- tbb-commits - lists.torproject.org

[Git][tpo/applications/tor-browser][tor-browser-115.1.0esr-13.0-1] 4 commits: Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs
by ma1 (＠ma1) 01 Aug '23

01 Aug '23

ma1 pushed to branch tor-browser-115.1.0esr-13.0-1 at The Tor Project / Applications / Tor Browser Commits: a9279174 by Edgar Chen at 2023-07-31T23:49:06+02:00 Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs Fullscreen/PointerLock warnings are initialized with hidden="true", but change to hidden="" after being shown and hidden again. I think this started happening when we began using HTML elements instead of XUL as they handle hidden attribute differently. Differential Revision: https://phabricator.services.mozilla.com/D177790 - - - - - 9f2eaedb by Edgar Chen at 2023-07-31T23:49:06+02:00 Bug 1821884 - Reshow initial fullscreen notification; r=Gijs Depends on D177790 Differential Revision: https://phabricator.services.mozilla.com/D178339 - - - - - 30d19aa0 by Eitan Isaacson at 2023-07-31T23:49:07+02:00 Bug 1819160 - Map Android ids to doc/accessible id pairs. r=Jamie Differential Revision: https://phabricator.services.mozilla.com/D179737 - - - - - efee8978 by Jon Coppeard at 2023-07-31T23:49:07+02:00 Bug 1828024 - Require the helper thread lock in the GC helper thread count getter r=sfink This makes us take a lock to read this state (we already lock when writing it). Also it adds a release assert in case something goes wrong with the thread count calculations, as a crash is preferable to the potential deadlock. Differential Revision: https://phabricator.services.mozilla.com/D181257 - - - - - 12 changed files: - accessible/android/SessionAccessibility.cpp - accessible/android/SessionAccessibility.h - accessible/ipc/DocAccessibleParent.cpp - accessible/ipc/DocAccessibleParent.h - accessible/ipc/moz.build - browser/base/content/browser-fullScreenAndPointerLock.js - browser/base/content/fullscreen-and-pointerlock.inc.xhtml - browser/base/content/test/fullscreen/browser_fullscreen_warning.js - dom/tests/browser/browser_pointerlock_warning.js - js/src/gc/GC.cpp - js/src/gc/ParallelMarking.cpp - js/src/vm/HelperThreadState.h Changes: ===================================== accessible/android/SessionAccessibility.cpp ===================================== @@ -269,12 +269,9 @@ RefPtr<SessionAccessibility> SessionAccessibility::GetInstanceFor( return GetInstanceFor(doc->GetPresShell()); } } else { - DocAccessibleParent* remoteDoc = aAccessible->AsRemote()->Document(); - if (remoteDoc->mSessionAccessibility) { - return remoteDoc->mSessionAccessibility; - } dom::CanonicalBrowsingContext* cbc = - static_cast<dom::BrowserParent*>(remoteDoc->Manager()) + static_cast<dom::BrowserParent*>( + aAccessible->AsRemote()->Document()->Manager()) ->GetBrowsingContext() ->Top(); dom::BrowserParent* bp = cbc->GetBrowserParent(); @@ -285,10 +282,7 @@ RefPtr<SessionAccessibility> SessionAccessibility::GetInstanceFor( if (auto element = bp->GetOwnerElement()) { if (auto doc = element->OwnerDoc()) { if (nsPresContext* presContext = doc->GetPresContext()) { - RefPtr<SessionAccessibility> sessionAcc = - GetInstanceFor(presContext->PresShell()); - remoteDoc->mSessionAccessibility = sessionAcc; - return sessionAcc; + return GetInstanceFor(presContext->PresShell()); } } else { MOZ_ASSERT_UNREACHABLE( @@ -684,14 +678,7 @@ void SessionAccessibility::PopulateNodeInfo( } Accessible* SessionAccessibility::GetAccessibleByID(int32_t aID) const { - Accessible* accessible = mIDToAccessibleMap.Get(aID); - if (accessible && accessible->IsLocal() && - accessible->AsLocal()->IsDefunct()) { - MOZ_ASSERT_UNREACHABLE("Registered accessible is defunct!"); - return nullptr; - } - - return accessible; + return mIDToAccessibleMap.Get(aID); } #ifdef DEBUG @@ -705,6 +692,58 @@ static bool IsDetachedDoc(Accessible* aAccessible) { } #endif +SessionAccessibility::IDMappingEntry::IDMappingEntry(Accessible* aAccessible) + : mInternalID(0) { + *this = aAccessible; +} + +SessionAccessibility::IDMappingEntry& +SessionAccessibility::IDMappingEntry::operator=(Accessible* aAccessible) { + mInternalID = aAccessible->ID(); + MOZ_ASSERT(!(mInternalID & IS_REMOTE), "First bit is used in accessible ID!"); + if (aAccessible->IsRemote()) { + mInternalID |= IS_REMOTE; + } + + Accessible* docAcc = nsAccUtils::DocumentFor(aAccessible); + MOZ_ASSERT(docAcc); + if (docAcc) { + MOZ_ASSERT(docAcc->IsRemote() == aAccessible->IsRemote()); + if (docAcc->IsRemote()) { + mDoc = docAcc->AsRemote()->AsDoc(); + } else { + mDoc = docAcc->AsLocal(); + } + } + + return *this; +} + +SessionAccessibility::IDMappingEntry::operator Accessible*() const { + if (mInternalID == 0) { + return static_cast<LocalAccessible*>(mDoc.get()); + } + + if (mInternalID == IS_REMOTE) { + return static_cast<DocAccessibleParent*>(mDoc.get()); + } + + if (mInternalID & IS_REMOTE) { + return static_cast<DocAccessibleParent*>(mDoc.get()) + ->GetAccessible(mInternalID & ~IS_REMOTE); + } + + Accessible* accessible = + static_cast<LocalAccessible*>(mDoc.get()) + ->AsDoc() + ->GetAccessibleByUniqueID(reinterpret_cast<void*>(mInternalID)); + // If the accessible is retrievable from the DocAccessible, it can't be + // defunct. + MOZ_ASSERT(!accessible->AsLocal()->IsDefunct()); + + return accessible; +} + void SessionAccessibility::RegisterAccessible(Accessible* aAccessible) { if (IPCAccessibilityActive()) { // Don't register accessible in content process. @@ -766,7 +805,6 @@ void SessionAccessibility::UnregisterAccessible(Accessible* aAccessible) { } RefPtr<SessionAccessibility> sessionAcc = GetInstanceFor(aAccessible); - MOZ_ASSERT(sessionAcc, "Need SessionAccessibility to unregister Accessible!"); if (sessionAcc) { Accessible* registeredAcc = sessionAcc->mIDToAccessibleMap.Get(virtualViewID); ===================================== accessible/android/SessionAccessibility.h ===================================== @@ -110,10 +110,34 @@ class SessionAccessibility final jni::NativeWeakPtr<widget::GeckoViewSupport> mWindow; // Parent only java::SessionAccessibility::NativeProvider::GlobalRef mSessionAccessibility; + class IDMappingEntry { + public: + explicit IDMappingEntry(Accessible* aAccessible); + + IDMappingEntry& operator=(Accessible* aAccessible); + + operator Accessible*() const; + + private: + // A strong reference to a DocAccessible or DocAccessibleParent. They don't + // share any useful base class except nsISupports, so we use that. + // When we retrieve the document from this reference we cast it to + // LocalAccessible in the DocAccessible case because DocAccessible has + // multiple inheritance paths for nsISupports. + RefPtr<nsISupports> mDoc; + // The ID of the accessible as used in the internal doc mapping. + // We rely on this ID being pointer derived and therefore divisible by two + // so we can use the first bit to mark if it is remote or not. + uint64_t mInternalID; + + static const uintptr_t IS_REMOTE = 0x1; + }; + /* * This provides a mapping from 32 bit id to accessible objects. */ - nsTHashMap<nsUint32HashKey, Accessible*> mIDToAccessibleMap; + nsBaseHashtable<nsUint32HashKey, IDMappingEntry, Accessible*> + mIDToAccessibleMap; }; } // namespace a11y ===================================== accessible/ipc/DocAccessibleParent.cpp ===================================== @@ -29,7 +29,6 @@ #endif #if defined(ANDROID) -# include "mozilla/a11y/SessionAccessibility.h" # define ACQUIRE_ANDROID_LOCK \ MonitorAutoLock mal(nsAccessibilityService::GetAndroidMonitor()); #else ===================================== accessible/ipc/DocAccessibleParent.h ===================================== @@ -29,10 +29,6 @@ class xpcAccessibleGeneric; class DocAccessiblePlatformExtParent; #endif -#ifdef ANDROID -class SessionAccessibility; -#endif - /* * These objects live in the main process and comunicate with and represent * an accessible document in a content process. @@ -348,10 +344,6 @@ class DocAccessibleParent : public RemoteAccessible, size_t SizeOfExcludingThis(MallocSizeOf aMallocSizeOf) override; -#ifdef ANDROID - RefPtr<SessionAccessibility> mSessionAccessibility; -#endif - private: ~DocAccessibleParent(); ===================================== accessible/ipc/moz.build ===================================== @@ -24,11 +24,6 @@ else: LOCAL_INCLUDES += [ "/accessible/mac", ] - elif CONFIG["MOZ_WIDGET_TOOLKIT"] == "android": - LOCAL_INCLUDES += [ - "/accessible/android", - "/widget/android", - ] else: LOCAL_INCLUDES += [ "/accessible/other", ===================================== browser/base/content/browser-fullScreenAndPointerLock.js ===================================== @@ -62,9 +62,14 @@ var PointerlockFsWarning = { this._element = document.getElementById(elementId); // Setup event listeners this._element.addEventListener("transitionend", this); + this._element.addEventListener("transitioncancel", this); window.addEventListener("mousemove", this, true); + window.addEventListener("activate", this); + window.addEventListener("deactivate", this); // The timeout to hide the warning box after a while. this._timeoutHide = new this.Timeout(() => { + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); this._state = "hidden"; }, timeout); // The timeout to show the warning box when the pointer is at the top @@ -116,11 +121,10 @@ var PointerlockFsWarning = { return; } - // Explicitly set the last state to hidden to avoid the warning - // box being hidden immediately because of mousemove. - this._state = "onscreen"; - this._lastState = "hidden"; - this._timeoutHide.start(); + if (Services.focus.activeWindow == window) { + this._state = "onscreen"; + this._timeoutHide.start(); + } }, /** @@ -148,7 +152,10 @@ var PointerlockFsWarning = { this._element.hidden = true; // Remove all event listeners this._element.removeEventListener("transitionend", this); + this._element.removeEventListener("transitioncancel", this); window.removeEventListener("mousemove", this, true); + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); // Clear fields this._element = null; this._timeoutHide = null; @@ -186,7 +193,7 @@ var PointerlockFsWarning = { } if (newState != "hidden") { if (currentState != "hidden") { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } else { // When the previous state is hidden, the display was none, // thus no box was constructed. We need to wait for the new @@ -197,7 +204,7 @@ var PointerlockFsWarning = { requestAnimationFrame(() => { requestAnimationFrame(() => { if (this._element) { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } }); }); @@ -217,7 +224,7 @@ var PointerlockFsWarning = { } else if (this._timeoutShow.delay >= 0) { this._timeoutShow.start(); } - } else { + } else if (state != "onscreen") { let elemRect = this._element.getBoundingClientRect(); if (state == "hiding" && this._lastState != "hidden") { // If we are on the hiding transition, and the pointer @@ -239,12 +246,23 @@ var PointerlockFsWarning = { } break; } - case "transitionend": { + case "transitionend": + case "transitioncancel": { if (this._state == "hiding") { this._element.hidden = true; } break; } + case "activate": { + this._state = "onscreen"; + this._timeoutHide.start(); + break; + } + case "deactivate": { + this._state = "hidden"; + this._timeoutHide.cancel(); + break; + } } }, }; ===================================== browser/base/content/fullscreen-and-pointerlock.inc.xhtml ===================================== @@ -3,7 +3,7 @@ # file, You can obtain one at http://mozilla.org/MPL/2.0/. <html:div id="fullscreen-and-pointerlock-wrapper"> - <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> @@ -20,7 +20,7 @@ </html:button> </html:div> - <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> ===================================== browser/base/content/test/fullscreen/browser_fullscreen_warning.js ===================================== @@ -3,14 +3,35 @@ "use strict"; -add_task(async function test_fullscreen_display_none() { +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + +add_setup(async function init() { await SpecialPowers.pushPrefEnv({ set: [ ["full-screen-api.enabled", true], ["full-screen-api.allow-trusted-requests-only", false], ], }); +}); +add_task(async function test_fullscreen_display_none() { await BrowserTestUtils.withNewTab( { gBrowser, @@ -30,11 +51,13 @@ add_task(async function test_fullscreen_display_none() { }, async function (browser) { let warning = document.getElementById("fullscreen-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( warning, - "true" + "hidden", + "Should not show full screen warning initially" ); + + let warningShownPromise = waitForWarningState(warning, "onscreen"); // Enter fullscreen await SpecialPowers.spawn(browser, [], async () => { let frame = content.document.querySelector("iframe"); @@ -54,39 +77,33 @@ add_task(async function test_fullscreen_display_none() { ); document.getElementById("fullscreen-exit-button").click(); await exitFullscreenPromise; + + checkWarningState( + warning, + "hidden", + "Should hide fullscreen warning after exiting fullscreen" + ); } ); }); add_task(async function test_fullscreen_pointerlock_conflict() { - await SpecialPowers.pushPrefEnv({ - set: [ - ["full-screen-api.enabled", true], - ["full-screen-api.allow-trusted-requests-only", false], - ], - }); - await BrowserTestUtils.withNewTab("https://example.com", async browser => { let fsWarning = document.getElementById("fullscreen-warning"); let plWarning = document.getElementById("pointerlock-warning"); - is( - fsWarning.getAttribute("onscreen"), - null, - "Should not show full screen warning initially." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning initially." - ); - - let fsWarningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( fsWarning, - "true" + "hidden", + "Should not show full screen warning initially" + ); + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning initially" ); + let fsWarningShownPromise = waitForWarningState(fsWarning, "onscreen"); info("Entering full screen and pointer lock."); await SpecialPowers.spawn(browser, [], async () => { await content.document.body.requestFullscreen(); @@ -94,15 +111,10 @@ add_task(async function test_fullscreen_pointerlock_conflict() { }); await fsWarningShownPromise; - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should show full screen warning." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); info("Exiting pointerlock"); @@ -110,18 +122,19 @@ add_task(async function test_fullscreen_pointerlock_conflict() { await content.document.exitPointerLock(); }); - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should still show full screen warning." + checkWarningState( + fsWarning, + "onscreen", + "Should still show full screen warning" ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); // Cleanup + info("Exiting fullscreen"); await document.exitFullscreen(); }); }); ===================================== dom/tests/browser/browser_pointerlock_warning.js ===================================== @@ -15,6 +15,25 @@ const FRAME_TEST_URL = encodeURI(BODY_URL) + '"></iframe></body>'; +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + // Make sure the pointerlock warning is shown and exited with the escape key add_task(async function show_pointerlock_warning_escape() { let urls = [TEST_URL, FRAME_TEST_URL]; @@ -24,11 +43,7 @@ add_task(async function show_pointerlock_warning_escape() { let tab = await BrowserTestUtils.openNewForegroundTab(gBrowser, url); let warning = document.getElementById("pointerlock-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", - warning, - "true" - ); + let warningShownPromise = waitForWarningState(warning, "onscreen"); let expectedWarningText; @@ -49,11 +64,7 @@ add_task(async function show_pointerlock_warning_escape() { ok(true, "Pointerlock warning shown"); - let warningHiddenPromise = BrowserTestUtils.waitForAttribute( - "hidden", - warning, - "" - ); + let warningHiddenPromise = waitForWarningState(warning, "hidden"); await BrowserTestUtils.waitForCondition( () => warning.innerText == expectedWarningText, ===================================== js/src/gc/GC.cpp ===================================== @@ -1331,6 +1331,11 @@ void GCRuntime::assertNoMarkingWork() const { } #endif +static size_t GetGCParallelThreadCount() { + AutoLockHelperThreadState lock; + return HelperThreadState().getGCParallelThreadCount(lock); +} + bool GCRuntime::updateMarkersVector() { MOZ_ASSERT(helperThreadCount >= 1, "There must always be at least one mark task"); @@ -1339,8 +1344,8 @@ bool GCRuntime::updateMarkersVector() { // Limit worker count to number of GC parallel tasks that can run // concurrently, otherwise one thread can deadlock waiting on another. - size_t targetCount = std::min(markingWorkerCount(), - HelperThreadState().getGCParallelThreadCount()); + size_t targetCount = + std::min(markingWorkerCount(), GetGCParallelThreadCount()); if (markers.length() > targetCount) { return markers.resize(targetCount); ===================================== js/src/gc/ParallelMarking.cpp ===================================== @@ -103,6 +103,10 @@ bool ParallelMarker::markOneColor(MarkColor color, SliceBudget& sliceBudget) { { AutoLockHelperThreadState lock; + // There should always be enough parallel tasks to run our marking work. + MOZ_RELEASE_ASSERT(HelperThreadState().getGCParallelThreadCount(lock) >= + workerCount()); + for (size_t i = 0; i < workerCount(); i++) { gc->startTask(*tasks[i], lock); } ===================================== js/src/vm/HelperThreadState.h ===================================== @@ -333,9 +333,11 @@ class GlobalHelperThreadState { GCParallelTaskList& gcParallelWorklist() { return gcParallelWorklist_; } - size_t getGCParallelThreadCount() const { return gcParallelThreadCount; } + size_t getGCParallelThreadCount(const AutoLockHelperThreadState& lock) const { + return gcParallelThreadCount; + } void setGCParallelThreadCount(size_t count, - const AutoLockHelperThreadState&) { + const AutoLockHelperThreadState& lock) { MOZ_ASSERT(count >= 1); MOZ_ASSERT(count <= threadCount); gcParallelThreadCount = count; View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/c4e8cd… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/c4e8cd… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser-build][main] Bug 40905: Avoid replacing .*browser-testbuild targets
by boklm (＠boklm) 01 Aug '23

01 Aug '23

boklm pushed to branch main at The Tor Project / Applications / tor-browser-build Commits: d8fe33ad by Nicolas Vigier at 2023-08-01T10:53:55+02:00 Bug 40905: Avoid replacing .*browser-testbuild targets In some places we use target_replace to replace .*browser-.* targets, in order to replace the torbrowser-$os-$arch targets. However the regexp we used would also replace the torbrowser-testbuild target, so we need to update the regexp to exclude the testbuild target. - - - - - 7 changed files: - projects/conjure/config - projects/go/config - projects/lyrebird/config - projects/snowflake/config - projects/tor-android-service/config - projects/tor-onion-proxy-library/config - projects/webtunnel/config Changes: ===================================== projects/conjure/config ===================================== @@ -21,4 +21,4 @@ steps: norec: sha256sum: 2b403d6edf075777003bf2194a43fb178a28a4eaa7d23ec8f104563d9bbd7e53 target_replace: - '^torbrowser-.*': 'torbrowser-linux-x86_64' + '^torbrowser-(?!testbuild).*': 'torbrowser-linux-x86_64' ===================================== projects/go/config ===================================== @@ -123,6 +123,6 @@ input_files: - project: go-bootstrap name: go-bootstrap target_replace: - '^.*browser-.*': 'basebrowser-linux-x86_64' + '^.*browser-(?!testbuild).*': 'basebrowser-linux-x86_64' - filename: 0001-Use-fixed-go-build-tmp-directory.patch enable: '[% c("var/android") %]' ===================================== projects/lyrebird/config ===================================== @@ -35,4 +35,4 @@ steps: norec: sha256sum: '[% c("var/go_vendor_sha256sum") %]' target_replace: - '^torbrowser-.*': 'torbrowser-linux-x86_64' + '^torbrowser-(?!testbuild).*': 'torbrowser-linux-x86_64' ===================================== projects/snowflake/config ===================================== @@ -22,4 +22,4 @@ steps: norec: sha256sum: 62f9065881f5a3cbe5ea5a9802e961e1821d9e468cb2a85ebab1e8afcc0cc953 target_replace: - '^torbrowser-.*': 'torbrowser-linux-x86_64' + '^torbrowser-(?!testbuild).*': 'torbrowser-linux-x86_64' ===================================== projects/tor-android-service/config ===================================== @@ -26,19 +26,19 @@ input_files: - project: tor-expert-bundle name: tor-expert-bundle-armv7 target_replace: - '^torbrowser-.*': 'torbrowser-android-armv7' + '^torbrowser-(?!testbuild).*': 'torbrowser-android-armv7' - project: tor-expert-bundle name: tor-expert-bundle-aarch64 target_replace: - '^torbrowser-.*': 'torbrowser-android-aarch64' + '^torbrowser-(?!testbuild).*': 'torbrowser-android-aarch64' - project: tor-expert-bundle name: tor-expert-bundle-x86 target_replace: - '^torbrowser-.*': 'torbrowser-android-x86' + '^torbrowser-(?!testbuild).*': 'torbrowser-android-x86' - project: tor-expert-bundle name: tor-expert-bundle-x86_64 target_replace: - '^torbrowser-.*': 'torbrowser-android-x86_64' + '^torbrowser-(?!testbuild).*': 'torbrowser-android-x86_64' - URL: 'https://dl.google.com/dl/android/studio/jetifier-zips/1.0.0-beta10/jetifier…' name: jetifier sha256sum: 38186db9c9d1b745890b3d35c0667da1cac146ceb3c26aae5bf0802119472c1b ===================================== projects/tor-onion-proxy-library/config ===================================== @@ -21,19 +21,19 @@ input_files: - project: tor-expert-bundle name: tor-expert-bundle-armv7 target_replace: - '^torbrowser-.*': 'torbrowser-android-armv7' + '^torbrowser-(?!testbuild).*': 'torbrowser-android-armv7' - project: tor-expert-bundle name: tor-expert-bundle-aarch64 target_replace: - '^torbrowser-.*': 'torbrowser-android-aarch64' + '^torbrowser-(?!testbuild).*': 'torbrowser-android-aarch64' - project: tor-expert-bundle name: tor-expert-bundle-x86 target_replace: - '^torbrowser-.*': 'torbrowser-android-x86' + '^torbrowser-(?!testbuild).*': 'torbrowser-android-x86' - project: tor-expert-bundle name: tor-expert-bundle-x86_64 target_replace: - '^torbrowser-.*': 'torbrowser-android-x86_64' + '^torbrowser-(?!testbuild).*': 'torbrowser-android-x86_64' - filename: 'gradle-dependencies-[% c("var/gradle_dependencies_version") %]' name: gradle-dependencies exec: '[% INCLUDE "fetch-gradle-dependencies" %]' ===================================== projects/webtunnel/config ===================================== @@ -21,4 +21,4 @@ steps: norec: sha256sum: e3b5a9b3c3939aafa5389246f3a7a7e78d70fe623bed495f99c39cc37bbbe645 target_replace: - '^torbrowser-.*': 'torbrowser-linux-x86_64' + '^torbrowser-(?!testbuild).*': 'torbrowser-linux-x86_64' View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/d… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/d… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser-build][main] Bug 40615: Add a README for the bundled font directory.
by Pier Angelo Vendrame (＠pierov) 01 Aug '23

01 Aug '23

Pier Angelo Vendrame pushed to branch main at The Tor Project / Applications / tor-browser-build Commits: 571d86aa by FlexFoot at 2023-08-01T08:47:20+00:00 Bug 40615: Add a README for the bundled font directory. Include a README in the final fonts directory to discourage users from modifying the bundled fonts and to warn them about the consequences this could have. - - - - - 3 changed files: - + projects/fonts/README.txt - projects/fonts/build - projects/fonts/config Changes: ===================================== projects/fonts/README.txt ===================================== @@ -0,0 +1,7 @@ +DO NOT MODIFY THE CONTENTS OF THIS DIRECTORY + +Any adjustment to bundled fonts will result in an altered fingerprint. Font +fingerprinting is more than just detecting what fonts you have, it also includes +font fallbacks and characters (unicode code points) and any change in those can +be measured. + ===================================== projects/fonts/build ===================================== @@ -32,6 +32,7 @@ mv noto-fonts-* noto-fonts [% IF c("var/linux") %] cp {NotoSansJP-Regular.otf,NotoSansKR-Regular.otf,NotoSansSC-Regular.otf,NotoSansTC-Regular.otf} $distdir/ [% END %] +cp README.txt "$distdir/000_README.txt" cd /var/tmp/dist [% c('tar', { tar_src => [ 'fonts' ], ===================================== projects/fonts/config ===================================== @@ -161,6 +161,7 @@ var: input_files: - project: container-image + - filename: README.txt - filename: 'noto-fonts-[% c("var/noto_git_hash") %]-[% c("version") %]' name: noto-fonts exec: '[% INCLUDE "fetch-noto-fonts" %]' View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/5… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/5… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser-build][maint-12.5] Bug 31546: Copy Firefox PDBs for Windows
by richard (＠richard) 31 Jul '23

31 Jul '23

richard pushed to branch maint-12.5 at The Tor Project / Applications / tor-browser-build Commits: 3ba21986 by Pier Angelo Vendrame at 2023-07-31T20:37:13+00:00 Bug 31546: Copy Firefox PDBs for Windows Also copy debug symbols after stripping on Linux 32-bit (we only copied them in Linux 64-bit) and for all our browsers (previously we copied them only for Tor Browser). Include the headers directory with the symbols, because some of them are generated during the build, but they are needed for debugging. - - - - - 2 changed files: - projects/browser/build - projects/firefox/build Changes: ===================================== projects/browser/build ===================================== @@ -397,13 +397,13 @@ SCRIPT_EOF [% IF c("var/updater_enabled") -%] cp $rootdir/[% c('input_files_by_name/firefox') %]/mar-tools-*.zip "$OUTDIR"/ [% END -%] -[% IF c("var/linux-x86_64") -%] - [% IF c("var/tor-browser") -%] - cp $rootdir/[% c('input_files_by_name/firefox') %]/browser-debug.tar.xz "$OUTDIR"/[% c("var/project-name") %]-[% c("var/mar_osname") %]-debug.tar.xz - [% END -%] - [% IF !c("var/asan") -%] - cp $rootdir/[% c('input_files_by_name/firefox') %]/geckodriver-linux64.tar.xz "$OUTDIR"/ +[% IF c("var/linux") -%] + cp $rootdir/[% c('input_files_by_name/firefox') %]/browser-debug.tar.xz "$OUTDIR/[% c('var/project-name') %]-[% c('var/mar_osname') %]-debug.tar.xz" + [% IF c("var/linux-x86_64") && !c("var/asan") -%] + cp $rootdir/[% c('input_files_by_name/firefox') %]/geckodriver.tar.xz "$OUTDIR/geckodriver-[% c('var/mar_osname') %].tar.xz" [% END -%] +[% ELSIF c("var/windows") -%] + cp $rootdir/[% c('input_files_by_name/firefox') %]/browser-debug.zip "$OUTDIR/[% c('var/project-name') %]-[% c('var/mar_osname') %]-debug.zip" [% END -%] [%IF c("var/tor-browser") -%] tor_expert_bundle_src="[% c("input_files_by_name/tor-expert-bundle") %]" ===================================== projects/firefox/build ===================================== @@ -211,6 +211,10 @@ export LANG=C.UTF-8 cp obj-*/testing/geckodriver/x86_64-unknown-linux-gnu/release/geckodriver $distdir [% END %] cp -a obj-*/dist/[% c('var/exe_name') %]/* $distdir/Browser/ + mkdir -p $distdir/Debug + # Some include files are symlinks, so use -Lr, or the tarball will fail + # silently. Also, on Linux we populate the debug symbols by stripping later. + cp -Lr obj-*/dist/include $distdir/Debug/ # Remove firefox-bin (we don't use it, see ticket #10126) rm -f "$distdir/Browser/[% c('var/exe_name') %]-bin" # TODO: There goes FIPS-140.. We could upload these somewhere unique and @@ -232,6 +236,11 @@ RBM_TB_EOF [% ELSE %] cp -a /var/tmp/dist/fxc2/bin/d3dcompiler_47.dll $distdir/Browser [% END %] + mkdir -p $distdir/Debug/Browser + pushd obj-* + cp -Lr dist/include $distdir/Debug/ + find . $ -path ./dist -o -path ./_tests $ -prune -o -name '*.pdb' -exec cp -l {} $distdir/Debug/Browser/ \; + popd [% END %] [% IF c("var/updater_enabled") -%] @@ -279,8 +288,8 @@ RBM_TB_EOF cd $distdir -[% IF c("var/linux-x86_64") %] - [% IF !c("var/asan") %] +[% IF c("var/linux") -%] + [% IF c("var/linux-x86_64") && !c("var/asan") -%] # No need for an unstripped geckodriver strip geckodriver [% END %] @@ -334,17 +343,22 @@ END; tar_args => '-czf ' _ dest_dir _ '/' _ c('filename') _ '/browser.tar.gz', }) %] -[% IF c("var/linux-x86_64") %] +[% IF c("var/linux") -%] [% c('tar', { tar_src => [ 'Debug' ], tar_args => '-cJf ' _ dest_dir _ '/' _ c('filename') _ '/browser-debug.tar.xz', }) %] - [% IF !c("var/asan") %] + [% IF c("var/linux-x86_64") && !c("var/asan") -%] [% c('tar', { tar_src => [ 'geckodriver' ], tar_args => '-cJf ' _ dest_dir _ '/' _ c('filename') _ '/geckodriver-linux64.tar.xz', }) %] [% END %] +[% ELSIF c("var/windows") -%] + [% c('zip', { + zip_src => [ 'Debug' ], + zip_args => dest_dir _ '/' _ c('filename') _ '/browser-debug.zip', + }) %] [% END %] [% IF c("var/updater_enabled") -%] View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/3… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/3… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser-build][maint-12.5] Bug 40909: Add dan_b and ma1 to list of taggers in relevant projects
by richard (＠richard) 31 Jul '23

31 Jul '23

richard pushed to branch maint-12.5 at The Tor Project / Applications / tor-browser-build Commits: 5f27b741 by Richard Pospesel at 2023-07-31T19:23:09+00:00 Bug 40909: Add dan_b and ma1 to list of taggers in relevant projects - - - - - 4 changed files: - projects/android-components/config - projects/fenix/config - projects/firefox/config - projects/geckoview/config Changes: ===================================== projects/android-components/config ===================================== @@ -5,6 +5,8 @@ git_hash: '[% project %]-[% c("var/android_components_version") %]-[% c("var/bro git_url: https://gitlab.torproject.org/tpo/applications/android-components.git tag_gpg_id: 1 gpg_keyring: + - dan_b.gpg + - ma1.gpg - pierov.gpg - richard.gpg variant: '[% IF c("var/release") %]Release[% ELSE %]Beta[% END %]' ===================================== projects/fenix/config ===================================== @@ -5,6 +5,8 @@ git_hash: 'tor-browser-[% c("var/fenix_version") %]-[% c("var/browser_branch") % git_url: https://gitlab.torproject.org/tpo/applications/fenix.git tag_gpg_id: 1 gpg_keyring: + - dan_b.gpg + - ma1.gpg - pierov.gpg - richard.gpg variant: Beta ===================================== projects/firefox/config ===================================== @@ -5,6 +5,8 @@ git_hash: '[% c("var/project-name") %]-[% c("var/firefox_version") %]-[% c("var/ tag_gpg_id: 1 git_url: https://gitlab.torproject.org/tpo/applications/tor-browser.git gpg_keyring: + - dan_b.gpg + - ma1.gpg - pierov.gpg - richard.gpg container: ===================================== projects/geckoview/config ===================================== @@ -5,6 +5,8 @@ git_hash: 'tor-browser-[% c("var/geckoview_version") %]-[% c("var/browser_branch tag_gpg_id: 1 git_url: https://gitlab.torproject.org/tpo/applications/tor-browser.git gpg_keyring: + - dan_b.gpg + - ma1.gpg - pierov.gpg - richard.gpg container: View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/5… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/5… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser-build][main] Bug 40909: Add dan_b and ma1 to list of taggers in relevant projects
by richard (＠richard) 31 Jul '23

31 Jul '23

richard pushed to branch main at The Tor Project / Applications / tor-browser-build Commits: 13122472 by Richard Pospesel at 2023-07-31T19:19:03+00:00 Bug 40909: Add dan_b and ma1 to list of taggers in relevant projects - - - - - 4 changed files: - projects/android-components/config - projects/fenix/config - projects/firefox/config - projects/geckoview/config Changes: ===================================== projects/android-components/config ===================================== @@ -5,6 +5,8 @@ git_hash: '[% project %]-[% c("var/android_components_version") %]-[% c("var/bro git_url: https://gitlab.torproject.org/tpo/applications/android-components.git tag_gpg_id: 1 gpg_keyring: + - dan_b.gpg + - ma1.gpg - pierov.gpg - richard.gpg variant: '[% IF c("var/release") %]Release[% ELSE %]Beta[% END %]' ===================================== projects/fenix/config ===================================== @@ -5,6 +5,8 @@ git_hash: 'tor-browser-[% c("var/fenix_version") %]-[% c("var/browser_branch") % git_url: https://gitlab.torproject.org/tpo/applications/fenix.git tag_gpg_id: 1 gpg_keyring: + - dan_b.gpg + - ma1.gpg - pierov.gpg - richard.gpg variant: Beta ===================================== projects/firefox/config ===================================== @@ -5,6 +5,8 @@ git_hash: '[% c("var/project-name") %]-[% c("var/firefox_version") %]-[% c("var/ tag_gpg_id: 1 git_url: https://gitlab.torproject.org/tpo/applications/tor-browser.git gpg_keyring: + - dan_b.gpg + - ma1.gpg - pierov.gpg - richard.gpg container: ===================================== projects/geckoview/config ===================================== @@ -5,6 +5,8 @@ git_hash: 'tor-browser-[% c("var/geckoview_version") %]-[% c("var/browser_branch tag_gpg_id: 1 git_url: https://gitlab.torproject.org/tpo/applications/tor-browser.git gpg_keyring: + - dan_b.gpg + - ma1.gpg - pierov.gpg - richard.gpg container: View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/1… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/commit/1… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser] Pushed new tag base-browser-102.14.0esr-12.5-1-build2
by ma1 (＠ma1) 31 Jul '23

31 Jul '23

ma1 pushed new tag base-browser-102.14.0esr-12.5-1-build2 at The Tor Project / Applications / Tor Browser -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/tree/base-brow… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][base-browser-102.14.0esr-12.5-1] 2 commits: Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs
by ma1 (＠ma1) 31 Jul '23

31 Jul '23

ma1 pushed to branch base-browser-102.14.0esr-12.5-1 at The Tor Project / Applications / Tor Browser Commits: f24d6cd4 by Edgar Chen at 2023-07-31T21:15:57+02:00 Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs Fullscreen/PointerLock warnings are initialized with hidden="true", but change to hidden="" after being shown and hidden again. I think this started happening when we began using HTML elements instead of XUL as they handle hidden attribute differently. Differential Revision: https://phabricator.services.mozilla.com/D177790 - - - - - 53142727 by Edgar Chen at 2023-07-31T21:16:28+02:00 Bug 1821884 - Reshow initial fullscreen notification; r=Gijs Depends on D177790 Differential Revision: https://phabricator.services.mozilla.com/D178339 - - - - - 4 changed files: - browser/base/content/browser-fullScreenAndPointerLock.js - browser/base/content/fullscreen-and-pointerlock.inc.xhtml - browser/base/content/test/fullscreen/browser_fullscreen_warning.js - dom/tests/browser/browser_pointerlock_warning.js Changes: ===================================== browser/base/content/browser-fullScreenAndPointerLock.js ===================================== @@ -62,9 +62,14 @@ var PointerlockFsWarning = { this._element = document.getElementById(elementId); // Setup event listeners this._element.addEventListener("transitionend", this); + this._element.addEventListener("transitioncancel", this); window.addEventListener("mousemove", this, true); + window.addEventListener("activate", this); + window.addEventListener("deactivate", this); // The timeout to hide the warning box after a while. this._timeoutHide = new this.Timeout(() => { + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); this._state = "hidden"; }, timeout); // The timeout to show the warning box when the pointer is at the top @@ -116,11 +121,10 @@ var PointerlockFsWarning = { return; } - // Explicitly set the last state to hidden to avoid the warning - // box being hidden immediately because of mousemove. - this._state = "onscreen"; - this._lastState = "hidden"; - this._timeoutHide.start(); + if (Services.focus.activeWindow == window) { + this._state = "onscreen"; + this._timeoutHide.start(); + } }, /** @@ -148,7 +152,10 @@ var PointerlockFsWarning = { this._element.hidden = true; // Remove all event listeners this._element.removeEventListener("transitionend", this); + this._element.removeEventListener("transitioncancel", this); window.removeEventListener("mousemove", this, true); + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); // Clear fields this._element = null; this._timeoutHide = null; @@ -186,7 +193,7 @@ var PointerlockFsWarning = { } if (newState != "hidden") { if (currentState != "hidden") { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } else { // When the previous state is hidden, the display was none, // thus no box was constructed. We need to wait for the new @@ -197,7 +204,7 @@ var PointerlockFsWarning = { requestAnimationFrame(() => { requestAnimationFrame(() => { if (this._element) { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } }); }); @@ -217,7 +224,7 @@ var PointerlockFsWarning = { } else if (this._timeoutShow.delay >= 0) { this._timeoutShow.start(); } - } else { + } else if (state != "onscreen") { let elemRect = this._element.getBoundingClientRect(); if (state == "hiding" && this._lastState != "hidden") { // If we are on the hiding transition, and the pointer @@ -239,12 +246,23 @@ var PointerlockFsWarning = { } break; } - case "transitionend": { + case "transitionend": + case "transitioncancel": { if (this._state == "hiding") { this._element.hidden = true; } break; } + case "activate": { + this._state = "onscreen"; + this._timeoutHide.start(); + break; + } + case "deactivate": { + this._state = "hidden"; + this._timeoutHide.cancel(); + break; + } } }, }; ===================================== browser/base/content/fullscreen-and-pointerlock.inc.xhtml ===================================== @@ -3,7 +3,7 @@ # file, You can obtain one at http://mozilla.org/MPL/2.0/. <html:div id="fullscreen-and-pointerlock-wrapper"> - <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> @@ -20,7 +20,7 @@ </html:button> </html:div> - <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> ===================================== browser/base/content/test/fullscreen/browser_fullscreen_warning.js ===================================== @@ -3,14 +3,35 @@ "use strict"; -add_task(async function test_fullscreen_display_none() { +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + +add_setup(async function init() { await SpecialPowers.pushPrefEnv({ set: [ ["full-screen-api.enabled", true], ["full-screen-api.allow-trusted-requests-only", false], ], }); +}); +add_task(async function test_fullscreen_display_none() { await BrowserTestUtils.withNewTab( { gBrowser, @@ -30,11 +51,13 @@ add_task(async function test_fullscreen_display_none() { }, async function(browser) { let warning = document.getElementById("fullscreen-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( warning, - "true" + "hidden", + "Should not show full screen warning initially" ); + + let warningShownPromise = waitForWarningState(warning, "onscreen"); // Enter fullscreen await SpecialPowers.spawn(browser, [], async () => { let frame = content.document.querySelector("iframe"); @@ -54,39 +77,33 @@ add_task(async function test_fullscreen_display_none() { ); document.getElementById("fullscreen-exit-button").click(); await exitFullscreenPromise; + + checkWarningState( + warning, + "hidden", + "Should hide fullscreen warning after exiting fullscreen" + ); } ); }); add_task(async function test_fullscreen_pointerlock_conflict() { - await SpecialPowers.pushPrefEnv({ - set: [ - ["full-screen-api.enabled", true], - ["full-screen-api.allow-trusted-requests-only", false], - ], - }); - await BrowserTestUtils.withNewTab("https://example.com", async browser => { let fsWarning = document.getElementById("fullscreen-warning"); let plWarning = document.getElementById("pointerlock-warning"); - is( - fsWarning.getAttribute("onscreen"), - null, - "Should not show full screen warning initially." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning initially." - ); - - let fsWarningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( fsWarning, - "true" + "hidden", + "Should not show full screen warning initially" + ); + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning initially" ); + let fsWarningShownPromise = waitForWarningState(fsWarning, "onscreen"); info("Entering full screen and pointer lock."); await SpecialPowers.spawn(browser, [], async () => { await content.document.body.requestFullscreen(); @@ -94,15 +111,10 @@ add_task(async function test_fullscreen_pointerlock_conflict() { }); await fsWarningShownPromise; - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should show full screen warning." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); info("Exiting pointerlock"); @@ -110,18 +122,19 @@ add_task(async function test_fullscreen_pointerlock_conflict() { await content.document.exitPointerLock(); }); - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should still show full screen warning." + checkWarningState( + fsWarning, + "onscreen", + "Should still show full screen warning" ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); // Cleanup + info("Exiting fullscreen"); await document.exitFullscreen(); }); }); ===================================== dom/tests/browser/browser_pointerlock_warning.js ===================================== @@ -15,6 +15,25 @@ const FRAME_TEST_URL = encodeURI(BODY_URL) + '"></iframe></body>'; +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + // Make sure the pointerlock warning is shown and exited with the escape key add_task(async function show_pointerlock_warning_escape() { let urls = [TEST_URL, FRAME_TEST_URL]; @@ -24,11 +43,7 @@ add_task(async function show_pointerlock_warning_escape() { let tab = await BrowserTestUtils.openNewForegroundTab(gBrowser, url); let warning = document.getElementById("pointerlock-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", - warning, - "true" - ); + let warningShownPromise = waitForWarningState(warning, "onscreen"); let expectedWarningText; @@ -49,11 +64,7 @@ add_task(async function show_pointerlock_warning_escape() { ok(true, "Pointerlock warning shown"); - let warningHiddenPromise = BrowserTestUtils.waitForAttribute( - "hidden", - warning, - "" - ); + let warningHiddenPromise = waitForWarningState(warning, "hidden"); await BrowserTestUtils.waitForCondition( () => warning.innerText == expectedWarningText, View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/fadc59… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/fadc59… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser] Pushed new tag tor-browser-102.14.0esr-12.5-1-build2
by ma1 (＠ma1) 31 Jul '23

31 Jul '23

ma1 pushed new tag tor-browser-102.14.0esr-12.5-1-build2 at The Tor Project / Applications / Tor Browser -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/tree/tor-brows… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][tor-browser-115.1.0esr-13.0-1] 9 commits: fixup! Add TorStrings module for localization
by Pier Angelo Vendrame (＠pierov) 31 Jul '23

