WebKit.git
11 years ago2010-05-09 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Sun, 9 May 2010 11:18:25 +0000 (11:18 +0000)]
2010-05-09  Maciej Stachowiak  <mjs@apple.com>

        Fixed version of: "Optimized o[s] where o is a cell and s is a string"
        https://bugs.webkit.org/show_bug.cgi?id=38815

        Fixed the previous patch for this from Geoff Garen.

        The two problems were a missing exception check and a PropertySlot
        initialized improperly, leading to crashes and failures in the case
        of getters accessed with bracket syntax.

    Previous patch:

        Optimized o[s] where o is a cell and s is a string, removing some old
        code that wasn't really tuned for the JIT.

        SunSpider says 0.8% faster.

        * jit/JITStubs.cpp:
        (JSC::DEFINE_STUB_FUNCTION):
        * runtime/JSCell.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59055 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-09 Robert Hogan <robert@webkit.org>
robert@webkit.org [Sun, 9 May 2010 10:24:58 +0000 (10:24 +0000)]
2010-05-09  Robert Hogan  <robert@webkit.org>

        [Qt] Update expected results for unskipped editing/inserting tests

        Unreviewed, remove platform-specific results for editing/inserting/6633727.html

        https://bugs.webkit.org/show_bug.cgi?id=38472

        * platform/qt/editing/inserting/6633727-expected.txt: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59054 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[Qt] Update expected results for unskipped editing/inserting tests
robert@webkit.org [Sun, 9 May 2010 10:15:16 +0000 (10:15 +0000)]
[Qt] Update expected results for unskipped editing/inserting tests

Unreviewed, add expected results for unskipped tests. Results are taken
from Qt bot.

https://bugs.webkit.org/show_bug.cgi?id=38472

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59053 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[Qt] Unskip editing/inserting tests that fail due to x/y differences
robert@webkit.org [Sun, 9 May 2010 09:52:31 +0000 (09:52 +0000)]
[Qt] Unskip editing/inserting tests that fail due to x/y differences

2010-05-03  Robert Hogan  <robert@webkit.org>

        Reviewed by Eric Seidel.

        [Qt] Unskip editing/inserting tests that fail due to x/y differences

        https://bugs.webkit.org/show_bug.cgi?id=38472

        Unskip the following tests:

        editing/inserting/12882.html
        editing/inserting/4840662.html
        editing/inserting/4875189-1.html
        editing/inserting/4959067.html
        editing/inserting/4960120-2.html
        editing/inserting/5002441.html
        editing/inserting/5058163-2.html
        editing/inserting/5156401-2.html
        editing/inserting/5418891.html
        editing/inserting/5510537.html
        editing/inserting/5549929-2.html
        editing/inserting/5549929-3.html
        editing/inserting/6633727.html
        editing/inserting/6703873.html
        editing/inserting/editable-inline-element.html
        editing/inserting/insert-3654864-fix.html
        editing/inserting/insert-3659587-fix.html
        editing/inserting/insert-3775316-fix.html
        editing/inserting/insert-3778059-fix.html
        editing/inserting/insert-3786362-fix.html
        editing/inserting/insert-3800346-fix.html
        editing/inserting/insert-3851164-fix.html
        editing/inserting/insert-3907422-fix.html
        editing/inserting/insert-after-delete-001.html
        editing/inserting/insert-at-end-01.html
        editing/inserting/insert-at-end-02.html
        editing/inserting/insert-before-link-1.html
        editing/inserting/insert-br-001.html
        editing/inserting/insert-br-002.html
        editing/inserting/insert-br-003.html
        editing/inserting/insert-br-004.html
        editing/inserting/insert-br-005.html
        editing/inserting/insert-br-006.html
        editing/inserting/insert-br-007.html
        editing/inserting/insert-br-008.html
        editing/inserting/insert-br-009.html
        editing/inserting/insert-br-at-tabspan-001.html
        editing/inserting/insert-br-at-tabspan-002.html
        editing/inserting/insert-br-at-tabspan-003.html
        editing/inserting/insert-br-quoted-001.html
        editing/inserting/insert-br-quoted-002.html
        editing/inserting/insert-br-quoted-003.html
        editing/inserting/insert-br-quoted-004.html
        editing/inserting/insert-br-quoted-005.html
        editing/inserting/insert-br-quoted-006.html
        editing/inserting/insert-div-001.html
        editing/inserting/insert-div-002.html
        editing/inserting/insert-div-003.html
        editing/inserting/insert-div-004.html
        editing/inserting/insert-div-005.html
        editing/inserting/insert-div-006.html
        editing/inserting/insert-div-007.html
        editing/inserting/insert-div-008.html
        editing/inserting/insert-div-009.html
        editing/inserting/insert-div-010.html
        editing/inserting/insert-div-011.html
        editing/inserting/insert-div-012.html
        editing/inserting/insert-div-013.html
        editing/inserting/insert-div-014.html
        editing/inserting/insert-div-015.html
        editing/inserting/insert-div-016.html
        editing/inserting/insert-div-017.html
        editing/inserting/insert-div-018.html
        editing/inserting/insert-div-019.html
        editing/inserting/insert-div-020.html
        editing/inserting/insert-div-021.html
        editing/inserting/insert-div-022.html
        editing/inserting/insert-div-024.html
        editing/inserting/insert-div-025.html
        editing/inserting/insert-div-026.html
        editing/inserting/insert-div-027.html
        editing/inserting/insert-paragraph-01.html
        editing/inserting/insert-paragraph-02.html
        editing/inserting/insert-paragraph-03.html
        editing/inserting/insert-paragraph-04.html
        editing/inserting/insert-paragraph-05.html
        editing/inserting/insert-space-in-empty-doc.html
        editing/inserting/insert-tab-001.html
        editing/inserting/insert-tab-002.html
        editing/inserting/insert-tab-004.html
        editing/inserting/insert-text-at-tabspan-001.html
        editing/inserting/insert-text-at-tabspan-002.html
        editing/inserting/insert-text-at-tabspan-003.html
        editing/inserting/line-break.html
        editing/inserting/paragraph-separator-01.html
        editing/inserting/paragraph-separator-02.html
        editing/inserting/paragraph-separator-in-table-1.html
        editing/inserting/paragraph-separator-in-table-2.html
        editing/inserting/redo.html
        editing/inserting/return-key-with-selection-003.html
        editing/inserting/typing-001.html
        editing/inserting/typing-002.html
        editing/inserting/typing-003.html
        editing/inserting/typing-around-br-001.html
        editing/inserting/typing-around-image-001.html

        editing/inserting/5058163-1.html
        editing/inserting/edited-whitespace-1.html
        editing/inserting/paragraph-separator-03.html
        editing/inserting/return-key-with-selection-001.html
        editing/inserting/return-key-with-selection-002.html
        editing/inserting/4875189-2.html

        * platform/qt/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59052 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-09 Daniel Bates <dbates@rim.com>
dbates@webkit.org [Sun, 9 May 2010 09:50:32 +0000 (09:50 +0000)]
2010-05-09  Daniel Bates  <dbates@rim.com>

        Reviewed by Chris Jerdonek.

        https://bugs.webkit.org/show_bug.cgi?id=38812

        Make the regular expressions for parsing the start of an SVN
        and Git header global variables since they are used throughout
        VCSUtils.pm.

        * Scripts/VCSUtils.pm:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59051 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Daniel Bates <dbates@rim.com>
dbates@webkit.org [Sun, 9 May 2010 09:08:55 +0000 (09:08 +0000)]
2010-05-07  Daniel Bates  <dbates@rim.com>

         Unreviewed, change log fix.

         Add back some entries inadvertently deleted in changeset 59000
         <http://trac.webkit.org/changeset/59000> and not restored (as
         intended) in changeset 59005 <http://trac.webkit.org/changeset/59005>.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59050 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoAdd back some entries accidentally deleted in r59000 and not restored as intended...
mjs@apple.com [Sun, 9 May 2010 08:42:22 +0000 (08:42 +0000)]
Add back some entries accidentally deleted in r59000 and not restored as intended in r59005.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59049 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-09 Chris Jerdonek <cjerdonek@webkit.org>
cjerdonek@webkit.org [Sun, 9 May 2010 07:21:35 +0000 (07:21 +0000)]
2010-05-09  Chris Jerdonek  <cjerdonek@webkit.org>

        Reviewed by Daniel Bates.

        Finished moving the header-parsing logic from svn-apply and -unapply
        to VCSUtils.pm's parsing methods.

        https://bugs.webkit.org/show_bug.cgi?id=38802

        * Scripts/VCSUtils.pm:
          - Added to parseGitDiffHeader() the ability to parse and store
            whether a file is new or deleted.
          - Also reordered in parseGitDiffHeader() some of the else statements
            to a more readable ordering.
          - Added to parseSvnDiffHeader() the ability to parse and store
            whether a file is new.
        * Scripts/svn-apply:
          - Changed handleGitBinaryChange() to use the new "isNew" and "isDeletion"
            diffHash key-values.
          - Changed patch() to use the new "isNew" diffHash key-value.
        * Scripts/svn-unapply:
          - Changed patch() to use the new "isNew" and "isDeletion" diffHash key-values.
        * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
          - Added unit tests for new and deleted files.
        * Scripts/webkitperl/VCSUtils_unittest/parseGitDiffHeader.pl:
          - Updated the unit tests as necessary.
          - Added a unit test for a deleted file.
        * Scripts/webkitperl/VCSUtils_unittest/parseSvnDiffHeader.pl:
          - Updated the unit tests as necessary.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59048 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Chris Jerdonek <cjerdonek@webkit.org>
cjerdonek@webkit.org [Sun, 9 May 2010 06:50:36 +0000 (06:50 +0000)]
2010-05-08  Chris Jerdonek  <cjerdonek@webkit.org>

        Reviewed by Maciej Stachowiak.

        Allow long comment lines to wrap to the window width instead
        of to 80 characters.

        https://bugs.webkit.org/show_bug.cgi?id=37792

        * Bugzilla/Constants.pm:
          - Introduced a new constant COMMENT_COLS_WRAP to replace COMMENT_COLS
            inside the wrap_comment() subroutine.  The new constant effectively
            disables text wrapping by increasing the value used in wrap_comment()
            from 80 characters to a very large value (8000).
        * Bugzilla/Util.pm:
          - Replaced COMMENT_COLS with COMMENT_COLS_WRAP inside wrap_comment().
        * skins/custom/global.css:
        (.bz_comment_text):
          - Eliminate the width styling on comments to allow long lines
            to wrap at the window width.
        (.bz_comment_text > span.quote):
          - Inherit the "white-space" property.  This prevents long quoted
            lines in comments from causing non-quoted lines to extend beyond
            the window width.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59047 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Antonio Gomes <tonikitoo@webkit.org>
tonikitoo@webkit.org [Sun, 9 May 2010 03:21:00 +0000 (03:21 +0000)]
2010-05-08  Antonio Gomes  <tonikitoo@webkit.org>

        Unreviewed attempt to fix Windows build.

        * page/SpatialNavigation.h:
        (WebCore::fudgeFactor):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59046 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoSpatial Navigation: create a getter for the "fudgeFactor"
tonikitoo@webkit.org [Sun, 9 May 2010 03:05:46 +0000 (03:05 +0000)]
Spatial Navigation: create a getter for the "fudgeFactor"
https://bugs.webkit.org/show_bug.cgi?id=38488

Reviewed by Kenneth Christiansen.
Patch by Antonio Gomes <tonikitoo@webkit.org>

A couple of places in the Spatial Navigation code make use of a "fudge factor"
to improve precision by working around outline focus metrics and such. Patch adds
a helper method for unify getter operations of this value, instead of having it
declared locally in the various methods it is used.

No behaviour change.

* page/SpatialNavigation.cpp:
(WebCore::scrollIntoView):
(WebCore::deflateIfOverlapped):
* page/SpatialNavigation.h:
(WebCore::fudgeFactor):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59045 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoAdded to svn-apply support for git renames and copies with similarity
cjerdonek@webkit.org [Sun, 9 May 2010 02:52:26 +0000 (02:52 +0000)]
Added to svn-apply support for git renames and copies with similarity
index less than 100%.

Reviewed by Daniel Bates.

https://bugs.webkit.org/show_bug.cgi?id=32834

* Scripts/VCSUtils.pm:
  - Added to parseGitDiffHeader() support for renames and similarity
    index less than 100%.
  - Added to parseDiff() support for processing renames, renames with
    changes, and copies with changes.
  - Added to parsePatch() the ability to process multiple return
    values from parseDiff().
* Scripts/svn-apply:
  - Added to patch() the ability to process diff hashes with the
    isDeletion key-value set.
* Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
  - Updated the unit tests as necessary.
  - Added unit tests for rename with similarity index 100%,
    rename with similarity index < 100%, and rename with a change
    to the executable bit.
* Scripts/webkitperl/VCSUtils_unittest/parseGitDiffHeader.pl:
  - Added unit tests for rename with similarity index 100%,
    rename with similarity index < 100%, and rename with a change
    to the executable bit.
* Scripts/webkitperl/VCSUtils_unittest/parsePatch.pl:
  - Updated the unit tests as necessary.
* Scripts/webkitperl/VCSUtils_unittest/parseSvnDiffHeader.pl:
  - Simplified the carriage-return unit test to more narrowly test
    only carriage returns.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59044 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoThis revision suppresses the misleading "error: pathspec..." messages
cjerdonek@webkit.org [Sun, 9 May 2010 02:37:10 +0000 (02:37 +0000)]
This revision suppresses the misleading "error: pathspec..." messages
when using svn-apply to add a binary file from a Git diff.

