aboutsummaryrefslogtreecommitdiff
path: root/crosperf/results_report_templates.py
AgeCommit message (Collapse)Author
2017-04-08[toolchain-utils] Fix remaining lint errors in toolchain-utils.Caroline Tice
In addition to fixing the lint errors, this also fixes the Python formatting issues (ran tc_pyformat on nearly all the files). BUG=chromium:570450 TEST=Ran all crosperf & bisect tool unit tests. Ran afe_lock_machine.py (check machine status) Ran full crosperf test (octane, speedometer, BootPerf) on alex. Change-Id: Ic86f9192801ac67769f3de30f1c5f0d203ce0831 Reviewed-on: https://chromium-review.googlesource.com/471886 Commit-Ready: Caroline Tice <cmtice@chromium.org> Tested-by: Caroline Tice <cmtice@chromium.org> Reviewed-by: Manoj Gupta <manojgupta@chromium.org>
2016-09-08crosperf: Make results_report more general.George Burgess IV
The goal of this patch is to allow us to use results_report to generate non-ChromeOS-specific results reports. It's meant to be a nop for regular crosperf runs (except that, in HTML report generation, whitespace in the HTML itself will look a bit different.) Moreover, results_report used to shuffle Experiments around, each of which, unsurprisingly, contained *tons* of data about an experiment. So, part of this patch was reducing results_report's reliance on Experiments, in favor of a new type, BenchmarkResult. Some parts of results_report still rely on Experiments, but only to provide *extra* data. The minimum amount of data needed to make a results report is contained in a BenchmarkResult, and there's a convenient API we can use to make a BenchmarkResult out of an Experiment. This patch also does a massive refactor of results_report, because lots of the code was mildly more icky than it is with this patch applied. The refactor-for-prettiness and refactor-for-BenchmarkResults kind of go hand-in-hand, so it's really hard to split them out. The only part that's not so difficult to split out is the refactor to results_report_templates, but the value of splitting that out is questionable (to me). Speaking of which, all HTML magicks that were in results_report is now in results_report_templates, and now use *actual* Templates. It -- and HTMLReportGenerator -- are hopefully more readable as a result. Next, this makes JSONRsultsReport's GetReport() return a string containing JSON data, rather than calling a callback to write that string to a file. Finally, this includes a change to perf_table. Namely, its removal. It was otherwise unused, and I couldn't get it to even work, so I cleaned it up, made a perf report parser (because I couldn't get the cros_utils one to work, either...), and made sure it functions. If we're able to produce perf reports that cros_utils can parse, I'm happy to back my new parser out; I was primarily using it so I could be sure I didn't break HTML report generation. BUG=chromium:641098 TEST=./run_tests.sh passes Change-Id: I437de9eb39e00c9dd5c223ecd27feaaab544a6fd Reviewed-on: https://chrome-internal-review.googlesource.com/282217 Commit-Ready: George Burgess <gbiv@google.com> Tested-by: George Burgess <gbiv@google.com> Reviewed-by: Caroline Tice <cmtice@google.com>