Fix PR8247: JumpThreading can cause a block to become unreachable while still having predecessor, if it is part of a self-loop.
Because of this, we cannot use the Simplify* APIs, as they can assert-fail on unreachable code.  Since it's not easy to determine
if a given threading will cause a block to become unreachable, simply defer simplifying simplification to later InstCombine and/or
DCE passes.

llvm-svn: 115082

llvm-svn: 115081

is not everything, but the remaining cases are less trivial.

llvm-svn: 115080

llvm-svn: 115079

Fix breakpoint id range testing to disallow ranges that specify breakpoint locations from
crossing major breakpoint boundaries (must be within a single breakpoint if specifying locations).

Add .* as a means of specifying all the breakpoint locations under a major breakpoint, e.g. "3.*"
means "all the breakpoint locations of breakpoint 3".

Fix error message to make more sense, if user attempts to specify a breakpoint command when there
isn't a target yet.

llvm-svn: 115077

llvm-svn: 115076

llvm-svn: 115075

llvm-svn: 115074

the Python runtime take over in case there isn't a match.

llvm-svn: 115073

llvm-svn: 115072

 - I don't like returning StringRef's ever, unless it is actually important for
   performance, which it isn't here.

 - Also, stop validating getClangRevision to be an integer, I don't see a good
   reason to do this.

llvm-svn: 115071

groks the data type of the variable.

llvm-svn: 115070

llvm-svn: 115068

llvm-svn: 115067

which require the use of the shifter-operand. This will be used to split
the ldr/str instructions such that those versions needing the shifter operand
can get a different scheduling itenerary, as in some cases, the use of the
shifter can cause different scheduling than the simpler forms.

llvm-svn: 115066

    # rdar://problem/8493023
    # test/types failures for Test*TypesExpr.py: element offset computed wrong?

llvm-svn: 115065

  lib/Target/X86/X86MCCodeEmitter.cpp: 190: error: suggest parentheses around '&&' within '||'

llvm-svn: 115064

llvm-svn: 115063

for LLVM 2.9

llvm-svn: 115062

llvm-svn: 115061

were causing the prompt to be stomped on, mangled or omitted occasionally.

llvm-svn: 115059

llvm-svn: 115058

 - Minor style tweaks by me.

llvm-svn: 115056

llvm-svn: 115055

Wrestling with the slowly dawning realization that <atomic> isn't implementable on any compiler at my disposal...

llvm-svn: 115054

llvm-svn: 115053

The failures are similar in nature to the radar already filed:

        # rdar://problem/8492646
        # test/foundation fails after updating to tot r115023
        # self->str displays nothing as output

llvm-svn: 115052

during template instantiation, from Martin Vejnar!

llvm-svn: 115051

test_data_type_and_expr_with_dwarf().

rdar://problem/8492646
test/foundation fails after updating to tot r115023: self->str displays nothing as output

llvm-svn: 115050

patch by Jonathan Mulder!

llvm-svn: 115049

 - Also, fix indention in GetSourceVersion while in the area.

llvm-svn: 115048

LDM/STM instructions can run one cycle faster on some ARM processors if the
memory address is 64-bit aligned.  Radar 8489376.

llvm-svn: 115047

llvm-svn: 115046

llvm-svn: 115045

llvm-svn: 115044

llvm-svn: 115043

Also chnaged the expected string for 'frame variable this' from '(class C *const) this ='
to 'C *const) this =' for the time being, while investigating the different output for
tot r115023.

runCmd: frame variable this
output: (struct C *const) this = 0x0000000100000c2e
llvm-svn: 115042

llvm-svn: 115041

shell runner.

We would inadvertently end up holding on to handles to the temporary files
longer than we should have been. On Win32, where open handles lock some file
operations, this caused problems in tests which would try to move temporary
files around (as Clang does by default now).

Many thanks to Francois Pichet for the excellent detective work on this.

llvm-svn: 115040

llvm-svn: 115039