Reviewed by Daniel Bates.

https://bugs.webkit.org/show_bug.cgi?id=38047

When adding a new binary file from a Git diff, svn-apply prints
a misleading error of the form -- "error: pathspec '<filename>' did
not match any file(s) known to git.  Did you forget to 'git add'?"
This patch suppresses these messages since they are normal.

* Scripts/VCSUtils.pm:
  Added the callSilently() subroutine from runPatchCommand.pl, which
  executes a Perl function while suppressing STDERR.
* Scripts/svn-apply:
  Refactored the Git portion of scmKnowsOfFile() into a
  gitKnowsOfFile(), and called this new subroutine using callSilently().
* Scripts/webkitperl/VCSUtils_unittest/runPatchCommand.pl:
  Removed callSilently() since it was moved to VCSUtils.pm in this patch.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59043 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Laszlo Gombos <laszlo.1.gombos@nokia.com>
laszlo.1.gombos@nokia.com [Sun, 9 May 2010 02:02:24 +0000 (02:02 +0000)]
2010-05-08  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Unreviewed, build fix.

        Fix missing terminating ' character warning

        The warning has been only seen on the Gtk buildbots.

        * wtf/Platform.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59042 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Laszlo Gombos <laszlo.1.gombos@nokia.com>
eric@webkit.org [Sun, 9 May 2010 00:38:43 +0000 (00:38 +0000)]
2010-05-08  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Reviewed by Gavin Barraclough.

        Refactor YARR and YARR_JIT default rules in Platform.h
        https://bugs.webkit.org/show_bug.cgi?id=38727

        Turn on YARR and YARR JIT by default of JIT is enabled.
        Ports can overrule this default decisions as they wish.

        * wtf/Platform.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59041 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Oliver Hunt <oliver@apple.com>
oliver@apple.com [Sat, 8 May 2010 22:59:03 +0000 (22:59 +0000)]
2010-05-08  Oliver Hunt  <oliver@apple.com>

        Reviewed by Gavin Barraclough.

        Split JSVALUE32_64 code out of JITOpcodes.cpp and into JITOpcodes32_64.cpp
        https://bugs.webkit.org/show_bug.cgi?id=38808

        * GNUmakefile.am:
        * JavaScriptCore.gypi:
        * JavaScriptCore.pro:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * jit/JITOpcodes.cpp:
        * jit/JITOpcodes32_64.cpp: Added.
        (JSC::JIT::privateCompileCTIMachineTrampolines):
        (JSC::JIT::emit_op_mov):
        (JSC::JIT::emit_op_end):
        (JSC::JIT::emit_op_jmp):
        (JSC::JIT::emit_op_loop_if_lesseq):
        (JSC::JIT::emitSlow_op_loop_if_lesseq):
        (JSC::JIT::emit_op_new_object):
        (JSC::JIT::emit_op_instanceof):
        (JSC::JIT::emitSlow_op_instanceof):
        (JSC::JIT::emit_op_new_func):
        (JSC::JIT::emit_op_get_global_var):
        (JSC::JIT::emit_op_put_global_var):
        (JSC::JIT::emit_op_get_scoped_var):
        (JSC::JIT::emit_op_put_scoped_var):
        (JSC::JIT::emit_op_tear_off_activation):
        (JSC::JIT::emit_op_tear_off_arguments):
        (JSC::JIT::emit_op_new_array):
        (JSC::JIT::emit_op_resolve):
        (JSC::JIT::emit_op_to_primitive):
        (JSC::JIT::emitSlow_op_to_primitive):
        (JSC::JIT::emit_op_strcat):
        (JSC::JIT::emit_op_resolve_base):
        (JSC::JIT::emit_op_resolve_skip):
        (JSC::JIT::emit_op_resolve_global):
        (JSC::JIT::emitSlow_op_resolve_global):
        (JSC::JIT::emit_op_not):
        (JSC::JIT::emitSlow_op_not):
        (JSC::JIT::emit_op_jfalse):
        (JSC::JIT::emitSlow_op_jfalse):
        (JSC::JIT::emit_op_jtrue):
        (JSC::JIT::emitSlow_op_jtrue):
        (JSC::JIT::emit_op_jeq_null):
        (JSC::JIT::emit_op_jneq_null):
        (JSC::JIT::emit_op_jneq_ptr):
        (JSC::JIT::emit_op_jsr):
        (JSC::JIT::emit_op_sret):
        (JSC::JIT::emit_op_eq):
        (JSC::JIT::emitSlow_op_eq):
        (JSC::JIT::emit_op_neq):
        (JSC::JIT::emitSlow_op_neq):
        (JSC::JIT::compileOpStrictEq):
        (JSC::JIT::emit_op_stricteq):
        (JSC::JIT::emitSlow_op_stricteq):
        (JSC::JIT::emit_op_nstricteq):
        (JSC::JIT::emitSlow_op_nstricteq):
        (JSC::JIT::emit_op_eq_null):
        (JSC::JIT::emit_op_neq_null):
        (JSC::JIT::emit_op_resolve_with_base):
        (JSC::JIT::emit_op_new_func_exp):
        (JSC::JIT::emit_op_throw):
        (JSC::JIT::emit_op_get_pnames):
        (JSC::JIT::emit_op_next_pname):
        (JSC::JIT::emit_op_push_scope):
        (JSC::JIT::emit_op_pop_scope):
        (JSC::JIT::emit_op_to_jsnumber):
        (JSC::JIT::emitSlow_op_to_jsnumber):
        (JSC::JIT::emit_op_push_new_scope):
        (JSC::JIT::emit_op_catch):
        (JSC::JIT::emit_op_jmp_scopes):
        (JSC::JIT::emit_op_switch_imm):
        (JSC::JIT::emit_op_switch_char):
        (JSC::JIT::emit_op_switch_string):
        (JSC::JIT::emit_op_new_error):
        (JSC::JIT::emit_op_debug):
        (JSC::JIT::emit_op_enter):
        (JSC::JIT::emit_op_enter_with_activation):
        (JSC::JIT::emit_op_create_arguments):
        (JSC::JIT::emit_op_init_arguments):
        (JSC::JIT::emit_op_convert_this):
        (JSC::JIT::emitSlow_op_convert_this):
        (JSC::JIT::emit_op_profile_will_call):
        (JSC::JIT::emit_op_profile_did_call):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59040 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Stuart Morgan <stuartmorgan@chromium.org>
eric@webkit.org [Sat, 8 May 2010 20:47:38 +0000 (20:47 +0000)]
2010-05-08  Stuart Morgan  <stuartmorgan@chromium.org>

        Reviewed by Darin Adler.

        Change the npapi.h header guards to match the npapi-headers repository
        copy.

        https://bugs.webkit.org/show_bug.cgi?id=38666

        * bridge/npapi.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59039 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Gabor Loki <loki@webkit.org>
eric@webkit.org [Sat, 8 May 2010 20:03:31 +0000 (20:03 +0000)]
2010-05-08  Gabor Loki  <loki@webkit.org>

        Reviewed by Gavin Barraclough.

        Fix halfword loads on ARM
        https://bugs.webkit.org/show_bug.cgi?id=38741

        The BaseIndex and ImplicitAddress are contain 32bit wide offset, but
        the load16 functions were working with 8 bit data (encoded in the
        instruction). If the offset cannot be encoded in an instruction, it
        should be stored in a temporary register.

        * assembler/ARMAssembler.h:
        (JSC::ARMAssembler::getOffsetForHalfwordDataTransfer):
        * assembler/MacroAssemblerARM.h:
        (JSC::MacroAssemblerARM::load16):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59038 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Gabor Loki <loki@webkit.org>
eric@webkit.org [Sat, 8 May 2010 19:51:23 +0000 (19:51 +0000)]
2010-05-08  Gabor Loki  <loki@webkit.org>

        Reviewed by Gavin Barraclough.

        Fix spanning branch instruction on Cortex-A8 with Thumb-2 JIT
        https://bugs.webkit.org/show_bug.cgi?id=38280

        If the 32-bit Thumb-2 branch instruction spans two 4KiB regions and
        the target of the branch falls within the first region it is
        possible for the processor to incorrectly determine the branch
        instruction, and it is also possible in some cases for the processor
        to enter a deadlock state.

        * assembler/ARMv7Assembler.h:
        (JSC::ARMv7Assembler::linkJumpAbsolute):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59037 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Victor Wang <victorw@chromium.org>
eric@webkit.org [Sat, 8 May 2010 19:36:42 +0000 (19:36 +0000)]
2010-05-08  Victor Wang  <victorw@chromium.org>

        Reviewed by Adam Barth.

        [chromium]: Upload test results json files to app engine server
        Add an option to run_webkit_tests.py to upload generated
        JSON files to app engine server. These JSON files will be used
        by chromium layout test falkiness dashboard.

        https://bugs.webkit.org/show_bug.cgi?id=36063

        * Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py: Added.
        * Scripts/webkitpy/layout_tests/run_webkit_tests.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59036 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Stuart Morgan <stuartmorgan@chromium.org>
eric@webkit.org [Sat, 8 May 2010 19:23:33 +0000 (19:23 +0000)]
2010-05-08  Stuart Morgan  <stuartmorgan@chromium.org>

        Reviewed by Darin Fisher.

        Resolve most of the whitespace/comment/order differences in npapi.h
        and npruntime.h relative to the npapi-headers repository versions.
        No changes to the substance of the code.

        https://bugs.webkit.org/show_bug.cgi?id=38666

        * bridge/npapi.h:
        (_NPPrint::):
        (_NPCocoaEvent::):
        * bridge/npruntime.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59035 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Laszlo Gombos <laszlo.1.gombos@nokia.com>
eric@webkit.org [Sat, 8 May 2010 19:05:46 +0000 (19:05 +0000)]
2010-05-08  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Reviewed by Darin Adler.

        Fix guard around IndexedDatabase
        https://bugs.webkit.org/show_bug.cgi?id=38710

        No new tests, there is no new functionality.

        * page/PageGroup.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59034 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Luiz Agostini <luiz.agostini@openbossa.org>
eric@webkit.org [Sat, 8 May 2010 18:51:48 +0000 (18:51 +0000)]
2010-05-08  Luiz Agostini  <luiz.agostini@openbossa.org>

        Reviewed by Simon Hausmann.

        [Qt] Platform plugin
        https://bugs.webkit.org/show_bug.cgi?id=38438

        Select popups delegate private API. To provide select popup delegates
        a plugin that implements QWebKitPlatformPlugin must be found
        in QCoreApplication::libraryPaths.

        * WebCore.pro:
2010-05-08  Luiz Agostini  <luiz.agostini@openbossa.org>

        Reviewed by Simon Hausmann.

        [Qt] Platform plugin
        https://bugs.webkit.org/show_bug.cgi?id=38438

        Select popups delegate private API. To provide select popup delegates
        a plugin that implements QWebKitPlatformPlugin must be found
        in QCoreApplication::libraryPaths.

        * Api/headers.pri:
        * Api/qwebkitplatformplugin.h: Added.
        (QWebSelectData::~QWebSelectData):
        (QWebSelectData::):
        (QWebSelectMethod::~QWebSelectMethod):
        (QWebKitPlatformPlugin::~QWebKitPlatformPlugin):
        (QWebKitPlatformPlugin::):
        * WebCoreSupport/ChromeClientQt.cpp:
        (WebCore::ChromeClientQt::createSelectPopup):
        * WebCoreSupport/ChromeClientQt.h:
        * WebCoreSupport/QtPlatformPlugin.cpp: Added.
        (WebCore::SelectData::SelectData):
        (WebCore::SelectData::itemText):
        (WebCore::SelectData::itemToolTip):
        (WebCore::SelectData::itemIsEnabled):
        (WebCore::SelectData::itemCount):
        (WebCore::SelectData::itemIsSelected):
        (WebCore::SelectData::multiple):
        (WebCore::SelectData::itemType):
        (WebCore::SelectInputMethodWrapper::SelectInputMethodWrapper):
        (WebCore::SelectInputMethodWrapper::show):
        (WebCore::SelectInputMethodWrapper::hide):
        (WebCore::SelectInputMethodWrapper::selectItem):
        (WebCore::SelectInputMethodWrapper::didHide):
        (WebCore::getPluginObject):
        (WebCore::QtPlatformPlugin::~QtPlatformPlugin):
        (WebCore::QtPlatformPlugin::plugin):
        (WebCore::QtPlatformPlugin::createSelectInputMethod):
        * WebCoreSupport/QtPlatformPlugin.h: Added.
        (WebCore::QtPlatformPlugin::QtPlatformPlugin):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59033 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Leandro Pereira <leandro@profusion.mobi>
eric@webkit.org [Sat, 8 May 2010 18:40:06 +0000 (18:40 +0000)]
2010-05-08  Leandro Pereira  <leandro@profusion.mobi>

        Reviewed by Laszlo Gombos.

