For Developers‎ > ‎

bisect-builds.py

Have you ever hit a regression bug like this: "In chromium 26.0.401.1, things were broken. Back in chromium 26.0.201.0, it was fine."A good way to attack bugs like this – where it's unclear what change could have caused the regression, but where you have a reliable repro – is to bisect.

tools/bisect-builds.py automates downloading builds of Chrome across a regression range, conducting a binary search for the problematic change. If you don't have a chromium checkout, download bisect-builds.py here.

Run it like this:

python tools/bisect-builds.py -a platform -g good-revision -b bad-revision -- flags-for-chrome

For example,

python tools/bisect-builds.py -a mac -g 3894 -b 4000 -- --no-first-run --user-data-dir=/tmp http://example.com

Valid archive types (the -a parameter) are mac, win, linux, and linux64.

You can also use the -p option to specify a profile. If no -p or --user-data-dir option is specified, a new profile will be created in a temporary directory each time you are asked to try a build.

The script will download a build in the revision range and execute it. You must then manually check if the bug still repros. Quit Chromium, and the script will ask you if the bug reproduced or not. It will use your answer to drive a binary search, and after just a few steps it will tell you "this regression happened somewhere between revisions 1234 and 1334". From that list, it's usually easy to spot the offending CL. If not, you can use the run-bisect-manual-test.py script which will further bisect down to a particular CL by syncing and building manually.  If you're adding the range as a comment to a bug, please always paste the output from bisect-builds.py, as this includes links to the chromium and webkit changes in the regression range.

View SVN changes in revision range with this Useful URL (replacing SUCCESS_REV and FAILURE_REV with the range start and end):

If there's a Blink roll in the range, it will all let you know that and recommend you do a Blink bisect (see below.)

Getting an initial revision range

If you have two Chrome binaries, one which doesn't work, one which does, you can check the chrome://version page for the revision that it was built at (look for the "(Official Build NNNNN)" text). You can also infer the revision number from the version number. The third set of numbers in a version number (e.g. the 835 in 14.0.835.202) is the branch number. You can look up http://src.chromium.org/viewvc/chrome/branches/<branch_number>/src/ and see find messages of the form "Branching for <branch_number> @<revision_number>".

Alternatively, you can download an old revision from the continuous build (go to build.chromium.org and then click "continuous" in the upper left corner), verify it doesn't have the problem, and use that.


Use an elevated command prompt for use on Windows

The script currently requires an elevated command prompt to extract the build on Windows.
You can run cmd or cygwin as administrator to work around the silent failure. Track this issue at crbug.com/98739.

If the offending CL is a Blink Roll:

You can run the same exact command and pass in the -l option to bisect Blink instead of Chromium. Note that you're still passing the same good and bad Chromium ranges you'd passed in the above Chromium bisect. Nothing changes except that you add a -l option. So the above example would change to:

python tools/bisect-builds.py -l -a mac -g 3894 -b 4000 -- --no-first-run --user-data-dir=/tmp

This will look at Blink's archived builds and should give you a small Blink range. 

It should give you the Blink revision range where the regression occurred.

As an example, to isolate the regression range for crbug.com/331899 we used:

python bisect-builds.py -l -a mac -g 241430 -b 242816 -- --no-first-run --user-data-dir=/tmp http://codepen.io/callmevlad/pen/JxhvF

Argh, the offending CL is a  Skia / V8 roll!

This can be annoying, especially if it's a big roll.  If that doesn't help, then you can use the run-bisect-manual-test.py script, which will recurse into the Blink, V8 or Skia repositories if you give it a start and end revision range which includes a roll from one of these repositories

If Pepper Flash is required to repro

You will have to locate a Flash binary from an official build. If you suspect a Chromium change causing the regression and the Flash version doesn't matter locate any binary on your machine. For instance:

./bisect-builds.py -f /opt/google/chrome/PepperFlash/libpepflashplayer.so -b 232915 -g 230425 -a linux64

python bisect-builds.py -f "C:\Program Files (x86)\Google\Chrome\Application\31.0.1650.39\PepperFlash\pepflashplayer.dll" -b 232915 -g 230425 -a win

./bisect-builds.py -f "/Applications/Google Chrome.app/Contents/Versions/33.0.1707.0/Google Chrome Framework.framework/Internet Plug-Ins/PepperFlash/PepperFlashPlayer.plugin" -b 232915 -g 230425 -a mac
Comments