This patch was generated by running:
```
perl -p -i \
-e 's/^(\s+)([a-zA-Z0-9.]+) = NS_ConvertUTF8toUTF16\((.*)\);/\1CopyUTF8toUTF16(\3, \2);/;' \
-e 's/^(\s+)([a-zA-Z0-9.]+) = NS_ConvertUTF16toUTF8\((.*)\);/\1CopyUTF16toUTF8(\3, \2);/;' \
$FILE
```
against every .cpp and .h in mozilla-central, and then fixing up the
inevitable errors that happen as a result of matching C++ expressions with
regexes. The errors fell into three categories:
1. Calling the convert functions with `std::string::c_str()`; these were
changed to simply pass the string instead, relying on implicit conversion
to `mozilla::Span`.
2. Calling the convert functions with raw pointers, which is not permitted
with the copy functions; these were changed to invoke `MakeStringSpan` first.
3. Other miscellaneous errors resulting from over-eager regexes and/or the
replacement not being type-aware. These changes were reverted.
Differential Revision: https://phabricator.services.mozilla.com/D88903
Now that we've concluded no immediate action is needed for slow
ShellExecuteByExplorer, `SHELLEXECUTEBYEXPLORER_DURATION_MS` is
no longer needed.
Differential Revision: https://phabricator.services.mozilla.com/D86143
Not just the one that takes a channel. This fixes the issue.
Also, as an optimization, avoid creating a timer if we're used already
or we have already one.
Differential Revision: https://phabricator.services.mozilla.com/D87267
Also improve detection of types needing overwrite by using the final extension settled on for the
type.
I didn't fold PDF fully into this part, as it has different conditions for showing the radio button vs.
viewing internally from the downloads list.
Also fix intermittent bug 1641774 (wouldn't run in verify) by saving handler info at the beginning
of browser_download_open_with_internal_handler, instead of at the start of a subtest.
Differential Revision: https://phabricator.services.mozilla.com/D86651
Also improve detection of types needing overwrite by using the final extension settled on for the
type.
I didn't fold PDF fully into this part, as it has different conditions for showing the radio button vs.
viewing internally from the downloads list.
Differential Revision: https://phabricator.services.mozilla.com/D86651
This patch adds a new measurement to the histogram about the elapsed time
of `ShellExecuteByExplorer`. If ShellExecuteByExplorer takes long in
a considerable number of instances, we need to consider fixing bug 1646986.
Differential Revision: https://phabricator.services.mozilla.com/D84407
Having two classes in the inheritance chain inherit from SupportsWeakPtr
now won't compile, but you can use WeakPtr<Derived> when any base class
inherits from SupportsWeakPtr.
Differential Revision: https://phabricator.services.mozilla.com/D83674
The aim of the code we're modifying here is to block things in one browsingcontext
tree from opening external links in another browsingcontext tree (and causing the
external protocol dialog to show up for that tab/window) -- except if the other
browsingcontext into which something is being loaded is same-origin.
Unfortunately the pre-patch code assumed that it would find currentWindowGlobal
objects for each browsingcontext, and it turns out that's not guaranteed,
especially in the case of hidden iframes, which turn out to be quite commonly
used for external protocol launches.
This patch fixes this by continuing to move towards the root of the browsingcontext
tree even if there's no currentWindowGlobal (though logically speaking, this
should only be necessary for the first iteration of the loop, it seems easier to
just always check this). It also adds a test for this behaviour working.
Differential Revision: https://phabricator.services.mozilla.com/D83015
The aim of the code we're modifying here is to block things in one browsingcontext
tree from opening external links in another browsingcontext tree (and causing the
external protocol dialog to show up for that tab/window) -- except if the other
browsingcontext into which something is being loaded is same-origin.
Unfortunately the pre-patch code assumed that it would find currentWindowGlobal
objects for each browsingcontext, and it turns out that's not guaranteed,
especially in the case of hidden iframes, which turn out to be quite commonly
used for external protocol launches.
