summaryrefslogtreecommitdiffstats
path: root/webkit/port
Commit message (Collapse)AuthorAgeFilesLines
* Merge: chromium side [42671:42725]levin@chromium.org2009-04-231-6/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | Rebaselines: http://trac.webkit.org/changeset/42722 resulted in LayoutTests/editing/inserting/insert-3907422-fix.html LayoutTests/editing/pasteboard/paste-text-015.html LayoutTests/editing/style/font-family-with-space.html http://trac.webkit.org/changeset/42723 resulted in LayoutTests/editing/selection/select-all-iframe.html LayoutTests/svg/custom/pointer-events-path.svg http://trac.webkit.org/changeset/42716 resulted in LayoutTests/fast/dom/HTMLSelectElement/named-options.html http://trac.webkit.org/changeset/42725 Broke the close event behavior and resulted in disabling these ui tests: * BrowserCloseBeforeUnloadOK and * BrowserCloseBeforeUnloadCancel Review URL: http://codereview.chromium.org/92051 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@14291 0039d316-1c4b-4281-b951-d872f2087c98
* Use upstreamed V8WorkerCustom and V8WorkerContextCustom. Remove duplicate ↵jianli@chromium.org2009-04-232-519/+0
| | | | | | | | V8Utilities in project. Review URL: http://codereview.chromium.org/92044 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@14275 0039d316-1c4b-4281-b951-d872f2087c98
* Save ~150k on a Linux release build by not inlining DOMWrapperToNode.deanm@chromium.org2009-04-212-9/+17
| | | | | | | | | | | | | | Creates DOMWrapperToNodeHelper to preform the actual work (in the .cpp) and DOMWrapperToNode just does the cast. This function has a lot of callers from the generated bindings, and I am skeptical an extra call instruction will hurt. 25740168 /tmp/chrome.after 25902672 /tmp/chrome.before Review URL: http://codereview.chromium.org/88029 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@14105 0039d316-1c4b-4281-b951-d872f2087c98
* - Exteranalize strings but not the results of toString() on non-strings.davemoore@chromium.org2009-04-212-13/+22
| | | | | | | | This increases our score on the DOM Perf benchmarks by about 10% Review URL: http://codereview.chromium.org/79055 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@14103 0039d316-1c4b-4281-b951-d872f2087c98
* Fix UMR, created by the merge.dglazkov@google.com2009-04-201-1/+1
| | | | | | | | TBR=eroman Review URL: http://codereview.chromium.org/88007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@14058 0039d316-1c4b-4281-b951-d872f2087c98
* WebKit Merge 42609:42671, Chromium Sidedglazkov@google.com2009-04-207-29/+12
| | | | | | | | | | | | | | | | | | Gross offenders: * http://trac.webkit.org/changeset/42633, event listener creation helpers moved to JSC-specific files. * http://trac.webkit.org/changeset/42647, our refactoring of XHR code. * http://trac.webkit.org/changeset/42671, unforking of DOMWindow.event R=levin Review URL: http://codereview.chromium.org/84002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@14056 0039d316-1c4b-4281-b951-d872f2087c98
* Null check V8 proxy before clearing it in out of memory situations.ager@chromium.org2009-04-201-4/+6
| | | | | | | BUG=10693 Review URL: http://codereview.chromium.org/67298 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@14026 0039d316-1c4b-4281-b951-d872f2087c98
* Unfork Inspector, Chromium side.dglazkov@google.com2009-04-174-1/+18
| | | | | | | | R=darin Review URL: http://codereview.chromium.org/67052 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13975 0039d316-1c4b-4281-b951-d872f2087c98
* Quickfix to the v8 code generator to properly handle combination of ↵ericroman@google.com2009-04-171-0/+8
| | | | | | | | "CustomGetter" + "Replaceable". Review URL: http://codereview.chromium.org/67261 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13968 0039d316-1c4b-4281-b951-d872f2087c98
* WebKit Merge 42580:42609, Chromium side.dglazkov@google.com2009-04-172-2/+2