        [EFL] Only compile database stuff if ENABLE_DATABASE is set.
        http://webkit.org/b/38777

        * efl/WebCoreSupport/ChromeClientEfl.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59032 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Laszlo Gombos <laszlo.1.gombos@nokia.com>
eric@webkit.org [Sat, 8 May 2010 18:29:02 +0000 (18:29 +0000)]
2010-05-08  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt] Build the ImageDiff tool for all platforms including Windows and Symbian
        https://bugs.webkit.org/show_bug.cgi?id=38706

        * WebKit.pro:
2010-05-08  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt] Build the ImageDiff tool for all platforms including Windows and Symbian
        https://bugs.webkit.org/show_bug.cgi?id=38706

        Use qmath.h instead of math.h to make it portable.

        * DumpRenderTree/qt/ImageDiff.cpp:
        (main):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59031 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Leandro Pereira <leandro@profusion.mobi>
eric@webkit.org [Sat, 8 May 2010 18:17:57 +0000 (18:17 +0000)]
2010-05-08  Leandro Pereira  <leandro@profusion.mobi>

        Reviewed by Gustavo Noronha Silva.

        [EFL] It does not make sense to use RenderThemeEfl::RenderThemeEfl, so
        use RenderThemeEfl (GCC 4.5 seems to be pickier about this).
        http://webkit.org/b/38770

        * platform/efl/RenderThemeEfl.cpp:
        (WebCore::renderThemeEflColorClassSelectionActive):
        (WebCore::renderThemeEflColorClassSelectionInactive):
        (WebCore::renderThemeEflColorClassFocusRing):
        (WebCore::renderThemeEflColorClassButtonText):
        (WebCore::renderThemeEflColorClassComboText):
        (WebCore::renderThemeEflColorClassEntryText):
        (WebCore::renderThemeEflColorClassSearchText):
        (WebCore::RenderThemeEfl::adjustCheckboxStyle):
        (WebCore::RenderThemeEfl::adjustRadioStyle):
        (WebCore::RenderThemeEfl::adjustButtonStyle):
        (WebCore::RenderThemeEfl::adjustMenuListStyle):
        (WebCore::RenderThemeEfl::adjustTextFieldStyle):
        (WebCore::RenderThemeEfl::adjustSearchFieldDecorationStyle):
        (WebCore::RenderThemeEfl::adjustSearchFieldResultsButtonStyle):
        (WebCore::RenderThemeEfl::adjustSearchFieldResultsDecorationStyle):
        (WebCore::RenderThemeEfl::adjustSearchFieldCancelButtonStyle):
        (WebCore::RenderThemeEfl::adjustSearchFieldStyle):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59030 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Jens Alfke <snej@chromium.org>
eric@webkit.org [Sat, 8 May 2010 18:06:37 +0000 (18:06 +0000)]
2010-05-08  Jens Alfke  <snej@chromium.org>

        Reviewed by Darin Fisher.

        [chromium] WebFrame::registerPasswordListener shouldn't assert on duplicate listener.
        https://bugs.webkit.org/show_bug.cgi?id=38765

        * public/WebFrame.h:
        (WebKit::WebFrameImpl::registerPasswordListener): Changed return type to bool. Updated comment.
        * src/WebFrameImpl.cpp:
        (WebKit::WebFrameImpl::registerPasswordListener): Return false instead of asserting.
        * src/WebFrameImpl.h:
        (WebKit::WebFrameImpl::registerPasswordListener): Changed return type to bool.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59029 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Leandro Pereira <leandro@profusion.mobi>
eric@webkit.org [Sat, 8 May 2010 17:53:41 +0000 (17:53 +0000)]
2010-05-08  Leandro Pereira  <leandro@profusion.mobi>

        Reviewed by Gustavo Noronha Silva.

        ResourceHandleSoup should depend only on GLib, not on GTK+.
        http://webkit.org/b/38763

        * platform/network/soup/ResourceHandleSoup.cpp: Include glib.h
        instead of gtk/gtk.h.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59028 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 16:49:51 +0000 (16:49 +0000)]
2010-05-08  Eric Seidel  <eric@webkit.org>

        Unreviewed, rolling out r59023.
        http://trac.webkit.org/changeset/59023
        https://bugs.webkit.org/show_bug.cgi?id=37874

        Caused Chromium Windows build to fail to link.

        * WebCore.gypi:
        * loader/CachedMetadata.h: Removed.
        * loader/CachedResource.cpp:
        * loader/CachedResource.h:
        * loader/ResourceLoader.h:
        * loader/SubresourceLoader.cpp:
        * loader/SubresourceLoader.h:
        * loader/SubresourceLoaderClient.h:
        * loader/loader.cpp:
        * loader/loader.h:
        * platform/network/ResourceHandle.cpp:
        * platform/network/ResourceHandle.h:
        * platform/network/ResourceHandleClient.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59027 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>
eric@webkit.org [Sat, 8 May 2010 13:31:12 +0000 (13:31 +0000)]
2010-05-08  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>

        Reviewed by Simon Hausmann.

        [Qt] Document a limitation of QWebPage::linkClicked
        https://bugs.webkit.org/show_bug.cgi?id=37694

        * Api/qwebpage.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59026 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Ilya Tikhonovsky <loislo@chromium.org>
eric@webkit.org [Sat, 8 May 2010 13:20:24 +0000 (13:20 +0000)]
2010-05-08  Ilya Tikhonovsky  <loislo@chromium.org>

        Reviewed by Timothy Hatcher.

        WebInspector: Autocompletion in console is not usable because it is not add to the entered text the common prefix of suggestions.
        As example if you entered $0.getE then getElementByClassName will be suggested as gray text.
        If you press tab then getElementByClassName be suggested again and lementByClassName becomes selected as black text with selection background.
        Really after pressing the TAB the next suggestion should be selected (getElementByTagName) and only flexible part of suggestions should be selected.
        In this case it should be TagName because getElementBy is the common part for all getE prefix suggestions.
        https://bugs.webkit.org/show_bug.cgi?id=38753

        * inspector/front-end/TextPrompt.js:
        (WebInspector.TextPrompt.prototype._completionsReady):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59025 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Adam Barth <abarth@webkit.org>
eric@webkit.org [Sat, 8 May 2010 13:09:37 +0000 (13:09 +0000)]
2010-05-08  Adam Barth  <abarth@webkit.org>

        Reviewed by Darin Fisher.

        [WTFURL] Add a buffer for managing the output of the canonicalizer
        https://bugs.webkit.org/show_bug.cgi?id=38671

        The canonicalizer produces a stream of canonicalized characters, which
        we will store in a URLBuffer.  URLBuffer is quite similar to Vector,
        but serves as an abstraction to isolate us from the underlying
        character storage.  In the case of WebKit, we'll probably implement a
        subclass of URLBuffer with Vector.  In Chromium, the backing store is a
        std::string, which avoids an unnecessary memcpy that would be caused by
        using Vector instead.

        * wtf/url/src/URLBuffer.h: Added.
        (WTF::URLBuffer::URLBuffer):
        (WTF::URLBuffer::~URLBuffer):
        (WTF::URLBuffer::at):
        (WTF::URLBuffer::set):
        (WTF::URLBuffer::capacity):
        (WTF::URLBuffer::length):
        (WTF::URLBuffer::data):
        (WTF::URLBuffer::setLength):
        (WTF::URLBuffer::append):
        (WTF::URLBuffer::grow):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59024 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Tony Gentilcore <tonyg@chromium.org>
eric@webkit.org [Sat, 8 May 2010 12:58:37 +0000 (12:58 +0000)]
2010-05-08  Tony Gentilcore  <tonyg@chromium.org>

        Reviewed by Darin Fisher.

        Provide mechanism to store cached metadata on a CachedResource.
        https://bugs.webkit.org/show_bug.cgi?id=37874

        No new tests because no new functionality.

        * WebCore.gypi:
        * loader/CachedMetadata.h: Added.
        (WebCore::CachedMetadata::~CachedMetadata):
        (WebCore::CachedMetadata::data):
        (WebCore::CachedMetadata::size):
        (WebCore::CachedMetadata::create):
        (WebCore::CachedMetadata::deserialize):
        (WebCore::CachedMetadata::serialize):
        (WebCore::CachedMetadata::dataTypeID):
        (WebCore::CachedMetadata::readUnsigned):
        (WebCore::CachedMetadata::appendUnsigned):
        (WebCore::CachedMetadata::CachedMetadata):
        * loader/CachedResource.cpp:
        (WebCore::CachedResource::setSerializedCachedMetadata):
        (WebCore::CachedResource::setCachedMetadata):
        (WebCore::CachedResource::cachedMetadata):
        * loader/CachedResource.h:
        * loader/ResourceLoader.h:
        (WebCore::ResourceLoader::didReceiveCachedMetadata):
        * loader/SubresourceLoader.cpp:
        (WebCore::SubresourceLoader::didReceiveCachedMetadata):
        * loader/SubresourceLoader.h:
        * loader/SubresourceLoaderClient.h:
        (WebCore::SubresourceLoaderClient::didReceiveCachedMetadata):
        * loader/loader.cpp:
        (WebCore::Loader::Host::didReceiveCachedMetadata):
        * loader/loader.h:
        * platform/network/ResourceHandle.cpp:
        (WebCore::ResourceHandle::cacheMetadata):
        * platform/network/ResourceHandle.h:
        * platform/network/ResourceHandleClient.h:
        (WebCore::ResourceHandleClient::didReceiveCachedMetadata):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59023 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Michael Nordman <michaeln@google.com>
eric@webkit.org [Sat, 8 May 2010 12:39:26 +0000 (12:39 +0000)]
2010-05-08  Michael Nordman  <michaeln@google.com>

        Reviewed by Dmitry Titov.

        Have WorkerScriptLoader provide access to the ResourceResponse for the loaded script
        through the WorkerScriptLoaderClient interface.
        https://bugs.webkit.org/show_bug.cgi?id=38605

        No new tests. There are no script visible changes.

        * workers/WorkerScriptLoader.cpp: Call the client with the response.
        (WebCore::WorkerScriptLoader::didReceiveResponse):
        * workers/WorkerScriptLoaderClient.h: Define a new method to receive the response.
        (WebCore::WorkerScriptLoaderClient::didReceiveResponse):
2010-05-08  Michael Nordman  <michaeln@google.com>

        Reviewed by Dmitry Titov.

        Add an additional parameter to the WebSharedWorker::startWorkerContext method
        to indicate which appcache the shared worker script resource was loaded from,
        and provide that value when calling out to chrome via this interface.
        https://bugs.webkit.org/show_bug.cgi?id=38605

        I'm keeping the old method around until patches can land in both repositories
        that make the old method obsolete and unneeded. There's a FIXME for that.

        Also somewhat related but independent change to propagate the runtime enabled
        flag thru to the page settings instance used by the 'shadow page'.

        * public/WebSharedWorker.h: Add the new param (but keep the old interface around for now too)
        (WebKit::WebSharedWorker::startWorkerContext):
        * src/SharedWorkerRepository.cpp: Provide the value when calling out to chrome.
        (WebCore::SharedWorkerScriptLoader::notifyFinished):
        * src/WebSharedWorkerImpl.cpp:
        (WebKit::WebSharedWorkerImpl::startWorkerContext):
        * src/WebSharedWorkerImpl.h:
        * src/WebWorkerBase.cpp: Propagate runtime enabled flag thru to the page settings.
        (WebKit::WebWorkerBase::initializeLoader):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59022 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Robert Hogan <robert@webkit.org>
robert@webkit.org [Sat, 8 May 2010 12:13:52 +0000 (12:13 +0000)]
2010-05-08  Robert Hogan  <robert@webkit.org>

        Unreviewed Qt bot fix: test unskipped accidentally in r59020.

        * platform/qt/Skipped: Uncommented http/tests/xmlhttprequest/failed-auth.html again.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59021 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[Qt] Fix http/tests/xmlhttprequest/cross-origin-no-authorization.html
robert@webkit.org [Sat, 8 May 2010 11:50:50 +0000 (11:50 +0000)]
[Qt] Fix http/tests/xmlhttprequest/cross-origin-no-authorization.html

2010-05-08  Robert Hogan  <robert@roberthogan.net>

        Reviewed by Simon Hausmann.

        [Qt] Fix http/tests/xmlhttprequest/cross-origin-no-authorization.html
             and http/tests/xmlhttprequest/cross-origin-authorization.html

        QHttpNetworkRequest adds Authorization and Cookie headers to XHRs
        without knowing if this is valid behaviour or not. In order to allow
        Qt to decide whether Cookie/Authorization headers should be added
        to an XHR QtWebKit needs to use an attribute added to QNetworkRequest.
        These new attributes are: QNetworkRequest::CookieLoadControlAttribute,
        QNetworkRequest::CookieSaveControlAttribute,and
        QNetworkRequest::AuthenticationReuseControlAttribute.

        QtWebKit will set QNetworkRequest::AuthenticationReuseControlAttribute to false
        unless withCredentials is set to true in the XHR.

        QtWebkit will set CookieLoad/SaveControlAttribute to false unless withCredentials
        is set to true in the XHR.

