summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Switch MessagePumpForIO to use completion ports on Windows.rvargas@google.com2008-11-0716-932/+816
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Cleanup the separation between MessagePumpForUI and MessagePumpForIO, and convert the latter to use Completion Ports instead of MsgWaitForMultipleobjects to sleep when idle. Remove all traces of Windows messages from MessagePumpForIO, remove the transitional API of completion port notifications and remove WatchObject API. Modify all callers of RegisterIOHandler so that they are no longer using RegisterIOContext, and also handle properly the new semantics of completion ports (notifications even when the IO completes immediately). Add a new interface to allow proper cleanup of disk cache (to replace code that was waiting for pending APCs from the destructor). Add a way for the message pump to perform cleanup of abandoned IO. BUG=B/1344358, 3497, 3630 TESt=unit tests R=darin Review URL: http://codereview.chromium.org/8156 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5021 0039d316-1c4b-4281-b951-d872f2087c98
* Restructure the keyboard test_shell test to properly init event namesevanm@google.com2008-11-071-63/+73
| | | | | | | | | | | | | | for all tests. Otherwise, depending on which tests ran and in which order, EventNames::init() would not get called. I lost a couple hours last night debugging to figure this out. :( While I was at it, I rearranged the test functions for cleanliness and fixed some minor style bits. Review URL: http://codereview.chromium.org/10213 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5019 0039d316-1c4b-4281-b951-d872f2087c98
* Landing this again as I cannot reproduce the perf regression locally.jcampan@chromium.org2008-11-0739-124/+739
| | | | | | | | | | | Will investigate on the bot. TBR=nsylvain Review URL: http://codereview.chromium.org/9700 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5018 0039d316-1c4b-4281-b951-d872f2087c98
* Turns back on tree's double buffering when RTL. For some reason asky@google.com2008-11-071-14/+11
| | | | | | | | | | | | | graphics mode of a ADVANCED doesn't work with a RTL layout for this code. BUG=4175 TEST=bring up the bookmark manager in the hebrew locale, make sure the text isn't mirrored. Review URL: http://codereview.chromium.org/10216 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5017 0039d316-1c4b-4281-b951-d872f2087c98
* Make AutomationProvider::WindowSimulateDrag use SendInput instead of sending ↵ben@chromium.org2008-11-075-65/+152
| | | | | | | | | | | messages directly to the targeted window. http://crbug.com/4067 Review URL: http://codereview.chromium.org/9485 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5014 0039d316-1c4b-4281-b951-d872f2087c98
* Port parts of chrome/common to Linux.evanm@google.com2008-11-073-33/+76
| | | | | | | | | | | | | This probably needs some bigger adjustments. Some things need to be decided, like where to store profile files on Linux/Mac. BUG=3649 Review URL: http://codereview.chromium.org/8059 Patch from Paweł Hajdan jr <phajdan.jr@gmail.com>. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5012 0039d316-1c4b-4281-b951-d872f2087c98
* Make the buildbots green.ager@google.com2008-11-071-1/+6
| | | | | | | | | | | | | A test list was accidentally commited with a line commented out. Aditionally, some of the animation tests are flaky. When running all tests, one of the tests seems to consistently time out. When running only the animation tests another test is timing out. Updating the test list with the test that currently times out on the buildbot. Review URL: http://codereview.chromium.org/10211 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5009 0039d316-1c4b-4281-b951-d872f2087c98
* Bringing in sct changes.bradnelson@google.com2008-11-073-73/+603
| | | | | | Review URL: http://codereview.chromium.org/10210 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5008 0039d316-1c4b-4281-b951-d872f2087c98
* Set cpus=1 on systems where we don't (yet) know how to deducesgk@google.com2008-11-071-2/+2
| | | | | | | | | | | the number of CPUs. This at least allows the build to function on unknown systems. Remove a now-unnecessary addition of $V8_DIR to LIBPATH. The v8 libraries now get installed in $LIBS_DIR like everyone else. Review URL: http://codereview.chromium.org/10208 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5007 0039d316-1c4b-4281-b951-d872f2087c98