| | | | | | | | | | | | | | | | Changes to ScriptController, glue due to http://trac.webkit.org/changeset/42583 Test rebaselines due to: * http://trac.webkit.org/changeset/42599 * http://trac.webkit.org/changeset/42583 * http://trac.webkit.org/changeset/42600 * http://trac.webkit.org/changeset/42586 R=levin Review URL: http://codereview.chromium.org/79038 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13965 0039d316-1c4b-4281-b951-d872f2087c98
* Chrome side of webkit merge from 42547 to 42580.tc@google.com2009-04-164-24/+2
| | | | | | | | | | | | | | | | In http://trac.webkit.org/changeset/42579 ScriptController::createInlineEventListener was changed to take 4 params instead of 3, so I updated the V8 ScriptController to do the same. We don't seem to need the extra arg because both v8_proxy::createSVGEventHandler and v8_proxy::createInlineEventListener are identical. Also one new .cpp/.h file from http://trac.webkit.org/changeset/42580 Review URL: http://codereview.chromium.org/79008 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13884 0039d316-1c4b-4281-b951-d872f2087c98
* Make HTMLSelectElement and HTMLOptionsCollection return collections of optionsdavemoore@chromium.org2009-04-163-9/+35
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | elements if there are multiple options that share a name. Currently Safari and chromium differ in their behavior in this area. Safari returns multiple options in a collection if necessary, but only for HTMLOptionsCollection. For HTMLSelectElement it always returns null. Chromium returns a single option no matter what, but for both HTMLSelectElement and HTMLOptionsCollection. It's also very slow at this, as it bypasses a webkit cache. The HTML5 doc states that a collection should be returned if necessary for both. The relevant parts of the spec are: http://dev.w3.org/html5/spec/Overview.html#the-select-element and http://dev.w3.org/html5/spec/Overview.html#htmloptionscollection-0 So if you have: <select id="sl1"><option value="Value" name="test" /></select> and in JS code you do: var test = document.getElementById("sl1").test test should be a single HTMLOptionElement. If instead you have: <select id="sl2"><option value="Value1" name="test" /><option value="Value2" name="test" /></select> test should be a collection with 2 HTMLOptionElements. This bug: https://bugs.webkit.org/show_bug.cgi?id=25191 has been filed against webkit, with a new layout test to confirm the behavior Review URL: http://codereview.chromium.org/69014 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13814 0039d316-1c4b-4281-b951-d872f2087c98
* Chromium side of webkit merge 42440:42523.ericroman@google.com2009-04-151-3/+3
| | | | | | Review URL: http://codereview.chromium.org/73066 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13774 0039d316-1c4b-4281-b951-d872f2087c98
* Reverting 13760.davemoore@chromium.org2009-04-153-34/+9
| | | | | | Review URL: http://codereview.chromium.org/75002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13762 0039d316-1c4b-4281-b951-d872f2087c98
* Make HTMLSelectElement and HTMLOptionsCollection return collections of optionsdavemoore@chromium.org2009-04-153-9/+34
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | elements if there are multiple options that share a name. Currently Safari and chromium differ in their behavior in this area. Safari returns multiple options in a collection if necessary, but only for HTMLOptionsCollection. For HTMLSelectElement it always returns null. Chromium returns a single option no matter what, but for both HTMLSelectElement and HTMLOptionsCollection. It's also very slow at this, as it bypasses a webkit cache. The HTML5 doc states that a collection should be returned if necessary for both. The relevant parts of the spec are: http://dev.w3.org/html5/spec/Overview.html#the-select-element and http://dev.w3.org/html5/spec/Overview.html#htmloptionscollection-0 So if you have: <select id="sl1"><option value="Value" name="test" /></select> and in JS code you do: var test = document.getElementById("sl1").test test should be a single HTMLOptionElement. If instead you have: <select id="sl2"><option value="Value1" name="test" /><option value="Value2" name="test" /></select> test should be a collection with 2 HTMLOptionElements. This bug: https://bugs.webkit.org/show_bug.cgi?id=25191 has been filed against webkit, with a new layout test to confirm the behavior Review URL: http://codereview.chromium.org/67148 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13760 0039d316-1c4b-4281-b951-d872f2087c98