        Qt will pass the values onto QHttpNetworkRequest and this will permit
        the Qt network access processing to decide whether or not to add either
        or both of the Cookie/Authorisation headers, and whether to save cookies
        returned from such requests. By default the attribute
        will always be true so unless QtWebKit sets it to false normal
        header processing is unaffected.

        There are platform-specific results for cross-origin-no-authorization.html:

        1. Qt's DRT does not capture the final 'DONE'.
        2. QNetworkReplyHandler.cpp's ignoreHttpError() means that async requests
           do not execute didFail() when they receive a 401 HTTP response. This
           results in async tests passing with '401 Authorisation required' rather
           than 'received error event'.

        The Qt part of these changes was merged at: http://gitorious.org/qt/qt/merge_requests/592

        In order to properly support the tests, Qt's DRT needs to use one
        NetworkAccessManager for all pages. This allows it to use cached
        credentials where appropriate.

        The tests now pass when run individually but there seems to be a problem with
        leaking the results of requests across tests when run with the others in
        http/tests. This will be addressed in a separate patch.

        https://bugs.webkit.org/show_bug.cgi?id=32967

        * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
        (WebCore::WebPage::WebPage):
        (WebCore::DumpRenderTree::DumpRenderTree):
        * DumpRenderTree/qt/DumpRenderTreeQt.h:
        * platform/network/qt/ResourceRequestQt.cpp:
        (WebCore::ResourceRequest::toNetworkRequest):
        * platform/qt/Skipped: Group together the tests that are waiting for 4.7 to be unskipped.
        * platform/qt/http/tests/xmlhttprequest/cross-origin-no-authorization-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59020 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Ilya Tikhonovsky <loislo@chromium.org>
eric@webkit.org [Sat, 8 May 2010 11:48:53 +0000 (11:48 +0000)]
2010-05-08  Ilya Tikhonovsky  <loislo@chromium.org>

        Reviewed by Pavel Feldman.

        WebInsector: Record expand UI element should be extracted from TimelineRecordGraphRow as separate class for future reuse.
        https://bugs.webkit.org/show_bug.cgi?id=38726

        * inspector/front-end/TimelinePanel.js:
        (WebInspector.TimelinePanel):
        (WebInspector.TimelinePanel.prototype._refreshRecords):
        (WebInspector.TimelineRecordGraphRow):
        (WebInspector.TimelineRecordGraphRow.prototype.update):
        (WebInspector.TimelineExpandableElement):
        (WebInspector.TimelineExpandableElement.prototype._update):
        (WebInspector.TimelineExpandableElement.prototype._dispose):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59019 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Nikolas Zimmermann <nzimmermann@rim.com>
zimmermann@webkit.org [Sat, 8 May 2010 11:08:21 +0000 (11:08 +0000)]
2010-05-08  Nikolas Zimmermann  <nzimmermann@rim.com>

        Not reviewed. Forgot to rebase fast/dom/prototype-inheritance-2.html after the <vkern> patches.

        * platform/win/fast/dom/prototype-inheritance-2-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59018 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[Qt] unskip http/tests/plugins/npapi-response-headers.html
robert@webkit.org [Sat, 8 May 2010 11:01:32 +0000 (11:01 +0000)]
[Qt] unskip http/tests/plugins/npapi-response-headers.html

2010-05-04  Robert Hogan  <robert@webkit.org>

        Reviewed by Eric Seidel.

        [Qt] unskip http/tests/plugins/npapi-response-headers.html

        Turns out this failed because run-webkit-tests was eating the first occurrence
        of 'Content-Type: text/plain' in the test output as a header. Strange but true.
        So do as Chromium does and preface all text dumps with the
        'Content-Type: text/plain' header.

        Take this opportunity to unskip plugins/private-browsing-mode.html too, which
        has passed since r55433. See https://bugs.webkit.org/show_bug.cgi?id=33180
        and https://bugs.webkit.org/show_bug.cgi?id=35576.

        https://bugs.webkit.org/show_bug.cgi?id=38541

        * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
        (WebCore::DumpRenderTree::dump):
        * platform/qt/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59017 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[Qt] Fix encoding of status message in Plugin API
robert@webkit.org [Sat, 8 May 2010 10:55:29 +0000 (10:55 +0000)]
[Qt] Fix encoding of status message in Plugin API

2010-05-05  Robert Hogan  <robert@webkit.org>

        Reviewed by Alexey Proskuryakov.

        [Qt] Fix encoding of status message in Plugin API

        Spotted by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=35144

        * platform/qt/Skipped: Unskip plugins/set-status.html
        * plugins/PluginView.cpp:
        (WebCore::PluginView::status):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59016 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Sat, 8 May 2010 10:11:21 +0000 (10:11 +0000)]
2010-05-08  Dirk Schulze  <krit@webkit.org>

        Unreviewed win build fix.

        JS bindings were missing in DerivedSources.cpp

        * DerivedSources.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59015 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Nikolas Zimmermann <nzimmermann@rim.com>
zimmermann@webkit.org [Sat, 8 May 2010 10:01:02 +0000 (10:01 +0000)]
2010-05-08  Nikolas Zimmermann  <nzimmermann@rim.com>

        Not reviewed. Rebaseline Win specific results after exposing JSSVGVKernElement/JSSVGHKernElement.

        * platform/win/fast/dom/Window/window-property-descriptors-expected.txt:
        * platform/win/fast/dom/prototype-inheritance-expected.txt:
        * platform/win/fast/js/global-constructors-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59014 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Sat, 8 May 2010 09:56:38 +0000 (09:56 +0000)]
2010-05-08  Dirk Schulze  <krit@webkit.org>

        Uneviewed fix of Win bot

        Added JS bindings to Win build system.

        * WebCore.vcproj/WebCore.vcproj:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59013 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Nikolas Zimmermann <nzimmermann@rim.com>
zimmermann@webkit.org [Sat, 8 May 2010 09:54:35 +0000 (09:54 +0000)]
2010-05-08  Nikolas Zimmermann  <nzimmermann@rim.com>

        Not reviewed. Rebaseline GTK specific results after exposing JSSVGVKernElement/JSSVGHKernElement.

        * platform/gtk/fast/dom/Window/window-property-descriptors-expected.txt:
        * platform/gtk/fast/dom/prototype-inheritance-expected.txt:
        * platform/gtk/fast/js/global-constructors-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59012 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Sat, 8 May 2010 09:28:53 +0000 (09:28 +0000)]
2010-05-08  Dirk Schulze  <krit@webkit.org>

        Unreviewed try to fix Win bot.

        SVGFont's VKern implementation missing
        https://bugs.webkit.org/show_bug.cgi?id=38663

        * bindings/scripts/CodeGeneratorJS.pm:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59011 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Sat, 8 May 2010 09:06:59 +0000 (09:06 +0000)]
2010-05-08  Dirk Schulze  <krit@webkit.org>

        Reviewed by Nikolas Zimmermann.

        SVGFont's VKern implementation missing
        https://bugs.webkit.org/show_bug.cgi?id=38663

        Implementation of vkern, needed by SVGFont. Also added DOM and JS bindings
        for vkern and hkern. Made some clean-up in SVGParserUtilities and SVGFontElement.
        SVGHKernElement and SVGVKernElement share most of the code now.

        Test: svg/text/text-vkern.svg

        * Android.derived.jscbindings.mk:
        * Android.derived.v8bindings.mk:
        * Android.mk:
        * DerivedSources.make:
        * GNUmakefile.am:
        * WebCore.gyp/WebCore.gyp:
        * WebCore.gypi:
        * WebCore.pri:
        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * page/DOMWindow.idl:
        * rendering/SVGRootInlineBox.cpp:
        (WebCore::applySVGKerning):
        (WebCore::SVGRootInlineBox::buildLayoutInformationForTextBox):
        * svg/SVGAllInOne.cpp:
        * svg/SVGFontElement.cpp:
        (WebCore::SVGFontElement::invalidateGlyphCache):
        (WebCore::SVGFontElement::ensureGlyphCache):
        (WebCore::matches):
        (WebCore::kerningForPairOfStringsAndGlyphs):
        (WebCore::SVGFontElement::horizontalKerningForPairOfStringsAndGlyphs):
        (WebCore::SVGFontElement::verticalKerningForPairOfStringsAndGlyphs):
        * svg/SVGFontElement.h:
        (WebCore::SVGKerningPair::SVGKerningPair):
        (WebCore::SVGFontElement::rendererIsNeeded):
        * svg/SVGHKernElement.cpp:
        (WebCore::SVGHKernElement::buildHorizontalKerningPair):
        * svg/SVGHKernElement.h:
        * svg/SVGParserUtilities.cpp:
        (WebCore::parseGlyphName):
        (WebCore::parseKerningUnicodeString):
        * svg/SVGParserUtilities.h:
        * svg/SVGVKernElement.cpp: Added.
        (WebCore::SVGVKernElement::SVGVKernElement):
        (WebCore::SVGVKernElement::~SVGVKernElement):
        (WebCore::SVGVKernElement::insertedIntoDocument):
        (WebCore::SVGVKernElement::removedFromDocument):
        (WebCore::SVGVKernElement::buildVerticalKerningPair):
        * svg/SVGVKernElement.h: Added.
        (WebCore::SVGVKernElement::rendererIsNeeded):
        * svg/SVGVKernElement.idl: Added.
        * svg/svgtags.in:

2010-05-08  Dirk Schulze  <krit@webkit.org>

        Reviewed by Nikolas Zimmermann.

        SVGFont's VKern implementation missing
        https://bugs.webkit.org/show_bug.cgi?id=38663

        The changes in DOMWindow.idl and the implementation of the DOM
        and JS bindings of VKern and HKern made some updates necessary.
        Added new test case for SVGFont's VKern.

        * fast/dom/Window/window-properties-expected.txt:
        * fast/dom/Window/window-property-descriptors-expected.txt:
        * fast/dom/prototype-inheritance-2-expected.txt:
        * fast/dom/prototype-inheritance-expected.txt:
        * fast/js/global-constructors-expected.txt:
        * platform/mac/svg/text/text-vkern-expected.checksum: Added.
        * platform/mac/svg/text/text-vkern-expected.png: Added.
        * platform/mac/svg/text/text-vkern-expected.txt: Added.
        * svg/custom/global-constructors-expected.txt:
        * svg/text/text-vkern.svg: Added.

2010-05-08  Dirk Schulze  <krit@webkit.org>

        Reviewed by Nikolas Zimmermann.

        SVGFont's VKern implementation missing
        https://bugs.webkit.org/show_bug.cgi?id=38663

        * projects/svg/status.xml:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59010 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoUnreviewed, just fixing windows build.
jer.noble@apple.com [Sat, 8 May 2010 08:44:41 +0000 (08:44 +0000)]
Unreviewed, just fixing windows build.

* platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.h: Add #if USE(ACCELERATED_COMPOSITING) guards around GraphicsLayer areas of the code.  Include GraphicsLayer.h directly.
* platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.cpp: As above.
* platform/graphics/win/QTMovieVisualContext.cpp: Remove the #include of d3d9types.h.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59009 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFix the windows build (by adding #if USE(ACCELERATED_COMPOSITING) guards).
jer.noble@apple.com [Sat, 8 May 2010 08:26:32 +0000 (08:26 +0000)]
Fix the windows build (by adding #if USE(ACCELERATED_COMPOSITING) guards).

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59008 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFix the windows build.
jer.noble@apple.com [Sat, 8 May 2010 07:38:07 +0000 (07:38 +0000)]
Fix the windows build.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59007 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 07:24:29 +0000 (07:24 +0000)]
2010-05-08  Eric Seidel  <eric@webkit.org>

        Reviewed by Daniel Bates.

        test-webkitpy fails under cygwin
        https://bugs.webkit.org/show_bug.cgi?id=38718

        * Scripts/webkitpy/common/system/executive.py:
         - Add _KILL_PROCESS_KILLED_PROCESS_EXIT_CODE and
           _KILL_ALL_KILLED_PROCESS_EXIT_CODE to store the expected
           exit codes of processes killed by kill_process and kill_all.
           These two constants are only used by the unit tests but are
           stored in executive.py so they can be right next to the platform ifs.
         - Remove unnecessary str() conversion, run_command does that for us.
         - Make os.kill retry on cygwin on EAGAIN.  It's unclear why CYGWIN
           throws EAGAIN, but it only does so sometimes.  3 may not be enough
           retries, but we'll try it to start with.
         - Add _windows_image_name to automatically convert "yes" to "yes.exe"
           for use with taskkill.exe /im.  Various callers to kill_all could
           be updated to remove the .exe, but that can be done in another patch.
         - Use taskkill.exe for killall on cygwin.
        * Scripts/webkitpy/common/system/executive_unittest.py:
         - Use the new *_KILLED_PROCESS_EXIT_CODE constants which are correctly
           set to 0 on windows/cygwin systems where taskkill.exe is used.
         - Test _windows_image_name
        * Scripts/webkitpy/layout_tests/port/chromium_win.py:
         - Add FIXME about including mac-snowleopard in baseline_search_path.
        * Scripts/webkitpy/layout_tests/port/webkit.py:
         - Make default_configuration actually read from the Configuration file.
        * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Added.
         - Test default_configuration
        * Scripts/webkitpy/layout_tests/port/win.py:
         - Need a basic baseline_search_path if --platform dryrun is to work.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59006 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-08 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 07:14:29 +0000 (07:14 +0000)]
2010-05-08  Eric Seidel  <eric@webkit.org>

        Unreviewed, just reverting commit.

