Datasets:

---

jag2023

/

10K_bug_reports

like 0

When using both the `no_rdm` and `reverse_split` properties in a spotlight screen with split layout, when responsive mode is entered at 800px the spotlight remains in it's full size (as intended), but the layout is un-reversed. This seems to be due to media queries throughout our CSS not appropriately handling the `no-rdm` property appropriately ([example](https://searchfox.org/mozilla-central/rev/8c3ca2f5a74e0ba59c3d9dddf5468a2ffab13467/browser/components/aboutwelcome/content-src/aboutwelcome.scss#827-837)). We should go through the css and add no-rdm negations to each of these media queries. Alternatively, we could bypass the need for `no-rdm` negations in each of these media queries by leveraging something like a [MediaQueryList change event listener](https://developer.mozilla.org/en-US/docs/Web/API/MediaQueryList/change_event), and applying an attribute or class in the case of the `no_rdm` property being present in content, and ridding the css of media query no-rdm negations.

Created attachment 9397896 Screenshot 2024-04-22 at 10.02.54 AM.png When both a title and subtitle aren't included in a screen's content in the split layout for spotlight, the element that contains them is still rendered, which makes the flex layout behave strangely when there is no title or subtitle.

In order to address a bug with auto light/dark theming on MacOS we are creating light/dark counterpoints for all wallpapers. The thumbnails for onboarding are in-tree so we need to update those with the new images.

Created attachment 9370190 poc.html #Reproduce OS:Win10 X64 123.0a1 (2023-12-19) (64-bit) #Step: 1. python -m http.server 1337 2. python -m ffpuppet firefox.exe -p prefs.js -d -u http://localhost:1337/poc.html #Analysis missing boundary checks for `ticks + start, detune + start, end - start` https://searchfox.org/mozilla-central/source/dom/media/webaudio/OscillatorNode.cpp#300 ``` void ProcessBlock(AudioNodeTrack* aTrack, GraphTime aFrom, const AudioBlock& aInput, AudioBlock* aOutput, bool* aFinished) override { MOZ_ASSERT(mSource == aTrack, "Invalid source track"); TRACE("OscillatorNodeEngine::ProcessBlock"); TrackTime ticks = mDestination->GraphTimeToTrackTime(aFrom); if (mStart == -1) { ComputeSilence(aOutput); return; } if (ticks + WEBAUDIO_BLOCK_SIZE <= mStart || ticks >= mStop) { ComputeSilence(aOutput); } else { aOutput->AllocateChannels(1); float* output = aOutput->ChannelFloatsForWrite(0); uint32_t start, end; FillBounds(output, ticks, start, end); float frequency[WEBAUDIO_BLOCK_SIZE]; float detune[WEBAUDIO_BLOCK_SIZE]; if (ParametersMayNeedUpdate()) { if (mFrequency.HasSimpleValue()) { std::fill_n(frequency, WEBAUDIO_BLOCK_SIZE, mFrequency.GetValue()); } else { mFrequency.GetValuesAtTime(ticks + start, frequency + start, end - start); } if (mDetune.HasSimpleValue()) { std::fill_n(detune, WEBAUDIO_BLOCK_SIZE, mDetune.GetValue()); } else { mDetune.GetValuesAtTime(ticks + start, detune + start, end - start); } } ```

Created attachment 9370812 Incorrect date displayed in fr-doorhanger-milestone-heading2 This was mentioned by a German localizer and I was able to reproduce on my machine locally. The string `cfr-doorhanger-milestone-heading2` should display that X number of trackers were blocked since the user profile began tracking blocked counters, e.g. `Nightly blocked over 1,000 trackers since May 2023!`. Instead, the string is displaying the current month and year (`Nightly blocked over 1,000 trackers since January 2024!`) Steps to reproduce: 1) Create a new profile and open in Firefox 2) Visit a page with a tracking cookie (recommend `amazon.com`). This will create an entry in `protections.sqlite` 3) Use `about:support` to find the Profile Folder where your `protections.sqlite` file is saved 4) Quit Firefox to ensure `protections.sqlite` is editable 5) Use a sqlite database editor such as `DB Browser for SQLite` to open `protections.sqlite` 6) Find the first entry in the `events` table, edit the `count` column to `1000`, and the `datetime` column to something other than January 2024 (e.g. `2023-11-03`). Write the changes to the database and close your editor. 7) Open Firefox with the profile created in step 1 8) Navigate to `about:config`, create and set a config `browser.contentblocking.cfr-milestone.update-interval` and set the value as a number `10` 9) Navigate to `amazon.com`. Refresh the page until the string is triggered. (Should take around 3 refreshes.) 10) Observe the doorhanger (screenshot attached) Similar strings (`graph-total-tracker-summary` and `protections-milestone`) show the date of earliest recorded blocked trackers as expected. I've checked a previous bug (Bug 1687910) in which this string was changed, but was this fixed more than 3 years ago, so something must have changed in the code calling this string.

Scope of this bug is to update shopping sidebar default width to 320px instead of 26.67em See https://bugzilla.mozilla.org/show_bug.cgi?id=1865351#c19

Client ID task-client/f-TYANruT36dl5ZJW7jgcg/0/on/us-central1/6121813105720314584/until/1701312830.229 does not have sufficient scopes and is missing the following scopes: ``` project:comm:thunderbird:releng:flathub:firefox:beta ``` This request requires the client to satisfy the following scope expression: ``` { "AllOf": [ "project:comm:thunderbird:releng:flathub:firefox:beta", "queue:route:tc-treeherder.v2.comm-beta.00853cf0251b46900dadfb74af4c7c38b5484dc7", "queue:create-task:project:none", "queue:scheduler-id:comm-level-3", { "AnyOf": [ "queue:create-task:highest:scriptworker-k8s/comm-3-pushflatpak", "queue:create-task:very-high:scriptworker-k8s/comm-3-pushflatpak", "queue:create-task:high:scriptworker-k8s/comm-3-pushflatpak" ] } ] } ``` That scope is being incorrectly assigned by gecko_taskgraph code, so the server is correctly rejecting the request. The scope will need to be adjusted for Thunderbird.

Created attachment 9368533 Screenshot 2023-12-13 at 4.53.11 PM.png **Steps to Reproduce** 1. In `about:config`, set `browser.firefox-view.search.enabled` to `true`. 2. Open Firefox View. 3. If necessary, sign into Mozilla accounts in order to populate tabs from other devices. 4. In the search box, enter a search term that does not match any synced tab. 5. Clear out the search box. **Expected Result** The original tab lists are restored. **Actual Result** The tab lists are blanked out. It appears that this has something to do with the new `updatesPaused` property introduced in [fxview-tab-list.mjs](https://searchfox.org/mozilla-central/source/browser/components/firefoxview/fxview-tab-list.mjs#240-247). When `shouldUpdate()` is commented out, the component appears to work as expected in this scenario.

Shutdown hang reported for Windows 10 and 11 which got more frequent with Firefox 111.0.x (6000 crash reports vs. 450 for 110.0.x). Crash report: https://crash-stats.mozilla.org/report/index/47e42dbc-f99c-474f-8154-bbf470230413 MOZ_CRASH Reason: ```Shutdown hanging at step XPCOMShutdown. Something is blocking the main-thread.``` Top 10 frames of crashing thread: ``` 0 ntdll.dll ZwWaitForAlertByThreadId 1 ntdll.dll RtlpWaitOnAddressWithTimeout 2 ntdll.dll RtlpWaitOnAddress 3 ntdll.dll RtlWaitOnAddress 4 KERNELBASE.dll WaitOnAddress 5 xul.dll std::sys::windows::thread_parker::Parker::park library/std/src/sys/windows/thread_parker.rs:117 5 xul.dll std::thread::park library/std/src/thread/mod.rs:999 6 xul.dll crossbeam_channel::context::Context::wait_until third_party/rust/crossbeam-channel/src/context.rs:177 7 xul.dll crossbeam_channel::flavors::zero::impl$3::recv::closure$1 third_party/rust/crossbeam-channel/src/flavors/zero.rs:323 7 xul.dll crossbeam_channel::context::impl$0::with::closure$0<crossbeam_channel::flavors::zero::impl$3::recv::closure_env$1<tuple$<> >, enum2$<core::result::Result<tuple$<>, crossbeam_channel::err::RecvTimeoutError> > > third_party/rust/crossbeam-channel/src/context.rs:52 ```

Created attachment 9319401 2023-02-23_09h12_00.png Affected versions: - Nightly (112.0a1/20230222214030) - Beta (111.0b4/20230221190142) Description: The RTAMO onboarding page design is cut at the bottom and along with it, the “Not Now” button. A small portion of the upper part of the button is still visible and can be clicked to expand the address bar i.e. functionality is not affected. Steps to reproduce: 1. While using a VM with Windows, launch Edge or Chrome and access https://addons.mozilla.org 2. On AMO, access the page of any add-on and click the “Download Firefox and get the extension” button 3. Once the installer is downloaded, run it to install Firefox (This will install the latest Firefox Release 110.0) 4. After installation Firefox Release will launch and the RTAMO onboarding page will be displayed, the “Not Now” button being fully visible and none of the page contents being cut. 5. Open the installation folder of Firefox Release, locate the “postSigningData” file and copy it to Desktop. 6. Download and install the latest Firefox Nightly 112.0a1 and Firefox Beta 111.0b4 7. Open the installation folders of Nightly and Beta and paste the previously copied “postSigningData” file inside the folders. 8. Launch both Nightly and Beta and access about:welcome on the browsers 9. The RTAMO onboarding page will be displayed on both browsers, but the page design will be cut at the bottom along with the “Not Now” button. The upper part of the button is still visible and can be clicked. Actual: RTAMO onboarding page design is cut at the bottom along with the “Not Now” button. Expected: The RTAMO onboarding page contents are fully displayed.

Created attachment 9359460 Screenshot of DefaultPrompt and Callout.png An edge case seen in dev that can be replicated with below steps: Pre-Req 1. Set Firefox Nightly as not default browser 2. Create a new profile in latest Firefox Nightly and Open Firefox Nightly with new profile 3. Open about:preferences -> General -> Check open previous windows and tabs checkbox 4. Open about:config and set `browser.shopping.experience2023.enabled` as true Steps 1. Open amazon PDP page and click `Yes try it` in shopping sidebar and exit by clicking x button, shows ‘Get back to review Checker whenever you see the price tag’ feature callout 2. Quit Firefox Nightly 3. Change System date and time to future say 2 days ahead 4. Launch again with the same profile 5. Default prompt and feature callout shows up together Expected Result Feature Callout and Default prompt doesn't display together Actual Result Feature Callout and Default prompt display together. Clicking on Feature Callout and Default Prompt or anywhere in web page doesn’t dismiss either of the dialog. Clicking partially visible `Not Now` of Default prompt dismisses default prompt and than let feature Callout dismiss

