summaryrefslogtreecommitdiffstats
path: root/tools/perf/benchmarks
Commit message (Collapse)AuthorAgeFilesLines
* Convert octane test score to float firstalexanderk2015-08-111-1/+1
| | | | | | | | | | | to prevent fails on slow devices BUG=508907 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect Review URL: https://codereview.chromium.org/1230063007 Cr-Commit-Position: refs/heads/master@{#342772}
* Add a thread_times.key_noop_cases benchmarkjdduke2015-08-071-0/+20
| | | | | | | | | | | | | | There are a number of common interaction scenarios that should have few expensive side effects. This includes no-op touch handlers, no-op rAF, scrolling unnscrollable pages, tapping empty space, etc... Create such a benchmark with the mentioned scenarios, ensuring they remain cheap. BUG=490465 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1274473002 Cr-Commit-Position: refs/heads/master@{#342412}
* Disable memory.top_7_stress on Mac 10.10magjed2015-08-071-0/+1
| | | | | | | | | | BUG=517806 TBR=sullivan CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect Review URL: https://codereview.chromium.org/1271203004 Cr-Commit-Position: refs/heads/master@{#342352}
* Disabled gpu_times.top_25_smooth, gpu_times.gpu_rasterization.top_25_smooth, ↵sullivan2015-08-062-3/+4
| | | | | | | | | | | | and dromaeo.jslibmodifyjquery on Android due to failures BUG=517476,503138 TBR=qyearsley@chromium.org CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1277913002 Cr-Commit-Position: refs/heads/master@{#342133}
* perf: Add more tough ad casesskyostil2015-08-061-0/+11
| | | | | | | | | | | | | | | | | | | | | | | This patch adds a bunch of pages with challenging ads and analytics content to the smoothness.tough_ad_cases benchmark. This set includes real world web sites and a synthetic page with 5 Swiffy ads on it. These pages also try to use a more realistic interaction model: 1. Instead of waiting for the page to load fully, try scrolling roughly 1 second after loading. This helps to catch more jank during loading. 2. Scroll multiple times instead of just once. This invokes any Javascript touch handlers multiple times and helps to measure their contribution to jank. BUG=487937,510398,500962,496791,447731,395245,515517 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1252243003 Cr-Commit-Position: refs/heads/master@{#342118}
* [Memory Health Benchmark] Migrate to use the DumpMemory APIperezju2015-08-051-1/+8
| | | | | | | | | | | | | | | | | | | As http://crbug.com/505826 is now fixed, we can make use of the DumpMemory API to request individual dumps. Resolve many TODOs: use the new api and, in the memory timeline metric, return values from *all* requested dumps (instead of a single selected one). We also update the names of pages to more descriptive ones so, e.g., 'chrome_backround_5' will become 'after_http_cnn_com'. BUG=505826 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1271493002 Cr-Commit-Position: refs/heads/master@{#341886}
* perf: Stop running some no longer useful benchmarksskyostil2015-08-042-73/+0
| | | | | | | | | | | | | | | | | | | This patch removes two classes of benchmarks to speed up perf bot iteration: - The task execution time benchmark will be supplanted by data from Slow Reports. This patch still keeps the measurement code around since it could be applied to Slow Report data. - The sync scroll served their purpose during the initial implementation of the Blink scheduler. Now that we are improving other benchmarks to be more realistic and to better surface scheduling improvements, these somewhat synthetic benchmarks are no longer necessary. CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1263233002 Cr-Commit-Position: refs/heads/master@{#341754}
* Re-record memory.idle_multi_tab with new google sign-in and deterministic ↵ulan2015-08-041-2/+1
| | | | | | | | | | | | | counters. This also removes gplus page and reduces number of tabs. BUG=510440,490841 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1254173006 Cr-Commit-Position: refs/heads/master@{#341719}
* Revert of Enable start_with_ext benchmark on Linux platforms. (patchset #2 ↵andrewhayden2015-08-041-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | id:20001 of https://codereview.chromium.org/1270743005/ ) Reason for revert: This is causing the Linux perf bot to consistently time out on start_with_ext.warm.blank_page. See issue 516623 for more information. Original issue's description: > Enable start_with_ext benchmark on Linux platforms. > > BUG=513881 > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect > > Committed: https://crrev.com/3b9444d1d42eb518bfb199d6b622d0b9ff0b4acf > Cr-Commit-Position: refs/heads/master@{#341596} TBR=erikchen@chromium.org,nednguyen@google.com,sydli@google.com NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=513881 Review URL: https://codereview.chromium.org/1262053005 Cr-Commit-Position: refs/heads/master@{#341707}