* Keep MessagePort wrappers alive for the life-time of the correspondingager@google.com2008-11-075-19/+45
| | | | | | | | | | | | | MessageChannel. I create the MessagePort wrappers when a MessageChannel is constructed and put references to them in internal fields in the MessageChannel. This is a fix for http://code.google.com/p/chromium/issues/detail?id=4195 Review URL: http://codereview.chromium.org/9508 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5006 0039d316-1c4b-4281-b951-d872f2087c98
* Updates one of the unit tests I disabled when I rewrote the popup blocker UI.erg@google.com2008-11-071-14/+61
| | | | | | Review URL: http://codereview.chromium.org/10206 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5005 0039d316-1c4b-4281-b951-d872f2087c98
* Remove KJS from the test lists since we don't keep itnsylvain@chromium.org2008-11-074-1543/+1411
| | | | | | | | up to date anymore. Review URL: http://codereview.chromium.org/9649 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5004 0039d316-1c4b-4281-b951-d872f2087c98
* Fix initialization order to fix Mac & Linux.brettw@google.com2008-11-071-2/+2
| | | | | | Review URL: http://codereview.chromium.org/10209 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5003 0039d316-1c4b-4281-b951-d872f2087c98
* Added a new interface IChromeHistoryIndexer which needs to be implemented bybrettw@google.com2008-11-0716-63/+447
| | | | | | | | | | | indexers which are interested in indexing Chrome's web history. These should register their CLSID in HKCU\Software\Google\Google Chrome\IndexerPlugins and this is be used by Chrome to publish its index. Added code to publish the index too. Checked in for Vijay Thadkal <veejay.t.s@gmail.com> (Google). Reitveld #9007. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5002 0039d316-1c4b-4281-b951-d872f2087c98
* Fix keyboard unit test.evanm@google.com2008-11-072-1/+4
| | | | | | | | | We work around all event names being null (!!!) by testing a enum instead. This code should be faster anyway, but I will also investigate the enum. Review URL: http://codereview.chromium.org/9495 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5001 0039d316-1c4b-4281-b951-d872f2087c98
* Initialize member on non-Windows platforms.evanm@google.com2008-11-071-0/+2
| | | | git-svn-id: svn://svn.chromium.org/chrome/trunk/src@5000 0039d316-1c4b-4281-b951-d872f2087c98
* Delete layout_test_results/kjs/. No automated scripts are using it, sopamg@google.com2008-11-074104-244594/+0
| | | | | | | | | | | it's not being kept up to date and is therefore not useful. We will continue to maintain the JSC build, just not its layout test results. BUG=4207 TEST=none git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4997 0039d316-1c4b-4281-b951-d872f2087c98
* Make wdiff work for Mac toopinkerton@google.com2008-11-073-5/+14
| | | | | | Review URL: http://codereview.chromium.org/9694 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4996 0039d316-1c4b-4281-b951-d872f2087c98
* Bring ignored/fixable test lists more in sync with win since the last merge ↵pinkerton@google.com2008-11-072-142/+351
| | | | | | | | or two. Review URL: http://codereview.chromium.org/10204 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4994 0039d316-1c4b-4281-b951-d872f2087c98
* Adds cut/copy/paste support to the bookmark manager tree/table andsky@google.com2008-11-079-5/+239
| | | | | | | | | | | | context menus. BUG=4186 TEST=try cut/copy/paste from table/tree in the bookmark manager as well as context menus in bookmark manager. Review URL: http://codereview.chromium.org/9481 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4993 0039d316-1c4b-4281-b951-d872f2087c98
* Make the linux test shell display the window title.agl@chromium.org2008-11-072-1/+2
| | | | | | | | | | I have only compiled this on linux and tested it there but that shouldn't be a problem as the file does seem to be linux-specific. Review URL: http://codereview.chromium.org/9685 Patch from Craig Schlenter. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4992 0039d316-1c4b-4281-b951-d872f2087c98
* Make PlatformCanvasLinux match up with recent changes in the Windows one, ↵brettw@google.com2008-11-072-3/+8
