- Nov 25, 2008
-
-
Chris Lattner authored
llvm-svn: 60009
-
Oscar Fuentes authored
llvm-svn: 60008
-
Douglas Gregor authored
llvm-svn: 60007
-
Scott Michel authored
they were too tight according to bug 3126. Fix bug 3126. llvm-svn: 60006
-
Douglas Gregor authored
llvm-svn: 60005
-
Zhongxing Xu authored
llvm-svn: 60002
-
Scott Michel authored
llvm-svn: 60001
-
Bill Wendling authored
llvm-svn: 60000
-
Dan Gohman authored
introduce any new spilling; it just uses unused registers. Refactor the SUnit topological sort code out of the RRList scheduler and make use of it to help with the post-pass scheduler. llvm-svn: 59999
-
Scott Michel authored
llvm-svn: 59998
-
Daniel Dunbar authored
llvm-svn: 59997
-
Zhongxing Xu authored
llvm-svn: 59995
-
Bill Wendling authored
llvm-svn: 59992
-
- Nov 24, 2008
-
-
Chris Lattner authored
llvm-svn: 59990
-
Chris Lattner authored
llvm-svn: 59989
-
Fariborz Jahanian authored
is imported from a protocol into the implementation. llvm-svn: 59988
-
Sebastian Redl authored
llvm-svn: 59987
-
Chris Lattner authored
llvm-svn: 59986
-
Chris Lattner authored
llvm-svn: 59985
-
Ted Kremenek authored
llvm-svn: 59983
-
Ted Kremenek authored
properly reversed once constructed. This fixes PR 3125: http://llvm.org/bugs/show_bug.cgi?id=3125 llvm-svn: 59982
-
Bill Wendling authored
llvm-svn: 59981
-
Sebastian Redl authored
llvm-svn: 59979
-
Dan Gohman authored
simplify header dependencies for front-ends that just want to choose a scheduler and don't need all the scheduling machinery declarations. llvm-svn: 59978
-
Dan Gohman authored
llvm-svn: 59977
-
Chris Lattner authored
llvm-svn: 59976
-
Sebastian Redl authored
llvm-svn: 59975
-
Sebastian Redl authored
There might be other, similar bugs lurking there. llvm-svn: 59974
-
Ted Kremenek authored
llvm-svn: 59973
-
Chris Lattner authored
with an undef. llvm-svn: 59972
-
Bill Wendling authored
- Mark "add with overflow" as having a custom lowering for X86. Give it a null lowering representation for now. llvm-svn: 59971
-
Scott Michel authored
(a) Slight rethink on i64 zero/sign/any extend code - use a shuffle to directly zero-extend i32 to i64, but use rotates and shifts for sign extension. Also ensure unified register consistency. (b) Add new test harness for i64 operations: i64ops.ll llvm-svn: 59970
-
Dan Gohman authored
to removePred because an SUnit can both data-depend and anti-depend on the same SUnit. llvm-svn: 59969
-
Dan Gohman authored
llvm-svn: 59968
-
Dan Gohman authored
obscure tail-merging opportunities. llvm-svn: 59967
-
Dan Gohman authored
llvm-svn: 59966
-
Scott Michel authored
(a) Improve the extract element code: there's no need to do gymnastics with rotates into the preferred slot if a shuffle will do the same thing. (b) Rename a couple of SPUISD pseudo-instructions for readability and better semantic correspondence. (c) Fix i64 sign/any/zero extension lowering. llvm-svn: 59965
-
Dan Gohman authored
llvm-svn: 59964
-
Dan Gohman authored
llvm-svn: 59963
-
Dan Gohman authored
llvm-svn: 59962
-