* Revert of Create classes_util API, change discover to return a list instead ↵jochen2015-08-043-9/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | of a dict. (patchset #2 id:280001 of https://codereview.chromium.org/1262623007/ ) Reason for revert: fails reliably on XP tests: https://build.chromium.org/p/chromium.win/builders/XP%20Tests%20(1)/builds/39209/steps/telemetry_perf_unittests/logs/stdio Original issue's description: > Reland of Create classes_util API, change discover to return a list instead of a dict. (patchset #1 id:1 of https://codereview.chromium.org/1263063003/) > > Reason for revert: > The telemetry tests aren't broken. The perf_unittests are because they don't pick tests in a deterministic way. Reverting after a change to make them deterministic. > > Original issue's description: > > Revert of Create classes_util API, change discover to return a list instead of a dict. (patchset #5 id:120001 of https://codereview.chromium.org/1244223002/) > > > > Reason for revert: > > This appears to break the XP telemetry tests. Details on the bug. > > > > Original issue's description: > > > Create classes_util API, change discover to return a list instead of a dict. > > > > > > BUG=498968 > > > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect > > > > > > Committed: https://crrev.com/e6cbec4747aa7fe3c96b1cdb89de21ae77b30ab0 > > > Cr-Commit-Position: refs/heads/master@{#341129} > > > > TBR=dtu@chromium.org,bengr@chromium.org,aiolos@chromium.org > > NOPRESUBMIT=true > > NOTREECHECKS=true > > NOTRY=true > > BUG=498968 > > > > Committed: https://crrev.com/c85e3e1af7e170d5f8ec2b012df6e9337d2352bb > > Cr-Commit-Position: refs/heads/master@{#341218} > > TBR=dtu@chromium.org,bengr@chromium.org,avi@chromium.org > BUG=498968 > > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect > > Committed: https://crrev.com/7e6fefcf3b839256c97e2e7bc0bfc3c53af75c8d > Cr-Commit-Position: refs/heads/master@{#341646} TBR=dtu@chromium.org,bengr@chromium.org,avi@chromium.org,aiolos@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=498968 Review URL: https://codereview.chromium.org/1267243002 Cr-Commit-Position: refs/heads/master@{#341692}
* Reland of Create classes_util API, change discover to return a list instead ↵aiolos2015-08-033-10/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | of a dict. (patchset #1 id:1 of https://codereview.chromium.org/1263063003/) Reason for revert: The telemetry tests aren't broken. The perf_unittests are because they don't pick tests in a deterministic way. Reverting after a change to make them deterministic. Original issue's description: > Revert of Create classes_util API, change discover to return a list instead of a dict. (patchset #5 id:120001 of https://codereview.chromium.org/1244223002/) > > Reason for revert: > This appears to break the XP telemetry tests. Details on the bug. > > Original issue's description: > > Create classes_util API, change discover to return a list instead of a dict. > > > > BUG=498968 > > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect > > > > Committed: https://crrev.com/e6cbec4747aa7fe3c96b1cdb89de21ae77b30ab0 > > Cr-Commit-Position: refs/heads/master@{#341129} > > TBR=dtu@chromium.org,bengr@chromium.org,aiolos@chromium.org > NOPRESUBMIT=true > NOTREECHECKS=true > NOTRY=true > BUG=498968 > > Committed: https://crrev.com/c85e3e1af7e170d5f8ec2b012df6e9337d2352bb > Cr-Commit-Position: refs/heads/master@{#341218} TBR=dtu@chromium.org,bengr@chromium.org,avi@chromium.org BUG=498968 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1262623007 Cr-Commit-Position: refs/heads/master@{#341646}
