llvm-svn: 94096

llvm-svn: 94095

llvm-svn: 94094

llvm-svn: 94093

llvm-svn: 94092

llvm-svn: 94091

ObjCObjectPointerType using the allocator associated with ASTContext.
Not only does this fix a memory leak, but it also makes these arrays
BumpPtrAllocated (in the typical case).

llvm-svn: 94090

llvm-svn: 94089

llvm-svn: 94088

Fixes radar 7562952.

llvm-svn: 94087

assignments.

llvm-svn: 94086

kinds, so that we see referenced types, protocols, classes, etc.

llvm-svn: 94085

unary operators.

llvm-svn: 94084

checks that the runtime library can be linked and works for compiling on 10.6
for 10.{4,5,6} targets. Currently test is only for x86, and misses some
compiler-rt functions.

llvm-svn: 94083

llvm-svn: 94082

assuming SVN. This should be fixed to not necessarily be an integer.

llvm-svn: 94081

one reason or another don't belong in the regular test suite.

Inside is utils/OptionalTests/Extra which is specifically for tests that match
the layout of the regular tests.

llvm-svn: 94080

llvm-svn: 94079

typedefs only (for now).

llvm-svn: 94078

repository. Feel free to add support for your favorite VCS.

llvm-svn: 94077

llvm-svn: 94076

llvm-svn: 94075

CallExprs as those edges help cause a n^2 explosion in the number of
destructor calls.  Other consumers, such as static analysis, that
would like to have more a more complete CFG can select the inclusion
of those edges as CFG build time.

This also fixes up the two compilation users of CFGs to be tolerant of
having or not having those edges.  All catch code is assumed be to
live if we didn't generate the exceptional edges for CallExprs.

llvm-svn: 94074

llvm-svn: 94073

message, to be more informative.

llvm-svn: 94072

llvm-svn: 94071

rememberInstruction so that future users of that user will be inserted
in the correct position. This fixes the Darwin selfhost.

llvm-svn: 94070

if one of the vectors didn't have elements (such as undef). Fixes PR 6096.

Fix an issue in the constant folder where fcmp (<2 x %ty>, <2 x %ty>) would
have <2 x i1> type if constant folding was successful and i1 type if it wasn't.
This exposed a related issue in the bitcode reader.

llvm-svn: 94069

llvm-svn: 94068

llvm-svn: 94067

they reach codegen.

llvm-svn: 94066

llvm-svn: 94065

for unreachable loops if all the blocks are already marked live by
this point.

llvm-svn: 94064

work, if we know we already marked all blocks as live.

llvm-svn: 94063

the CFG.  WIP.

llvm-svn: 94062

This new version is much more aggressive about doing "full" reduction in
cases where it reduces register pressure, and also more aggressive about
rewriting induction variables to count down (or up) to zero when doing so
reduces register pressure.

It currently uses fairly simplistic algorithms for finding reuse
opportunities, but it introduces a new framework allows it to combine
multiple strategies at once to form hybrid solutions, instead of doing
all full-reduction or all base+index.

llvm-svn: 94061

llvm-svn: 94060

answers.  This will update as object size checking gets better information.

llvm-svn: 94059

Fix a minor issue in x86 load / store folding table. movups does an unaligned load so it doesn't require 16-byte alignment.

llvm-svn: 94058

order for SjLj style exception handling.

llvm-svn: 94055