Title: [194373] trunk/Source/_javascript_Core
- Revision
- 194373
- Author
- [email protected]
- Date
- 2015-12-22 14:19:31 -0800 (Tue, 22 Dec 2015)
Log Message
Work around issue in bug #152510
https://bugs.webkit.org/show_bug.cgi?id=152511
Reviewed by Filip Pizlo.
* runtime/Options.cpp:
(JSC::recomputeDependentOptions):
Modified Paths
Diff
Modified: trunk/Source/_javascript_Core/ChangeLog (194372 => 194373)
--- trunk/Source/_javascript_Core/ChangeLog 2015-12-22 21:59:59 UTC (rev 194372)
+++ trunk/Source/_javascript_Core/ChangeLog 2015-12-22 22:19:31 UTC (rev 194373)
@@ -1,3 +1,13 @@
+2015-12-22 Saam barati <[email protected]>
+
+ Work around issue in bug #152510
+ https://bugs.webkit.org/show_bug.cgi?id=152511
+
+ Reviewed by Filip Pizlo.
+
+ * runtime/Options.cpp:
+ (JSC::recomputeDependentOptions):
+
2015-12-22 Filip Pizlo <[email protected]>
FTL B3 does not logicalNot correctly
Modified: trunk/Source/_javascript_Core/runtime/Options.cpp (194372 => 194373)
--- trunk/Source/_javascript_Core/runtime/Options.cpp 2015-12-22 21:59:59 UTC (rev 194372)
+++ trunk/Source/_javascript_Core/runtime/Options.cpp 2015-12-22 22:19:31 UTC (rev 194373)
@@ -26,6 +26,7 @@
#include "config.h"
#include "Options.h"
+#include "CCallHelpers.h"
#include <algorithm>
#include <limits>
#include <math.h>
@@ -316,6 +317,16 @@
Options::useOSREntryToFTL() = false;
}
+ if (isARM64()) {
+ // FIXME: https://bugs.webkit.org/show_bug.cgi?id=152510
+ // We're running into a bug where some ARM64 tests are failing in the FTL
+ // with what appears to be an llvm bug where llvm is miscalculating the
+ // live-out variables of a patchpoint. This causes us to not keep a
+ // volatile register alive across a C call in a patchpoint, even though
+ // that register is used immediately after the patchpoint.
+ Options::assumeAllRegsInFTLICAreLive() = true;
+ }
+
// Compute the maximum value of the reoptimization retry counter. This is simply
// the largest value at which we don't overflow the execute counter, when using it
// to left-shift the execution counter by this amount. Currently the value ends
_______________________________________________
webkit-changes mailing list
[email protected]
https://lists.webkit.org/mailman/listinfo/webkit-changes