* Revert of Create classes_util API, change discover to return a list instead ↵aiolos2015-08-033-9/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | of a dict. (patchset #1 id:1 of https://codereview.chromium.org/1262623007/) Reason for revert: Landed even though the patch failed. Original issue's description: > Reland of Create classes_util API, change discover to return a list instead of a dict. (patchset #1 id:1 of https://codereview.chromium.org/1263063003/) > > Reason for revert: > The telemetry tests aren't broken. The perf_unittests are because they don't pick tests in a deterministic way. Reverting after a change to make them deterministic. > > Original issue's description: > > Revert of Create classes_util API, change discover to return a list instead of a dict. (patchset #5 id:120001 of https://codereview.chromium.org/1244223002/) > > > > Reason for revert: > > This appears to break the XP telemetry tests. Details on the bug. > > > > Original issue's description: > > > Create classes_util API, change discover to return a list instead of a dict. > > > > > > BUG=498968 > > > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect > > > > > > Committed: https://crrev.com/e6cbec4747aa7fe3c96b1cdb89de21ae77b30ab0 > > > Cr-Commit-Position: refs/heads/master@{#341129} > > > > TBR=dtu@chromium.org,bengr@chromium.org,aiolos@chromium.org > > NOPRESUBMIT=true > > NOTREECHECKS=true > > NOTRY=true > > BUG=498968 > > > > Committed: https://crrev.com/c85e3e1af7e170d5f8ec2b012df6e9337d2352bb > > Cr-Commit-Position: refs/heads/master@{#341218} > > TBR=dtu@chromium.org,bengr@chromium.org,avi@chromium.org > BUG=498968 > > Committed: https://crrev.com/2c1531af216c026be0b4aed672dc89d264f95a72 > Cr-Commit-Position: refs/heads/master@{#341603} TBR=dtu@chromium.org,bengr@chromium.org,avi@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=498968 Review URL: https://codereview.chromium.org/1272433002 Cr-Commit-Position: refs/heads/master@{#341621}
* Reland of Create classes_util API, change discover to return a list instead ↵aiolos2015-08-033-10/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | of a dict. (patchset #1 id:1 of https://codereview.chromium.org/1263063003/) Reason for revert: The telemetry tests aren't broken. The perf_unittests are because they don't pick tests in a deterministic way. Reverting after a change to make them deterministic. Original issue's description: > Revert of Create classes_util API, change discover to return a list instead of a dict. (patchset #5 id:120001 of https://codereview.chromium.org/1244223002/) > > Reason for revert: > This appears to break the XP telemetry tests. Details on the bug. > > Original issue's description: > > Create classes_util API, change discover to return a list instead of a dict. > > > > BUG=498968 > > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect > > > > Committed: https://crrev.com/e6cbec4747aa7fe3c96b1cdb89de21ae77b30ab0 > > Cr-Commit-Position: refs/heads/master@{#341129} > > TBR=dtu@chromium.org,bengr@chromium.org,aiolos@chromium.org > NOPRESUBMIT=true > NOTREECHECKS=true > NOTRY=true > BUG=498968 > > Committed: https://crrev.com/c85e3e1af7e170d5f8ec2b012df6e9337d2352bb > Cr-Commit-Position: refs/heads/master@{#341218} TBR=dtu@chromium.org,bengr@chromium.org,avi@chromium.org BUG=498968 Review URL: https://codereview.chromium.org/1262623007 Cr-Commit-Position: refs/heads/master@{#341603}
* Enable start_with_ext benchmark on Linux platforms.sydli2015-08-031-4/+4
| | | | | | | | | BUG=513881 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1270743005 Cr-Commit-Position: refs/heads/master@{#341596}
* Add long running gmail memory benchmark for background tab.ulan2015-07-311-1/+12
| | | | | | | | | | | | This benchmark is for catching regressions in background GC scheduling, which differs from foreground GC scheduling. BUG=513149 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1266963003 Cr-Commit-Position: refs/heads/master@{#341369}
* telemetry: Increase page_set repeat for large profile tests.erikchen2015-07-311-2/+2
| | | | | | | | | BUG=442546 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1247423002 Cr-Commit-Position: refs/heads/master@{#341274}
* telemetry: Add tough image decode page set and smoothness benchmark.vmpstr2015-07-311-0/+8
| | | | | | | | | | | | | This patch adds a page set that has 20 large images that are scaled down. The action is to scroll down for a bit, then scroll up. The intent of the page is to measure improvements for image caching and image decoding that are planned in the near future. CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1184763010 Cr-Commit-Position: refs/heads/master@{#341260}
