Mercurial > hg > truffle
comparison src/cpu/sparc/vm/sparc.ad @ 1274:2883969d09e7
6910664: C2: java/util/Arrays/Sorting.java fails with DeoptimizeALot flag
Summary: Matcher::float_in_double should be true only when FPU is used for floats.
Reviewed-by: never, twisti
author | kvn |
---|---|
date | Fri, 19 Feb 2010 10:04:16 -0800 |
parents | f24201449cac |
children | 9e321dcfa5b7 |
comparison
equal
deleted
inserted
replaced
1273:877a14af58e1 | 1274:2883969d09e7 |
---|---|
1801 | 1801 |
1802 // Advertise here if the CPU requires explicit rounding operations | 1802 // Advertise here if the CPU requires explicit rounding operations |
1803 // to implement the UseStrictFP mode. | 1803 // to implement the UseStrictFP mode. |
1804 const bool Matcher::strict_fp_requires_explicit_rounding = false; | 1804 const bool Matcher::strict_fp_requires_explicit_rounding = false; |
1805 | 1805 |
1806 // Do floats take an entire double register or just half? | 1806 // Are floats conerted to double when stored to stack during deoptimization? |
1807 const bool Matcher::float_in_double = false; | 1807 // Sparc does not handle callee-save floats. |
1808 bool Matcher::float_in_double() { return false; } | |
1808 | 1809 |
1809 // Do ints take an entire long register or just half? | 1810 // Do ints take an entire long register or just half? |
1810 // Note that we if-def off of _LP64. | 1811 // Note that we if-def off of _LP64. |
1811 // The relevant question is how the int is callee-saved. In _LP64 | 1812 // The relevant question is how the int is callee-saved. In _LP64 |
1812 // the whole long is written but de-opt'ing will have to extract | 1813 // the whole long is written but de-opt'ing will have to extract |