31 Jul '23

Pier Angelo Vendrame pushed to branch tor-browser-115.1.0esr-13.0-1 at The Tor Project / Applications / Tor Browser Commits: c3d2496b by Pier Angelo Vendrame at 2023-07-27T21:07:52+02:00 fixup! Add TorStrings module for localization Move the `getLocale` function here from TorButton util.js and stop importing it. - - - - - 642df684 by Pier Angelo Vendrame at 2023-07-27T21:07:53+02:00 fixup! Bug 40933: Add tor-launcher functionality Fix a couple of problems in TorLauncherUtil and TorParsers. Also, moved the function to parse bridges in TorParsers. - - - - - 9e825b61 by Pier Angelo Vendrame at 2023-07-27T21:07:53+02:00 fixup! Bug 31286: Implementation of bridge, proxy, and firewall settings in about:preferences#connection Use the bridge line parser from TorParsers. - - - - - 8061f810 by Pier Angelo Vendrame at 2023-07-27T21:07:53+02:00 fixup! Bug 40933: Add tor-launcher functionality Use actual private members for TorProcess. - - - - - f5a3f4af by Pier Angelo Vendrame at 2023-07-27T21:07:54+02:00 fixup! Bug 40933: Add tor-launcher functionality Group arg pushes in one line (keys and values together). - - - - - 71b08c4e by Pier Angelo Vendrame at 2023-07-27T21:07:54+02:00 fixup! Bug 40933: Add tor-launcher functionality TorProcess: use real private properties instead of _, and removed the dependency on TorProtocolService (temporarily moved it to TorMonitorService, but eventually we should unify TorProtocolService and TorMonitorService, to then split them again in a smarter way). - - - - - e1a69b4e by Pier Angelo Vendrame at 2023-07-27T21:07:55+02:00 fixup! Bug 10760: Integrate TorButton to TorBrowser core Move the SOCKS preference updater to TorProtocolService. We will need to refactor all this kind of stuff, but at least let's get it in a single place. Also, since this was the last bit of the startup service, remove the file, the component and what else was needed to add them. - - - - - d457b6f8 by Pier Angelo Vendrame at 2023-07-31T20:42:54+02:00 fixup! Bug 40933: Add tor-launcher functionality Hashing the control port password is needed only on the process, so move this function to TorProcess. - - - - - c4e8cd0f by Pier Angelo Vendrame at 2023-07-31T20:42:57+02:00 fixup! Bug 10760: Integrate TorButton to TorBrowser core The hashpassword parameter has been removed. - - - - - 13 changed files: - browser/components/torpreferences/content/connectionPane.js - browser/installer/package-manifest.in - browser/modules/TorStrings.jsm - toolkit/components/tor-launcher/TorLauncherUtil.sys.mjs - toolkit/components/tor-launcher/TorMonitorService.sys.mjs - toolkit/components/tor-launcher/TorParsers.sys.mjs - toolkit/components/tor-launcher/TorProcess.sys.mjs - toolkit/components/tor-launcher/TorProtocolService.sys.mjs - toolkit/torbutton/chrome/content/torbutton.js - toolkit/torbutton/components.conf - − toolkit/torbutton/modules/TorbuttonStartupObserver.jsm - toolkit/torbutton/moz.build - − toolkit/torbutton/torbutton.manifest Changes: ===================================== browser/components/torpreferences/content/connectionPane.js ===================================== @@ -14,8 +14,11 @@ const { setTimeout, clearTimeout } = ChromeUtils.import( const { TorSettings, TorSettingsTopics, TorSettingsData, TorBridgeSource } = ChromeUtils.import("resource:///modules/TorSettings.jsm"); -const { TorProtocolService } = ChromeUtils.import( - "resource://gre/modules/TorProtocolService.jsm" +const { TorParsers } = ChromeUtils.importESModule( + "resource://gre/modules/TorParsers.sys.mjs" +); +const { TorProtocolService } = ChromeUtils.importESModule( + "resource://gre/modules/TorProtocolService.sys.mjs" ); const { TorMonitorService, TorMonitorTopics } = ChromeUtils.import( "resource://gre/modules/TorMonitorService.jsm" @@ -495,7 +498,7 @@ const gConnectionPane = (function () { }); const idString = TorStrings.settings.bridgeId; const id = card.querySelector(selectors.bridges.cardId); - const details = parseBridgeLine(bridgeString); + const details = TorParsers.parseBridgeLine(bridgeString); if (details && details.id !== undefined) { card.setAttribute("data-bridge-id", details.id); } @@ -1111,23 +1114,3 @@ function makeBridgeId(bridgeString) { hash & 0x000000ff, ]; } - -function parseBridgeLine(line) { - const re = - /^\s*(\S+\s+)?([0-9a-fA-F\.\[\]\:]+:\d{1,5})(\s+[0-9a-fA-F]{40})?(\s+.+)?/; - const matches = line.match(re); - if (!matches) { - return null; - } - let bridge = { addr: matches[2] }; - if (matches[1] !== undefined) { - bridge.transport = matches[1].trim(); - } - if (matches[3] !== undefined) { - bridge.id = matches[3].trim().toUpperCase(); - } - if (matches[4] !== undefined) { - bridge.args = matches[4].trim(); - } - return bridge; -} ===================================== browser/installer/package-manifest.in ===================================== @@ -228,7 +228,6 @@ @RESPATH@/components/tor-launcher.manifest @RESPATH@/chrome/torbutton.manifest @RESPATH@/chrome/torbutton/* -@RESPATH@/components/torbutton.manifest @RESPATH@/chrome/toolkit@JAREXT@ @RESPATH@/chrome/toolkit.manifest #ifdef MOZ_GTK ===================================== browser/modules/TorStrings.jsm ===================================== @@ -11,9 +11,11 @@ const { Services } = ChromeUtils.import("resource://gre/modules/Services.jsm"); const { AppConstants } = ChromeUtils.import( "resource://gre/modules/AppConstants.jsm" ); -const { getLocale } = ChromeUtils.import( - "resource://torbutton/modules/utils.js" -); + +function getLocale() { + const locale = Services.locale.appLocaleAsBCP47; + return locale === "ja-JP-macos" ? "ja" : locale; +} /* Tor Property String Bundle ===================================== toolkit/components/tor-launcher/TorLauncherUtil.sys.mjs ===================================== @@ -5,6 +5,12 @@ * Tor Launcher Util JS Module *************************************************************************/ +const lazy = {}; + +ChromeUtils.defineESModuleGetters(lazy, { + FileUtils: "resource://gre/modules/FileUtils.sys.jsm", +}); + const kPropBundleURI = "chrome://torbutton/locale/torlauncher.properties"; const kPropNamePrefix = "torlauncher."; const kIPCDirPrefName = "extensions.torlauncher.tmp_ipc_dir"; @@ -209,14 +215,15 @@ class TorFile { // and return a file object. The control and SOCKS IPC objects will be // created by tor. normalize() { - if (!this.file.exists() && !this.isIPC) { + if (this.file.exists()) { + try { + this.file.normalize(); + } catch (e) { + console.warn("Normalization of the path failed", e); + } + } else if (!this.isIPC) { throw new Error(`${this.fileType} file not found: ${this.file.path}`); } - try { - this.file.normalize(); - } catch (e) { - console.warn("Normalization of the path failed", e); - } // Ensure that the IPC path length is short enough for use by the // operating system. If not, create and use a unique directory under @@ -452,6 +459,154 @@ export const TorLauncherUtil = Object.freeze({ return result ? result : ""; }, + /** + * Determine what kind of SOCKS port has been requested for this session or + * the browser has been configured for. + * On Windows (where Unix domain sockets are not supported), TCP is always + * used. + * + * The following environment variables are supported and take precedence over + * preferences: + * TOR_TRANSPROXY (do not use a proxy) + * TOR_SOCKS_IPC_PATH (file system path; ignored on Windows) + * TOR_SOCKS_HOST + * TOR_SOCKS_PORT + * + * The following preferences are consulted: + * network.proxy.socks + * network.proxy.socks_port + * extensions.torlauncher.socks_port_use_ipc (Boolean) + * extensions.torlauncher.socks_ipc_path (file system path) + * If extensions.torlauncher.socks_ipc_path is empty, a default path is used. + * + * When using TCP, if a value is not defined via an env variable it is + * taken from the corresponding browser preference if possible. The + * exceptions are: + * If network.proxy.socks contains a file: URL, a default value of + * "127.0.0.1" is used instead. + * If the network.proxy.socks_port value is not valid (outside the + * (0; 65535] range), a default value of 9150 is used instead. + * + * The SOCKS configuration will not influence the launch of a tor daemon and + * the configuration of the control port in any way. + * When a SOCKS configuration is required without TOR_SKIP_LAUNCH, the browser + * will try to configure the tor instance to use the required configuration. + * This also applies to TOR_TRANSPROXY (at least for now): tor will be + * launched with its defaults. + * + * TODO: add a preference to ignore the current configuration, and let tor + * listen on any free port. Then, the browser will prompt the daemon the port + * to use through the control port (even though this is quite dangerous at the + * moment, because with network disabled tor will disable also the SOCKS + * listeners, so it means that we will have to check it every time we change + * the network status). + */ + getPreferredSocksConfiguration() { + if (Services.env.exists("TOR_TRANSPROXY")) { + Services.prefs.setBoolPref("network.proxy.socks_remote_dns", false); + Services.prefs.setIntPref("network.proxy.type", 0); + Services.prefs.setIntPref("network.proxy.socks_port", 0); + Services.prefs.setCharPref("network.proxy.socks", ""); + return { transproxy: true }; + } + + let useIPC; + const socksPortInfo = { + transproxy: false, + }; + + if (!this.isWindows && Services.env.exists("TOR_SOCKS_IPC_PATH")) { + useIPC = true; + const ipcPath = Services.env.get("TOR_SOCKS_IPC_PATH"); + if (ipcPath) { + socksPortInfo.ipcFile = new lazy.FileUtils.File(ipcPath); + } + } else { + // Check for TCP host and port environment variables. + if (Services.env.exists("TOR_SOCKS_HOST")) { + socksPortInfo.host = Services.env.get("TOR_SOCKS_HOST"); + useIPC = false; + } + if (Services.env.exists("TOR_SOCKS_PORT")) { + const port = parseInt(Services.env.get("TOR_SOCKS_PORT"), 10); + if (Number.isInteger(port) && port > 0 && port <= 65535) { + socksPortInfo.port = port; + useIPC = false; + } + } + } + + if (useIPC === undefined) { + socksPortInfo.useIPC = + !this.isWindows && + Services.prefs.getBoolPref( + "extensions.torlauncher.socks_port_use_ipc", + false + ); + } + + // Fill in missing SOCKS info from prefs. + if (socksPortInfo.useIPC) { + if (!socksPortInfo.ipcFile) { + socksPortInfo.ipcFile = TorLauncherUtil.getTorFile("socks_ipc", false); + } + } else { + if (!socksPortInfo.host) { + let socksAddr = Services.prefs.getCharPref( + "network.proxy.socks", + "127.0.0.1" + ); + let socksAddrHasHost = socksAddr && !socksAddr.startsWith("file:"); + socksPortInfo.host = socksAddrHasHost ? socksAddr : "127.0.0.1"; + } + + if (!socksPortInfo.port) { + let socksPort = Services.prefs.getIntPref( + "network.proxy.socks_port", + 0 + ); + // This pref is set as 0 by default in Firefox, use 9150 if we get 0. + socksPortInfo.port = + socksPort > 0 && socksPort <= 65535 ? socksPort : 9150; + } + } + + return socksPortInfo; + }, + + setProxyConfiguration(socksPortInfo) { + if (socksPortInfo.transproxy) { + return; + } + + if (socksPortInfo.useIPC) { + const fph = Services.io + .getProtocolHandler("file") + .QueryInterface(Ci.nsIFileProtocolHandler); + const fileURI = fph.newFileURI(socksPortInfo.ipcFile); + Services.prefs.setCharPref("network.proxy.socks", fileURI.spec); + Services.prefs.setIntPref("network.proxy.socks_port", 0); + } else { + if (socksPortInfo.host) { + Services.prefs.setCharPref("network.proxy.socks", socksPortInfo.host); + } + if (socksPortInfo.port) { + Services.prefs.setIntPref( + "network.proxy.socks_port", + socksPortInfo.port + ); + } + } + + if (socksPortInfo.ipcFile || socksPortInfo.host || socksPortInfo.port) { + Services.prefs.setBoolPref("network.proxy.socks_remote_dns", true); + Services.prefs.setIntPref("network.proxy.type", 1); + } + + // Force prefs to be synced to disk + Services.prefs.savePrefFile(null); + }, + get shouldStartAndOwnTor() { const kPrefStartTor = "extensions.torlauncher.start_tor"; try { ===================================== toolkit/components/tor-launcher/TorMonitorService.sys.mjs ===================================== @@ -13,6 +13,10 @@ import { TorLauncherUtil } from "resource://gre/modules/TorLauncherUtil.sys.mjs" const lazy = {}; +ChromeUtils.defineESModuleGetters(lazy, { + TorProtocolService: "resource://gre/modules/TorProtocolService.sys.mjs", +}); + ChromeUtils.defineModuleGetter( lazy, "controller", @@ -233,7 +237,10 @@ export const TorMonitorService = { // TorProcess should be instanced once, then always reused and restarted // only through the prompt it exposes when the controlled process dies. if (!this._torProcess) { - this._torProcess = new TorProcess(); + this._torProcess = new TorProcess( + lazy.TorProtocolService.torControlPortInfo, + lazy.TorProtocolService.torSOCKSPortInfo + ); this._torProcess.onExit = () => { this._shutDownEventMonitor(); Services.obs.notifyObservers(null, TorTopics.ProcessExited); @@ -254,6 +261,7 @@ export const TorMonitorService = { await this._torProcess.start(); if (this._torProcess.isRunning) { logger.info("tor started"); + this._torProcessStartTime = Date.now(); } } catch (e) { // TorProcess already logs the error. ===================================== toolkit/components/tor-launcher/TorParsers.sys.mjs ===================================== @@ -267,4 +267,18 @@ export const TorParsers = Object.freeze({ rv += aStr.substring(lastAdded, aStr.length - 1); return rv; }, + + parseBridgeLine(line) { + const re = + /\s*(?:(?<transport>\S+)\s+)?(?<addr>[0-9a-fA-F\.\[\]\:]+:\d{1,5})(?:\s+(?<id>[0-9a-fA-F]{40}))?(?:\s+(?<args>.+))?/; + const match = re.exec(line); + if (!match) { + throw new Error("Invalid bridge line."); + } + const bridge = match.groups; + if (!bridge.transport) { + bridge.transport = "vanilla"; + } + return bridge; + }, }); ===================================== toolkit/components/tor-launcher/TorProcess.sys.mjs ===================================== @@ -1,21 +1,17 @@ +/* This Source Code Form is subject to the terms of the Mozilla Public + * License, v. 2.0. If a copy of the MPL was not distributed with this + * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ + import { setTimeout } from "resource://gre/modules/Timer.sys.mjs"; import { ConsoleAPI } from "resource://gre/modules/Console.sys.mjs"; import { Subprocess } from "resource://gre/modules/Subprocess.sys.mjs"; const lazy = {}; -ChromeUtils.defineModuleGetter( - lazy, - "TorProtocolService", - "resource://gre/modules/TorProtocolService.jsm" -); -const { TorLauncherUtil } = ChromeUtils.import( - "resource://gre/modules/TorLauncherUtil.jsm" -); - -const { TorParsers } = ChromeUtils.import( - "resource://gre/modules/TorParsers.jsm" -); +ChromeUtils.defineESModuleGetters(lazy, { + TorLauncherUtil: "resource://gre/modules/TorLauncherUtil.sys.mjs", + TorParsers: "resource://gre/modules/TorParsers.sys.mjs", +}); const TorProcessStatus = Object.freeze({ Unknown: 0, @@ -30,53 +26,86 @@ const logger = new ConsoleAPI({ }); export class TorProcess { - _exeFile = null; - _dataDir = null; - _args = []; - _subprocess = null; - _status = TorProcessStatus.Unknown; - _torProcessStartTime = null; // JS Date.now() - _didConnectToTorControlPort = false; // Have we ever made a connection? + #controlSettings; + #socksSettings; + #exeFile = null; + #dataDir = null; + #args = []; + #subprocess = null; + #status = TorProcessStatus.Unknown; + // Have we ever made a connection on the control port? + #didConnectToTorControlPort = false; + + onExit = exitCode => {}; + onRestart = () => {}; + + constructor(controlSettings, socksSettings) { + if ( + controlSettings && + !controlSettings.password && + !controlSettings.cookieFilePath + ) { + throw new Error("Unauthenticated control port is not supported"); + } - onExit = null; - onRestart = null; + const checkPort = port => + port === undefined || + (Number.isInteger(controlSettings.port) && + controlSettings.port > 0 && + controlSettings.port < 65535); + if (!checkPort(controlSettings?.port)) { + throw new Error("Invalid control port"); + } + if (!checkPort(socksSettings.port)) { + throw new Error("Invalid port specified for the SOCKS port"); + } + + this.#controlSettings = { ...controlSettings }; + const ipcFileToString = file => + "unix:" + lazy.TorParsers.escapeString(file.path); + if (controlSettings.ipcFile) { + this.#controlSettings.ipcFile = ipcFileToString(controlSettings.ipcFile); + } + this.#socksSettings = { ...socksSettings }; + if (socksSettings.ipcFile) { + this.#socksSettings.ipcFile = ipcFileToString(socksSettings.ipcFile); + } + } get status() { - return this._status; + return this.#status; } get isRunning() { return ( - this._status === TorProcessStatus.Starting || - this._status === TorProcessStatus.Running + this.#status === TorProcessStatus.Starting || + this.#status === TorProcessStatus.Running ); } async start() { - if (this._subprocess) { + if (this.#subprocess) { return; } - this._status = TorProcessStatus.Unknown; + this.#status = TorProcessStatus.Unknown; try { - this._makeArgs(); - this._addControlPortArg(); - this._addSocksPortArg(); + this.#makeArgs(); + this.#addControlPortArgs(); + this.#addSocksPortArg(); const pid = Services.appinfo.processID; if (pid !== 0) { - this._args.push("__OwningControllerProcess"); - this._args.push("" + pid); + this.#args.push("__OwningControllerProcess", pid.toString()); } - if (TorLauncherUtil.shouldShowNetworkSettings) { - this._args.push("DisableNetwork"); - this._args.push("1"); + if (lazy.TorLauncherUtil.shouldShowNetworkSettings) { + this.#args.push("DisableNetwork", "1"); } - this._status = TorProcessStatus.Starting; - this._didConnectToTorControlPort = false; + this.#status = TorProcessStatus.Starting; + this.#didConnectToTorControlPort = false; // useful for simulating slow tor daemon launch const kPrefTorDaemonLaunchDelay = "extensions.torlauncher.launch_delay"; @@ -88,29 +117,31 @@ export class TorProcess { await new Promise(resolve => setTimeout(() => resolve(), launchDelay)); } - logger.debug(`Starting ${this._exeFile.path}`, this._args); + logger.debug(`Starting ${this.#exeFile.path}`, this.#args); const options = { - command: this._exeFile.path, - arguments: this._args, + command: this.#exeFile.path, + arguments: this.#args, stderr: "stdout", - workdir: TorLauncherUtil.getTorFile("pt-startup-dir", false).path, + workdir: lazy.TorLauncherUtil.getTorFile("pt-startup-dir", false).path, }; - this._subprocess = await Subprocess.call(options); - this._dumpStdout(); - this._watchProcess(); - this._status = TorProcessStatus.Running; - this._torProcessStartTime = Date.now(); + this.#subprocess = await Subprocess.call(options); + this.#status = TorProcessStatus.Running; } catch (e) { - this._status = TorProcessStatus.Exited; - this._subprocess = null; + this.#status = TorProcessStatus.Exited; + this.#subprocess = null; logger.error("startTor error:", e); throw e; } + + // Do not await the following functions, as they will return only when the + // process exits. + this.#dumpStdout(); + this.#watchProcess(); } // Forget about a process. // - // Instead of killing the tor process, we rely on the TAKEOWNERSHIP feature + // Instead of killing the tor process, we rely on the TAKEOWNERSHIP feature // to shut down tor when we close the control port connection. // // Previously, we sent a SIGNAL HALT command to the tor control port, @@ -123,36 +154,38 @@ export class TorProcess { // Still, before closing the owning connection, this class should forget about // the process, so that future notifications will be ignored. forget() { - this._subprocess = null; - this._status = TorProcessStatus.Exited; + this.#subprocess = null; + this.#status = TorProcessStatus.Exited; } // The owner of the process can use this function to tell us that they // successfully connected to the control port. This information will be used // only to decide which text to show in the confirmation dialog if tor exits. connectionWorked() { - this._didConnectToTorControlPort = true; + this.#didConnectToTorControlPort = true; } - async _dumpStdout() { + async #dumpStdout() { let string; while ( - this._subprocess && - (string = await this._subprocess.stdout.readString()) + this.#subprocess && + (string = await this.#subprocess.stdout.readString()) ) { dump(string); } } - async _watchProcess() { - const watched = this._subprocess; + async #watchProcess() { + const watched = this.#subprocess; if (!watched) { return; } + let processExitCode; try { const { exitCode } = await watched.wait(); + processExitCode = exitCode; - if (watched !== this._subprocess) { + if (watched !== this.#subprocess) { logger.debug(`A Tor process exited with code ${exitCode}.`); } else if (exitCode) { logger.warn(`The watched Tor process exited with code ${exitCode}.`); @@ -163,30 +196,31 @@ export class TorProcess { logger.error("Failed to watch the tor process", e); } - if (watched === this._subprocess) { - this._processExitedUnexpectedly(); + if (watched === this.#subprocess) { + this.#processExitedUnexpectedly(processExitCode); } } - _processExitedUnexpectedly() { - this._subprocess = null; - this._status = TorProcessStatus.Exited; + #processExitedUnexpectedly(exitCode) { + this.#subprocess = null; + this.#status = TorProcessStatus.Exited; // TODO: Move this logic somewhere else? let s; - if (!this._didConnectToTorControlPort) { + if (!this.#didConnectToTorControlPort) { // tor might be misconfigured, becauser we could never connect to it const key = "tor_exited_during_startup"; - s = TorLauncherUtil.getLocalizedString(key); + s = lazy.TorLauncherUtil.getLocalizedString(key); } else { // tor exited suddenly, so configuration should be okay s = - TorLauncherUtil.getLocalizedString("tor_exited") + + lazy.TorLauncherUtil.getLocalizedString("tor_exited") + "\n\n" + - TorLauncherUtil.getLocalizedString("tor_exited2"); + lazy.TorLauncherUtil.getLocalizedString("tor_exited2"); } logger.info(s); - const defaultBtnLabel = TorLauncherUtil.getLocalizedString("restart_tor"); + const defaultBtnLabel = + lazy.TorLauncherUtil.getLocalizedString("restart_tor"); let cancelBtnLabel = "OK"; try { const kSysBundleURI = "chrome://global/locale/commonDialogs.properties"; @@ -196,51 +230,43 @@ export class TorProcess { logger.warn("Could not localize the cancel button", e); } - const restart = TorLauncherUtil.showConfirm( + const restart = lazy.TorLauncherUtil.showConfirm( null, s, defaultBtnLabel, cancelBtnLabel ); if (restart) { - this.start().then(() => { - if (this.onRestart) { - this.onRestart(); - } - }); - } else if (this.onExit) { - this.onExit(); + this.start().then(this.onRestart); + } else { + this.onExit(exitCode); } } - _makeArgs() { - // Ideally, we would cd to the Firefox application directory before - // starting tor (but we don't know how to do that). Instead, we - // rely on the TBB launcher to start Firefox from the right place. - + #makeArgs() { + this.#exeFile = lazy.TorLauncherUtil.getTorFile("tor", false); + const torrcFile = lazy.TorLauncherUtil.getTorFile("torrc", true); // Get the Tor data directory first so it is created before we try to // construct paths to files that will be inside it. - this._exeFile = TorLauncherUtil.getTorFile("tor", false); - const torrcFile = TorLauncherUtil.getTorFile("torrc", true); - this._dataDir = TorLauncherUtil.getTorFile("tordatadir", true); - const onionAuthDir = TorLauncherUtil.getTorFile("toronionauthdir", true); - const hashedPassword = lazy.TorProtocolService.torGetPassword(true); + this.#dataDir = lazy.TorLauncherUtil.getTorFile("tordatadir", true); + const onionAuthDir = lazy.TorLauncherUtil.getTorFile( + "toronionauthdir", + true + ); let detailsKey; - if (!this._exeFile) { + if (!this.#exeFile) { detailsKey = "tor_missing"; } else if (!torrcFile) { detailsKey = "torrc_missing"; - } else if (!this._dataDir) { + } else if (!this.#dataDir) { detailsKey = "datadir_missing"; } else if (!onionAuthDir) { detailsKey = "onionauthdir_missing"; - } else if (!hashedPassword) { - detailsKey = "password_hash_missing"; } if (detailsKey) { - const details = TorLauncherUtil.getLocalizedString(detailsKey); + const details = lazy.TorLauncherUtil.getLocalizedString(detailsKey); const key = "unable_to_start_tor"; - const err = TorLauncherUtil.getFormattedLocalizedString( + const err = lazy.TorLauncherUtil.getFormattedLocalizedString( key, [details], 1 @@ -248,7 +274,7 @@ export class TorProcess { throw new Error(err); } - const torrcDefaultsFile = TorLauncherUtil.getTorFile( + const torrcDefaultsFile = lazy.TorLauncherUtil.getTorFile( "torrc-defaults", false ); @@ -258,77 +284,131 @@ export class TorProcess { const geoip6File = torrcDefaultsFile.clone(); geoip6File.leafName = "geoip6"; - this._args = []; + this.#args = []; if (torrcDefaultsFile) { - this._args.push("--defaults-torrc"); - this._args.push(torrcDefaultsFile.path); + this.#args.push("--defaults-torrc", torrcDefaultsFile.path); } - this._args.push("-f"); - this._args.push(torrcFile.path); - this._args.push("DataDirectory"); - this._args.push(this._dataDir.path); - this._args.push("ClientOnionAuthDir"); - this._args.push(onionAuthDir.path); - this._args.push("GeoIPFile"); - this._args.push(geoipFile.path); - this._args.push("GeoIPv6File"); - this._args.push(geoip6File.path); - this._args.push("HashedControlPassword"); - this._args.push(hashedPassword); + this.#args.push("-f", torrcFile.path); + this.#args.push("DataDirectory", this.#dataDir.path); + this.#args.push("ClientOnionAuthDir", onionAuthDir.path); + this.#args.push("GeoIPFile", geoipFile.path); + this.#args.push("GeoIPv6File", geoip6File.path); } - _addControlPortArg() { - // Include a ControlPort argument to support switching between - // a TCP port and an IPC port (e.g., a Unix domain socket). We - // include a "+__" prefix so that (1) this control port is added - // to any control ports that the user has defined in their torrc - // file and (2) it is never written to torrc. + /** + * Add all the arguments related to the control port. + * We use the + prefix so that the the port is added to any other port already + * defined in the torrc, and the __ prefix so that it is never written to + * torrc. + */ + #addControlPortArgs() { + if (!this.#controlSettings) { + return; + } + let controlPortArg; - const controlIPCFile = lazy.TorProtocolService.torGetControlIPCFile(); - const controlPort = lazy.TorProtocolService.torGetControlPort(); - if (controlIPCFile) { - controlPortArg = this._ipcPortArg(controlIPCFile); - } else if (controlPort) { - controlPortArg = "" + controlPort; + if (this.#controlSettings.ipcFile) { + controlPortArg = this.#controlSettings.ipcFile; + } else if (this.#controlSettings.port) { + controlPortArg = this.#controlSettings.host + ? `${this.#controlSettings.host}:${this.#controlSettings.port}` + : this.#controlSettings.port.toString(); } if (controlPortArg) { - this._args.push("+__ControlPort"); - this._args.push(controlPortArg); + this.#args.push("+__ControlPort", controlPortArg); + } + + if (this.#controlSettings.password) { + this.#args.push( + "HashedControlPassword", + this.#hashPassword(this.#controlSettings.password) + ); + } + if (this.#controlSettings.cookieFilePath) { + this.#args.push("CookieAuthentication", "1"); + this.#args.push("CookieAuthFile", this.#controlSettings.cookieFilePath); } } - _addSocksPortArg() { - // Include a SocksPort argument to support switching between - // a TCP port and an IPC port (e.g., a Unix domain socket). We - // include a "+__" prefix so that (1) this SOCKS port is added - // to any SOCKS ports that the user has defined in their torrc - // file and (2) it is never written to torrc. - const socksPortInfo = lazy.TorProtocolService.torGetSOCKSPortInfo(); - if (socksPortInfo) { - let socksPortArg; - if (socksPortInfo.ipcFile) { - socksPortArg = this._ipcPortArg(socksPortInfo.ipcFile); - } else if (socksPortInfo.host && socksPortInfo.port != 0) { - socksPortArg = socksPortInfo.host + ":" + socksPortInfo.port; - } - if (socksPortArg) { - let socksPortFlags = Services.prefs.getCharPref( - "extensions.torlauncher.socks_port_flags", - "IPv6Traffic PreferIPv6 KeepAliveIsolateSOCKSAuth" - ); - if (socksPortFlags) { - socksPortArg += " " + socksPortFlags; - } - this._args.push("+__SocksPort"); - this._args.push(socksPortArg); + /** + * Add the argument related to the control port. + * We use the + prefix so that the the port is added to any other port already + * defined in the torrc, and the __ prefix so that it is never written to + * torrc. + */ + #addSocksPortArg() { + let socksPortArg; + if (this.#socksSettings.ipcFile) { + socksPortArg = this.#socksSettings.ipcFile; + } else if (this.#socksSettings.port != 0) { + socksPortArg = this.#socksSettings.host + ? `${this.#socksSettings.host}:${this.#socksSettings.port}` + : this.#socksSettings.port.toString(); + } + if (socksPortArg) { + const socksPortFlags = Services.prefs.getCharPref( + "extensions.torlauncher.socks_port_flags", + "IPv6Traffic PreferIPv6 KeepAliveIsolateSOCKSAuth" + ); + if (socksPortFlags) { + socksPortArg += " " + socksPortFlags; } + this.#args.push("+__SocksPort", socksPortArg); + } + } + + // Based on Vidalia's TorSettings::hashPassword(). + #hashPassword(aHexPassword) { + if (!aHexPassword) { + return null; } + + // Generate a random, 8 byte salt value. + const salt = Array.from(crypto.getRandomValues(new Uint8Array(8))); + + // Convert hex-encoded password to an array of bytes. + const password = []; + for (let i = 0; i < aHexPassword.length; i += 2) { + password.push(parseInt(aHexPassword.substring(i, i + 2), 16)); + } + + // Run through the S2K algorithm and convert to a string. + const toHex = v => v.toString(16).padStart(2, "0"); + const arrayToHex = aArray => aArray.map(toHex).join(""); + const kCodedCount = 96; + const hashVal = this.#cryptoSecretToKey(password, salt, kCodedCount); + return "16:" + arrayToHex(salt) + toHex(kCodedCount) + arrayToHex(hashVal); } - // Return a ControlPort or SocksPort argument for aIPCFile (an nsIFile). - // The result is unix:/path or unix:"/path with spaces" with appropriate - // C-style escaping within the path portion. - _ipcPortArg(aIPCFile) { - return "unix:" + TorParsers.escapeString(aIPCFile.path); + // #cryptoSecretToKey() is similar to Vidalia's crypto_secret_to_key(). + // It generates and returns a hash of aPassword by following the iterated + // and salted S2K algorithm (see RFC 2440 section 3.6.1.3). + // See also https://gitlab.torproject.org/tpo/core/torspec/-/blob/main/control-spec.txt…. + // Returns an array of bytes. + #cryptoSecretToKey(aPassword, aSalt, aCodedCount) { + const inputArray = aSalt.concat(aPassword); + + // Subtle crypto only has the final digest, and does not allow incremental + // updates. + const hasher = Cc["@mozilla.org/security/hash;1"].createInstance( + Ci.nsICryptoHash + ); + hasher.init(hasher.SHA1); + const kEXPBIAS = 6; + let count = (16 + (aCodedCount & 15)) << ((aCodedCount >> 4) + kEXPBIAS); + while (count > 0) { + if (count > inputArray.length) { + hasher.update(inputArray, inputArray.length); + count -= inputArray.length; + } else { + const finalArray = inputArray.slice(0, count); + hasher.update(finalArray, finalArray.length); + count = 0; + } + } + return hasher + .finish(false) + .split("") + .map(b => b.charCodeAt(0)); } } ===================================== toolkit/components/tor-launcher/TorProtocolService.sys.mjs ===================================== @@ -289,9 +289,8 @@ export const TorProtocolService = { // are also used in torbutton. // Returns Tor password string or null if an error occurs. - torGetPassword(aPleaseHash) { - const pw = this._controlPassword; - return aPleaseHash ? this._hashPassword(pw) : pw; + torGetPassword() { + return this._controlPassword; }, torGetControlIPCFile() { @@ -306,6 +305,24 @@ export const TorProtocolService = { return this._SOCKSPortInfo; }, + get torControlPortInfo() { + const info = { + password: this._controlPassword, + }; + if (this._controlIPCFile) { + info.ipcFile = this._controlIPCFile?.clone(); + } + if (this._controlPort) { + info.host = this._controlHost; + info.port = this._controlPort; + } + return info; + }, + + get torSOCKSPortInfo() { + return this._SOCKSPortInfo; + }, + // Public, but called only internally // Executes a command on the control port. @@ -469,115 +486,8 @@ export const TorProtocolService = { this._controlPassword = this._generateRandomPassword(); } - // Determine what kind of SOCKS port Tor and the browser will use. - // On Windows (where Unix domain sockets are not supported), TCP is - // always used. - // - // The following environment variables are supported and take - // precedence over preferences: - // TOR_SOCKS_IPC_PATH (file system path; ignored on Windows) - // TOR_SOCKS_HOST - // TOR_SOCKS_PORT - // - // The following preferences are consulted: - // network.proxy.socks - // network.proxy.socks_port - // extensions.torlauncher.socks_port_use_ipc (Boolean) - // extensions.torlauncher.socks_ipc_path (file system path) - // If extensions.torlauncher.socks_ipc_path is empty, a default - // path is used (<tor-data-directory>/socks.socket). - // - // When using TCP, if a value is not defined via an env variable it is - // taken from the corresponding browser preference if possible. The - // exceptions are: - // If network.proxy.socks contains a file: URL, a default value of - // "127.0.0.1" is used instead. - // If the network.proxy.socks_port value is 0, a default value of - // 9150 is used instead. - // - // Supported scenarios: - // 1. By default, an IPC object at a default path is used. - // 2. If extensions.torlauncher.socks_port_use_ipc is set to false, - // a TCP socket at 127.0.0.1:9150 is used, unless different values - // are set in network.proxy.socks and network.proxy.socks_port. - // 3. If the TOR_SOCKS_IPC_PATH env var is set, an IPC object at that - // path is used (e.g., a Unix domain socket). - // 4. If the TOR_SOCKS_HOST and/or TOR_SOCKS_PORT env vars are set, TCP - // is used. Values not set via env vars will be taken from the - // network.proxy.socks and network.proxy.socks_port prefs as described - // above. - // 5. If extensions.torlauncher.socks_port_use_ipc is true and - // extensions.torlauncher.socks_ipc_path is set, an IPC object at - // the specified path is used. - // 6. Tor Launcher is disabled. Torbutton will respect the env vars if - // present; if not, the values in network.proxy.socks and - // network.proxy.socks_port are used without modification. - - let useIPC; - this._SOCKSPortInfo = { ipcFile: undefined, host: undefined, port: 0 }; - if (!isWindows && Services.env.exists("TOR_SOCKS_IPC_PATH")) { - let ipcPath = Services.env.get("TOR_SOCKS_IPC_PATH"); - this._SOCKSPortInfo.ipcFile = new lazy.FileUtils.File(ipcPath); - useIPC = true; - } else { - // Check for TCP host and port environment variables. - if (Services.env.exists("TOR_SOCKS_HOST")) { - this._SOCKSPortInfo.host = Services.env.get("TOR_SOCKS_HOST"); - useIPC = false; - } - if (Services.env.exists("TOR_SOCKS_PORT")) { - this._SOCKSPortInfo.port = parseInt( - Services.env.get("TOR_SOCKS_PORT"), - 10 - ); - useIPC = false; - } - } - - if (useIPC === undefined) { - useIPC = - !isWindows && - Services.prefs.getBoolPref( - "extensions.torlauncher.socks_port_use_ipc", - false - ); - } - - // Fill in missing SOCKS info from prefs. - if (useIPC) { - if (!this._SOCKSPortInfo.ipcFile) { - this._SOCKSPortInfo.ipcFile = TorLauncherUtil.getTorFile( - "socks_ipc", - false - ); - } - } else { - if (!this._SOCKSPortInfo.host) { - let socksAddr = Services.prefs.getCharPref( - "network.proxy.socks", - "127.0.0.1" - ); - let socksAddrHasHost = socksAddr && !socksAddr.startsWith("file:"); - this._SOCKSPortInfo.host = socksAddrHasHost ? socksAddr : "127.0.0.1"; - } - - if (!this._SOCKSPortInfo.port) { - let socksPort = Services.prefs.getIntPref( - "network.proxy.socks_port", - 0 - ); - // This pref is set as 0 by default in Firefox, use 9150 if we get 0. - this._SOCKSPortInfo.port = socksPort != 0 ? socksPort : 9150; - } - } - - logger.info("SOCKS port type: " + (useIPC ? "IPC" : "TCP")); - if (useIPC) { - logger.info(`ipcFile: ${this._SOCKSPortInfo.ipcFile.path}`); - } else { - logger.info(`SOCKS host: ${this._SOCKSPortInfo.host}`); - logger.info(`SOCKS port: ${this._SOCKSPortInfo.port}`); - } + this._SOCKSPortInfo = TorLauncherUtil.getPreferredSocksConfiguration(); + TorLauncherUtil.setProxyConfiguration(this._SOCKSPortInfo); // Set the global control port info parameters. // These values may be overwritten by torbutton when it initializes, but @@ -781,38 +691,6 @@ export const TorProtocolService = { return pwd; }, - // Based on Vidalia's TorSettings::hashPassword(). - _hashPassword(aHexPassword) { - if (!aHexPassword) { - return null; - } - - // Generate a random, 8 byte salt value. - const salt = Array.from(crypto.getRandomValues(new Uint8Array(8))); - - // Convert hex-encoded password to an array of bytes. - const password = []; - for (let i = 0; i < aHexPassword.length; i += 2) { - password.push(parseInt(aHexPassword.substring(i, i + 2), 16)); - } - - // Run through the S2K algorithm and convert to a string. - const kCodedCount = 96; - const hashVal = this._cryptoSecretToKey(password, salt, kCodedCount); - if (!hashVal) { - logger.error("_cryptoSecretToKey() failed"); - return null; - } - - const arrayToHex = aArray => - aArray.map(item => this._toHex(item, 2)).join(""); - let rv = "16:"; - rv += arrayToHex(salt); - rv += this._toHex(kCodedCount, 2); - rv += arrayToHex(hashVal); - return rv; - }, - // Returns -1 upon failure. _cryptoRandInt(aMax) { // Based on tor's crypto_rand_int(). @@ -831,43 +709,6 @@ export const TorProtocolService = { return val % aMax; }, - // _cryptoSecretToKey() is similar to Vidalia's crypto_secret_to_key(). - // It generates and returns a hash of aPassword by following the iterated - // and salted S2K algorithm (see RFC 2440 section 3.6.1.3). - // Returns an array of bytes. - _cryptoSecretToKey(aPassword, aSalt, aCodedCount) { - if (!aPassword || !aSalt) { - return null; - } - - const inputArray = aSalt.concat(aPassword); - - // Subtle crypto only has the final digest, and does not allow incremental - // updates. Also, it is async, so we should hash and keep the hash in a - // variable if we wanted to switch to getters. - // So, keeping this implementation should be okay for now. - const hasher = Cc["@mozilla.org/security/hash;1"].createInstance( - Ci.nsICryptoHash - ); - hasher.init(hasher.SHA1); - const kEXPBIAS = 6; - let count = (16 + (aCodedCount & 15)) << ((aCodedCount >> 4) + kEXPBIAS); - while (count > 0) { - if (count > inputArray.length) { - hasher.update(inputArray, inputArray.length); - count -= inputArray.length; - } else { - const finalArray = inputArray.slice(0, count); - hasher.update(finalArray, finalArray.length); - count = 0; - } - } - return hasher - .finish(false) - .split("") - .map(b => b.charCodeAt(0)); - }, - _toHex(aValue, aMinLen) { return aValue.toString(16).padStart(aMinLen, "0"); }, ===================================== toolkit/torbutton/chrome/content/torbutton.js ===================================== @@ -60,7 +60,7 @@ var torbutton_init; } else { try { // Try to get password from Tor Launcher. - m_tb_control_pass = TorProtocolService.torGetPassword(false); + m_tb_control_pass = TorProtocolService.torGetPassword(); } catch (e) {} } ===================================== toolkit/torbutton/components.conf ===================================== @@ -1,12 +1,4 @@ Classes = [ - { - "cid": "{06322def-6fde-4c06-aef6-47ae8e799629}", - "contract_ids": [ - "@torproject.org/startup-observer;1" - ], - "jsm": "resource://torbutton/modules/TorbuttonStartupObserver.jsm", - "constructor": "StartupObserver", - }, { "cid": "{f36d72c9-9718-4134-b550-e109638331d7}", "contract_ids": [ ===================================== toolkit/torbutton/modules/TorbuttonStartupObserver.jsm deleted ===================================== @@ -1,138 +0,0 @@ -// Bug 1506 P1-3: This code is mostly hackish remnants of session store -// support. There are a couple of observer events that *might* be worth -// listening to. Search for 1506 in the code. - -/************************************************************************* - * Startup observer (JavaScript XPCOM component) - * - * Cases tested (each during Tor and Non-Tor, FF4 and FF3.6) - * 1. Crash - * 2. Upgrade - * 3. Fresh install - * - *************************************************************************/ - -var EXPORTED_SYMBOLS = ["StartupObserver"]; - -const { Services } = ChromeUtils.import("resource://gre/modules/Services.jsm"); -const { XPCOMUtils } = ChromeUtils.import( - "resource://gre/modules/XPCOMUtils.jsm" -); - -const { TorProtocolService } = ChromeUtils.import( - "resource://gre/modules/TorProtocolService.jsm" -); - -const lazy = {}; - -XPCOMUtils.defineLazyModuleGetters(lazy, { - FileUtils: "resource://gre/modules/FileUtils.jsm", -}); - -function StartupObserver() { - this.logger = Cc["@torproject.org/torbutton-logger;1"].getService( - Ci.nsISupports - ).wrappedJSObject; - this._prefs = Services.prefs; - this.logger.log(3, "Startup Observer created"); - - try { - // XXX: We're in a race with HTTPS-Everywhere to update our proxy settings - // before the initial SSL-Observatory test... If we lose the race, Firefox - // caches the old proxy settings for check.tp.o somehwere, and it never loads :( - this.setProxySettings(); - } catch (e) { - this.logger.log( - 4, - "Early proxy change failed. Will try again at profile load. Error: " + e - ); - } -} - -StartupObserver.prototype = { - // Bug 6803: We need to get the env vars early due to - // some weird proxy caching code that showed up in FF15. - // Otherwise, homepage domain loads fail forever. - setProxySettings() { - // Bug 1506: Still want to get these env vars - if (Services.env.exists("TOR_TRANSPROXY")) { - this.logger.log(3, "Resetting Tor settings to transproxy"); - this._prefs.setBoolPref("network.proxy.socks_remote_dns", false); - this._prefs.setIntPref("network.proxy.type", 0); - this._prefs.setIntPref("network.proxy.socks_port", 0); - this._prefs.setCharPref("network.proxy.socks", ""); - } else { - // Try to retrieve SOCKS proxy settings from Tor Launcher. - let socksPortInfo; - try { - socksPortInfo = TorProtocolService.torGetSOCKSPortInfo(); - } catch (e) { - this.logger.log(3, "tor launcher failed " + e); - } - - // If Tor Launcher is not available, check environment variables. - if (!socksPortInfo) { - socksPortInfo = { ipcFile: undefined, host: undefined, port: 0 }; - - let isWindows = Services.appinfo.OS === "WINNT"; - if (!isWindows && Services.env.exists("TOR_SOCKS_IPC_PATH")) { - socksPortInfo.ipcFile = new lazy.FileUtils.File( - Services.env.get("TOR_SOCKS_IPC_PATH") - ); - } else { - if (Services.env.exists("TOR_SOCKS_HOST")) { - socksPortInfo.host = Services.env.get("TOR_SOCKS_HOST"); - } - if (Services.env.exists("TOR_SOCKS_PORT")) { - socksPortInfo.port = parseInt(Services.env.get("TOR_SOCKS_PORT")); - } - } - } - - // Adjust network.proxy prefs. - if (socksPortInfo.ipcFile) { - let fph = Services.io - .getProtocolHandler("file") - .QueryInterface(Ci.nsIFileProtocolHandler); - let fileURI = fph.newFileURI(socksPortInfo.ipcFile); - this.logger.log(3, "Reset socks to " + fileURI.spec); - this._prefs.setCharPref("network.proxy.socks", fileURI.spec); - this._prefs.setIntPref("network.proxy.socks_port", 0); - } else { - if (socksPortInfo.host) { - this._prefs.setCharPref("network.proxy.socks", socksPortInfo.host); - this.logger.log(3, "Reset socks host to " + socksPortInfo.host); - } - if (socksPortInfo.port) { - this._prefs.setIntPref( - "network.proxy.socks_port", - socksPortInfo.port - ); - this.logger.log(3, "Reset socks port to " + socksPortInfo.port); - } - } - - if (socksPortInfo.ipcFile || socksPortInfo.host || socksPortInfo.port) { - this._prefs.setBoolPref("network.proxy.socks_remote_dns", true); - this._prefs.setIntPref("network.proxy.type", 1); - } - } - - // Force prefs to be synced to disk - Services.prefs.savePrefFile(null); - - this.logger.log(3, "Synced network settings to environment."); - }, - - observe(subject, topic, data) { - if (topic == "profile-after-change") { - this.setProxySettings(); - } - - // In all cases, force prefs to be synced to disk - Services.prefs.savePrefFile(null); - }, - - // Hack to get us registered early to observe recovery - _xpcom_categories: [{ category: "profile-after-change" }], -}; ===================================== toolkit/torbutton/moz.build ===================================== @@ -8,7 +8,3 @@ JAR_MANIFESTS += ['jar.mn'] XPCOM_MANIFESTS += [ "components.conf", ] - -EXTRA_COMPONENTS += [ - "torbutton.manifest", -] ===================================== toolkit/torbutton/torbutton.manifest deleted ===================================== @@ -1 +0,0 @@ -category profile-after-change StartupObserver @torproject.org/startup-observer;1 View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/f0493f… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/f0493f… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][base-browser-115.0.2esr-13.0-1] Deleted 2 commits: Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs
by richard (＠richard) 31 Jul '23