* Revert of Create classes_util API, change discover to return a list instead ↵avi2015-07-303-9/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | of a dict. (patchset #5 id:120001 of https://codereview.chromium.org/1244223002/) Reason for revert: This appears to break the XP telemetry tests. Details on the bug. Original issue's description: > Create classes_util API, change discover to return a list instead of a dict. > > BUG=498968 > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect > > Committed: https://crrev.com/e6cbec4747aa7fe3c96b1cdb89de21ae77b30ab0 > Cr-Commit-Position: refs/heads/master@{#341129} TBR=dtu@chromium.org,bengr@chromium.org,aiolos@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=498968 Review URL: https://codereview.chromium.org/1263063003 Cr-Commit-Position: refs/heads/master@{#341218}
* Create classes_util API, change discover to return a list instead of a dict.aiolos2015-07-303-10/+9
| | | | | | | | | BUG=498968 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1244223002 Cr-Commit-Position: refs/heads/master@{#341129}
* Revert of Disable rasterize_and_record_micro.key_mobile_sites_smooth on ↵simonhatch2015-07-301-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | linux. (patchset #1 id:1 of https://codereview.chromium.org/1259113005/) Reason for revert: CL causing timeouts was reverted in https://codereview.chromium.org/1263923002/. Original issue's description: > Disable rasterize_and_record_micro.key_mobile_sites_smooth on linux. > > This is timing out (and taking 1h 40 mins to do so). > > BUG=515094 > > Committed: https://crrev.com/4e2733b5808cf8e652540c74f2edbcf333f0523f > Cr-Commit-Position: refs/heads/master@{#340918} TBR=sullivan@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=515094 Review URL: https://codereview.chromium.org/1264873004 Cr-Commit-Position: refs/heads/master@{#341127}
* Adding telemetry tests for touch text selection performance.mfomitchev2015-07-302-0/+64
| | | | | | | | | | | | | | | | This CL adds a new page set |text_selection_sites|, which contains pages from top_10_mobile page set and peforms text selection and selection handle drag on the pages. There's two benchmarks: one for "direction" selection strategy, and another one for "character" selection strategy. BUG=497779 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect Review URL: https://codereview.chromium.org/1190423003 Cr-Commit-Position: refs/heads/master@{#341124}
* Disable rasterize_and_record_micro.key_mobile_sites_smooth on linux.simonhatch2015-07-291-1/+2
| | | | | | | | | | This is timing out (and taking 1h 40 mins to do so). BUG=515094 Review URL: https://codereview.chromium.org/1259113005 Cr-Commit-Position: refs/heads/master@{#340918}
* Enable out-of-process iframes benchmark using custom page set.Nasko Oskov2015-07-291-9/+13
| | | | | | | | | | | | | | This is a reland of the original CL from https://codereview.chromium.org/1250653002/ The difference is that now the benchmark is disabled for Android and the reference build, as we are fixing --site-per-process issues fairly quickly and the reference build is almost guaranteed not to be stable enough. BUG=462323,514742 R=aiolos@chromium.org Review URL: https://codereview.chromium.org/1261663003 . Cr-Commit-Position: refs/heads/master@{#340891}
* Revert of Enable out-of-process iframes benchmark using custom page set. ↵oysteine2015-07-281-12/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (patchset #2 id:20001 of https://codereview.chromium.org/1250653002/) Reason for revert: Failing on all perf bots. https://code.google.com/p/chromium/issues/detail?id=514742 Original issue's description: > Enable out-of-process iframes benchmark using custom page set. > > Since out-of-process iframes have different characteristics for memory > usage and page load times, we should have a separate basic page set to > have more control over performance measurements. > > This CL adds such a page set and re-enables the benchmark. It also adds > another benchmark that executes without out-of-process iframes enabled > with the same new page set, which will allow us to compare regular Chrome > vs out-of-process enabled Chrome. > > BUG=462323 > > Committed: https://crrev.com/f558eec2af3294c4fdb9bb8094567ad73edd1a84 > Cr-Commit-Position: refs/heads/master@{#340592} TBR=aiolos@chromium.org,sullivan@chromium.org,nasko@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=462323 Review URL: https://codereview.chromium.org/1262663002 Cr-Commit-Position: refs/heads/master@{#340739}
* Revert "Reland of Disable smoothness.top_25_smooth on win/mac/linux. ↵hpayer2015-07-281-1/+0
| | | | | | | | | | | | | | (patchset #1 id:1 of https://codereview.chromium.org/1245483002/)" This reverts commit 27caabde676d75984390e355ebe185bea501b020. BUG=chromium:510818 LOG=N CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect; Review URL: https://codereview.chromium.org/1260033002 Cr-Commit-Position: refs/heads/master@{#340671}