### Preconditions: - Set the `browser.shopping.experience2023.enabled` - `TRUE` - Set the `browser.shopping.experience2023.optedIn` - `0` - Turn on a screen reader (i.e. NVDA on Windows OS) ### Steps to reproduce: 1. While the screen reader is running, reach the following link: https://www.amazon.com/dp/B07V6ZSHF4?th=1 2. Observe the announcement made by a screen reader ### Expected result: * The user who cannot see the screen is made aware that there is an opt-in message: * the screen reader announced the text of the opti-in message * (preferred) and the keyboard focus is placed on the `Yes, try it` control (so pressing `Enter` would activate this control) ### Actual result: * Screen reader is not announcing the opt-in message and the user who cannot see the sidebar would not be aware that there is, in fact, a new feature sidebar available for them to use (the screen reader would read the sidebar info at the very end of announcing the entire long product page content) * NVDA user feedback: > It turns out that it [the sidebar] did appear. I didn't realise it was a sidebar and there was no indication it had appeared, so I had no indication that I needed to focus the sidebar with f6 to access it. We probably need to figure out how to fix that; e.g. role="alert" or something so that it gets reported like doorhanger alerts. ### Recommended solution: 1. Move the programmatic (and visual) focus to the `Yes, try it` button on the sidebar when the sidebar appears for the user and shows the opt-in message. Make sure to keep the `role=alertdialog` ARIA role on the message container so the message is announced by a screen reader when the focus lands on the container's control.

``` 0:09.71 error: failed to open file `/home/alex/codaz/Mozilla/gecko-cinnabar/third_party/rust/error-support/Cargo.toml` 0:09.71 --> /home/alex/codaz/Mozilla/gecko-cinnabar/third_party/rust/error-support/src/lib.rs:162:1 0:09.71 | 0:09.71 162 | uniffi::include_scaffolding!("errorsupport"); 0:09.71 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 0:09.71 | 0:09.71 = note: this error originates in the macro `uniffi::include_scaffolding` (in Nightly builds, run with -Z macro-backtrace for more info) 0:09.71 note: erroneous constant used 0:09.71 --> /home/alex/codaz/Mozilla/gecko-cinnabar/third_party/rust/error-support/src/lib.rs:162:1 0:09.71 | 0:09.71 162 | uniffi::include_scaffolding!("errorsupport"); 0:09.71 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 0:09.71 | 0:09.71 = note: this note originates in the macro `uniffi::include_scaffolding` (in Nightly builds, run with -Z macro-backtrace for more info) 0:09.71 sccache: Compiler killed by signal 1 0:09.71 error: could not compile `error-support` due to previous error 0:09.71 Caused by: 0:09.71 process didn't exit successfully: `/home/alex/codaz/Mozilla/sccache/target/release/sccache /home/alex/.rustup/toolchains/stable-x86_64-unknown-linux-gnu/bin/rustc --crate-name error_support --edition=2021 /home/alex/codaz/Mozilla/gecko-cinnabar/third_party/rust/error-support/src/lib.rs --error-format=json --json=diagnostic-ren dered-ansi,artifacts,future-incompat --crate-type lib --emit=dep-info,metadata,link -C opt-level=2 -C panic=abort -C embed-bitcode=no -C debuginfo=2 -C debug-assertions=on -C metadata=c65501e5aa5afc9b -C extra-filename=-c65501e5aa5afc9b --out-dir /home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps --target x86_64-unknown-linux-gnu -C linker=/home/alex/codaz/Mozilla/gecko-cinnabar/build/cargo-linker -L dependency=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps -L dependency=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/debug/deps --extern error_support_macros=/home/alex/codaz/Mo zilla/gecko-cinnabar/obj-browser-dbg/debug/deps/liberror_support_macros-31d34d0b886b254d.so --extern lazy_static=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/liblazy_static-e75e842cb90ee29b.rmeta --extern log=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/de bug/deps/liblog-e9d4752019f925f5.rmeta --extern parking_lot=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libparking_lot-67d8bc657c73e131.rmeta --extern uniffi=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libuniffi-3d32c1c5406f89c4.rmeta --cap-li nts allow -C debuginfo=2 -C force-frame-pointers=yes --cap-lints warn -Clto=off` (exit status: 254) 0:09.71 warning: build failed, waiting for other jobs to finish... 0:10.99 error: failed to open file `/home/alex/codaz/Mozilla/gecko-cinnabar/third_party/rust/glean-core/Cargo.toml` 0:11.00 --> /home/alex/codaz/Mozilla/gecko-cinnabar/third_party/rust/glean-core/src/lib.rs:1136:5 0:11.00 | 0:11.00 1136 | uniffi::include_scaffolding!("glean"); 0:11.00 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 0:11.00 | 0:11.00 = note: this error originates in the macro `uniffi::include_scaffolding` (in Nightly builds, run with -Z macro-backtrace for more info) 0:11.00 note: erroneous constant used 0:11.00 --> /home/alex/codaz/Mozilla/gecko-cinnabar/third_party/rust/glean-core/src/lib.rs:1136:5 0:11.00 | 0:11.00 1136 | uniffi::include_scaffolding!("glean"); 0:11.00 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 0:11.00 | 0:11.00 = note: this note originates in the macro `uniffi::include_scaffolding` (in Nightly builds, run with -Z macro-backtrace for more info) 0:11.00 sccache: Compiler killed by signal 1 0:11.00 error: could not compile `glean-core` due to previous error 0:11.00 Caused by: 0:11.00 process didn't exit successfully: `/home/alex/codaz/Mozilla/sccache/target/release/sccache /home/alex/.rustup/toolchains/stable-x86_64-unknown-linux-gnu/bin/rustc --crate-name glean_core --edition=2021 /home/alex/codaz/Mozilla/gecko-cinnabar/third_party/rust/glean-core/src/lib.rs --error-format=json --json=diagnostic-rendered- ansi,artifacts,future-incompat --crate-type lib --emit=dep-info,metadata,link -C opt-level=2 -C panic=abort -C embed-bitcode=no -C debuginfo=2 -C debug-assertions=on --cfg 'feature="preinit_million_queue"' -C metadata=de6c859d0e0fa087 -C extra-filename=-de6c859d0e0fa087 --out-dir /home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x8 6_64-unknown-linux-gnu/debug/deps --target x86_64-unknown-linux-gnu -C linker=/home/alex/codaz/Mozilla/gecko-cinnabar/build/cargo-linker -L dependency=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps -L dependency=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/debug/deps --extern bincod e=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libbincode-ee6aff6e184367c5.rmeta --extern chrono=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libchrono-6a40d7884def0c6c.rmeta --extern crossbeam_channel=/home/alex/codaz/Mozilla/gecko-cinnabar/obj -browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libcrossbeam_channel-a77aeafd4fd45e5c.rmeta --extern flate2=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libflate2-b78b97e9376aa8e6.rmeta --extern log=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/ liblog-e9d4752019f925f5.rmeta --extern once_cell=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libonce_cell-f19b32376b05f724.rmeta --extern rkv=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/librkv-c3f52c096735b36c.rmeta --extern serde=/home/alex/c odaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libserde-62852348b9c45ac9.rmeta --extern serde_json=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libserde_json-903ac9daa2b805b9.rmeta --extern thiserror=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86 _64-unknown-linux-gnu/debug/deps/libthiserror-9bb2113f7c77916f.rmeta --extern time=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libtime-eb8a554a9f1a5bf3.rmeta --extern uniffi=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libuniffi-3d32c1c5406f89c 4.rmeta --extern uuid=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libuuid-d32b787b69ebc85f.rmeta --extern zeitstempel=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/deps/libzeitstempel-8f2eb5008d586665.rmeta --cap-lints allow -C debuginfo=2 -C force-f rame-pointers=yes --cap-lints warn -Clto=off -L native=/home/alex/codaz/Mozilla/gecko-cinnabar/obj-browser-dbg/x86_64-unknown-linux-gnu/debug/build/lmdb-rkv-sys-bb847ef5d99655bf/out` (exit status: 254) ``` Likely `sccache-dist` does not sent some files or we should do the compile locally

**Steps to reproduce** 1. Navigate to `data:text/html,<meta charset=utf8><style>h1 { color: gold; @media screen {border: 1px solid gold} }</style><h1>Hello</h1>` 1. Open the style editor **Expected results** The at-rules panel is displayed, showing the `@media screen` rule **Actual results** The at-rules panel is not displayed

This is a clone of bug 1832891 where a fix has been landed for the initial navigation when opening new windows and tabs. But there is still an issue with the very first initial navigation during `WebDriver:NewSession` right after the browser has been started. Because we do not wait long enough and miss the additional page load, all commands that get run immediately after the new session has been created could fail because the underlying browsing context gets replaced. There are a couple of different intermittent failures that are caused by this broken behavior. We are going to add them all to the dependency list. We should consider this bug for M7 inclusion.

Created attachment 9341549 Screen Shot 2023-06-28 at 8.30.10 AM.png CFR default 'icon_and_message' layout is showing ratings section by default. This breaks all messages live with layout as 'icon_and_message' https://experimenter.services.mozilla.com/nimbus/device-migration-existing-users-sumo-switch-device-cfr-rollout/summary https://searchfox.org/mozilla-central/source/browser/components/newtab/lib/OnboardingMessageProvider.jsm#81 This issue seems to happen in same session after a CFR message with layout `addon_recommendation` is shown and is replicable in Fx114 and could be possible fallout of Bug 1825658 fix STR: 1. Open about:newtab#devtools and [copy message](https://gist.github.com/punamdahiya/4a8d7808a96c0619dbef0a4558cb8c90) with layout 'addon_recommendation' and hit Show https://experimenter.services.mozilla.com/nimbus/dictionary-addon-all-users/summary 2. Search for CFR message e.g `CFR_FULL_VIDEO_SUPPORT_EN ` with layout as `icon_and_message`. Click Show Expected Result Icon and Message shouldn't show ratings section

[This string](https://searchfox.org/mozilla-central/rev/0c2945ad4769e2d4428c72e6ddd78d60eb920394/browser/locales/en-US/browser/newtab/asrouter.ftl#247) has double en dashes, but ([per Acorn](https://acorn.firefox.com/latest/content/punctuation.html#section-hyphens-and-dashes-c9)) it should have a single em dash `—`. This isn't a critical issue and won't block the experiment. The string has already been translated, and translators have done their own thing with the dashes. So there's no need to change the ID to restart translation. We can just change the EN strings in-place, and let it ride the trains for 115 or 116. The needed changes are very simple: ```diff diff --git a/browser/locales/en-US/browser/newtab/asrouter.ftl b/browser/locales/en-US/browser/newtab/asrouter.ftl index 6615a601fa70c..53b6515cec100 100644 --- a/browser/locales/en-US/browser/newtab/asrouter.ftl +++ b/browser/locales/en-US/browser/newtab/asrouter.ftl @@ -244,6 +244,6 @@ fxa-sync-cfr-secondary = Remind me later ## Device Migration FxA Spotlight device-migration-fxa-spotlight-header = Using an older device? -device-migration-fxa-spotlight-body = Back up your data to make sure you don’t lose important info like bookmarks and passwords –– especially if you switch to a new device. +device-migration-fxa-spotlight-body = Back up your data to make sure you don’t lose important info like bookmarks and passwords — especially if you switch to a new device. device-migration-fxa-spotlight-primary-button = How to back up my data device-migration-fxa-spotlight-link = Remind me later ```

Original report at https://github.com/mozilla/web-ext/issues/2780; I confirmed that this is happening. After starting Firefox Nightly 116 (latest version, buildid 20230607033033), I get `web-ext run -f nightly` > Error: Could not install add-on at '/private/tmp/shared-worker-port': [Exception... "AddonManager is not initialized" nsresult: "0xc1f30001 (NS_ERROR_NOT_INITIALIZED)" location: "JS frame :: resource://gre/modules/AddonManager.sys.mjs :: installTemporaryAddon :: line 2658" data: no]