31 Jul '23

richard pushed to branch base-browser-115.0.2esr-13.0-1 at The Tor Project / Applications / Tor Browser WARNING: The push did not contain any new commits, but force pushed to delete the commits and changes below. Deleted commits: 60fefea4 by Edgar Chen at 2023-07-31T17:56:34+00:00 Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs Fullscreen/PointerLock warnings are initialized with hidden="true", but change to hidden="" after being shown and hidden again. I think this started happening when we began using HTML elements instead of XUL as they handle hidden attribute differently. Differential Revision: https://phabricator.services.mozilla.com/D177790 - - - - - 524c7622 by Edgar Chen at 2023-07-31T17:56:38+00:00 Bug 1821884 - Reshow initial fullscreen notification; r=Gijs Depends on D177790 Differential Revision: https://phabricator.services.mozilla.com/D178339 - - - - - 4 changed files: - browser/base/content/browser-fullScreenAndPointerLock.js - browser/base/content/fullscreen-and-pointerlock.inc.xhtml - browser/base/content/test/fullscreen/browser_fullscreen_warning.js - dom/tests/browser/browser_pointerlock_warning.js Changes: ===================================== browser/base/content/browser-fullScreenAndPointerLock.js ===================================== @@ -62,9 +62,14 @@ var PointerlockFsWarning = { this._element = document.getElementById(elementId); // Setup event listeners this._element.addEventListener("transitionend", this); + this._element.addEventListener("transitioncancel", this); window.addEventListener("mousemove", this, true); + window.addEventListener("activate", this); + window.addEventListener("deactivate", this); // The timeout to hide the warning box after a while. this._timeoutHide = new this.Timeout(() => { + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); this._state = "hidden"; }, timeout); // The timeout to show the warning box when the pointer is at the top @@ -116,11 +121,10 @@ var PointerlockFsWarning = { return; } - // Explicitly set the last state to hidden to avoid the warning - // box being hidden immediately because of mousemove. - this._state = "onscreen"; - this._lastState = "hidden"; - this._timeoutHide.start(); + if (Services.focus.activeWindow == window) { + this._state = "onscreen"; + this._timeoutHide.start(); + } }, /** @@ -148,7 +152,10 @@ var PointerlockFsWarning = { this._element.hidden = true; // Remove all event listeners this._element.removeEventListener("transitionend", this); + this._element.removeEventListener("transitioncancel", this); window.removeEventListener("mousemove", this, true); + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); // Clear fields this._element = null; this._timeoutHide = null; @@ -186,7 +193,7 @@ var PointerlockFsWarning = { } if (newState != "hidden") { if (currentState != "hidden") { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } else { // When the previous state is hidden, the display was none, // thus no box was constructed. We need to wait for the new @@ -197,7 +204,7 @@ var PointerlockFsWarning = { requestAnimationFrame(() => { requestAnimationFrame(() => { if (this._element) { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } }); }); @@ -217,7 +224,7 @@ var PointerlockFsWarning = { } else if (this._timeoutShow.delay >= 0) { this._timeoutShow.start(); } - } else { + } else if (state != "onscreen") { let elemRect = this._element.getBoundingClientRect(); if (state == "hiding" && this._lastState != "hidden") { // If we are on the hiding transition, and the pointer @@ -239,12 +246,23 @@ var PointerlockFsWarning = { } break; } - case "transitionend": { + case "transitionend": + case "transitioncancel": { if (this._state == "hiding") { this._element.hidden = true; } break; } + case "activate": { + this._state = "onscreen"; + this._timeoutHide.start(); + break; + } + case "deactivate": { + this._state = "hidden"; + this._timeoutHide.cancel(); + break; + } } }, }; ===================================== browser/base/content/fullscreen-and-pointerlock.inc.xhtml ===================================== @@ -3,7 +3,7 @@ # file, You can obtain one at http://mozilla.org/MPL/2.0/. <html:div id="fullscreen-and-pointerlock-wrapper"> - <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> @@ -20,7 +20,7 @@ </html:button> </html:div> - <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> ===================================== browser/base/content/test/fullscreen/browser_fullscreen_warning.js ===================================== @@ -3,14 +3,35 @@ "use strict"; -add_task(async function test_fullscreen_display_none() { +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + +add_setup(async function init() { await SpecialPowers.pushPrefEnv({ set: [ ["full-screen-api.enabled", true], ["full-screen-api.allow-trusted-requests-only", false], ], }); +}); +add_task(async function test_fullscreen_display_none() { await BrowserTestUtils.withNewTab( { gBrowser, @@ -30,11 +51,13 @@ add_task(async function test_fullscreen_display_none() { }, async function (browser) { let warning = document.getElementById("fullscreen-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( warning, - "true" + "hidden", + "Should not show full screen warning initially" ); + + let warningShownPromise = waitForWarningState(warning, "onscreen"); // Enter fullscreen await SpecialPowers.spawn(browser, [], async () => { let frame = content.document.querySelector("iframe"); @@ -54,39 +77,33 @@ add_task(async function test_fullscreen_display_none() { ); document.getElementById("fullscreen-exit-button").click(); await exitFullscreenPromise; + + checkWarningState( + warning, + "hidden", + "Should hide fullscreen warning after exiting fullscreen" + ); } ); }); add_task(async function test_fullscreen_pointerlock_conflict() { - await SpecialPowers.pushPrefEnv({ - set: [ - ["full-screen-api.enabled", true], - ["full-screen-api.allow-trusted-requests-only", false], - ], - }); - await BrowserTestUtils.withNewTab("https://example.com", async browser => { let fsWarning = document.getElementById("fullscreen-warning"); let plWarning = document.getElementById("pointerlock-warning"); - is( - fsWarning.getAttribute("onscreen"), - null, - "Should not show full screen warning initially." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning initially." - ); - - let fsWarningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( fsWarning, - "true" + "hidden", + "Should not show full screen warning initially" + ); + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning initially" ); + let fsWarningShownPromise = waitForWarningState(fsWarning, "onscreen"); info("Entering full screen and pointer lock."); await SpecialPowers.spawn(browser, [], async () => { await content.document.body.requestFullscreen(); @@ -94,15 +111,10 @@ add_task(async function test_fullscreen_pointerlock_conflict() { }); await fsWarningShownPromise; - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should show full screen warning." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); info("Exiting pointerlock"); @@ -110,18 +122,19 @@ add_task(async function test_fullscreen_pointerlock_conflict() { await content.document.exitPointerLock(); }); - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should still show full screen warning." + checkWarningState( + fsWarning, + "onscreen", + "Should still show full screen warning" ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); // Cleanup + info("Exiting fullscreen"); await document.exitFullscreen(); }); }); ===================================== dom/tests/browser/browser_pointerlock_warning.js ===================================== @@ -15,6 +15,25 @@ const FRAME_TEST_URL = encodeURI(BODY_URL) + '"></iframe></body>'; +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + // Make sure the pointerlock warning is shown and exited with the escape key add_task(async function show_pointerlock_warning_escape() { let urls = [TEST_URL, FRAME_TEST_URL]; @@ -24,11 +43,7 @@ add_task(async function show_pointerlock_warning_escape() { let tab = await BrowserTestUtils.openNewForegroundTab(gBrowser, url); let warning = document.getElementById("pointerlock-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", - warning, - "true" - ); + let warningShownPromise = waitForWarningState(warning, "onscreen"); let expectedWarningText; @@ -49,11 +64,7 @@ add_task(async function show_pointerlock_warning_escape() { ok(true, "Pointerlock warning shown"); - let warningHiddenPromise = BrowserTestUtils.waitForAttribute( - "hidden", - warning, - "" - ); + let warningHiddenPromise = waitForWarningState(warning, "hidden"); await BrowserTestUtils.waitForCondition( () => warning.innerText == expectedWarningText, View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/f3faeb… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/f3faeb… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][tor-browser-115.0.2esr-13.0-1] Deleted 2 commits: Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs
by richard (＠richard) 31 Jul '23

31 Jul '23

richard pushed to branch tor-browser-115.0.2esr-13.0-1 at The Tor Project / Applications / Tor Browser WARNING: The push did not contain any new commits, but force pushed to delete the commits and changes below. Deleted commits: 6a6c68cf by Edgar Chen at 2023-07-31T17:56:02+00:00 Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs Fullscreen/PointerLock warnings are initialized with hidden="true", but change to hidden="" after being shown and hidden again. I think this started happening when we began using HTML elements instead of XUL as they handle hidden attribute differently. Differential Revision: https://phabricator.services.mozilla.com/D177790 - - - - - 5e8c1b08 by Edgar Chen at 2023-07-31T17:56:07+00:00 Bug 1821884 - Reshow initial fullscreen notification; r=Gijs Depends on D177790 Differential Revision: https://phabricator.services.mozilla.com/D178339 - - - - - 4 changed files: - browser/base/content/browser-fullScreenAndPointerLock.js - browser/base/content/fullscreen-and-pointerlock.inc.xhtml - browser/base/content/test/fullscreen/browser_fullscreen_warning.js - dom/tests/browser/browser_pointerlock_warning.js Changes: ===================================== browser/base/content/browser-fullScreenAndPointerLock.js ===================================== @@ -62,9 +62,14 @@ var PointerlockFsWarning = { this._element = document.getElementById(elementId); // Setup event listeners this._element.addEventListener("transitionend", this); + this._element.addEventListener("transitioncancel", this); window.addEventListener("mousemove", this, true); + window.addEventListener("activate", this); + window.addEventListener("deactivate", this); // The timeout to hide the warning box after a while. this._timeoutHide = new this.Timeout(() => { + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); this._state = "hidden"; }, timeout); // The timeout to show the warning box when the pointer is at the top @@ -116,11 +121,10 @@ var PointerlockFsWarning = { return; } - // Explicitly set the last state to hidden to avoid the warning - // box being hidden immediately because of mousemove. - this._state = "onscreen"; - this._lastState = "hidden"; - this._timeoutHide.start(); + if (Services.focus.activeWindow == window) { + this._state = "onscreen"; + this._timeoutHide.start(); + } }, /** @@ -148,7 +152,10 @@ var PointerlockFsWarning = { this._element.hidden = true; // Remove all event listeners this._element.removeEventListener("transitionend", this); + this._element.removeEventListener("transitioncancel", this); window.removeEventListener("mousemove", this, true); + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); // Clear fields this._element = null; this._timeoutHide = null; @@ -186,7 +193,7 @@ var PointerlockFsWarning = { } if (newState != "hidden") { if (currentState != "hidden") { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } else { // When the previous state is hidden, the display was none, // thus no box was constructed. We need to wait for the new @@ -197,7 +204,7 @@ var PointerlockFsWarning = { requestAnimationFrame(() => { requestAnimationFrame(() => { if (this._element) { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } }); }); @@ -217,7 +224,7 @@ var PointerlockFsWarning = { } else if (this._timeoutShow.delay >= 0) { this._timeoutShow.start(); } - } else { + } else if (state != "onscreen") { let elemRect = this._element.getBoundingClientRect(); if (state == "hiding" && this._lastState != "hidden") { // If we are on the hiding transition, and the pointer @@ -239,12 +246,23 @@ var PointerlockFsWarning = { } break; } - case "transitionend": { + case "transitionend": + case "transitioncancel": { if (this._state == "hiding") { this._element.hidden = true; } break; } + case "activate": { + this._state = "onscreen"; + this._timeoutHide.start(); + break; + } + case "deactivate": { + this._state = "hidden"; + this._timeoutHide.cancel(); + break; + } } }, }; ===================================== browser/base/content/fullscreen-and-pointerlock.inc.xhtml ===================================== @@ -3,7 +3,7 @@ # file, You can obtain one at http://mozilla.org/MPL/2.0/. <html:div id="fullscreen-and-pointerlock-wrapper"> - <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> @@ -20,7 +20,7 @@ </html:button> </html:div> - <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> ===================================== browser/base/content/test/fullscreen/browser_fullscreen_warning.js ===================================== @@ -3,14 +3,35 @@ "use strict"; -add_task(async function test_fullscreen_display_none() { +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + +add_setup(async function init() { await SpecialPowers.pushPrefEnv({ set: [ ["full-screen-api.enabled", true], ["full-screen-api.allow-trusted-requests-only", false], ], }); +}); +add_task(async function test_fullscreen_display_none() { await BrowserTestUtils.withNewTab( { gBrowser, @@ -30,11 +51,13 @@ add_task(async function test_fullscreen_display_none() { }, async function (browser) { let warning = document.getElementById("fullscreen-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( warning, - "true" + "hidden", + "Should not show full screen warning initially" ); + + let warningShownPromise = waitForWarningState(warning, "onscreen"); // Enter fullscreen await SpecialPowers.spawn(browser, [], async () => { let frame = content.document.querySelector("iframe"); @@ -54,39 +77,33 @@ add_task(async function test_fullscreen_display_none() { ); document.getElementById("fullscreen-exit-button").click(); await exitFullscreenPromise; + + checkWarningState( + warning, + "hidden", + "Should hide fullscreen warning after exiting fullscreen" + ); } ); }); add_task(async function test_fullscreen_pointerlock_conflict() { - await SpecialPowers.pushPrefEnv({ - set: [ - ["full-screen-api.enabled", true], - ["full-screen-api.allow-trusted-requests-only", false], - ], - }); - await BrowserTestUtils.withNewTab("https://example.com", async browser => { let fsWarning = document.getElementById("fullscreen-warning"); let plWarning = document.getElementById("pointerlock-warning"); - is( - fsWarning.getAttribute("onscreen"), - null, - "Should not show full screen warning initially." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning initially." - ); - - let fsWarningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( fsWarning, - "true" + "hidden", + "Should not show full screen warning initially" + ); + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning initially" ); + let fsWarningShownPromise = waitForWarningState(fsWarning, "onscreen"); info("Entering full screen and pointer lock."); await SpecialPowers.spawn(browser, [], async () => { await content.document.body.requestFullscreen(); @@ -94,15 +111,10 @@ add_task(async function test_fullscreen_pointerlock_conflict() { }); await fsWarningShownPromise; - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should show full screen warning." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); info("Exiting pointerlock"); @@ -110,18 +122,19 @@ add_task(async function test_fullscreen_pointerlock_conflict() { await content.document.exitPointerLock(); }); - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should still show full screen warning." + checkWarningState( + fsWarning, + "onscreen", + "Should still show full screen warning" ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); // Cleanup + info("Exiting fullscreen"); await document.exitFullscreen(); }); }); ===================================== dom/tests/browser/browser_pointerlock_warning.js ===================================== @@ -15,6 +15,25 @@ const FRAME_TEST_URL = encodeURI(BODY_URL) + '"></iframe></body>'; +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + // Make sure the pointerlock warning is shown and exited with the escape key add_task(async function show_pointerlock_warning_escape() { let urls = [TEST_URL, FRAME_TEST_URL]; @@ -24,11 +43,7 @@ add_task(async function show_pointerlock_warning_escape() { let tab = await BrowserTestUtils.openNewForegroundTab(gBrowser, url); let warning = document.getElementById("pointerlock-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", - warning, - "true" - ); + let warningShownPromise = waitForWarningState(warning, "onscreen"); let expectedWarningText; @@ -49,11 +64,7 @@ add_task(async function show_pointerlock_warning_escape() { ok(true, "Pointerlock warning shown"); - let warningHiddenPromise = BrowserTestUtils.waitForAttribute( - "hidden", - warning, - "" - ); + let warningHiddenPromise = waitForWarningState(warning, "hidden"); await BrowserTestUtils.waitForCondition( () => warning.innerText == expectedWarningText, View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/93ef5b… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/93ef5b… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][tor-browser-115.0.2esr-13.0-1] 2 commits: Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs
by richard (＠richard) 31 Jul '23

31 Jul '23

richard pushed to branch tor-browser-115.0.2esr-13.0-1 at The Tor Project / Applications / Tor Browser Commits: 6a6c68cf by Edgar Chen at 2023-07-31T17:56:02+00:00 Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs Fullscreen/PointerLock warnings are initialized with hidden="true", but change to hidden="" after being shown and hidden again. I think this started happening when we began using HTML elements instead of XUL as they handle hidden attribute differently. Differential Revision: https://phabricator.services.mozilla.com/D177790 - - - - - 5e8c1b08 by Edgar Chen at 2023-07-31T17:56:07+00:00 Bug 1821884 - Reshow initial fullscreen notification; r=Gijs Depends on D177790 Differential Revision: https://phabricator.services.mozilla.com/D178339 - - - - - 4 changed files: - browser/base/content/browser-fullScreenAndPointerLock.js - browser/base/content/fullscreen-and-pointerlock.inc.xhtml - browser/base/content/test/fullscreen/browser_fullscreen_warning.js - dom/tests/browser/browser_pointerlock_warning.js Changes: ===================================== browser/base/content/browser-fullScreenAndPointerLock.js ===================================== @@ -62,9 +62,14 @@ var PointerlockFsWarning = { this._element = document.getElementById(elementId); // Setup event listeners this._element.addEventListener("transitionend", this); + this._element.addEventListener("transitioncancel", this); window.addEventListener("mousemove", this, true); + window.addEventListener("activate", this); + window.addEventListener("deactivate", this); // The timeout to hide the warning box after a while. this._timeoutHide = new this.Timeout(() => { + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); this._state = "hidden"; }, timeout); // The timeout to show the warning box when the pointer is at the top @@ -116,11 +121,10 @@ var PointerlockFsWarning = { return; } - // Explicitly set the last state to hidden to avoid the warning - // box being hidden immediately because of mousemove. - this._state = "onscreen"; - this._lastState = "hidden"; - this._timeoutHide.start(); + if (Services.focus.activeWindow == window) { + this._state = "onscreen"; + this._timeoutHide.start(); + } }, /** @@ -148,7 +152,10 @@ var PointerlockFsWarning = { this._element.hidden = true; // Remove all event listeners this._element.removeEventListener("transitionend", this); + this._element.removeEventListener("transitioncancel", this); window.removeEventListener("mousemove", this, true); + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); // Clear fields this._element = null; this._timeoutHide = null; @@ -186,7 +193,7 @@ var PointerlockFsWarning = { } if (newState != "hidden") { if (currentState != "hidden") { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } else { // When the previous state is hidden, the display was none, // thus no box was constructed. We need to wait for the new @@ -197,7 +204,7 @@ var PointerlockFsWarning = { requestAnimationFrame(() => { requestAnimationFrame(() => { if (this._element) { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } }); }); @@ -217,7 +224,7 @@ var PointerlockFsWarning = { } else if (this._timeoutShow.delay >= 0) { this._timeoutShow.start(); } - } else { + } else if (state != "onscreen") { let elemRect = this._element.getBoundingClientRect(); if (state == "hiding" && this._lastState != "hidden") { // If we are on the hiding transition, and the pointer @@ -239,12 +246,23 @@ var PointerlockFsWarning = { } break; } - case "transitionend": { + case "transitionend": + case "transitioncancel": { if (this._state == "hiding") { this._element.hidden = true; } break; } + case "activate": { + this._state = "onscreen"; + this._timeoutHide.start(); + break; + } + case "deactivate": { + this._state = "hidden"; + this._timeoutHide.cancel(); + break; + } } }, }; ===================================== browser/base/content/fullscreen-and-pointerlock.inc.xhtml ===================================== @@ -3,7 +3,7 @@ # file, You can obtain one at http://mozilla.org/MPL/2.0/. <html:div id="fullscreen-and-pointerlock-wrapper"> - <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> @@ -20,7 +20,7 @@ </html:button> </html:div> - <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> ===================================== browser/base/content/test/fullscreen/browser_fullscreen_warning.js ===================================== @@ -3,14 +3,35 @@ "use strict"; -add_task(async function test_fullscreen_display_none() { +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + +add_setup(async function init() { await SpecialPowers.pushPrefEnv({ set: [ ["full-screen-api.enabled", true], ["full-screen-api.allow-trusted-requests-only", false], ], }); +}); +add_task(async function test_fullscreen_display_none() { await BrowserTestUtils.withNewTab( { gBrowser, @@ -30,11 +51,13 @@ add_task(async function test_fullscreen_display_none() { }, async function (browser) { let warning = document.getElementById("fullscreen-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( warning, - "true" + "hidden", + "Should not show full screen warning initially" ); + + let warningShownPromise = waitForWarningState(warning, "onscreen"); // Enter fullscreen await SpecialPowers.spawn(browser, [], async () => { let frame = content.document.querySelector("iframe"); @@ -54,39 +77,33 @@ add_task(async function test_fullscreen_display_none() { ); document.getElementById("fullscreen-exit-button").click(); await exitFullscreenPromise; + + checkWarningState( + warning, + "hidden", + "Should hide fullscreen warning after exiting fullscreen" + ); } ); }); add_task(async function test_fullscreen_pointerlock_conflict() { - await SpecialPowers.pushPrefEnv({ - set: [ - ["full-screen-api.enabled", true], - ["full-screen-api.allow-trusted-requests-only", false], - ], - }); - await BrowserTestUtils.withNewTab("https://example.com", async browser => { let fsWarning = document.getElementById("fullscreen-warning"); let plWarning = document.getElementById("pointerlock-warning"); - is( - fsWarning.getAttribute("onscreen"), - null, - "Should not show full screen warning initially." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning initially." - ); - - let fsWarningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( fsWarning, - "true" + "hidden", + "Should not show full screen warning initially" + ); + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning initially" ); + let fsWarningShownPromise = waitForWarningState(fsWarning, "onscreen"); info("Entering full screen and pointer lock."); await SpecialPowers.spawn(browser, [], async () => { await content.document.body.requestFullscreen(); @@ -94,15 +111,10 @@ add_task(async function test_fullscreen_pointerlock_conflict() { }); await fsWarningShownPromise; - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should show full screen warning." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); info("Exiting pointerlock"); @@ -110,18 +122,19 @@ add_task(async function test_fullscreen_pointerlock_conflict() { await content.document.exitPointerLock(); }); - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should still show full screen warning." + checkWarningState( + fsWarning, + "onscreen", + "Should still show full screen warning" ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); // Cleanup + info("Exiting fullscreen"); await document.exitFullscreen(); }); }); ===================================== dom/tests/browser/browser_pointerlock_warning.js ===================================== @@ -15,6 +15,25 @@ const FRAME_TEST_URL = encodeURI(BODY_URL) + '"></iframe></body>'; +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + // Make sure the pointerlock warning is shown and exited with the escape key add_task(async function show_pointerlock_warning_escape() { let urls = [TEST_URL, FRAME_TEST_URL]; @@ -24,11 +43,7 @@ add_task(async function show_pointerlock_warning_escape() { let tab = await BrowserTestUtils.openNewForegroundTab(gBrowser, url); let warning = document.getElementById("pointerlock-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", - warning, - "true" - ); + let warningShownPromise = waitForWarningState(warning, "onscreen"); let expectedWarningText; @@ -49,11 +64,7 @@ add_task(async function show_pointerlock_warning_escape() { ok(true, "Pointerlock warning shown"); - let warningHiddenPromise = BrowserTestUtils.waitForAttribute( - "hidden", - warning, - "" - ); + let warningHiddenPromise = waitForWarningState(warning, "hidden"); await BrowserTestUtils.waitForCondition( () => warning.innerText == expectedWarningText, View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/93ef5b… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/93ef5b… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][base-browser-115.0.2esr-13.0-1] 2 commits: Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs
by richard (＠richard) 31 Jul '23

31 Jul '23

richard pushed to branch base-browser-115.0.2esr-13.0-1 at The Tor Project / Applications / Tor Browser Commits: 60fefea4 by Edgar Chen at 2023-07-31T17:56:34+00:00 Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs Fullscreen/PointerLock warnings are initialized with hidden="true", but change to hidden="" after being shown and hidden again. I think this started happening when we began using HTML elements instead of XUL as they handle hidden attribute differently. Differential Revision: https://phabricator.services.mozilla.com/D177790 - - - - - 524c7622 by Edgar Chen at 2023-07-31T17:56:38+00:00 Bug 1821884 - Reshow initial fullscreen notification; r=Gijs Depends on D177790 Differential Revision: https://phabricator.services.mozilla.com/D178339 - - - - - 4 changed files: - browser/base/content/browser-fullScreenAndPointerLock.js - browser/base/content/fullscreen-and-pointerlock.inc.xhtml - browser/base/content/test/fullscreen/browser_fullscreen_warning.js - dom/tests/browser/browser_pointerlock_warning.js Changes: ===================================== browser/base/content/browser-fullScreenAndPointerLock.js ===================================== @@ -62,9 +62,14 @@ var PointerlockFsWarning = { this._element = document.getElementById(elementId); // Setup event listeners this._element.addEventListener("transitionend", this); + this._element.addEventListener("transitioncancel", this); window.addEventListener("mousemove", this, true); + window.addEventListener("activate", this); + window.addEventListener("deactivate", this); // The timeout to hide the warning box after a while. this._timeoutHide = new this.Timeout(() => { + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); this._state = "hidden"; }, timeout); // The timeout to show the warning box when the pointer is at the top @@ -116,11 +121,10 @@ var PointerlockFsWarning = { return; } - // Explicitly set the last state to hidden to avoid the warning - // box being hidden immediately because of mousemove. - this._state = "onscreen"; - this._lastState = "hidden"; - this._timeoutHide.start(); + if (Services.focus.activeWindow == window) { + this._state = "onscreen"; + this._timeoutHide.start(); + } }, /** @@ -148,7 +152,10 @@ var PointerlockFsWarning = { this._element.hidden = true; // Remove all event listeners this._element.removeEventListener("transitionend", this); + this._element.removeEventListener("transitioncancel", this); window.removeEventListener("mousemove", this, true); + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); // Clear fields this._element = null; this._timeoutHide = null; @@ -186,7 +193,7 @@ var PointerlockFsWarning = { } if (newState != "hidden") { if (currentState != "hidden") { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } else { // When the previous state is hidden, the display was none, // thus no box was constructed. We need to wait for the new @@ -197,7 +204,7 @@ var PointerlockFsWarning = { requestAnimationFrame(() => { requestAnimationFrame(() => { if (this._element) { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } }); }); @@ -217,7 +224,7 @@ var PointerlockFsWarning = { } else if (this._timeoutShow.delay >= 0) { this._timeoutShow.start(); } - } else { + } else if (state != "onscreen") { let elemRect = this._element.getBoundingClientRect(); if (state == "hiding" && this._lastState != "hidden") { // If we are on the hiding transition, and the pointer @@ -239,12 +246,23 @@ var PointerlockFsWarning = { } break; } - case "transitionend": { + case "transitionend": + case "transitioncancel": { if (this._state == "hiding") { this._element.hidden = true; } break; } + case "activate": { + this._state = "onscreen"; + this._timeoutHide.start(); + break; + } + case "deactivate": { + this._state = "hidden"; + this._timeoutHide.cancel(); + break; + } } }, }; ===================================== browser/base/content/fullscreen-and-pointerlock.inc.xhtml ===================================== @@ -3,7 +3,7 @@ # file, You can obtain one at http://mozilla.org/MPL/2.0/. <html:div id="fullscreen-and-pointerlock-wrapper"> - <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> @@ -20,7 +20,7 @@ </html:button> </html:div> - <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> ===================================== browser/base/content/test/fullscreen/browser_fullscreen_warning.js ===================================== @@ -3,14 +3,35 @@ "use strict"; -add_task(async function test_fullscreen_display_none() { +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + +add_setup(async function init() { await SpecialPowers.pushPrefEnv({ set: [ ["full-screen-api.enabled", true], ["full-screen-api.allow-trusted-requests-only", false], ], }); +}); +add_task(async function test_fullscreen_display_none() { await BrowserTestUtils.withNewTab( { gBrowser, @@ -30,11 +51,13 @@ add_task(async function test_fullscreen_display_none() { }, async function (browser) { let warning = document.getElementById("fullscreen-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( warning, - "true" + "hidden", + "Should not show full screen warning initially" ); + + let warningShownPromise = waitForWarningState(warning, "onscreen"); // Enter fullscreen await SpecialPowers.spawn(browser, [], async () => { let frame = content.document.querySelector("iframe"); @@ -54,39 +77,33 @@ add_task(async function test_fullscreen_display_none() { ); document.getElementById("fullscreen-exit-button").click(); await exitFullscreenPromise; + + checkWarningState( + warning, + "hidden", + "Should hide fullscreen warning after exiting fullscreen" + ); } ); }); add_task(async function test_fullscreen_pointerlock_conflict() { - await SpecialPowers.pushPrefEnv({ - set: [ - ["full-screen-api.enabled", true], - ["full-screen-api.allow-trusted-requests-only", false], - ], - }); - await BrowserTestUtils.withNewTab("https://example.com", async browser => { let fsWarning = document.getElementById("fullscreen-warning"); let plWarning = document.getElementById("pointerlock-warning"); - is( - fsWarning.getAttribute("onscreen"), - null, - "Should not show full screen warning initially." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning initially." - ); - - let fsWarningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( fsWarning, - "true" + "hidden", + "Should not show full screen warning initially" + ); + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning initially" ); + let fsWarningShownPromise = waitForWarningState(fsWarning, "onscreen"); info("Entering full screen and pointer lock."); await SpecialPowers.spawn(browser, [], async () => { await content.document.body.requestFullscreen(); @@ -94,15 +111,10 @@ add_task(async function test_fullscreen_pointerlock_conflict() { }); await fsWarningShownPromise; - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should show full screen warning." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); info("Exiting pointerlock"); @@ -110,18 +122,19 @@ add_task(async function test_fullscreen_pointerlock_conflict() { await content.document.exitPointerLock(); }); - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should still show full screen warning." + checkWarningState( + fsWarning, + "onscreen", + "Should still show full screen warning" ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); // Cleanup + info("Exiting fullscreen"); await document.exitFullscreen(); }); }); ===================================== dom/tests/browser/browser_pointerlock_warning.js ===================================== @@ -15,6 +15,25 @@ const FRAME_TEST_URL = encodeURI(BODY_URL) + '"></iframe></body>'; +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + // Make sure the pointerlock warning is shown and exited with the escape key add_task(async function show_pointerlock_warning_escape() { let urls = [TEST_URL, FRAME_TEST_URL]; @@ -24,11 +43,7 @@ add_task(async function show_pointerlock_warning_escape() { let tab = await BrowserTestUtils.openNewForegroundTab(gBrowser, url); let warning = document.getElementById("pointerlock-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", - warning, - "true" - ); + let warningShownPromise = waitForWarningState(warning, "onscreen"); let expectedWarningText; @@ -49,11 +64,7 @@ add_task(async function show_pointerlock_warning_escape() { ok(true, "Pointerlock warning shown"); - let warningHiddenPromise = BrowserTestUtils.waitForAttribute( - "hidden", - warning, - "" - ); + let warningHiddenPromise = waitForWarningState(warning, "hidden"); await BrowserTestUtils.waitForCondition( () => warning.innerText == expectedWarningText, View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/f3faeb… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/f3faeb… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][tor-browser-102.14.0esr-12.5-1] 2 commits: Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs
by richard (＠richard) 31 Jul '23

31 Jul '23

richard pushed to branch tor-browser-102.14.0esr-12.5-1 at The Tor Project / Applications / Tor Browser Commits: f70fa908 by Edgar Chen at 2023-07-28T23:08:31+02:00 Bug 1821884 - Ensure consistent state for fullscreen/pointerlock warnings; r=Gijs Fullscreen/PointerLock warnings are initialized with hidden="true", but change to hidden="" after being shown and hidden again. I think this started happening when we began using HTML elements instead of XUL as they handle hidden attribute differently. Differential Revision: https://phabricator.services.mozilla.com/D177790 - - - - - d27c6a1b by Edgar Chen at 2023-07-28T23:09:46+02:00 Bug 1821884 - Reshow initial fullscreen notification; r=Gijs Depends on D177790 Differential Revision: https://phabricator.services.mozilla.com/D178339 - - - - - 4 changed files: - browser/base/content/browser-fullScreenAndPointerLock.js - browser/base/content/fullscreen-and-pointerlock.inc.xhtml - browser/base/content/test/fullscreen/browser_fullscreen_warning.js - dom/tests/browser/browser_pointerlock_warning.js Changes: ===================================== browser/base/content/browser-fullScreenAndPointerLock.js ===================================== @@ -62,9 +62,14 @@ var PointerlockFsWarning = { this._element = document.getElementById(elementId); // Setup event listeners this._element.addEventListener("transitionend", this); + this._element.addEventListener("transitioncancel", this); window.addEventListener("mousemove", this, true); + window.addEventListener("activate", this); + window.addEventListener("deactivate", this); // The timeout to hide the warning box after a while. this._timeoutHide = new this.Timeout(() => { + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); this._state = "hidden"; }, timeout); // The timeout to show the warning box when the pointer is at the top @@ -116,11 +121,10 @@ var PointerlockFsWarning = { return; } - // Explicitly set the last state to hidden to avoid the warning - // box being hidden immediately because of mousemove. - this._state = "onscreen"; - this._lastState = "hidden"; - this._timeoutHide.start(); + if (Services.focus.activeWindow == window) { + this._state = "onscreen"; + this._timeoutHide.start(); + } }, /** @@ -148,7 +152,10 @@ var PointerlockFsWarning = { this._element.hidden = true; // Remove all event listeners this._element.removeEventListener("transitionend", this); + this._element.removeEventListener("transitioncancel", this); window.removeEventListener("mousemove", this, true); + window.removeEventListener("activate", this); + window.removeEventListener("deactivate", this); // Clear fields this._element = null; this._timeoutHide = null; @@ -186,7 +193,7 @@ var PointerlockFsWarning = { } if (newState != "hidden") { if (currentState != "hidden") { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } else { // When the previous state is hidden, the display was none, // thus no box was constructed. We need to wait for the new @@ -197,7 +204,7 @@ var PointerlockFsWarning = { requestAnimationFrame(() => { requestAnimationFrame(() => { if (this._element) { - this._element.setAttribute(newState, true); + this._element.setAttribute(newState, ""); } }); }); @@ -217,7 +224,7 @@ var PointerlockFsWarning = { } else if (this._timeoutShow.delay >= 0) { this._timeoutShow.start(); } - } else { + } else if (state != "onscreen") { let elemRect = this._element.getBoundingClientRect(); if (state == "hiding" && this._lastState != "hidden") { // If we are on the hiding transition, and the pointer @@ -239,12 +246,23 @@ var PointerlockFsWarning = { } break; } - case "transitionend": { + case "transitionend": + case "transitioncancel": { if (this._state == "hiding") { this._element.hidden = true; } break; } + case "activate": { + this._state = "onscreen"; + this._timeoutHide.start(); + break; + } + case "deactivate": { + this._state = "hidden"; + this._timeoutHide.cancel(); + break; + } } }, }; ===================================== browser/base/content/fullscreen-and-pointerlock.inc.xhtml ===================================== @@ -3,7 +3,7 @@ # file, You can obtain one at http://mozilla.org/MPL/2.0/. <html:div id="fullscreen-and-pointerlock-wrapper"> - <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="fullscreen-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> @@ -20,7 +20,7 @@ </html:button> </html:div> - <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden="true"> + <html:div id="pointerlock-warning" class="pointerlockfswarning" hidden=""> <html:div class="pointerlockfswarning-domain-text"> <html:span class="pointerlockfswarning-domain" data-l10n-name="domain"/> </html:div> ===================================== browser/base/content/test/fullscreen/browser_fullscreen_warning.js ===================================== @@ -3,14 +3,35 @@ "use strict"; -add_task(async function test_fullscreen_display_none() { +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + +add_setup(async function init() { await SpecialPowers.pushPrefEnv({ set: [ ["full-screen-api.enabled", true], ["full-screen-api.allow-trusted-requests-only", false], ], }); +}); +add_task(async function test_fullscreen_display_none() { await BrowserTestUtils.withNewTab( { gBrowser, @@ -30,11 +51,13 @@ add_task(async function test_fullscreen_display_none() { }, async function(browser) { let warning = document.getElementById("fullscreen-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( warning, - "true" + "hidden", + "Should not show full screen warning initially" ); + + let warningShownPromise = waitForWarningState(warning, "onscreen"); // Enter fullscreen await SpecialPowers.spawn(browser, [], async () => { let frame = content.document.querySelector("iframe"); @@ -54,39 +77,33 @@ add_task(async function test_fullscreen_display_none() { ); document.getElementById("fullscreen-exit-button").click(); await exitFullscreenPromise; + + checkWarningState( + warning, + "hidden", + "Should hide fullscreen warning after exiting fullscreen" + ); } ); }); add_task(async function test_fullscreen_pointerlock_conflict() { - await SpecialPowers.pushPrefEnv({ - set: [ - ["full-screen-api.enabled", true], - ["full-screen-api.allow-trusted-requests-only", false], - ], - }); - await BrowserTestUtils.withNewTab("https://example.com", async browser => { let fsWarning = document.getElementById("fullscreen-warning"); let plWarning = document.getElementById("pointerlock-warning"); - is( - fsWarning.getAttribute("onscreen"), - null, - "Should not show full screen warning initially." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning initially." - ); - - let fsWarningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", + checkWarningState( fsWarning, - "true" + "hidden", + "Should not show full screen warning initially" + ); + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning initially" ); + let fsWarningShownPromise = waitForWarningState(fsWarning, "onscreen"); info("Entering full screen and pointer lock."); await SpecialPowers.spawn(browser, [], async () => { await content.document.body.requestFullscreen(); @@ -94,15 +111,10 @@ add_task(async function test_fullscreen_pointerlock_conflict() { }); await fsWarningShownPromise; - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should show full screen warning." - ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); info("Exiting pointerlock"); @@ -110,18 +122,19 @@ add_task(async function test_fullscreen_pointerlock_conflict() { await content.document.exitPointerLock(); }); - is( - fsWarning.getAttribute("onscreen"), - "true", - "Should still show full screen warning." + checkWarningState( + fsWarning, + "onscreen", + "Should still show full screen warning" ); - is( - plWarning.getAttribute("onscreen"), - null, - "Should not show pointer lock warning." + checkWarningState( + plWarning, + "hidden", + "Should not show pointer lock warning" ); // Cleanup + info("Exiting fullscreen"); await document.exitFullscreen(); }); }); ===================================== dom/tests/browser/browser_pointerlock_warning.js ===================================== @@ -15,6 +15,25 @@ const FRAME_TEST_URL = encodeURI(BODY_URL) + '"></iframe></body>'; +function checkWarningState(aWarningElement, aExpectedState, aMsg) { + ["hidden", "ontop", "onscreen"].forEach(state => { + is( + aWarningElement.hasAttribute(state), + state == aExpectedState, + `${aMsg} - check ${state} attribute.` + ); + }); +} + +async function waitForWarningState(aWarningElement, aExpectedState) { + await BrowserTestUtils.waitForAttribute(aExpectedState, aWarningElement, ""); + checkWarningState( + aWarningElement, + aExpectedState, + `Wait for ${aExpectedState} state` + ); +} + // Make sure the pointerlock warning is shown and exited with the escape key add_task(async function show_pointerlock_warning_escape() { let urls = [TEST_URL, FRAME_TEST_URL]; @@ -24,11 +43,7 @@ add_task(async function show_pointerlock_warning_escape() { let tab = await BrowserTestUtils.openNewForegroundTab(gBrowser, url); let warning = document.getElementById("pointerlock-warning"); - let warningShownPromise = BrowserTestUtils.waitForAttribute( - "onscreen", - warning, - "true" - ); + let warningShownPromise = waitForWarningState(warning, "onscreen"); let expectedWarningText; @@ -49,11 +64,7 @@ add_task(async function show_pointerlock_warning_escape() { ok(true, "Pointerlock warning shown"); - let warningHiddenPromise = BrowserTestUtils.waitForAttribute( - "hidden", - warning, - "" - ); + let warningHiddenPromise = waitForWarningState(warning, "hidden"); await BrowserTestUtils.waitForCondition( () => warning.innerText == expectedWarningText, View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/855b69… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/855b69… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][tor-browser-115.1.0esr-13.0-1] fixup! Bug 40597: Implement TorSettings module
by richard (＠richard) 27 Jul '23

27 Jul '23

richard pushed to branch tor-browser-115.1.0esr-13.0-1 at The Tor Project / Applications / Tor Browser Commits: f0493f9f by Pier Angelo Vendrame at 2023-07-27T18:20:08+00:00 fixup! Bug 40597: Implement TorSettings module Bug 41907: Change state after the process becomes ready only when we are in the initial state. - - - - - 1 changed file: - browser/modules/TorConnect.jsm Changes: ===================================== browser/modules/TorConnect.jsm ===================================== @@ -874,6 +874,18 @@ const TorConnect = (() => { switch (topic) { /* We need to wait until TorSettings have been loaded and applied before we can Quickstart */ case TorSettingsTopics.Ready: { + // tor-browser#41907: This is only a workaround to avoid users being + // bounced back to the initial panel without any explanation. + // Longer term we should disable the clickable elements, or find a UX + // to prevent this from happening (e.g., allow buttons to be clicked, + // but show an intermediate starting state, or a message that tor is + // starting while the butons are disabled, etc...). + if (this.state !== TorConnectState.Initial) { + console.warn( + "TorConnect: Seen the torsettings:ready after the state has already changed, ignoring the notification." + ); + break; + } if (this.shouldQuickStart) { // Quickstart this._changeState(TorConnectState.Bootstrapping); View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/commit/f0493f9… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/commit/f0493f9… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][tor-browser-115.1.0esr-13.0-1] 7 commits: fixup! Bug 40933: Add tor-launcher functionality
by richard (＠richard) 27 Jul '23

