encode/decode function alignment in bc files

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24218 91177308-0d34-0410-b5e6-96231b3b80d8

describe extensions to the .bc format for function/global alignment

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24217 91177308-0d34-0410-b5e6-96231b3b80d8

Read/write global variable alignments if present

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24216 91177308-0d34-0410-b5e6-96231b3b80d8

add alignment info for globals and functions

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24213 91177308-0d34-0410-b5e6-96231b3b80d8

print alignment info for globals and functions

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24212 91177308-0d34-0410-b5e6-96231b3b80d8

regenerate

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24211 91177308-0d34-0410-b5e6-96231b3b80d8

Allow globals to have an alignment specified.  Switch to using isPowerOf2_32
at Jim's request for the checking code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24210 91177308-0d34-0410-b5e6-96231b3b80d8

allow functions and modules to have an explicit alignment

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24209 91177308-0d34-0410-b5e6-96231b3b80d8

regenerate

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24208 91177308-0d34-0410-b5e6-96231b3b80d8

factor optional alignment

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24207 91177308-0d34-0410-b5e6-96231b3b80d8

ask for 16-byte aligned jmpbufs. This should unbreak C++ on IA64 (and
a bunch of other things) but is currently ignored by the code
generator.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24206 91177308-0d34-0410-b5e6-96231b3b80d8

enumerate non-standard argument encoding cases, such as alignment info for
allocations

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24205 91177308-0d34-0410-b5e6-96231b3b80d8

rearrange some info about the instruction encoding

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24204 91177308-0d34-0410-b5e6-96231b3b80d8

Write/read allocation instruction alignment info to .bc files.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24203 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24202 91177308-0d34-0410-b5e6-96231b3b80d8

add an accessor

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24201 91177308-0d34-0410-b5e6-96231b3b80d8

verify that alignments are always a power of 2

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24200 91177308-0d34-0410-b5e6-96231b3b80d8

regenerate

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24199 91177308-0d34-0410-b5e6-96231b3b80d8

Verify that alignment amounts are a power of 2

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24198 91177308-0d34-0410-b5e6-96231b3b80d8

fix printing the alignment directive

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24197 91177308-0d34-0410-b5e6-96231b3b80d8

Add support alignment of allocation instructions.
Add support for specifying alignment and size of setjmp jmpbufs.

No targets currently do anything with this information, nor is it presrved
in the bytecode representation.  That's coming up next.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24196 91177308-0d34-0410-b5e6-96231b3b80d8

add a case Nate sent me

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24195 91177308-0d34-0410-b5e6-96231b3b80d8

Implement Transforms/TailCallElim/return-undef.ll, a trivial case
that has been sitting in my inbox since May 18. :)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24194 91177308-0d34-0410-b5e6-96231b3b80d8

New testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24191 91177308-0d34-0410-b5e6-96231b3b80d8

Turn sdiv into udiv if both operands have a clear sign bit.  This occurs
a few times in crafty:

OLD:    %tmp.36 = div int %tmp.35, 8            ; <int> [#uses=1]
NEW:    %tmp.36 = div uint %tmp.35, 8           ; <uint> [#uses=0]
OLD:    %tmp.19 = div int %tmp.18, 8            ; <int> [#uses=1]
NEW:    %tmp.19 = div uint %tmp.18, 8           ; <uint> [#uses=0]
OLD:    %tmp.117 = div int %tmp.116, 8          ; <int> [#uses=1]
NEW:    %tmp.117 = div uint %tmp.116, 8         ; <uint> [#uses=0]
OLD:    %tmp.92 = div int %tmp.91, 8            ; <int> [#uses=1]
NEW:    %tmp.92 = div uint %tmp.91, 8           ; <uint> [#uses=0]

Which all turn into shrs.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24190 91177308-0d34-0410-b5e6-96231b3b80d8

Turn srem -> urem when neither input has their sign bit set.  This triggers
8 times in vortex, allowing the srems to be turned into shrs:

OLD:    %tmp.104 = rem int %tmp.5.i37, 16               ; <int> [#uses=1]
NEW:    %tmp.104 = rem uint %tmp.5.i37, 16              ; <uint> [#uses=0]
OLD:    %tmp.98 = rem int %tmp.5.i24, 16                ; <int> [#uses=1]
NEW:    %tmp.98 = rem uint %tmp.5.i24, 16               ; <uint> [#uses=0]
OLD:    %tmp.91 = rem int %tmp.5.i19, 8         ; <int> [#uses=1]
NEW:    %tmp.91 = rem uint %tmp.5.i19, 8                ; <uint> [#uses=0]
OLD:    %tmp.88 = rem int %tmp.5.i14, 8         ; <int> [#uses=1]
NEW:    %tmp.88 = rem uint %tmp.5.i14, 8                ; <uint> [#uses=0]
OLD:    %tmp.85 = rem int %tmp.5.i9, 1024               ; <int> [#uses=2]
NEW:    %tmp.85 = rem uint %tmp.5.i9, 1024              ; <uint> [#uses=0]
OLD:    %tmp.82 = rem int %tmp.5.i, 512         ; <int> [#uses=2]
NEW:    %tmp.82 = rem uint %tmp.5.i1, 512               ; <uint> [#uses=0]
OLD:    %tmp.48.i = rem int %tmp.5.i.i161, 4            ; <int> [#uses=1]
NEW:    %tmp.48.i = rem uint %tmp.5.i.i161, 4           ; <uint> [#uses=0]
OLD:    %tmp.20.i2 = rem int %tmp.5.i.i, 4              ; <int> [#uses=1]
NEW:    %tmp.20.i2 = rem uint %tmp.5.i.i, 4             ; <uint> [#uses=0]

it also occurs 9 times in gcc, but with odd constant divisors (1009 and 61)
so the payoff isn't as great.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24189 91177308-0d34-0410-b5e6-96231b3b80d8