| | | | | | | | which is able to report allocation failures. Review URL: http://codereview.chromium.org/9510 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4991 0039d316-1c4b-4281-b951-d872f2087c98
* Fix Linux build bustage by making the ImageBufferSkia code compilable on all ↵brettw@google.com2008-11-072-3/+6
| | | | | | | | Skia platforms. I do this by making the calling of PlatformCanvas.initialize() able to be the same by using an evil default parameter. Review URL: http://codereview.chromium.org/9695 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4990 0039d316-1c4b-4281-b951-d872f2087c98
* Fix crash in the icon manager. The decoding could be NULL, which would then getbrettw@google.com2008-11-073-2/+12
| | | | | | | | | | | dereferenced in specific cases. We would have to create the failed icon twice, so that it would think the icon was already in the cache, when in fact a NULL value was being cached. BUG=1259695 Review URL: http://codereview.chromium.org/9672 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4988 0039d316-1c4b-4281-b951-d872f2087c98
* Make Canvas and other ImageBuffer uses able to detect failure and not createbrettw@google.com2008-11-072-1/+20
| | | | | | | the canvas. Review URL: http://codereview.chromium.org/9668 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4987 0039d316-1c4b-4281-b951-d872f2087c98
* No code change. Force a build.maruel@chromium.org2008-11-071-5/+6
| | | | git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4986 0039d316-1c4b-4281-b951-d872f2087c98
* Rebaselined the following 6 layout tests.sgjesse@chromium.org2008-11-077-13/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | LayoutTests/http/tests/security/listener/xss-XMLHttpRequest-shortcut.html LayoutTests/http/tests/security/listener/xss-XMLHttpRequest-addEventListener.html LayoutTests/http/tests/security/listener/xss-window-onclick-shortcut.html LayoutTests/http/tests/security/listener/xss-window-onclick-addEventListener.html LayoutTests/http/tests/security/listener/xss-JSTargetNode-onclick-shortcut.html LayoutTests/http/tests/security/listener/xss-JSTargetNode-onclick-addEventListener.html In WebKit r33979 all these layout tests had alert("document.domain: " + window.document.domain); changed to try { alert("document.domain: " + window.document.domain); } catch(e) { } causing the document.domain access error to no longer be reported. Rebaselining the tests does not compromize what they are testing as if window.domain was accessible the output would change causing failures. Review URL: http://codereview.chromium.org/9691 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4985 0039d316-1c4b-4281-b951-d872f2087c98
* Disallow shadowing of top since shadowing top can break frame-bustingager@google.com2008-11-0715-22/+340
| | | | | | | | | | | | code. Also, make sure that top, window, location, and location.href are not deletable. I had to fix a couple of layout tests that rely on top being shadowable. Review URL: http://codereview.chromium.org/9506 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4984 0039d316-1c4b-4281-b951-d872f2087c98
* Do not allow shadowing of location.href since that can breakager@google.com2008-11-073-2/+8
| | | | | | | frame-busting code. Review URL: http://codereview.chromium.org/9688 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4981 0039d316-1c4b-4281-b951-d872f2087c98
* Make the buildbots happy. These started failing after a completelyager@google.com2008-11-072-0/+8
| | | | | | | | unrelated change. I think there is a real GC issue here and I will look into it. Review URL: http://codereview.chromium.org/9687 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4980 0039d316-1c4b-4281-b951-d872f2087c98
* Another attempt at fixing the flash security issue surroundingager@google.com2008-11-0713-120/+320
| | | | | | | | | | | | | | | | window.location. This change disallows shadowing of window and location both when using var declarations and when defining accessors. Using 'var location = stuff' will cause a navigation. This is consistent with what Firefox does. Also, introduce an explicit valueOf function on location objects so that implicit string conversions cannot be intercepted to fake the current location reported by window.location. Review URL: http://codereview.chromium.org/9636 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4979 0039d316-1c4b-4281-b951-d872f2087c98