## Steps to reproduce - In Preferences / Calendar, set `Start the week on` to `Monday` - Restart Thunderbird - Look at the Mini-Month ## Expected behavior Mini-Month first day of the week should be **Monday**, as requested ## Actual behavior - Mini-Month first day of the week is **Sunday** - If flipping the pref to anything then back to **Monday**, then Mini-Month temporarily renders correctly with Monday as first day... - ... but then behaves incorrectly again after a restart. ## Information - 86.0b1-20210128005905 - Bug is a regression, pref used to be honored in 85.0b3 - Linux x86_64 - Available for more debug info, feel free to ask

Spec: https://www.figma.com/file/hGOaTi5ufs2Ese6agbUN4G/New-New-Tab?node-id=985%3A10147

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:85.0) Gecko/20100101 Firefox/85.0 Steps to reproduce: Open my calendar to today and sit back and watch. (i.e. do nothing, but don't switch to another tab like email or tasks, etc.) Windows 10; Thunderbird 78.7; Calendar "Week" view; about 9 calendars enabled; All are Google sync'd. Actual results: Over the course of a day, the red bar indicating the current time of day loses time. (falls behind) It corrects itself upon leaving and activating the calendar tab. This has two problematic effects: 1. The red bar indicates it is 11am when in fact I just missed my 12:30 appointment. 2. Reminders pop-up late. It seems it's not just the display, but the internal clock in the calendar. So if I leave the calendar page and sit on the inbox page for 4 hours, I will rarely get a reminder. When I return to the calendar page, all the missed reminders suddenly pop up. Same thing happens when sitting on the calendar page but not doing anything that causes a refresh. Expected results: The time (as shown by the red horizontal bar in today's calendar) falls behind as much as 15/minutes per hour. This adds up quickly. It corrects itself upon refresh of the calendar tab. Expected results - calendar clock should keep time.

With the pref: browser.newtabpage.activity-stream.newNewtabExperience.enabled=true or browser.newtabpage.activity-stream.customizationMenu.enabled=true When opening the customize menu, sometimes it jumps and stutters. We should fix this and uplioft to 85 for testing.

This only affects Firefox 85.0b1 so far. It's frequent. Crash report: https://crash-stats.mozilla.org/report/index/13e1daf2-1669-48ff-9114-90f0f0201215 MOZ_CRASH Reason: ```Global Glean object not initialized``` Top 10 frames of crashing thread: ``` 0 xul.dll RustMozCrash mozglue/static/rust/wrappers.cpp:16 1 xul.dll mozglue_static::panic_hook mozglue/static/rust/lib.rs:89 2 xul.dll core::ops::function::Fn::call<fn ../7eac88abb2e57e752f3302f02be5f3ce3d7adfb4/library/core/src/ops/function.rs:227 3 xul.dll std::panicking::rust_panic_with_hook ../7eac88abb2e57e752f3302f02be5f3ce3d7adfb4//library/std/src/panicking.rs:581 4 xul.dll std::panicking::begin_panic_handler::{{closure}} ../7eac88abb2e57e752f3302f02be5f3ce3d7adfb4//library/std/src/panicking.rs:484 5 xul.dll std::sys_common::backtrace::__rust_end_short_backtrace<closure-0, !> ../7eac88abb2e57e752f3302f02be5f3ce3d7adfb4//library/std/src/sys_common/backtrace.rs:153 6 xul.dll std::panicking::begin_panic_handler ../7eac88abb2e57e752f3302f02be5f3ce3d7adfb4//library/std/src/panicking.rs:483 7 xul.dll core::panicking::panic_fmt ../7eac88abb2e57e752f3302f02be5f3ce3d7adfb4//library/core/src/panicking.rs:85 8 xul.dll core::option::expect_failed ../7eac88abb2e57e752f3302f02be5f3ce3d7adfb4//library/core/src/option.rs:1226 9 xul.dll core::ops::function::FnOnce::call_once<closure-0, tuple<>> ../7eac88abb2e57e752f3302f02be5f3ce3d7adfb4/library/core/src/ops/function.rs:227 ```

Remove extra added space with newtab update 1 on top of search. It's pushing Pocket down.

about:ion watches for add-on changes, but it is not comparing the cached vs. the add-on listener event correctly.

Crash report: https://crash-stats.mozilla.org/report/index/582e27ee-8961-46a8-b750-c0cc40201109 Reason: ```EXCEPTION_ACCESS_VIOLATION_READ``` Top 10 frames of crashing thread: ``` 0 xul.dll mozilla::gfx::DrawTargetRecording::DrawTargetRecording gfx/2d/DrawTargetRecording.cpp:185 1 xul.dll mozilla::layers::CanvasChild::CreateDrawTarget gfx/layers/ipc/CanvasChild.cpp:252 2 xul.dll mozilla::layers::RecordedTextureData::Lock gfx/layers/client/TextureRecorded.cpp:47 3 xul.dll mozilla::layers::TextureClient::Lock gfx/layers/client/TextureClient.cpp:700 4 xul.dll mozilla::layers::ShareableCanvasRenderer::UpdateCompositableClient gfx/layers/ShareableCanvasRenderer.cpp:213 5 xul.dll nsDisplayCanvas::CreateWebRenderCommands layout/generic/nsHTMLCanvasFrame.cpp:136 6 xul.dll mozilla::layers::WebRenderCommandBuilder::CreateWebRenderCommands gfx/layers/wr/WebRenderCommandBuilder.cpp:1653 7 xul.dll mozilla::layers::WebRenderCommandBuilder::CreateWebRenderCommandsFromDisplayList gfx/layers/wr/WebRenderCommandBuilder.cpp:1758 8 xul.dll nsDisplayOwnLayer::CreateWebRenderCommands layout/painting/nsDisplayList.cpp:6416 9 xul.dll mozilla::layers::WebRenderCommandBuilder::CreateWebRenderCommands gfx/layers/wr/WebRenderCommandBuilder.cpp:1653 ```

When using the dispatcher from RLB in FOG from bug 1662868, [TSAN started complaining](https://treeherder.mozilla.org/jobs?repo=autoland&fromchange=1e6b3c1288586627d5905d6ce87764daf4f7bc2c&searchStr=linux%2C18.04%2Cx64%2Ctsan%2Copt%2Cxpcshell%2Ctests%2Ctest-linux1804-64-tsan%2Fopt-xpcshell-e10&tochange=18db050b9f5606ba641067039e1182d262b137f9&group_state=expanded&selectedTaskRun=TzfhvhleQFOtjKGPjxZoPA.0). Following the steps [here](https://firefox-source-docs.mozilla.org/tools/sanitizer/tsan.html) I was able to reproduce the issue locally, but still not able to understand what's going on, since the dispatcher shutdown procedure is correctly being called. This bug should be take right after bug 1662868 to either confirm the suppression or fix the underlying issue.

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:85.0) Gecko/20100101 Thunderbird/85.0 Steps to reproduce: 1. Preferences → General → Default Search Engine → Add from file 2. Picked a file with following contents: <SearchPlugin xmlns="http://www.mozilla.org/2006/browser/search/"> <ShortName>DuckDuckGo</ShortName> <Description>DuckDuckGo</Description> <InputEncoding>UTF-8</InputEncoding> <Image width="16" height="16">https://duckduckgo.com/favicon.ico</Image> <Url type="text/html" method="GET" template="https://duckduckgo.com/"> <Param name="q" value="{searchTerms}"/> </Url> <SearchForm>https://duckduckgo.com</SearchForm> </SearchPlugin> Actual results: Uncaught (in promise) TypeError: Services.search.addEngine is not a function addSearchEngine chrome://messenger/content/preferences/general.js:701 Expected results: The search engine successfully added.

Spec: https://www.figma.com/file/hGOaTi5ufs2Ese6agbUN4G/New-New-Tab?node-id=1227%3A12445

When resizing the screen, sometimes the search bar and the personalize button overlap. We need some more work here to make the button properly responsive.

## Steps to reproduce 0. Install the official Linux flatpak of Firefox from Flathub: https://flathub.org/apps/details/org.mozilla.firefox 1. Go to any website, press Ctrl+P/select to print. 2. It now opens the system dialog for printing (likely because it uses a portal or so it cannot show Firefox's own dialog) 3. Choose "Print". You see a progress bar as usual (it is quite fast, but sending a printing job may be fast). You also see a notification about a finished printing job. However, it does not print it. Also no print job is shown in the printer settings. What does work, however, is everything else: If I print something else, it even shows the current state in the printing dialog – maybe, because that's just a system component. ## System Firefox 84.0.2 Fedora 33 Silverblue GNOME here Reproduced with: * two different systems (devices) * two different printers (Canon and HP) * two different printer connections (USB, WiFi) When I print the website into a PDF file (which works in the dialog) and then use another app (Evince e.g.) to print the PDF, it does work, however. So it's not a system error. Also, in the non-Flatpak versions it also seems to work.

Google reported to us that download protection requests made from Firefox do not set the API key correctly. The current request is: https://sb-ssl.google.com/safebrowsing/clientreport/download?key=%GOOGLE_SAFEBROWSING_API_KEY% The right one should look like: https://sb-ssl.google.com/safebrowsing/clientreport/download?key=AIzaSyC7jspt... From a quick look, this seems to be related to the fix in Bug 1237103, so this bug might have been here since Firefox 44. This bug didn’t affect the functionality of the download protection because Google didn’t drop requests without the API key previously. However, they are migrating the backends onto a newer infrastructure (1% of traffic at this point), which will drop requests without the correct API key set. This means users will not be protected by download protection. We should fix this issue as soon as possible. For the older version of Firefox, we are still discussing if there is any possible workaround for this issue.

As brought up by @prathiksha and @JulianWels , there is no focus styling that is visible when tabbing to a highlighted tiles right under the search area. The pocket tiles still seem to have the styling. This is not the case without the newnew tab pref so it seems like a recent regression.

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:85.0) Gecko/20100101 Firefox/85.0 Steps to reproduce: 1. Open a zoom invite link 2. Click "Launch meeting" Actual results: Nothing launched. Expected results: In version 84 and below, the flatpak zoom client would always launch. It also doesn't work with the Zoom .rpm.

User Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:80.0) Gecko/20100101 Firefox/80.0 Steps to reproduce: I tried to pass the automatic tests of moz-phab with the instructions provided here: https://github.com/mozilla-conduit/review Basically, I tried to execute the following command: venv/bin/pytest -vv tests/test_bmo.py Actual results: ========================================================================== test session starts ========================================================================== platform linux -- Python 3.8.5, pytest-5.2.1, py-1.8.0, pluggy-0.13.0 -- /.../venv/bin/python3 cachedir: .pytest_cache rootdir: /.../review-mozphab, inifile: pytest.ini collected 242 items tests/test_bmo.py::test_get PASSED [ 0%] tests/test_bmo.py::test_whoami ^C !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! KeyboardInterrupt !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! The second test will be stuck forever until I would do a ctrl+c. In fact, it will be locked mozphab fixture_reset_glean that is trying to call the following Glean_sdk function: glean.testing.reset_glean(application_id="mozphab", application_version="0.1.86") And reset_glean will be stuck waiting for all "glean" process to terminate. And it is stuck, because one process with PingUploadWorker will be stuck Expected results: PingUploadWorker should have returned in a few seconds maximum. And so the subprocess would have been terminated, and reset_glean could return and not be stuck for half a day.

We are seeing these crash being reported on sentry multiple times Fenix related issue https://github.com/mozilla-mobile/fenix/issues/16623 Sentry crash https://sentry.prod.mozaws.net/operations/firefox-nightly/issues/9812825/ ``` java.lang.NumberFormatException: For input string: "undefined" at java.lang.Integer.parseInt(Integer.java:615) at java.lang.Integer.valueOf(Integer.java:801) at org.mozilla.geckoview.MediaSession$Metadata.fromBundle(MediaSession.java:16) at org.mozilla.geckoview.MediaSession$Handler.handleMessage(MediaSession.java:9) at org.mozilla.geckoview.MediaSession$Handler.handleMessage(MediaSession.java:1) at org.mozilla.geckoview.GeckoSessionHandler.handleMessage(GeckoSessionHandler.java:2) at org.mozilla.gecko.EventDispatcher$2.run(EventDispatcher.java:1) at android.os.Handler.handleCallback(Handler.java:883) at android.os.Handler.dispatchMessage(Handler.java:100) at android.os.Looper.loop(Looper.java:214) at android.app.ActivityThread.main(ActivityThread.java:7682) at java.lang.reflect.Method.invoke(Method.java) at com.android.internal.os.RuntimeInit$MethodAndArgsCaller.run(RuntimeInit.java:516) at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:950) ```