27 Jul '23

richard pushed to branch tor-browser-115.1.0esr-13.0-1 at The Tor Project / Applications / Tor Browser Commits: 66229717 by Pier Angelo Vendrame at 2023-07-27T18:11:55+02:00 fixup! Bug 40933: Add tor-launcher functionality Bug 41844: Added a couple of wrappers for Onion Auth on TorProtocolService. - - - - - 2cde9fc3 by Pier Angelo Vendrame at 2023-07-27T18:11:56+02:00 fixup! Bug 30237: Add v3 onion services client authentication prompt Bug 41844: Stop using the control port directly - - - - - b2cd0ee8 by Pier Angelo Vendrame at 2023-07-27T18:11:57+02:00 fixup! Bug 40933: Add tor-launcher functionality Small improvements on event registration. - - - - - 26152fa9 by Pier Angelo Vendrame at 2023-07-27T18:11:57+02:00 fixup! Bug 40933: Add tor-launcher functionality Bug 41844: Do not use a the control port directly. Collect the bridge node for the about:preferences#connection page in TorMonitorService. Also, move parts of the circuit display to TorMonitorService and TorProtocolService. - - - - - 749aeaca by Pier Angelo Vendrame at 2023-07-27T18:11:58+02:00 fixup! Bug 31286: Implementation of bridge, proxy, and firewall settings in about:preferences#connection Bug 41844: Do not use the control port directly Do not use the controller in the settings frontend. Instead, let TorMonitorService collect the first node's fingerprint. - - - - - 20641450 by Pier Angelo Vendrame at 2023-07-27T18:11:58+02:00 fixup! Bug 3455: Add DomainIsolator, for isolating circuit by domain. Bug 41844: Do not use the control port directly. Use TorDomainIsolator also as a backend for the circuit display. - - - - - 1a8be7b1 by Pier Angelo Vendrame at 2023-07-27T18:11:59+02:00 fixup! Bug 41600: Add a tor circuit display panel. Bug 41844: Have a separate backend for the tor circuits Remove the backend stuff from the circuit display. - - - - - 11 changed files: - browser/base/content/browser.js - browser/components/onionservices/content/authPrompt.js - browser/components/onionservices/content/savedKeysDialog.js - browser/components/torcircuit/content/torCircuitPanel.js - browser/components/torpreferences/content/connectionPane.js - toolkit/components/tor-launcher/TorDomainIsolator.jsm → toolkit/components/tor-launcher/TorDomainIsolator.sys.mjs - toolkit/components/tor-launcher/TorMonitorService.sys.mjs - toolkit/components/tor-launcher/TorParsers.sys.mjs - toolkit/components/tor-launcher/TorProtocolService.sys.mjs - toolkit/components/tor-launcher/TorStartupService.sys.mjs - toolkit/components/tor-launcher/moz.build Changes: ===================================== browser/base/content/browser.js ===================================== @@ -66,6 +66,7 @@ ChromeUtils.defineESModuleGetters(this, { TabsSetupFlowManager: "resource:///modules/firefox-view-tabs-setup-manager.sys.mjs", TelemetryEnvironment: "resource://gre/modules/TelemetryEnvironment.sys.mjs", + TorDomainIsolator: "resource://gre/modules/TorDomainIsolator.sys.mjs", TranslationsParent: "resource://gre/actors/TranslationsParent.sys.mjs", UITour: "resource:///modules/UITour.sys.mjs", UpdateUtils: "resource://gre/modules/UpdateUtils.sys.mjs", @@ -100,7 +101,6 @@ XPCOMUtils.defineLazyModuleGetters(this, { TorConnect: "resource:///modules/TorConnect.jsm", TorConnectState: "resource:///modules/TorConnect.jsm", TorConnectTopics: "resource:///modules/TorConnect.jsm", - TorDomainIsolator: "resource://gre/modules/TorDomainIsolator.jsm", Translation: "resource:///modules/translation/TranslationParent.jsm", webrtcUI: "resource:///modules/webrtcUI.jsm", ZoomUI: "resource:///modules/ZoomUI.jsm", ===================================== browser/components/onionservices/content/authPrompt.js ===================================== @@ -7,6 +7,7 @@ XPCOMUtils.defineLazyModuleGetters(this, { OnionAuthUtil: "chrome://browser/content/onionservices/authUtil.jsm", CommonUtils: "resource://services-common/utils.js", + TorProtocolService: "resource://gre/modules/TorProtocolService.jsm", TorStrings: "resource:///modules/TorStrings.jsm", }); @@ -192,10 +193,6 @@ const OnionAuthPrompt = (function () { let controllerFailureMsg = TorStrings.onionServices.authPrompt.failedToSetKey; try { - let { controller } = ChromeUtils.import( - "resource://torbutton/modules/tor-control-port.js" - ); - let torController = await controller(); // ^(subdomain.)*onionserviceid.onion$ (case-insensitive) const onionServiceIdRegExp = /^(.*\.)*(?<onionServiceId>[a-z2-7]{56})\.onion$/i; @@ -206,8 +203,7 @@ const OnionAuthPrompt = (function () { let checkboxElem = this._getCheckboxElement(); let isPermanent = checkboxElem && checkboxElem.checked; - torController - .onionAuthAdd(onionServiceId, base64key, isPermanent) + TorProtocolService.onionAuthAdd(onionServiceId, base64key, isPermanent) .then(aResponse => { // Success! Reload the page. this._browser.sendMessageToActor( ===================================== browser/components/onionservices/content/savedKeysDialog.js ===================================== @@ -10,8 +10,8 @@ ChromeUtils.defineModuleGetter( ChromeUtils.defineModuleGetter( this, - "controller", - "resource://torbutton/modules/tor-control-port.js" + "TorProtocolService", + "resource://gre/modules/TorProtocolService.jsm" ); var gOnionServicesSavedKeysDialog = { @@ -49,11 +49,9 @@ var gOnionServicesSavedKeysDialog = { const controllerFailureMsg = TorStrings.onionServices.authPreferences.failedToRemoveKey; try { - const torController = await controller(); - // Remove in reverse index order to avoid issues caused by index changes. for (let i = indexesToDelete.length - 1; i >= 0; --i) { - await this._deleteOneKey(torController, indexesToDelete[i]); + await this._deleteOneKey(indexesToDelete[i]); } } catch (e) { if (e.torMessage) { @@ -127,8 +125,7 @@ var gOnionServicesSavedKeysDialog = { try { this._tree.view = this; - const torController = await controller(); - const keyInfoList = await torController.onionAuthViewKeys(); + const keyInfoList = await TorProtocolService.onionAuthViewKeys(); if (keyInfoList) { // Filter out temporary keys. this._keyInfoList = keyInfoList.filter(aKeyInfo => { @@ -165,9 +162,9 @@ var gOnionServicesSavedKeysDialog = { }, // This method may throw; callers should catch errors. - async _deleteOneKey(aTorController, aIndex) { + async _deleteOneKey(aIndex) { const keyInfoObj = this._keyInfoList[aIndex]; - await aTorController.onionAuthRemove(keyInfoObj.hsAddress); + await TorProtocolService.onionAuthRemove(keyInfoObj.hsAddress); this._tree.view.selection.clearRange(aIndex, aIndex); this._keyInfoList.splice(aIndex, 1); this._tree.rowCountChanged(aIndex + 1, -1); ===================================== browser/components/torcircuit/content/torCircuitPanel.js ===================================== @@ -1,18 +1,5 @@ /* eslint-env mozilla/browser-window */ -/** - * Stores the data associated with a circuit node. - * - * @typedef NodeData - * @property {string[]} ipAddrs - The ip addresses associated with this node. - * @property {string?} bridgeType - The bridge type for this node, or "" if the - * node is a bridge but the type is unknown, or null if this is not a bridge - * node. - * @property {string?} regionCode - An upper case 2-letter ISO3166-1 code for - * the first ip address, or null if there is no region. This should also be a - * valid BCP47 Region subtag. - */ - /** * Data about the current domain and circuit for a xul:browser. * @@ -35,29 +22,6 @@ var gTorCircuitPanel = { * @type {Element} */ toolbarButton: null, - /** - * A list of IDs for "mature" circuits (those that have conveyed a stream). - * - * @type {string[]} - */ - _knownCircuitIDs: [], - /** - * Stores the circuit nodes for each SOCKS username/password pair. The keys - * are of the form "<username>|<password>". - * - * @type {Map<string, NodeData[]>} - */ - _credentialsToCircuitNodes: new Map(), - /** - * Browser data for their currently shown page. - * - * This data may be stale for a given browser since we only update this data - * when loading a new page in the currently selected browser, when switching - * tabs, or if we find a new circuit for the current browser. - * - * @type {WeakMap<MozBrowser, BrowserCircuitData>} - */ - _browserData: new WeakMap(), /** * The data for the currently shown browser. * @@ -71,6 +35,13 @@ var gTorCircuitPanel = { */ _isActive: false, + /** + * The topic on which circuit changes are broadcast. + * + * @type {string} + */ + TOR_CIRCUIT_TOPIC: "TorCircuitChange", + /** * Initialize the panel. */ @@ -86,31 +57,6 @@ var gTorCircuitPanel = { maxLogLevelPref: "browser.torcircuitpanel.loglevel", }); - const { wait_for_controller } = ChromeUtils.import( - "resource://torbutton/modules/tor-control-port.js" - ); - wait_for_controller().then( - controller => { - if (!this._isActive) { - // uninit() was called before resolution. - return; - } - // FIXME: We should be using some dedicated integrated back end to - // store circuit information, rather than collecting it all here in the - // front end. See tor-browser#41700. - controller.watchEvent( - "STREAM", - streamEvent => streamEvent.StreamStatus === "SENTCONNECT", - streamEvent => this._collectCircuit(controller, streamEvent) - ); - }, - error => { - this._log.error( - `Not collecting circuits because of an error: ${error.message}` - ); - } - ); - this.panel = document.getElementById("tor-circuit-panel"); this._panelElements = { heading: document.getElementById("tor-circuit-heading"), @@ -245,6 +191,9 @@ var gTorCircuitPanel = { // Notified of new locations for the currently selected browser (tab) *and* // switching selected browser. gBrowser.addProgressListener(this._locationListener); + + // Get notifications for circuit changes. + Services.obs.addObserver(this, this.TOR_CIRCUIT_TOPIC); }, /** @@ -253,6 +202,17 @@ var gTorCircuitPanel = { uninit() { this._isActive = false; gBrowser.removeProgressListener(this._locationListener); + Services.obs.removeObserver(this, this.TOR_CIRCUIT_TOPIC); + }, + + /** + * Observe circuit changes. + */ + observe(subject, topic, data) { + if (topic === this.TOR_CIRCUIT_TOPIC) { + // TODO: Maybe check if we actually need to do something earlier. + this._updateCurrentBrowser(); + } }, /** @@ -286,109 +246,6 @@ var gTorCircuitPanel = { window.openWebLinkIn(this._panelElements.aliasLink.href, where); }, - /** - * Collect circuit data for the found circuits, to be used later for display. - * - * @param {controller} controller - The tor controller. - * @param {object} streamEvent - The streamEvent for the new circuit. - */ - async _collectCircuit(controller, streamEvent) { - const id = streamEvent.CircuitID; - if (this._knownCircuitIDs.includes(id)) { - return; - } - this._log.debug(`New streamEvent.CircuitID: ${id}.`); - // FIXME: This list grows and is never freed. See tor-browser#41700. - this._knownCircuitIDs.push(id); - const circuitStatus = (await controller.getInfo("circuit-status"))?.find( - circuit => circuit.id === id - ); - if (!circuitStatus?.SOCKS_USERNAME || !circuitStatus?.SOCKS_PASSWORD) { - return; - } - const nodes = await Promise.all( - circuitStatus.circuit.map(names => - this._nodeDataForCircuit(controller, names) - ) - ); - // Remove quotes from the strings. - const username = circuitStatus.SOCKS_USERNAME.replace(/^"(.*)"$/, "$1"); - const password = circuitStatus.SOCKS_PASSWORD.replace(/^"(.*)"$/, "$1"); - const credentials = `${username}|${password}`; - // FIXME: This map grows and is never freed. We cannot simply request this - // information when needed because it is no longer available once the - // circuit is dropped, even if the web page is still displayed. - // See tor-browser#41700. - this._credentialsToCircuitNodes.set(credentials, nodes); - // Update the circuit in case the current page gains a new circuit whilst - // the popup is still open. - this._updateCurrentBrowser(credentials); - }, - - /** - * Fetch the node data for the given circuit node. - * - * @param {controller} controller - The tor controller. - * @param {string[]} circuitNodeNames - The names for the circuit node. Only - * the first name, the node id, will be used. - * - * @returns {NodeData} - The data for this circuit node. - */ - async _nodeDataForCircuit(controller, circuitNodeNames) { - // The first "name" in circuitNodeNames is the id. - // Remove the leading '$' if present. - const id = circuitNodeNames[0].replace(/^\$/, ""); - let result = { ipAddrs: [], bridgeType: null, regionCode: null }; - const bridge = (await controller.getConf("bridge"))?.find( - foundBridge => foundBridge.ID?.toUpperCase() === id.toUpperCase() - ); - const addrRe = /^\[?([^\]]+)\]?:\d+$/; - if (bridge) { - result.bridgeType = bridge.type ?? ""; - // Attempt to get an IP address from bridge address string. - const ip = bridge.address.match(addrRe)?.[1]; - if (ip && !ip.startsWith("0.")) { - result.ipAddrs.push(ip); - } - } else { - // Either dealing with a relay, or a bridge whose fingerprint is not saved - // in torrc. - let statusMap; - try { - statusMap = await controller.getInfo("ns/id/" + id); - } catch { - // getInfo will throw if the given id is not a relay. - // This probably means we are dealing with a user-provided bridge with - // no fingerprint. - // We don't know the ip/ipv6 or type, so leave blank. - result.bridgeType = ""; - return result; - } - if (statusMap.IP && !statusMap.IP.startsWith("0.")) { - result.ipAddrs.push(statusMap.IP); - } - const ip6 = statusMap.IPv6?.match(addrRe)?.[1]; - if (ip6) { - result.ipAddrs.push(ip6); - } - } - if (result.ipAddrs.length) { - // Get the country code for the node's IP address. - let regionCode; - try { - // Expect a 2-letter ISO3166-1 code, which should also be a valid BCP47 - // Region subtag. - regionCode = await controller.getInfo( - "ip-to-country/" + result.ipAddrs[0] - ); - } catch {} - if (regionCode && regionCode !== "??") { - result.regionCode = regionCode.toUpperCase(); - } - } - return result; - }, - /** * A list of schemes to never show the circuit display for. * @@ -398,71 +255,50 @@ var gTorCircuitPanel = { * * @type {string[]} */ - // FIXME: Have a back end that handles this instead. See tor-browser#41700. + // FIXME: Check if we find a UX to handle some of these cases, and if we + // manage to solve some technical issues. + // See tor-browser#41700 and tor-browser!699. _ignoredSchemes: ["about", "file", "chrome", "resource"], /** * Update the current circuit and domain data for the currently selected * browser, possibly changing the UI. - * - * @param {string?} [matchingCredentials=null] - If given, only update the - * current browser data if the current browser's credentials match. */ - _updateCurrentBrowser(matchingCredentials = null) { + _updateCurrentBrowser() { const browser = gBrowser.selectedBrowser; const domain = TorDomainIsolator.getDomainForBrowser(browser); + const nodes = TorDomainIsolator.getCircuit( + browser, + domain, + browser.contentPrincipal.originAttributes.userContextId + ); // We choose the currentURI, which matches what is shown in the URL bar and // will match up with the domain. // In contrast, documentURI corresponds to the shown page. E.g. it could // point to "about:certerror". const scheme = browser.currentURI?.scheme; - let credentials = TorDomainIsolator.getSocksProxyCredentials( - domain, - browser.contentPrincipal.originAttributes.userContextId - ); - if (credentials) { - credentials = `${credentials.username}|${credentials.password}`; - } - - if (matchingCredentials && matchingCredentials !== credentials) { - // This update was triggered by the circuit update for some other browser - // or process. - return; - } - - let nodes = this._credentialsToCircuitNodes.get(credentials) ?? []; - - const prevData = this._browserData.get(browser); - if ( - prevData && - prevData.domain && - prevData.domain === domain && - prevData.scheme === scheme && - prevData.nodes.length && - !nodes.length - ) { - // Since this is the same domain, for the same browser, and we used to - // have circuit nodes, we *assume* we are re-generating a circuit. So we - // keep the old circuit data around for the time being. - // FIXME: Have a back end that makes this explicit, rather than an - // assumption. See tor-browser#41700. - nodes = prevData.nodes; - this._log.debug(`Keeping old circuit for ${domain}.`); - } - - this._browserData.set(browser, { domain, scheme, nodes }); if ( this._currentBrowserData && this._currentBrowserData.domain === domain && this._currentBrowserData.scheme === scheme && - this._currentBrowserData.nodes === nodes + this._currentBrowserData.nodes.length === nodes.length && + // If non-null, the fingerprints of the nodes match. + (!nodes || + nodes.every( + (n, index) => + n.fingerprint === this._currentBrowserData.nodes[index].fingerprint + )) ) { // No change. + this._log.debug( + "Skipping browser update because the data is already up to date." + ); return; } - this._currentBrowserData = this._browserData.get(browser); + this._currentBrowserData = { domain, scheme, nodes }; + this._log.debug("Updating current browser.", this._currentBrowserData); if ( // Schemes where we always want to hide the display. ===================================== browser/components/torpreferences/content/connectionPane.js ===================================== @@ -17,6 +17,9 @@ const { TorSettings, TorSettingsTopics, TorSettingsData, TorBridgeSource } = const { TorProtocolService } = ChromeUtils.import( "resource://gre/modules/TorProtocolService.jsm" ); +const { TorMonitorService, TorMonitorTopics } = ChromeUtils.import( + "resource://gre/modules/TorMonitorService.jsm" +); const { TorConnect, TorConnectTopics, TorConnectState, TorCensorshipLevel } = ChromeUtils.import("resource:///modules/TorConnect.jsm"); @@ -144,8 +147,6 @@ const gConnectionPane = (function () { _internetStatus: InternetStatus.Unknown, - _controller: null, - _currentBridgeId: null, // populate xul with strings and cache the relevant elements @@ -727,9 +728,10 @@ const gConnectionPane = (function () { }; // Use a promise to avoid blocking the population of the page // FIXME: Stop using a JSON file, and switch to properties - fetch( + const annotationPromise = fetch( "chrome://browser/content/torpreferences/bridgemoji/annotations.json" - ).then(async res => { + ); + annotationPromise.then(async res => { const annotations = await res.json(); const bcp47 = Services.locale.appLocaleAsBCP47; const dash = bcp47.indexOf("-"); @@ -749,6 +751,7 @@ const gConnectionPane = (function () { ".currently-connected" )) { card.classList.remove("currently-connected"); + card.querySelector(selectors.bridges.cardQrGrid).style.height = ""; } if (!this._currentBridgeId) { return; @@ -769,72 +772,17 @@ const gConnectionPane = (function () { placeholder.replaceWith(...cards); this._checkBridgeCardsHeight(); }; - try { - const { controller } = ChromeUtils.import( - "resource://torbutton/modules/tor-control-port.js" - ); - // Avoid the cache because we set our custom event watcher, and at the - // moment, watchers cannot be removed from a controller. - controller(true).then(aController => { - this._controller = aController; - // Getting the circuits may be enough, if we have bootstrapped for a - // while, but at the beginning it gives many bridges as connected, - // because tor pokes all the bridges to find the best one. - // Also, watching circuit events does not work, at the moment, but in - // any case, checking the stream has the advantage that we can see if - // it really used for a connection, rather than tor having created - // this circuit to check if the bridge can be used. We do this by - // checking if the stream has SOCKS username, which actually contains - // the destination of the stream. - // FIXME: We only know the currentBridge *after* a circuit event, but - // if the circuit event is sent *before* about:torpreferences is - // opened we will miss it. Therefore this approach only works if a - // circuit is created after opening about:torconnect. A dedicated - // backend outside of about:preferences would help, and could be - // shared with gTorCircuitPanel. See tor-browser#41700. - this._controller.watchEvent( - "STREAM", - event => - event.StreamStatus === "SUCCEEDED" && "SOCKS_USERNAME" in event, - async event => { - const circuitStatuses = await this._controller.getInfo( - "circuit-status" - ); - if (!circuitStatuses) { - return; - } - for (const status of circuitStatuses) { - if (status.id === event.CircuitID && status.circuit.length) { - // The id in the circuit begins with a $ sign. - const id = status.circuit[0][0].replace(/^\$/, ""); - if (id !== this._currentBridgeId) { - const bridge = ( - await this._controller.getConf("bridge") - )?.find( - foundBridge => - foundBridge.ID?.toUpperCase() === id.toUpperCase() - ); - if (!bridge) { - // Either there is no bridge, or bridge with no - // fingerprint. - this._currentBridgeId = null; - } else { - this._currentBridgeId = id; - } - this._updateConnectedBridges(); - } - break; - } - } - } - ); - }); - } catch (err) { - console.warn( - "We could not load torbutton, bridge statuses will not be updated", - err - ); - } + this._checkConnectedBridge = () => { + // TODO: We could make sure TorSettings is in sync by monitoring also + // changes of settings. At that point, we could query it, instead of + // doing a query over the control port. + const bridge = TorMonitorService.currentBridge; + if (bridge?.fingerprint !== this._currentBridgeId) { + this._currentBridgeId = bridge?.fingerprint ?? null; + this._updateConnectedBridges(); + } + }; + annotationPromise.then(this._checkConnectedBridge.bind(this)); // Add a new bridge prefpane.querySelector(selectors.bridges.addHeader).textContent = @@ -927,6 +875,7 @@ const gConnectionPane = (function () { }); Services.obs.addObserver(this, TorConnectTopics.StateChange); + Services.obs.addObserver(this, TorMonitorTopics.BridgeChanged); }, init() { @@ -950,11 +899,7 @@ const gConnectionPane = (function () { // unregister our observer topics Services.obs.removeObserver(this, TorSettingsTopics.SettingChanged); Services.obs.removeObserver(this, TorConnectTopics.StateChange); - - if (this._controller !== null) { - this._controller.close(); - this._controller = null; - } + Services.obs.removeObserver(this, TorMonitorTopics.BridgeChanged); }, // whether the page should be present in about:preferences @@ -985,6 +930,12 @@ const gConnectionPane = (function () { this.onStateChange(); break; } + case TorMonitorTopics.BridgeChanged: { + if (data?.fingerprint !== this._currentBridgeId) { + this._checkConnectedBridge(); + } + break; + } } }, @@ -1028,7 +979,7 @@ const gConnectionPane = (function () { onRemoveAllBridges() { TorSettings.bridges.enabled = false; TorSettings.bridges.bridge_strings = ""; - if (TorSettings.bridges.source == TorBridgeSource.BuiltIn) { + if (TorSettings.bridges.source === TorBridgeSource.BuiltIn) { TorSettings.bridges.builtin_type = ""; } TorSettings.saveToPrefs(); ===================================== toolkit/components/tor-launcher/TorDomainIsolator.jsm → toolkit/components/tor-launcher/TorDomainIsolator.sys.mjs ===================================== @@ -1,13 +1,14 @@ -// A component for Tor Browser that puts requests from different -// first party domains on separate Tor circuits. - -var EXPORTED_SYMBOLS = ["TorDomainIsolator"]; +/** + * A component for Tor Browser that puts requests from different first party + * domains on separate Tor circuits. + */ -const { Services } = ChromeUtils.import("resource://gre/modules/Services.jsm"); -const { XPCOMUtils } = ChromeUtils.import( - "resource://gre/modules/XPCOMUtils.jsm" -); -const { ConsoleAPI } = ChromeUtils.import("resource://gre/modules/Console.jsm"); +import { XPCOMUtils } from "resource://gre/modules/XPCOMUtils.sys.mjs"; +import { ConsoleAPI } from "resource://gre/modules/Console.sys.mjs"; +import { + clearInterval, + setInterval, +} from "resource://gre/modules/Timer.sys.mjs"; const lazy = {}; @@ -18,11 +19,10 @@ XPCOMUtils.defineLazyServiceGetters(lazy, { ], }); -ChromeUtils.defineModuleGetter( - lazy, - "TorProtocolService", - "resource://gre/modules/TorProtocolService.jsm" -); +ChromeUtils.defineESModuleGetters(lazy, { + TorMonitorTopics: "resource://gre/modules/TorMonitorService.sys.mjs", + TorProtocolService: "resource://gre/modules/TorProtocolService.sys.mjs", +}); const logger = new ConsoleAPI({ prefix: "TorDomainIsolator", @@ -33,6 +33,12 @@ const logger = new ConsoleAPI({ // The string to use instead of the domain when it is not known. const CATCHALL_DOMAIN = "--unknown--"; +// The maximum lifetime for the catch-all circuit in milliseconds. +// When the catch-all circuit is needed, we check if more than this amount of +// time has passed since we last changed it nonce, and in case we change it +// again. +const CATCHALL_MAX_LIFETIME = 600_000; + // The preference to observe, to know whether isolation should be enabled or // disabled. const NON_TOR_PROXY_PREF = "extensions.torbutton.use_nontor_proxy"; @@ -40,23 +46,92 @@ const NON_TOR_PROXY_PREF = "extensions.torbutton.use_nontor_proxy"; // The topic of new identity, to observe to cleanup all the nonces. const NEW_IDENTITY_TOPIC = "new-identity-requested"; +// The topic on which we broacast circuit change notifications. +const TOR_CIRCUIT_TOPIC = "TorCircuitChange"; + +// We have an interval to delete circuits that are not reclaimed by any browser. +const CLEAR_TIMEOUT = 600_000; + +/** + * @typedef {string} CircuitId A string that we use to identify a circuit. + * Currently, it is a string that combines SOCKS credentials, to make it easier + * to use as a map key. + * It is not related to Tor's CircuitIDs. + */ +/** + * @typedef {number} BrowserId + */ +/** + * @typedef {NodeData[]} CircuitData The data about the nodes, ordered from + * guard (or bridge) to exit. + */ +/** + * @typedef BrowserCircuits Circuits related to a certain combination of + * isolators (first-party domain and user context ID, currently). + * @property {CircuitId} current The id of the last known circuit that has been + * used to fetch data for the isolated context. + * @property {CircuitId?} pending The id of the last used circuit for this + * isolation context. We might or might not know data about it, yet. But if we + * know it, we should move this id into current. + */ + class TorDomainIsolatorImpl { - // A mutable map that records what nonce we are using for each domain. + /** + * A mutable map that records what nonce we are using for each domain. + * + * @type {Map<string, string>} + */ #noncesForDomains = new Map(); - // A mutable map that records what nonce we are using for each tab container. + /** + * A mutable map that records what nonce we are using for each tab container. + * + * @type {Map<string, string>} + */ #noncesForUserContextId = new Map(); - // A bool that controls if we use SOCKS auth for isolation or not. + /** + * Tell whether we use SOCKS auth for isolation or not. + * + * @type {boolean} + */ #isolationEnabled = true; - // Specifies when the current catch-all circuit was first used + /** + * Specifies when the current catch-all circuit was first used. + * + * @type {integer} + */ #catchallDirtySince = Date.now(); + /** + * A map that associates circuit ids to the circuit information. + * + * @type {Map<CircuitId, CircuitData>} + */ + #knownCircuits = new Map(); + + /** + * A map that associates a certain browser to all the circuits it used or it + * is going to use. + * The circuits are keyed on the SOCKS username, which we take for granted + * being a combination of the first-party domain and the user context id. + * + * @type {Map<BrowserId, Map<string, BrowserCircuits>>} + */ + #browsers = new Map(); + + /** + * The handle of the interval we use to cleanup old circuit data. + * + * @type {number?} + */ + #cleanupIntervalId = null; + /** * Initialize the domain isolator. - * This function will setup the proxy filter that injects the credentials and - * register some observers. + * This function will setup the proxy filter that injects the credentials, + * register some observers, and setup the cleaning interval. */ init() { logger.info("Setup circuit isolation by domain and user context"); @@ -68,14 +143,25 @@ class TorDomainIsolatorImpl { Services.prefs.addObserver(NON_TOR_PROXY_PREF, this); Services.obs.addObserver(this, NEW_IDENTITY_TOPIC); + Services.obs.addObserver(this, lazy.TorMonitorTopics.StreamSucceeded); + + this.#cleanupIntervalId = setInterval( + this.#clearKnownCircuits.bind(this), + CLEAR_TIMEOUT + ); } /** - * Removes the observers added in the initialization. + * Removes the observers added in the initialization and stops the cleaning + * interval. */ uninit() { Services.prefs.removeObserver(NON_TOR_PROXY_PREF, this); Services.obs.removeObserver(this, NEW_IDENTITY_TOPIC); + Services.obs.removeObserver(this, lazy.TorMonitorTopics.StreamSucceeded); + clearInterval(this.#cleanupIntervalId); + this.#cleanupIntervalId = null; + this.clearIsolation(); } enable() { @@ -89,52 +175,52 @@ class TorDomainIsolatorImpl { } /** - * Return the credentials to use as username and password for the SOCKS proxy, - * given a certain domain and userContextId. Optionally, create them. + * Get the last circuit used in a certain browser. + * The returned data is created when the circuit is first seen, therefore it + * could be stale (i.e., the circuit might not be available anymore). * - * @param {string} firstPartyDomain The first party domain associated to the requests - * @param {string} userContextId The context ID associated to the request - * @param {bool} create Whether to create the nonce, if it is not available - * @returns {object|null} Either the credential, or null if we do not have them and create is - * false. + * @param {MozBrowser} browser The browser to get data for + * @param {string} domain The first party domain we want to get the circuit + * for + * @param {number} userContextId The user context domain we want to get the + * circuit for + * @returns {NodeData[]} The node data, or an empty array if we do not have + * data for the requested key. */ - getSocksProxyCredentials(firstPartyDomain, userContextId, create = false) { - if (!this.#noncesForDomains.has(firstPartyDomain)) { - if (!create) { - return null; - } - const nonce = this.#nonce(); - logger.info(`New nonce for first party ${firstPartyDomain}: ${nonce}`); - this.#noncesForDomains.set(firstPartyDomain, nonce); + getCircuit(browser, domain, userContextId) { + const username = this.#makeUsername(domain, userContextId); + const circuits = this.#browsers.get(browser.browserId)?.get(username); + // This is the only place where circuit data can go out, so the only place + // where it makes a difference to check whether the pending circuit is still + // pending, or it has actually got data. + const pending = this.#knownCircuits.get(circuits?.pending); + if (pending?.length) { + circuits.current = circuits.pending; + circuits.pending = null; + return pending; } - if (!this.#noncesForUserContextId.has(userContextId)) { - if (!create) { - return null; - } - const nonce = this.#nonce(); - logger.info(`New nonce for userContextId ${userContextId}: ${nonce}`); - this.#noncesForUserContextId.set(userContextId, nonce); - } - return { - username: this.#makeUsername(firstPartyDomain, userContextId), - password: - this.#noncesForDomains.get(firstPartyDomain) + - this.#noncesForUserContextId.get(userContextId), - }; + // TODO: At this point we already know if we expect a circuit change for + // this key: (circuit?.pending && !pending). However, we do not consume this + // data yet in the frontend, so do not send it for now. + return this.#knownCircuits.get(circuits?.current) ?? []; } /** * Create a new nonce for the FP domain of the selected browser and reload the * tab with a new circuit. * - * @param {object} browser Should be the gBrowser from the context of the - * caller + * @param {object} globalBrowser Should be the gBrowser from the context of + * the caller */ - newCircuitForBrowser(browser) { - const firstPartyDomain = getDomainForBrowser(browser.selectedBrowser); + newCircuitForBrowser(globalBrowser) { + const browser = globalBrowser.selectedBrowser; + const firstPartyDomain = getDomainForBrowser(browser); this.#newCircuitForDomain(firstPartyDomain); - // TODO: How to properly handle the user context? Should we use - // (domain, userContextId) pairs, instead of concatenating nonces? + const { username, password } = this.#getSocksProxyCredentials( + firstPartyDomain, + browser.contentPrincipal.originAttributes.userContextId + ); + this.#trackBrowser(browser, username, password); browser.reloadWithFlags(Ci.nsIWebNavigation.LOAD_FLAGS_BYPASS_CACHE); } @@ -147,12 +233,15 @@ class TorDomainIsolatorImpl { // Per-domain and per contextId nonces are stored in maps, so simply clear // them. + // Notice that the catch-all circuit is included in #noncesForDomains, so we + // are implicilty cleaning it. Should this change, we should change its + // nonce explicitly here. this.#noncesForDomains.clear(); this.#noncesForUserContextId.clear(); + this.#catchallDirtySince = Date.now(); - // Force a rotation on the next catch-all circuit use by setting the - // creation time to the epoch. - this.