* Don't assume WebCore::UChar is equivalent to wchar_t, and don't assume '\\' ↵estade@chromium.org2008-11-073-8/+9
| | | | | | | | is a valid file path separator. Review URL: http://codereview.chromium.org/9489 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4972 0039d316-1c4b-4281-b951-d872f2087c98
* Adds import/export of bookmarks to bookmarks.html file.sky@google.com2008-11-0728-153/+1066
| | | | | | | | | | | | BUG=1649 TEST=bring up bookmark manager and try out import/export from the tools menu. Note that import ALWAYS creates a new folder under the 'Other bookmarks folder' with the name of Imported (x). This is by design. Review URL: http://codereview.chromium.org/9471 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4968 0039d316-1c4b-4281-b951-d872f2087c98
* Potential fix for FF importing issuecpu@google.com2008-11-072-3/+6
| | | | | | | | | | | | | | | - Select the highest FF version, not just the the HCKU one. - I don't have a repro case, it is a blind fix Seems like leftovers of old FF installs in HCKU are tripping us, that is my guess, now we would find FF3 in HKLM. BUG=3109 Review URL: http://codereview.chromium.org/9675 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4960 0039d316-1c4b-4281-b951-d872f2087c98
* Implement a gtk test shell function. Pass dom_operations test shell test.estade@chromium.org2008-11-075-16/+25
| | | | | | Review URL: http://codereview.chromium.org/8230 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4956 0039d316-1c4b-4281-b951-d872f2087c98
* This fixes bug http://code.google.com/p/chromium/issues/detail?id=4137, ananta@chromium.org2008-11-072-11/+15
| | | | | | | | | | | | | | | | | | which was an issue with the site failing to detect Windows media player. The site would run a script which would enumerate the list of plugins from the navigator object and compare each plugin name for "Windows Media" and "Plug". This is under the assumption that the windows media player plugin is typically named as Microsoft� Windows Media Player Firefox Plugin and it would find both the strings in the name. I had changed the name of activex shim instance for media player previously to Windows Media to fix a similar issue with bloomberg.com. Should have changed it to the full name at that point. The fix is to fully match the windows media player plugin name in the shim. The other change is to perform a case insensitive comparison of the plugin file names while looking for matching plugins. R=jam Bug=4137 Review URL: http://codereview.chromium.org/9482 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4950 0039d316-1c4b-4281-b951-d872f2087c98
* Implement the new mechanism we are going to use for handling Chromium updateskuchhal@chromium.org2008-11-0723-65/+514
| | | | | | | | | | | | | | | | | | | | while it is in use. This should work for per user as well as system level installs (after some additional changes once Google Update changes are ready). The following scenarios should work now: - If Chromium is using two different profiles at the same time we do not switch chrome executables until all of them are closed. - The old version of Chromium can be run after a restart if the renaming of executables fails. - We will not use environment variable any more but we need to keep it until all the users get this change on their machines. - opv/rename registry keys and new_chrome.exe should always together. If one exist all three should exist because they are created and deleted as one atomic operation (as much as possible given laws of physics). BUG=1463346 Review URL: http://codereview.chromium.org/9436 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4949 0039d316-1c4b-4281-b951-d872f2087c98
* * Never launch Chrome in system level install mode.kuchhal@chromium.org2008-11-071-1/+3
| | | | | | Review URL: http://codereview.chromium.org/9479 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4946 0039d316-1c4b-4281-b951-d872f2087c98
* Rename SConscript.dll to SConscript.browser, to more accurately reflect itsmpcomplete@google.com2008-11-072-2/+2
| | | | | | | purpose (targets built there are per-browser). Review URL: http://codereview.chromium.org/9483 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4945 0039d316-1c4b-4281-b951-d872f2087c98
* Some basic keyboard and focus support.evanm@google.com2008-11-075-6/+51