        REGRESSION(59000): r59000 contained all sorts of changes it should not have, needs revert.
        https://bugs.webkit.org/show_bug.cgi?id=38798

        * bytecompiler/BytecodeGenerator.cpp:
        (JSC::BytecodeGenerator::emitResolve):
        * jit/JITOpcodes.cpp:
        (JSC::JIT::emit_op_resolve_global):
2010-05-08  Eric Seidel  <eric@webkit.org>

        Unreviewed, just reverting commit.

        REGRESSION(59000): r59000 contained all sorts of changes it should not have, needs revert.
        https://bugs.webkit.org/show_bug.cgi?id=38798

        * fast/js/global-resolve-through-eval-expected.txt: Added.
        * fast/js/global-resolve-through-eval.html: Added.
        * fast/js/script-tests/global-resolve-through-eval.js: Added.
        (accessGlobal):
        (accessLocal):
        * platform/win/Skipped:
2010-05-08  Eric Seidel  <eric@webkit.org>

        Unreviewed, just reverting commit.

        REGRESSION(59000): r59000 contained all sorts of changes it should not have, needs revert.
        https://bugs.webkit.org/show_bug.cgi?id=38798

        Test: fast/js/global-resolve-through-eval.html

        * bindings/js/JSCallbackData.h:
        (WebCore::JSCallbackData::JSCallbackData):
        (WebCore::JSCallbackData::~JSCallbackData):
        (WebCore::DeleteCallbackDataTask::create):
        (WebCore::DeleteCallbackDataTask::performTask):
        (WebCore::DeleteCallbackDataTask::isCleanupTask):
        (WebCore::DeleteCallbackDataTask::DeleteCallbackDataTask):
        * bindings/js/JSCustomVoidCallback.cpp:
        (WebCore::JSCustomVoidCallback::JSCustomVoidCallback):
        (WebCore::JSCustomVoidCallback::~JSCustomVoidCallback):
        * bindings/js/JSCustomVoidCallback.h:
        * bindings/scripts/CodeGeneratorJS.pm:
        * bindings/scripts/test/JS/JSTestInterface.cpp:
        (WebCore::):
        * bindings/scripts/test/JS/JSTestObj.cpp:
        (WebCore::):
        * workers/WorkerThread.cpp:
        (WebCore::WorkerThreadShutdownFinishTask::performTask):
        (WebCore::WorkerThreadShutdownStartTask::performTask):
2010-05-08  Eric Seidel  <eric@webkit.org>

        Unreviewed, just reverting commit.

        REGRESSION(59000): r59000 contained all sorts of changes it should not have, needs revert.
        https://bugs.webkit.org/show_bug.cgi?id=38798

        * Scripts/webkitpy/common/system/executive.py:
        * Scripts/webkitpy/common/system/executive_unittest.py:
        * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
        * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
        * Scripts/webkitpy/layout_tests/port/chromium_win.py:
        * Scripts/webkitpy/layout_tests/port/mac.py:
        * Scripts/webkitpy/layout_tests/port/webkit.py:
        * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Removed.
        * Scripts/webkitpy/layout_tests/port/win.py:
        * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
        * Scripts/webkitpy/tool/commands/download.py:
        * Scripts/webkitpy/tool/commands/download_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59005 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[Qt] Unreviewed, QtWebKit versioning for the trunk.
hausmann@webkit.org [Sat, 8 May 2010 07:08:09 +0000 (07:08 +0000)]
[Qt] Unreviewed, QtWebKit versioning for the trunk.

The next version is going to be 2.1, based off the trunk.

* Api/qwebkitglobal.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59004 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFix the windows build.
jer.noble@apple.com [Sat, 8 May 2010 06:56:38 +0000 (06:56 +0000)]
Fix the windows build.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59003 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFixed the inverse function of f(x) = ((x + 0.055) / 1.055) ^ 2.4
zherczeg@webkit.org [Sat, 8 May 2010 06:27:44 +0000 (06:27 +0000)]
Fixed the inverse function of f(x) = ((x + 0.055) / 1.055) ^ 2.4
https://bugs.webkit.org/show_bug.cgi?id=38735

Reviewed by Dirk Schulze

It is ((x ^ (1 / 2.4)) * 1.055) - 0.055

Chromium pixel tests cover the issue.

* platform/graphics/ImageBuffer.cpp:
(WebCore::ImageBuffer::transformColorSpace):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59002 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoJavaScriptCore: Safari pegs CPU and drops tons of frames using HTML5 Vimeo player
jer.noble@apple.com [Sat, 8 May 2010 06:24:12 +0000 (06:24 +0000)]
JavaScriptCore: Safari pegs CPU and drops tons of frames using HTML5 Vimeo player
<https://bugs.webkit.org/show_bug.cgi?id=34005>
<rdar://problem/7569713>

Reviewed by Maciej Stachowiak

Added cancelCallOnMainThread.  callOnMainThread should always now be paired
with cancelCallOnMainThread in situations where the refcon passed to callOnMainThread
may be dealloced before the main thread function can be dispatched.

* wtf/MainThread.cpp:
(WTF::FunctionWithContext::operator == ): Supports the FunctionWithContextFinder predicate functor.
(WTF::FunctionWithContextFinder::FunctionWithContextFinder):  Predicate functor for use with Dequeue::findIf
(WTF::FunctionWithContextFinder::operator()):
(WTF::cancelCallOnMainThread):
* wtf/MainThread.h:

WebCore: Safari pegs CPU and drops tons of frames using HTML5 Vimeo player
<https://bugs.webkit.org/show_bug.cgi?id=34005>
<rdar://problem/7569713>

Reviewed by Eric Carlson

The original functionality of QTMovieWin has been split between QTMovieGWorld,
and a new class: QTMovie.  QTMovie contains all the "controller" parts (changing
the rate, seeking, etc) while QTMovieGWorld retains all the drawing code.  QTMovieGWorld
now takes a QTMovie, and as such QTMovie is now retainable.  QTMovieGWorld registers
itself as a client of QTMovie, so that it can receive load-state notifications,
and thus QTMovie must now support multiple clients.  Movie tasking timer support
has been moved into its own class (QTMovieTask) and will be addressed in a future
patch.  Most of the functions listed below only changed so much as their class name changed.

* platform/graphics/win/QTMovie.cpp: Copied from WebCore/platform/graphics/win/QTMovieWin.cpp.
(QTMoviePrivate::QTMoviePrivate):
(QTMoviePrivate::~QTMoviePrivate):
(QTMoviePrivate::startTask):
(QTMoviePrivate::endTask):
(QTMoviePrivate::task):
(QTMoviePrivate::createMovieController):
(QTMovie::QTMovie):
(QTMovie::~QTMovie):
(QTMovie::addClient):
(QTMovie::removeClient):
(QTMovie::play):
(QTMovie::pause):
(QTMovie::rate):
(QTMovie::setRate):
(QTMovie::duration):
(QTMovie::currentTime):
(QTMovie::setCurrentTime):
(QTMovie::setVolume):
(QTMovie::setPreservesPitch):
(QTMovie::dataSize):
(QTMovie::maxTimeLoaded):
(QTMovie::loadState):
(QTMovie::getNaturalSize):
(QTMovie::load):
(QTMovie::disableUnsupportedTracks):
(QTMovie::isDisabled):
(QTMovie::setDisabled):
(QTMovie::hasVideo):
(QTMovie::hasAudio):
(QTMovie::hasClosedCaptions):
(QTMovie::setClosedCaptionsVisible):
(QTMovie::countSupportedTypes):
(QTMovie::getSupportedType):
(QTMovie::initializeQuickTime):
(QTMovie::getMovieHandle):
* platform/graphics/win/QTMovie.h: Copied from WebCore/platform/graphics/win/QTMovieWin.h.
* platform/graphics/win/QTMovieGWorld.cpp: Copied from WebCore/platform/graphics/win/QTMovieWin.cpp.
(QTMovieGWorldPrivate::QTMovieGWorldPrivate):
(QTMovieGWorldPrivate::~QTMovieGWorldPrivate):
(QTMovieGWorldPrivate::cacheMovieScale):
(movieDrawingCompleteProc):
(QTMovieGWorldPrivate::registerDrawingCallback):
(QTMovieGWorldPrivate::unregisterDrawingCallback):
(QTMovieGWorldPrivate::drawingComplete):
(QTMovieGWorldPrivate::updateGWorld):
(QTMovieGWorldPrivate::createGWorld):
(QTMovieGWorldPrivate::clearGWorld):
(QTMovieGWorldPrivate::setSize):
(QTMovieGWorldPrivate::updateMovieSize):
(QTMovieGWorldPrivate::deleteGWorld):
(QTMovieGWorldPrivate::movieEnded):
(QTMovieGWorldPrivate::movieLoadStateChanged):
(QTMovieGWorldPrivate::movieTimeChanged):
(QTMovieGWorld::QTMovieGWorld):
(QTMovieGWorld::~QTMovieGWorld):
(QTMovieGWorld::setSize):
(QTMovieGWorld::setVisible):
(QTMovieGWorld::getCurrentFrameInfo):
(QTMovieGWorld::paint):
(QTMovieGWorld::setDisabled):
(QTMovieGWorld::isDisabled):
(QTMovieGWorld::fullscreenWndProc):
(QTMovieGWorld::enterFullscreen):
(QTMovieGWorld::exitFullscreen):
(QTMovieGWorld::setMovie):
(QTMovieGWorld::movie):
* platform/graphics/win/QTMovieGWorld.h: Copied from WebCore/platform/graphics/win/QTMovieWin.h.
* platform/graphics/win/QTMovieTask.cpp: Added.
(QTMovieTask::QTMovieTask):
(QTMovieTask::~QTMovieTask):
(QTMovieTask::sharedTask):
(QTMovieTask::updateTaskTimer):
(QTMovieTask::fireTaskClients):
(QTMovieTask::addTaskClient):
(QTMovieTask::removeTaskClient):
(QTMovieTask::setTaskTimerFuncs):
* platform/graphics/win/QTMovieTask.h: Added.

WebKit/win: Safari pegs CPU and drops tons of frames using HTML5 Vimeo player
https://bugs.webkit.org/show_bug.cgi?id=34005

Reviewed by Adele Peterson.

QTMovieWin is now QTMovieGWorld.
* FullscreenVideoController.cpp:
(FullscreenVideoController::movie):
* FullscreenVideoController.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59001 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-06 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 06:10:59 +0000 (06:10 +0000)]
2010-05-06  Eric Seidel  <eric@webkit.org>

        Reviewed by Daniel Bates.

        test-webkitpy fails under cygwin
        https://bugs.webkit.org/show_bug.cgi?id=38718

        * Scripts/webkitpy/common/system/executive.py:
         - Add _KILL_PROCESS_KILLED_PROCESS_EXIT_CODE and
           _KILL_ALL_KILLED_PROCESS_EXIT_CODE to store the expected
           exit codes of processes killed by kill_process and kill_all.
           These two constants are only used by the unit tests but are
           stored in executive.py so they can be right next to the platform ifs.
         - Remove unnecessary str() conversion, run_command does that for us.
         - Make os.kill retry on cygwin on EAGAIN.  It's unclear why CYGWIN
           throws EAGAIN, but it only does so sometimes.  3 may not be enough
           retries, but we'll try it to start with.
         - Add _windows_image_name to automatically convert "yes" to "yes.exe"
           for use with taskkill.exe /im.  Various callers to kill_all could
           be updated to remove the .exe, but that can be done in another patch.
         - Use taskkill.exe for killall on cygwin.
        * Scripts/webkitpy/common/system/executive_unittest.py:
         - Use the new *_KILLED_PROCESS_EXIT_CODE constants which are correctly
           set to 0 on windows/cygwin systems where taskkill.exe is used.
         - Test _windows_image_name
        * Scripts/webkitpy/layout_tests/port/chromium_win.py:
         - Add FIXME about including mac-snowleopard in baseline_search_path.
        * Scripts/webkitpy/layout_tests/port/webkit.py:
         - Make default_configuration actually read from the Configuration file.
        * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Added.
         - Test default_configuration
        * Scripts/webkitpy/layout_tests/port/win.py:
         - Need a basic baseline_search_path if --platform dryrun is to work.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@59000 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 06:03:22 +0000 (06:03 +0000)]
2010-05-07  Eric Seidel  <eric@webkit.org>

        Reviewed by Nikolas Zimmermann.

        DryrunTest fails on every platform other than mac
        https://bugs.webkit.org/show_bug.cgi?id=38796

        The test uses the port detection logic to find a suitable
        port to use results from.  However that detection logic assumes
        chromium on linux, which requires a chromium checkout which the
        bots don't have.  The test is broken and we'll need to fix it.
        For now I'm just going to disable the test on all platforms besides mac.

        * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58999 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Eric Uhrhane <ericu@chromium.org>
eric@webkit.org [Sat, 8 May 2010 06:01:07 +0000 (06:01 +0000)]
2010-05-07  Eric Uhrhane  <ericu@chromium.org>

        Reviewed by Dmitry Titov.

        JSCallbackData is deleted on the Main thread, not the Context thread.
        https://bugs.webkit.org/show_bug.cgi?id=38623

        No new tests; you'll only see a problem after the worker-access-to-DB
        code goes in.  Without this fix, the GTK bots assert, but it's
        timing-sensitive and not truly platform-specific.

        * bindings/js/JSCallbackData.h:
        (WebCore::JSCallbackData::JSCallbackData):
        (WebCore::JSCallbackData::~JSCallbackData):
        Store the allocating thread, then assert that we're on it at deletion.

        (WebCore::DeleteCallbackDataTask):
        This is the task that can get posted to the context thread for cleanup.

        * bindings/js/JSCustomVoidCallback.cpp:
        * bindings/js/JSCustomVoidCallback.h:
        This callback isn't autogenerated, so we have to do the cleanup manually.

        * bindings/scripts/CodeGeneratorJS.pm:
        Add the proper cleanup to all autogenerated callbacks.

        * workers/WorkerThread.cpp:
        (WebCore::WorkerThreadShutdownFinishTask::performTask):
        (WebCore::WorkerThreadShutdownStartTask::performTask):
        WorkerContext::clearScript can't be called until the posted deletion tasks have completed; move the call from shutdown start to shutdown finish.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58998 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Nikolas Zimmermann <nzimmermann@rim.com>
zimmermann@webkit.org [Sat, 8 May 2010 05:51:54 +0000 (05:51 +0000)]
2010-05-07  Nikolas Zimmermann  <nzimmermann@rim.com>

        Not reviewed. Attempt to turn Windows bots green again.

        Skip two failing tests:

        https://bugs.webkit.org/show_bug.cgi?id=38794
        fast/css/display-none-inline-style-change-crash.html (fails on Windows Release bot)

        https://bugs.webkit.org/show_bug.cgi?id=38795
        inspector/timeline-paint.html

        * platform/win/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58997 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 05:35:05 +0000 (05:35 +0000)]
2010-05-07  Eric Seidel  <eric@webkit.org>

        Reviewed by Daniel Bates.

        Clean up baseline_search_path to use map to reduce copy/paste code
        https://bugs.webkit.org/show_bug.cgi?id=38792

        Reading which portnames a port falls back to is easier if
        we convert port names to paths with map instead of using copy/paste code.

        * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
        * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
        * Scripts/webkitpy/layout_tests/port/chromium_win.py:
        * Scripts/webkitpy/layout_tests/port/mac.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58996 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 04:17:48 +0000 (04:17 +0000)]
2010-05-07  Eric Seidel  <eric@webkit.org>

        Unreviewed, rolling out r58990.
        http://trac.webkit.org/changeset/58990

        Broke 4 tests on Snow Leopard and Qt.

        * jit/JITStubs.cpp:
        (JSC::DEFINE_STUB_FUNCTION):
        * runtime/JSCell.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58995 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 04:00:38 +0000 (04:00 +0000)]
2010-05-07  Eric Seidel  <eric@webkit.org>

        Reviewed by Daniel Bates.

        rollout commands fail when commit is missing bug number
        https://bugs.webkit.org/show_bug.cgi?id=38791

        * Scripts/webkitpy/tool/commands/download.py:
         - _commit_info failed to actually return the CommitInfo in the no-bug-id case.
        * Scripts/webkitpy/tool/commands/download_unittest.py:
         - Test that the fix worked.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58994 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoOptimize access to the global object from a function that uses eval
oliver@apple.com [Sat, 8 May 2010 03:19:14 +0000 (03:19 +0000)]
Optimize access to the global object from a function that uses eval
https://bugs.webkit.org/show_bug.cgi?id=38644

Reviewed by Gavin Barraclough.

JavaScriptCore:

Fix bug where cross scope access to a global var (vs. property) would
be allowed without checking for intervening dynamic scopes.

* bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::emitResolve):

LayoutTests:

Add test case to ensure we don't incorrectly allow dynamic scopes
to be skipped when doing direct var access.

* fast/js/global-resolve-through-eval-expected.txt: Added.
* fast/js/global-resolve-through-eval.html: Added.
* fast/js/script-tests/global-resolve-through-eval.js: Added.
(accessGlobal):
(accessLocal):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58993 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 8 May 2010 03:16:13 +0000 (03:16 +0000)]
2010-05-07  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Fix JSC binding generation to match webkit style
        https://bugs.webkit.org/show_bug.cgi?id=38173

        Add spaces around "|" to match webkit style.

        * bindings/scripts/CodeGeneratorJS.pm:
        * bindings/scripts/test/JS/JSTestObj.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58992 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoMacro expansion I stab at thee!
oliver@apple.com [Sat, 8 May 2010 02:49:55 +0000 (02:49 +0000)]
Macro expansion I stab at thee!

32bit test fix

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoOptimized o[s] where o is a cell and s is a string, removing some old
ggaren@apple.com [Sat, 8 May 2010 02:18:45 +0000 (02:18 +0000)]
Optimized o[s] where o is a cell and s is a string, removing some old
code that wasn't really tuned for the JIT.

Reviewed by Darin Adler.

SunSpider says 0.8% faster.

* jit/JITStubs.cpp:
(JSC::DEFINE_STUB_FUNCTION):
* runtime/JSCell.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58990 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoAdding the IDL files for the sync DB API and the stubs for the JS and V8 bindings.
dumi@chromium.org [Sat, 8 May 2010 01:27:42 +0000 (01:27 +0000)]
Adding the IDL files for the sync DB API and the stubs for the JS and V8 bindings.
https://bugs.webkit.org/show_bug.cgi?id=34994

Reviewed by Brady Eidson.

* DerivedSources.cpp:
* DerivedSources.make:
* GNUmakefile.am:
* WebCore.gypi:
* WebCore.pri:
* WebCore.pro:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* bindings/js/JSBindingsAllInOne.cpp:
* bindings/js/JSDatabaseSyncCustom.cpp: Added.
(WebCore::JSDatabaseSync::changeVersion):
(WebCore::createTransaction):
(WebCore::JSDatabaseSync::transaction):
(WebCore::JSDatabaseSync::readTransaction):
* bindings/js/JSSQLTransactionSyncCustom.cpp: Added.
(WebCore::JSSQLTransactionSync::executeSql):
* bindings/js/JSWorkerContextCustom.cpp:
(WebCore::JSWorkerContext::openDatabaseSync):
* bindings/v8/custom/V8BindingMacros.h: Added.
* bindings/v8/custom/V8DatabaseSyncCustom.cpp: Added.
(WebCore::V8DatabaseSync::changeVersionCallback):
(WebCore::createTransaction):
(WebCore::V8DatabaseSync::transactionCallback):
(WebCore::V8DatabaseSync::readTransactionCallback):
* bindings/v8/custom/V8SQLTransactionSyncCustom.cpp: Added.
(WebCore::V8SQLTransactionSync::executeSqlCallback):
* bindings/v8/custom/V8WorkerContextCustom.cpp:
(WebCore::V8WorkerContext::openDatabaseSyncCallback):
* storage/Database.idl:
* storage/DatabaseCallback.h:
* storage/DatabaseCallback.idl:
* storage/DatabaseSync.cpp:
(WebCore::DatabaseSync::changeVersion):
(WebCore::DatabaseSync::transaction):
* storage/DatabaseSync.h:
* storage/DatabaseSync.idl: Added.
* storage/SQLError.idl:
* storage/SQLResultSet.idl:
* storage/SQLResultSetRowList.idl:
* storage/SQLTransaction.idl:
* storage/SQLTransactionSync.idl: Added.
* storage/SQLTransactionSyncCallback.h:
* storage/SQLTransactionSyncCallback.idl: Added.
* workers/WorkerContext.cpp:
(WebCore::WorkerContext::openDatabaseSync):
* workers/WorkerContext.h:
* workers/WorkerContext.idl:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58989 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWindows build fix
oliver@apple.com [Sat, 8 May 2010 01:05:14 +0000 (01:05 +0000)]
Windows build fix

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWindows build fix.
oliver@apple.com [Sat, 8 May 2010 00:51:14 +0000 (00:51 +0000)]
Windows build fix.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Oliver Hunt <oliver@apple.com>
oliver@apple.com [Sat, 8 May 2010 00:05:00 +0000 (00:05 +0000)]
2010-05-07  Oliver Hunt  <oliver@apple.com>

        Reviewed by Geoffrey Garen.

        Optimize access to the global object from a function that uses eval
        https://bugs.webkit.org/show_bug.cgi?id=38644

        Add op_resolve_global_dynamic, a variant of op_resolve_global that
        checks each node in the scope chain for dynamically inserted properties
        and falls back to the normal resolve logic in that case.

        * JavaScriptCore.exp:
        * bytecode/CodeBlock.cpp:
        (JSC::isGlobalResolve):
        (JSC::CodeBlock::printStructures):
        (JSC::CodeBlock::dump):
        (JSC::CodeBlock::derefStructures):
        * bytecode/Opcode.h:
        * bytecompiler/BytecodeGenerator.cpp:
        (JSC::BytecodeGenerator::findScopedProperty):
           Now take an additional reference parameter to used to indicate that
           there were nodes that may gain dynamic properties
        (JSC::BytecodeGenerator::emitResolve):
        (JSC::BytecodeGenerator::emitResolveBase):
        (JSC::BytecodeGenerator::emitResolveWithBase):
           deal with additional argument to findScopedProperty
        * bytecompiler/BytecodeGenerator.h:
        * bytecompiler/NodesCodegen.cpp:
        (JSC::FunctionCallResolveNode::emitBytecode):
        (JSC::PostfixResolveNode::emitBytecode):
        (JSC::PrefixResolveNode::emitBytecode):
        (JSC::ReadModifyResolveNode::emitBytecode):
        (JSC::AssignResolveNode::emitBytecode):
           These functions use findScopedProperty directly in order to
           optimise lookup.  They cannot trivially handle any degree of
           dynamism in the lookup so we just give up in such case.
        * interpreter/Interpreter.cpp:
        (JSC::Interpreter::resolveGlobalDynamic):
        (JSC::Interpreter::execute):
        (JSC::Interpreter::privateExecute):
        * interpreter/Interpreter.h:
        * jit/JIT.cpp:
        (JSC::JIT::privateCompileMainPass):
        (JSC::JIT::privateCompileSlowCases):
        * jit/JIT.h:
        * jit/JITOpcodes.cpp:
        (JSC::JIT::emit_op_resolve_global):
        (JSC::JIT::emit_op_resolve_global_dynamic):
        (JSC::JIT::emitSlow_op_resolve_global):
        (JSC::JIT::emitSlow_op_resolve_global_dynamic):
           Happily resolve_global_dynamic can share the slow case!
        * jit/JITStubs.h:
        (JSC::):
        * runtime/JSActivation.cpp:
        (JSC::JSActivation::isDynamicScope):
        * runtime/JSActivation.h:
        * runtime/JSGlobalObject.cpp:
        (JSC::JSGlobalObject::isDynamicScope):
        * runtime/JSGlobalObject.h:
        * runtime/JSStaticScopeObject.cpp:
        (JSC::JSStaticScopeObject::isDynamicScope):
        * runtime/JSStaticScopeObject.h:
        * runtime/JSVariableObject.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Darin Fisher <darin@chromium.org>
darin@chromium.org [Fri, 7 May 2010 23:53:07 +0000 (23:53 +0000)]
2010-05-07  Darin Fisher  <darin@chromium.org>

        Fix build bustage.  Update to Chromium code that avoids a bad
        dependency from net/ on chrome/.

        * DEPS:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-06 Dirk Pranke <dpranke@chromium.org>
dpranke@chromium.org [Fri, 7 May 2010 23:51:15 +0000 (23:51 +0000)]
2010-05-06  Dirk Pranke  <dpranke@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Rewrite test to fail faster if the inner frame doesn't load at all
        for some reason (previous version would raise a TypeError before
        the timeout was set).

        https://bugs.webkit.org/show_bug.cgi?id=38705

        * http/tests/security/sandbox-inherit-to-initial-document-2.html:
        * platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFix for <rdar://problem/7956972> REGRESSION: Can't focus and type
bdakin@apple.com [Fri, 7 May 2010 23:35:27 +0000 (23:35 +0000)]
Fix for <rdar://problem/7956972> REGRESSION: Can't focus and type
in GMail due to bad repainting
-and corresponding-
https://bugs.webkit.org/show_bug.cgi?id=38782

Reviewed by Simon Fraser.

This patch reverts back to pre-r58797 behavior when
shouldPropagateCompositingToIFrameParent() is false.
* rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::paintingGoesToWindow):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago<rdar://problem/7947356>
andersca@apple.com [Fri, 7 May 2010 23:23:36 +0000 (23:23 +0000)]
<rdar://problem/7947356>
QT Plug-in in hardware-accelerated WebKit is missing the controller

Reviewed by Sam Weinig and Simon Fraser.

Replace the useSoftwareRenderer boolean with an enum that lets the plug-in host opt into using a layer
backed NSView instead of inserting the layer into the WebCore layer hierarchy.