Created attachment 9188586 thunderbird-dark-icons.png User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:83.0) Gecko/20100101 Firefox/83.0 Steps to reproduce: After upgrading Thunderbid to v.78.4.3 on Manjaro Linux, which ships dark theme for Thunderbird by default Actual results: Now the new dark-bluish icons for each email folder is not easily visible, as it does very little contrast with the dark background. Also, those brighter icons for each account is not as easy as before. Expected results: Those icons should have more contrast, having a brighter color or a white border.

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_14_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.198 Safari/537.36 Steps to reproduce: 1.Go to firefox version Firefox 83.0 2. Login to facebook account using credentials [email protected]/t3st@123 3. Open our game https://apps.facebook.com/mesmogames/ 4. Open console logs before opening game. Actual results: In console provided error is seen XDR encoding failure in file https://cdn.trackjs.com/agent/v3/latest/t.js 13 : 288. Because of this, app is not loading for certain players in firefox. Note: This issue is seen only in firefox and in particular versions Firefox 83.0 and Firefox 80.0. In no other browser issue is seen. Expected results: This console error should not been seen and game should have loaded properly

On new profiles - in Pocket locales - there is a brief moment on the New Tab page where the Firefox logo will show, then when the user's location is determined, it will hide and Pocket will load. This can be fixed by not showing the logo until the location is determined. Conversely this will cause a slight stutter on non pocket locales, where the logo will load slightly later. But loading a moment later is better than showing and being removed.

Created attachment 9189373 site FF desktop.png User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.198 Safari/537.36 Steps to reproduce: 1. Open web-link: https://www.mozilla.org/ru/firefox/channel/desktop/ 2. Look at the top of the site, where the menu is located Actual results: There is an error in the phrase "Firefox Browsers", in the word "Browsers" Actual result: Баузеры Firefox Expected results: expected Result: (in russian) Браузеры Firefox

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:84.0) Gecko/20100101 Firefox/84.0 Steps to reproduce: No user changes Actual results: This bug was fixed last year but has returned in a different form in 78.6.1 (64-bit) The task list "SHOW" filtering does not work correctly. Also the Today pane task list in not filtering for the time period or task status. Expected results: Proper filter of date and task status

I've many sync pings for my personal account which have, say: ``` "devices": [ { "id": "00000000000000000000000000000000", "type": "desktop", "os": "Darwin", "version": "77.0a1" }, { "id": "00000000000000000000000000000000", "type": "desktop" } ], ... "uid": "2b...", ``` The ping reports a sync did happen (so we should have a token) and given we report the uid, we should be able to report the device IDs. This is true for all recent pings, then they start appearing (on May 2). Without digging deeper and without real evidence, I'm suspecting bug 1604844

Reproducible in Firefox Developer Edition 77.0b3 (64-bit) version. **STR** 1. Ensure that `devtools.responsive.reloadConditions.touchSimulation` is true and `devtools.responsive.browserUI.enabled` is false. 2. Open RDM. 3. Try closing RDM. **Expected** RDM closes without issue. **Actual** RDM doesn't close. The following error is reported in the Browser Console: ``` Exception when calling onAvailable handler this.getViewportBrowser().reload is not a function TypeError: "this.getViewportBrowser().reload is not a function" restoreActorState resource://devtools/client/responsive/ui.js:945 onTargetAvailable resource://devtools/client/responsive/ui.js:1208 promises resource://devtools/shared/resources/target-list.js:527 watchTargets resource://devtools/shared/resources/target-list.js:522 connectToServer resource://devtools/client/responsive/ui.js:476 init resource://devtools/client/responsive/ui.js:224 ResponsiveUI resource://devtools/client/responsive/ui.js:115 openIfNeeded resource://devtools/client/responsive/manager.js:129 toggle resource://devtools/client/responsive/manager.js:92 onKeyShortcut resource://devtools/client/framework/devtools-browser.js:342 onKey resource:///modules/DevToolsStartup.jsm:876 xulKey resource:///modules/DevToolsStartup.jsm:798 createKey resource:///modules/DevToolsStartup.jsm:910 attachKeys resource:///modules/DevToolsStartup.jsm:798 hookKeyShortcuts resource:///modules/DevToolsStartup.jsm:777 onBeforeCreated resource:///modules/DevToolsStartup.jsm:591 aEventName resource:///modules/CustomizableUI.jsm:2862 buildWidget resource:///modules/CustomizableUI.jsm:1698 getWidgetNode resource:///modules/CustomizableUI.jsm:1165 buildArea resource:///modules/CustomizableUI.jsm:975 registerToolbarNode resource:///modules/CustomizableUI.jsm:889 registerToolbarNode resource:///modules/CustomizableUI.jsm:3599 onDOMContentLoaded chrome://browser/content/browser.js:1794 EventListener.handleEvent* chrome://browser/content/browser.xhtml:135 target-list.js:536:19 promises resource://devtools/shared/resources/target-list.js:536 watchTargets resource://devtools/shared/resources/target-list.js:522 connectToServer resource://devtools/client/responsive/ui.js:476 init resource://devtools/client/responsive/ui.js:224 ResponsiveUI resource://devtools/client/responsive/ui.js:115 openIfNeeded resource://devtools/client/responsive/manager.js:129 toggle resource://devtools/client/responsive/manager.js:92 onKeyShortcut resource://devtools/client/framework/devtools-browser.js:342 onKey resource:///modules/DevToolsStartup.jsm:876 xulKey resource:///modules/DevToolsStartup.jsm:798 (Async: EventListener.handleEvent) createKey resource:///modules/DevToolsStartup.jsm:910 attachKeys resource:///modules/DevToolsStartup.jsm:798 hookKeyShortcuts resource:///modules/DevToolsStartup.jsm:777 onBeforeCreated resource:///modules/DevToolsStartup.jsm:591 aEventName resource:///modules/CustomizableUI.jsm:2862 buildWidget resource:///modules/CustomizableUI.jsm:1698 getWidgetNode resource:///modules/CustomizableUI.jsm:1165 buildArea resource:///modules/CustomizableUI.jsm:975 registerToolbarNode resource:///modules/CustomizableUI.jsm:889 registerToolbarNode resource:///modules/CustomizableUI.jsm:3599 onDOMContentLoaded chrome://browser/content/browser.js:1794 <anonymous> chrome://browser/content/browser.xhtml:135 ```

Created attachment 9149249 Dark theme User Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:78.0) Gecko/20100101 Firefox/78.0 Steps to reproduce: Launched TB 77 Beta with a fresh profile, and added one of my email addresses. System is Xubuntu 20.04, tested with Adwaita/Adwaita Dark themes. Actual results: Icons in side pane are barely visible (dark blue on black) Expected results: Icons should be visible (maybe light blue for dark systems themes?) Unrelated, but not sure if it's worth a new bug report or if it's already known, but the color in the active tab is missing until you switch tabs.

# The Problem Since some beta versions of Thunderbird it is no longer possible to edit email addresses that have already been entered. Also, the input fields for the screen reader are not labeled correctly. # steps to reproduce 1. Launch Thunderbird. 1. Compose a new mail with CTRL+N You are now in the "To" field. The screen reader should say "To Edit Field", but it says something like "Empty To input field Edit Field". 1. Enter an e-mail address, press TAB and then press Shift+TAB You are back in the "To" field, the screen reader should say the name of the edit field and the e-mail address you just entered. Instead, it says something like "To input field with one address Edit Field". Now try to edit the e-mail address with the arrow keys. Unfortunately this fails. # Note The tests just described were performed using JAWS 2020.2003.13 and Thunderbird 76.

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:77.0) Gecko/20100101 Firefox/77.0 Steps to reproduce: install Addons in Thunderbird 77 Actual results: i tried to remove them , i can disable them but i cannot remove them Expected results: i should be able to remove the Addons

Created attachment 9140124 example.png On 77 Nightly, most of the sharing icons in the macOS are not displaying for me. This creates visual imbalance, and also makes it challenging to scan the list and find which option you're looking to choose.

This is a blocker to turning on this feature. During the cache load processes there is a new object that is setup but doesn't tear down the old worker. This new object then goes and sets up a new worker. Pretty sure this is a memory leak of sorts. The fix should be easy, either make sure, if there is a provider already available, use it and not make a new one. Or teardown the old one. I prefer the option to use the old one.

**Problem description** When typing text in a plain old textarea in the latest build of Fenix, Talkback speaks too much information. For each letter I type, Talkback announces "Editing. [all the text I've entered ] Entry. [the label for the textarea] Edit options available. Use swipe up then right to view." when it probably should just speak the letter I typed. **Steps to reproduce** Load a page with this content: `<label for="message">Good Times:</label><textarea id="message" name="message"></textarea>` https://codepen.io/asadotzler/pen/bGVVevY Now focus the textarea and double tap to bring up the keyboard. Type the word "fun" **Results** Talback announces "Editing. Fun. Entry. Good Times. Edit options available. Use swipe up then right to view." **Expected results** Talkback announces each letter as I type it and nothing else.

The font-family telemetry added by bug 1619349 is broken for Windows because of a copy/paste error in the DirectWrite impl of GetVisibilityForFamily -- it never returns the LangPack value. :(

Created attachment 9144445 Logcat captured while reproducing issue. STR: 1) Visit https://www.ezdrm.com/html/drm-demos.asp and play video 2) Open new tab and visit youtube.com 3) Play any video Actual: Youtube videos are broken Expected: Youtube videos will play This reproduces with and DRM video such as amazon prime. Related FxR issue: https://github.com/MozillaReality/FirefoxReality/issues/3285 The FxR issue contains a screen capture of the issue. I was able to reproduce this in GeckoView Example. I've attached a logcat captured while reproducing the issue.

This bug is for crash report bp-a18b3a1d-63f9-4bfb-bb97-f422c0200503. ``` Top 10 frames of crashing thread: 0 xul.dll nsGlobalWindowInner::StorageAccessGranted dom/base/nsGlobalWindowInner.cpp:7202 1 xul.dll static mozilla::ContentBlocking::OnAllowAccessFor toolkit/components/antitracking/ContentBlocking.cpp:567 2 xul.dll mozilla::ContentBlocking::AllowAccessFor::<unnamed-tag>::operator const toolkit/components/antitracking/ContentBlocking.cpp:363 3 xul.dll mozilla::MozPromise<mozilla::Maybe<mozilla::ContentBlocking::StorageAccessPromptChoices>, mozilla::ipc::ResponseRejectReason, 1>::ThenValue<`lambda at /builds/worker/checkouts/gecko/toolkit/components/antitracking/ContentBlocking.cpp:354:14'>::DoResolveOrRejectInternal xpcom/threads/MozPromise.h:798 4 xul.dll mozilla::MozPromise<CopyableTArray<bool>, nsresult, 0>::ThenValueBase::ResolveOrRejectRunnable::Run xpcom/threads/MozPromise.h:405 5 xul.dll nsThread::ProcessNextEvent xpcom/threads/nsThread.cpp:1200 6 xul.dll NS_ProcessNextEvent xpcom/threads/nsThreadUtils.cpp:481 7 xul.dll mozilla::ipc::MessagePump::Run ipc/glue/MessagePump.cpp:87 8 xul.dll MessageLoop::RunHandler ipc/chromium/src/base/message_loop.cc:308 9 xul.dll MessageLoop::Run ipc/chromium/src/base/message_loop.cc:290 ``` This looks like a a dereference of a `NULL` `this` pointer. The oldest buildid for this crash is 20200429215539. Tentatively filing under Core > Privacy: Anti-Tracking because most of the changes around this code seem to be happening in this component. Please move to the right component if this is wrong.