| | | | | | | Review URL: http://codereview.chromium.org/9626 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4944 0039d316-1c4b-4281-b951-d872f2087c98
* Plop gears 0.5.3.0 binaries in the Chrome tree.mpcomplete@google.com2008-11-072-0/+0
| | | | | | Review URL: http://codereview.chromium.org/9667 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4943 0039d316-1c4b-4281-b951-d872f2087c98
* Fix gears-scons linux buildmpcomplete@chromium.org2008-11-072-42/+45
| | | | | | Review URL: http://codereview.chromium.org/9480 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4942 0039d316-1c4b-4281-b951-d872f2087c98
* Insert histograms to support latency experiments with SDCHjar@google.com2008-11-075-1/+42
| | | | | | | | | | | | Gather data to show the duration of time between the first byte sent, to last byte read. Use a second histogram to look at latency when the content-encoding is not really sdch (due to server side experiment, or proxy interference). r=huanr,wtc Review URL: http://codereview.chromium.org/9448 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4941 0039d316-1c4b-4281-b951-d872f2087c98
* A crasher would happen if you donwload an exe and the download has completed ↵jcampan@chromium.org2008-11-072-2/+8
| | | | | | | | | | | | | by the time the download item is created, you have the download tab opened and you press on the discard button. This is because such downloads gets added to a vector that is used to unregister the DownloadTabView as an observer. We were deleting the download item before notifying the DownloadTabView, which would cause it to access the deleted download item. BUG=4134 TEST=Open the download tab, start the download of a short exe. Click the discard button in the dangerous download prompt. Review URL: http://codereview.chromium.org/9435 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4940 0039d316-1c4b-4281-b951-d872f2087c98
* Pull input file lists from SConscript.common and SConscript.dll tompcomplete@google.com2008-11-074-994/+1010
| | | | | | | | | | SConscript.inputs. This file will be shared by the symbian port. SConscript.inputs is mostly a cut+paste of the original files. The few changes are to stick resources on the 'env' var so they can be shared. Review URL: http://codereview.chromium.org/9666 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4939 0039d316-1c4b-4281-b951-d872f2087c98
* Add prefixes to I/O data trace names so they can be distinguished on thepamg@google.com2008-11-071-8/+16
| | | | | | | | | | waterfall. BUG=4173 TEST=look at page-cycler step on waterfall Review URL: http://codereview.chromium.org/9658 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4936 0039d316-1c4b-4281-b951-d872f2087c98
* Continue with crosssite navigations if alerts are suppressedojan@google.com2008-11-075-6/+21
| | | | | | | in an unload handler (i.e. we have an infinite-ish alert loop). Review URL: http://codereview.chromium.org/9477 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4935 0039d316-1c4b-4281-b951-d872f2087c98
* Prevent window.close() working from Javascript when a blocked popup ↵erg@google.com2008-11-073-2/+15
| | | | | | | | | notification exists. This is a copy of Firefox's behaviour. BUG=crbug.com/4007 Review URL: http://codereview.chromium.org/9656 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4934 0039d316-1c4b-4281-b951-d872f2087c98
* Implement the idle timer for Linux using libxss.dsh@google.com2008-11-063-0/+66
| | | | | | | | BUG=2183 Review URL: http://codereview.chromium.org/8806 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4930 0039d316-1c4b-4281-b951-d872f2087c98
* Adaptively identify URL subresources and pre-resolve hosts via DNSjar@google.com2008-11-0615-164/+645
| | | | | | | | | | | | | | | | | | | | Use the HTTP "referer" header to identify subresources used during a page load. Store that info, and use it when next visiting the referenced hosts to pre-resolve the (probably) needed subresources. This set of changes will surely evolve as we see how it plays out on broader distribution (via histogram measurments), but this should be the foundation of the change. In design specs, this was previously referred to as "adaptive correlated DNS prefetching." r=mbelshe Review URL: http://codereview.chromium.org/9168 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4929 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-01-27 13:55:59 +0000