Giovanni Pezzino would like Devlin Cronin, David Yeung, Oleg Davydov, Chromium LUCI CQ, Simon Hangl, David Yeung and Rubber Stamper to review this change.
Reland "Force the update of corrupted extensions."
This reverts commit 1c94bad1e6507cb40eebd8e07c44b5f6afbfc230.
Reason for revert: The broken test is now fixed.
Original change's description:
> Revert "Force the update of corrupted extensions."
>
> This reverts commit ecdd40c9ab12b9df2ac1e09302ac95e7eb4627d5.
>
> Reason for revert: Caused a regression in SingleClientExtensionAppsSyncTest.InstallSomeApps
> See
> https://ci.chromium.org/ui/p/chromium/builders/ci/Linux%20MSan%20Tests/60857/test-results
>
> Original change's description:
> > Force the update of corrupted extensions.
> >
> > Rebooting a device with Refresh+Power results in the
> > computed_hashes.json file being truncated to 0 bytes length. The next
> > time the extension is loaded, the contents verification fails and the
> > extension is marked as corrupted and disabled. The extension update flow
> > will try and reinstall the extension, but it fails as the version of the
> > corrupted extension and the version of the to-be-installed extension
> > are the same. This CL modifies the update logic so that same version
> > updates are allowed when the update is caused by a detected corruption.
> >
> > Bug: 301307156, 395009770, 396458740
> > Change-Id: I116f53b79c6d94d98c7f2781c067cb6af285953c
> > Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/6830426
> > Reviewed-by: Oleg Davydov <buru...@chromium.org>
> > Auto-Submit: Giovanni Pezzino <gio...@google.com>
> > Reviewed-by: Devlin Cronin <rdevlin...@chromium.org>
> > Reviewed-by: Simon Hangl <sim...@google.com>
> > Commit-Queue: Devlin Cronin <rdevlin...@chromium.org>
> > Cr-Commit-Position: refs/heads/main@{#1563199}
>
> Bug: 301307156, 395009770, 396458740
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Change-Id: Ief248898fbfb3675be63e34b85b0a962d02caf9c
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7354774
> Bot-Commit: Rubber Stamper <rubber-...@appspot.gserviceaccount.com>
> Owners-Override: David Yeung <day...@google.com>
> Commit-Queue: David Yeung <day...@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#1563260}
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Bot-Commit | +1 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Gerrit GetPureRevert API does not mark this CL as a pure revert. Learn more: go/rubber-stamper-user-guide.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Devlin, PTAL.
Patchset 2 includes the new check in extension_protocols.cc to make the broken test pass. I could not use the IsShuttingDown() method as we discussed, as it returned false.
Thanks, Gio! The new check seems conceptually correct, but it looks like there are failing tests?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Hi Hidehiko,
could you PTAL at the change to the single_client_extension_apps_sync_test.cc file?
I have still some ongoing discussions with Devlin about the approach, so there is no ush for this 😊
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
extensions::SetShouldVerifyContentOverrideForTesting(false);could you document why this is needed?
static std::optional<bool> should_verify_content_override_for_testing =could you use anonymous namespace?