* Revert changes causing memory_test problems. asargent@chromium.org2009-04-141-11/+2
| | | | | | | | | | | | This undoes changes from: http://codereview.chromium.org/66045 Review URL: http://codereview.chromium.org/67114 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13639 0039d316-1c4b-4281-b951-d872f2087c98
* Fix problems in event listener teardownasargent@chromium.org2009-04-131-2/+11
| | | | | | | | | | | The problem is sometimes the functions to remove event listeners get called outside of a ContextScope (even though we still have a valid context). BUG=9775 Review URL: http://codereview.chromium.org/66045 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13622 0039d316-1c4b-4281-b951-d872f2087c98
* Fix worker constructor so that nested worker can be created.jianli@chromium.org2009-04-131-6/+13
| | | | | | Review URL: http://codereview.chromium.org/67093 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13619 0039d316-1c4b-4281-b951-d872f2087c98
* Chromium side of merge 42364:42403.pam@chromium.org2009-04-111-1/+1
| | | | | | | | Includes change to v8_proxy.cpp to account for addition of MessagePortProxy class. Review URL: http://codereview.chromium.org/66062 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13553 0039d316-1c4b-4281-b951-d872f2087c98
* Remove unneeded ASSERT -- this code path is valid and should not be ASSERTing.dglazkov@google.com2009-04-101-1/+0
| | | | | | | | R=mbelshe Review URL: http://codereview.chromium.org/67040 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13523 0039d316-1c4b-4281-b951-d872f2087c98
* Chrome side of WEbKit merge 42244:42287.brettw@chromium.org2009-04-081-2/+10
| | | | | | Review URL: http://codereview.chromium.org/63110 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13337 0039d316-1c4b-4281-b951-d872f2087c98
* Chrome side change to use V8EventListenerList from upstream. This involves ↵jianli@chromium.org2009-04-032-114/+1
| | | | | | | | removing V8EventListenerList from v8_proxy and change the project files to compile V8EventListenerList. Review URL: http://codereview.chromium.org/60102 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@13107 0039d316-1c4b-4281-b951-d872f2087c98
* Implement custom methods for some Node dom callsmbelshe@google.com2009-04-011-2/+5
| | | | | | | | | | | | | so that we can take advantage of lazy wrapper initialization. Dependent on the webkit fix. will roll the deps for webkit as well. Review URL: http://codereview.chromium.org/57086 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12958 0039d316-1c4b-4281-b951-d872f2087c98
* Keep JS wrappers for non-node objects associated with a DOMWindowager@chromium.org2009-04-012-7/+64
| | | | | | | | | | alive across GCs. This change needs a WebKit change to go in first which adds frame and type accessors to these objects. Review URL: http://codereview.chromium.org/49038 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12942 0039d316-1c4b-4281-b951-d872f2087c98
* Reland changes to fix the problem that DOM object is deleted in the GC ↵jianli@chromium.org2009-03-294-382/+218
| | | | | | | | | | | | | thread, not owning thread. This patch contains the fix to the Purify failure: memory leaks of thread specific data for main thread. The fix is to switch to using the static maps for main thread. Compared with previous patch being reviewed, the main change in this patch is in V8DOMMap.cpp: I introduce NonMainThreadSpecificDOMData and MainThreadSpecificDOMData that are both derived from ThreadSpecificDOMData. Review URL: http://codereview.chromium.org/49044 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12757 0039d316-1c4b-4281-b951-d872f2087c98
* Search box support, Chromium side.BUG=9210pkasting@chromium.org2009-03-271-1/+1
| | | | | | Review URL: http://codereview.chromium.org/42678 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12680 0039d316-1c4b-4281-b951-d872f2087c98