* Re-enable v8.top_25_smooth tests on Linux and Mac.hpayer2015-07-281-2/+1
| | | | | | | | | | BUG=chromium:510911 LOG=n CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect; Review URL: https://codereview.chromium.org/1252873012 Cr-Commit-Position: refs/heads/master@{#340670}
* Enable out-of-process iframes benchmark using custom page set.nasko2015-07-271-9/+12
| | | | | | | | | | | | | | | | | Since out-of-process iframes have different characteristics for memory usage and page load times, we should have a separate basic page set to have more control over performance measurements. This CL adds such a page set and re-enables the benchmark. It also adds another benchmark that executes without out-of-process iframes enabled with the same new page set, which will allow us to compare regular Chrome vs out-of-process enabled Chrome. BUG=462323 Review URL: https://codereview.chromium.org/1250653002 Cr-Commit-Position: refs/heads/master@{#340592}
* Disable v8.detached_context_age_in_gc on mac/win.simonhatch2015-07-271-0/+1
| | | | | | | | | BUG=514198 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1254253002 Cr-Commit-Position: refs/heads/master@{#340465}
* Blacklist top_7_stress for BenchmarkSmokeTestChristopher Cameron2015-07-241-1/+3
| | | | | | | | | | | | Test is timing out on the bots. BUG=513767 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect R=nednguyen@google.com Review URL: https://codereview.chromium.org/1259693002 . Cr-Commit-Position: refs/heads/master@{#340339}
* Disable dromaeo.jslibmodifyjquery on linux.bokan2015-07-241-1/+2
| | | | | | | | | | | | | | This test has been crashing on the Linux Perf 3 bot for a couple of runs now. BUG=513853 TBR=haraken@chromium.org NOTRY=true CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1252983004 Cr-Commit-Position: refs/heads/master@{#340336}
* Disable smoothness.image_decoding_cases.bokan2015-07-241-1/+2
| | | | | | | | | | | | | | | Disable smoothness.image_decoding_cases and smoothness.gpu_rasterization_and_decoding.image_decoding_cases on Android since they're crashing. BUG=513699 TBR=sugoi@chromium.org NOTRY=true CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1255883003 Cr-Commit-Position: refs/heads/master@{#340282}
* Disable http://m.taobao.com in page_cycler.top_10_mobilebokan2015-07-231-4/+7
| | | | | | | | | | | This page was the only one crashing on the page_cycler benchmark. I'm disabling the specific page and reenabling the page_cycler benchmark generally. BUG=509690 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1248413002 Cr-Commit-Position: refs/heads/master@{#340204}
* Add long running idle gmail memory benchmark.ulan2015-07-231-0/+11
| | | | | | | | | | | google_login.py is from http://crrev.com/1234453003 BUG=513149 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1228373010 Cr-Commit-Position: refs/heads/master@{#340099}
* Extension profile generator + benchmark for startup with profile.sydli2015-07-221-0/+56
| | | | | | | | | | | | | | Extension profile extender for profile generator and a new benchmark for startup time on a profile that loads extensions. Also for this purpose, a script to maintain a static set of extensions in cloud storage, and small additions to python gsutil wrapper (cloud_storage). BUG=444230 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect Review URL: https://codereview.chromium.org/1240703003 Cr-Commit-Position: refs/heads/master@{#339807}
* Re-enable session_restore on macsullivan2015-07-211-2/+2
| | | | | | | | | BUG=511273 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect Review URL: https://codereview.chromium.org/1250453005 Cr-Commit-Position: refs/heads/master@{#339702}
* Disable session_restore benchamrk on mac until they can be fixed.majidvp2015-07-201-4/+4
| | | | | | | | | BUG=511273 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1245723002 Cr-Commit-Position: refs/heads/master@{#339520}