* Plugins/Hosted/NetscapePluginHostManager.mm:
(WebKit::NetscapePluginHostManager::instantiatePlugin):
* Plugins/Hosted/NetscapePluginHostProxy.mm:
(WKPCInstantiatePluginReply):
* Plugins/Hosted/NetscapePluginInstanceProxy.h:
(WebKit::NetscapePluginInstanceProxy::rendererType):
(WebKit::NetscapePluginInstanceProxy::setRendererType):
(WebKit::NetscapePluginInstanceProxy::InstantiatePluginReply::InstantiatePluginReply):
* Plugins/Hosted/NetscapePluginInstanceProxy.mm:
(WebKit::NetscapePluginInstanceProxy::NetscapePluginInstanceProxy):
Replace the useSoftwareRenderer boolean everywhere with the RendererType enum.

* Plugins/Hosted/WebHostedNetscapePluginView.mm:
(-[WebHostedNetscapePluginView createPlugin]):
Only go into full compositing mode when the renderer type returned is UseAcceleratedCompositing.
Otherwise, use a layer backed NSView.

* Plugins/Hosted/WebKitPluginHostTypes.h:
Add a RendererType enum.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Darin Fisher <darin@chromium.org>
darin@chromium.org [Fri, 7 May 2010 23:15:18 +0000 (23:15 +0000)]
2010-05-07  Darin Fisher  <darin@chromium.org>

        Fix build bustage.  Update to Chromium code that no longer uses
        deprecated methods.

        * DEPS:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Darin Fisher <darin@chromium.org>
darin@chromium.org [Fri, 7 May 2010 22:50:54 +0000 (22:50 +0000)]
2010-05-07  Darin Fisher  <darin@chromium.org>

        Fix build bustage: toElement<T> should be to<T>.

        * DumpRenderTree/chromium/LayoutTestController.cpp:
        (LayoutTestController::elementDoesAutoCompleteForElementWithId):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFixed Sputnik failure seen on buildbot.
ggaren@apple.com [Fri, 7 May 2010 22:18:41 +0000 (22:18 +0000)]
Fixed Sputnik failure seen on buildbot.

Reviewed by Gavin Barraclough.

* runtime/NumberPrototype.cpp:
(JSC::numberProtoFuncToString): Cast to unsigned before checking magnitude,
to efficiently exclude negative numbers, which require two characters
instead of one.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Sheriff Bot <webkit.review.bot@gmail.com>
johnnyg@google.com [Fri, 7 May 2010 21:44:10 +0000 (21:44 +0000)]
2010-05-07  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r58963.
        http://trac.webkit.org/changeset/58963
        https://bugs.webkit.org/show_bug.cgi?id=38773

        Broke Chromium layout tests. (Requested by dcheng on #webkit).

        * platform/chromium/DragDataChromium.cpp:
        (WebCore::DragData::asURL):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 John Gregg <johnnyg@google.com>
johnnyg@google.com [Fri, 7 May 2010 21:18:46 +0000 (21:18 +0000)]
2010-05-07  John Gregg  <johnnyg@google.com>

        Unreviewed [re-]baseline of chromium tests.

        * platform/chromium-linux/fast/frames/take-focus-from-iframe-expected.checksum: Added.
        * platform/chromium-linux/fast/frames/take-focus-from-iframe-expected.png: Added.
        * platform/chromium-mac/fast/frames/take-focus-from-iframe-expected.checksum: Added.
        * platform/chromium-win/fast/frames/take-focus-from-iframe-expected.checksum: Added.
        * platform/chromium-win/fast/frames/take-focus-from-iframe-expected.png: Added.
        * platform/chromium-win/fast/frames/take-focus-from-iframe-expected.txt: Added.
        * platform/chromium-win/svg/hixie/error/017-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Fri, 7 May 2010 20:28:11 +0000 (20:28 +0000)]
2010-05-07  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Gustavo Noronha Silva.

        Adjust the position of the comment on the WOFF test and
        link to the bug report. This should fix a spurious error
        while running layout tests.

        * platform/gtk/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoSlightly more effective way to guarantee a compile-time constant, at
ggaren@apple.com [Fri, 7 May 2010 20:13:54 +0000 (20:13 +0000)]
Slightly more effective way to guarantee a compile-time constant, at
least on gcc.

Reviewed by Alexey Proskuryakov.

* runtime/NumberPrototype.cpp:
(JSC::numberProtoFuncToString): Use pointer notation instead of array
notation, so the compiler treats the string and its contents as constant.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Fri, 7 May 2010 20:13:31 +0000 (20:13 +0000)]
2010-05-07  Dirk Schulze  <krit@webkit.org>

        Unreviewed fix of Windows test.

        Windows need an platform dependent DRT result for clip-path-referencing-use2.svg,
        Win doesn't have console output during DRT.

        * platform/win/svg/custom/clip-path-referencing-use2-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Fri, 7 May 2010 20:03:45 +0000 (20:03 +0000)]
2010-05-07  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Gustavo Noronha Silva.

        https://bugs.webkit.org/show_bug.cgi?id=38769
        ClipboardGtk should write the URL label in the text portion of the DataObject

        No tests as this code path is currently unused in the GTK+ port.

        * platform/gtk/ClipboardGtk.cpp:
        (WebCore::ClipboardGtk::writeURL): Write the label to the text portion of the DataObject, instead of the URL.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Darin Fisher <darin@chromium.org>
darin@chromium.org [Fri, 7 May 2010 19:57:50 +0000 (19:57 +0000)]
2010-05-07  Darin Fisher  <darin@chromium.org>

        Reviewed by Dimitri Glazkov.

        [chromium] clear out more deprecated methods that are no longer used
        https://bugs.webkit.org/show_bug.cgi?id=38702

        * public/WebFrame.h:
        * public/WebNode.h:
        * src/WebFrameImpl.cpp:
        * src/WebFrameImpl.h:
        * src/WebNode.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Fri, 7 May 2010 19:56:59 +0000 (19:56 +0000)]
2010-05-07  Adam Barth  <abarth@webkit.org>

        Reviewed by Alexey Proskuryakov.

        [WTFURL] Add core URL parser
        https://bugs.webkit.org/show_bug.cgi?id=38572

        This patch adds the core of the URL parser.  The URL parser uses a
        templated notion of a code unit to support different string types.
        Later we'll add some non-templated APIs with concrete types that
        clients can use more conveniently.

        The URLParser has a couple calls to ASSERT(), which I've commented out
        for now.  I'm not 100% sure how to handle the dependency issues there
        yet, so I've punted on the issue for now.

        Also, there are a number of methods that are declared public in this
        patch that are only used by the canonicalizer.  My plan is to make
        those private or protected and make the canonicalizer a friend of the
        parser.  The details will be in a later patch.

        * wtf/url/src/URLComponent.h:
        (WTF::URLComponent::fromRange):
        (WTF::URLComponent::isNonEmpty):
        (WTF::URLComponent::isEmptyOrInvalid):
        * wtf/url/src/URLParser.h: Added.
        (WTF::URLParser::):
        (WTF::URLParser::isPossibleAuthorityTerminator):
        (WTF::URLParser::parseAuthority):
        (WTF::URLParser::extractScheme):
        (WTF::URLParser::parseAfterScheme):
        (WTF::URLParser::parseStandardURL):
        (WTF::URLParser::parsePath):
        (WTF::URLParser::parsePathURL):
        (WTF::URLParser::parseMailtoURL):
        (WTF::URLParser::parsePort):
        (WTF::URLParser::extractFileName):
        (WTF::URLParser::extractQueryKeyValue):
        (WTF::URLParser::isURLSlash):
        (WTF::URLParser::shouldTrimFromURL):
        (WTF::URLParser::trimURL):
        (WTF::URLParser::consecutiveSlashes):
        (WTF::URLParser::isPortDigit):
        (WTF::URLParser::nextAuthorityTerminator):
        (WTF::URLParser::parseUserInfo):
        (WTF::URLParser::parseServerInfo):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58966 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Nikolas Zimmermann <nzimmermann@rim.com>
zimmermann@webkit.org [Fri, 7 May 2010 19:35:47 +0000 (19:35 +0000)]
2010-05-07  Nikolas Zimmermann  <nzimmermann@rim.com>

        Reviewed by Dirk Schulze.

        RenderSVGResourceClipper assigns a temporary mutated RenderStyle but does not correctly preserve the old style
        https://bugs.webkit.org/show_bug.cgi?id=38767

        Fix small logic error leading to a problem in RenderSVGResourceClipper. It assigns a new temporary RenderStyle
        to the target object, but fails to reset it correctly to the old style. Fixes all svg/clip-path errors on the windows bots.

        * rendering/RenderSVGResourceClipper.cpp:
        (WebCore::RenderSVGResourceClipper::createClipData):
2010-05-07  Nikolas Zimmermann  <nzimmermann@rim.com>

        Reviewed by Dirk Schulze.

        RenderSVGResourceClipper assigns a temporary mutated RenderStyle but does not correctly preserve the old style
        https://bugs.webkit.org/show_bug.cgi?id=38767

        Update baseline after fixing a logic error in RenderSVGResourceClipper.

        * platform/mac/svg/clip-path/clip-path-child-clipped-expected.txt:
        * platform/mac/svg/clip-path/clip-path-childs-clipped-expected.txt:
        * platform/mac/svg/clip-path/clip-path-clipped-expected.txt:
        * platform/mac/svg/clip-path/clip-path-evenodd-expected.txt:
        * platform/mac/svg/clip-path/clip-path-evenodd-nonzero-expected.txt:
        * platform/mac/svg/clip-path/clip-path-nonzero-evenodd-expected.txt:
        * platform/mac/svg/clip-path/clip-path-text-and-shape-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58965 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoAdded a fast path for number-to-character conversion via
ggaren@apple.com [Fri, 7 May 2010 19:30:31 +0000 (19:30 +0000)]
Added a fast path for number-to-character conversion via
Number.prototype.toString base 36.

Reviewed by Oliver Hunt.

0.7% speedup on SunSpider.

* runtime/NumberPrototype.cpp:
(JSC::numberProtoFuncToString): Made radix interpretation a little more
efficient by keeping it in int space. Turned "const char" into
"static const char" just in case. Added a fast path for base 36
conversion of something that will turn into a character.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoAdd a few more files with Windows linefeeds to the .gitattributes file.
jianli@chromium.org [Fri, 7 May 2010 19:23:24 +0000 (19:23 +0000)]
Add a few more files with Windows linefeeds to the .gitattributes file.

Patch by Simon Hausmann <simon.hausmann@nokia.com> on 2010-05-07
* .gitattributes:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Fri, 7 May 2010 18:46:10 +0000 (18:46 +0000)]
2010-05-07  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein.

        Fix debug-only render tree output to show object addresses in hex.

        * rendering/RenderTreeAsText.cpp:
        (WebCore::writeRenderObject):
        (WebCore::write):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Adele Peterson.
ap@apple.com [Fri, 7 May 2010 18:40:55 +0000 (18:40 +0000)]
    Reviewed by Adele Peterson.

        https://bugs.webkit.org/show_bug.cgi?id=38546
        Node.focus() fails to move focus from subframe properly

        Test: fast/frames/take-focus-from-iframe.html

        * html/HTMLFrameElementBase.cpp: (WebCore::HTMLFrameElementBase::setFocus): Don't clear
        focus if this frame doesn't have it. This can happen if page's and HTMLFrameElement's ideas
        of focused frame get out of sync temporarily.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Nikolas Zimmermann <nzimmermann@rim.com>
zimmermann@webkit.org [Fri, 7 May 2010 18:24:50 +0000 (18:24 +0000)]
2010-05-07  Nikolas Zimmermann  <nzimmermann@rim.com>

        Reviewed by Dirk Schulze.

        svg/custom/use-instanceRoot-as-event-target.xhtml crashes randomly
        https://bugs.webkit.org/show_bug.cgi?id=37798

        When creating JSEventListeners through attributes (onclick, etc..) or add/removeEventListener
        calls on a SVGElementInstance, do NOT pass the element instance itself as JS wrapper object
        but the correspondingElement(). SVGElementInstance redirects all event listener registrations
        to the correspondingElement(), as they share an event listener list, per SVG spec. The old
        code was very dangerous, leading to random assertions, when garbage collection teared down
        the JSSVGElementInstance and event listeners fired, whose listeners were registered with
        it, but residing in the correspondingElement() event listener list.

        Removes the need for CustomToJS code for JSSVGElementInstance, which was wrong anyways. We hoped
        to keep the event listeers alive by just creating a js wrapper of the correspondingElement(), that
        could only work as long as garbage collection didn't tear it down, just luck. Also remove the
        CustomPushEventHandlerScope marker, as it is only used for JSLazyEventListeners and only works
        for JSNode derived objects, it was a no-op for SVGElementInstance, thus removed it.

        Should fix all random crashes/assertions seen with svg/custom/use-instanceRoot-as-event-target.xhtml
        See bug report for a detailed crash analysis.