This bug is for crash report bp-3d1796a9-b704-4049-abed-1ff470200413. ``` Top 10 frames of crashing thread: 0 XUL RustMozCrash mozglue/static/rust/wrappers.cpp:17 1 XUL mozglue_static::panic_hook mozglue/static/rust/lib.rs:89 2 XUL core::ops::function::Fn::call src/libcore/ops/function.rs:72 3 XUL std::panicking::rust_panic_with_hook src/libstd/panicking.rs:475 4 XUL rust_begin_unwind src/libstd/panicking.rs:375 5 XUL core::panicking::panic_fmt src/libcore/panicking.rs:84 6 XUL core::panicking::panic src/libcore/panicking.rs:51 7 XUL cubeb_coreaudio::backend::buffer_manager::BufferManager::push_silent_data third_party/rust/cubeb-coreaudio/src/backend/buffer_manager.rs 8 XUL cubeb_coreaudio::backend::audiounit_output_callback third_party/rust/cubeb-coreaudio/src/backend/mod.rs:615 9 CoreAudio CoreAudio@0xca1e ```

If we re initialize a personality provider using cache, it's not properly passing the previous cached score into the new insistence, causing it to make them fresh which is not really ideal.

This bug is for crash report bp-8479b20a-ebbf-4cad-a4bc-424330200415. ``` Java stack trace: java.lang.AssertionError at org.mozilla.gecko.process.GeckoProcessManager$ChildConnection.bind(GeckoProcessManager.java:19) at org.mozilla.gecko.process.GeckoProcessManager.start(GeckoProcessManager.java:7) at org.mozilla.gecko.process.GeckoProcessManager.lambda$start$5(GeckoProcessManager.java:2) at org.mozilla.gecko.process.-$$Lambda$GeckoProcessManager$49HYY0DiXCj5gwg0g9ecfGUvYqc.run(Unknown Source:20) at org.mozilla.gecko.util.XPCOMEventTarget$JNIRunnable.run(XPCOMEventTarget.java:1)

STR: Start new profile of Firefox in the UK, engine list will look like Google Amazon.com DuckDuckGo Bing Twitter Wikipedia eBay Amazon.co.uk Then restart, engine order looks like: eBay Amazon.co.uk Bing DuckDuckGo Google Twitter Wikipedia

## Step to Reproduce 1. Open Firefox Nightly and go to youtube 2. Open any video #1 - Now the metadata shown on Notification bar meets the info on youtube 3. Open another video #2 via another tab, or click other videos on same page - The metadata shown on Notification bar isn't updated ## Expect Result The metadata shown on Notification bar is updated to the info of video #2 ## Expect Result The metadata shown on Notification bar stays unchanged. It's still the info of video #1 I am on Ubuntu 18.04

It's _way_ too low for real-world usage, 32MB: https://searchfox.org/mozilla-central/source/dom/worklet/WorkletThread.cpp#24. This is making https://gogins.github.io/csound-extended/html/examples/instr.csd.html error out with OOM. This is port of [csound](https://en.wikipedia.org/wiki/Csound) to the web using emscripten, upstream is https://github.com/gogins/csound-extended. Bumping this to 64MB makes the worklet thread GC continuously, but 128MB makes it work. (note that this page has a few of other issues that makes it not work, I'm attaching a patched copy. I'll file those upstream)

Created attachment 9142257 Screen Shot 2020-04-21 at 11.58.41 PM.png Steps to reproduce: 1. Enable dark mode in macOS (I am using Catalina) 2. Enable separate search box 3. Type anything into search box What happens: The search suggestions panel background color is a gray that has bad contrast. Expected result: As previous. 18:05.50 INFO: Narrowed inbound regression window from [b87c6162, e7f852ba] (3 builds) to [b87c6162, 6c312e49] (2 builds) (~1 steps left) 18:05.50 INFO: No more inbound revisions, bisection finished. 18:05.50 INFO: Last good revision: b87c616243ab9b9b5210ba23317cc0668bb5a27b 18:05.50 INFO: First bad revision: 6c312e49ee094e66684abedc5ee314dfa595bfa1 18:05.50 INFO: Pushlog: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=b87c616243ab9b9b5210ba23317cc0668bb5a27b&tochange=6c312e49ee094e66684abedc5ee314dfa595bfa1

Steps to reproduce: Download the testcase and run: `python3 -m http.server` Open the developer tools (Network) and navigate to http://localhost:8000/testcase.html This will send a XHR request to the same origin on a page where Referrer-Policy: no-referrer is applied. What happened: Request headers include: ``` Origin: null Sec-Fetch-Dest: empty Sec-Fetch-Mode: cors Sec-Fetch-Site: same-origin ``` Expected results: The request header should be this (matches Firefox 75.0): ``` Origin: http://localhost:8000 ``` Additional information: This change was bisected to https://hg.mozilla.org/integration/autoland/rev/a925b80f029e3f2107a91bde02e9799fb9821eb3 (application_buildid: 20200420201907) It breaks search functionality on a Confluence wiki page where this referrer-policy is set.

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/76.0.3809.132 Safari/537.36 Steps to reproduce: Click on a link to display a PDF in i iFrame in a new window. Actual results: The PDF.js black windows remains blank. We have the following messages : Content Security Policy: Les paramètres de la page ont empêché le chargement d’une ressource à resource://pdf.js/web/viewer.js, (« script-src »).Content Security Policy: Les paramètres de la page ont empêché le chargement d’une ressource à resource://pdf.js/build/pdf.js, (« script-src »). CSP in configured in .htaccess file as followed : Header always set Content-Security-Policy "script-src 'self';" Expected results: the pdf.js files should be allowed to be loaded

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:81.0) Gecko/20100101 Firefox/81.0 Steps to reproduce: Fix for Bug 1656155 caused on Solaris build failure: ``` gmake[3]: Entering directory '/builds/psumbera/FIREFOX/obj-x86_64-pc-solaris2.11/memory/replace/logalloc' memory/replace/logalloc/FdPrintf.o /builds/psumbera/FIREFOX/epoll/g++ -std=gnu++17 -o FdPrintf.o -c -I/builds/psumbera/FIREFOX/obj-x86_64-pc-solaris2.11/dist/system_wrappers -include /builds/psumbera/FIREFOX/config/gcc_hidden.h -fstack-protector-strong -DDEBUG=1 -DMOZ_REPLACE_MALLOC_PREFIX=logalloc -DMOZ_NO_MOZALLOC -DMOZ_HAS_MOZGLUE -DIMPL_MFBT -I/builds/psumbera/FIREFOX/memory/replace/logalloc -I/builds/psumbera/FIREFOX/obj-x86_64-pc-solaris2.11/memory/replace/logalloc -I/builds/psumbera/FIREFOX/memory/build -I/builds/psumbera/FIREFOX/obj-x86_64-pc-solaris2.11/dist/include -I/builds/psumbera/FIREFOX/obj-x86_64-pc-solaris2.11/dist/include/nspr -I/builds/psumbera/FIREFOX/obj-x86_64-pc-solaris2.11/dist/include/nss -fPIC -DMOZILLA_CLIENT -include /builds/psumbera/FIREFOX/obj-x86_64-pc-solaris2.11/mozilla-config.h -Wall -Wempty-body -Wignored-qualifiers -Woverloaded-virtual -Wpointer-arith -Wsign-compare -Wtype-limits -Wunreachable-code -Wwrite-strings -Wno-invalid-offsetof -Wduplicated-cond -Wimplicit-fallthrough -Wunused-function -Wunused-variable -Wno-error=maybe-uninitialized -Wno-error=deprecated-declarations -Wno-error=array-bounds -Wno-error=coverage-mismatch -Wno-error=free-nonheap-object -Wno-multistatement-macros -Wno-error=class-memaccess -Wno-error=deprecated-copy -Wformat -Wformat-overflow=2 -Werror=implicit-function-declaration -Wno-psabi -fno-sized-deallocation -fno-aligned-new -fno-exceptions -fno-strict-aliasing -fno-rtti -ffunction-sections -fdata-sections -fno-exceptions -fno-math-errno -pthread -pipe -g -fno-omit-frame-pointer -funwind-tables -MD -MP -MF .deps/FdPrintf.o.pp /builds/psumbera/FIREFOX/memory/replace/logalloc/FdPrintf.cpp cc1plus: warning: ‘-Werror=’ argument ‘-Werror=implicit-function-declaration’ is not valid for C++ /builds/psumbera/FIREFOX/memory/replace/logalloc/FdPrintf.cpp: In instantiation of ‘unsigned int NumDigits(T) [with T = long unsigned int]’: /builds/psumbera/FIREFOX/memory/replace/logalloc/FdPrintf.cpp:123:46: required from here /builds/psumbera/FIREFOX/memory/replace/logalloc/FdPrintf.cpp:70:19: error: call of overloaded ‘log10(long unsigned int&)’ is ambiguous 70 | double l = log10(n); | ~~~~~^~~ In file included from /usr/gcc/9/lib/gcc/x86_64-pc-solaris2.11/9.3.0/include-fixed/math.h:24, from /usr/gcc/9/include/c++/9.3.0/cmath:45, from /builds/psumbera/FIREFOX/obj-x86_64-pc-solaris2.11/dist/system_wrappers/cmath:3, from /builds/psumbera/FIREFOX/memory/replace/logalloc/FdPrintf.cpp:14: /usr/gcc/9/lib/gcc/x86_64-pc-solaris2.11/9.3.0/include-fixed/iso/math_iso.h:209:21: note: candidate: ‘long double std::log10(long double)’ 209 | inline long double log10(long double __X) { return __log10l(__X); } | ^~~~~ /usr/gcc/9/lib/gcc/x86_64-pc-solaris2.11/9.3.0/include-fixed/iso/math_iso.h:170:15: note: candidate: ‘float std::log10(float)’ 170 | inline float log10(float __X) { return __log10f(__X); } | ^~~~~ /usr/gcc/9/lib/gcc/x86_64-pc-solaris2.11/9.3.0/include-fixed/iso/math_iso.h:70:15: note: candidate: ‘double std::log10(double)’ 70 | extern double log10 __P((double)); | ^~~~~ gmake[3]: *** [/builds/psumbera/FIREFOX/config/rules.mk:725: FdPrintf.o] Error 1 ```

Created attachment 9182135 install.log On a Windows 8.1 machine, installing Firefox 82.0 RC 2 and 83.0a1 into a new folder requires a reboot to complete. The issue does not reproduce on a Windows 10 machine. The full .exe installers are used (en-US) versions, both 32-bit and 64-bit reproduce the issue. The installation dialog shows "A Little Housekeeping…" for several seconds. The installation folder is outside C:\Program Files After the install and before the reboot, there is a firefox.exe.moz-upgrade file in the install folder which is gone after the reboot.

This seems to be a regression in GV 82. STR: - Load any page - Observe onProgressChange 15 - Observe onProgressChange 100 In 81 and before we received onProgressChange calls for 55 and 80 as well. This is causing the progress bar in Fenix to never finish and hurts perceived performance.

I don't investigate regression range yet. I found this yesterday. ### Env - Fenix Nightly 201018 17:01 - Android 10 (Xperia XZ2 Compact) - GBoard ### Step 1. Browse https://www.google.com 2. Type any string (ex. google) 3. set caret between characters. 4. Type any character. (ex. a) and commit this. 5. Set caret to tail 6. Type any character ### Result 6.'s character is inserted after 3.'s character. This doesn't occurs on Fenix beta (82-beta6)

I saw a crash from this location in a try push.

From github: https://github.com/mozilla-mobile/fenix/issues/16037. > ## Steps to reproduce > 1. Open any web site with Thai characters, for example https://th.wikipedia.org/ > 2. Long tap on any Thai text, as if you wanted to select some text. > > ### Expected behavior > > Some text is being selected. > > ### Actual behavior > > 1. No text is being selected by the long tap. > 2. No text, not even with latin script, can be selected any more. > 3. Loading a page in new tabs will do nothing at all. > > ### Device information > > * Android device: Xiaomi Mi A2 Lite, Android 10, last system update 2020-09-01 > * Fenix version: 81.1.4 (Build #2015768387), reproducible with today's nightly build (from Google Play Store) > Change performed by the [Move to Bugzilla add-on](https://addons.mozilla.org/en-US/firefox/addon/move-to-bugzilla/).

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.75 Safari/537.36 Steps to reproduce: Opened page, clicked on print Actual results: Only header was visible Expected results: Contents of the page should have been visible

User Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:81.0) Gecko/20100101 Firefox/81.0 Steps to reproduce: I try to print receipt to Star TSP 100 & TSP650II model printers. I set headers & footers blank, and margins to zero. Then print. Receipt printing works with Firefox 81.0.2 Actual results: Printer prints a 1,5cm blank slice. Print preview is also blank. Expected results: Receipt should have printed corectly.

Created attachment 9183057 Screen Shot 2020-10-21 at 2.08.01 PM.png User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:82.0) Gecko/20100101 Firefox/82.0 Steps to reproduce: *My FF version is not using the Chromium print prompt. *macOS Catalina (10.15.7) 1) Open example PDF: https://food-guide.canada.ca/static/assets/pdf/HEPs-Guide-nw-en.pdf 2) Click printer icon. 3) Set Paper size to "Letter". Set scale to 100%. 4) Print OR Open with Preview. Actual results: There appear to be extra margins at the top and left sides, pushing some content off the printed document. Please see screenshot attached. Expected results: The printed document should match the onscreen PDF, since paper size is the same and scale is 100%.

