Bad usage of base::SysInfo::GetLsbReleaseBoard(). |
|||||||||
Issue descriptionGetLsbReleaseBoard() has confusing semantics and is insufficiently documented. On developer builds, it contains the board name, however in release builds, the board name is appended with signature information (eg. "-signed-mp-v2keys"). As devs develop and test with test builds they never see the appended signature information and as a consequence there's several places where special-casing of boards happens by string match against the board name which works on dev images but fails on production builds. There's wrong use at least in LockStateController::PreLockAnimationFinished() and DriveMetricsProvider::HasSeekPenalty(), I haven't looked in detail at the other call sites.
,
Nov 10 2016
We should add a warning to the description of GetLsbReleaseBoard() and provider a safer way to obtain the board name.
,
Nov 17 2016
,
Dec 1 2016
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/1a5a3f57210bc089a74d37ef6cc50f034d281c41 commit 1a5a3f57210bc089a74d37ef6cc50f034d281c41 Author: igorcov <igorcov@chromium.org> Date: Thu Dec 01 19:42:32 2016 Added function to provide the short board name of the device To be used in Chrome OS when separating functionality based on board. Fixed the bad usage of GetLsbReleaseBoard(). BUG=653814, 663857 Review-Url: https://codereview.chromium.org/2489853002 Cr-Commit-Position: refs/heads/master@{#435679} [modify] https://crrev.com/1a5a3f57210bc089a74d37ef6cc50f034d281c41/ash/wm/lock_state_controller.cc [modify] https://crrev.com/1a5a3f57210bc089a74d37ef6cc50f034d281c41/base/sys_info.h [modify] https://crrev.com/1a5a3f57210bc089a74d37ef6cc50f034d281c41/base/sys_info_chromeos.cc [modify] https://crrev.com/1a5a3f57210bc089a74d37ef6cc50f034d281c41/base/sys_info_unittest.cc [modify] https://crrev.com/1a5a3f57210bc089a74d37ef6cc50f034d281c41/components/metrics/drive_metrics_provider_linux.cc
,
Dec 2 2016
,
Mar 4 2017
,
Apr 17 2017
,
May 30 2017
,
Aug 1 2017
,
Oct 14 2017
|
|||||||||
►
Sign in to add a comment |
|||||||||
Comment 1 by tnagel@chromium.org
, Nov 9 2016