This makes the IPC messages a little bigger, but that's unavoidable.
MozReview-Commit-ID: 1oPz2Yjjd9y
--HG--
extra : rebase_source : 0cff8cf5b25f66b73f6864ce50c1e5f575026ec3
Preferences::SetPreference() is used when setting prefs in the content process
from dom::Pref values that are passed from the parent process. Currently we
use the high-level Set*InAnyProcess() methods to do this -- basically the same
code path as sets done via the API -- but this has several problems.
- It is subtly broken. If the content process already has a pref of type T with
a default value and then we get a SetPreference() call that tries to change
it to type U, it will erroneously fail. In practice this never(?) happens,
but it shows that things aren't arranged very well.
- SetPreference() also looks up the hashtable twice to get the same pref if
both a default value and a user value are present in the dom::Pref.
- This happens in content processes, while all other pref set operations occur
in the parent process. This is the main reason we have the Set*InAnyProcess()
functions.
In short, the setting of prefs via IPC is quite different to the setting of
prefs via other means -- because it happens in content processes, and it's more
of a clobber that can set both values at once -- and so should be handled
differently.
The solution is to make Preferences::SetPreference() use lower-level operations
to do the update. It now does the hash table lookup/add itself, and then calls
the new Pref::FromDomPref() method. That method then possibly calls the new
PrefValue::FromDomPrefValue() method for both kinds of value, and overwrites
them as necessary. SetValueFromDom() is no longer used and the patch removes it.
MozReview-Commit-ID: 2Rg8VMOc0Cl
--HG--
extra : rebase_source : 0eddc3a4b694a79af3e173aefa7758f8e2ae776b
And remove the type argument from PrefValue's constructor. This is needed
for the next patch.
MozReview-Commit-ID: Ls8hEU2uRQQ
--HG--
extra : rebase_source : 115828e219f6bbe04677ffc106068a662458481a
The nice thing about this is that the memory management of the strings
(moz_xstrdup() and free()) is now entirely within PrefValue.
MozReview-Commit-ID: KJjnURpmgfE
--HG--
extra : rebase_source : 39c058cddf5ebf9e19f9151f40fd507f6909a289
It's something of an obfuscation, and it forces together various bool values
that don't necessarily have to be together (and won't be together after future
refactorings).
The patch also reorders some function arguments for consistency: PrefType, then
PrefValueKind, then PrefValue.
MozReview-Commit-ID: KNY0Pxo0Nxf
--HG--
extra : rebase_source : d46d228c3b13549b2159757dcdaf9583cca828f7
Although it is a subclass of PLDHashEntryHdr, it's the main representation of a
pref, so the name should reflect that.
MozReview-Commit-ID: 5qJNQtjbFmH
--HG--
extra : rebase_source : f2bd77a57c4e2a48e24ead736f15856fbeb9f718
As is done in pref_SavePrefs().
The confusion here is because a Vector can fill 100% of its capacity, but a
hash table cannot go past 75% of its capacity.
MozReview-Commit-ID: 5JMbmtrxMGN
--HG--
extra : rebase_source : 5ce1ce9dd0259588a0df924c2b45c39497b1ce71
It represents a pref, so `Pref` is a better name. Within Preferences.cpp the
patch uses domPref/aDomPref to distinguish it from PrefHashEntry values.
MozReview-Commit-ID: HXTl0GX4BtO
--HG--
extra : rebase_source : c1e0726c55e7577720f669f0ed2dbc38627d853e
This factors out some common code from SetValue(), making it easier to read.
The patch also improves the comments in SetValue().
MozReview-Commit-ID: 60JnBlIS1q6
--HG--
extra : rebase_source : cc0e47eb556ab87549137777625856db12782702
High resolution, high framerate was disabled by default on old AMD cards on the provisio that it was bad. But this assumes that the CPU decoder could do it better.
This assumption appears fragile at best, as CPU with those old adapter are likely to be old and underpower to start with.
Chrome doesn't appear to restrict use of those cards to a given resolution.
So we disable this restriction, while making it user configurable.
MozReview-Commit-ID: HhADHNR0FdJ
--HG--
extra : rebase_source : ece39cd9b84c6e372d1002ee12e72523cee3d04d
Currently, you can create a pref that only has a user value, and then later
give it a default value with a different type. The entire pref is then recorded
as having this second type. This causes problems later when interpreting the
user value.
This patch makes SetValue() fail if it tries to set a default value whose type
differs from an existing user value. It also expands an existing test to cover
this case and some similar ones.
MozReview-Commit-ID: 89tvISQ7RNT
--HG--
extra : rebase_source : 6cf34da0ff24f5b90a88003445a4a7c88b1f3907
Before bug 1367813, the code in test_stickyprefs.js, using
readUserPrefs, would end up changing the file under which the
preferences service would save user prefs, making xpcshell
possibly overwrite the test data files at the end of the test run.