* Reland of Disable smoothness.top_25_smooth on win/mac/linux. (patchset #1 ↵majidvp2015-07-171-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | id:1 of https://codereview.chromium.org/1245483002/) Reason for revert: Disabling the test smoothness.top_25_smooth on win/mac/linux again. The test is still causing perf bot failure. Original issue's description: > Revert of Disable smoothness.top_25_smooth on win/mac/linux. (patchset #1 id:1 of https://codereview.chromium.org/1238073003/) > > Reason for revert: > This was caused by https://codereview.chromium.org/1234223003, which has been reverted. > > BUG=510818 > > Original issue's description: > > Disable smoothness.top_25_smooth on win/mac/linux. > > > > BUG=510818 > > > > Committed: https://crrev.com/d143ef088ec1069825eb9a585d82be3b4ccc9ce6 > > Cr-Commit-Position: refs/heads/master@{#339041} > > TBR=sullivan@chromium.org > NOPRESUBMIT=true > NOTREECHECKS=true > NOTRY=true > BUG=510818 > > Committed: https://crrev.com/9e04ff6e861e62e2023c1f74fe1a479356e78e73 > Cr-Commit-Position: refs/heads/master@{#339259} TBR=sullivan@chromium.org,simonhatch@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=510818 Review URL: https://codereview.chromium.org/1241483003 Cr-Commit-Position: refs/heads/master@{#339306}
* Disable failing blob_storage.blob_storage and v8.top_25_smooth tests until ↵sullivan2015-07-172-3/+5
| | | | | | | | | | | they can be fixed. BUG=510815,510911 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1239493006 Cr-Commit-Position: refs/heads/master@{#339261}
* Revert of Disable smoothness.top_25_smooth on win/mac/linux. (patchset #1 ↵simonhatch2015-07-171-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | id:1 of https://codereview.chromium.org/1238073003/) Reason for revert: This was caused by https://codereview.chromium.org/1234223003, which has been reverted. BUG=510818 Original issue's description: > Disable smoothness.top_25_smooth on win/mac/linux. > > BUG=510818 > > Committed: https://crrev.com/d143ef088ec1069825eb9a585d82be3b4ccc9ce6 > Cr-Commit-Position: refs/heads/master@{#339041} TBR=sullivan@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=510818 Review URL: https://codereview.chromium.org/1245483002 Cr-Commit-Position: refs/heads/master@{#339259}
* Refactored TestExpectations out of Telemetry and into content/test/gpu.kbr2015-07-163-26/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (Re-land #2 with fixes for handling crashed renderers and other bug fixes) (Original CL: https://codereview.chromium.org/1224233005/ ) (Re-land #1: https://codereview.chromium.org/1231863008/ ) The GPU tests are the primary user of Telemetry's TestExpectations, and it turns out they can be implemented entirely in the code using Telemetry, rather than Telemetry itself. For this reason, move them into content/test/gpu. Refactored the GPU tests to all derive from a common set of base classes implementing support for test expectations. Verified that skipping tests, expected failures, and flaky test handling all work. More tests of test expectations are needed, and will be added in later CLs; this one is too large already, but can't be broken up further. Fixed up two WebGL perf tests that were relying on test expectations to skip the pages on specific GPUs, and added a hook to Telemetry's browser_info to enable this. This CL loses one valuable test, testHandlingOfCrashedTabWithExpectedFailure, from page_run_end_to_end_unittest.py. It will be reintroduced in a later CL. A subsequent CL will redesign TestExpectations. This CL is intended to be a refactoring with no behavioral changes. Ran telemetry_unittests, telemetry_gpu_unittests, telemetry_perf_unittests, and all GPU tests locally to verify. BUG=495870 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect TBR=bajones@chromium.org,nednguyen@google.com,dtu@chromium.org,zmo@chromium.org Review URL: https://codereview.chromium.org/1233133003 Cr-Commit-Position: refs/heads/master@{#339069}
* Disable smoothness.top_25_smooth on win/mac/linux.simonhatch2015-07-161-0/+1
| | | | | | | | BUG=510818 Review URL: https://codereview.chromium.org/1238073003 Cr-Commit-Position: refs/heads/master@{#339041}
* Remove inbox benchmark.nednguyen2015-07-151-19/+0
| | | | | | | | | | | | | This is making cros perf bot fails. Clearly, the credential file for cros doesn't exist anymore: $./depot_tools/gsutil.py ls gs://chromium-telemetry/01ff3ee8b3e06ade7f96bdef52b528422e71251e CommandException: No such object gs://chromium-telemetry/01ff3ee8b3e06ade7f96bdef52b528422e71251e Review URL: https://codereview.chromium.org/1232463004 Cr-Commit-Position: refs/heads/master@{#338939}