#catchallDirtySince = 0; + this.#knownCircuits.clear(); + this.#browsers.clear(); } async observe(subject, topic, data) { @@ -173,55 +262,20 @@ class TorDomainIsolatorImpl { logger.error("Could not send the newnym command", e); // TODO: What UX to use here? See tor-browser#41708 } + } else if (topic === lazy.TorMonitorTopics.StreamSucceeded) { + const { username, password, circuit } = subject.wrappedJSObject; + this.#updateCircuit(username, password, circuit); } } /** - * Setup a filter that for every HTTPChannel, replaces the default SOCKS proxy - * with one that authenticates to the SOCKS server (the tor client process) - * with a username (the first party domain and userContextId) and a nonce - * password. - * Tor provides a separate circuit for each username+password combination. + * Setup a filter that for every HTTPChannel. */ #setupProxyFilter() { - const filterFunction = (aChannel, aProxy) => { - if (!this.#isolationEnabled) { - return aProxy; - } - try { - const channel = aChannel.QueryInterface(Ci.nsIChannel); - let firstPartyDomain = - channel.loadInfo.originAttributes.firstPartyDomain; - const userContextId = channel.loadInfo.originAttributes.userContextId; - if (firstPartyDomain === "") { - firstPartyDomain = CATCHALL_DOMAIN; - if (Date.now() - this.#catchallDirtySince > 1000 * 10 * 60) { - logger.info( - "tor catchall circuit has been dirty for over 10 minutes. Rotating." - ); - this.#newCircuitForDomain(CATCHALL_DOMAIN); - this.#catchallDirtySince = Date.now(); - } - } - const replacementProxy = this.#applySocksProxyCredentials( - aProxy, - firstPartyDomain, - userContextId - ); - logger.debug( - `Requested ${channel.URI.spec} via ${replacementProxy.username}:${replacementProxy.password}` - ); - return replacementProxy; - } catch (e) { - logger.error("Error while setting a new proxy", e); - return null; - } - }; - lazy.ProtocolProxyService.registerChannelFilter( { - applyFilter(aChannel, aProxy, aCallback) { - aCallback.onProxyFilterResult(filterFunction(aChannel, aProxy)); + applyFilter: (aChannel, aProxy, aCallback) => { + aCallback.onProxyFilterResult(this.#proxyFilter(aChannel, aProxy)); }, }, 0 @@ -229,33 +283,96 @@ class TorDomainIsolatorImpl { } /** - * Takes a proxyInfo object (originalProxy) and returns a new proxyInfo - * object with the same properties, except the username is set to the - * the domain and userContextId, and the password is a nonce. + * Replaces the default SOCKS proxy with one that authenticates to the SOCKS + * server (the tor client process) with a username (the first party domain and + * userContextId) and a nonce password. + * Tor provides a separate circuit for each username+password combination. + * + * @param {nsIChannel} aChannel The channel we are setting the proxy for + * @param {nsIProxyInfo} aProxy The original proxy + * @returns {nsIProxyInfo} The new proxy to use */ - #applySocksProxyCredentials(originalProxy, domain, userContextId) { - const proxy = originalProxy.QueryInterface(Ci.nsIProxyInfo); - const { username, password } = this.getSocksProxyCredentials( - domain, - userContextId, - true - ); - return lazy.ProtocolProxyService.newProxyInfoWithAuth( - "socks", - proxy.host, - proxy.port, - username, - password, - "", // aProxyAuthorizationHeader - "", // aConnectionIsolationKey - proxy.flags, - proxy.failoverTimeout, - proxy.failoverProxy - ); + #proxyFilter(aChannel, aProxy) { + if (!this.#isolationEnabled) { + return aProxy; + } + try { + const channel = aChannel.QueryInterface(Ci.nsIChannel); + let firstPartyDomain = channel.loadInfo.originAttributes.firstPartyDomain; + const userContextId = channel.loadInfo.originAttributes.userContextId; + if (!firstPartyDomain) { + firstPartyDomain = CATCHALL_DOMAIN; + if (Date.now() - this.#catchallDirtySince > CATCHALL_MAX_LIFETIME) { + logger.info( + "tor catchall circuit has reached its maximum lifetime. Rotating." + ); + this.#newCircuitForDomain(CATCHALL_DOMAIN); + } + } + const { username, password } = this.#getSocksProxyCredentials( + firstPartyDomain, + userContextId + ); + const browser = this.#getBrowserForChannel(channel); + if (browser) { + this.#trackBrowser(browser, username, password); + } + logger.debug(`Requested ${channel.URI.spec} via ${username}:${password}`); + const proxy = aProxy.QueryInterface(Ci.nsIProxyInfo); + return lazy.ProtocolProxyService.newProxyInfoWithAuth( + "socks", + proxy.host, + proxy.port, + username, + password, + "", // aProxyAuthorizationHeader + "", // aConnectionIsolationKey + proxy.flags, + proxy.failoverTimeout, + proxy.failoverProxy + ); + } catch (e) { + logger.error("Error while setting a new proxy", e); + return null; + } + } + + /** + * Return the credentials to use as username and password for the SOCKS proxy, + * given a certain domain and userContextId. + * A new random password will be created if not available yet. + * + * @param {string} firstPartyDomain The first party domain associated to the + * requests + * @param {number} userContextId The context ID associated to the request + * @returns {object} The credentials + */ + #getSocksProxyCredentials(firstPartyDomain, userContextId) { + if (!this.#noncesForDomains.has(firstPartyDomain)) { + const nonce = this.#nonce(); + logger.info(`New nonce for first party ${firstPartyDomain}: ${nonce}`); + this.#noncesForDomains.set(firstPartyDomain, nonce); + } + if (!this.#noncesForUserContextId.has(userContextId)) { + const nonce = this.#nonce(); + logger.info(`New nonce for userContextId ${userContextId}: ${nonce}`); + this.#noncesForUserContextId.set(userContextId, nonce); + } + // TODO: How to properly handle the user-context? Should we use + // (domain, userContextId) pairs, instead of concatenating nonces? + return { + username: this.#makeUsername(firstPartyDomain, userContextId), + password: + this.#noncesForDomains.get(firstPartyDomain) + + this.#noncesForUserContextId.get(userContextId), + }; } /** * Combine the needed data into a username for the proxy. + * + * @param {string} domain The first-party domain associated to the request + * @param {integer} userContextId The userContextId associated to the request */ #makeUsername(domain, userContextId) { if (!domain) { @@ -264,12 +381,26 @@ class TorDomainIsolatorImpl { return `${domain}:${userContextId}`; } + /** + * Combine SOCKS username and password into a string to use as ID. + * + * @param {string} username The SOCKS username + * @param {string} password The SOCKS password + * @returns {CircuitId} A string that combines username and password and can + * be used for map lookups. + */ + #credentialsToId(username, password) { + return `${username}|${password}`; + } + /** * Generate a new 128 bit random tag. * * Strictly speaking both using a cryptographic entropy source and using 128 * bits of entropy for the tag are likely overkill, as correct behavior only * depends on how unlikely it is for there to be a collision. + * + * @returns {string} The random nonce */ #nonce() { return Array.from(crypto.getRandomValues(new Uint8Array(16)), byte => @@ -279,12 +410,18 @@ class TorDomainIsolatorImpl { /** * Re-generate the nonce for a certain domain. + * + * @param {string?} domain The first-party domain to re-create the nonce for. + * If empty or null, the catchall domain will be used. */ #newCircuitForDomain(domain) { if (!domain) { domain = CATCHALL_DOMAIN; } this.#noncesForDomains.set(domain, this.#nonce()); + if (domain === CATCHALL_DOMAIN) { + this.#catchallDirtySince = Date.now(); + } logger.info( `New domain isolation for ${domain}: ${this.#noncesForDomains.get( domain @@ -296,6 +433,8 @@ class TorDomainIsolatorImpl { * Re-generate the nonce for a userContextId. * * Currently, this function is not hooked to anything. + * + * @param {integer} userContextId The userContextId to re-create the nonce for */ #newCircuitForUserContextId(userContextId) { this.#noncesForUserContextId.set(userContextId, this.#nonce()); @@ -305,13 +444,182 @@ class TorDomainIsolatorImpl { )}` ); } + + /** + * Try to extract a browser from a channel. + * + * @param {nsIChannel} channel The channel to extract the browser from + * @returns {MozBrowser?} The browser the channel is associated to + */ + #getBrowserForChannel(channel) { + const browsers = + channel.loadInfo.browsingContext?.topChromeWindow?.gBrowser.browsers; + if (!browsers || !channel.loadInfo.browsingContext?.browserId) { + return null; + } + for (const browser of browsers) { + if (browser.browserId === channel.loadInfo.browsingContext.browserId) { + logger.debug( + "Matched browser with browserId", + channel.loadInfo.browsingContext.browserId + ); + return browser; + } + } + // Expected to arrive here for example for the update checker. + // If we find a way to check that, we could raise the level to a warn. + logger.debug("Browser not matched", channel); + return null; + } + + /** + * Associate the SOCKS credentials to a browser. + * If needed (the browser is associated for the first time, or it was already + * known but its credential changed), notify the related circuit display. + * + * @param {MozBrowser} browser The browser to track + * @param {string} username The SOCKS username + * @param {string} password The SOCKS password + */ + #trackBrowser(browser, username, password) { + let browserCircuits = this.#browsers.get(browser.browserId); + if (!browserCircuits) { + browserCircuits = new Map(); + this.#browsers.set(browser.browserId, browserCircuits); + } + const circuitIds = browserCircuits.get(username) ?? {}; + const id = this.#credentialsToId(username, password); + if (circuitIds.current === id) { + // The circuit with these credentials was already built (we already knew + // its nodes, or we would not have promoted it to the current circuit). + // We do not need to do anything else, because we cannot detect a change + // of nodes here. + return; + } + + logger.debug( + `Found new credentials ${username} ${password} for browser`, + browser + ); + const circuit = this.#knownCircuits.get(id); + if (circuit?.length) { + circuitIds.current = id; + if (circuitIds.pending === id) { + circuitIds.pending = null; + } + browserCircuits.set(username, circuitIds); + // FIXME: We only notify the circuit display when we have a change that + // involves circuits whose nodes are known, for now. We need to resolve a + // few other techical problems (e.g., associate the circuit to the + // document?) and develop a UX with some animation to notify the circuit + // display more often. + // See tor-browser#41700 and tor-browser!699. + // In any case, notify the circuit display only after the internal map has + // been updated. + this.#notifyCircuitDisplay(); + } else if (circuitIds.pending !== id) { + // We do not have node data, so we store that we might need to track this. + // Otherwise, when a circuit is ready, we do not know which browser was it + // used for. + circuitIds.pending = id; + browserCircuits.set(username, circuitIds); + } + } + + /** + * Update a circuit, and notify the related circuit displays if it changed. + * + * This function is called when a certain stream has succeeded and so we can + * associate its SOCKS credential to the circuit it is using. + * We receive only the fingerprints of the circuit nodes, but they are enough + * to check if the circuit has changed. If it has, we also get the nodes' + * information through the control port. + * + * @param {string} username The SOCKS username + * @param {string} password The SOCKS password + * @param {NodeFingerprint[]} circuit The fingerprints of the nodes that + * compose the circuit + */ + async #updateCircuit(username, password, circuit) { + const id = this.#credentialsToId(username, password); + let data = this.#knownCircuits.get(id) ?? []; + // Should we modify the lower layer to send a circuit identifier, instead? + if ( + circuit.length === data.length && + circuit.every((id, index) => id === data[index].fingerprint) + ) { + return; + } + + data = await Promise.all( + circuit.map(fingerprint => + lazy.TorProtocolService.getNodeInfo(fingerprint) + ) + ); + this.#knownCircuits.set(id, data); + // We know that something changed, but we cannot know if anyone is + // interested in this change. So, we have to notify all the possible + // consumers of the data in any case. + // Not being specific and let them check if they need to do something allows + // us to keep a simpler structure. + this.#notifyCircuitDisplay(); + } + + /** + * Broadcast a notification when a circuit changed, or a browser is changing + * circuit (which might happen also in case of navigation). + */ + #notifyCircuitDisplay() { + Services.obs.notifyObservers(null, TOR_CIRCUIT_TOPIC); + } + + /** + * Clear the known circuit information, when they are not needed anymore. + * + * We keep circuit data around for a while. We decouple it from the underlying + * tor circuit management in case the user clicks on the circuit display when + * circuit has long gone. + * However, data accumulate during a session. So, since we store all the + * browsers that used a circuit anyway, every now and then we check if we + * still know browsers using a certain circuits. If there are not, we forget + * about it. + * + * This function is run by an interval. + */ + #clearKnownCircuits() { + logger.info("Running the circuit cleanup"); + const windows = []; + const enumerator = Services.wm.getEnumerator("navigator:browser"); + while (enumerator.hasMoreElements()) { + windows.push(enumerator.getNext()); + } + const browsers = windows + .flatMap(win => win.gBrowser.browsers.map(b => b.browserId)) + .filter(id => this.#browsers.has(id)); + this.#browsers = new Map(browsers.map(id => [id, this.#browsers.get(id)])); + this.#knownCircuits = new Map( + Array.from(this.#browsers.values(), circuits => + Array.from(circuits.values(), ids => { + const r = []; + const current = this.#knownCircuits.get(ids.current); + if (current) { + r.push([ids.current, current]); + } + const pending = this.#knownCircuits.get(ids.pending); + if (pending) { + r.push([ids.pending, pending]); + } + return r; + }) + ).flat(2) + ); + } } /** * Get the first party domain for a certain browser. * - * @param browser The browser to get the FP-domain for. - * + * @param {MozBrowser} browser The browser to get the FP-domain for. * Please notice that it should be gBrowser.selectedBrowser, because * browser.documentURI is the actual shown page, and might be an error page. * In this case, we rely on currentURI, which for gBrowser is an alias of @@ -358,6 +666,6 @@ function getDomainForBrowser(browser) { return fpd; } -const TorDomainIsolator = new TorDomainIsolatorImpl(); +export const TorDomainIsolator = new TorDomainIsolatorImpl(); // Reduce global vars pollution TorDomainIsolator.getDomainForBrowser = getDomainForBrowser; ===================================== toolkit/components/tor-launcher/TorMonitorService.sys.mjs ===================================== @@ -19,6 +19,10 @@ ChromeUtils.defineModuleGetter( "resource://torbutton/modules/tor-control-port.js" ); +ChromeUtils.defineESModuleGetters(lazy, { + TorProtocolService: "resource://gre/modules/TorProtocolService.sys.mjs", +}); + const logger = new ConsoleAPI({ maxLogLevel: "warn", maxLogLevelPref: "browser.tor_monitor_service.log_level", @@ -37,12 +41,34 @@ const TorTopics = Object.freeze({ ProcessRestarted: "TorProcessRestarted", }); +export const TorMonitorTopics = Object.freeze({ + BridgeChanged: "TorBridgeChanged", + StreamSucceeded: "TorStreamSucceeded", +}); + const ControlConnTimings = Object.freeze({ initialDelayMS: 25, // Wait 25ms after the process has started, before trying to connect maxRetryMS: 10000, // Retry at most every 10 seconds timeoutMS: 5 * 60 * 1000, // Wait at most 5 minutes for tor to start }); +/** + * From control-spec.txt: + * CircuitID = 1*16 IDChar + * IDChar = ALPHA / DIGIT + * Currently, Tor only uses digits, but this may change. + * + * @typedef {string} CircuitID + */ +/** + * The fingerprint of a node. + * From control-spec.txt: + * Fingerprint = "$" 40*HEXDIG + * However, we do not keep the $ in our structures. + * + * @typedef {string} NodeFingerprint + */ + /** * This service monitors an existing Tor instance, or starts one, if needed, and * then starts monitoring it. @@ -52,7 +78,7 @@ const ControlConnTimings = Object.freeze({ */ export const TorMonitorService = { _connection: null, - _eventsToMonitor: Object.freeze(["STATUS_CLIENT", "NOTICE", "WARN", "ERR"]), + _eventHandlers: {}, _torLog: [], // Array of objects with date, type, and msg properties. _startTimeout: null, @@ -64,6 +90,28 @@ export const TorMonitorService = { _inited: false, + /** + * Stores the nodes of a circuit. Keys are cicuit IDs, and values are the node + * fingerprints. + * + * Theoretically, we could hook this map up to the new identity notification, + * but in practice it does not work. Tor pre-builds circuits, and the NEWNYM + * signal does not affect them. So, we might end up using a circuit that was + * built before the new identity but not yet used. If we cleaned the map, we + * risked of not having the data about it. + * + * @type {Map<CircuitID, NodeFingerprint[]>} + */ + _circuits: new Map(), + /** + * The last used bridge, or null if bridges are not in use or if it was not + * possible to detect the bridge. This needs the user to have specified bridge + * lines with fingerprints to work. + * + * @type {NodeFingerprint?} + */ + _currentBridge: null, + // Public methods // Starts Tor, if needed, and starts monitoring for events @@ -72,14 +120,28 @@ export const TorMonitorService = { return; } this._inited = true; + + // We always liten to these events, because they are needed for the circuit + // display. + this._eventHandlers = new Map([ + ["CIRC", this._processCircEvent.bind(this)], + ["STREAM", this._processStreamEvent.bind(this)], + ]); + if (this.ownsTorDaemon) { + // When we own the tor daemon, we listen to more events, that are used + // for about:torconnect or for showing the logs in the settings page. + this._eventHandlers.set("STATUS_CLIENT", (_eventType, lines) => + this._processBootstrapStatus(lines[0], false) + ); + this._eventHandlers.set("NOTICE", this._processLog.bind(this)); + this._eventHandlers.set("WARN", this._processLog.bind(this)); + this._eventHandlers.set("ERR", this._processLog.bind(this)); this._controlTor(); } else { - logger.info( - "Not starting the event monitor, as we do not own the Tor daemon." - ); + this._startEventMonitor(); } - logger.debug("TorMonitorService initialized"); + logger.info("TorMonitorService initialized"); }, // Closes the connection that monitors for events. @@ -153,6 +215,18 @@ export const TorMonitorService = { return !!this._connection; }, + /** + * Return the data about the current bridge, if any, or null. + * We can detect bridge only when the configured bridge lines include the + * fingerprints. + * + * @returns {NodeData?} The node information, or null if the first node + * is not a bridge, or no circuit has been opened, yet. + */ + get currentBridge() { + return this._currentBridge; + }, + // Private methods async _startProcess() { @@ -272,7 +346,7 @@ export const TorMonitorService = { // TODO: optionally monitor INFO and DEBUG log messages. let reply = await conn.sendCommand( - "SETEVENTS " + this._eventsToMonitor.join(" ") + "SETEVENTS " + Array.from(this._eventHandlers.keys()).join(" ") ); reply = TorParsers.parseCommandResponse(reply); if (!TorParsers.commandSucceeded(reply)) { @@ -281,14 +355,10 @@ export const TorMonitorService = { return false; } - // FIXME: At the moment it is not possible to start the event monitor - // when we do start the tor process. So, does it make sense to keep this - // control? if (this._torProcess) { this._torProcess.connectionWorked(); } - - if (!TorLauncherUtil.shouldOnlyConfigureTor) { + if (this.ownsTorDaemon && !TorLauncherUtil.shouldOnlyConfigureTor) { try { await this._takeTorOwnership(conn); } catch (e) { @@ -297,7 +367,31 @@ export const TorMonitorService = { } this._connection = conn; - this._waitForEventData(); + + for (const [type, callback] of this._eventHandlers.entries()) { + this._monitorEvent(type, callback); + } + + // Populate the circuit map already, in case we are connecting to an + // external tor daemon. + try { + const reply = await this._connection.sendCommand( + "GETINFO circuit-status" + ); + const lines = reply.split(/\r?\n/); + if (lines.shift() === "250+circuit-status=") { + for (const line of lines) { + if (line === ".") { + break; + } + // _processCircEvent processes only one line at a time + this._processCircEvent("CIRC", [line]); + } + } + } catch (e) { + logger.warn("Could not populate the initial circuit map", e); + } + return true; }, @@ -318,65 +412,49 @@ export const TorMonitorService = { } }, - _waitForEventData() { - if (!this._connection) { - return; - } - logger.debug("Start watching events:", this._eventsToMonitor); + _monitorEvent(type, callback) { + logger.info(`Watching events of type ${type}.`); let replyObj = {}; - for (const torEvent of this._eventsToMonitor) { - this._connection.watchEvent( - torEvent, - null, - line => { - if (!line) { - return; - } - logger.debug("Event response: ", line); - const isComplete = TorParsers.parseReplyLine(line, replyObj); - if (isComplete) { - this._processEventReply(replyObj); - replyObj = {}; - } - }, - true - ); - } + this._connection.watchEvent( + type, + null, + line => { + if (!line) { + return; + } + logger.debug("Event response: ", line); + const isComplete = TorParsers.parseReplyLine(line, replyObj); + if (!isComplete || replyObj._parseError || !replyObj.lineArray.length) { + return; + } + const reply = replyObj; + replyObj = {}; + if (reply.statusCode !== TorStatuses.EventNotification) { + logger.error("Unexpected event status code:", reply.statusCode); + return; + } + if (!reply.lineArray[0].startsWith(`${type} `)) { + logger.error("Wrong format for the first line:", reply.lineArray[0]); + return; + } + reply.lineArray[0] = reply.lineArray[0].substring(type.length + 1); + try { + callback(type, reply.lineArray); + } catch (e) { + logger.error("Exception while handling an event", reply, e); + } + }, + true + ); }, - _processEventReply(aReply) { - if (aReply._parseError || !aReply.lineArray.length) { - return; - } - - if (aReply.statusCode !== TorStatuses.EventNotification) { - logger.warn("Unexpected event status code:", aReply.statusCode); - return; - } - - // TODO: do we need to handle multiple lines? - const s = aReply.lineArray[0]; - const idx = s.indexOf(" "); - if (idx === -1) { - return; - } - const eventType = s.substring(0, idx); - const msg = s.substring(idx + 1).trim(); - - if (eventType === "STATUS_CLIENT") { - this._processBootstrapStatus(msg, false); - return; - } else if (!this._eventsToMonitor.includes(eventType)) { - logger.debug(`Dropping unlistened event ${eventType}`); - return; - } - - if (eventType === "WARN" || eventType === "ERR") { + _processLog(type, lines) { + if (type === "WARN" || type === "ERR") { // Notify so that Copy Log can be enabled. Services.obs.notifyObservers(null, TorTopics.HasWarnOrErr); } - const now = new Date(); + const date = new Date(); const maxEntries = Services.prefs.getIntPref( "extensions.torlauncher.max_tor_log_entries", 1000 @@ -384,8 +462,10 @@ export const TorMonitorService = { if (maxEntries > 0 && this._torLog.length >= maxEntries) { this._torLog.splice(0, 1); } - this._torLog.push({ date: now, type: eventType, msg }); - const logString = `Tor ${eventType}: ${msg}`; + + const msg = lines.join("\n"); + this._torLog.push({ date, type, msg }); + const logString = `Tor ${type}: ${msg}`; logger.info(logString); }, @@ -461,8 +541,108 @@ export const TorMonitorService = { } }, + async _processCircEvent(_type, lines) { + const builtEvent = + /^(?<CircuitID>[a-zA-Z0-9]{1,16})\sBUILT\s(?<Path>(?:,?\$[0-9a-fA-F]{40}(?:~[a-zA-Z0-9]{1,19})?)+)/.exec( + lines[0] + ); + const closedEvent = /^(?<ID>[a-zA-Z0-9]{1,16})\sCLOSED/.exec(lines[0]); + if (builtEvent) { + const fp = /\$([0-9a-fA-F]{40})/g; + const nodes = Array.from(builtEvent.groups.Path.matchAll(fp), g => + g[1].toUpperCase() + ); + this._circuits.set(builtEvent.groups.CircuitID, nodes); + // Ignore circuits of length 1, that are used, for example, to probe + // bridges. So, only store them, since we might see streams that use them, + // but then early-return. + if (nodes.length === 1) { + return; + } + // In some cases, we might already receive SOCKS credentials in the line. + // However, this might be a problem with onion services: we get also a + // 4-hop circuit that we likely do not want to show to the user, + // especially because it is used only temporarily, and it would need a + // technical explaination. + // this._checkCredentials(lines[0], nodes); + if (this._currentBridge?.fingerprint !== nodes[0]) { + const nodeInfo = await lazy.TorProtocolService.getNodeInfo(nodes[0]); + let notify = false; + if (nodeInfo?.bridgeType) { + logger.info(`Bridge changed to ${nodes[0]}`); + this._currentBridge = nodeInfo; + notify = true; + } else if (this._currentBridge) { + logger.info("Bridges disabled"); + this._currentBridge = null; + notify = true; + } + if (notify) { + Services.obs.notifyObservers( + null, + TorMonitorTopics.BridgeChanged, + this._currentBridge + ); + } + } + } else if (closedEvent) { + this._circuits.delete(closedEvent.groups.ID); + } + }, + + _processStreamEvent(_type, lines) { + // The first block is the stream ID, which we do not need at the moment. + const succeeedEvent = + /^[a-zA-Z0-9]{1,16}\sSUCCEEDED\s(?<CircuitID>[a-zA-Z0-9]{1,16})/.exec( + lines[0] + ); + if (!succeeedEvent) { + return; + } + const circuit = this._circuits.get(succeeedEvent.groups.CircuitID); + if (!circuit) { + logger.error( + "Seen a STREAM SUCCEEDED with an unknown circuit. Not notifying observers.", + lines[0] + ); + return; + } + this._checkCredentials(lines[0], circuit); + }, + + /** + * Check if a STREAM or CIRC response line contains SOCKS_USERNAME and + * SOCKS_PASSWORD. In case, notify observers that we could associate a certain + * circuit to these credentials. + * + * @param {string} line The circ or stream line to check + * @param {NodeFingerprint[]} circuit The fingerprints of the nodes in the + * circuit. + */ + _checkCredentials(line, circuit) { + const username = /SOCKS_USERNAME=("(?:[^"\\]|\\.)*")/.exec(line); + const password = /SOCKS_PASSWORD=("(?:[^"\\]|\\.)*")/.exec(line); + if (!username || !password) { + return; + } + Services.obs.notifyObservers( + { + wrappedJSObject: { + username: TorParsers.unescapeString(username[1]), + password: TorParsers.unescapeString(password[1]), + circuit, + }, + }, + TorMonitorTopics.StreamSucceeded + ); + }, + _shutDownEventMonitor() { - this._connection?.close(); + try { + this._connection?.close(); + } catch (e) { + logger.error("Could not close the connection to the control port", e); + } this._connection = null; if (this._startTimeout !== null) { clearTimeout(this._startTimeout); ===================================== toolkit/components/tor-launcher/TorParsers.sys.mjs ===================================== @@ -181,12 +181,12 @@ export const TorParsers = Object.freeze({ return aStr; } const escaped = aStr - .replace("\\", "\\\\") - .replace('"', '\\"') - .replace("\n", "\\n") - .replace("\r", "\\r") - .replace("\t", "\\t") - .replace(/[^\x20-\x7e]+/g, text => { + .replaceAll("\\", "\\\\") + .replaceAll('"', '\\"') + .replaceAll("\n", "\\n") + .replaceAll("\r", "\\r") + .replaceAll("\t", "\\t") + .replaceAll(/[^\x20-\x7e]+/g, text => { const encoder = new TextEncoder(); return Array.from( encoder.encode(text), ===================================== toolkit/components/tor-launcher/TorProtocolService.sys.mjs ===================================== @@ -40,6 +40,20 @@ const logger = new ConsoleAPI({ prefix: "TorProtocolService", }); +/** + * Stores the data associated with a circuit node. + * + * @typedef NodeData + * @property {string} fingerprint The node fingerprint. + * @property {string[]} ipAddrs - The ip addresses associated with this node. + * @property {string?} bridgeType - The bridge type for this node, or "" if the + * node is a bridge but the type is unknown, or null if this is not a bridge + * node. + * @property {string?} regionCode - An upper case 2-letter ISO3166-1 code for + * the first ip address, or null if there is no region. This should also be a + * valid BCP47 Region subtag. + */ + // Manage the connection to tor's control port, to update its settings and query // other useful information. // @@ -188,6 +202,89 @@ export const TorProtocolService = { return TorParsers.parseReply(cmd, keyword, response); }, + async getBridges() { + // Ideally, we would not need this function, because we should be the one + // setting them with TorSettings. However, TorSettings is not notified of + // change of settings. So, asking tor directly with the control connection + // is the most reliable way of getting the configured bridges, at the + // moment. Also, we are using this for the circuit display, which should + // work also when we are not configuring the tor daemon, but just using it. + return this._withConnection(conn => { + return conn.getConf("bridge"); + }); + }, + + /** + * Returns tha data about a relay or a bridge. + * + * @param {string} id The fingerprint of the node to get data about + * @returns {NodeData} + */ + async getNodeInfo(id) { + return this._withConnection(async conn => { + const node = { + fingerprint: id, + ipAddrs: [], + bridgeType: null, + regionCode: null, + }; + const bridge = (await conn.getConf("bridge"))?.find( + foundBridge => foundBridge.ID?.toUpperCase() === id.toUpperCase() + ); + const addrRe = /^\[?([^\]]+)\]?:\d+$/; + if (bridge) { + node.bridgeType = bridge.type ?? ""; + // Attempt to get an IP address from bridge address string. + const ip = bridge.address.match(addrRe)?.[1]; + if (ip && !ip.startsWith("0.")) { + node.ipAddrs.push(ip); + } + } else { + // Either dealing with a relay, or a bridge whose fingerprint is not + // saved in torrc. + const info = await conn.getInfo(`ns/id/${id}`); + if (info.IP && !info.IP.startsWith("0.")) { + node.ipAddrs.push(info.IP); + } + const ip6 = info.IPv6?.match(addrRe)?.[1]; + if (ip6) { + node.ipAddrs.push(ip6); + } + } + if (node.ipAddrs.length) { + // Get the country code for the node's IP address. + let regionCode; + try { + // Expect a 2-letter ISO3166-1 code, which should also be a valid + // BCP47 Region subtag. + regionCode = await conn.getInfo("ip-to-country/" + node.ipAddrs[0]); + } catch {} + if (regionCode && regionCode !== "??") { + node.regionCode = regionCode.toUpperCase(); + } + } + return node; + }); + }, + + async onionAuthAdd(hsAddress, b64PrivateKey, isPermanent) { + return this._withConnection(conn => { + return conn.onionAuthAdd(hsAddress, b64PrivateKey, isPermanent); + }); + }, + + async onionAuthRemove(hsAddress) { + return this._withConnection(conn => { + return conn.onionAuthRemove(hsAddress); + }); + }, + + async onionAuthViewKeys() { + return this._withConnection(conn => { + return conn.onionAuthViewKeys(); + }); + }, + // TODO: transform the following 4 functions in getters. At the moment they // are also used in torbutton. @@ -630,6 +727,16 @@ export const TorProtocolService = { } }, + async _withConnection(func) { + // TODO: Make more robust? + const conn = await this._getConnection(); + try { + return await func(conn); + } finally { + this._returnConnection(); + } + }, + // If aConn is omitted, the cached connection is closed. _closeConnection() { if (this._controlConnection) { ===================================== toolkit/components/tor-launcher/TorStartupService.sys.mjs ===================================== @@ -3,6 +3,7 @@ const lazy = {}; // We will use the modules only when the profile is loaded, so prefer lazy // loading ChromeUtils.defineESModuleGetters(lazy, { + TorDomainIsolator: "resource://gre/modules/TorDomainIsolator.sys.mjs", TorLauncherUtil: "resource://gre/modules/TorLauncherUtil.sys.mjs", TorMonitorService: "resource://gre/modules/TorMonitorService.sys.mjs", TorProtocolService: "resource://gre/modules/TorProtocolService.sys.mjs", @@ -19,12 +20,6 @@ ChromeUtils.defineModuleGetter( "resource:///modules/TorSettings.jsm" ); -ChromeUtils.defineModuleGetter( - lazy, - "TorDomainIsolator", - "resource://gre/modules/TorDomainIsolator.jsm" -); - /* Browser observer topis */ const BrowserTopics = Object.freeze({ ProfileAfterChange: "profile-after-change", ===================================== toolkit/components/tor-launcher/moz.build ===================================== @@ -1,6 +1,6 @@ EXTRA_JS_MODULES += [ "TorBootstrapRequest.sys.mjs", - "TorDomainIsolator.jsm", + "TorDomainIsolator.sys.mjs", "TorLauncherUtil.sys.mjs", "TorMonitorService.sys.mjs", "TorParsers.sys.mjs", View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/bdda46… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/bdda46… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser] Pushed new branch base-browser-115.1.0esr-13.0-1
by Pier Angelo Vendrame (＠pierov) 27 Jul '23