A hack was put in place to avoid this, that is not required anymore.
--HG--
extra : rebase_source : 6a4b1442c0d1593322226ad162c421df4f9c7f87
This requires adding an aPriority argument (defaulting to false) to
Preferences::RegisterCallback(). And RegisterVarCacheCallback() is no longer
necessary.
MozReview-Commit-ID: BMDk3HuaQVV
--HG--
extra : rebase_source : 17a61cfd9a82f24854162fc993223691041ea46d
- Ensure ovr_GetSessionStatus is polled even when a VR presentation
is not active.
- When we fail to initialize an Oculus Session or detect VR hardware,
immediately unload the Oculus Library as we can't poll for ShouldQuit
without a valid Oculus session.
- When we poll ovr_GetSessionStatus, we are now updating the mounted state
in VRDisplayInfo::mIsMounted.
- Added prefs to control enumeration throttling and timeout to release
VR hardware when inactive.
- Some refactoring to make frame loop more understandable and less
brittle.
- When throttling enumeration, we ensure that all other VR apis
also throttle enumeration so that they don't pick up the same device
during throttling.
- Some long functions in VRManager have been broken up and
had their inner-workings documented in more detail.
MozReview-Commit-ID: CEYwwQ9mYd0
--HG--
extra : rebase_source : b82cd9f200e6311528d4bed469d1b8044e9fc7f2
extra : amend_source : e74f56f5ec95641fd478deb71d6c7ba5d2cba7b1
In practice we always use the same functions for these purposes.
MozReview-Commit-ID: 4Be9pRhUeff
--HG--
extra : rebase_source : 3dfafd9479371d3a47ec263a66942ddbfbefdb46
This patch makes it a proper class, and moves existing functions into it.
MozReview-Commit-ID: 5pbT3ljq44R
--HG--
extra : rebase_source : ac7ba98f9d39b3cd6f71498a5e108cb6757034e0
And use new/delete for them. And make mDomain a unique pointer so it doesn't
need explicit deallocation.
MozReview-Commit-ID: E1jLccXaSwT
--HG--
extra : rebase_source : 5a64135d5471297ab98f8ec4557f66dac8b7eff9
Maybe<PrefType> is a better way of representing "no type".
MozReview-Commit-ID: Fnha5RxbNg4
--HG--
extra : rebase_source : 8e8322b0443305ab71acd6d98ea2607f626c5bce
This field isn't accessed outside PrefHashEntry.
MozReview-Commit-ID: IvwQe5UtjjJ
--HG--
extra : rebase_source : 5447d4e24bbf0d0637ee29377cc749b9b77b4e1e
We can move the setting into ReplaceValue() and ClearValue(), and then those
methods aren't needed any more.
The patch also removes some getter calls within PrefHashEntry by directly using
field names.
MozReview-Commit-ID: 42EAx1Kh9Et
--HG--
extra : rebase_source : 3393a80a9c5d2d7c660171cdda8d3914c35e96ea
There's an "XXX" comment suggesting a possible leak when string user values are
cleared. Turns out a lot of the time there won't be a leak, because the string
pointer isn't overwritten and ClearEntry() frees the string when the pref is
destroyed. However, if the user value is subsequently overwritten with a
different string, there will be a leak. Also, even in the non-leak case, we
currently hold onto the string for longer than necessary.
This patch introduces ClearUserValue() -- which frees the string when
appropriate -- and uses it in all the places where values are cleared.
MozReview-Commit-ID: ARuWUNzPTfy
--HG--
extra : rebase_source : 4567e37ba96ba3b4ae9b1972d887eeaed1257cb0
Specifically:
- rename it as NotifyCallbacks();
- remove the return value, because it is always NS_OK;
- some minor naming and declaration clean-ups.
MozReview-Commit-ID: GcH81owPLsp
--HG--
extra : rebase_source : 501a85f76bb823cc45dba8e4601584f5218b1a9e
This removes from all.js some dom, browser, autocomplete, ui, and other prefs
that are no longer accessed in our code.
MozReview-Commit-ID: JGoBZ2GUbPD
--HG--
extra : rebase_source : cc87f28ca547ac5568a6df0dd5c3561712dc149d
This removes prefs for printing that are only declared in all.js but not accessed
elsewhere. Also correct the name of "print.print_in_color".
MozReview-Commit-ID: GrHNUjCiijA
--HG--
extra : rebase_source : 2d778dd6e517c10b569ff98af8da1ed9b80308e0
It's a horrible method, but the horribleness was pre-existing. I hope to make
it nicer in a follow-up bug.