* Revert of Refactored TestExpectations out of Telemetry and into ↵tsergeant2015-07-153-6/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | content/test/gpu. (patchset #2 id:20001 of https://codereview.chromium.org/1231863008/) Reason for revert: This CL appears to be causing flakes on android_webview_webgl_conformance_tests, webgl_conformance_tests and gpu_rasterization_tests. See, for example, https://build.chromium.org/p/chromium.gpu/builders/Android%20Debug%20%28Nexus%209%29/builds/1079 https://build.chromium.org/p/chromium.gpu/builders/Android%20Debug%20%28Nexus%205%29/builds/975 Original issue's description: > Refactored TestExpectations out of Telemetry and into content/test/gpu. > > (Re-land of https://codereview.chromium.org/1224233005/ after fix for > blink_canvas.perf benchmark) > > The GPU tests are the primary user of Telemetry's TestExpectations, > and it turns out they can be implemented entirely in the code using > Telemetry, rather than Telemetry itself. For this reason, move them > into content/test/gpu. > > Refactored the GPU tests to all derive from a common set of base > classes implementing support for test expectations. Verified that > skipping tests, expected failures, and flaky test handling all work. > More tests of test expectations are needed, and will be added in later > CLs; this one is too large already, but can't be broken up further. > > Fixed up two WebGL perf tests that were relying on test expectations > to skip the pages on specific GPUs, and added a hook to Telemetry's > browser_info to enable this. > > This CL loses one valuable test, > testHandlingOfCrashedTabWithExpectedFailure, from > page_run_end_to_end_unittest.py. It will be reintroduced in a later CL. > > A subsequent CL will redesign TestExpectations. This CL is intended to > be a refactoring with no behavioral changes. > > Ran telemetry_unittests, telemetry_gpu_unittests, > telemetry_perf_unittests, and all GPU tests locally to verify. > > BUG=495870 > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect > TBR=bajones@chromium.org,nednguyen@google.com,dtu@chromium.org,zmo@chromium.org > > Committed: https://crrev.com/296319ccdb96f3f31df22f3b4ced4e040a864210 > Cr-Commit-Position: refs/heads/master@{#338768} TBR=zhenw@chromium.org,bajones@chromium.org,nednguyen@google.com,dtu@chromium.org,aiolos@chromium.org,zmo@chromium.org,kbr@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=495870 Review URL: https://codereview.chromium.org/1233303002 Cr-Commit-Position: refs/heads/master@{#338811}
* Refactored TestExpectations out of Telemetry and into content/test/gpu.kbr2015-07-143-26/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (Re-land of https://codereview.chromium.org/1224233005/ after fix for blink_canvas.perf benchmark) The GPU tests are the primary user of Telemetry's TestExpectations, and it turns out they can be implemented entirely in the code using Telemetry, rather than Telemetry itself. For this reason, move them into content/test/gpu. Refactored the GPU tests to all derive from a common set of base classes implementing support for test expectations. Verified that skipping tests, expected failures, and flaky test handling all work. More tests of test expectations are needed, and will be added in later CLs; this one is too large already, but can't be broken up further. Fixed up two WebGL perf tests that were relying on test expectations to skip the pages on specific GPUs, and added a hook to Telemetry's browser_info to enable this. This CL loses one valuable test, testHandlingOfCrashedTabWithExpectedFailure, from page_run_end_to_end_unittest.py. It will be reintroduced in a later CL. A subsequent CL will redesign TestExpectations. This CL is intended to be a refactoring with no behavioral changes. Ran telemetry_unittests, telemetry_gpu_unittests, telemetry_perf_unittests, and all GPU tests locally to verify. BUG=495870 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect TBR=bajones@chromium.org,nednguyen@google.com,dtu@chromium.org,zmo@chromium.org Review URL: https://codereview.chromium.org/1231863008 Cr-Commit-Position: refs/heads/master@{#338768}
* Re-enable blink_style.top_25 on Androideakuefner2015-07-141-1/+1
| | | | | | | | | | | | According to dstockwell@ blink_style.top_25 is working again. BUG=506322 R=nednguyen CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1241693002 Cr-Commit-Position: refs/heads/master@{#338763}
* [Telemetry Memory] Include usage by allocator in reported metricsperezju2015-07-141-1/+6
| | | | | | | | | | | | | | | | Include by default all memory metrics, including allocators, in the web_perf.metrics.memory_timeline. Individual benchmarks may filter out and only keep a subset if desired. Also make the memory_health_plan benchmark separate between "foreground" and "background" measurements. BUG=486009 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1235583002 Cr-Commit-Position: refs/heads/master@{#338672}