* Start using the upstream versions of V8NodeFilter and V8XML*.levin@chromium.org2009-03-278-1083/+1
| | | | | | Review URL: http://codereview.chromium.org/42673 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12626 0039d316-1c4b-4281-b951-d872f2087c98
* Speedup event performance by caching V8EventListener on the wrapped object. asargent@chromium.org2009-03-262-27/+120
| | | | | | Review URL: http://codereview.chromium.org/49026 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12574 0039d316-1c4b-4281-b951-d872f2087c98
* Fix window.history GC issue. Since the history object is not part ofager@chromium.org2009-03-262-14/+27
| | | | | | | | the DOM, we need to explicitly keep the history wrapper object alive for the lifetime of the DOMWindow object with which it is associated. Review URL: http://codereview.chromium.org/53092 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12543 0039d316-1c4b-4281-b951-d872f2087c98
* Revert change 12507 and 12532 because it breaks the nsylvain@chromium.org2009-03-265-787/+388
| | | | | | | tree and the purify tests. Review URL: http://codereview.chromium.org/42634 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12534 0039d316-1c4b-4281-b951-d872f2087c98
* Fix the problem that DOM object is deleted in the GC thread, not owning thread.jianli@chromium.org2009-03-255-388/+787
| | | | | | Review URL: http://codereview.chromium.org/40088 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12507 0039d316-1c4b-4281-b951-d872f2087c98
* Fix data url security tests.ager@chromium.org2009-03-251-4/+7
| | | | | | | | | | | | | | The problem was that the security token in V8 was set to the string "null" for all data urls. Therefore, the access check succeeded. The cause of the regression was that securityOrigin's toString implementation was changed in WebKit to return "null" instead of String(). R=kasperl@chromium.org TBR=scherkus@chromium.org Review URL: http://codereview.chromium.org/53049 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12444 0039d316-1c4b-4281-b951-d872f2087c98
* Start using upstreamed worker files and remove the old ones from webkit/port.dimich@google.com2009-03-244-642/+0
| | | | | | | Before submitting, I'll add DEPS with corresponding WebKit rev (see http://codereview.chromium.org/52007) Review URL: http://codereview.chromium.org/45035 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12377 0039d316-1c4b-4281-b951-d872f2087c98
* Add basic infrastructure for sending async browser API request and receiving ↵aa@chromium.org2009-03-242-12/+28
| | | | | | | | results. Review URL: http://codereview.chromium.org/42262 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12347 0039d316-1c4b-4281-b951-d872f2087c98
* Chrome side of webkit merge from r41860 to 41906.tc@google.com2009-03-232-2/+2
| | | | | | | | | | | | | The encoding parameter got dropped from Workers (r41871) so drop it through the whole stack. The upstream WorkerContextExecutionProxy has trackEvent with a lower case t, but we pull in the wrong header in some places so make both headers match. Review URL: http://codereview.chromium.org/42538 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12326 0039d316-1c4b-4281-b951-d872f2087c98
* Back out string morphing.mbelshe@google.com2009-03-231-5/+10
| | | | | | Review URL: http://codereview.chromium.org/52027 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12320 0039d316-1c4b-4281-b951-d872f2087c98
* Rework the v8 binding string handling to be more efficient.mbelshe@google.com2009-03-232-79/+128
| | | | | | | | | | | Break out ToAtomicWebCoreString in addition to ToWebCoreString. Re-enables the v8 string morphing; there is no longer degradation on intl1 page cycler. Review URL: http://codereview.chromium.org/42445 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12309 0039d316-1c4b-4281-b951-d872f2087c98
* Change chromium to use the V8Collection from WebKit/WebCore/bindings/v8.levin@chromium.org2009-03-232-192/+1
| | | | | | Review URL: http://codereview.chromium.org/52007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12293 0039d316-1c4b-4281-b951-d872f2087c98