This patch fixes this by continuing to move towards the root of the browsingcontext
tree even if there's no currentWindowGlobal (though logically speaking, this
should only be necessary for the first iteration of the loop, it seems easier to
just always check this). It also adds a test for this behaviour working.
Differential Revision: https://phabricator.services.mozilla.com/D83015
CLOSED TREE
Backed out changeset 51d7c644a1e6 (bug 1650163)
Backed out changeset 3d2b6908447a (bug 1650163)
Backed out changeset 79141707d47b (bug 1650163)
The WPT a-download-click-404.html requires this behaviour for links with the download attribute, and this is also the current behaviour for Content-Disposition: Attachment (see bug 1604308).
This previously worked because the parent process version of nsDocumentOpenInfo failed (with NS_ERROR_FILE_NOT_FOUND), but the error code was discarded and we forwarded the channel to the content process. The content process version then would then return NS_ERROR_WONT_HANDLE_CONTENT since the load requires downloading, but we don't allow that in the content process. This new error code is one that doesn't have an associated error page (unlilke the original error), so was silently discarded.
Differential Revision: https://phabricator.services.mozilla.com/D81014
When we hit the stylesheet cache for an existing stylesheet in a
different document:
* There's no existing preload in the PreloadService (because the sheet
was loaded in another document).
* We don't create a PreloaderBase (because we don't trigger a load at
all).
* And thus we believe the load has just errored, which is wrong...
Fix it by properly passing through the "already complete" status from
PreloadStyle.
Note that there's still another issue, which is that we'd still hit this
broken path if two stylesheets with the same URI are loading in
different documents and the CSS loader coalesces them.
For that, I think we'd have to make SheetLoadData the thing that
actually implements PreloaderBase. This is kind of an obscure case, and
SheetLoadData has a pretty complex lifetime already (keeps alive a whole
lot of stuff), so I'd prefer to do this in a separate bug.
Differential Revision: https://phabricator.services.mozilla.com/D80289
We probably should remove the referrer policy check altogether in
bug 1642591, but the preload code still uses the whole referrer as the
key for css, which the css loader stopped doing in bug 1599160.
So this should be an uncontroversial simplification.
Differential Revision: https://phabricator.services.mozilla.com/D79810
This removes the diagnostic warnings which used to be logged when the
Large-Allocation header was present, but failed to switch into a
Large-Allocation process. Due to the low adoption of the header, this shouldn't
be too large of a problem, but we can look into re-adding the diagnostics if
needed in the future.
The new codepath no longer performs multiple network requests for
Large-Allocation resources, and now relies on the battle-tested
DocumentLoadListener codepath for process switching.
Differential Revision: https://phabricator.services.mozilla.com/D78998
This removes the diagnostic warnings which used to be logged when the
Large-Allocation header was present, but failed to switch into a
Large-Allocation process. Due to the low adoption of the header, this shouldn't
be too large of a problem, but we can look into re-adding the diagnostics if
needed in the future.
The new codepath no longer performs multiple network requests for
Large-Allocation resources, and now relies on the battle-tested
DocumentLoadListener codepath for process switching.
Differential Revision: https://phabricator.services.mozilla.com/D78998
This patch adds the following constraints:
- close the dialog when the originating context navigates;
- close the dialog when the originating context closes / goes away;
- do not allow more than 1 dialog at a time;
Differential Revision: https://phabricator.services.mozilla.com/D77029
I'm about to introduce the concept of "Loader principal" (as in "the
principal of the CSS loader"), and SheetLoadData already has an
mLoaderPrincipal.
However SheetLoadData's principal is just the triggering principal (the
principal that initiated the load). So name it that with consistency
with SheetInfo::mTriggeringPrincipal etc.
Differential Revision: https://phabricator.services.mozilla.com/D77613
There's no use case for stateful comparators, so they can be just plain
function pointers.
This is used in some hot places like CSS selector matching.
Differential Revision: https://phabricator.services.mozilla.com/D77084
This was a bit of pain to track down, because some code was allowing an ID of 0, and creating a request context to match, but other code was treating 0 as invalid.
Differential Revision: https://phabricator.services.mozilla.com/D76652
Which is the spec term. nsIStyleSheetLinkingElement is even more
confusing since it may not be an element at all (see: processing
instructions).
Differential Revision: https://phabricator.services.mozilla.com/D76071
In this bug we're moving away from monolithic JNI headers to class-specific
headers so that we don't have to rebuild the world every time we make a change
to a JNI interface.
Differential Revision: https://phabricator.services.mozilla.com/D75367
Previously this would typically set GRE_HOME to /data/data/org.mozilla.geckoview.test,
even though org.mozilla.geckoview.test is not normally installed when running
xpcshell tests -- a non-existent directory in a privileged location!
The new location, remoteBinDir, is typically /data/local/xpcb, the location of the
xpcshell executable.
Differential Revision: https://phabricator.services.mozilla.com/D74938
Prior to this patch, PDF.js tracks both its own 'disabled' pref (which is used
by enterprise policy) and whether it is the default handler per the handler
service - but it tracks both in one bool, which determines whether its
streamconverter registers.
Really, what we want is to never use PDF.js if it's preffed off.
However, if there is some other default, it should be acceptable to use PDF.js
in some circumstances, like for <embed> or <object>s where otherwise we
would show no content at all.
Even for toplevel PDFs, if the user has configured Firefox to open PDFs in
an external helper app which is Firefox (which is currently an easy mistake
to make in the unknownContentType dialog), or has it set to the OS default,
but has changed their OS default to Firefox, we really still want to open
those PDFs with PDF.js.
This patch fixes all of this by splitting out the pref tracking from the
handler state tracking. Only the pref will completely disable PDF.js.
Then, in the streamconverter code, we check whether PDF.js should be used for
PDFs, and if there's a misconfiguration that we can correct. This code is
invoked from the parent process when we load PDFs in frames or toplevel
documents, and will prevent us from invoking PDF.js in the child if the user
would prefer that not to happen.
As a driveby, this cleans up how we track the pref inside PDF.js, and how we
get notified of changes to the handler - we were missing changes made in the
unknown content type dialog, so it seemed worth making it generic.
Differential Revision: https://phabricator.services.mozilla.com/D73510
This also renames `saveToDisk` to the more accurate `promptForSaveDestination`
given that it doesn't actually do the saving itself.
Differential Revision: https://phabricator.services.mozilla.com/D73507
Raw Cr.ERROR don't get stack information, same as throwing JS literals instead
of `new Error()`s.
This was done automatically with a new eslint rule that will be introduced in
the next commit. One instance of a raw Cr.ERROR was not replaced since it is
used in a test that specifically checks the preservation of raw Cr values in
XPCJS. The rule will be disabled for that instance.
Differential Revision: https://phabricator.services.mozilla.com/D28073
There are many limitations currently, but this prototype should successfully
render most basic AVIF images. Known limitations:
- No support for any derived image items (crop, rotate, etc.)
- No support for alpha planes
- No support for ICC profiles (bug 1634741)
- The primary image item must be an av01 (no grid support)
- HDR images aren't tone-mapped
Differential Revision: https://phabricator.services.mozilla.com/D68498
There are many limitations currently, but this prototype should successfully
render most basic AVIF images. Known limitations:
- No support for any derived image items (crop, rotate, etc.)
- No support for alpha planes
- The primary image item must be an av01 (no grid support)
- HDR images aren't tone-mapped
Differential Revision: https://phabricator.services.mozilla.com/D68498
There are many limitations currently, but this prototype should successfully
render most basic AVIF images. Known limitations:
- No support for any derived image items (crop, rotate, etc.)
- No support for alpha planes
- The primary image item must be an av01 (no grid support)
- HDR images aren't tone-mapped
Differential Revision: https://phabricator.services.mozilla.com/D68498
Also move MOZ_MUST_USE before function declarations' specifiers and return type. While clang and gcc's __attribute__((warn_unused_result)) can appear before, between, or after function specifiers and return types, the [[nodiscard]] attribute must precede the function specifiers.
I also needed to wrap the nsDocLoader::RemoveChildLoader() function declaration with // clang-format off and // clang-format on. clang-format insists on indenting the function declaration because it is confused by the NS_DECL_ macros without trailing semicolons before the [[nodiscard]] attribute.
Differential Revision: https://phabricator.services.mozilla.com/D70478
--HG--
extra : moz-landing-system : lando
Mostly a matter of:
rg -l '\->LoadingPrincipal' | xargs sed -i 's/->LoadingPrincipal/->GetLoadingPrincipal/g'
And then clang-format. But I tweaked manually nsHttpChannelAuthProvider (move
the variable where it's used, don't take a useless strong ref),
AddonContentPolicy (move the declaration of the variable to the if condition),
and BackgroundUtils (same).
Differential Revision: https://phabricator.services.mozilla.com/D69828
--HG--
extra : moz-landing-system : lando
Mostly a matter of:
rg -l '\->LoadingPrincipal' | xargs sed -i 's/->LoadingPrincipal/->GetLoadingPrincipal/g'
And then clang-format. But I tweaked manually nsHttpChannelAuthProvider (move
the variable where it's used, don't take a useless strong ref),
AddonContentPolicy (move the declaration of the variable to the if condition),
and BackgroundUtils (same).
Differential Revision: https://phabricator.services.mozilla.com/D69828
--HG--
extra : moz-landing-system : lando
This fixes 1597154 because the lists of URIs here and in nsDocShell now
match.
Note that they don't exactly match, there's an exception for about:newtab
which we'll fix in a follow-up bug.
Differential Revision: https://phabricator.services.mozilla.com/D69136
--HG--
extra : moz-landing-system : lando
It's only used to check whether the docshell is gone already, and we need to
deal with that in other places anyway. No other useful implementation.
In comm-central there are a couple, but they're useless too.
Differential Revision: https://phabricator.services.mozilla.com/D68731
--HG--
extra : moz-landing-system : lando
We use `ShellExecuteByExplorer` to open a uri because applications may not
support the mitigation policies inherited from our process.
Unlike Firefox, Thunderbird runs `ShellExecuteByExplorer` to launch a browser
to open an HTTP/HTTPS uri. If a uri includes credentials like
http://user:password@domain.com/, `ShellExecuteByExplorer` succceds, but
explorer.exe refuses to open the uri by design. In this case the fallback to
`ShellExecuteExW` does not happen because the API is async and no error is
returned
The proposed fix is to skip `ShellExecuteByExplorer` if it's Thunderbird.
Because Thunderbird does not apply additional mitigation policies which may
cause compat issues, we don't need to run `ShellExecuteByExplorer` in the
first place.
Differential Revision: https://phabricator.services.mozilla.com/D67339
--HG--
extra : moz-landing-system : lando
Only 3 callers are using a non-UTF-8 charset as the first parameter.
* MediaDocument.cpp: This does not make sense because the "filename" part of
URLs will always be encoded with UTF-8.
* nsContextMenu.js: This is wrong because "mailto:" URLs don't care about the
document charset.
* Finder.jsm: This caused bug 1623222.
Differential Revision: https://phabricator.services.mozilla.com/D67386
--HG--
extra : moz-landing-system : lando
Note that this also suppresses notifications from the initial about:blank in the new process, and updates the tabbrowser to not expect those.
Differential Revision: https://phabricator.services.mozilla.com/D65923
--HG--
extra : moz-landing-system : lando
If a system uses a custom shell instead of Windows Explorer,
`ShellExecuteByExplorer` always fails because it relies on explorer.exe,
but `ShellExecute` stil works because it's in-proc implementation.