STR the crash: 1. Open https://www.w3schools.com/tags/tryit.asp?filename=tryhtml_iframe with new print UI 2. Highlight "The iframe element" on right side of page 3. "Print selection" using context menu 4. Tab Crashes STR the error: 1. Open https://www.w3schools.com/tags/tryit.asp?filename=tryhtml_iframe with new print UI 2. Highlight anything on the left side of page 3. "Print selection" using context menu 4. Notice print preview error The error is being caught from `await previewBrowser.frameLoader.printPreview(settings, sourceWinId));` is `UnknownError("Print preview failed", (void 0))`. Here are the warnings I see: ``` [Parent 59430, Main Thread] WARNING: 'nsContentUtils::GetCommonBrowserParentAncestor( remote, oldRemote) != remote', file /Users/emmamalysz/Desktop/firefox-source/dom/events/EventStateManager.cpp:1367 [Child 59437, Main Thread] WARNING: '!endContent', file /Users/emmamalysz/Desktop/firefox-source/dom/base/Selection.cpp:1511 [Child 59437, Main Thread] WARNING: '!mPrt->mPrintObject->mPresContext', file /Users/emmamalysz/Desktop/firefox-source/layout/printing/nsPrintJob.cpp:1239 [Child 59437, Main Thread] WARNING: nsPrintJob::ResumePrintAfterResourcesLoaded failed: 'rv == NS_ERROR_ABORT', file /Users/emmamalysz/Desktop/firefox-source/layout/printing/nsPrintJob.cpp:1545 ``` I'm not sure if this should be split into two separate bugs. This blocks the work I'm doing in Bug 1670122 because adding this option directly into the settings UI makes it much easier for a user to encounter this error.

Created attachment 9183628 firefox-debug_0e74_2020-10-24_17-24-01-965.log User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.75 Safari/537.36 Steps to reproduce: - <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event"> - <System> <Provider Name="Application Error" /> <EventID Qualifiers="0">1000</EventID> <Level>2</Level> <Task>100</Task> <Keywords>0x80000000000000</Keywords> <TimeCreated SystemTime="2020-10-24T13:08:22.372506800Z" /> <EventRecordID>19589</EventRecordID> <Channel>Application</Channel> <Computer>DESKTOP-FQ35S98</Computer> <Security /> </System> - <EventData> <Data>firefox.exe</Data> <Data>84.0.0.7602</Data> <Data>5f940f9b</Data> <Data>ntdll.dll</Data> <Data>10.0.14393.3986</Data> <Data>5f77fd0d</Data> <Data>c0000005</Data> <Data>000000000003469c</Data> <Data>824</Data> <Data>01d6aa06c00a6e80</Data> <Data>C:\Program Files\Firefox Nightly\firefox.exe</Data> <Data>C:\WINDOWS\SYSTEM32\ntdll.dll</Data> <Data>9f952e82-7c06-430e-acb3-db4ea80f36b3</Data> <Data /> <Data /> </EventData> </Event> I try to use windbg and mozregression. Tested autoland build: 2a46f29c (verdict: g) app_name: firefox build_date: 2020-10-21 00:08:25.137000 build_file: C:\Users\amix\.mozilla\mozregression\persist\2a46f29c8cdd--autoland--target.zip build_type: integration build_url: https://firefox-ci-tc.services.mozilla.com/api/queue/v1/task/cMjDIhv6QXumrBVi86wxhQ/runs/0/artifacts/public%2Fbuild%2Ftarget.zip changeset: 2a46f29c8cdd6f6269483e86c6e21e649ffa7127 pushlog_url: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=2a46f29c8cdd6f6269483e86c6e21e649ffa7127&tochange=e88986658685538d5ab62d66c31e183b43e813ce repo_name: autoland repo_url: https://hg.mozilla.org/integration/autoland task_id: cMjDIhv6QXumrBVi86wxhQ Tested autoland build: 4c3e6fb9 (verdict: b) app_name: firefox build_date: 2020-10-21 00:24:50.430000 build_file: C:\Users\amix\.mozilla\mozregression\persist\4c3e6fb95aa6-pgo--autoland--target.zip build_type: integration build_url: https://firefox-ci-tc.services.mozilla.com/api/queue/v1/task/Okfv9Gf6T-GgBvUhGNEU1g/runs/0/artifacts/public%2Fbuild%2Ftarget.zip changeset: 4c3e6fb95aa6184651f215d466e702ff7d1660d3 pushlog_url: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=2a46f29c8cdd6f6269483e86c6e21e649ffa7127&tochange=4c3e6fb95aa6184651f215d466e702ff7d1660d3 repo_name: autoland repo_url: https://hg.mozilla.org/integration/autoland task_id: Okfv9Gf6T-GgBvUhGNEU1g Differential Revision: https://phabricator.services.mozilla.com/D93709 Actual results: Windows event log full of app errors since 22.10.2020.

Currently we badge the Pioneer icon if the `pioneer-study-addons` collection changes, but we switched this to `pioneer-study-addons-v1`. We could work around it on the server, but it'd be much cleaner to just have this in the client.

Created attachment 9175992 screenshot: tb78, first install dialog with Win high-contrast unreadable (white on white) +++ This bug was initially created as a clone of Bug #1661229 +++ Steps to reproduce: 1. Enable high-contrast mode in Windows 10, by doing the following: 1.1. Open Windows Settings (Windows+I). 1.2. Go to Ease of access -> High-contrast. 1.3. Turn on the high-contrast mode, using the respective toggle. 1.4. From the "High-contrast theme" combo-box, select "High-Contrast Black". 2. Install and launch Thunderbird 78.2. Actual results: - Intro text on first installer is invisible (probably white high-contrast text on white preset bg) - Subsequent screens seem mostly readable, but we should also change the top background stripe on install dialog screens to be black (under the icon of Thunderbird) to look less out of place. Expected results: In such dialogs and windows as in the examples above, the background color should be taken from OS high contrast colors (black) so that text is readable.

Created attachment 9176874 Untitled.png **Affected versions** * 82.0a1(2020-09-20); **Affected platforms** * Windows 10, macOS 11.0; **Steps to reproduce** 1. Launch Firefox, click on the Hamburger menu; 2. Click on the What's new option; 3. Press the Tab key a couple of times; **Expected result** * no issues; **Actual result** * tab-selection indicator cuts the date for listed entries; **Regression range** * First bad build_date: 2020-05-10; * Last good build_date: 2020-05-11; * Potential regressor: 1630907 **Additional notes** * attached screenshot with the issue; * S3 as suggested severity.

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Firefox/68.0 Steps to reproduce: FTR, I attempted to report this issue to RNP via https://www.ribose.com/feedback/security, but got an error. Summary ======= When rnp uses a signature subpacket, it doesn't check whether the subpacket came from the hashed subpacket area (signature metadata that is integrity protected by the signature), or the unhashed subpacket area (signature metadata that is *not* integrity protected by the signature). This allows an attacker to modify how rnp interprets a signature without invalidating the signature. The scope of this vulnerability is partially mitigated by the way rnp looks up subpackets. First, if there are multiple subpackets of the same type, it returns the first instance that it finds. Second, it considers the hashed subpacket area before it considers the unhashed subpacket area. Thus, an attacker can only control the value of subpackets that are not present in the hashed subpacket area. An attacker can exploit this vulnerability by causing rnp to think that a key is expired when it doesn't actually expire by adding a Key Expiration Time subpacket to the key's self-signature. Because rnp considers the key to be expired, the user can't encrypt a message using the key, and may decide to send the message unencrypted. Further, once a signature has been inserted into rnp's keyring, it is non-trivial to remove it. Specifically, if Mallory has modified Alice's key, and Bob imports the modified version, then importing an unmodified version of the key will not update the modified signature; the update will be ignored. This works in both directions: if a clean copy of a signature is present in the user's keyring, importing a corrupted signature will not cause the saved version to be replaced with the corrupted version. Note: this vulnerability applies to all subpackets that belong in the hashed subpacket area, not just the Key Expiration Time subpacket. For instance, rnp includes an accessor to get the Trust subpacket (`signature_get_trust`). Like the other accessors, this function does not check whether the subpacket came from the hashed subpacket area, nor does the accessor return this information so that the caller can check. An attacker can use this to effectively change a third-party certification into a trusted signature, i.e., it can turn the key into a CA. Background ========== OpenPGP v4 signatures include so-called subpackets. A subpacket describes some property of the signature. For instance, the Signature Creation Time subpacket indicates when the signature was allegedly created, and the Expiration Time subpacket indicates when the Signature expires. Whereas the first subpacket is mandatory, the second subpacket is only present if the signature should expire. An OpenPGP v4 signature includes two so-called subpacket areas: a so-called hashed subpacket area, and a so-called unhashed subpacket area. The hashed subpacket area's integrity is protected by the signature; the unhashed subpacket area's is not. As such, the unhashed area should only be used to store self-authenticating or advisory data, and OpenPGP implementations should ignore subpackets in the unhashed area that require integrity projection. Vulnerability ============= When RNP parses a signature, it calls [`signature_read_v4` (`src/librepgp/stream-packet.cpp`)][1] to parse the subpackets. That function parses the subpackets and adds them to a vector, [`sig->subpkts` (`lib/types.h`)][2], in the order that they are encountered. It parses subpackets in the hashed subpacket area first and then the subpackets in the unhashed subpacket area. Thus, packets from the hashed area occur before packets from the unhashed area in the `subpkts` vector. The saved information includes not only the subpacket, but also whether the packet came from the hashed subpacket area. [1] https://github.com/rnpgp/rnp/blob/6a5fb890eb30a59f7aa46b3e915a236d4bbe4176/src/librepgp/stream-packet.cpp#L1639 [2] https://github.com/rnpgp/rnp/blob/6a5fb890eb30a59f7aa46b3e915a236d4bbe4176/src/lib/types.h#L319 rnp provides several functions to look up a subpacket. [`signature_get_subpkt` (`src/librepgp/stream-sig.cpp`)][3] returns the first subpacket with the specified tag. If there are multiple subpackets with the same type, the first one is returned. The subpacket is not filtered for whether the subpacket area that it comes from is acceptable for that kind of packet. It is possible for the caller to check whether the subpacket came from the unhashed area, however, no callers in the rnp source code appear to do this. For instance, [`pgp_key_latest_selfsig` (`pgp-key.cpp`)][4] uses the accessor to check for the Signature Creation Time subpacket without checking that the returned subpacket comes from the hashed subpacket area. Given that the documentation also does not appear to mention that the subpackets need to be filtered, it is unreasonable to expect that any callers are using this function correctly. [3] https://github.com/rnpgp/rnp/blob/6a5fb890eb30a59f7aa46b3e915a236d4bbe4176/src/librepgp/stream-sig.cpp#L64 [4] https://github.com/rnpgp/rnp/blob/6a5fb890eb30a59f7aa46b3e915a236d4bbe4176/src/lib/pgp-key.cpp#L859 The higher-level subpacket accessor functions are defined in [`librepgp/stream-sig.cpp`] starting at line 148. There are accessors for the following packets: - [Issuer](https://tools.ietf.org/html/rfc4880#section-5.2.3.5) - [Signature Creation Time](https://tools.ietf.org/html/rfc4880#section-5.2.3.4) - [Signature Expiration Time](https://tools.ietf.org/html/rfc4880#section-5.2.3.10) - [Key Expiration Time](https://tools.ietf.org/html/rfc4880#section-5.2.3.6) - [Key Flags](https://tools.ietf.org/html/rfc4880#section-5.2.3.21) - [Primary User ID](https://tools.ietf.org/html/rfc4880#section-5.2.3.19) - [Preferred Symmetric Algorithms](https://tools.ietf.org/html/rfc4880#section-5.2.3.7) - [Preferred Hash Algorithms](https://tools.ietf.org/html/rfc4880#section-5.2.3.8) - [Preferred Compression Algorithms](https://tools.ietf.org/html/rfc4880#section-5.2.3.9) - [Key Server Preferences](https://tools.ietf.org/html/rfc4880#section-5.2.3.17) - [Preferred Key