* Hook up V8 Database Bindings.dglazkov@google.com2009-03-225-3/+29
| | | | | | | | R=darin Review URL: http://codereview.chromium.org/50071 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12264 0039d316-1c4b-4281-b951-d872f2087c98
* Remove the DOM constructor cache and instead use an object wrapper prototypembelshe@google.com2009-03-202-63/+76
| | | | | | | | | | which we just clone. This fix cannot land until we get the latest v8. Review URL: http://codereview.chromium.org/50056 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12225 0039d316-1c4b-4281-b951-d872f2087c98
* Preupstreaming cleanup for xhr (part 2): levin@chromium.org2009-03-205-325/+492
| | | | | | | | Split V8XMLHttpRequestCustom into appropriate files. Review URL: http://codereview.chromium.org/45018 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12216 0039d316-1c4b-4281-b951-d872f2087c98
* Pre-upstreaming style changes to match webkit style.levin@chromium.org2009-03-193-761/+743
| | | | | | Review URL: http://codereview.chromium.org/48169 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12169 0039d316-1c4b-4281-b951-d872f2087c98
* Enable IDL attribute [V8ReadOnly] to have the same mbelshe@google.com2009-03-191-3/+4
| | | | | | | | | | | | | semantics as having specified an attribute as readonly. This allows us to unfork IDL differences where V8 wants readonly and JSC does not by using the V8ReadOnly tag. Review URL: http://codereview.chromium.org/50037 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12165 0039d316-1c4b-4281-b951-d872f2087c98
* Linux: fix include filename caseagl@chromium.org2009-03-191-2/+2
| | | | | | | Review URL: http://codereview.chromium.org/42383 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12075 0039d316-1c4b-4281-b951-d872f2087c98
* Pre-upstream style changes: rename enableWebWorkers to ↵dimich@google.com2009-03-192-5/+5
| | | | | | | | | setIsWebWorkersEnabled to follow isFoo/setIsFoo pattern. Also pull DEPS to bring in the WebKit/chromium/src part. Review URL: http://codereview.chromium.org/48157 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@12066 0039d316-1c4b-4281-b951-d872f2087c98
* Unfork WebInspector frontend take 2: New DEPS and webkit/...pfeldman@chromium.org2009-03-181-0/+9
| | | | | | Review URL: http://codereview.chromium.org/48132 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@11967 0039d316-1c4b-4281-b951-d872f2087c98
* Add a handle scope to V8Proxy::initContextIfNeeded.ager@chromium.org2009-03-181-1/+4
| | | | | | | | | | | | initContextIfNeeded can be called without a handle scope in place which will lead to a renderer crash when creating the first handle. Also V8Proxy::initContextIfNeeded creates a number of local handles that should be deallocated on exit. BUG=8922 Review URL: http://codereview.chromium.org/48131 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@11965 0039d316-1c4b-4281-b951-d872f2087c98
* Revert the src/webkit part of r11958 since it broke the build.maruel@chromium.org2009-03-181-9/+0
| | | | | | | TBR=pfeldman Review URL: http://codereview.chromium.org/42332 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@11960 0039d316-1c4b-4281-b951-d872f2087c98
* Unfork WebInspector frontend.pfeldman@chromium.org2009-03-181-0/+9
| | | | | | Review URL: http://codereview.chromium.org/48063 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@11958 0039d316-1c4b-4281-b951-d872f2087c98
* Make worker constructor return error if it is enabled by the flag.jianli@chromium.org2009-03-183-0/+22
| | | | | | | This is to move corresponding files out of http://codereview.chromium.org/48106/ so that this can be checked in before the change to enable worker. The reason for doing this is to avoid build break since the already checked-in third_party/WebKit/WebKit/chromium/src/WebKit.cpp has the ifdef guarded code depend on this. Review URL: http://codereview.chromium.org/42313 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@11952 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-02-01 08:42:17 +0000