27 Jul '23

Pier Angelo Vendrame pushed new branch base-browser-115.1.0esr-13.0-1 at The Tor Project / Applications / Tor Browser -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/tree/base-brow… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser] Pushed new tag tor-browser-115.1.0esr-13.0-1-build1
by Pier Angelo Vendrame (＠pierov) 27 Jul '23

27 Jul '23

Pier Angelo Vendrame pushed new tag tor-browser-115.1.0esr-13.0-1-build1 at The Tor Project / Applications / Tor Browser -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/tree/tor-brows… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][tor-browser-115.1.0esr-13.0-1] 116 commits: Bug 1832523 - Allow using NSS to sign and verify MAR signatures....
by Pier Angelo Vendrame (＠pierov) 27 Jul '23

27 Jul '23

Pier Angelo Vendrame pushed to branch tor-browser-115.1.0esr-13.0-1 at The Tor Project / Applications / Tor Browser Commits: 5039cc76 by Pier Angelo Vendrame at 2023-07-25T10:23:19+02:00 Bug 1832523 - Allow using NSS to sign and verify MAR signatures. r=application-update-reviewers,glandium,bytesized Allow using NSS for checking MAR signatures also in platforms where OS-native APIs are used by default, i.e., macOS and Windows. Differential Revision: https://phabricator.services.mozilla.com/D177743 - - - - - c35d6545 by Henry Wilkes at 2023-07-25T10:23:22+02:00 Bug 41454: Move focus after calling openPreferences for a sub-category. Temporary fix until mozilla bug 1799153 gets a patch upstream. - - - - - bf8d4092 by Henry Wilkes at 2023-07-25T10:23:23+02:00 Bug 41483: Remove the firefox override for appstrings.properties Remove this patch after upstream bugzilla bug 1790187 - - - - - d0d369ac by Pier Angelo Vendrame at 2023-07-25T11:10:35+02:00 Bug 41116: Normalize system fonts. System fonts are an enormous fingerprinting vector. Even with font allow lists and with our custom configuration on Linux, which counter metrics measurements, getComputedStyle leaks several details. This patch counters both these kinds of attacks. - - - - - f8ef7d1e by Marco Simonelli at 2023-07-25T11:10:36+02:00 Bug 41459: WebRTC fails to build under mingw (Part 1) - properly define NOMINMAX for just MSVC builds - - - - - 2617640b by Marco Simonelli at 2023-07-25T11:10:37+02:00 Bug 41459: WebRTC fails to build under mingw (Part 2) - fixes required to build third_party/libwebrtc - - - - - 1cf88911 by Marco Simonelli at 2023-07-25T11:10:37+02:00 Bug 41459: WebRTC fails to build under mingw (Part 3) - fixes required to build third_party/sipcc - - - - - 8443845b by Marco Simonelli at 2023-07-25T11:10:38+02:00 Bug 41459: WebRTC fails to build under mingw (Part 4) - fixes requried to build netwerk/sctp - - - - - 9f2a9cfb by Marco Simonelli at 2023-07-25T11:10:38+02:00 Bug 41459: WebRTC fails to build under mingw (Part 5) - fixes required to build dom/media/webrtc - - - - - 0b45cd9a by Marco Simonelli at 2023-07-25T11:10:39+02:00 Bug 41459: WebRTC fails to build under mingw (Part 6) - fixes required to build dom/media/systemservices - - - - - d98e0066 by hackademix at 2023-07-25T11:10:39+02:00 Bug 41854: Allow overriding download spam protection. - - - - - ab596fbe by Gaba at 2023-07-25T11:10:40+02:00 Adding issue and merge request templates - - - - - fea58a11 by Pier Angelo Vendrame at 2023-07-25T11:10:43+02:00 Base Browser's .mozconfigs. Bug 17858: Cannot create incremental MARs for hardened builds. Define HOST_CFLAGS, etc. to avoid compiling programs such as mbsdiff (which is part of mar-tools and is not distributed to end-users) with ASan. Bug 21849: Don't allow SSL key logging. Bug 25741 - TBA: Disable features at compile-time Define MOZ_ANDROID_NETWORK_STATE and MOZ_ANDROID_LOCATION Bug 27623 - Export MOZILLA_OFFICIAL during desktop builds This fixes a problem where some preferences had the wrong default value. Also see bug 27472 where we made a similar fix for Android. Bug 29859: Disable HLS support for now Bug 30463: Explicitly disable MOZ_TELEMETRY_REPORTING Bug 32493: Disable MOZ_SERVICES_HEALTHREPORT Bug 33734: Set MOZ_NORMANDY to False Bug 33851: Omit Parental Controls. Bug 40252: Add --enable-rust-simd to our tor-browser mozconfig files Bug 41584: Move some configuration options to base-browser level - - - - - 6aef387e by Pier Angelo Vendrame at 2023-07-25T11:19:14+02:00 fixup! Base Browser's .mozconfigs. Always use lld on Android. - - - - - 8d61dbfc by Pier Angelo Vendrame at 2023-07-25T11:19:15+02:00 fixup! Base Browser's .mozconfigs. Bug 41890: Let tor-browser-build set the update channel it prefers. Forgot to update the mozconfig-android-all in the previous commit. - - - - - 74bfde5d by Pier Angelo Vendrame at 2023-07-25T11:19:15+02:00 Tweaks to the build system Bug 40857: Modified the fat .aar creation file This is a workaround to build fat .aars with the compiling enviornment disabled. Mozilla does not use a similar configuration, but either runs a Firefox build and discards its output, or uses artifacts build. We might switch to artifact builds too, and drop this patch, or write a better one to upstream. But until then we need this patch. See also https://bugzilla.mozilla.org/show_bug.cgi?id=1763770. Bug 41458: Prevent `mach package-multi-locale` from actually creating a package macOS builds need some files to be moved around with ./mach package-multi-locale to create multi-locale packages. The required command isn't exposed through any other mach command. So, we patch package-multi-locale both to prevent it from failing when doing official builds and to detect any future changes on it. - - - - - f5d7bb8e by Pier Angelo Vendrame at 2023-07-25T11:19:16+02:00 Bug 41108: Remove privileged macOS installation from 102 - - - - - 81b084ac by Dan Ballard at 2023-07-25T11:19:16+02:00 Bug 41149: Re-enable DLL injection protection in all builds not just nightlies - - - - - e2af470f by Matthew Finkel at 2023-07-25T11:19:16+02:00 Bug 24796: Comment out excess permissions from GeckoView The GeckoView AndroidManifest.xml is not preprocessed unlike Fennec's manifest, so we can't use the ifdef preprocessor guards around the permissions we do not want. Commenting the permissions is the next-best-thing. - - - - - 8382899e by Matthew Finkel at 2023-07-25T11:19:17+02:00 Bug 28125: Prevent non-Necko network connections - - - - - 3572aad9 by Mike Perry at 2023-07-25T11:19:17+02:00 Bug 12974: Disable NTLM and Negotiate HTTP Auth The Mozilla bugs: https://bugzilla.mozilla.org/show_bug.cgi?id=1046421, https://bugzilla.mozilla.org/show_bug.cgi?id=1261591, tor-browser#27602 - - - - - dddf73d3 by Alex Catarineu at 2023-07-25T11:19:18+02:00 Bug 40166: Disable security.certerrors.mitm.auto_enable_enterprise_roots Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1768899 - - - - - 7cff6ac8 by Georg Koppen at 2023-07-25T11:19:18+02:00 Bug 16285: Exclude ClearKey system for now In the past the ClearKey system had not been compiled when specifying --disable-eme. But that changed and it is even bundled nowadays (see: Mozilla's bug 1300654). We don't want to ship it right now as the use case for it is not really visible while the code had security vulnerabilities in the past. - - - - - 618fba42 by Kathy Brade at 2023-07-25T11:19:19+02:00 Bug 21431: Clean-up system extensions shipped in Firefox Only ship the pdfjs extension. - - - - - 0b405c2a by Kathy Brade at 2023-07-25T11:19:19+02:00 Bug 33852: Clean up about:logins (LockWise) to avoid mentioning sync, etc. Hide elements on about:logins that mention sync, "Firefox LockWise", and Mozilla's LockWise mobile apps. Disable the "Create New Login" button when security.nocertdb is true. - - - - - fc84956d by Alex Catarineu at 2023-07-25T11:19:19+02:00 Bug 41457: Remove Mozilla permissions Bug 40025: Remove Mozilla add-on install permissions - - - - - d5f1422c by Kathy Brade at 2023-07-25T11:19:20+02:00 Bug 40002: Remove about:ion Firefox Ion (previously Firefox Pioneer) is an opt-in program in which people volunteer to participate in studies that collect detailed, sensitive data about how they use their browser. Bug 41662: Disable about:sync-logs Even though we disable sync by default with `identity.fxaccounts.enabled`, this about: page is still avilable. We could throw an exception on the constructor of the related component, but it would result only in an error in the console, without a visible "this address does not look right" error page. If we fix the issues with MOZ_SERVICES_SYNC, we can restore the component. - - - - - e4304ef6 by Arthur Edelstein at 2023-07-25T11:19:20+02:00 Bug 26353: Prevent speculative connect that violated FPI. Connections were observed in the catch-all circuit when the user entered an https or http URL in the URL bar, or typed a search term. - - - - - 155fb94c by Alex Catarineu at 2023-07-25T11:19:21+02:00 Bug 31740: Remove some unnecessary RemoteSettings instances More concretely, SearchService.jsm 'hijack-blocklists' and url-classifier-skip-urls. Avoid creating instance for 'anti-tracking-url-decoration'. If prefs are disabling their usage, avoid creating instances for 'cert-revocations' and 'intermediates'. Do not ship JSON dumps for collections we do not expect to need. For the ones in the 'main' bucket, this prevents them from being synced unnecessarily (the code in remote-settings does so for collections in the main bucket for which a dump or local data exists). For the collections in the other buckets, we just save some size by not shipping their dumps. We also clear the collections database on the v2 -> v3 migration. - - - - - 223fa2f4 by cypherpunks1 at 2023-07-25T11:19:21+02:00 Bug 41092: Add a RemoteSettings JSON dump for query-stripping - - - - - 34e93dea by Pier Angelo Vendrame at 2023-07-25T11:19:22+02:00 Bug 41635: Disable the Normandy component Do not include Normandy at all whenever MOZ_NORMANDY is False. - - - - - 6af17016 by Georg Koppen at 2023-07-25T11:19:22+02:00 Bug 30541: Disable WebGL readPixel() for web content Related Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1428034 - - - - - 498786e7 by Alex Catarineu at 2023-07-25T11:19:22+02:00 Bug 28369: Stop shipping pingsender executable - - - - - 918d7a34 by cypherpunks1 at 2023-07-25T11:19:23+02:00 Bug 41568: Disable LaterRun - - - - - 188293e5 by cypherpunks1 at 2023-07-25T11:19:23+02:00 Bug 40717: Hide Windows SSO in settings - - - - - a7b44945 by Pier Angelo Vendrame at 2023-07-25T11:19:24+02:00 Bug 41599: Always return an empty string as network ID Firefox computes an internal network ID used to detect network changes and act consequently (e.g., to improve WebSocket UX). However, there are a few ways to get this internal network ID, so we patch them out, to be sure any new code will not be able to use them and possibly link users. We also sent a patch to Mozilla to seed the internal network ID, to prevent any accidental leak in the future. Upstream: https://bugzilla.mozilla.org/show_bug.cgi?id=1817756 - - - - - 4c2ebaeb by Mike Perry at 2023-07-25T11:19:24+02:00 Firefox preference overrides. This hack directly includes our preference changes in omni.ja. Bug 18292: Staged updates fail on Windows Temporarily disable staged updates on Windows. Bug 18297: Use separate Noto JP,KR,SC,TC fonts Bug 23404: Add Noto Sans Buginese to the macOS whitelist Bug 23745: Set dom.indexedDB.enabled = true Bug 13575: Disable randomised Firefox HTTP cache decay user tests. (Fernando Fernandez Mancera <ffmancera(a)riseup.net>) Bug 17252: Enable session identifiers with FPI Session tickets and session identifiers were isolated by OriginAttributes, so we can re-enable them by allowing the default value (true) of "security.ssl.disable_session_identifiers". The pref "security.enable_tls_session_tickets" is obsolete (removed in https://bugzilla.mozilla.org/917049) Bug 14952: Enable http/2 and AltSvc In Firefox, SPDY/HTTP2 now uses Origin Attributes for isolation of connections, push streams, origin frames, etc. That means we get first-party isolation provided "privacy.firstparty.isolate" is true. So in this patch, we stop overriding "network.http.spdy.enabled" and "network.http.spdy.enabled.http2". Alternate Services also use Origin Attributes for isolation. So we stop overriding "network.http.altsvc.enabled" and "network.http.altsvc.oe" as well. (All 4 of the abovementioned "network.http.*" prefs adopt Firefox 60ESR's default value of true.) However, we want to disable HTTP/2 push for now, so we set "network.http.spdy.allow-push" to false. "network.http.spdy.enabled.http2draft" was removed in Bug 1132357. "network.http.sped.enabled.v2" was removed in Bug 912550. "network.http.sped.enabled.v3" was removed in Bug 1097944. "network.http.sped.enabled.v3-1" was removed in Bug 1248197. Bug 26114: addons.mozilla.org is not special * Don't expose navigator.mozAddonManager on any site * Don't block NoScript from modifying addons.mozilla.org or other sites Enable ReaderView mode again (#27281). Bug 29916: Make sure enterprise policies are disabled Bug 2874: Block Components.interfaces from content Bug 26146: Spoof HTTP User-Agent header for desktop platforms In Tor Browser 8.0, the OS was revealed in both the HTTP User-Agent header and to JavaScript code via navigator.userAgent. To avoid leaking the OS inside each HTTP request (which many web servers log), always use the Windows 7 OS value in the desktop User-Agent header. We continue to allow access to the actual OS via JavaScript, since doing so improves compatibility with web applications such as GitHub and Google Docs. Bug 12885: Windows Jump Lists fail for Tor Browser Jumplist entries are stored in a binary file in: %APPDATA%\\Microsoft\Windows\Recent\CustomDestinations\ and has a name in the form [a-f0-9]+.customDestinations-ms The hex at the front is unique per app, and is ultimately derived from something called the 'App User Model ID' (AUMID) via some unknown hashing method. The AUMID is provided as a key when programmatically creating, updating, and deleting a jumplist. The default behaviour in firefox is for the installer to define an AUMID for an app, and save it in the registry so that the jumplist data can be removed by the uninstaller. However, the Tor Browser does not set this (or any other) regkey during installation, so this codepath fails and the app's AUMID is left undefined. As a result the app's AUMID ends up being defined by windows, but unknowable by Tor Browser. This unknown AUMID is used to create and modify the jumplist, but the delete API requires that we provide the app's AUMID explicitly. Since we don't know what the AUMID is (since the expected regkey where it is normally stored does not exist) jumplist deletion will fail and we will leave behind a mostly empty customDestinations-ms file. The name of the file is derived from the binary path, so an enterprising person could reverse engineer how that hex name is calculated, and generate the name for Tor Browser's default Desktop installation path to determine whether a person had used Tor Browser in the past. The 'taskbar.grouping.useprofile' option that is enabled by this patch works around this AUMID problem by having firefox.exe create it's own AUMID based on the profile path (rather than looking for a regkey). This way, if a user goes in and enables and disables jumplist entries, the backing store is properly deleted. Unfortunately, all windows users currently have this file lurking in the above mentioned directory and this patch will not remove it since it was created with an unknown AUMID. However, another patch could be written which goes to that directory and deletes any item containing the 'Tor Browser' string. See bug 28996. Bug 30845: Make sure default themes and other internal extensions are enabled Bug 28896: Enable extensions in private browsing by default Bug 31065: Explicitly allow proxying localhost Bug 31598: Enable letterboxing Disable Presentation API everywhere Bug 21549 - Use Firefox's WASM default pref. It is disabled at safer security levels. Bug 32321: Disable Mozilla's MitM pings Bug 19890: Disable installation of system addons By setting the URL to "" we make sure that already installed system addons get deleted as well. Bug 22548: Firefox downgrades VP9 videos to VP8. On systems where H.264 is not available or no HWA, VP9 is preferred. But in Tor Browser 7.0 all youtube videos are degraded to VP8. This behaviour can be turned off by setting media.benchmark.vp9.threshold to 0. All clients will get better experience and lower traffic, beause TBB doesn't use "Use hardware acceleration when available". Bug 25741 - TBA: Add mobile-override of 000-tor-browser prefs Bug 16441: Suppress "Reset Tor Browser" prompt. Bug 29120: Use the in-memory media cache and increase its maximum size. Bug 33697: use old search config based on list.json Bug 33855: Ensure that site-specific browser mode is disabled. Bug 30682: Disable Intermediate CA Preloading. Bug 40061: Omit the Windows default browser agent from the build Bug 40322: Consider disabling network.connectivity-service.enabled Bug 40408: Disallow SVG Context Paint in all web content Bug 40308: Disable network partitioning until we evaluate dFPI Bug 40322: Consider disabling network.connectivity-service.enabled Bug 40383: Disable dom.enable_event_timing Bug 40423: Disable http/3 Bug 40177: Update prefs for Fx91esr Bug 40700: Disable addons and features recommendations Bug 40682: Disable network.proxy.allow_bypass Bug 40736: Disable third-party cookies in PBM Bug 19850: Enabled HTTPS-Only by default Bug 40912: Hide the screenshot menu Bug 41292: Disable moreFromMozilla in preferences page Bug 40057: Ensure the CSS4 system colors are not a fingerprinting vector Bug 24686: Set network.http.tailing.enabled to true Bug 40183: Disable TLS ciphersuites using SHA-1 Bug 40783: Review 000-tor-browser.js and 001-base-profile.js for 102 We reviewed all the preferences we set for 102, and remove a few old ones. See the description of that issue to see all the preferences we believed were still valid for 102, and some brief description for the reasons to keep them. - - - - - 31a3cea4 by Pier Angelo Vendrame at 2023-07-25T11:19:25+02:00 fixup! Firefox preference overrides. Bug 41806: Prevent Private Browsing start menu item to be automatically added. - - - - - 83d0c51c by Pier Angelo Vendrame at 2023-07-27T09:55:58+02:00 fixup! Firefox preference overrides. Bug 41732: Use font.system.whitelist also on Linux as a defense-in-depth - - - - - e3df3023 by Richard Pospesel at 2023-07-27T09:55:59+02:00 Bug 41659: Add canonical color definitions to base-browser - - - - - c5b76528 by Pier Angelo Vendrame at 2023-07-27T09:56:00+02:00 Bug 41043: Hardcode the UI font on Linux The mechanism to choose the UI font does not play well with our fontconfig configuration. As a result, the final criterion to choose the font for the UI was its version. Since we hardcode Arimo as a default sans-serif on preferences, we use it also for the UI. FontConfig will fall back to some other font for scripts Arimo does not cover as expected (we tested with Japanese). - - - - - 5eacfcf5 by Alex Catarineu at 2023-07-27T09:56:00+02:00 Bug 30605: Honor privacy.spoof_english in Android This checks `privacy.spoof_english` whenever `setLocales` is called from Fenix side and sets `intl.accept_languages` accordingly. Bug 40198: Expose privacy.spoof_english pref in GeckoView - - - - - cc97c1ea by Alex Catarineu at 2023-07-27T09:56:00+02:00 Bug 40199: Avoid using system locale for intl.accept_languages in GeckoView - - - - - 880a41c9 by Alex Catarineu at 2023-07-27T09:56:01+02:00 Bug 40171: Make WebRequest and GeckoWebExecutor First-Party aware - - - - - 183c5d6a by Alex Catarineu at 2023-07-27T09:56:01+02:00 Bug 26345: Hide tracking protection UI - - - - - a3ab67d4 by Pier Angelo Vendrame at 2023-07-27T09:56:02+02:00 Bug 9173: Change the default Firefox profile directory to be relative. This commit makes Firefox look for the default profile directory in a directory relative to the binary path. The directory can be specified through the --with-relative-data-dir. This is relative to the same directory as the firefox main binary for Linux and Windows. On macOS, we remove Contents/MacOS from it. Or, in other words, the directory is relative to the application bundle. This behavior can be overriden at runtime, by placing a file called system-install adjacent to the firefox main binary (also on macOS). - - - - - ab6cec43 by Alex Catarineu at 2023-07-27T09:56:02+02:00 Bug 27604: Fix addon issues when moving the profile directory Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1429838 - - - - - f42a137b by Mike Perry at 2023-07-27T09:56:03+02:00 Bug 13028: Prevent potential proxy bypass cases. It looks like these cases should only be invoked in the NSS command line tools, and not the browser, but I decided to patch them anyway because there literally is a maze of network function pointers being passed around, and it's very hard to tell if some random code might not pass in the proper proxied versions of the networking code here by accident. Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1433509 - - - - - 39819e87 by Igor Oliveira at 2023-07-27T09:56:03+02:00 Bug 23104: Add a default line height compensation Many fonts have issues with their vertical metrics. they are used to influence the height of ascenders and depth of descenders. Gecko uses it to calculate the line height (font height + ascender + descender), however because of that idiosyncratic behavior across multiple operating systems, it can be used to identify the user's OS. The solution proposed in the patch uses a default factor to be multiplied with the font size, simulating the concept of ascender and descender. This way all operating systems will have the same line height. - - - - - a848a42e by Pier Angelo Vendrame at 2023-07-27T09:56:04+02:00 Bug 40309: Avoid using regional OS locales Avoid regional OS locales if the pref `intl.regional_prefs.use_os_locales` is false but RFP is enabled. - - - - - 32e8a75c by Matthew Finkel at 2023-07-27T09:56:04+02:00 Bug 40432: Prevent probing installed applications Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1711084 - - - - - 964f9e58 by cypherpunks1 at 2023-07-27T09:56:04+02:00 Bug 33955: When copying an image only copy the image contents to the clipboard - - - - - b228060e by cypherpunks1 at 2023-07-27T09:56:05+02:00 Bug 41791: Omit the source URL when copying page contents to the clipboard - - - - - 96571e9b by Pier Angelo Vendrame at 2023-07-27T09:56:05+02:00 Base Browser strings This commit adds all the strings needed by following Base Browser patches. - - - - - 570965c9 by hackademix at 2023-07-27T09:56:06+02:00 Bug 41434: Letterboxing, preemptively apply margins in a global CSS rule to mitigate race conditions on newly created windows and tabs. - - - - - cc650be6 by hackademix at 2023-07-27T09:56:06+02:00 Bug 41434: Letterboxing, improve logging. - - - - - a1a205f3 by hackademix at 2023-07-27T09:56:07+02:00 Bug 31064: Letterboxing, exempt browser extensions. - - - - - 1e2e47c8 by hackademix at 2023-07-27T09:56:07+02:00 Bug 32411: Letterboxing, exempt view-source: URIs. - - - - - 30c23116 by hackademix at 2023-07-27T09:56:07+02:00 Bug 32308: use direct browser sizing for letterboxing. - - - - - c4066fed by hackademix at 2023-07-27T09:56:08+02:00 Bug 41631: Prevent weird initial window dimensions caused by subpixel computations - - - - - 1d4901db by Pier Angelo Vendrame at 2023-07-27T09:56:08+02:00 Bug 41369: Improve Firefox language settings for multi-lingual packages Change the language selector to be sorted by language code, rather than name, and to display the language code to the user. Bug 41372: Handle Japanese as a special case in preferences on macOS Japanese is treated in a special way on macOS. However, seeing the Japanese language tag could be confusing for users, and moreover the language name is not localized correctly like other langs. Bug 41378: Tell users that they can change their language at the first start With multi-lingual builds, Tor Browser matches the user's system language, but some users might want to change it. So, we tell them that it is possible, but only once. - - - - - 55ebb736 by p13dz at 2023-07-27T09:56:09+02:00 Bug 40283: Workaround for the file upload bug - - - - - f9f46be2 by Arthur Edelstein at 2023-07-27T09:56:09+02:00 Bug 18905: Hide unwanted items from help menu Bug 25660: Remove the "New Private Window" option - - - - - 7ed790e6 by Pier Angelo Vendrame at 2023-07-27T09:56:10+02:00 Bug 41603: Customize the creation of MOZ_SOURCE_URL MOZ_SOURCE_URL is created by combining MOZ_SOURCE_REPO and MOZ_SOURCE_CHANGESET. But the code takes for granted that it refers to a Hg instance, so it combines them as `$MOZ_SOURCE_REPO/rev/$MOZ_SOURCE_CHANGESET`. With this commit, we change this logic to combine them to create a URL that is valid for GitLab. $MOZ_SOURCE_CHANGESET needs to be a commit hash, not a branch or a tag. If that is needed, we could use /-/tree/, instead of /-/commit/. - - - - - c11e65ca by Henry Wilkes at 2023-07-27T09:56:10+02:00 Bug 31575: Disable Firefox Home (Activity Stream) Treat about:blank as the default home page and new tab page. Avoid loading AboutNewTab in BrowserGlue.sys.mjs in order to avoid several network requests that we do not need. Bug 41624: Disable about:pocket-* pages. Bug 40144: Redirect about:privatebrowsing to the user's home - - - - - 4ef92224 by Kathy Brade at 2023-07-27T09:56:10+02:00 Bug 4234: Use the Firefox Update Process for Base Browser. Windows: disable "runas" code path in updater (15201). Windows: avoid writing to the registry (16236). Also includes fixes for tickets 13047, 13301, 13356, 13594, 15406, 16014, 16909, 24476, and 25909. Also fix bug 27221: purge the startup cache if the Base Browser version changed (even if the Firefox version and build ID did not change), e.g., after a minor Base Browser update. Also fix 32616: Disable GetSecureOutputDirectoryPath() functionality. Bug 26048: potentially confusing "restart to update" message Within the update doorhanger, remove the misleading message that mentions that windows will be restored after an update is applied, and replace the "Restart and Restore" button label with an existing "Restart to update Tor Browser" string. Bug 28885: notify users that update is downloading Add a "Downloading Base Browser update" item which appears in the hamburger (app) menu while the update service is downloading a MAR file. Before this change, the browser did not indicate to the user that an update was in progress, which is especially confusing in Tor Browser because downloads often take some time. If the user clicks on the new menu item, the about dialog is opened to allow the user to see download progress. As part of this fix, the update service was changed to always show update-related messages in the hamburger menu, even if the update was started in the foreground via the about dialog or via the "Check for Tor Browser Update" toolbar menu item. This change is consistent with the Tor Browser goal of making sure users are informed about the update process. Removed #28885 parts of this patch which have been uplifted to Firefox. - - - - - 195de779 by Nicolas Vigier at 2023-07-27T09:56:11+02:00 Bug 41682: Add base-browser nightly mar signing key - - - - - 259032dd by hackademix at 2023-07-27T09:56:11+02:00 Bug 41695: Warn on window maximization without letterboxing in RFPHelper module - - - - - 2a09e39e by Pier Angelo Vendrame at 2023-07-27T09:56:12+02:00 Bug 41698: Reword the recommendation badges in about:addons Firefox strings use { -brand-product-name }. As a result, it seems that the fork is recommending extensions, whereas AMO curators are doing that. So, we replace the strings with custom ones that clarify that Mozilla is recommending them. We assign the strings with JS because our translation backend does not support Fluent attributes, yet, but once it does, we should switch to them, instead. Upstream bug: https://bugzilla.mozilla.org/show_bug.cgi?id=1825033 - - - - - d3b506c7 by Alex Catarineu at 2023-07-27T09:56:12+02:00 Bug 40069: Add helpers for message passing with extensions - - - - - c46db8d1 by Matthew Finkel at 2023-07-27T09:56:13+02:00 Bug 41598: Prevent NoScript from being removed/disabled. Bug 40253: Explicitly allow NoScript in Private Browsing mode. - - - - - 2c215baf by hackademix at 2023-07-27T09:56:13+02:00 Bug 41834: Hide "Can't Be Removed - learn more" menu line for uninstallable add-ons - - - - - 0e037600 by Pier Angelo Vendrame at 2023-07-27T09:56:13+02:00 Bug 40925: Implemented the Security Level component This component adds a new Security Level toolbar button which visually indicates the current global security level via icon (as defined by the extensions.torbutton.security_slider pref), a drop-down hanger with a short description of the current security level, and a new section in the about:preferences#privacy page where users can change their current security level. In addition, the hanger and the preferences page will show a visual warning when the user has modified prefs associated with the security level and provide a one-click 'Restore Defaults' button to get the user back on recommended settings. Bug 40125: Expose Security Level pref in GeckoView - - - - - 1615b2fd by Pier Angelo Vendrame at 2023-07-27T09:56:16+02:00 Bug 40926: Implemented the New Identity feature - - - - - c847c74b by Henry Wilkes at 2023-07-27T09:56:16+02:00 Bug 41736: Customize toolbar for base-browser. - - - - - 970ae0d8 by Richard Pospesel at 2023-07-27T09:56:17+02:00 Bug 41649: Create rebase and security backport gitlab issue templates - - - - - 2546c1ce by Richard Pospesel at 2023-07-27T09:56:17+02:00 Bug 41089: Add tor-browser build scripts + Makefile to tor-browser - - - - - a29b93c6 by Henry Wilkes at 2023-07-27T09:56:17+02:00 Bug 41803: Add some developer tools for working on tor-browser. - - - - - 845e2a11 by Kathy Brade at 2023-07-27T09:56:18+02:00 Bug 11641: Disable remoting by default. Unless the -osint command line flag is used, the browser now defaults to the equivalent of -no-remote. There is a new -allow-remote flag that may be used to restore the original (Firefox-like) default behavior. - - - - - cf1d39d2 by Mike Perry at 2023-07-27T09:56:18+02:00 Bug 2176: Rebrand Firefox to TorBrowser See also Bugs #5194, #7187, #8115, #8219. This patch does some basic renaming of Firefox to TorBrowser. The rest of the branding is done by images and icons. Also fix bug 27905. Bug 25702: Update Tor Browser icon to follow design guidelines - Updated all of the branding in /browser/branding/official with new 'stable' icon series. - Updated /extensions/onboarding/content/img/tor-watermark.png with new icon and add the source svg in the same directory - Copied /browser/branding/official over /browser/branding/nightly and the new /browser/branding/alpha directories. Replaced content with 'nightly' and 'alpha' icon series. Updated VisualElements_70.png and VisualElements_150.png with updated icons in each branding directory (fixes #22654) - Updated firefox.VisualElementsManfiest.xml with updated colors in each branding directory - Added firefox.svg to each branding directory from which all the other icons are derived (apart from document.icns and document.ico) - Added default256.png and default512.png icons - Updated aboutTBUpdate.css to point to branding-aware icon128.png and removed original icon - Use the Tor Browser icon within devtools/client/themes/images/. Bug 30631: Blurry Tor Browser icon on macOS app switcher It would seem the png2icns tool does not generate correct icns files and so on macOS the larger icons were missing resulting in blurry icons in the OS chrome. Regenerated the padded icons in a macOS VM using iconutil. Bug 28196: preparations for using torbutton tor-browser-brand.ftl A small change to Fluent FileSource class is required so that we can register a new source without its supported locales being counted as available locales for the browser. Bug 31803: Replaced about:debugging logo with flat version Bug 21724: Make Firefox and Tor Browser distinct macOS apps When macOS opens a document or selects a default browser, it sometimes uses the CFBundleSignature. Changing from the Firefox MOZB signature to a different signature TORB allows macOS to distinguish between Firefox and Tor Browser. Bug 32092: Fix Tor Browser Support link in preferences For bug 40562, we moved onionPattern* from bug 27476 to here, as about:tor needs these files but it is included earlier. Bug 41278: Create Tor Browser styled pdf logo similar to the vanilla Firefox one - - - - - 2730e4d8 by sanketh at 2023-07-27T09:56:19+02:00 Bug 40209: Implement Basic Crypto Safety Adds a CryptoSafety actor which detects when you've copied a crypto address from a HTTP webpage and shows a warning. Closes #40209. Bug 40428: Fix string attribute names - - - - - 86e41729 by Mike Perry at 2023-07-27T09:56:19+02:00 TB3: Tor Browser's official .mozconfigs. Also: Add an --enable-tor-browser-data-outside-app-dir configure option Add --with-tor-browser-version configure option Bug 31457: disable per-installation profiles The dedicated profiles (per-installation) feature does not interact well with our bundled profiles on Linux and Windows, and it also causes multiple profiles to be created on macOS under TorBrowser-Data. Bug 31935: Disable profile downgrade protection. Since Tor Browser does not support more than one profile, disable the prompt and associated code that offers to create one when a version downgrade situation is detected. Add --enable-tor-browser-update build option Bug 40793: moved Tor configuration options from old-configure.in to moz.configure Bug 41584: Move some configuration options to base-browser level - - - - - c8357ebe by Henry Wilkes at 2023-07-27T09:56:20+02:00 Bug 41340: Enable TOR_BROWSER_NIGHTLY_BUILD features for dev and nightly builds tor-browser#41285: Enable fluent warnings. - - - - - 8deed10e by Pier Angelo Vendrame at 2023-07-27T09:56:20+02:00 Bug 40562: Added Tor Browser preferences to 000-tor-browser.js Before reordering patches, we used to keep the Tor-related patches (torbutton and tor-launcher) at the beginning. After that issue, we decided to move them towards the end. In addition to that, we have decided to move Tor Browser-only preferences there, too, to make Base Browser-only fixups easier to apply. - - - - - c824681a by Pier Angelo Vendrame at 2023-07-27T09:56:21+02:00 Bug 13252: Customize profile management on macOS On macOS we allow both portable mode and system installation. However, in the latter case, we customize Firefox's directories to match the hierarchy we use for the portable mode. Also, display an informative error message if the TorBrowser-Data directory cannot be created due to an "access denied" or a "read only volume" error. - - - - - ed2f93d4 by Pier Angelo Vendrame at 2023-07-27T09:56:21+02:00 Bug 40933: Add tor-launcher functionality - - - - - daf07d2b by Richard Pospesel at 2023-07-27T09:56:21+02:00 Bug 40597: Implement TorSettings module - migrated in-page settings read/write implementation from about:preferences#tor to the TorSettings module - TorSettings initially loads settings from the tor daemon, and saves them to firefox prefs - TorSettings notifies observers when a setting has changed; currently only QuickStart notification is implemented for parity with previous preference notify logic in about:torconnect and about:preferences#tor - about:preferences#tor, and about:torconnect now read and write settings thorugh the TorSettings module - all tor settings live in the torbrowser.settings.* preference branch - removed unused pref modify permission for about:torconnect content page from AsyncPrefs.jsm Bug 40645: Migrate Moat APIs to Moat.jsm module - - - - - 42f5b8d7 by Alex Catarineu at 2023-07-27T09:56:22+02:00 Bug 10760: Integrate TorButton to TorBrowser core Because of the non-restartless nature of Torbutton, it required a two-stage installation process. On mobile, it was a problem, because it was not loading when the user opened the browser for the first time. Moving it to tor-browser and making it a system extension allows it to load when the user opens the browser for first time. Additionally, this patch also fixes Bug 27611. Bug 26321: New Circuit and New Identity menu items Bug 14392: Make about:tor behave like other initial pages. Bug 25013: Add torbutton as a tor-browser submodule Bug 31575: Replace Firefox Home (newtab) with about:tor - - - - - 493ce8c9 by Arthur Edelstein at 2023-07-27T09:56:22+02:00 Bug 3455: Add DomainIsolator, for isolating circuit by domain. Add an XPCOM component that registers a ProtocolProxyChannelFilter which sets the username/password for each web request according to url bar domain. Bug 9442: Add New Circuit button Bug 13766: Set a 10 minute circuit dirty timeout for the catch-all circ. Bug 19206: Include a 128 bit random tag as part of the domain isolator nonce. Bug 19206: Clear out the domain isolator state on `New Identity`. Bug 21201.2: Isolate by firstPartyDomain from OriginAttributes Bug 21745: Fix handling of catch-all circuit Bug 41741: Refactor the domain isolator and new circuit - - - - - dad80d9a by Henry Wilkes at 2023-07-27T09:56:23+02:00 Bug 41600: Add a tor circuit display panel. - - - - - be4deedc by hackademix at 2023-07-27T09:56:23+02:00 Bug 8324: Prevent DNS proxy bypasses caused by Drag&Drop Bug 41613: Skip Drang & Drop filtering for DNS-safe URLs - - - - - a34a0036 by Amogh Pradeep at 2023-07-27T09:56:24+02:00 Orfox: Centralized proxy applied to AbstractCommunicator and BaseResources. See Bug 1357997 for partial uplift. Also: Bug 28051 - Use our Orbot for proxying our connections Bug 31144 - ESR68 Network Code Review - - - - - f84eea90 by Matthew Finkel at 2023-07-27T09:56:24+02:00 Bug 25741: TBA: Disable GeckoNetworkManager The browser should not need information related to the network interface or network state, tor should take care of that. - - - - - fd4486c3 by Alex Catarineu at 2023-07-27T09:56:25+02:00 Add TorStrings module for localization - - - - - de3172d3 by Kathy Brade at 2023-07-27T09:56:25+02:00 Bug 14631: Improve profile access error messages. Instead of always reporting that the profile is locked, display specific messages for "access denied" and "read-only file system". To allow for localization, get profile-related error strings from Torbutton. Use app display name ("Tor Browser") in profile-related error alerts. - - - - - b18f8901 by Pier Angelo Vendrame at 2023-07-27T09:56:25+02:00 Bug 40807: Added QRCode.js to toolkit/modules - - - - - b4cf177e by Richard Pospesel at 2023-07-27T09:56:26+02:00 Bug 31286: Implementation of bridge, proxy, and firewall settings in about:preferences#connection This patch adds a new about:preferences#connection page which allows modifying bridge, proxy, and firewall settings from within Tor Browser. All of the functionality present in tor-launcher's Network Configuration panel is present: - Setting built-in bridges - Requesting bridges from BridgeDB via moat - Using user-provided bridges - Configuring SOCKS4, SOCKS5, and HTTP/HTTPS proxies - Setting firewall ports - Viewing and Copying Tor's logs - The Networking Settings in General preferences has been removed Bug 40774: Update about:preferences page to match new UI designs - - - - - d60d4929 by Richard Pospesel at 2023-07-27T09:56:26+02:00 Bug 27476: Implement about:torconnect captive portal within Tor Browser - implements new about:torconnect page as tor-launcher replacement - adds new torconnect component to browser - tor process management functionality remains implemented in tor-launcher through the TorProtocolService module - adds warning/error box to about:preferences#tor when not connected to tor Bug 40773: Update the about:torconnect frontend page to match additional UI flows. Bug 41608: Add a toolbar status button and a urlbar "Connect" button. - - - - - bf71bd68 by Kathleen Brade at 2023-07-27T09:56:27+02:00 Bug 7494: Create local home page for TBB. - - - - - 64597639 by Arthur Edelstein at 2023-07-27T09:56:27+02:00 Bug 12620: TorBrowser regression tests Regression tests for Bug #2950: Make Permissions Manager memory-only Regression tests for TB4: Tor Browser's Firefox preference overrides. Note: many more functional tests could be made here Regression tests for #2874: Block Components.interfaces from content Bug 18923: Add a script to run all Tor Browser specific tests Regression tests for Bug #16441: Suppress "Reset Tor Browser" prompt. - - - - - 1b0c70f5 by Pier Angelo Vendrame at 2023-07-27T09:56:28+02:00 Bug 41668: Tweaks to the Base Browser updater for Tor Browser This commit was once part of "Bug 4234: Use the Firefox Update Process for Tor Browser.". However, some parts of it were not needed for Base Browser and some derivative browsers. Therefore, we extracted from that commit the parts for Tor Browser legacy, and we add them back to the patch set with this commit. - - - - - c0fadea5 by Kathy Brade at 2023-07-27T09:56:28+02:00 Bug 12647: Support symlinks in the updater. - - - - - a04d84b0 by Kathy Brade at 2023-07-27T09:56:29+02:00 Bug 19121: reinstate the update.xml hash check This is a partial revert of commit f1241db6986e4b54473a1ed870f7584c75d51122. Revert most changes from Mozilla Bug 862173 "don't verify mar file hash when using mar signing to verify the mar file (lessens main thread I/O)." We kept the addition to the AppConstants API in case other JS code references it in the future. - - - - - 1be01b50 by Kathy Brade at 2023-07-27T09:56:29+02:00 Bug 16940: After update, load local change notes. Add an about:tbupdate page that displays the first section from TorBrowser/Docs/ChangeLog.txt and includes a link to the remote post-update page (typically our blog entry for the release). Always load about:tbupdate in a content process, but implement the code that reads the file system (changelog) in the chrome process for compatibility with future sandboxing efforts. Also fix bug 29440. Now about:tbupdate is styled as a fairly simple changelog page that is designed to be displayed via a link that is on about:tor. - - - - - 767b3431 by Georg Koppen at 2023-07-27T09:56:29+02:00 Bug 32658: Create a new MAR signing key It's time for our rotation again: Move the backup key in the front position and add a new backup key. Bug 33803: Move our primary nightly MAR signing key to tor-browser Bug 33803: Add a secondary nightly MAR signing key - - - - - 933b4bfe by Mike Perry at 2023-07-27T09:56:30+02:00 Omnibox: Add DDG, Startpage, Disconnect, Youtube, Twitter; remove Amazon, eBay, bing eBay and Amazon don't treat Tor users very well. Accounts often get locked and payments reversed. Also: Bug 16322: Update DuckDuckGo search engine We are replacing the clearnet URL with an onion service one (thanks to a patch by a cypherpunk) and are removing the duplicated DDG search engine. Duplicating DDG happend due to bug 1061736 where Mozilla included DDG itself into Firefox. Interestingly, this caused breaking the DDG search if JavaScript is disabled as the Mozilla engine, which gets loaded earlier, does not use the html version of the search page. Moreover, the Mozilla engine tracked where the users were searching from by adding a respective parameter to the search query. We got rid of that feature as well. Also: This fixes bug 20809: the DuckDuckGo team has changed its server-side code in a way that lets users with JavaScript enabled use the default landing page while those without JavaScript available get redirected directly to the non-JS page. We adapt the search engine URLs accordingly. Also fixes bug 29798 by making sure we only specify the Google search engine we actually ship an .xml file for. Also regression tests. squash! Omnibox: Add DDG, Startpage, Disconnect, Youtube, Twitter; remove Amazon, eBay, bing Bug 40494: Update Startpage search provider squash! Omnibox: Add DDG, Startpage, Disconnect, Youtube, Twitter; remove Amazon, eBay, bing Bug 40438: Add Blockchair as a search engine Bug 33342: Avoid disconnect search addon error after removal. We removed the addon in #32767, but it was still being loaded from addonStartup.json.lz4 and throwing an error on startup because its resource: location is not available anymore. - - - - - 8167258f by Alex Catarineu at 2023-07-27T09:56:34+02:00 Bug 40073: Disable remote Public Suffix List fetching In https://bugzilla.mozilla.org/show_bug.cgi?id=1563246 Firefox implemented fetching the Public Suffix List via RemoteSettings and replacing the default one at runtime, which we do not want. - - - - - b234e10c by Richard Pospesel at 2023-07-27T09:56:35+02:00 Bug 23247: Communicating security expectations for .onion Encrypting pages hosted on Onion Services with SSL/TLS is redundant (in terms of hiding content) as all traffic within the Tor network is already fully encrypted. Therefore, serving HTTP pages from an Onion Service is more or less fine. Prior to this patch, Tor Browser would mostly treat pages delivered via Onion Services as well as pages delivered in the ordinary fashion over the internet in the same way. This created some inconsistencies in behaviour and misinformation presented to the user relating to the security of pages delivered via Onion Services: - HTTP Onion Service pages did not have any 'lock' icon indicating the site was secure - HTTP Onion Service pages would be marked as unencrypted in the Page Info screen - Mixed-mode content restrictions did not apply to HTTP Onion Service pages embedding Non-Onion HTTP content This patch fixes the above issues, and also adds several new 'Onion' icons to the mix to indicate all of the various permutations of Onion Services hosted HTTP or HTTPS pages with HTTP or HTTPS content. Strings for Onion Service Page Info page are pulled from Torbutton's localization strings. - - - - - 3002d6cb by Kathy Brade at 2023-07-27T09:56:35+02:00 Bug 30237: Add v3 onion services client authentication prompt When Tor informs the browser that client authentication is needed, temporarily load about:blank instead of about:neterror and prompt for the user's key. If a correctly formatted key is entered, use Tor's ONION_CLIENT_AUTH_ADD control port command to add the key (via Torbutton's control port module) and reload the page. If the user cancels the prompt, display the standard about:neterror "Unable to connect" page. This requires a small change to browser/actors/NetErrorChild.jsm to account for the fact that the docShell no longer has the failedChannel information. The failedChannel is used to extract TLS-related error info, which is not applicable in the case of a canceled .onion authentication prompt. Add a leaveOpen option to PopupNotifications.show so we can display error messages within the popup notification doorhanger without closing the prompt. Add support for onion services strings to the TorStrings module. Add support for Tor extended SOCKS errors (Tor proposal 304) to the socket transport and SOCKS layers. Improved display of all of these errors will be implemented as part of bug 30025. Also fixes bug 19757: Add a "Remember this key" checkbox to the client auth prompt. Add an "Onion Services Authentication" section within the about:preferences "Privacy & Security section" to allow viewing and removal of v3 onion client auth keys that have been stored on disk. Also fixes bug 19251: use enhanced error pages for onion service errors. - - - - - 1e3d7af6 by Alex Catarineu at 2023-07-27T09:56:36+02:00 Bug 21952: Implement Onion-Location Whenever a valid Onion-Location HTTP header (or corresponding HTML <meta> http-equiv attribute) is found in a document load, we either redirect to it (if the user opted-in via preference) or notify the presence of an onionsite alternative with a badge in the urlbar. - - - - - 1f387301 by Pier Angelo Vendrame at 2023-07-27T09:56:36+02:00 Bug 40458: Implement .tor.onion aliases We have enabled HTTPS-Only mode, therefore we do not need HTTPS-Everywhere anymore. However, we want to keep supporting .tor.onion aliases (especially for securedrop). Therefore, in this patch we implemented the parsing of HTTPS-Everywhere rulesets, and the redirect of .tor.onion domains. Actually, Tor Browser believes they are actual domains. We change them on the fly on the SOCKS proxy requests to resolve the domain, and on the code that verifies HTTPS certificates. - - - - - 6a83fe48 by Pier Angelo Vendrame at 2023-07-27T09:56:37+02:00 Bug 11698: Incorporate Tor Browser Manual pages into Tor Browser This patch associates the about:manual page to a translated page that must be injected to browser/omni.ja after the build. The content must be placed in chrome/browser/content/browser/manual/, so that is then available at chrome://browser/content/manual/. We preferred giving absolute freedom to the web team, rather than having to change the patch in case of changes on the documentation. - - - - - 25fa2aec by Pier Angelo Vendrame at 2023-07-27T09:56:37+02:00 Bug 41435: Add a Tor Browser migration function For now this function only deletes old language packs for which we are already packaging the strings with the application. - - - - - 450b7e8e by Dan Ballard at 2023-07-27T09:56:37+02:00 Bug 40701: Add security warning when downloading a file Shown in the downloads panel, about:downloads and places.xhtml. - - - - - 03da74d0 by Henry Wilkes at 2023-07-27T09:56:38+02:00 Bug 41736: Customize toolbar for tor-browser. - - - - - bdda467f by hackademix at 2023-07-27T09:56:38+02:00 Bug 41728: Pin bridges.torproject.org domains to Let's Encrypt's root cert public key - - - - - 30 changed files: - .eslintignore - .gitignore - + .gitlab/issue_templates/Backport Android Security Fixes.md - + .gitlab/issue_templates/Rebase Browser - Alpha.md - + .gitlab/issue_templates/Rebase Browser - Stable.md - + .gitlab/issue_templates/UXBug.md - + .gitlab/issue_templates/bug.md - + .gitlab/merge_request_templates/default.md - + browser/actors/AboutTBUpdateChild.jsm - + browser/actors/AboutTBUpdateParent.jsm - + browser/actors/CryptoSafetyChild.jsm - + browser/actors/CryptoSafetyParent.jsm - − browser/actors/RFPHelperChild.sys.mjs - − browser/actors/RFPHelperParent.sys.mjs - browser/actors/moz.build - browser/app/Makefile.in - browser/app/macbuild/Contents/Info.plist.in - browser/app/macbuild/Contents/MacOS-files.in - browser/app/permissions - + browser/app/profile/000-tor-browser.js - + browser/app/profile/001-base-profile.js - browser/app/profile/firefox.js - browser/base/content/aboutDialog-appUpdater.js - browser/base/content/aboutDialog.js - browser/base/content/aboutDialog.xhtml - + browser/base/content/aboutDialogTor.css - + browser/base/content/abouttbupdate/aboutTBUpdate.css - + browser/base/content/abouttbupdate/aboutTBUpdate.js - + browser/base/content/abouttbupdate/aboutTBUpdate.xhtml - browser/base/content/appmenu-viewcache.inc.xhtml The diff was not included because it is too large. View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/58bb97… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/58bb97… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser-build][main] 2 commits: Move the macOS lld patch to llvm-project.
by Pier Angelo Vendrame (＠pierov) 27 Jul '23

27 Jul '23

Pier Angelo Vendrame pushed to branch main at The Tor Project / Applications / tor-browser-build Commits: a812dd07 by Pier Angelo Vendrame at 2023-07-26T18:04:01+02:00 Move the macOS lld patch to llvm-project. To solve a macOS reproducibility project, we backported a patch from LLVM, but we applied it only to clang and only to macOS to prevent a rebuild for all the other platforms. However, a new rebuild is now needed, so we can apply the patch to all the platforms and all the project using the LLVM source code for consistency, even though the previous solution was completely fine. - - - - - 98860cf9 by Pier Angelo Vendrame at 2023-07-26T18:10:51+02:00 Bug 40907: Backport the fix to LLVM #62546 We encountered the same problem as llvm-project#62546: we did not have complete reproducibility on debug symbols. So, we need to backport the patch that fixes the problem upstream. - - - - - 6 changed files: - projects/clang/build - projects/clang/config - projects/llvm-project/build - projects/llvm-project/config - + projects/llvm-project/partially-revert-llvm-less_second.patch - projects/clang/use-fixed-chunk-size-for-uuid.patch → projects/llvm-project/use-fixed-chunk-size-for-uuid.patch Changes: ===================================== projects/clang/build ===================================== @@ -16,12 +16,6 @@ mkdir -p /var/tmp/build cd /var/tmp/build tar -xf $rootdir/[% c('input_files_by_name/clang-source') %] cd clang-source -[% IF c("var/macos") -%] - # Remove once we update to a version that includes commit - # 359f170f5f712ee714193b46bad45a45656b2c59 (> 16.0.6) - # https://github.com/llvm/llvm-project/issues/63961 - patch -p1 < $rootdir/use-fixed-chunk-size-for-uuid.patch -[% END -%] export LLVM_HOME=$(pwd) mkdir build cd build ===================================== projects/clang/config ===================================== @@ -24,5 +24,3 @@ input_files: - project: python name: python enable: '[% c("var/linux") %]' - - filename: use-fixed-chunk-size-for-uuid.patch - enable: '[% c("var/macos") %]' ===================================== projects/llvm-project/build ===================================== @@ -2,6 +2,18 @@ [% c("var/set_default_env") -%] tar -xf [% project %]-[% c("version") %].tar.[% c('compress_tar') %] mv [% project %]-[% c("version") %] clang-source + +pushd clang-source +# Remove once we update to a version that includes commit +# 359f170f5f712ee714193b46bad45a45656b2c59 (> 16.0.6) +# https://github.com/llvm/llvm-project/issues/63961 +patch -p1 < $rootdir/use-fixed-chunk-size-for-uuid.patch +# Remove once we update to a version that includes commit +# 32ab0978dc3f7f7036df2038ee96a4ab89196255 +# https://github.com/llvm/llvm-project/issues/62546 +patch -p1 < $rootdir/partially-revert-llvm-less_second.patch +popd + [% c('tar', { tar_src => [ 'clang-source' ], tar_args => '-caf ' _ dest_dir _ '/' _ c('filename'), ===================================== projects/llvm-project/config ===================================== @@ -6,3 +6,7 @@ filename: '[% project %]-[% c("version") %]-[% c("var/build_id") %].tar.[% c("co var: llvm_major: 16 + +input_files: + - filename: use-fixed-chunk-size-for-uuid.patch + - filename: partially-revert-llvm-less_second.patch ===================================== projects/llvm-project/partially-revert-llvm-less_second.patch ===================================== @@ -0,0 +1,29 @@ +From 32ab0978dc3f7f7036df2038ee96a4ab89196255 Mon Sep 17 00:00:00 2001 +From: Kazu Hirata <kazu(a)google.com> +Date: Tue, 16 May 2023 14:49:32 -0700 +Subject: [PATCH] Partially revert "Use llvm::less_second (NFC)" + +This reverts part of commit e0039b8d6a5bd05e70203962f448569f2d2ef1c2. + +This should fix the issue reported in: + +https://github.com/llvm/llvm-project/issues/62546 +--- + llvm/lib/IR/Metadata.cpp | 4 +++- + 1 file changed, 3 insertions(+), 1 deletion(-) + +diff --git a/llvm/lib/IR/Metadata.cpp b/llvm/lib/IR/Metadata.cpp +index 13fb2b650f89e9..face1ba58cfbbd 100644 +--- a/llvm/lib/IR/Metadata.cpp ++++ b/llvm/lib/IR/Metadata.cpp +@@ -282,7 +282,9 @@ void ReplaceableMetadataImpl::replaceAllUsesWith(Metadata *MD) { + // Copy out uses since UseMap will get touched below. + using UseTy = std::pair<void *, std::pair<OwnerTy, uint64_t>>; + SmallVector<UseTy, 8> Uses(UseMap.begin(), UseMap.end()); +- llvm::sort(Uses, llvm::less_second()); ++ llvm::sort(Uses, [](const UseTy &L, const UseTy &R) { ++ return L.second.second < R.second.second; ++ }); + for (const auto &Pair : Uses) { + // Check that this Ref hasn't disappeared after RAUW (when updating a + // previous Ref). ===================================== projects/clang/use-fixed-chunk-size-for-uuid.patch → projects/llvm-project/use-fixed-chunk-size-for-uuid.patch ===================================== View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/compare/… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/compare/… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser] Pushed new tag base-browser-102.14.0esr-12.5-1-build1
by ma1 (＠ma1) 26 Jul '23

26 Jul '23

ma1 pushed new tag base-browser-102.14.0esr-12.5-1-build1 at The Tor Project / Applications / Tor Browser -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/tree/base-brow… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser] Pushed new tag tor-browser-102.14.0esr-12.5-1-build1
by ma1 (＠ma1) 26 Jul '23

26 Jul '23

ma1 pushed new tag tor-browser-102.14.0esr-12.5-1-build1 at The Tor Project / Applications / Tor Browser -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/tree/tor-brows… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][base-browser-102.14.0esr-12.5-1] 97 commits: Bug 1774413: Bundled fonts are not picked up on macOS r=jfkthame
by ma1 (＠ma1) 26 Jul '23