Server](https://tools.ietf.org/html/rfc4880#section-5.2.3.18) - [Trust](https://tools.ietf.org/html/rfc4880#section-5.2.3.13) - [Revocable](https://tools.ietf.org/html/rfc4880#section-5.2.3.12) - [Features](https://tools.ietf.org/html/rfc4880#section-5.2.3.24) - [Signer's User ID](https://tools.ietf.org/html/rfc4880#section-5.2.3.22) - [Embedded Signature](https://tools.ietf.org/html/rfc4880#section-5.2.3.26) - [Reason For Revocation](https://tools.ietf.org/html/rfc4880#section-5.2.3.23) [5] https://github.com/rnpgp/rnp/blob/c10e33ec75393eb810417be8d447a11d5da289f0/src/librepgp/stream-sig.cpp#L148 These functions don't check whether subpackets that must come from the hashed subpacket area actually do nor do they return this information to the caller. If an attacker modifies the Issuer subpacket or Embedded Signature subpacket, then the user cannot use the signature. But, since an attacker could just as easily invalidate the signature by flipping an arbitrary bit, this attack on these subpackets is not interesting. Because rnp returns the first instance of a subpacket and prefers packets in the hashed subpacket area, subpackets that are almost always or always set (in the relevant content, e.g., revocation) will be difficult or impossible for an attacker to exploit. These include the Signature Creation Time, Key Flags, and Reason for Revocation. This leaves the following subpackets that an attacker can sometimes control: - Signature Expiration Time - Key Expiration Time - Primary User ID - Cryptographic preferences (Preferred Symmetric Algorithms, Preferred Hash Algorithms, Preferred Compression Algorithms). - Key Server Preferences - Preferred Key Server - Trust - Revocable - Features - Signer's User ID (I'm ignoring subpackets that rnp doesn't provide accessors for.) The following subsections describe some possible attacks that abuse these subpackets. The attacks are sorted by severity. Key Expiration Time ------------------- By setting this on a key that doesn't otherwise expire, an attacker can cause the key to be unusable. In particular, if rnp considers a key be expired, it won't encrypt to it: ``` $ rm /tmp/tmp.MhdvXtVrFA/* $ src/rnpkeys/rnpkeys --homedir=$T --import key-expiration-bad.pgp [init_file_src() /home/us/neal/work/pep/rnp/src/librepgp/stream-common.cpp:408] can't stat '/tmp/tmp.MhdvXtVrFA/pubring.gpg' wrong pubring path pub 255/EdDSA ce381c215870a0b8 2020-09-22 [SC] [EXPIRED 2020-09-22] a42e374d8e6863b073dd5e97ce381c215870a0b8 uid Key Expiration <[email protected]> sub 255/ECDH c0ff86940f2b070e 2020-09-22 [E] [EXPIRED 2020-09-22] 24676b5208646f91c71cba33c0ff86940f2b070e $ src/rnp/rnp --homedir=$T --encrypt -r ce381c215870a0b8 foo.txt [encrypted_add_recipient() /home/us/neal/work/pep/rnp/src/librepgp/stream-write.cpp:518] attempt to use invalid key as recipient ``` But, gpg has no problem encrypting to it, because it corruptly ignores the invalid subpacket: ``` $ gpg --import key-expiration-bad.pgp gpg: key 0xCE381C215870A0B8: public key "Key Expiration <[email protected]>" imported gpg: Total number processed: 1 gpg: imported: 1 $ echo | gpg -r 0xCE381C215870A0B8 > /dev/null gpg: WARNING: no command supplied. Trying to guess what you mean ... gpg: [don't know]: invalid packet (ctb=0a) $ echo | gpg -e -r 0xCE381C215870A0B8 > /dev/null $ echo $? 0 ``` Importing the good version after importing the bad version doesn't fix the problem: ``` $ src/rnpkeys/rnpkeys --homedir=$T --import key-expiration-good.pgp sec keyring '/tmp/tmp.MhdvXtVrFA/secring.gpg' is empty $ src/rnp/rnp --homedir=$T --encrypt -r ce381c215870a0b8 foo.txt [encrypted_add_recipient() /home/us/neal/work/pep/rnp/src/librepgp/stream-write.cpp:518] attempt to use invalid key as recipient ``` As such, it is extremely difficult to recover from this type of attack. Note: an attacker can upload the signature to a key server, and it may not be filtered out. For instance, AFAICT, SKS will delivery both signatures, because they appear to be different. If the corrupted signature is delivered first, then rnp will use it, and the key will appear to be broken. Trust ----- An attacker can use this to turn a plain certification into a tsig. Consider: Alice certifies Bob's OpenPGP Certificate. Mallory can change her certification into a tsig. Then, everyone who considers Alice to be a meta-trusted introducer (CA) will also consider Bob to be a trusted introducer. Although this attack appears to be mitigated by the fact that no one uses trusted introducers, if RNP were used in a caff-style work flow, then Mallory could cause Alice to believe that her own certification is actually a tsig. Consider: - Alice signs Bob's key, and sends the certification to Bob's email - Bob adds a Trust subpacket to Alice's certification - Bob uploads it to a key server - Alice refreshes her key and now considers Bob to be a CA As far as I know no one is currently using web of trust functionality with RNP, however, at least Thunderbird plans to. Signature Expiration Time ------------------------- By setting this on a signature that doesn't otherwise expire, an attacker can cause rnp to indicate that a signature is expired when, in reality, it doesn't expire. When applied to a signature over a document, the signature will appear invalid, because rnp will incorrectly consider the signature to be expired. Because most users will be unable to distinguish this from an invalid signature, which an attacker can create by flipping a bit in the data that is part of the hash, we consider the severity to be low. When applied to a self signature, this will have no effect. This is because rnp (incorrectly) appears to ignore the Signature Expiration Time when looking up a self signature. See [`pgp_key_latest_selfsig` (`src/lib/pgp-key.cpp`)][6]. [6] https://github.com/rnpgp/rnp/blob/6a5fb890eb30a59f7aa46b3e915a236d4bbe4176/src/lib/pgp-key.cpp#L840 Primary User ID --------------- An attacker can change what User ID is considered the primary User ID. This changes how the certificate is displayed. Revocable --------- An attacker could set this flag to indicate that a signature is irrevocable thereby causing rnp to consider a signature that has been revoked to be valid. Although the consequences of this attack are severe, signatures are not revoked in practice. As such, the impact is minor. Preferred Key Server -------------------- An attacker can set this to a URL that they control. This allows the attacker to observe when a user refreshes a certificate, and serve potentially malicious content. rnp does not appear to support fetching updates, so this would only impact applications that use rnp. Further, most applications do not respect this setting, because it can be abused as a type of tracker. Cryptographic preferences (Preferred Symmetric, Hash, Compression Algorithms) ----------------------------------------------------------------------------- An attacker can control what algorithms a sender will use when encrypting or signing a message. This could result in a downgrade attack. In practice all relevant signatures include these subpackets. As such, there are few, if any, keys where rnp's accessor would return an attacker-controller subpacket. Key Server Preferences ---------------------- This is only used by key servers, and only in an advisory manner. Features -------- In theory it could be used to perform a downgrade attack. But, the only feature flag that is currently defined is whether the key holder's OpenPGP implementation supports MDC packets, and, as far as I know, all implementations assume that this is the case anyway. Signer's User ID ---------------- This subpacket allows a keyholder to state which User ID is responsible for the signing. Most applications do not use this field. rnp does not itself use this field.

Created attachment 9169659 Screenshot 1: Error message comes too late, two accounts with same name already exist Seen on 78.1.1 (64-bit) As seen in screenshot (and in 3-pane folder list), I already succeeded to create two accounts with exactly the same name (POP & IMAP, both test accounts with disfunctional fake servers). At some random point in time, much later, when going through account settings, TB suddenly starts complaining after the fact. Too late, and very irritating. Expected: warn immediately when creating / renaming an account with a duplicate account name. I don't think this is a duplicate of some similar bugs, which seem to be about the technical impossibility of creating similar accounts.

From https://github.com/mozilla-mobile/fenix/issues/11785 ### Steps to reproduce 1. Go to www.Onlinegdb.com/online_c++_compiler 2. click on the centre text box 3. keyboard appears 4. press back 5. click again on the centre text box ### Expected behavior Keyboard appears. Note that for most other text boxes this is the current behaviour ### Actual behavior Nothing. ### Device information Android device: Asus Zenfone Fenix version: Nightly

It appears that if you have an existing `GeckoView` with an existing session on the screen, then swap that out to a new session via `GeckoView.setSession()`, the compositor in the new session is sometimes not correctly initialized and the `GeckoView` displays the content of the old session. We see the following in logs: ``` 2020-08-21 15:48:19.232 628-2439/? E/BufferQueueProducer: [SurfaceView - org.mozilla.fenix.debug/org.mozilla.fenix.debug.App#0] connect: already connected (cur=1 req=1) 2020-08-21 15:48:19.233 4369-4546/org.mozilla.fenix.debug E/libEGL: eglCreateWindowSurface: native_window_api_connect (win=0x76e7e9f010) failed (0xffffffea) (already connected to another API?) 2020-08-21 15:48:19.233 4369-4546/org.mozilla.fenix.debug E/libEGL: eglCreateWindowSurfaceTmpl:730 error 3003 (EGL_BAD_ALLOC) 2020-08-21 15:48:19.814 788-4684/? E/ResolverController: No valid NAT64 prefix (305, <unspecified>/0) 2020-08-21 15:48:31.663 2378-2486/? E/FirebaseMessaging: Topic operation failed: SERVICE_NOT_AVAILABLE. Will retry Topic operation. ``` This seems to indicate that we are not destroying the `EGLSurface` for the outgoing session before the incoming one is initialized.

It's pretty bad that we don't cover this in any existing unit tests.

Created attachment 9173482 consent_modal.html Attached is the text for the privacy notice acceptance modal that all users must accept (in the modal) in order to first enroll in Pioneer. I was given two URLs for the Manifesto and for the overall Mozilla Privacy Policy that do not have localization paths in them. Please check to ensure these are correct.

Created attachment 9173937 PrintToPdf.gif **Affected versions** * 82.0a1 (BuildId:20200903151816) * 81.0b6 (BuildId:20200903205131) **Affected platforms** * Windows 10 64bit * Ubuntu 18.04 64bit * macOS 10.14 **Steps to reproduce** 1. Launch Firefox. 2. Access any webpage. 3. Hit Ctrl + P 4. Select the Save to Pdf, Microsoft Print to PDF or the Microsoft XPS document destination. 5. Click Print/Save 6. Cancel the Save As window modal. **Expected result** * The UI remains intractable and displayed since the print action was canceled by the user. **Actual result** * The UI options are locked and the user has to close the UI by pressing the esc keyboard button. **Regression Window** * This is a regression. * Pushlog: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=284e3c053bf57ee8f331b21a90751936ee604482&tochange=0de10c26da552953484a56f0fbdb3ab70196988e * Possible Regressor: Bug 1659624 **Additional Information** * For further information regarding this issue please observe the attached screencast. * [Suggested Severity] S2

Created attachment 9173960 pdfJS.gif **Affected versions** * 82.0a1 (BuildId:20200903151816) * 81.0b6 (BuildId:20200903205131) **Affected platforms** * Windows 10 64bit * Ubuntu 18.04 64bit * macOS 10.14 **Steps to reproduce** 1. Launch Firefox 2. Access the following [link](http://www.dbsparks.com/research/html/SNRN02.pdf) 3. Click the print button. **Expected result** * The print preview successfully displays the pdf file. **Actual result** * The print preview displays a blank pdf and it cannot be printed. **Regression Window** * This seems to be a regression but I can’t reproduce this issue using mozregression (odd..) I’ll try to search for a manual regression as soon as time permits. **Additional Information** * For further information regarding this issue please observe the attached screencast.

