Title: [123393] trunk
Revision
123393
Author
[email protected]
Date
2012-07-23 15:38:30 -0700 (Mon, 23 Jul 2012)

Log Message

REGRESSION(r123169): It made fast/js/dfg-inline-arguments-use-from-uninlined-code.html fail on 32 bit platforms
https://bugs.webkit.org/show_bug.cgi?id=92002

Source/_javascript_Core: 

Reviewed by Mark Hahnenberg.
        
In the process of changing the nature of local variable typing, I forgot to modify one of the places where
we glue the DFG's notion of variable prediction to the runtime's notion of variable tagging.

* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compile):

LayoutTests: 

Unskip test that this change fixes, as requested by Ossy.

* platform/qt/Skipped:

Modified Paths

Diff

Modified: trunk/LayoutTests/ChangeLog (123392 => 123393)


--- trunk/LayoutTests/ChangeLog	2012-07-23 22:36:43 UTC (rev 123392)
+++ trunk/LayoutTests/ChangeLog	2012-07-23 22:38:30 UTC (rev 123393)
@@ -1,3 +1,12 @@
+2012-07-23  Filip Pizlo  <[email protected]>
+
+        REGRESSION(r123169): It made fast/js/dfg-inline-arguments-use-from-uninlined-code.html fail on 32 bit platforms
+        https://bugs.webkit.org/show_bug.cgi?id=92002
+
+        Unskip test that this change fixes, as requested by Ossy.
+
+        * platform/qt/Skipped:
+
 2012-07-23  Andrew Wilson  <[email protected]>
 
         More unreviewed chromium expectation changes to mark tests as failing/slow.

Modified: trunk/LayoutTests/platform/qt/Skipped (123392 => 123393)


--- trunk/LayoutTests/platform/qt/Skipped	2012-07-23 22:36:43 UTC (rev 123392)
+++ trunk/LayoutTests/platform/qt/Skipped	2012-07-23 22:38:30 UTC (rev 123393)
@@ -2595,10 +2595,6 @@
 # https://bugs.webkit.org/show_bug.cgi?id=91968
 fast/sub-pixel/table-rows-have-stable-height.html
 
-# REGRESSION(r123169): It made fast/js/dfg-inline-arguments-use-from-uninlined-code.html fail on 32 bit platforms
-# https://bugs.webkit.org/show_bug.cgi?id=92002
-fast/js/dfg-inline-arguments-use-from-uninlined-code.html
-
 # [Qt] svg/zoom/{page,text}/zoom-hixie-mixed-009.xml fails
 # https://bugs.webkit.org/show_bug.cgi?id=92018
 svg/zoom/text/zoom-hixie-mixed-009.xml

Modified: trunk/Source/_javascript_Core/ChangeLog (123392 => 123393)


--- trunk/Source/_javascript_Core/ChangeLog	2012-07-23 22:36:43 UTC (rev 123392)
+++ trunk/Source/_javascript_Core/ChangeLog	2012-07-23 22:38:30 UTC (rev 123393)
@@ -1,3 +1,16 @@
+2012-07-23  Filip Pizlo  <[email protected]>
+
+        REGRESSION(r123169): It made fast/js/dfg-inline-arguments-use-from-uninlined-code.html fail on 32 bit platforms
+        https://bugs.webkit.org/show_bug.cgi?id=92002
+
+        Reviewed by Mark Hahnenberg.
+        
+        In the process of changing the nature of local variable typing, I forgot to modify one of the places where
+        we glue the DFG's notion of variable prediction to the runtime's notion of variable tagging.
+
+        * dfg/DFGSpeculativeJIT.cpp:
+        (JSC::DFG::SpeculativeJIT::compile):
+
 2012-07-23  Simon Fraser  <[email protected]>
 
         Part 2 of: Implement sticky positioning

Modified: trunk/Source/_javascript_Core/dfg/DFGSpeculativeJIT.cpp (123392 => 123393)


--- trunk/Source/_javascript_Core/dfg/DFGSpeculativeJIT.cpp	2012-07-23 22:36:43 UTC (rev 123392)
+++ trunk/Source/_javascript_Core/dfg/DFGSpeculativeJIT.cpp	2012-07-23 22:38:30 UTC (rev 123393)
@@ -1063,7 +1063,7 @@
                             valueSource = ValueSource(DoubleInRegisterFile);
                         else if (isInt32Speculation(argumentPosition.prediction()))
                             valueSource = ValueSource(Int32InRegisterFile);
-                        else if (isArraySpeculation(argumentPosition.prediction()))
+                        else if (isArraySpeculation(argumentPosition.prediction()) || isCellSpeculation(argumentPosition.prediction()))
                             valueSource = ValueSource(CellInRegisterFile);
                         else if (isBooleanSpeculation(argumentPosition.prediction()))
                             valueSource = ValueSource(BooleanInRegisterFile);
_______________________________________________
webkit-changes mailing list
[email protected]
http://lists.webkit.org/mailman/listinfo/webkit-changes

Reply via email to