MozReview-Commit-ID: 3tMTEeBNVax
--HG--
extra : rebase_source : b21d80ed6468b7c22689df64c1ae1bdda49a2a06
This moves part of pref_savePrefs() into PrefHashEntry.
This requires moving StrEscape() higher up to avoid a forward declaration.
Note that clang-format insists on indenting an unrelated comment, I don't know
why.
MozReview-Commit-ID: 7gww3r7t9y4
--HG--
extra : rebase_source : 63f19bb2c556cc806cc6ce43e70ac6c95e6d67bc
And invert its sense, renaming as Equals(), because that is easier to think
about.
The patch also reorders some conditions so that
HasDefaultValue()/HasUserValue() is tested before Equals(). This isn't strictly
necessary, but it reads better.
MozReview-Commit-ID: JeGrevDwqKz
--HG--
extra : rebase_source : 575758e11d7e66226fe196f92f2dd355b1444d1d
This requires adding a constructor, which is used in a placement new.
MozReview-Commit-ID: 3v6843D5DMu
--HG--
extra : rebase_source : 780902ae8e761bdf1141f24bf5594ce910ea32b4
And rename it as ToSetting(). This requires also moving
AssignPrefValueToDomPrefValue into PrefHashEntry.
MozReview-Commit-ID: 8ntNZJt9rhv
--HG--
extra : rebase_source : 5dd5a2c570b2ceefb506b4051c414b06d27a95f2
This patch:
- partially inlines GetPrefValueFromEntry() at its two call sites, to avoid
repeating the kind check;
- renames it as AssignPrefValueToDomPrefValue().
MozReview-Commit-ID: BUBk3ML6h5O
--HG--
extra : rebase_source : eb12adf4af2ee05356af6b2b4f346aa3425ff438
PrefTypeFlags is a class with a lot of smarts (i.e. methods). PrefHashEntry is
a class with little smarts. This is silly, because PrefTypeFlags is essentially
an internal implementation detail of PrefHashEntry.
This patch merges PrefTypeFlags into PrefHashEntry, so that PrefHashEntry has
all the smarts. This means lots of `pref->mPrefFlags.Foo()` calls become
`pref->Foo()`.
The patch also changes the representation of the type and flags within
PrefHashEntry to use bitfields, which avoids the need for a Flags type and is
much simpler than the old approach.
MozReview-Commit-ID: 4Yt9OtBzh9e
--HG--
extra : rebase_source : 18fc01217aeaee7057b3d9a54b18cae45be5538f
They are guaranteed to be zeroed already by PLDHashTable.
MozReview-Commit-ID: LBeGa9ABYc3
--HG--
extra : rebase_source : f767d0d20cfb1c187496be20fff4cce43c768ccd
This commit removes the `layers.omtp.force-sync` preference and replaces
it with a preprocessor define that is commented out. This commit also
changes the behavior of force-sync so that it also does synchronization
with CompositorBridgeChild like normal OMTP. This simplifies the code
and makes using a preprocessor define easier.
MozReview-Commit-ID: 6RfuFTFBdMh
--HG--
extra : rebase_source : 0778a3087324b9e87f44587efbb49c71edf47f23
Because they are the pref's values, not the pref itself.
MozReview-Commit-ID: 1N4Fh2TF8lL
--HG--
extra : rebase_source : 5645a0e5897fc825e6bbc6da2c6d40fb852a9efb
It's the user value that's sticky, not the default value. Though we typically
talk about the entire pref being sticky, so that's what this patch does.
MozReview-Commit-ID: 8THuRCTZ7uw
--HG--
extra : rebase_source : 8164f71065bcae1376c3e2bf241593c62c90b702
This matches the corresponding string "sticky_pref".
MozReview-Commit-ID: 8bqM5x8bIoG
--HG--
extra : rebase_source : 20b5b8b8df8b936a22849766552d11f659ffc84b
So that "default value" and "user value" are used consistently.
MozReview-Commit-ID: Hmfct8STu33
--HG--
extra : rebase_source : 3f27992630e6e834ce470693a15d8f6b99706cfe
This makes it like the VarCache functions for int, uint, atomic unit, and float.
MozReview-Commit-ID: 4CRuvAjqGCj
--HG--
extra : rebase_source : 5780456f1f0cf18f98ab3b42efcaec514324a171
Because it holds 8-bit strings, which only need 1 byte alignment.
For the profile on my Linux box, in every process this reduces the size of the
arena by 1 chunk, from 120 KiB to 112 KiB.
MozReview-Commit-ID: 8ozrLIq5ZZ4
--HG--
extra : rebase_source : 1f7d337875d5f554772bec6395e7f4241900a26c
This patch renames it as gInstallingCallback, adds a comment, and reduces the
scope of the AUTO_INSTALLING_CALLBACK() objects to the minimum.