In bug 1657220 I added new telemetry probes, but mistakenly set them to `record_in_processes: 'content'`. They need to be recorded in the main process.

Created attachment 9179379 58769.html This was initially reported in https://github.com/webcompat/web-bugs/issues/58769 Steps to reproduce (this requires an account on ebay.com): 1. Visit https://bulksell.ebay.com/ws/eBayISAPI.dll?SingleList in Firefox Nightly 201001 on Android and sign in 2. Click on "Create a template" button 3. Scroll down to the Photos section and click Add photos 4. Select a photo taken with a camera from a photo gallery Expected: Photo uploaded Actual: Photo doesn't upload, the upload process remains in loading state I've attached a reduced test case. The problem seems to be that in Firefox on Android `img.onload` never gets called, neither does `img.onerror` with large data URIs. The image I'm using is 3.7MB in size. This is probably related to https://bugzilla.mozilla.org/show_bug.cgi?id=1626687. I've noticed that uploading works for screenshots, but not for the photos, since they're generally greater in size.

Created attachment 9181520 Screenshot 2020-10-14 at 14.44.37.png User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.0 Safari/605.1.15 Steps to reproduce: 1. Go to https://badssl.com/ 2. Find http-login option and click it. You can observe this behavior in all GeckoView-based browsers so far. I see the same in Focus, Nightly and Fenix. Actual results: Application crashes. Expected results: Application must not crash.

Once nsIWebBrowserPrint gets a `rawNumPages` attribute (being added in bug 1660502), we need to update the page-range validation code to use that. Otherwise we end up in a catch-22 scenario. The existing `printPreviewNumPages` attribute is really poorly named given the way that this interface is evolving -- nowadays, it is really "how many sheets will be printed", and that's not the right value to use for page range validation. To see this, here's one scenario where this matters: Suppose e.g. you've got a 12-page document and the user chooses a range "from:6 to:10" (5 pages). Then after we've rendered their chosen selection, we'll suddenly start reporting a `printPreviewNumPages` of `5`, which then will cause the user's selected range to be invalid, since 6 and 10 are both out-of-bounds. Anyway - there's no better option right now, but once bug 1660502 has landed, we should migrate the range validation code to use the new `rawNumPages` API.

**Affected versions** * 82.0a1 (BuildId:20200824215021) * 81.0b1 (BuildId:20200824150741) **Affected platforms** * macOS 10.14 **Unaffected platforms** * Ubuntu 20.04 64bit * Windows 10 64bit **Steps to reproduce** 1. Launch Firefox. 2. Open any webpage. 3. Hit CTRL + P. 4. Tab through the options until you reach the Copies section. 5. Press the -> arrow key to remove the focus from the default value. 6. Try to delete the value via backspace. **Expected result** * Like in Windows & Ubuntu, the value can be successfully deleted even though it is not selected and the leading value starting with a different number (like ex: 2 for 20 copies) can be easily entered. **Actual result** * The default value (1) cannot be easily delete (the keyboard only user has to enter another value next to the default one and navigate to the first one in order to delete it || the keyboard user has to refocus the Copies field and to directly input the leading value). **Regression Window** * This seems to be a regression: * Pushlog: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=89bbbe5e6369e1ce2b75ddf75cd201782ce362b2&tochange=3ea0ca6e62185703a20b27fe1e7b7a062e9c4d2a * Potential regressor: Bug 1660296

Created attachment 9172109 Zipped example screenshots User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/84.0.4147.135 Safari/537.36 Edg/84.0.522.63 Steps to reproduce: 1. Enable high-contrast mode in Windows 10, by doing the following: 1.1. Open Windows Settings (Windows+I). 1.2. Go to Ease of access -> High-contrast. 1.3. Turn on the high-contrast mode, using the respective toggle. 1.4. From the "High-contrast theme" combo-box, select "High-Contrast Black". 2. Install and launch Thunderbird 78.2. Actual results: Observe how some dialogs and windows of Thunderbird do not obey the colors defined by the currently active high-contrast black theme. Mostly, where the background should be black, it is gray instead. I'll attach some example screenshots in a ZIP archive. Examples of such windows include, but may not be limited to, the following: - The main window of Thunderbird. - The dialogs for subscribing to IMAP folders and RSS feeds. - The dialogs for the properties of RSS and mail folders. - The dialog for customizing the toolbars in Thunderbird. - The dialog for creating a new folder. - The dialog/window for adding an account to Thunderbird. Expected results: In such dialogs and windows as in the examples above, the background color should be black, instead of being gray as it is now.

STR; 1 Open https://www.irs.gov/pub/irs-pdf/f1040.pdf 2 Click the print icon button in the pdf viewer

Created attachment 9172445 Screencast from 08-27-2020 09:43:40 AM.webm User Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:80.0) Gecko/20100101 Firefox/80.0 Steps to reproduce: When right clicking in the bottom right corner of the screen, the last entry of the context menu is automatically selected when releasing the right click. Happens only under wayland; when using xwayland the context menu shows as expected. Reproduced with firefox 80.0 and firefox nightly 82.0a1 on Fedora 32. Actual results: The screenshot tool show up on right click (see attached video). Expected results: The context menu should show up.

Created attachment 9177468 Saving_screenshot.png ### Steps to reproduce 1. Click the "Save to Pocket" address bar button. 2. BUG: The "Saving" door hanger" has a big white rectangle at the bottom. 3. Click the "Saved to Pocket" door hanger's "Remove Page" button. 4. BUG: The "Page Removed" door hanger has a small gray rectangle at the bottom. See the attached screenshots. This is a regression in Fx82.

Crash report: https://crash-stats.mozilla.org/report/index/018bc5a5-8939-4c79-bc26-fc2e20200924 Top 10 frames of crashing thread: ``` 0 xul.dll NS_ABORT_OOM xpcom/base/nsDebugImpl.cpp:620 1 xul.dll nsTArray_base<nsTArrayInfallibleAllocator, nsTArray_RelocateUsingMemutils>::EnsureCapacity<nsTArrayInfallibleAllocator> xpcom/ds/nsTArray-inl.h:154 2 xul.dll nsTArray_Impl<unsigned char, nsTArrayInfallibleAllocator>::SetLength<nsTArrayInfallibleAllocator> xpcom/ds/nsTArray.h:2198 3 xul.dll nsPrinterWin::DefaultSettings const widget/windows/nsPrinterWin.cpp:77 4 xul.dll static std::_C__Invoker_functor::_Call<`lambda at /builds/worker/checkouts/gecko/widget/PrintBackgroundTask.h:54:17'> 5 xul.dll mozilla::SpawnPrintBackgroundTask<nsPrinterBase, mozilla::PrintSettingsInitializer>::<unnamed-tag>::operator const widget/PrintBackgroundTask.h:53 6 xul.dll mozilla::detail::RunnableFunction<`lambda at /builds/worker/checkouts/gecko/widget/PrintBackgroundTask.h:48:11'>::Run xpcom/threads/nsThreadUtils.h:577 7 xul.dll nsThreadPool::Run xpcom/threads/nsThreadPool.cpp:299 8 xul.dll nsThread::ProcessNextEvent xpcom/threads/nsThread.cpp:1234 9 xul.dll mozilla::ipc::MessagePumpForNonMainThreads::Run ipc/glue/MessagePump.cpp:332 ``` These printing related OOM crashes look new in 82.

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:80.0) Gecko/20100101 Firefox/80.0 Steps to reproduce: 1. Open Thunderbird 78.3.0 (candidate). 2. Open `Tools|OpenPGP Key Manager`. 3. Select `File|Import Secret Key(s) From File`. 4. Select gpg-exported secret key file. 5. Ensure, that one key is found and marked with checkbox. 5. Click `Continue`. 6. On dialog with (successful results) click `Key Properties`. Actual results: Properties of some random public key from key storage are shown. Expected results: Properties of freshly-imported personal secret key are shows,

Created attachment 9178078 Ctrl+Click vs Right-click new-tab behavior User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:83.0) Gecko/20100101 Firefox/83.0 Steps to reproduce: 1. Ctrl+Click on "Top sites" from the default home page. Actual results: The newly opened tab got focused. "When you open a link in a new tab, switch to it immediately" is disabled in about:preferences Right click -> open in new tab did not focus on new tab. [mozregression directed](https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=fe060ab6199d15711dff0f2862554698651e939b&tochange=6e92ff608742ace8573f94bf867340a1401bd491) me to bug 1599368. Expected results: New tab should not have gained focus on Ctrl+Click.

At some point, the build system extracts symbols from binaries it's just built and then later uploads them to Tecken. For Windows builds, it runs makecab on files to compress them. This happens before uploading. At some point in the last 3 months (I can spend more time to figure out exactly when things changed), .dll files stopped being compressed. I did some spot checking and found these two uploads--somewhere between those two, things changed. This upload from July 8th, 2020 has compressed dlls: https://symbols.mozilla.org/uploads/upload/145567 It has this record: ``` v1/xul.dll/5F064109699B000/xul.dl_ 41.7 MB ``` This upload from August 7th, 2020 does not have compressed dlls: https://symbols.mozilla.org/uploads/upload/165577 It has this record: ``` v1/xul.dll/5F2D86426EB9000/xul.dll 109.9 MB ``` The consequence of this is that VS can no longer download dlls. It's trying to download: https://symbols.mozilla.org/xul.dll/5F517CAB659C000/xul.dll But "dll" is not in the allowed extensions list, so it gets back a 404 (ignored) message. This bug covers fixing something so VS works again.

Created attachment 9167053 about_welcome_00.gif * Opened on request on [bug 1653315](https://bugzilla.mozilla.org/show_bug.cgi?id=1653315#c9) *Affected versions* * 81.0a1 (20200729213824) *Affected platforms* * Windows 10x64 * Ubuntu 18.04 *Steps to reproduce* 1. Open Firefox and go to about:welcome. 2. Navigate through the welcome screens to the "Choose A Look" screen. 3. Select any random theme and then change it to another one. *Expected result* * Themes are changed as expected. *Actual result* * Themes section/ Save Theme/ Not now buttons are moved when changing the selection. *Regression range* * https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=3894c97b6b9479a3f189424d52c74281ecb5d17b&tochange=aafd087ec6adaf20816250dc410be7138dc5099f *Notes* * Attached a screen recording. * This happens when focusing/ defocusing the themes. * On macOS 10.12 the items are moving only when using the TAB key to select them and they are not moving when clicked. **Suggested Severity: S3**

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:79.0) Gecko/20100101 Firefox/79.0 Steps to reproduce: * Have a PGP-key with a User-ID packet Prename Surname <[email protected]> Notice the capitalization of the mail address. * Write a mail to "[email protected]" and turn on PGP encryption Actual results: * "View Security Info" shows the key with capitalized letters and the acceptance/trust level can be set. This suggests that encrypted mail can be sent to this person regardless of capitalization of their mail address * However, when clicking on "Send", an unspecific error message Sending of the message failed is shown. The console reveals an error in finishCryptoEncapsulation chrome://openpgp/content/modules/mimeEncrypt.jsm:588 Expected results: The encrypted message should have been sent. As an additional observation: when using the capitalized mail address (i.e. "[email protected]"), the correct security information is shown AND the encrypted message can be sent. Also note that this bug might be related to bug 1653763.

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:81.0) Gecko/20100101 Firefox/81.0 Steps to reproduce: Update from Android Gradle Plugin 3.5 to 4.0 and try to compile an app using Glean to generate code. See https://github.com/mozilla-mobile/fenix/pull/13271 for an example with logs, where all generated classes cannot be found. Actual results: Glean code is not generated Expected results: Glean code should be generated