        * bindings/js/JSSVGElementInstanceCustom.cpp: Remove custom toJS()/pushEventHandlerScope() handling, not necessary anymore.
        (WebCore::JSSVGElementInstance::markChildren):
        * bindings/scripts/CodeGeneratorJS.pm: For JSSVGElementInstance pass the correspondingElement() as JS wrapper object, not itself.
        * svg/SVGElementInstance.cpp:
        (WebCore::SVGElementInstance::invalidateAllInstancesOfElement): Be sure to trigger a style update here, so dirty shadow trees for <use> get rebuild. Otherwhise DOM may be out-of-sync.
        (WebCore::SVGElementInstance::eventTargetData): Add ASSERT_NOT_REACHED(), all event listener calls are forwarded to the correspondingElement().
        (WebCore::SVGElementInstance::ensureEventTargetData): Ditto.
        * svg/SVGElementInstance.idl: Remove CustomToJS, CustomPushEventHandlerScope markers.
2010-05-07  Nikolas Zimmermann  <nzimmermann@rim.com>

        Reviewed by Dirk Schulze.

        svg/custom/use-instanceRoot-as-event-target.xhtml crashes randomly
        https://bugs.webkit.org/show_bug.cgi?id=37798

        Make use-instanceRoot-as-event-target.xhtml behave more correctly. Due a copy & paste problem test #7 was not executed. Fixed.
        Make destruction more reliable, by using the standard js-test-post.js injection through <script> before calling notifyDone(), and "TEST COMPLETE" magically appears.

        * platform/mac/svg/hixie/error/017-expected.txt: Update result, now that invalidateAllInstancesOfElement(), correctly triggers <use> shadow tree rebuilds.
        * svg/custom/use-instanceRoot-as-event-target-expected.txt: Update result, was missing a "TEST COMPLETE" marker, and Test #7 was not executed before.
        * svg/custom/use-instanceRoot-as-event-target.xhtml: Fix typo, test #7 is now executed, it also had a bug which has been corrected.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Adam Barth <abarth@webkit.org>
eric@webkit.org [Fri, 7 May 2010 18:20:31 +0000 (18:20 +0000)]
2010-05-07  Adam Barth  <abarth@webkit.org>

        Reviewed by Darin Fisher.

        [WTFURL] Add a malloc-backed URLBuffer
        https://bugs.webkit.org/show_bug.cgi?id=38694

        This patch adds a URLBuffer that uses new/delete to manage the backing
        store for the buffer.

        * wtf/url/src/RawURLBuffer.h: Added.
        (WTF::RawURLBuffer::RawURLBuffer):
        (WTF::RawURLBuffer::~RawURLBuffer):
        (WTF::RawURLBuffer::resize):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58959 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Finnur Thorarinsson <finnur.webkit@gmail.com>
eric@webkit.org [Fri, 7 May 2010 18:05:52 +0000 (18:05 +0000)]
2010-05-07  Finnur Thorarinsson  <finnur.webkit@gmail.com>

        Reviewed by Dimitri Glazkov.

        [chromium]: Fixing Find in page crash on sites.google.com
        https://bugs.webkit.org/show_bug.cgi?id=38591

        setMarkerActive should not try to set a collapsed marker as
        active. This can happen when you Edit a Sites page after doing
        a Find operation and then you press FindNext.

        * src/WebFrameImpl.cpp:
        (WebKit::WebFrameImpl::setMarkerActive):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-06 Kenneth Russell <kbr@google.com>
kbr@google.com [Fri, 7 May 2010 17:52:29 +0000 (17:52 +0000)]
2010-05-06  Kenneth Russell  <kbr@google.com>

        Reviewed by Dimitri Glazkov.

        WebCore::WebGLUnsignedIntArrayInternal::getCallback ReadAV@Arbitrary (deef89ee3d0345edebeaf13cf974c47c)
        https://bugs.webkit.org/show_bug.cgi?id=38039

        Web IDL now allows indexed getters and setters to be unnamed. Per
        discussion in WebGL working group and recent update to spec,
        removed the buggy get() and single-element set() methods from the
        JavaScript bindings to the WebGL array types. Refactored set()
        implementation in JSC bindings to share more code and modified V8
        binding to look more like it. Added unit tests for indexed getter
        with out-of-range indices and verifying removal of get and
        single-element set methods. Updated existing WebGL array tests.

        Tests: fast/canvas/webgl/array-get-and-set-method-removal.html
               fast/canvas/webgl/array-get-out-of-bounds.html

        * bindings/js/JSWebGLArrayHelper.h:
        (WebCore::setWebGLArrayHelper):
        * bindings/js/JSWebGLByteArrayCustom.cpp:
        (WebCore::JSWebGLByteArray::set):
        * bindings/js/JSWebGLFloatArrayCustom.cpp:
        (WebCore::JSWebGLFloatArray::set):
        * bindings/js/JSWebGLIntArrayCustom.cpp:
        (WebCore::JSWebGLIntArray::set):
        * bindings/js/JSWebGLShortArrayCustom.cpp:
        (WebCore::JSWebGLShortArray::set):
        * bindings/js/JSWebGLUnsignedByteArrayCustom.cpp:
        (WebCore::JSWebGLUnsignedByteArray::set):
        * bindings/js/JSWebGLUnsignedIntArrayCustom.cpp:
        (WebCore::JSWebGLUnsignedIntArray::set):
        * bindings/js/JSWebGLUnsignedShortArrayCustom.cpp:
        (WebCore::JSWebGLUnsignedShortArray::set):
        * bindings/v8/custom/V8WebGLArrayCustom.h:
        (WebCore::setWebGLArrayHelper):
        * bindings/v8/custom/V8WebGLByteArrayCustom.cpp:
        (WebCore::V8WebGLByteArray::setCallback):
        * bindings/v8/custom/V8WebGLFloatArrayCustom.cpp:
        (WebCore::V8WebGLFloatArray::setCallback):
        * bindings/v8/custom/V8WebGLIntArrayCustom.cpp:
        (WebCore::V8WebGLIntArray::setCallback):
        * bindings/v8/custom/V8WebGLShortArrayCustom.cpp:
        (WebCore::V8WebGLShortArray::setCallback):
        * bindings/v8/custom/V8WebGLUnsignedByteArrayCustom.cpp:
        (WebCore::V8WebGLUnsignedByteArray::setCallback):
        * bindings/v8/custom/V8WebGLUnsignedIntArrayCustom.cpp:
        (WebCore::V8WebGLUnsignedIntArray::setCallback):
        * bindings/v8/custom/V8WebGLUnsignedShortArrayCustom.cpp:
        (WebCore::V8WebGLUnsignedShortArray::setCallback):
        * html/canvas/WebGLByteArray.h:
        * html/canvas/WebGLByteArray.idl:
        * html/canvas/WebGLFloatArray.h:
        * html/canvas/WebGLFloatArray.idl:
        * html/canvas/WebGLIntArray.h:
        * html/canvas/WebGLIntArray.idl:
        * html/canvas/WebGLShortArray.h:
        * html/canvas/WebGLShortArray.idl:
        * html/canvas/WebGLUnsignedByteArray.h:
        * html/canvas/WebGLUnsignedByteArray.idl:
        * html/canvas/WebGLUnsignedIntArray.h:
        * html/canvas/WebGLUnsignedIntArray.idl:
        * html/canvas/WebGLUnsignedShortArray.h:
        * html/canvas/WebGLUnsignedShortArray.idl:
2010-05-06  Kenneth Russell  <kbr@google.com>

        Reviewed by Dimitri Glazkov.

        WebCore::WebGLUnsignedIntArrayInternal::getCallback ReadAV@Arbitrary (deef89ee3d0345edebeaf13cf974c47c)
        https://bugs.webkit.org/show_bug.cgi?id=38039

        Web IDL now allows indexed getters and setters to be unnamed. Per
        discussion in WebGL working group and recent update to spec,
        removed the buggy get() and single-element set() methods from the
        JavaScript bindings to the WebGL array types. Refactored set()
        implementation in JSC bindings to share more code and modified V8
        binding to look more like it. Added unit tests for indexed getter
        with out-of-range indices and verifying removal of get and
        single-element set methods. Updated existing WebGL array tests.

        * fast/canvas/webgl/array-get-and-set-method-removal-expected.txt: Added.
        * fast/canvas/webgl/array-get-and-set-method-removal.html: Added.
        * fast/canvas/webgl/array-get-out-of-bounds-expected.txt: Added.
        * fast/canvas/webgl/array-get-out-of-bounds.html: Added.
        * fast/canvas/webgl/array-setters-expected.txt:
        * fast/canvas/webgl/array-setters.html:
        * fast/canvas/webgl/array-unit-tests-expected.txt:
        * fast/canvas/webgl/array-unit-tests.html:
        * fast/canvas/webgl/bug-32456-expected.txt:
        * fast/canvas/webgl/bug-32456.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Fri, 7 May 2010 17:32:24 +0000 (17:32 +0000)]
2010-05-07  Pavel Feldman  <pfeldman@chromium.org>

        Not reviewed: chromium dev tools tests fix.
        Add a more strict check prior to accessing constructor property in
        Web Inspector frontend.

        * inspector/front-end/InjectedScript.js:
        (injectedScriptConstructor):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 John Gregg <johnnyg@google.com>
johnnyg@google.com [Fri, 7 May 2010 17:30:57 +0000 (17:30 +0000)]
2010-05-07  John Gregg  <johnnyg@google.com>

        Unreviewed, rebaseline for chromium.

        * platform/chromium-mac/fast/layers/video-layer-expected.checksum:
        * platform/chromium-mac/fast/layers/video-layer-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Fri, 7 May 2010 17:03:27 +0000 (17:03 +0000)]
2010-05-07  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Joseph Pecoraro.

        Web Inspector: On Linux/Windows panel history is
        traversed while iterating over words in text prompt.

        https://bugs.webkit.org/show_bug.cgi?id=38740

        * inspector/front-end/TextPrompt.js:
        (WebInspector.TextPrompt):
        * inspector/front-end/inspector.js:
        (WebInspector.documentKeyDown):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Fri, 7 May 2010 16:59:29 +0000 (16:59 +0000)]
2010-05-07  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Timothy Hatcher.

        Web Inspector: come up with InjectedScript._className that works for both JSC and V8.

        https://bugs.webkit.org/show_bug.cgi?id=38755

        * inspector/front-end/InjectedScript.js:
        (injectedScriptConstructor):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2010-05-07 John Gregg <johnnyg@google.com>
johnnyg@google.com [Fri, 7 May 2010 16:32:16 +0000 (16:32 +0000)]
2010-05-07  John Gregg  <johnnyg@google.com>

        Unreviewed, rebaseline for chromium.

        * platform/chromium-mac/media/audio-controls-rendering-expected.checksum:
        * platform/chromium-mac/media/audio-controls-rendering-expected.png:
        * platform/chromium-mac/media/controls-after-reload-expected.checksum:
        * platform/chromium-mac/media/controls-after-reload-expected.png:
        * platform/chromium-mac/media/controls-strict-expected.checksum:
        * platform/chromium-mac/media/controls-strict-expected.png:
        * platform/chromium-mac/media/video-controls-rendering-expected.checksum:
        * platform/chromium-mac/media/video-controls-rendering-expected.png:
        * platform/chromium-mac/media/video-display-toggle-expected.checksum:
        * platform/chromium-mac/media/video-display-toggle-expected.png:
        * platform/chromium-mac/media/video-empty-source-expected.checksum:
        * platform/chromium-mac/media/video-empty-source-expected.png:
        * platform/chromium-mac/media/video-no-audio-expected.checksum:
        * platform/chromium-mac/media/video-no-audio-expected.png:
        * platform/chromium-mac/media/video-volume-slider-expected.checksum:
        * platform/chromium-mac/media/video-volume-slider-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58952 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoRefactored VCSUtils.pm's parse-related methods to leave inapplicable
cjerdonek@webkit.org [Fri, 7 May 2010 15:54:10 +0000 (15:54 +0000)]
Refactored VCSUtils.pm's parse-related methods to leave inapplicable
hash values unset instead of setting them to "undef".

Reviewed by Daniel Bates.

https://bugs.webkit.org/show_bug.cgi?id=38724

Preferring "not set" over "undef" keeps the unit tests smaller and
easier to maintain.  Otherwise, we would have to update every unit
test case each time we add support for a new key-value pair --
instead of just the relevant ones.

* Scripts/VCSUtils.pm:
  - In parseGitDiffHeader(), adjusted the handling of these key-values:
    executableBitDelta and isBinary.
  - In parseSvnDiffHeader(), adjusted the handling of these key-values:
    copiedFromPath, isBinary, and sourceRevision.
  - In parseDiffHeader(), adjusted the handling of these key-values:
    isGit and isSvn.
  - In parseDiff(), adjusted the handling of these key-values:
    isBinary, isGit, isSvn, and sourceRevision.

* Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
  - Updated the unit tests as necessary.

* Scripts/webkitperl/VCSUtils_unittest/parseDiffHeader.pl:
  - Updated the unit tests as necessary.

* Scripts/webkitperl/VCSUtils_unittest/parseGitDiffHeader.pl:
  - Updated the unit tests as necessary.

* Scripts/webkitperl/VCSUtils_unittest/parseSvnDiffHeader.pl:
  - Updated the unit tests as necessary.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@58951 268f45cc-cd09-0410-ab3c-d52691b4dbfc