26 Jul '23

ma1 pushed to branch base-browser-102.14.0esr-12.5-1 at The Tor Project / Applications / Tor Browser Commits: 66650a2e by Pier Angelo Vendrame at 2023-07-25T23:31:49+02:00 Bug 1774413: Bundled fonts are not picked up on macOS r=jfkthame Bundled fonts were not picked up because also child processes need to register them. Also, they were assigned User visibility, instead of Base, which was not coherent with other platforms. Differential Revision: https://phabricator.services.mozilla.com/D150400 - - - - - 891ce49f by Pier Angelo Vendrame at 2023-07-25T23:31:56+02:00 Bug 1769030: Add a configure flag to load policies only from the local policies.json file r=mkaply,glandium Add a configuration flag to make Enterprise Policies mechanism only consult a policies.json file (avoiding the Windows Registry, macOS's file system attributes, and /etc/firefox/policies/policies.json on other OS). Differential Revision: https://phabricator.services.mozilla.com/D146300 - - - - - 0167b97a by Jamie Nicol at 2023-07-25T23:31:57+02:00 Bug 1784588 - Fix SplitDriverVersion when the string has less than 4 numbers. r=aosmond Differential Revision: https://phabricator.services.mozilla.com/D154897 - - - - - b1829371 by Pier Angelo Vendrame at 2023-07-25T23:31:58+02:00 Bug 1768907 - Part 1: Make browser.privatebrowsing.autostart a static pref. r=handyman,necko-reviewers Differential Revision: https://phabricator.services.mozilla.com/D157843 - - - - - 1125e718 by Neill Miller at 2023-07-25T23:31:59+02:00 Bug 1768907 - Part 2: nsTransferable disk-leaks when private browsing mode is enabled by default. r=handyman nsTransferable was modified to prevent disk leakings when copying data in private browsing mode with Bug 1123480. However, the context is nullptr when it is initialized, so it still leaks if PBM is enabled by default. Our solution is to check the browser.privatebrowsing.autostart in this condition. Differential Revision: https://phabricator.services.mozilla.com/D157800 - - - - - e3afc014 by Gabriele Svelto at 2023-07-25T23:32:00+02:00 Bug 1670885 - Fix post-fork() handlers for PHC/LogAlloc to work on macOS using unfair locks r=glandium macOS unfair locks enforce that a lock can only be released by the thread which locked it. On macOS 11+ this caused the fork()'d child process to raise a SIGILL signal. Confusingly enough this behavior seems to be different on macOS 10.15 and possibly interacted in odd ways with our exception handler if it was installed before fork()-ing. Differential Revision: https://phabricator.services.mozilla.com/D148287 - - - - - f4cf7ede by Gabriele Svelto at 2023-07-25T23:32:01+02:00 Bug 1670885 - Replace deprecated NSSpinLocks with os_unfair_locks in the memory allocator r=glandium Depends on D148287 Differential Revision: https://phabricator.services.mozilla.com/D99280 - - - - - 3d94ece4 by Kershaw Chang at 2023-07-25T23:32:02+02:00 Bug 1685403 - Use NS_GetFinalChannelURI at some places, r=necko-reviewers,valentin Differential Revision: https://phabricator.services.mozilla.com/D164656 - - - - - 24236471 by Kathy Brade at 2023-07-25T23:32:02+02:00 Bug 1769028: Disable DNS lookup in nsProfileLock when the proxy bypass protection is enabled r=mossop Instead of using the local computer's IP address within symlink-based profile lock signatures, always use 127.0.0.1 when the proxy bypass protection is enabled. Differential Revision: https://phabricator.services.mozilla.com/D161264 - - - - - c3235edc by Gabriele Svelto at 2023-07-25T23:32:03+02:00 Bug 1774458 - Use undocumented, non-public adaptive spinlocks on macOS 10.15+, revert to user-space spinlocks on older versions r=pbone Differential Revision: https://phabricator.services.mozilla.com/D149599 - - - - - b22ed534 by Gabriele Svelto at 2023-07-25T23:32:04+02:00 Bug 1784018 - Remove deprecated OSSpinLocks r=glandium On macOS versions prior to 10.15 os_unfair_locks cannot spin in kernel-space which degrades performance significantly. To obviate for this we spin in user-space like OSSpinLock does, for the same number of times and invoking x86-specific pause instructions in-between the locking attempts to avoid starving a thread that might be running on the same physical core. Differential Revision: https://phabricator.services.mozilla.com/D154205 - - - - - e5e87caa by Hubert Boma Manilla at 2023-07-25T23:32:05+02:00 Bug 1790542 - [devtools] Do not allow redirects on the 'sourceMappingUrl' r=jdescottes,devtools-reviewers - Block redirects on sourceMappingUrl - Add debugger test Differential Revision: https://phabricator.services.mozilla.com/D168648 - - - - - 9d39b9d6 by Hubert Boma Manilla at 2023-07-25T23:32:06+02:00 Bug 1790542 - [devtools] Convert browser_source_map-reload.js test from using .sjs files to createTestHTTPServer r=ochameau,devtools-reviewers .sjs files depend on redirects to load static files. In D168648 we block redirects on sourceMappingUrl, so this browser_source_map-reload.js fails because we use .sjs to serve the content. This patch refactors to no longer use .sjs files Differential Revision: https://phabricator.services.mozilla.com/D169739 - - - - - c02ec6c6 by Jonathan Kew at 2023-07-25T23:32:07+02:00 Bug 1792157 - Ensure correct font-visibility for bundled fonts on Linux when they 'shadow' system-installed families. r=gfx-reviewers,lsalzman Differential Revision: https://phabricator.services.mozilla.com/D175350 - - - - - a72fbee6 by Rob Wu at 2023-07-25T23:32:08+02:00 Bug 1802385 - Use NS_GetFinalChannelURI in FetchDriver r=rpl,valentin Depends on D164656 Differential Revision: https://phabricator.services.mozilla.com/D166108 - - - - - a424f1cf by Kash Shampur at 2023-07-25T23:32:09+02:00 Bug 1803109 - Discard blocks of data that are too big for two chunks. r=canaltinova Currently, `ReserveAndPutRaw` allocates a second span even if the data would be too big for the chunk. Here a second conditional is added to check if the block of data is too big in this scenario and silently discard the data if so. Differential Revision: https://phabricator.services.mozilla.com/D167167 - - - - - de73c560 by Pier Angelo Vendrame at 2023-07-25T23:32:09+02:00 Bug 1814416: Generalize the app name in about:buildconfig. r=ahochheiden Now it displays app display name, instead of having Firefox hardcoded. Differential Revision: https://phabricator.services.mozilla.com/D168562 - - - - - b74c0791 by Jon Coppeard at 2023-07-25T23:32:10+02:00 Bug 1816158 - Part 1: Disallow GC while iterating global's debugger vector r=sfink GC can mutate this vector so don't allow that while we are iterating. I think it would be safe to use index-based iteration but it's safer to just ban it entirely. This fixes the crash produced by the testcase. Differential Revision: https://phabricator.services.mozilla.com/D169701 - - - - - d6c35ccf by Jon Coppeard at 2023-07-25T23:32:11+02:00 Bug 1816158 - Part 2: Require no GC when giving out references to the realm's debugger vector r=sfink To prevent any other instances of this problme we can update the getDebuggers() methods on the global and the realm to require no GC. Depends on D169701 Differential Revision: https://phabricator.services.mozilla.com/D169702 - - - - - 4651288e by Pier Angelo Vendrame at 2023-07-25T23:32:12+02:00 Bug 1817756 - Add a seed to the network ID. r=valentin,necko-reviewers This helps to prevent linkability of users in the same network. Differential Revision: https://phabricator.services.mozilla.com/D170373 - - - - - 72140b84 by John Lin at 2023-07-25T23:42:26+02:00 Bug 1832832 - protect encoder accesses in Java callbacks. r=media-playback-reviewers,alwu a=pascalc Differential Revision: https://phabricator.services.mozilla.com/D178564 - - - - - cffad159 by Arsen Arsenović at 2023-07-25T23:42:27+02:00 Bug 1811714 - Add a few missing <cstdint> includes r=gfx-reviewers,nical GCC 13s libstdc++ reduced its dependency on some headers like <cstdint>, so it's no longer transitively included through various headers. Include it explicitly. See also: https://gcc.gnu.org/gcc-13/porting_to.html#header-dep-changes 11:16.32 .../components/telemetry/pingsender/pingsender.cpp:30:7: error: unknown type name 'uint32_t' 11:16.33 const uint32_t kConnectionTimeoutMs = 30 * 1000; 11:16.33 ^ 11:16.33 ... /components/telemetry/pingsender/pingsender.cpp:76:9: error: unknown type name 'uint32_t' 11:16.33 const uint32_t kBufferSize = 4 * 1024 - 1; 11:16.34 ^ 11:16.44 2 errors generated. (et al) Differential Revision: https://phabricator.services.mozilla.com/D167491 - - - - - 2f4a54a3 by James Teh at 2023-07-26T21:52:19+02:00 Bug 1772170: Ensure cached Accessibles and offsets are valid before returning them in DocAccessibleParent::SelectionRanges. r=eeejay Differential Revision: https://phabricator.services.mozilla.com/D148745 - - - - - e460aee8 by James Teh at 2023-07-26T21:52:20+02:00 Bug 1781169: Ensure that the cached start/end containers contain characters before returning them in DocAccessibleParent::SelectionRanges. r=eeejay In bug 1772170, I added code to ensure that the cached offsets are valid. However, this didn't account for the possibility that one of the containers contained 0 characters. In that case, offset 0 is itself invalid. If there were also no children, this might in GetChildAtOffset being called and returning null, causing a crash when retrieving selection. Differential Revision: https://phabricator.services.mozilla.com/D154964 - - - - - 7f4de3a3 by Georg Koppen at 2023-07-26T21:52:21+02:00 Revert "Bug 1724777, optimize suppressed MicroTask handling, r=mccr8 a=RyanVM" This reverts commit 1eb1364357ac5bc2a4531337fb5416af39c3793f. This fixes tor-browser#40721, tor-browser#40698, and tor-browser#40706. However, it is a temporary workaround, that we should revert once https://bugzilla.mozilla.org/show_bug.cgi?id=1744719 is fixed. - - - - - 5388480a by Henry Wilkes at 2023-07-26T21:52:22+02:00 Bug 41454: Move focus after calling openPreferences for a sub-category. Temporary fix until mozilla bug 1799153 gets a patch upstream. - - - - - 113974c0 by Henry Wilkes at 2023-07-26T21:52:23+02:00 Bug 41483: Remove the firefox override for appstrings.properties Remove this patch after upstream bugzilla bug 1790187 - - - - - bab3b502 by Pier Angelo Vendrame at 2023-07-26T21:52:24+02:00 Bug 41116: Normalize system fonts. System fonts are an enormous fingerprinting vector. Even with font allow lists and with our custom configuration on Linux, which counter metrics measurements, getComputedStyle leaks several details. This patch counters both these kinds of attacks. - - - - - 7bf81d37 by Marco Simonelli at 2023-07-26T21:52:25+02:00 Bug 41459: WebRTC fails to build under mingw (Part 1) - properly define NOMINMAX for just MSVC builds - - - - - 28aeacd5 by Marco Simonelli at 2023-07-26T21:52:26+02:00 Bug 41459: WebRTC fails to build under mingw (Part 2) - fixes required to build third_party/libwebrtc - - - - - 17fd7119 by Marco Simonelli at 2023-07-26T21:52:26+02:00 Bug 41459: WebRTC fails to build under mingw (Part 3) - fixes required to build third_party/sipcc - - - - - 49244d4b by Marco Simonelli at 2023-07-26T21:52:27+02:00 Bug 41459: WebRTC fails to build under mingw (Part 4) - fixes requried to build netwerk/sctp - - - - - 516a096f by Marco Simonelli at 2023-07-26T21:52:28+02:00 Bug 41459: WebRTC fails to build under mingw (Part 5) - fixes required to build dom/media/webrtc - - - - - e08f2fc3 by Marco Simonelli at 2023-07-26T21:52:29+02:00 Bug 41459: WebRTC fails to build under mingw (Part 6) - fixes required to build dom/media/systemservices - - - - - 447b1ded by Gaba at 2023-07-26T21:52:30+02:00 Adding issue and merge request templates - - - - - 91460f85 by Pier Angelo Vendrame at 2023-07-26T21:52:31+02:00 Base Browser's .mozconfigs. Bug 17858: Cannot create incremental MARs for hardened builds. Define HOST_CFLAGS, etc. to avoid compiling programs such as mbsdiff (which is part of mar-tools and is not distributed to end-users) with ASan. Bug 21849: Don't allow SSL key logging. Bug 25741 - TBA: Disable features at compile-time Define MOZ_ANDROID_NETWORK_STATE and MOZ_ANDROID_LOCATION Bug 27623 - Export MOZILLA_OFFICIAL during desktop builds This fixes a problem where some preferences had the wrong default value. Also see bug 27472 where we made a similar fix for Android. Bug 29859: Disable HLS support for now Bug 30463: Explicitly disable MOZ_TELEMETRY_REPORTING Bug 32493: Disable MOZ_SERVICES_HEALTHREPORT Bug 33734: Set MOZ_NORMANDY to False Bug 33851: Omit Parental Controls. Bug 40252: Add --enable-rust-simd to our tor-browser mozconfig files Bug 41584: Move some configuration options to base-browser level - - - - - 8fcbe886 by Pier Angelo Vendrame at 2023-07-26T21:52:32+02:00 Tweaks to the build system Bug 40857: Modified the fat .aar creation file This is a workaround to build fat .aars with the compiling enviornment disabled. Mozilla does not use a similar configuration, but either runs a Firefox build and discards its output, or uses artifacts build. We might switch to artifact builds too, and drop this patch, or write a better one to upstream. But until then we need this patch. See also https://bugzilla.mozilla.org/show_bug.cgi?id=1763770. Bug 41458: Prevent `mach package-multi-locale` from actually creating a package macOS builds need some files to be moved around with ./mach package-multi-locale to create multi-locale packages. The required command isn't exposed through any other mach command. So, we patch package-multi-locale both to prevent it from failing when doing official builds and to detect any future changes on it. - - - - - 5b2779a5 by Dan Ballard at 2023-07-26T21:52:32+02:00 Bug 41117: using MOZ_PROXY_BYPASS_PROTECTION disable share menu on mac and windows, and restore email link on mac - - - - - 5a3057bf by Pier Angelo Vendrame at 2023-07-26T21:52:33+02:00 Bug 41108: Remove privileged macOS installation from 102 - - - - - c78741f2 by Dan Ballard at 2023-07-26T21:52:34+02:00 Bug 41149: Re-enable DLL injection protection in all builds not just nightlies - - - - - b4a4bec8 by Matthew Finkel at 2023-07-26T21:52:35+02:00 Bug 24796: Comment out excess permissions from GeckoView The GeckoView AndroidManifest.xml is not preprocessed unlike Fennec's manifest, so we can't use the ifdef preprocessor guards around the permissions we do not want. Commenting the permissions is the next-best-thing. - - - - - 81122dd6 by Matthew Finkel at 2023-07-26T21:52:36+02:00 Bug 28125: Prevent non-Necko network connections - - - - - 617b19d0 by Mike Perry at 2023-07-26T21:52:37+02:00 Bug 12974: Disable NTLM and Negotiate HTTP Auth The Mozilla bugs: https://bugzilla.mozilla.org/show_bug.cgi?id=1046421, https://bugzilla.mozilla.org/show_bug.cgi?id=1261591, tor-browser#27602 - - - - - 5107f970 by Alex Catarineu at 2023-07-26T21:52:38+02:00 Bug 40166: Disable security.certerrors.mitm.auto_enable_enterprise_roots Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1768899 - - - - - 5ccd2869 by Georg Koppen at 2023-07-26T21:52:39+02:00 Bug 16285: Exclude ClearKey system for now In the past the ClearKey system had not been compiled when specifying --disable-eme. But that changed and it is even bundled nowadays (see: Mozilla's bug 1300654). We don't want to ship it right now as the use case for it is not really visible while the code had security vulnerabilities in the past. - - - - - 972b550c by Kathy Brade at 2023-07-26T21:52:39+02:00 Bug 21431: Clean-up system extensions shipped in Firefox Only ship the pdfjs extension. - - - - - 9cfa53dc by Kathy Brade at 2023-07-26T21:52:40+02:00 Bug 33852: Clean up about:logins (LockWise) to avoid mentioning sync, etc. Hide elements on about:logins that mention sync, "Firefox LockWise", and Mozilla's LockWise mobile apps. Disable the "Create New Login" button when security.nocertdb is true. - - - - - 4afff71e by Alex Catarineu at 2023-07-26T21:52:41+02:00 Bug 41457: Remove Mozilla permissions Bug 40025: Remove Mozilla add-on install permissions - - - - - f3231564 by Kathy Brade at 2023-07-26T21:52:42+02:00 Bug 40002: Remove about:ion Firefox Ion (previously Firefox Pioneer) is an opt-in program in which people volunteer to participate in studies that collect detailed, sensitive data about how they use their browser. Bug 41662: Disable about:sync-logs Even though we disable sync by default with `identity.fxaccounts.enabled`, this about: page is still avilable. We could throw an exception on the constructor of the related component, but it would result only in an error in the console, without a visible "this address does not look right" error page. If we fix the issues with MOZ_SERVICES_SYNC, we can restore the component. - - - - - 42f24da1 by Arthur Edelstein at 2023-07-26T21:52:43+02:00 Bug 26353: Prevent speculative connect that violated FPI. Connections were observed in the catch-all circuit when the user entered an https or http URL in the URL bar, or typed a search term. - - - - - 727a1d8b by Alex Catarineu at 2023-07-26T21:52:44+02:00 Bug 31740: Remove some unnecessary RemoteSettings instances More concretely, SearchService.jsm 'hijack-blocklists' and url-classifier-skip-urls. Avoid creating instance for 'anti-tracking-url-decoration'. If prefs are disabling their usage, avoid creating instances for 'cert-revocations' and 'intermediates'. Do not ship JSON dumps for collections we do not expect to need. For the ones in the 'main' bucket, this prevents them from being synced unnecessarily (the code in remote-settings does so for collections in the main bucket for which a dump or local data exists). For the collections in the other buckets, we just save some size by not shipping their dumps. We also clear the collections database on the v2 -> v3 migration. - - - - - e5bc5b12 by Pier Angelo Vendrame at 2023-07-26T21:52:45+02:00 Bug 41635: Disable the Normandy component Do not include Normandy at all whenever MOZ_NORMANDY is False. - - - - - dfc706b5 by Georg Koppen at 2023-07-26T21:52:46+02:00 Bug 30541: Disable WebGL readPixel() for web content Related Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1428034 - - - - - fae932dd by Alex Catarineu at 2023-07-26T21:52:46+02:00 Bug 28369: Stop shipping pingsender executable - - - - - 5d5d652e by cypherpunks1 at 2023-07-26T21:52:47+02:00 Bug 41565: Gate Telemetry Tasks behind MOZ_TELEMETRY_REPORTING - - - - - 35cb57e6 by cypherpunks1 at 2023-07-26T21:52:48+02:00 Bug 41568: Disable LaterRun - - - - - 75bb1ef6 by cypherpunks1 at 2023-07-26T21:52:49+02:00 Bug 40717: Hide Windows SSO in settings - - - - - 683e266e by Pier Angelo Vendrame at 2023-07-26T21:52:50+02:00 Bug 40788: Prevent nsIURLQueryStrippingListService from calling home when it is not enabled. The URL query stripping service is enabled only in nightly builds, still it is initialized and remote settings are downloaded. This adds a condition that prevents the service from being initialized if disabled. Upstream Bug: https://bugzilla.mozilla.org/show_bug.cgi?id=1812594 We should remove this patch if Mozilla decides that this is a problem also for them, or if they do not but we decide to use the feature. - - - - - c10ca88e by Pier Angelo Vendrame at 2023-07-26T21:52:51+02:00 Bug 41599: Always return an empty string as network ID Firefox computes an internal network ID used to detect network changes and act consequently (e.g., to improve WebSocket UX). However, there are a few ways to get this internal network ID, so we patch them out, to be sure any new code will not be able to use them and possibly link users. We also sent a patch to Mozilla to seed the internal network ID, to prevent any accidental leak in the future. Upstream: https://bugzilla.mozilla.org/show_bug.cgi?id=1817756 - - - - - 08d32d39 by Mike Perry at 2023-07-26T21:52:52+02:00 Firefox preference overrides. This hack directly includes our preference changes in omni.ja. Bug 18292: Staged updates fail on Windows Temporarily disable staged updates on Windows. Bug 18297: Use separate Noto JP,KR,SC,TC fonts Bug 23404: Add Noto Sans Buginese to the macOS whitelist Bug 23745: Set dom.indexedDB.enabled = true Bug 13575: Disable randomised Firefox HTTP cache decay user tests. (Fernando Fernandez Mancera <ffmancera(a)riseup.net>) Bug 17252: Enable session identifiers with FPI Session tickets and session identifiers were isolated by OriginAttributes, so we can re-enable them by allowing the default value (true) of "security.ssl.disable_session_identifiers". The pref "security.enable_tls_session_tickets" is obsolete (removed in https://bugzilla.mozilla.org/917049) Bug 14952: Enable http/2 and AltSvc In Firefox, SPDY/HTTP2 now uses Origin Attributes for isolation of connections, push streams, origin frames, etc. That means we get first-party isolation provided "privacy.firstparty.isolate" is true. So in this patch, we stop overriding "network.http.spdy.enabled" and "network.http.spdy.enabled.http2". Alternate Services also use Origin Attributes for isolation. So we stop overriding "network.http.altsvc.enabled" and "network.http.altsvc.oe" as well. (All 4 of the abovementioned "network.http.*" prefs adopt Firefox 60ESR's default value of true.) However, we want to disable HTTP/2 push for now, so we set "network.http.spdy.allow-push" to false. "network.http.spdy.enabled.http2draft" was removed in Bug 1132357. "network.http.sped.enabled.v2" was removed in Bug 912550. "network.http.sped.enabled.v3" was removed in Bug 1097944. "network.http.sped.enabled.v3-1" was removed in Bug 1248197. Bug 26114: addons.mozilla.org is not special * Don't expose navigator.mozAddonManager on any site * Don't block NoScript from modifying addons.mozilla.org or other sites Enable ReaderView mode again (#27281). Bug 29916: Make sure enterprise policies are disabled Bug 2874: Block Components.interfaces from content Bug 26146: Spoof HTTP User-Agent header for desktop platforms In Tor Browser 8.0, the OS was revealed in both the HTTP User-Agent header and to JavaScript code via navigator.userAgent. To avoid leaking the OS inside each HTTP request (which many web servers log), always use the Windows 7 OS value in the desktop User-Agent header. We continue to allow access to the actual OS via JavaScript, since doing so improves compatibility with web applications such as GitHub and Google Docs. Bug 12885: Windows Jump Lists fail for Tor Browser Jumplist entries are stored in a binary file in: %APPDATA%\\Microsoft\Windows\Recent\CustomDestinations\ and has a name in the form [a-f0-9]+.customDestinations-ms The hex at the front is unique per app, and is ultimately derived from something called the 'App User Model ID' (AUMID) via some unknown hashing method. The AUMID is provided as a key when programmatically creating, updating, and deleting a jumplist. The default behaviour in firefox is for the installer to define an AUMID for an app, and save it in the registry so that the jumplist data can be removed by the uninstaller. However, the Tor Browser does not set this (or any other) regkey during installation, so this codepath fails and the app's AUMID is left undefined. As a result the app's AUMID ends up being defined by windows, but unknowable by Tor Browser. This unknown AUMID is used to create and modify the jumplist, but the delete API requires that we provide the app's AUMID explicitly. Since we don't know what the AUMID is (since the expected regkey where it is normally stored does not exist) jumplist deletion will fail and we will leave behind a mostly empty customDestinations-ms file. The name of the file is derived from the binary path, so an enterprising person could reverse engineer how that hex name is calculated, and generate the name for Tor Browser's default Desktop installation path to determine whether a person had used Tor Browser in the past. The 'taskbar.grouping.useprofile' option that is enabled by this patch works around this AUMID problem by having firefox.exe create it's own AUMID based on the profile path (rather than looking for a regkey). This way, if a user goes in and enables and disables jumplist entries, the backing store is properly deleted. Unfortunately, all windows users currently have this file lurking in the above mentioned directory and this patch will not remove it since it was created with an unknown AUMID. However, another patch could be written which goes to that directory and deletes any item containing the 'Tor Browser' string. See bug 28996. Bug 30845: Make sure default themes and other internal extensions are enabled Bug 28896: Enable extensions in private browsing by default Bug 31065: Explicitly allow proxying localhost Bug 31598: Enable letterboxing Disable Presentation API everywhere Bug 21549 - Use Firefox's WASM default pref. It is disabled at safer security levels. Bug 32321: Disable Mozilla's MitM pings Bug 19890: Disable installation of system addons By setting the URL to "" we make sure that already installed system addons get deleted as well. Bug 22548: Firefox downgrades VP9 videos to VP8. On systems where H.264 is not available or no HWA, VP9 is preferred. But in Tor Browser 7.0 all youtube videos are degraded to VP8. This behaviour can be turned off by setting media.benchmark.vp9.threshold to 0. All clients will get better experience and lower traffic, beause TBB doesn't use "Use hardware acceleration when available". Bug 25741 - TBA: Add mobile-override of 000-tor-browser prefs Bug 16441: Suppress "Reset Tor Browser" prompt. Bug 29120: Use the in-memory media cache and increase its maximum size. Bug 33697: use old search config based on list.json Bug 33855: Ensure that site-specific browser mode is disabled. Bug 30682: Disable Intermediate CA Preloading. Bug 40061: Omit the Windows default browser agent from the build Bug 40322: Consider disabling network.connectivity-service.enabled Bug 40408: Disallow SVG Context Paint in all web content Bug 40308: Disable network partitioning until we evaluate dFPI Bug 40322: Consider disabling network.connectivity-service.enabled Bug 40383: Disable dom.enable_event_timing Bug 40423: Disable http/3 Bug 40177: Update prefs for Fx91esr Bug 40700: Disable addons and features recommendations Bug 40682: Disable network.proxy.allow_bypass Bug 40736: Disable third-party cookies in PBM Bug 19850: Enabled HTTPS-Only by default Bug 40912: Hide the screenshot menu Bug 41292: Disable moreFromMozilla in preferences page Bug 40057: Ensure the CSS4 system colors are not a fingerprinting vector Bug 24686: Set network.http.tailing.enabled to true Bug 40183: Disable TLS ciphersuites using SHA-1 Bug 40783: Review 000-tor-browser.js and 001-base-profile.js for 102 We reviewed all the preferences we set for 102, and remove a few old ones. See the description of that issue to see all the preferences we believed were still valid for 102, and some brief description for the reasons to keep them. - - - - - b469f0bf by Richard Pospesel at 2023-07-26T21:52:53+02:00 Bug 41659: Add canonical color definitions to base-browser - - - - - 92e56693 by Pier Angelo Vendrame at 2023-07-26T21:52:53+02:00 Bug 41043: Hardcode the UI font on Linux The mechanism to choose the UI font does not play well with our fontconfig configuration. As a result, the final criterion to choose the font for the UI was its version. Since we hardcode Arimo as a default sans-serif on preferences, we use it also for the UI. FontConfig will fall back to some other font for scripts Arimo does not cover as expected (we tested with Japanese). - - - - - e4faff8e by Alex Catarineu at 2023-07-26T21:52:54+02:00 Bug 30605: Honor privacy.spoof_english in Android This checks `privacy.spoof_english` whenever `setLocales` is called from Fenix side and sets `intl.accept_languages` accordingly. Bug 40198: Expose privacy.spoof_english pref in GeckoView - - - - - 62ae3e48 by Alex Catarineu at 2023-07-26T21:52:55+02:00 Bug 40199: Avoid using system locale for intl.accept_languages in GeckoView - - - - - d8a6dad5 by Alex Catarineu at 2023-07-26T21:52:56+02:00 Bug 40171: Make WebRequest and GeckoWebExecutor First-Party aware - - - - - 5e7fdf93 by Alex Catarineu at 2023-07-26T21:52:57+02:00 Bug 26345: Hide tracking protection UI - - - - - bf0600b5 by Pier Angelo Vendrame at 2023-07-26T21:52:58+02:00 Bug 9173: Change the default Firefox profile directory to be relative. This commit makes Firefox look for the default profile directory in a directory relative to the binary path. The directory can be specified through the --with-relative-data-dir. This is relative to the same directory as the firefox main binary for Linux and Windows. On macOS, we remove Contents/MacOS from it. Or, in other words, the directory is relative to the application bundle. This behavior can be overriden at runtime, by placing a file called system-install adjacent to the firefox main binary (also on macOS). - - - - - f86be196 by Alex Catarineu at 2023-07-26T21:52:59+02:00 Bug 27604: Fix addon issues when moving the profile directory Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1429838 - - - - - 54a51fd9 by Mike Perry at 2023-07-26T21:53:00+02:00 Bug 13028: Prevent potential proxy bypass cases. It looks like these cases should only be invoked in the NSS command line tools, and not the browser, but I decided to patch them anyway because there literally is a maze of network function pointers being passed around, and it's very hard to tell if some random code might not pass in the proper proxied versions of the networking code here by accident. Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1433509 - - - - - d1149500 by Igor Oliveira at 2023-07-26T21:53:00+02:00 Bug 23104: Add a default line height compensation Many fonts have issues with their vertical metrics. they are used to influence the height of ascenders and depth of descenders. Gecko uses it to calculate the line height (font height + ascender + descender), however because of that idiosyncratic behavior across multiple operating systems, it can be used to identify the user's OS. The solution proposed in the patch uses a default factor to be multiplied with the font size, simulating the concept of ascender and descender. This way all operating systems will have the same line height only and only if the frame is outside the chrome. - - - - - b294d2f2 by Pier Angelo Vendrame at 2023-07-26T21:53:01+02:00 Bug 40309: Avoid using regional OS locales Avoid regional OS locales if the pref `intl.regional_prefs.use_os_locales` is false but RFP is enabled. - - - - - 2de3c3e1 by Matthew Finkel at 2023-07-26T21:53:02+02:00 Bug 40432: Prevent probing installed applications Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1711084 - - - - - ae956fde by Pier Angelo Vendrame at 2023-07-26T21:53:03+02:00 Base Browser strings This commit adds all the strings needed by following Base Browser patches. - - - - - 7dc8381d by hackademix at 2023-07-26T21:53:04+02:00 Bug 41434: Letterboxing, preemptively apply margins in a global CSS rule to mitigate race conditions on newly created windows and tabs. - - - - - b69832f6 by hackademix at 2023-07-26T21:53:05+02:00 Bug 41434: Letterboxing, improve logging. - - - - - 29d68b60 by hackademix at 2023-07-26T21:53:06+02:00 Bug 31064: Letterboxing, exempt browser extensions. - - - - - 83e0ee54 by hackademix at 2023-07-26T21:53:06+02:00 Bug 32411: Letterboxing, exempt view-source: URIs. - - - - - 0b245d7d by hackademix at 2023-07-26T21:53:07+02:00 Bug 32308: use direct browser sizing for letterboxing. - - - - - b9381cc9 by hackademix at 2023-07-26T21:53:08+02:00 Bug 41631: Prevent weird initial window dimensions caused by subpixel computations - - - - - 31d7df07 by Pier Angelo Vendrame at 2023-07-26T21:53:09+02:00 Bug 41369: Improve Firefox language settings for multi-lingual packages Change the language selector to be sorted by language code, rather than name, and to display the language code to the user. Bug 41372: Handle Japanese as a special case in preferences on macOS Japanese is treated in a special way on macOS. However, seeing the Japanese language tag could be confusing for users, and moreover the language name is not localized correctly like other langs. Bug 41378: Tell users that they can change their language at the first start With multi-lingual builds, Tor Browser matches the user's system language, but some users might want to change it. So, we tell them that it is possible, but only once. - - - - - 60deb682 by Henry Wilkes at 2023-07-26T21:53:10+02:00 Bug 41371: Temporary hack to fix language selector popup position - - - - - 43be7b06 by p13dz at 2023-07-26T21:53:11+02:00 Bug 40283: Workaround for the file upload bug - - - - - 0a7e7744 by Arthur Edelstein at 2023-07-26T21:53:12+02:00 Bug 18905: Hide unwanted items from help menu Bug 25660: Remove the "New Private Window" option - - - - - f799f6d8 by Pier Angelo Vendrame at 2023-07-26T21:53:13+02:00 Bug 41603: Customize the creation of MOZ_SOURCE_URL MOZ_SOURCE_URL is created by combining MOZ_SOURCE_REPO and MOZ_SOURCE_CHANGESET. But the code takes for granted that it refers to a Hg instance, so it combines them as `$MOZ_SOURCE_REPO/rev/$MOZ_SOURCE_CHANGESET`. With this commit, we change this logic to combine them to create a URL that is valid for GitLab. $MOZ_SOURCE_CHANGESET needs to be a commit hash, not a branch or a tag. If that is needed, we could use /-/tree/, instead of /-/commit/. - - - - - 83594880 by Henry Wilkes at 2023-07-26T21:53:14+02:00 Bug 31575: Disable Firefox Home (Activity Stream) Treat about:blank as the default home page and new tab page. Avoid loading AboutNewTab in BrowserGlue.jsm in order to avoid several network requests that we do not need. Bug 41624: Disable about:pocket-* pages and about:firefoxview. Bug 40144: Redirect about:privatebrowsing to the user's home - - - - - 88b637f7 by Pier Angelo Vendrame at 2023-07-26T21:53:14+02:00 Bug 41542: Disable the creation of a default profile Firefox creates a profile as a fallback for old versions. However, we do not need this, so we want to prevent Firefox from creating the related directories. We comment out the code, but should Mozilla be interested in a compile-time flag, we could rewrite the patch to add one. See also https://bugzilla.mozilla.org/show_bug.cgi?id=1770174. - - - - - 71e39eb0 by Kathy Brade at 2023-07-26T21:53:15+02:00 Bug 13379: Allow using NSS to sign and verify MAR signatures Allow using NSS on all platforms for checking MAR signatures (instead of using OS-native APIs, the default on Mac OS and Windows). So that the NSS and NSPR libraries the updater depends on can be found at runtime, we add the firefox directory to the shared library search path on macOS. On Linux, rpath is used to solve that problem, but that approach won't work on macOS because the updater executable is copied during the update process to a location that can vary. - - - - - 7315ed21 by Kathy Brade at 2023-07-26T21:53:16+02:00 Bug 4234: Use the Firefox Update Process for Base Browser. Windows: disable "runas" code path in updater (15201). Windows: avoid writing to the registry (16236). Also includes fixes for tickets 13047, 13301, 13356, 13594, 15406, 16014, 16909, 24476, and 25909. Also fix bug 27221: purge the startup cache if the Base Browser version changed (even if the Firefox version and build ID did not change), e.g., after a minor Base Browser update. Also fix 32616: Disable GetSecureOutputDirectoryPath() functionality. Bug 26048: potentially confusing "restart to update" message Within the update doorhanger, remove the misleading message that mentions that windows will be restored after an update is applied, and replace the "Restart and Restore" button label with an existing "Restart to update Tor Browser" string. Bug 28885: notify users that update is downloading Add a "Downloading Base Browser update" item which appears in the hamburger (app) menu while the update service is downloading a MAR file. Before this change, the browser did not indicate to the user that an update was in progress, which is especially confusing in Tor Browser because downloads often take some time. If the user clicks on the new menu item, the about dialog is opened to allow the user to see download progress. As part of this fix, the update service was changed to always show update-related messages in the hamburger menu, even if the update was started in the foreground via the about dialog or via the "Check for Tor Browser Update" toolbar menu item. This change is consistent with the Tor Browser goal of making sure users are informed about the update process. Removed #28885 parts of this patch which have been uplifted to Firefox. - - - - - c6cf8afc by Nicolas Vigier at 2023-07-26T21:53:17+02:00 Bug 41682: Add base-browser nightly mar signing key - - - - - 698ee78e by hackademix at 2023-07-26T21:53:18+02:00 Bug 41695: Warn on window maximization without letterboxing in RFPHelper module - - - - - 835a208b by hackademix at 2023-07-26T21:53:19+02:00 Bug 41854: Allow overriding download spam protection. - - - - - 7a250899 by Pier Angelo Vendrame at 2023-07-26T21:53:20+02:00 Bug 41698: Reword the recommendation badges in about:addons Firefox strings use { -brand-product-name }. As a result, it seems that the fork is recommending extensions, whereas AMO curators are doing that. So, we replace the strings with custom ones that clarify that Mozilla is recommending them. We assign the strings with JS because our translation backend does not support Fluent attributes, yet, but once it does, we should switch to them, instead. Upstream bug: https://bugzilla.mozilla.org/show_bug.cgi?id=1825033 - - - - - c963884b by Alex Catarineu at 2023-07-26T21:53:21+02:00 Bug 40069: Add helpers for message passing with extensions - - - - - d164facc by Matthew Finkel at 2023-07-26T21:53:21+02:00 Bug 41598: Prevent NoScript from being removed/disabled. Bug 40253: Explicitly allow NoScript in Private Browsing mode. - - - - - c4322c27 by Pier Angelo Vendrame at 2023-07-26T21:53:22+02:00 Bug 40925: Implemented the Security Level component This component adds a new Security Level toolbar button which visually indicates the current global security level via icon (as defined by the extensions.torbutton.security_slider pref), a drop-down hanger with a short description of the current security level, and a new section in the about:preferences#privacy page where users can change their current security level. In addition, the hanger and the preferences page will show a visual warning when the user has modified prefs associated with the security level and provide a one-click 'Restore Defaults' button to get the user back on recommended settings. Bug 40125: Expose Security Level pref in GeckoView - - - - - f23ceeeb by Pier Angelo Vendrame at 2023-07-26T21:53:23+02:00 Bug 40926: Implemented the New Identity feature - - - - - fadc5932 by Henry Wilkes at 2023-07-26T21:53:24+02:00 Bug 41736: Customize toolbar for base-browser. - - - - - 30 changed files: - .eslintignore - + .gitlab/issue_templates/UXBug.md - + .gitlab/issue_templates/bug.md - + .gitlab/merge_request_templates/default.md - accessible/ipc/DocAccessibleParent.cpp - − browser/actors/RFPHelperChild.jsm - − browser/actors/RFPHelperParent.jsm - browser/actors/moz.build - browser/app/Makefile.in - browser/app/macbuild/Contents/MacOS-files.in - browser/app/permissions - + browser/app/profile/001-base-profile.js - browser/app/profile/firefox.js - browser/base/content/aboutDialog-appUpdater.js - browser/base/content/aboutDialog.js - browser/base/content/appmenu-viewcache.inc.xhtml - browser/base/content/browser-menubar.inc - browser/base/content/browser-safebrowsing.js - browser/base/content/browser-sets.inc - browser/base/content/browser-siteIdentity.js - browser/base/content/browser.css - browser/base/content/browser.js - browser/base/content/browser.xhtml - + browser/base/content/languageNotification.js - browser/base/content/main-popupset.inc.xhtml - browser/base/content/navigator-toolbox.inc.xhtml - browser/base/jar.mn - browser/base/moz.build - browser/components/BrowserContentHandler.jsm - browser/components/BrowserGlue.jsm The diff was not included because it is too large. View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/4ce612… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/4ce612… You're receiving this email because of your account on gitlab.torproject.org.

1 0

[Git][tpo/applications/tor-browser][tor-browser-102.14.0esr-12.5-1] 138 commits: Bug 1774413: Bundled fonts are not picked up on macOS r=jfkthame
by ma1 (＠ma1) 26 Jul '23