We added a fallback for a local file to address Bug 1602726. This patch adds
a fallback for a uri.
Differential Revision: https://phabricator.services.mozilla.com/D63095
--HG--
extra : moz-landing-system : lando
This also fixes the Window lookup in nsExternalAppHandler, which was the reason for previously needing the process-top BrowsingContext.
Differential Revision: https://phabricator.services.mozilla.com/D64910
--HG--
extra : moz-landing-system : lando
Given that we are going to add ContentBlockingAllowList in
CookieSettings, so CookieSettings will be responsible for more stuff than the
cookie behavior and cookie permission. We should use a proper name to
reflect the purpose of it. The name 'CookieSettings' is misleading that
this is only for cookie related stuff. So, we decide to rename
'CookieSettins' to 'CookieJarSettings' which serves better meaning here.
Differential Revision: https://phabricator.services.mozilla.com/D63935
--HG--
rename : netwerk/cookie/CookieSettings.cpp => netwerk/cookie/CookieJarSettings.cpp
rename : netwerk/cookie/nsICookieSettings.idl => netwerk/cookie/nsICookieJarSettings.idl
extra : moz-landing-system : lando
This covers most cycle collected objects which support weak references, but
not the ones which inherit from a cycle collected class and don't do any cycle
collection on their own.
Differential Revision: https://phabricator.services.mozilla.com/D63962
--HG--
extra : moz-landing-system : lando
This removes the need for explicit #ifdef NS_BUILD_REFCNT_LOGGING without
introducing user-defined destructors when it is not defined.
Also, some uses of virtual for declaring destructors are replaced by the
appropriate override declaration through these changes.
Differential Revision: https://phabricator.services.mozilla.com/D62604
--HG--
extra : moz-landing-system : lando
This removes the need for explicit #ifdef NS_BUILD_REFCNT_LOGGING without
introducing user-defined destructors when it is not defined.
Also, some uses of virtual for declaring destructors are replaced by the
appropriate override declaration through these changes.
Differential Revision: https://phabricator.services.mozilla.com/D62604
--HG--
extra : moz-landing-system : lando
Thunderbird has no default handlers so test_check_defaults_get_added does
nothing and test_check_default_modification fails looking for a mailto handler.
Differential Revision: https://phabricator.services.mozilla.com/D62525
--HG--
extra : moz-landing-system : lando
And fix a bunch of fallout as we now don't include nsDocShell.h everywhere.
Differential Revision: https://phabricator.services.mozilla.com/D62169
--HG--
extra : moz-landing-system : lando
Sebastian Streich
Matt Woodrow
Bogdan Tara
Adam Gashlin
Barret Rennie
Dimi Lee
Henri Sivonen
Razvan Maries
Nathan Froyd
pbz
Gijs Kruitbosch
Toshihito Kikuchi
Narcis Beleuzu
Tom Schuster
Kershaw Chang
Emilio Cobos Álvarez
Cosmin Sabou
Dorel Luca
Frederik Braun
Simon Giesecke
Noemi Erli
Joel Maher
owlishDeveloper
Cameron McCormack
Marco Bonardo
Kris Maglione
Nika Layzell
Mihai Alexandru Michis
Butkovits Atila
Honza Bambas
Marco Castelluccio
Alexis Beingessner
Jared Wein
Mark Striemer
Valentin Gosu
Aaron Klotz
Geoff Brown
ssengupta
Geoff Lankow
Edgar Chen
Ian Moody
Jon Bauman
sagudev
Jean-Yves Avenard
Kagami Sascha Rosylight
Chris Peterson
Dylan Roeh
Eric Rahm
Christoph Kerschbaumer
Paul Bone
sonakshi
Andrea Marchesini
Masatoshi Kimura
Coroiu Cristina
Boris Zbarsky
Sylvestre Ledru
Mark Banner
Tim Huang
Andreea Pavel
shindli