MozReview-Commit-ID: EGfnBPVFtmw
--HG--
extra : rebase_source : 9feafbb4c92c6b9dbe37466fefb777a8638b1c6e
This splits the measurements into several buckets, like so:
> 718,528 B (00.40%) -- preferences
> ├──262,176 B (00.14%) ── hash-table
> ├──181,952 B (00.10%) ── callbacks
> ├──122,880 B (00.07%) ── pref-name-arena
> ├───91,872 B (00.05%) ── root-branches
> ├───38,296 B (00.02%) ── string-values
> ├───21,272 B (00.01%) ── cache-data
> └───────80 B (00.00%) ── misc
The patch also measures some things that were previously overlooked.
- String pref values. (The old code had a comment that incorrectly claimed they
were allocated out of an arena.)
- The PrefCallback objects pointed to by entries in nsPrefBranch::mObservers.
And it makes the code more like typical reporters.
- It removes the "AndOtherStuff" from Preferences' measuring method, and
measuring those global structures in
PreferenceServiceReporter::CollectReports().
- It adds `const` where appropriate.
MozReview-Commit-ID: dyNg7ldQdh
This patch also adds some Set*InAnyProcess() methods, and makes nsPrefBranch a
friend of Preferences so it can call those methods.
And it moves the thin Set*() wrapper functions to Preferences.h, alongside
SetUint().
MozReview-Commit-ID: 88HhmcTFZNc
--HG--
extra : rebase_source : 88a854d52afce86d93008a6e1a4b5f32bcf24a1a
This will allow other functions to be moved into Preferences and be marked as
`private` in subsequent patches.
The patch also renames SetPrefValue() as SetValueFromDom(), because that's a
clearer name.
MozReview-Commit-ID: CB1xmPSmac6
--HG--
extra : rebase_source : 0d597a800f2295c04af26d5abaac4aea0e0d3373
It's a `final` class, so there's no need for `protected`.
MozReview-Commit-ID: 7n4DLpXo0el
--HG--
extra : rebase_source : b2d3eb9cf0e912689efa29d2255cc018fd8a7c64
This is unused for now, but will be necessary for nsPrefBranch::Set*() to call
into Preferences::Set*().
The patch also renames some arguments from aPref to aPrefName, because that's a
better name.
MozReview-Commit-ID: 2OPB7CHOgpw
--HG--
extra : rebase_source : 232b7be3c33d185f13ce86d91feea3b55069a4e6
This is nicer than a bool for tracking the Default vs. User distinction, and it
replaces the Preferences.cpp-only WhichValue type.
MozReview-Commit-ID: 8CrdDN2vBJQ
--HG--
extra : rebase_source : 0d49148c73e5aeb0a355a6d4189232c89295d2a7
This also changes URIUtils.cpp:DeserializeURI() to use the mutator to instantiate new URIs, instead of using their default constructor.
MozReview-Commit-ID: JQOvIquuQAP
--HG--
extra : rebase_source : e146624c5ae423f7f69a738aaaafaa55dd0940d9
Given that we only support macOS >= 10.9 nowadays, we don't need
to consider fallbacks on old systems anymore.
Removing LiHei Pro & LiSong Pro because they are old.
Removing STSong & STHeiti because they are simply old font names aliased to new Songti/Heiti SC.
MozReview-Commit-ID: BNYBjXCpQOr
--HG--
extra : rebase_source : 4c19607d183ac4033c8bdc05bd05b92c906a5e9d
libpref only allows pref modifications in the parent process. This patch
tightens up the checking, which is a bit inconsistent.
- It removes ENSURE_MAIN_PROCESS_WITH_WARNING, which does NS_WARNING on
failure, and replaces its uses with ENSURE_MAIN_PROCESS, which does NS_ERROR
on failure. This required adding an XRE_IsParentProcess() check to one place
in editor/.
- It converts XRE_IsContentProcess() tests to !XRE_IsParentProcess(), because
we now have multiple kinds of non-parent process.
- It uses ENSURE_MAIN_PROCESS to replace other checking code in a few places.
- It improves a comment in HandleDirty().
MozReview-Commit-ID: D8znQWH7ery
--HG--
extra : rebase_source : ea0fc095b31525bde82a1be217923512d030b76d
InitStaticMembers() is the hottest function within libpref, because it is
called from all the getters and setters. This patch tweaks it so the common
case is handled with a single comparison, instead of two. This is valid because
sPreferences is nulled at the same time that sShutdown is set.
The patch also tweaks some other conditions to take advantage of this fact.
MozReview-Commit-ID: C74fLWOw7bE