26 Jul '23

ma1 pushed to branch tor-browser-102.14.0esr-12.5-1 at The Tor Project / Applications / Tor Browser Commits: 66650a2e by Pier Angelo Vendrame at 2023-07-25T23:31:49+02:00 Bug 1774413: Bundled fonts are not picked up on macOS r=jfkthame Bundled fonts were not picked up because also child processes need to register them. Also, they were assigned User visibility, instead of Base, which was not coherent with other platforms. Differential Revision: https://phabricator.services.mozilla.com/D150400 - - - - - 891ce49f by Pier Angelo Vendrame at 2023-07-25T23:31:56+02:00 Bug 1769030: Add a configure flag to load policies only from the local policies.json file r=mkaply,glandium Add a configuration flag to make Enterprise Policies mechanism only consult a policies.json file (avoiding the Windows Registry, macOS's file system attributes, and /etc/firefox/policies/policies.json on other OS). Differential Revision: https://phabricator.services.mozilla.com/D146300 - - - - - 0167b97a by Jamie Nicol at 2023-07-25T23:31:57+02:00 Bug 1784588 - Fix SplitDriverVersion when the string has less than 4 numbers. r=aosmond Differential Revision: https://phabricator.services.mozilla.com/D154897 - - - - - b1829371 by Pier Angelo Vendrame at 2023-07-25T23:31:58+02:00 Bug 1768907 - Part 1: Make browser.privatebrowsing.autostart a static pref. r=handyman,necko-reviewers Differential Revision: https://phabricator.services.mozilla.com/D157843 - - - - - 1125e718 by Neill Miller at 2023-07-25T23:31:59+02:00 Bug 1768907 - Part 2: nsTransferable disk-leaks when private browsing mode is enabled by default. r=handyman nsTransferable was modified to prevent disk leakings when copying data in private browsing mode with Bug 1123480. However, the context is nullptr when it is initialized, so it still leaks if PBM is enabled by default. Our solution is to check the browser.privatebrowsing.autostart in this condition. Differential Revision: https://phabricator.services.mozilla.com/D157800 - - - - - e3afc014 by Gabriele Svelto at 2023-07-25T23:32:00+02:00 Bug 1670885 - Fix post-fork() handlers for PHC/LogAlloc to work on macOS using unfair locks r=glandium macOS unfair locks enforce that a lock can only be released by the thread which locked it. On macOS 11+ this caused the fork()'d child process to raise a SIGILL signal. Confusingly enough this behavior seems to be different on macOS 10.15 and possibly interacted in odd ways with our exception handler if it was installed before fork()-ing. Differential Revision: https://phabricator.services.mozilla.com/D148287 - - - - - f4cf7ede by Gabriele Svelto at 2023-07-25T23:32:01+02:00 Bug 1670885 - Replace deprecated NSSpinLocks with os_unfair_locks in the memory allocator r=glandium Depends on D148287 Differential Revision: https://phabricator.services.mozilla.com/D99280 - - - - - 3d94ece4 by Kershaw Chang at 2023-07-25T23:32:02+02:00 Bug 1685403 - Use NS_GetFinalChannelURI at some places, r=necko-reviewers,valentin Differential Revision: https://phabricator.services.mozilla.com/D164656 - - - - - 24236471 by Kathy Brade at 2023-07-25T23:32:02+02:00 Bug 1769028: Disable DNS lookup in nsProfileLock when the proxy bypass protection is enabled r=mossop Instead of using the local computer's IP address within symlink-based profile lock signatures, always use 127.0.0.1 when the proxy bypass protection is enabled. Differential Revision: https://phabricator.services.mozilla.com/D161264 - - - - - c3235edc by Gabriele Svelto at 2023-07-25T23:32:03+02:00 Bug 1774458 - Use undocumented, non-public adaptive spinlocks on macOS 10.15+, revert to user-space spinlocks on older versions r=pbone Differential Revision: https://phabricator.services.mozilla.com/D149599 - - - - - b22ed534 by Gabriele Svelto at 2023-07-25T23:32:04+02:00 Bug 1784018 - Remove deprecated OSSpinLocks r=glandium On macOS versions prior to 10.15 os_unfair_locks cannot spin in kernel-space which degrades performance significantly. To obviate for this we spin in user-space like OSSpinLock does, for the same number of times and invoking x86-specific pause instructions in-between the locking attempts to avoid starving a thread that might be running on the same physical core. Differential Revision: https://phabricator.services.mozilla.com/D154205 - - - - - e5e87caa by Hubert Boma Manilla at 2023-07-25T23:32:05+02:00 Bug 1790542 - [devtools] Do not allow redirects on the 'sourceMappingUrl' r=jdescottes,devtools-reviewers - Block redirects on sourceMappingUrl - Add debugger test Differential Revision: https://phabricator.services.mozilla.com/D168648 - - - - - 9d39b9d6 by Hubert Boma Manilla at 2023-07-25T23:32:06+02:00 Bug 1790542 - [devtools] Convert browser_source_map-reload.js test from using .sjs files to createTestHTTPServer r=ochameau,devtools-reviewers .sjs files depend on redirects to load static files. In D168648 we block redirects on sourceMappingUrl, so this browser_source_map-reload.js fails because we use .sjs to serve the content. This patch refactors to no longer use .sjs files Differential Revision: https://phabricator.services.mozilla.com/D169739 - - - - - c02ec6c6 by Jonathan Kew at 2023-07-25T23:32:07+02:00 Bug 1792157 - Ensure correct font-visibility for bundled fonts on Linux when they 'shadow' system-installed families. r=gfx-reviewers,lsalzman Differential Revision: https://phabricator.services.mozilla.com/D175350 - - - - - a72fbee6 by Rob Wu at 2023-07-25T23:32:08+02:00 Bug 1802385 - Use NS_GetFinalChannelURI in FetchDriver r=rpl,valentin Depends on D164656 Differential Revision: https://phabricator.services.mozilla.com/D166108 - - - - - a424f1cf by Kash Shampur at 2023-07-25T23:32:09+02:00 Bug 1803109 - Discard blocks of data that are too big for two chunks. r=canaltinova Currently, `ReserveAndPutRaw` allocates a second span even if the data would be too big for the chunk. Here a second conditional is added to check if the block of data is too big in this scenario and silently discard the data if so. Differential Revision: https://phabricator.services.mozilla.com/D167167 - - - - - de73c560 by Pier Angelo Vendrame at 2023-07-25T23:32:09+02:00 Bug 1814416: Generalize the app name in about:buildconfig. r=ahochheiden Now it displays app display name, instead of having Firefox hardcoded. Differential Revision: https://phabricator.services.mozilla.com/D168562 - - - - - b74c0791 by Jon Coppeard at 2023-07-25T23:32:10+02:00 Bug 1816158 - Part 1: Disallow GC while iterating global's debugger vector r=sfink GC can mutate this vector so don't allow that while we are iterating. I think it would be safe to use index-based iteration but it's safer to just ban it entirely. This fixes the crash produced by the testcase. Differential Revision: https://phabricator.services.mozilla.com/D169701 - - - - - d6c35ccf by Jon Coppeard at 2023-07-25T23:32:11+02:00 Bug 1816158 - Part 2: Require no GC when giving out references to the realm's debugger vector r=sfink To prevent any other instances of this problme we can update the getDebuggers() methods on the global and the realm to require no GC. Depends on D169701 Differential Revision: https://phabricator.services.mozilla.com/D169702 - - - - - 4651288e by Pier Angelo Vendrame at 2023-07-25T23:32:12+02:00 Bug 1817756 - Add a seed to the network ID. r=valentin,necko-reviewers This helps to prevent linkability of users in the same network. Differential Revision: https://phabricator.services.mozilla.com/D170373 - - - - - 72140b84 by John Lin at 2023-07-25T23:42:26+02:00 Bug 1832832 - protect encoder accesses in Java callbacks. r=media-playback-reviewers,alwu a=pascalc Differential Revision: https://phabricator.services.mozilla.com/D178564 - - - - - cffad159 by Arsen Arsenović at 2023-07-25T23:42:27+02:00 Bug 1811714 - Add a few missing <cstdint> includes r=gfx-reviewers,nical GCC 13s libstdc++ reduced its dependency on some headers like <cstdint>, so it's no longer transitively included through various headers. Include it explicitly. See also: https://gcc.gnu.org/gcc-13/porting_to.html#header-dep-changes 11:16.32 .../components/telemetry/pingsender/pingsender.cpp:30:7: error: unknown type name 'uint32_t' 11:16.33 const uint32_t kConnectionTimeoutMs = 30 * 1000; 11:16.33 ^ 11:16.33 ... /components/telemetry/pingsender/pingsender.cpp:76:9: error: unknown type name 'uint32_t' 11:16.33 const uint32_t kBufferSize = 4 * 1024 - 1; 11:16.34 ^ 11:16.44 2 errors generated. (et al) Differential Revision: https://phabricator.services.mozilla.com/D167491 - - - - - 2f4a54a3 by James Teh at 2023-07-26T21:52:19+02:00 Bug 1772170: Ensure cached Accessibles and offsets are valid before returning them in DocAccessibleParent::SelectionRanges. r=eeejay Differential Revision: https://phabricator.services.mozilla.com/D148745 - - - - - e460aee8 by James Teh at 2023-07-26T21:52:20+02:00 Bug 1781169: Ensure that the cached start/end containers contain characters before returning them in DocAccessibleParent::SelectionRanges. r=eeejay In bug 1772170, I added code to ensure that the cached offsets are valid. However, this didn't account for the possibility that one of the containers contained 0 characters. In that case, offset 0 is itself invalid. If there were also no children, this might in GetChildAtOffset being called and returning null, causing a crash when retrieving selection. Differential Revision: https://phabricator.services.mozilla.com/D154964 - - - - - 7f4de3a3 by Georg Koppen at 2023-07-26T21:52:21+02:00 Revert "Bug 1724777, optimize suppressed MicroTask handling, r=mccr8 a=RyanVM" This reverts commit 1eb1364357ac5bc2a4531337fb5416af39c3793f. This fixes tor-browser#40721, tor-browser#40698, and tor-browser#40706. However, it is a temporary workaround, that we should revert once https://bugzilla.mozilla.org/show_bug.cgi?id=1744719 is fixed. - - - - - 5388480a by Henry Wilkes at 2023-07-26T21:52:22+02:00 Bug 41454: Move focus after calling openPreferences for a sub-category. Temporary fix until mozilla bug 1799153 gets a patch upstream. - - - - - 113974c0 by Henry Wilkes at 2023-07-26T21:52:23+02:00 Bug 41483: Remove the firefox override for appstrings.properties Remove this patch after upstream bugzilla bug 1790187 - - - - - bab3b502 by Pier Angelo Vendrame at 2023-07-26T21:52:24+02:00 Bug 41116: Normalize system fonts. System fonts are an enormous fingerprinting vector. Even with font allow lists and with our custom configuration on Linux, which counter metrics measurements, getComputedStyle leaks several details. This patch counters both these kinds of attacks. - - - - - 7bf81d37 by Marco Simonelli at 2023-07-26T21:52:25+02:00 Bug 41459: WebRTC fails to build under mingw (Part 1) - properly define NOMINMAX for just MSVC builds - - - - - 28aeacd5 by Marco Simonelli at 2023-07-26T21:52:26+02:00 Bug 41459: WebRTC fails to build under mingw (Part 2) - fixes required to build third_party/libwebrtc - - - - - 17fd7119 by Marco Simonelli at 2023-07-26T21:52:26+02:00 Bug 41459: WebRTC fails to build under mingw (Part 3) - fixes required to build third_party/sipcc - - - - - 49244d4b by Marco Simonelli at 2023-07-26T21:52:27+02:00 Bug 41459: WebRTC fails to build under mingw (Part 4) - fixes requried to build netwerk/sctp - - - - - 516a096f by Marco Simonelli at 2023-07-26T21:52:28+02:00 Bug 41459: WebRTC fails to build under mingw (Part 5) - fixes required to build dom/media/webrtc - - - - - e08f2fc3 by Marco Simonelli at 2023-07-26T21:52:29+02:00 Bug 41459: WebRTC fails to build under mingw (Part 6) - fixes required to build dom/media/systemservices - - - - - 447b1ded by Gaba at 2023-07-26T21:52:30+02:00 Adding issue and merge request templates - - - - - 91460f85 by Pier Angelo Vendrame at 2023-07-26T21:52:31+02:00 Base Browser's .mozconfigs. Bug 17858: Cannot create incremental MARs for hardened builds. Define HOST_CFLAGS, etc. to avoid compiling programs such as mbsdiff (which is part of mar-tools and is not distributed to end-users) with ASan. Bug 21849: Don't allow SSL key logging. Bug 25741 - TBA: Disable features at compile-time Define MOZ_ANDROID_NETWORK_STATE and MOZ_ANDROID_LOCATION Bug 27623 - Export MOZILLA_OFFICIAL during desktop builds This fixes a problem where some preferences had the wrong default value. Also see bug 27472 where we made a similar fix for Android. Bug 29859: Disable HLS support for now Bug 30463: Explicitly disable MOZ_TELEMETRY_REPORTING Bug 32493: Disable MOZ_SERVICES_HEALTHREPORT Bug 33734: Set MOZ_NORMANDY to False Bug 33851: Omit Parental Controls. Bug 40252: Add --enable-rust-simd to our tor-browser mozconfig files Bug 41584: Move some configuration options to base-browser level - - - - - 8fcbe886 by Pier Angelo Vendrame at 2023-07-26T21:52:32+02:00 Tweaks to the build system Bug 40857: Modified the fat .aar creation file This is a workaround to build fat .aars with the compiling enviornment disabled. Mozilla does not use a similar configuration, but either runs a Firefox build and discards its output, or uses artifacts build. We might switch to artifact builds too, and drop this patch, or write a better one to upstream. But until then we need this patch. See also https://bugzilla.mozilla.org/show_bug.cgi?id=1763770. Bug 41458: Prevent `mach package-multi-locale` from actually creating a package macOS builds need some files to be moved around with ./mach package-multi-locale to create multi-locale packages. The required command isn't exposed through any other mach command. So, we patch package-multi-locale both to prevent it from failing when doing official builds and to detect any future changes on it. - - - - - 5b2779a5 by Dan Ballard at 2023-07-26T21:52:32+02:00 Bug 41117: using MOZ_PROXY_BYPASS_PROTECTION disable share menu on mac and windows, and restore email link on mac - - - - - 5a3057bf by Pier Angelo Vendrame at 2023-07-26T21:52:33+02:00 Bug 41108: Remove privileged macOS installation from 102 - - - - - c78741f2 by Dan Ballard at 2023-07-26T21:52:34+02:00 Bug 41149: Re-enable DLL injection protection in all builds not just nightlies - - - - - b4a4bec8 by Matthew Finkel at 2023-07-26T21:52:35+02:00 Bug 24796: Comment out excess permissions from GeckoView The GeckoView AndroidManifest.xml is not preprocessed unlike Fennec's manifest, so we can't use the ifdef preprocessor guards around the permissions we do not want. Commenting the permissions is the next-best-thing. - - - - - 81122dd6 by Matthew Finkel at 2023-07-26T21:52:36+02:00 Bug 28125: Prevent non-Necko network connections - - - - - 617b19d0 by Mike Perry at 2023-07-26T21:52:37+02:00 Bug 12974: Disable NTLM and Negotiate HTTP Auth The Mozilla bugs: https://bugzilla.mozilla.org/show_bug.cgi?id=1046421, https://bugzilla.mozilla.org/show_bug.cgi?id=1261591, tor-browser#27602 - - - - - 5107f970 by Alex Catarineu at 2023-07-26T21:52:38+02:00 Bug 40166: Disable security.certerrors.mitm.auto_enable_enterprise_roots Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1768899 - - - - - 5ccd2869 by Georg Koppen at 2023-07-26T21:52:39+02:00 Bug 16285: Exclude ClearKey system for now In the past the ClearKey system had not been compiled when specifying --disable-eme. But that changed and it is even bundled nowadays (see: Mozilla's bug 1300654). We don't want to ship it right now as the use case for it is not really visible while the code had security vulnerabilities in the past. - - - - - 972b550c by Kathy Brade at 2023-07-26T21:52:39+02:00 Bug 21431: Clean-up system extensions shipped in Firefox Only ship the pdfjs extension. - - - - - 9cfa53dc by Kathy Brade at 2023-07-26T21:52:40+02:00 Bug 33852: Clean up about:logins (LockWise) to avoid mentioning sync, etc. Hide elements on about:logins that mention sync, "Firefox LockWise", and Mozilla's LockWise mobile apps. Disable the "Create New Login" button when security.nocertdb is true. - - - - - 4afff71e by Alex Catarineu at 2023-07-26T21:52:41+02:00 Bug 41457: Remove Mozilla permissions Bug 40025: Remove Mozilla add-on install permissions - - - - - f3231564 by Kathy Brade at 2023-07-26T21:52:42+02:00 Bug 40002: Remove about:ion Firefox Ion (previously Firefox Pioneer) is an opt-in program in which people volunteer to participate in studies that collect detailed, sensitive data about how they use their browser. Bug 41662: Disable about:sync-logs Even though we disable sync by default with `identity.fxaccounts.enabled`, this about: page is still avilable. We could throw an exception on the constructor of the related component, but it would result only in an error in the console, without a visible "this address does not look right" error page. If we fix the issues with MOZ_SERVICES_SYNC, we can restore the component. - - - - - 42f24da1 by Arthur Edelstein at 2023-07-26T21:52:43+02:00 Bug 26353: Prevent speculative connect that violated FPI. Connections were observed in the catch-all circuit when the user entered an https or http URL in the URL bar, or typed a search term. - - - - - 727a1d8b by Alex Catarineu at 2023-07-26T21:52:44+02:00 Bug 31740: Remove some unnecessary RemoteSettings instances More concretely, SearchService.jsm 'hijack-blocklists' and url-classifier-skip-urls. Avoid creating instance for 'anti-tracking-url-decoration'. If prefs are disabling their usage, avoid creating instances for 'cert-revocations' and 'intermediates'. Do not ship JSON dumps for collections we do not expect to need. For the ones in the 'main' bucket, this prevents them from being synced unnecessarily (the code in remote-settings does so for collections in the main bucket for which a dump or local data exists). For the collections in the other buckets, we just save some size by not shipping their dumps. We also clear the collections database on the v2 -> v3 migration. - - - - - e5bc5b12 by Pier Angelo Vendrame at 2023-07-26T21:52:45+02:00 Bug 41635: Disable the Normandy component Do not include Normandy at all whenever MOZ_NORMANDY is False. - - - - - dfc706b5 by Georg Koppen at 2023-07-26T21:52:46+02:00 Bug 30541: Disable WebGL readPixel() for web content Related Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1428034 - - - - - fae932dd by Alex Catarineu at 2023-07-26T21:52:46+02:00 Bug 28369: Stop shipping pingsender executable - - - - - 5d5d652e by cypherpunks1 at 2023-07-26T21:52:47+02:00 Bug 41565: Gate Telemetry Tasks behind MOZ_TELEMETRY_REPORTING - - - - - 35cb57e6 by cypherpunks1 at 2023-07-26T21:52:48+02:00 Bug 41568: Disable LaterRun - - - - - 75bb1ef6 by cypherpunks1 at 2023-07-26T21:52:49+02:00 Bug 40717: Hide Windows SSO in settings - - - - - 683e266e by Pier Angelo Vendrame at 2023-07-26T21:52:50+02:00 Bug 40788: Prevent nsIURLQueryStrippingListService from calling home when it is not enabled. The URL query stripping service is enabled only in nightly builds, still it is initialized and remote settings are downloaded. This adds a condition that prevents the service from being initialized if disabled. Upstream Bug: https://bugzilla.mozilla.org/show_bug.cgi?id=1812594 We should remove this patch if Mozilla decides that this is a problem also for them, or if they do not but we decide to use the feature. - - - - - c10ca88e by Pier Angelo Vendrame at 2023-07-26T21:52:51+02:00 Bug 41599: Always return an empty string as network ID Firefox computes an internal network ID used to detect network changes and act consequently (e.g., to improve WebSocket UX). However, there are a few ways to get this internal network ID, so we patch them out, to be sure any new code will not be able to use them and possibly link users. We also sent a patch to Mozilla to seed the internal network ID, to prevent any accidental leak in the future. Upstream: https://bugzilla.mozilla.org/show_bug.cgi?id=1817756 - - - - - 08d32d39 by Mike Perry at 2023-07-26T21:52:52+02:00 Firefox preference overrides. This hack directly includes our preference changes in omni.ja. Bug 18292: Staged updates fail on Windows Temporarily disable staged updates on Windows. Bug 18297: Use separate Noto JP,KR,SC,TC fonts Bug 23404: Add Noto Sans Buginese to the macOS whitelist Bug 23745: Set dom.indexedDB.enabled = true Bug 13575: Disable randomised Firefox HTTP cache decay user tests. (Fernando Fernandez Mancera <ffmancera(a)riseup.net>) Bug 17252: Enable session identifiers with FPI Session tickets and session identifiers were isolated by OriginAttributes, so we can re-enable them by allowing the default value (true) of "security.ssl.disable_session_identifiers". The pref "security.enable_tls_session_tickets" is obsolete (removed in https://bugzilla.mozilla.org/917049) Bug 14952: Enable http/2 and AltSvc In Firefox, SPDY/HTTP2 now uses Origin Attributes for isolation of connections, push streams, origin frames, etc. That means we get first-party isolation provided "privacy.firstparty.isolate" is true. So in this patch, we stop overriding "network.http.spdy.enabled" and "network.http.spdy.enabled.http2". Alternate Services also use Origin Attributes for isolation. So we stop overriding "network.http.altsvc.enabled" and "network.http.altsvc.oe" as well. (All 4 of the abovementioned "network.http.*" prefs adopt Firefox 60ESR's default value of true.) However, we want to disable HTTP/2 push for now, so we set "network.http.spdy.allow-push" to false. "network.http.spdy.enabled.http2draft" was removed in Bug 1132357. "network.http.sped.enabled.v2" was removed in Bug 912550. "network.http.sped.enabled.v3" was removed in Bug 1097944. "network.http.sped.enabled.v3-1" was removed in Bug 1248197. Bug 26114: addons.mozilla.org is not special * Don't expose navigator.mozAddonManager on any site * Don't block NoScript from modifying addons.mozilla.org or other sites Enable ReaderView mode again (#27281). Bug 29916: Make sure enterprise policies are disabled Bug 2874: Block Components.interfaces from content Bug 26146: Spoof HTTP User-Agent header for desktop platforms In Tor Browser 8.0, the OS was revealed in both the HTTP User-Agent header and to JavaScript code via navigator.userAgent. To avoid leaking the OS inside each HTTP request (which many web servers log), always use the Windows 7 OS value in the desktop User-Agent header. We continue to allow access to the actual OS via JavaScript, since doing so improves compatibility with web applications such as GitHub and Google Docs. Bug 12885: Windows Jump Lists fail for Tor Browser Jumplist entries are stored in a binary file in: %APPDATA%\\Microsoft\Windows\Recent\CustomDestinations\ and has a name in the form [a-f0-9]+.customDestinations-ms The hex at the front is unique per app, and is ultimately derived from something called the 'App User Model ID' (AUMID) via some unknown hashing method. The AUMID is provided as a key when programmatically creating, updating, and deleting a jumplist. The default behaviour in firefox is for the installer to define an AUMID for an app, and save it in the registry so that the jumplist data can be removed by the uninstaller. However, the Tor Browser does not set this (or any other) regkey during installation, so this codepath fails and the app's AUMID is left undefined. As a result the app's AUMID ends up being defined by windows, but unknowable by Tor Browser. This unknown AUMID is used to create and modify the jumplist, but the delete API requires that we provide the app's AUMID explicitly. Since we don't know what the AUMID is (since the expected regkey where it is normally stored does not exist) jumplist deletion will fail and we will leave behind a mostly empty customDestinations-ms file. The name of the file is derived from the binary path, so an enterprising person could reverse engineer how that hex name is calculated, and generate the name for Tor Browser's default Desktop installation path to determine whether a person had used Tor Browser in the past. The 'taskbar.grouping.useprofile' option that is enabled by this patch works around this AUMID problem by having firefox.exe create it's own AUMID based on the profile path (rather than looking for a regkey). This way, if a user goes in and enables and disables jumplist entries, the backing store is properly deleted. Unfortunately, all windows users currently have this file lurking in the above mentioned directory and this patch will not remove it since it was created with an unknown AUMID. However, another patch could be written which goes to that directory and deletes any item containing the 'Tor Browser' string. See bug 28996. Bug 30845: Make sure default themes and other internal extensions are enabled Bug 28896: Enable extensions in private browsing by default Bug 31065: Explicitly allow proxying localhost Bug 31598: Enable letterboxing Disable Presentation API everywhere Bug 21549 - Use Firefox's WASM default pref. It is disabled at safer security levels. Bug 32321: Disable Mozilla's MitM pings Bug 19890: Disable installation of system addons By setting the URL to "" we make sure that already installed system addons get deleted as well. Bug 22548: Firefox downgrades VP9 videos to VP8. On systems where H.264 is not available or no HWA, VP9 is preferred. But in Tor Browser 7.0 all youtube videos are degraded to VP8. This behaviour can be turned off by setting media.benchmark.vp9.threshold to 0. All clients will get better experience and lower traffic, beause TBB doesn't use "Use hardware acceleration when available". Bug 25741 - TBA: Add mobile-override of 000-tor-browser prefs Bug 16441: Suppress "Reset Tor Browser" prompt. Bug 29120: Use the in-memory media cache and increase its maximum size. Bug 33697: use old search config based on list.json Bug 33855: Ensure that site-specific browser mode is disabled. Bug 30682: Disable Intermediate CA Preloading. Bug 40061: Omit the Windows default browser agent from the build Bug 40322: Consider disabling network.connectivity-service.enabled Bug 40408: Disallow SVG Context Paint in all web content Bug 40308: Disable network partitioning until we evaluate dFPI Bug 40322: Consider disabling network.connectivity-service.enabled Bug 40383: Disable dom.enable_event_timing Bug 40423: Disable http/3 Bug 40177: Update prefs for Fx91esr Bug 40700: Disable addons and features recommendations Bug 40682: Disable network.proxy.allow_bypass Bug 40736: Disable third-party cookies in PBM Bug 19850: Enabled HTTPS-Only by default Bug 40912: Hide the screenshot menu Bug 41292: Disable moreFromMozilla in preferences page Bug 40057: Ensure the CSS4 system colors are not a fingerprinting vector Bug 24686: Set network.http.tailing.enabled to true Bug 40183: Disable TLS ciphersuites using SHA-1 Bug 40783: Review 000-tor-browser.js and 001-base-profile.js for 102 We reviewed all the preferences we set for 102, and remove a few old ones. See the description of that issue to see all the preferences we believed were still valid for 102, and some brief description for the reasons to keep them. - - - - - b469f0bf by Richard Pospesel at 2023-07-26T21:52:53+02:00 Bug 41659: Add canonical color definitions to base-browser - - - - - 92e56693 by Pier Angelo Vendrame at 2023-07-26T21:52:53+02:00 Bug 41043: Hardcode the UI font on Linux The mechanism to choose the UI font does not play well with our fontconfig configuration. As a result, the final criterion to choose the font for the UI was its version. Since we hardcode Arimo as a default sans-serif on preferences, we use it also for the UI. FontConfig will fall back to some other font for scripts Arimo does not cover as expected (we tested with Japanese). - - - - - e4faff8e by Alex Catarineu at 2023-07-26T21:52:54+02:00 Bug 30605: Honor privacy.spoof_english in Android This checks `privacy.spoof_english` whenever `setLocales` is called from Fenix side and sets `intl.accept_languages` accordingly. Bug 40198: Expose privacy.spoof_english pref in GeckoView - - - - - 62ae3e48 by Alex Catarineu at 2023-07-26T21:52:55+02:00 Bug 40199: Avoid using system locale for intl.accept_languages in GeckoView - - - - - d8a6dad5 by Alex Catarineu at 2023-07-26T21:52:56+02:00 Bug 40171: Make WebRequest and GeckoWebExecutor First-Party aware - - - - - 5e7fdf93 by Alex Catarineu at 2023-07-26T21:52:57+02:00 Bug 26345: Hide tracking protection UI - - - - - bf0600b5 by Pier Angelo Vendrame at 2023-07-26T21:52:58+02:00 Bug 9173: Change the default Firefox profile directory to be relative. This commit makes Firefox look for the default profile directory in a directory relative to the binary path. The directory can be specified through the --with-relative-data-dir. This is relative to the same directory as the firefox main binary for Linux and Windows. On macOS, we remove Contents/MacOS from it. Or, in other words, the directory is relative to the application bundle. This behavior can be overriden at runtime, by placing a file called system-install adjacent to the firefox main binary (also on macOS). - - - - - f86be196 by Alex Catarineu at 2023-07-26T21:52:59+02:00 Bug 27604: Fix addon issues when moving the profile directory Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1429838 - - - - - 54a51fd9 by Mike Perry at 2023-07-26T21:53:00+02:00 Bug 13028: Prevent potential proxy bypass cases. It looks like these cases should only be invoked in the NSS command line tools, and not the browser, but I decided to patch them anyway because there literally is a maze of network function pointers being passed around, and it's very hard to tell if some random code might not pass in the proper proxied versions of the networking code here by accident. Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1433509 - - - - - d1149500 by Igor Oliveira at 2023-07-26T21:53:00+02:00 Bug 23104: Add a default line height compensation Many fonts have issues with their vertical metrics. they are used to influence the height of ascenders and depth of descenders. Gecko uses it to calculate the line height (font height + ascender + descender), however because of that idiosyncratic behavior across multiple operating systems, it can be used to identify the user's OS. The solution proposed in the patch uses a default factor to be multiplied with the font size, simulating the concept of ascender and descender. This way all operating systems will have the same line height only and only if the frame is outside the chrome. - - - - - b294d2f2 by Pier Angelo Vendrame at 2023-07-26T21:53:01+02:00 Bug 40309: Avoid using regional OS locales Avoid regional OS locales if the pref `intl.regional_prefs.use_os_locales` is false but RFP is enabled. - - - - - 2de3c3e1 by Matthew Finkel at 2023-07-26T21:53:02+02:00 Bug 40432: Prevent probing installed applications Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1711084 - - - - - ae956fde by Pier Angelo Vendrame at 2023-07-26T21:53:03+02:00 Base Browser strings This commit adds all the strings needed by following Base Browser patches. - - - - - 7dc8381d by hackademix at 2023-07-26T21:53:04+02:00 Bug 41434: Letterboxing, preemptively apply margins in a global CSS rule to mitigate race conditions on newly created windows and tabs. - - - - - b69832f6 by hackademix at 2023-07-26T21:53:05+02:00 Bug 41434: Letterboxing, improve logging. - - - - - 29d68b60 by hackademix at 2023-07-26T21:53:06+02:00 Bug 31064: Letterboxing, exempt browser extensions. - - - - - 83e0ee54 by hackademix at 2023-07-26T21:53:06+02:00 Bug 32411: Letterboxing, exempt view-source: URIs. - - - - - 0b245d7d by hackademix at 2023-07-26T21:53:07+02:00 Bug 32308: use direct browser sizing for letterboxing. - - - - - b9381cc9 by hackademix at 2023-07-26T21:53:08+02:00 Bug 41631: Prevent weird initial window dimensions caused by subpixel computations - - - - - 31d7df07 by Pier Angelo Vendrame at 2023-07-26T21:53:09+02:00 Bug 41369: Improve Firefox language settings for multi-lingual packages Change the language selector to be sorted by language code, rather than name, and to display the language code to the user. Bug 41372: Handle Japanese as a special case in preferences on macOS Japanese is treated in a special way on macOS. However, seeing the Japanese language tag could be confusing for users, and moreover the language name is not localized correctly like other langs. Bug 41378: Tell users that they can change their language at the first start With multi-lingual builds, Tor Browser matches the user's system language, but some users might want to change it. So, we tell them that it is possible, but only once. - - - - - 60deb682 by Henry Wilkes at 2023-07-26T21:53:10+02:00 Bug 41371: Temporary hack to fix language selector popup position - - - - - 43be7b06 by p13dz at 2023-07-26T21:53:11+02:00 Bug 40283: Workaround for the file upload bug - - - - - 0a7e7744 by Arthur Edelstein at 2023-07-26T21:53:12+02:00 Bug 18905: Hide unwanted items from help menu Bug 25660: Remove the "New Private Window" option - - - - - f799f6d8 by Pier Angelo Vendrame at 2023-07-26T21:53:13+02:00 Bug 41603: Customize the creation of MOZ_SOURCE_URL MOZ_SOURCE_URL is created by combining MOZ_SOURCE_REPO and MOZ_SOURCE_CHANGESET. But the code takes for granted that it refers to a Hg instance, so it combines them as `$MOZ_SOURCE_REPO/rev/$MOZ_SOURCE_CHANGESET`. With this commit, we change this logic to combine them to create a URL that is valid for GitLab. $MOZ_SOURCE_CHANGESET needs to be a commit hash, not a branch or a tag. If that is needed, we could use /-/tree/, instead of /-/commit/. - - - - - 83594880 by Henry Wilkes at 2023-07-26T21:53:14+02:00 Bug 31575: Disable Firefox Home (Activity Stream) Treat about:blank as the default home page and new tab page. Avoid loading AboutNewTab in BrowserGlue.jsm in order to avoid several network requests that we do not need. Bug 41624: Disable about:pocket-* pages and about:firefoxview. Bug 40144: Redirect about:privatebrowsing to the user's home - - - - - 88b637f7 by Pier Angelo Vendrame at 2023-07-26T21:53:14+02:00 Bug 41542: Disable the creation of a default profile Firefox creates a profile as a fallback for old versions. However, we do not need this, so we want to prevent Firefox from creating the related directories. We comment out the code, but should Mozilla be interested in a compile-time flag, we could rewrite the patch to add one. See also https://bugzilla.mozilla.org/show_bug.cgi?id=1770174. - - - - - 71e39eb0 by Kathy Brade at 2023-07-26T21:53:15+02:00 Bug 13379: Allow using NSS to sign and verify MAR signatures Allow using NSS on all platforms for checking MAR signatures (instead of using OS-native APIs, the default on Mac OS and Windows). So that the NSS and NSPR libraries the updater depends on can be found at runtime, we add the firefox directory to the shared library search path on macOS. On Linux, rpath is used to solve that problem, but that approach won't work on macOS because the updater executable is copied during the update process to a location that can vary. - - - - - 7315ed21 by Kathy Brade at 2023-07-26T21:53:16+02:00 Bug 4234: Use the Firefox Update Process for Base Browser. Windows: disable "runas" code path in updater (15201). Windows: avoid writing to the registry (16236). Also includes fixes for tickets 13047, 13301, 13356, 13594, 15406, 16014, 16909, 24476, and 25909. Also fix bug 27221: purge the startup cache if the Base Browser version changed (even if the Firefox version and build ID did not change), e.g., after a minor Base Browser update. Also fix 32616: Disable GetSecureOutputDirectoryPath() functionality. Bug 26048: potentially confusing "restart to update" message Within the update doorhanger, remove the misleading message that mentions that windows will be restored after an update is applied, and replace the "Restart and Restore" button label with an existing "Restart to update Tor Browser" string. Bug 28885: notify users that update is downloading Add a "Downloading Base Browser update" item which appears in the hamburger (app) menu while the update service is downloading a MAR file. Before this change, the browser did not indicate to the user that an update was in progress, which is especially confusing in Tor Browser because downloads often take some time. If the user clicks on the new menu item, the about dialog is opened to allow the user to see download progress. As part of this fix, the update service was changed to always show update-related messages in the hamburger menu, even if the update was started in the foreground via the about dialog or via the "Check for Tor Browser Update" toolbar menu item. This change is consistent with the Tor Browser goal of making sure users are informed about the update process. Removed #28885 parts of this patch which have been uplifted to Firefox. - - - - - c6cf8afc by Nicolas Vigier at 2023-07-26T21:53:17+02:00 Bug 41682: Add base-browser nightly mar signing key - - - - - 698ee78e by hackademix at 2023-07-26T21:53:18+02:00 Bug 41695: Warn on window maximization without letterboxing in RFPHelper module - - - - - 835a208b by hackademix at 2023-07-26T21:53:19+02:00 Bug 41854: Allow overriding download spam protection. - - - - - 7a250899 by Pier Angelo Vendrame at 2023-07-26T21:53:20+02:00 Bug 41698: Reword the recommendation badges in about:addons Firefox strings use { -brand-product-name }. As a result, it seems that the fork is recommending extensions, whereas AMO curators are doing that. So, we replace the strings with custom ones that clarify that Mozilla is recommending them. We assign the strings with JS because our translation backend does not support Fluent attributes, yet, but once it does, we should switch to them, instead. Upstream bug: https://bugzilla.mozilla.org/show_bug.cgi?id=1825033 - - - - - c963884b by Alex Catarineu at 2023-07-26T21:53:21+02:00 Bug 40069: Add helpers for message passing with extensions - - - - - d164facc by Matthew Finkel at 2023-07-26T21:53:21+02:00 Bug 41598: Prevent NoScript from being removed/disabled. Bug 40253: Explicitly allow NoScript in Private Browsing mode. - - - - - c4322c27 by Pier Angelo Vendrame at 2023-07-26T21:53:22+02:00 Bug 40925: Implemented the Security Level component This component adds a new Security Level toolbar button which visually indicates the current global security level via icon (as defined by the extensions.torbutton.security_slider pref), a drop-down hanger with a short description of the current security level, and a new section in the about:preferences#privacy page where users can change their current security level. In addition, the hanger and the preferences page will show a visual warning when the user has modified prefs associated with the security level and provide a one-click 'Restore Defaults' button to get the user back on recommended settings. Bug 40125: Expose Security Level pref in GeckoView - - - - - f23ceeeb by Pier Angelo Vendrame at 2023-07-26T21:53:23+02:00 Bug 40926: Implemented the New Identity feature - - - - - fadc5932 by Henry Wilkes at 2023-07-26T21:53:24+02:00 Bug 41736: Customize toolbar for base-browser. - - - - - 401c2be8 by Richard Pospesel at 2023-07-26T21:53:25+02:00 Bug 41649: Create rebase and security backport gitlab issue templates - - - - - d2ccf82a by Richard Pospesel at 2023-07-26T21:53:26+02:00 Bug 41089: Add tor-browser build scripts + Makefile to tor-browser - - - - - ea2391d0 by Henry Wilkes at 2023-07-26T21:53:27+02:00 Bug 41803: Add some developer tools for working on tor-browser. - - - - - 9b88d8cb by Kathy Brade at 2023-07-26T21:53:27+02:00 Bug 11641: Disable remoting by default. Unless the -osint command line flag is used, the browser now defaults to the equivalent of -no-remote. There is a new -allow-remote flag that may be used to restore the original (Firefox-like) default behavior. - - - - - 130aa927 by Mike Perry at 2023-07-26T21:53:28+02:00 Bug 2176: Rebrand Firefox to TorBrowser See also Bugs #5194, #7187, #8115, #8219. This patch does some basic renaming of Firefox to TorBrowser. The rest of the branding is done by images and icons. Also fix bug 27905. Bug 25702: Update Tor Browser icon to follow design guidelines - Updated all of the branding in /browser/branding/official with new 'stable' icon series. - Updated /extensions/onboarding/content/img/tor-watermark.png with new icon and add the source svg in the same directory - Copied /browser/branding/official over /browser/branding/nightly and the new /browser/branding/alpha directories. Replaced content with 'nightly' and 'alpha' icon series. Updated VisualElements_70.png and VisualElements_150.png with updated icons in each branding directory (fixes #22654) - Updated firefox.VisualElementsManfiest.xml with updated colors in each branding directory - Added firefox.svg to each branding directory from which all the other icons are derived (apart from document.icns and document.ico) - Added default256.png and default512.png icons - Updated aboutTBUpdate.css to point to branding-aware icon128.png and removed original icon - Use the Tor Browser icon within devtools/client/themes/images/. Bug 30631: Blurry Tor Browser icon on macOS app switcher It would seem the png2icns tool does not generate correct icns files and so on macOS the larger icons were missing resulting in blurry icons in the OS chrome. Regenerated the padded icons in a macOS VM using iconutil. Bug 28196: preparations for using torbutton tor-browser-brand.ftl A small change to Fluent FileSource class is required so that we can register a new source without its supported locales being counted as available locales for the browser. Bug 31803: Replaced about:debugging logo with flat version Bug 21724: Make Firefox and Tor Browser distinct macOS apps When macOS opens a document or selects a default browser, it sometimes uses the CFBundleSignature. Changing from the Firefox MOZB signature to a different signature TORB allows macOS to distinguish between Firefox and Tor Browser. Bug 32092: Fix Tor Browser Support link in preferences For bug 40562, we moved onionPattern* from bug 27476 to here, as about:tor needs these files but it is included earlier. Bug 41278: Create Tor Browser styled pdf logo similar to the vanilla Firefox one - - - - - 74d4cead by Alex Catarineu at 2023-07-26T21:53:29+02:00 Bring back old Firefox onboarding Revert "Bug 1462415 - Delete onboarding system add-on r=Standard8,k88hudson" This reverts commit f7ffd78b62541d44d0102f8051d2f4080bdbc432. Revert "Bug 1498378 - Actually remove the old onboarding add-on's prefs r=Gijs" This reverts commit 057fe36fc6f3e93e265505c7dcc703a0941778e2. Bug 28822: Convert onboarding to webextension Partially revert 1564367 (controlCenter in UITour.jsm) - - - - - 28647eb6 by Kathy Brade at 2023-07-26T21:53:30+02:00 Bug 26961: New user onboarding. Reuse the Firefox onboarding mechanism with minimal changes. Localizable strings are pulled in from Torbutton (if Torbutton is not installed, we lack about:tor and no tour will be shown). Replace SVG images with PNGs (see bug 27002), For defense in depth, omit include OnboardingTelemetry.jsm entirely. Added support for the following UITour page event: torBrowserOpenSecuritySettings Also fix bug 27403: the onboarding bubble is not always displayed. Arthur suggested to make the onboarding bubble visible on displays with less than 960px width available, so we choose 200px instead. Also fix bug 28628: Change onboarding Security panel to open new Security Level panel. Also fix bug 27484: Improve navigation within onboarding. Bug 27082: enable a limited UITour Disallow access to UITour functionality from all pages other than about:home, about:newtab, and about:tor. Implement a whitelist mechanism for page actions. Bug 26962 - implement new features onboarding (part 1). Add an "Explore" button to the "Circuit Display" panel within new user onboarding which opens the DuckDuckGo .onion and then guides users through a short circuit display tutorial. Allow a few additional UITour actions while limiting as much as possible how it can be used. Tweak the UITour styles to match the Tor Browser branding. All user interface strings are retrieved from Torbutton's browserOnboarding.properties file. Bug 27486 Avoid about:blank tabs when opening onboarding pages. Instead of using a simple <a href>, programmatically open onboarding web pages by using tabBrowser.addTab(). The same technique is now used for "See My Path", "See FAQs", and "Visit an Onion". Bug 29768: Introduce new features to users Add an "update" tour for the Tor Browser 8.5 release that contains two panels: Toolbar and Security (with appropriate description text and images). Display an attention-grabbing dot on the onboarding text bubble when the update tour is active. The animation lasts for 14 seconds. Bug 31768: Introduce toolbar and network settings changes in onboarding Update the "Tor Network" onboarding page to include a note that settings can now be accessed via the application preferences and add an "Adjust Your Tor Network Settings" action button which opens about:preferences#tor. Replace the Tor Browser 8.5 "update" onboarding tour with a 9.0 one that includes the revised "Tor Network" page and a revised "Toolbar" page. The latter explains that Torbutton's toolbar item has been removed ("Goodbye Onion Button") and explains how to access the New Identity feature using the hamburger menu and new toolbar item. Bug 34321 - Add Learn More onboarding item Bug 40429: Update Onboarding for 10.5 - - - - - a418b858 by sanketh at 2023-07-26T21:53:31+02:00 Bug 40209: Implement Basic Crypto Safety Adds a CryptoSafety actor which detects when you've copied a crypto address from a HTTP webpage and shows a warning. Closes #40209. Bug 40428: Fix string attribute names - - - - - 9a9c9cd0 by Mike Perry at 2023-07-26T21:53:32+02:00 TB3: Tor Browser's official .mozconfigs. Also: Add an --enable-tor-browser-data-outside-app-dir configure option Add --with-tor-browser-version configure option Bug 31457: disable per-installation profiles The dedicated profiles (per-installation) feature does not interact well with our bundled profiles on Linux and Windows, and it also causes multiple profiles to be created on macOS under TorBrowser-Data. Bug 31935: Disable profile downgrade protection. Since Tor Browser does not support more than one profile, disable the prompt and associated code that offers to create one when a version downgrade situation is detected. Add --enable-tor-browser-update build option Bug 40793: moved Tor configuration options from old-configure.in to moz.configure Bug 41584: Move some configuration options to base-browser level - - - - - 6ded175d by Henry Wilkes at 2023-07-26T21:53:33+02:00 Bug 41340: Enable TOR_BROWSER_NIGHTLY_BUILD features for dev and nightly builds tor-browser#41285: Enable fluent warnings. - - - - - 51dbd05f by Pier Angelo Vendrame at 2023-07-26T21:53:34+02:00 Bug 40562: Added Tor Browser preferences to 000-tor-browser.js Before reordering patches, we used to keep the Tor-related patches (torbutton and tor-launcher) at the beginning. After that issue, we decided to move them towards the end. In addition to that, we have decided to move Tor Browser-only preferences there, too, to make Base Browser-only fixups easier to apply. - - - - - e5a9e4bd by Pier Angelo Vendrame at 2023-07-26T21:53:35+02:00 Bug 13252: Customize profile management on macOS On macOS we allow both portable mode and system installation. However, in the latter case, we customize Firefox's directories to match the hierarchy we use for the portable mode. Also, display an informative error message if the TorBrowser-Data directory cannot be created due to an "access denied" or a "read only volume" error. - - - - - ac395e44 by Pier Angelo Vendrame at 2023-07-26T21:53:36+02:00 Bug 40933: Add tor-launcher functionality - - - - - b618f102 by Richard Pospesel at 2023-07-26T21:53:36+02:00 Bug 40597: Implement TorSettings module - migrated in-page settings read/write implementation from about:preferences#tor to the TorSettings module - TorSettings initially loads settings from the tor daemon, and saves them to firefox prefs - TorSettings notifies observers when a setting has changed; currently only QuickStart notification is implemented for parity with previous preference notify logic in about:torconnect and about:preferences#tor - about:preferences#tor, and about:torconnect now read and write settings thorugh the TorSettings module - all tor settings live in the torbrowser.settings.* preference branch - removed unused pref modify permission for about:torconnect content page from AsyncPrefs.jsm Bug 40645: Migrate Moat APIs to Moat.jsm module - - - - - 7ac8e77e by Alex Catarineu at 2023-07-26T21:53:37+02:00 Bug 10760: Integrate TorButton to TorBrowser core Because of the non-restartless nature of Torbutton, it required a two-stage installation process. On mobile, it was a problem, because it was not loading when the user opened the browser for the first time. Moving it to tor-browser and making it a system extension allows it to load when the user opens the browser for first time. Additionally, this patch also fixes Bug 27611. Bug 26321: New Circuit and New Identity menu items Bug 14392: Make about:tor behave like other initial pages. Bug 25013: Add torbutton as a tor-browser submodule Bug 31575: Replace Firefox Home (newtab) with about:tor - - - - - 7c6d4f43 by hackademix at 2023-07-26T21:53:38+02:00 Bug 8324: Prevent DNS proxy bypasses caused by Drag&Drop Bug 41613: Skip Drang & Drop filtering for DNS-safe URLs - - - - - 819219bd by Henry Wilkes at 2023-07-26T21:53:39+02:00 Bug 41600: Add a tor circuit display panel. - - - - - e576ab0c by Amogh Pradeep at 2023-07-26T21:53:40+02:00 Orfox: Centralized proxy applied to AbstractCommunicator and BaseResources. See Bug 1357997 for partial uplift. Also: Bug 28051 - Use our Orbot for proxying our connections Bug 31144 - ESR68 Network Code Review - - - - - 65039927 by Matthew Finkel at 2023-07-26T21:53:41+02:00 Bug 25741: TBA: Disable GeckoNetworkManager The browser should not need information related to the network interface or network state, tor should take care of that. - - - - - b4619b3c by Alex Catarineu at 2023-07-26T21:53:42+02:00 Add TorStrings module for localization - - - - - 59d28fa9 by Kathy Brade at 2023-07-26T21:53:43+02:00 Bug 14631: Improve profile access error messages. Instead of always reporting that the profile is locked, display specific messages for "access denied" and "read-only file system". To allow for localization, get profile-related error strings from Torbutton. Use app display name ("Tor Browser") in profile-related error alerts. - - - - - 7b4886ad by Pier Angelo Vendrame at 2023-07-26T21:53:43+02:00 Bug 40807: Added QRCode.js to toolkit/modules - - - - - b7b938b2 by Richard Pospesel at 2023-07-26T21:53:44+02:00 Bug 31286: Implementation of bridge, proxy, and firewall settings in about:preferences#connection This patch adds a new about:preferences#connection page which allows modifying bridge, proxy, and firewall settings from within Tor Browser. All of the functionality present in tor-launcher's Network Configuration panel is present: - Setting built-in bridges - Requesting bridges from BridgeDB via moat - Using user-provided bridges - Configuring SOCKS4, SOCKS5, and HTTP/HTTPS proxies - Setting firewall ports - Viewing and Copying Tor's logs - The Networking Settings in General preferences has been removed Bug 40774: Update about:preferences page to match new UI designs - - - - - 6b31946b by Richard Pospesel at 2023-07-26T21:53:45+02:00 Bug 27476: Implement about:torconnect captive portal within Tor Browser - implements new about:torconnect page as tor-launcher replacement - adds new torconnect component to browser - tor process management functionality remains implemented in tor-launcher through the TorProtocolService module - adds warning/error box to about:preferences#tor when not connected to tor Bug 40773: Update the about:torconnect frontend page to match additional UI flows. Bug 41608: Add a toolbar status button and a urlbar "Connect" button. - - - - - 3cbd1f50 by Arthur Edelstein at 2023-07-26T21:53:46+02:00 Bug 12620: TorBrowser regression tests Regression tests for Bug #2950: Make Permissions Manager memory-only Regression tests for TB4: Tor Browser's Firefox preference overrides. Note: many more functional tests could be made here Regression tests for #2874: Block Components.interfaces from content Bug 18923: Add a script to run all Tor Browser specific tests Regression tests for Bug #16441: Suppress "Reset Tor Browser" prompt. - - - - - 94185044 by Pier Angelo Vendrame at 2023-07-26T21:53:47+02:00 Bug 41668: Tweaks to the Base Browser updater for Tor Browser This commit was once part of "Bug 4234: Use the Firefox Update Process for Tor Browser.". However, some parts of it were not needed for Base Browser and some derivative browsers. Therefore, we extracted from that commit the parts for Tor Browser legacy, and we add them back to the patch set with this commit. - - - - - f56a18ae by Kathy Brade at 2023-07-26T21:53:48+02:00 Bug 12647: Support symlinks in the updater. - - - - - bc81fa25 by Kathy Brade at 2023-07-26T21:53:49+02:00 Bug 19121: reinstate the update.xml hash check This is a partial revert of commit f1241db6986e4b54473a1ed870f7584c75d51122. Revert most changes from Mozilla Bug 862173 "don't verify mar file hash when using mar signing to verify the mar file (lessens main thread I/O)." We kept the addition to the AppConstants API in case other JS code references it in the future. - - - - - 9e7bcd78 by Kathy Brade at 2023-07-26T21:53:50+02:00 Bug 16940: After update, load local change notes. Add an about:tbupdate page that displays the first section from TorBrowser/Docs/ChangeLog.txt and includes a link to the remote post-update page (typically our blog entry for the release). Always load about:tbupdate in a content process, but implement the code that reads the file system (changelog) in the chrome process for compatibility with future sandboxing efforts. Also fix bug 29440. Now about:tbupdate is styled as a fairly simple changelog page that is designed to be displayed via a link that is on about:tor. - - - - - c5fcad3e by Georg Koppen at 2023-07-26T21:53:50+02:00 Bug 32658: Create a new MAR signing key It's time for our rotation again: Move the backup key in the front position and add a new backup key. Bug 33803: Move our primary nightly MAR signing key to tor-browser Bug 33803: Add a secondary nightly MAR signing key - - - - - 90fefc2a by Mike Perry at 2023-07-26T21:53:51+02:00 Omnibox: Add DDG, Startpage, Disconnect, Youtube, Twitter; remove Amazon, eBay, bing eBay and Amazon don't treat Tor users very well. Accounts often get locked and payments reversed. Also: Bug 16322: Update DuckDuckGo search engine We are replacing the clearnet URL with an onion service one (thanks to a patch by a cypherpunk) and are removing the duplicated DDG search engine. Duplicating DDG happend due to bug 1061736 where Mozilla included DDG itself into Firefox. Interestingly, this caused breaking the DDG search if JavaScript is disabled as the Mozilla engine, which gets loaded earlier, does not use the html version of the search page. Moreover, the Mozilla engine tracked where the users were searching from by adding a respective parameter to the search query. We got rid of that feature as well. Also: This fixes bug 20809: the DuckDuckGo team has changed its server-side code in a way that lets users with JavaScript enabled use the default landing page while those without JavaScript available get redirected directly to the non-JS page. We adapt the search engine URLs accordingly. Also fixes bug 29798 by making sure we only specify the Google search engine we actually ship an .xml file for. Also regression tests. squash! Omnibox: Add DDG, Startpage, Disconnect, Youtube, Twitter; remove Amazon, eBay, bing Bug 40494: Update Startpage search provider squash! Omnibox: Add DDG, Startpage, Disconnect, Youtube, Twitter; remove Amazon, eBay, bing Bug 40438: Add Blockchair as a search engine Bug 33342: Avoid disconnect search addon error after removal. We removed the addon in #32767, but it was still being loaded from addonStartup.json.lz4 and throwing an error on startup because its resource: location is not available anymore. - - - - - 8db157b0 by Alex Catarineu at 2023-07-26T21:53:52+02:00 Bug 40073: Disable remote Public Suffix List fetching In https://bugzilla.mozilla.org/show_bug.cgi?id=1563246 Firefox implemented fetching the Public Suffix List via RemoteSettings and replacing the default one at runtime, which we do not want. - - - - - e9a16e05 by Richard Pospesel at 2023-07-26T21:53:53+02:00 Bug 23247: Communicating security expectations for .onion Encrypting pages hosted on Onion Services with SSL/TLS is redundant (in terms of hiding content) as all traffic within the Tor network is already fully encrypted. Therefore, serving HTTP pages from an Onion Service is more or less fine. Prior to this patch, Tor Browser would mostly treat pages delivered via Onion Services as well as pages delivered in the ordinary fashion over the internet in the same way. This created some inconsistencies in behaviour and misinformation presented to the user relating to the security of pages delivered via Onion Services: - HTTP Onion Service pages did not have any 'lock' icon indicating the site was secure - HTTP Onion Service pages would be marked as unencrypted in the Page Info screen - Mixed-mode content restrictions did not apply to HTTP Onion Service pages embedding Non-Onion HTTP content This patch fixes the above issues, and also adds several new 'Onion' icons to the mix to indicate all of the various permutations of Onion Services hosted HTTP or HTTPS pages with HTTP or HTTPS content. Strings for Onion Service Page Info page are pulled from Torbutton's localization strings. - - - - - 55675a15 by Kathy Brade at 2023-07-26T21:53:54+02:00 Bug 30237: Add v3 onion services client authentication prompt When Tor informs the browser that client authentication is needed, temporarily load about:blank instead of about:neterror and prompt for the user's key. If a correctly formatted key is entered, use Tor's ONION_CLIENT_AUTH_ADD control port command to add the key (via Torbutton's control port module) and reload the page. If the user cancels the prompt, display the standard about:neterror "Unable to connect" page. This requires a small change to browser/actors/NetErrorChild.jsm to account for the fact that the docShell no longer has the failedChannel information. The failedChannel is used to extract TLS-related error info, which is not applicable in the case of a canceled .onion authentication prompt. Add a leaveOpen option to PopupNotifications.show so we can display error messages within the popup notification doorhanger without closing the prompt. Add support for onion services strings to the TorStrings module. Add support for Tor extended SOCKS errors (Tor proposal 304) to the socket transport and SOCKS layers. Improved display of all of these errors will be implemented as part of bug 30025. Also fixes bug 19757: Add a "Remember this key" checkbox to the client auth prompt. Add an "Onion Services Authentication" section within the about:preferences "Privacy & Security section" to allow viewing and removal of v3 onion client auth keys that have been stored on disk. Also fixes bug 19251: use enhanced error pages for onion service errors. - - - - - bb22ce38 by Alex Catarineu at 2023-07-26T21:53:55+02:00 Bug 21952: Implement Onion-Location Whenever a valid Onion-Location HTTP header (or corresponding HTML <meta> http-equiv attribute) is found in a document load, we either redirect to it (if the user opted-in via preference) or notify the presence of an onionsite alternative with a badge in the urlbar. - - - - - 9d8c5bda by Pier Angelo Vendrame at 2023-07-26T21:53:56+02:00 Bug 40458: Implement .tor.onion aliases We have enabled HTTPS-Only mode, therefore we do not need HTTPS-Everywhere anymore. However, we want to keep supporting .tor.onion aliases (especially for securedrop). Therefore, in this patch we implemented the parsing of HTTPS-Everywhere rulesets, and the redirect of .tor.onion domains. Actually, Tor Browser believes they are actual domains. We change them on the fly on the SOCKS proxy requests to resolve the domain, and on the code that verifies HTTPS certificates. - - - - - 29f22cba by Pier Angelo Vendrame at 2023-07-26T21:53:57+02:00 Bug 11698: Incorporate Tor Browser Manual pages into Tor Browser This patch associates the about:manual page to a translated page that must be injected to browser/omni.ja after the build. The content must be placed in chrome/browser/content/browser/manual/, so that is then available at chrome://browser/content/manual/. We preferred giving absolute freedom to the web team, rather than having to change the patch in case of changes on the documentation. - - - - - 08860a4e by Pier Angelo Vendrame at 2023-07-26T21:53:58+02:00 Bug 41435: Add a Tor Browser migration function For now this function only deletes old language packs for which we are already packaging the strings with the application. - - - - - 6c1f46dd by Dan Ballard at 2023-07-26T21:53:59+02:00 Bug 40701: Add security warning when downloading a file Shown in the downloads panel, about:downloads and places.xhtml. - - - - - 44a9f9a8 by Henry Wilkes at 2023-07-26T21:53:59+02:00 Bug 41736: Customize toolbar for tor-browser. - - - - - 855b6979 by hackademix at 2023-07-26T21:54:00+02:00 Bug 41728: Pin bridges.torproject.org domains to Let's Encrypt's root cert public key - - - - - 30 changed files: - .eslintignore - .gitignore - + .gitlab/issue_templates/Backport Android Security Fixes.md - + .gitlab/issue_templates/Rebase Browser - Alpha.md - + .gitlab/issue_templates/Rebase Browser - Stable.md - + .gitlab/issue_templates/UXBug.md - + .gitlab/issue_templates/bug.md - + .gitlab/merge_request_templates/default.md - accessible/ipc/DocAccessibleParent.cpp - + browser/actors/AboutTBUpdateChild.jsm - + browser/actors/AboutTBUpdateParent.jsm - + browser/actors/CryptoSafetyChild.jsm - + browser/actors/CryptoSafetyParent.jsm - browser/actors/NetErrorChild.jsm - browser/actors/NetErrorParent.jsm - − browser/actors/RFPHelperChild.jsm - − browser/actors/RFPHelperParent.jsm - browser/actors/moz.build - browser/app/Makefile.in - browser/app/macbuild/Contents/Info.plist.in - browser/app/macbuild/Contents/MacOS-files.in - browser/app/permissions - + browser/app/profile/000-tor-browser.js - + browser/app/profile/001-base-profile.js - browser/app/profile/firefox.js - browser/base/content/aboutDialog-appUpdater.js - browser/base/content/aboutDialog.js - browser/base/content/aboutDialog.xhtml - + browser/base/content/abouttbupdate/aboutTBUpdate.css - + browser/base/content/abouttbupdate/aboutTBUpdate.js The diff was not included because it is too large. View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/4ce612… -- View it on GitLab: https://gitlab.torproject.org/tpo/applications/tor-browser/-/compare/4ce612… You're receiving this email because of your account on gitlab.torproject.org.

1 0