* Revert of Refactored TestExpectations out of Telemetry and into ↵alexclarke2015-07-142-0/+25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | content/test/gpu. (patchset #5 id:70001 of https://codereview.chromium.org/1224233005/) Reason for revert: I believe this patch broke the blink_perf.canvas perf test on Linux, Mac and Windows: http://build.chromium.org/p/chromium.perf/builders/Linux%20Perf%20%284%29/builds/8101/steps/blink_perf.canvas/logs/stdio http://build.chromium.org/p/chromium.perf/builders/Mac%2010.8%20Perf%20%284%29/builds/5636/steps/blink_perf.canvas/logs/stdio http://build.chromium.org/p/chromium.perf/builders/Win%207%20Nvidia%20GPU%20Perf%20%284%29/builds/815/steps/blink_perf.canvas/logs/stdio Original issue's description: > Refactored TestExpectations out of Telemetry and into content/test/gpu. > > The GPU tests are the primary user of Telemetry's TestExpectations, > and it turns out they can be implemented entirely in the code using > Telemetry, rather than Telemetry itself. For this reason, move them > into content/test/gpu. > > Refactored the GPU tests to all derive from a common set of base > classes implementing support for test expectations. Verified that > skipping tests, expected failures, and flaky test handling all work. > More tests of test expectations are needed, and will be added in later > CLs; this one is too large already, but can't be broken up further. > > Fixed up two WebGL perf tests that were relying on test expectations > to skip the pages on specific GPUs, and added a hook to Telemetry's > browser_info to enable this. > > This CL loses one valuable test, > testHandlingOfCrashedTabWithExpectedFailure, from > page_run_end_to_end_unittest.py. It will be reintroduced in a later CL. > > A subsequent CL will redesign TestExpectations. This CL is intended to > be a refactoring with no behavioral changes. > > Ran telemetry_unittests, telemetry_gpu_unittests, > telemetry_perf_unittests, and all GPU tests locally to verify. > > BUG=495870 > CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect > > Committed: https://crrev.com/5b3151cd707dee82b947ae096b12967e3c4a7c26 > Cr-Commit-Position: refs/heads/master@{#338623} TBR=aiolos@chromium.org,dtu@chromium.org,nednguyen@google.com,bajones@chromium.org,zmo@chromium.org,zhenw@chromium.org,kbr@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=495870 Review URL: https://codereview.chromium.org/1238653002 Cr-Commit-Position: refs/heads/master@{#338662}
* Refactored TestExpectations out of Telemetry and into content/test/gpu.kbr2015-07-142-25/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The GPU tests are the primary user of Telemetry's TestExpectations, and it turns out they can be implemented entirely in the code using Telemetry, rather than Telemetry itself. For this reason, move them into content/test/gpu. Refactored the GPU tests to all derive from a common set of base classes implementing support for test expectations. Verified that skipping tests, expected failures, and flaky test handling all work. More tests of test expectations are needed, and will be added in later CLs; this one is too large already, but can't be broken up further. Fixed up two WebGL perf tests that were relying on test expectations to skip the pages on specific GPUs, and added a hook to Telemetry's browser_info to enable this. This CL loses one valuable test, testHandlingOfCrashedTabWithExpectedFailure, from page_run_end_to_end_unittest.py. It will be reintroduced in a later CL. A subsequent CL will redesign TestExpectations. This CL is intended to be a refactoring with no behavioral changes. Ran telemetry_unittests, telemetry_gpu_unittests, telemetry_perf_unittests, and all GPU tests locally to verify. BUG=495870 CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect Review URL: https://codereview.chromium.org/1224233005 Cr-Commit-Position: refs/heads/master@{#338623}
* Disable page_cycler.top_10_mobile to make the perf bots go greenalexclarke2015-07-131-2/+2
| | | | | | | | | | | | | | Something is causing the page to timeout http://m.taobao.com/channel/act/mobile/20131111-women.html It's not immediatly obvious what is causing this so I'm disabling the test for now. BUG=509690 Review URL: https://codereview.chromium.org/1236913002 Cr-Commit-Position: refs/heads/master@{#338525}
* Remove doubled license header from blob_storage.pynasko2015-07-101-3/+0
| | | | | | | | | BUG=none CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect Review URL: https://codereview.chromium.org/1227913004 Cr-Commit-Position: refs/heads/master@{#338316}
generated by cgit v1.1 at 2025-03-31 08:24:10 +0000