failure to remember to push a new version of this commit.
2016-03-03 15:51:03 -08:00
af9d4342eb
Merge pull request #1529 from gottesmm/sanity-check-perftest-statistics
...
[perftests] Add some sanity asserts to the low level perf test suite …
2016-03-03 15:49:01 -08:00
b9dd782e4b
[gardening] Fix recently introduced file header. Also: "–" → "--".
2016-03-03 23:10:43 +01:00
d921b4e1bc
[benchmark][SR-871] Convert duration strings to integers before calculating min/max
...
Benchmark_Driver used string comparison to calculate the minimum and maximum
of durations in benchmark results, sometimes leading to wildly inaccurate reports.
2016-03-03 22:34:54 +01:00
cf4bafe9e3
Merge remote-tracking branch 'origin/master' into swift-3-api-guidelines
2016-03-03 13:22:03 -08:00
afece1c8c2
[perftests] Add some sanity asserts to the low level perf test suite driver results.
...
This just consists of ensuring that:
1. min <= max.
2. min <= mean.
3. min <= median.
4. max >= mean.
5. max >= median.
2016-03-03 13:11:33 -08:00
3d58bc0d82
Adjust UTF8Decode benchmark to run in <1s
...
Under Debug -Onone
2016-03-03 18:17:09 +11:00
e05fc61533
Merge pull request #1493 from PatrickPijnappel/utf8-benchmark
...
Add benchmark for UTF-8 decoding
2016-03-02 14:44:12 -08:00
d37ddcf8bd
Merge pull request #1514 from PatrickPijnappel/patch-1
...
Clarify usage of N in benchmark README
2016-03-02 14:41:48 -08:00
4754285879
Clarify usage of N in benchmark README
2016-03-03 08:55:21 +11:00
Arnold Schwaighofer
practicalswift
Mishal Shah
Mishal Shah
Max Moiseev
Michael Gottesman
Brian Gesiak
Chris Lattner
Michael Gottesman
Karoly Lorentey
Patrick Pijnappel
lplarson