| Commit-Queue | +1 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| 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. |
This is a reland of commit 6aaad9dc9d6d602d9415612ca5199e9fba068b8bWhat the difference between this CL and that one?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
This is a reland of commit 6aaad9dc9d6d602d9415612ca5199e9fba068b8bWhat the difference between this CL and that one?
I found it myself with the following command. One good practice is to make PS1 the original code, PS2+ are the changes on top, and add a review request comment stating that for clarity.
`$ git diff change-8103530 change-8259429` ``` diff --git a/chrome/browser/custom_handlers/protocol_handler_registry_browsertest.cc b/chrome/browser/custom_handlers/protocol_handler_registry_browsertest.cc index b0e75ba59936d..07bf17a8ee87e 100644
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
This is a reland of commit 6aaad9dc9d6d602d9415612ca5199e9fba068b8bOh, it's exactly like that, PS2 are the changes that fixed the regression. The PS3 is just an edit in the description.
Thanks anyway.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Reland "Custom handlers: tie extension-registered handlers to their extension"
This is a reland of commit 6aaad9dc9d6d602d9415612ca5199e9fba068b8b
Original change's description:
> Custom handlers: tie extension-registered handlers to their extension
>
> A protocol handler at the kExtensionFeatures security level relaxes the
> HTML-spec restrictions (allowing cross-origin target URLs and the
> ext+/chrome-extension schemes). Extension cleanup -- uninstall/disable
> handling and ProtocolHandlersSanityCheck -- keys off the handler's
> extension_id, so a handler with the elevated level but no extension_id
> can never be removed: its privileges survive the registering extension's
> removal, and a build that predates recording the extension_id at
> registration time could have persisted such an orphan to prefs.
>
> In this CL we associate such handlers with the registering extension:
>
> - In Browser, centralize handler construction in a new
> CreateProtocolHandlerForFrame() helper. For a privileged extension
> frame it builds the handler with the owning extension's id.
>
> - In ProtocolHandlersManager::OnExtensionUnloaded, remove every handler
> returned by GetExtensionProtocolHandlers(extension->id()) instead of
> only the manifest-declared ones.
>
> Additionally, a new method ProtocolHandler::IsAllowedExtensionHandler()
> adds an extra security check which drops orphaned handlers on reload
> rather than re-registering them with their elevated privileges intact.
>
> Add a browser test, HandlerRemovedOnExtensionUninstall, that registers a
> handler from an extension page through the JS API and verifies it is
> associated with the extension and removed on uninstall.
>
> Add also a regression test covering the orphan rejection and the pref
> round-trip, and update the two scheme-validation tests that previously
> registered untagged kExtensionFeatures handlers to use realistic
> extension-tagged handlers.
>
> Bug: 511754574
> Change-Id: I404c00a7c69a06443c8ece8f75ebe9fbef065a48
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/8103530
> Reviewed-by: Ted Choc <ted...@chromium.org>
> Reviewed-by: Antonio Sartori <antonio...@chromium.org>
> Reviewed-by: Solomon Kinard <solomo...@chromium.org>
> Commit-Queue: Javier Fernandez <jfern...@igalia.com>
> Cr-Commit-Position: refs/heads/main@{#1678351}
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |