same Value without breaking things.

llvm-svn: 83861

llvm-svn: 83860

llvm-svn: 83857

llvm-svn: 83856

llvm-svn: 83854

it to hold the address of an sret return value, for x86-64 ABI purposes.

Also, fix the test that was originally intended to test this to actually
test it, using FileCheck.

llvm-svn: 83853

llvm-svn: 83852

llvm-svn: 83851

llvm-svn: 83850

llvm-svn: 83849

llvm-svn: 83848

llvm-svn: 83837

llvm-svn: 83835

llvm-svn: 83834

llvm-svn: 83833

llvm-svn: 83828

unbreaks make check after evocallaghan's changes.

llvm-svn: 83827

Fix syntax error missed in converting zext.ll test. Convert 2003-11-13-ConstExprCastCall.ll to FileCheck from notcast.

llvm-svn: 83826

llvm-svn: 83825

compile time penalty on gnugo, the worst case in MultiSource, is down to
about 2.5% from 30%

llvm-svn: 83824

llvm-svn: 83823

llvm-svn: 83822

llvm-svn: 83821

llvm-svn: 83820

llvm-svn: 83819

it to visit instructions from the start of the function to the
end of the function in the first path.  This greatly speeds up
some pathological cases (e.g. PR5150).

Try #3, this time with some unneeded debug info stuff removed
which was causing dead pointers to be added to the worklist.

llvm-svn: 83818

llvm-svn: 83817

llvm-svn: 83815

it to visit instructions from the start of the function to the
end of the function in the first path.  This greatly speeds up
some pathological cases (e.g. PR5150). 

llvm-svn: 83814

llvm-svn: 83813

llvm-svn: 83812

llvm-svn: 83811

into a shuffle even if it was used by another insertelement.  If the
visitation order of instcombine was wrong, this would turn a chain of
insertelements into a chain of shufflevectors, which was quite painful.

Since CollectShuffleElements handles these cases, the code can just
be nuked.

llvm-svn: 83810

llvm-svn: 83809

llvm-svn: 83807

llvm-svn: 83806

llvm-svn: 83805

llvm-svn: 83804

llvm-svn: 83803

llvm-svn: 83802