Fix logic bug in finding retry slot in tally.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24188 91177308-0d34-0410-b5e6-96231b3b80d8

Fix a warning

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24187 91177308-0d34-0410-b5e6-96231b3b80d8

oops, forgot to load GP for indirect calls, though the old code now commented
out failed (e.g. methcall) - now the code compiles, though it's not quite
right just yet (tm) ;)

would fix this but it's 3am! :O

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24186 91177308-0d34-0410-b5e6-96231b3b80d8

kill redundant SP/GP/RP save/restores across calls

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24183 91177308-0d34-0410-b5e6-96231b3b80d8

add support for loading bools

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24182 91177308-0d34-0410-b5e6-96231b3b80d8

Scheduling now uses itinerary data.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24180 91177308-0d34-0410-b5e6-96231b3b80d8

<cassert> no longer required to make VC++ happy.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24177 91177308-0d34-0410-b5e6-96231b3b80d8

change NULL to 0, unbreaks the ppc target when building on ia64

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24176 91177308-0d34-0410-b5e6-96231b3b80d8

fun with predicates! (add TRUNC i64->i1, AND i1 i1, fix XOR i1 i1)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24175 91177308-0d34-0410-b5e6-96231b3b80d8

1. Remove ranges from itinerary data.

2. Tidy up the subtarget emittined code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24172 91177308-0d34-0410-b5e6-96231b3b80d8

Fix mac os spelling

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24171 91177308-0d34-0410-b5e6-96231b3b80d8

Add more bison versions, thanks to Vladimir

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24169 91177308-0d34-0410-b5e6-96231b3b80d8

Per bug 655, give people more options in case 1.35 doesn't build on their
system.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24168 91177308-0d34-0410-b5e6-96231b3b80d8

add pattern to load constant 0 into a predicate reg

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24164 91177308-0d34-0410-b5e6-96231b3b80d8

add a hack that fixes:
llvm-gcc main.c -Wl,-native -o a.out -g

This is important because it used by many configure scripts.

John, please pull this onto the 1.6 branch.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24163 91177308-0d34-0410-b5e6-96231b3b80d8

Reject integer literals that are out of range for their type.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24162 91177308-0d34-0410-b5e6-96231b3b80d8

Fix a bug that prevented this pattern from matching

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24161 91177308-0d34-0410-b5e6-96231b3b80d8

Fix a crash that Andrew noticed, and add a pair of braces to unfconfuse
XCode's indenting.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24159 91177308-0d34-0410-b5e6-96231b3b80d8

make this 64 bit clean, fixed test30 of /Regression/Transforms/InstCombine/add.ll

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24158 91177308-0d34-0410-b5e6-96231b3b80d8

This is missed by InstCombine, patch comming

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24157 91177308-0d34-0410-b5e6-96231b3b80d8

Fix a QOI issue noticed by Markus F.X.J. Oberhumer.
This fixes PR641

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24154 91177308-0d34-0410-b5e6-96231b3b80d8

"fix" support for FP constants (this code asserts in the scheduler,
though)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24152 91177308-0d34-0410-b5e6-96231b3b80d8

add F0 and F1 to the FP register class

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24151 91177308-0d34-0410-b5e6-96231b3b80d8

This works now

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24150 91177308-0d34-0410-b5e6-96231b3b80d8

Add support for immediates directly in the pattern, this allows itanium to
define:

def : Pat<(i1 1), (CMPEQ r0, r0)>;

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24149 91177308-0d34-0410-b5e6-96231b3b80d8

Keep VC++ happy.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24148 91177308-0d34-0410-b5e6-96231b3b80d8

add support for SELECT to TargetSelectionDAG.td, add support for
selecting ints to IA64, and a few other ia64 bits and pieces

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24147 91177308-0d34-0410-b5e6-96231b3b80d8

add support for loading FP constants +0.0 and +1.0 to the dag isel,
stop pretending -0.0 and -1.0 are machine constants

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24146 91177308-0d34-0410-b5e6-96231b3b80d8

Fix a source of undefined behavior when dealing with 64-bit types. This
may fix PR652.  Thanks to Andrew for tracking down the problem.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24145 91177308-0d34-0410-b5e6-96231b3b80d8

* Replace ampersands in section titles with more formal ``and''
* Surround C++ template operators with <tt>
* Add <> after templated operators for consistency

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24144 91177308-0d34-0410-b5e6-96231b3b80d8

Re-generated to fix copy-paste typo noticed by Marco Matthies.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24143 91177308-0d34-0410-b5e6-96231b3b80d8

Fixed copy-paste typo, patch by Marco Matthies.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24142 91177308-0d34-0410-b5e6-96231b3b80d8

Allow itineraries to be passed through the Target Machine.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24139 91177308-0d34-0410-b5e6-96231b3b80d8

Keep VC++ happy.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24137 91177308-0d34-0410-b5e6-96231b3b80d8

Let people who run the nightly tester specify paths to external tests, instead
of having to hack the nightly tester script itself.

as an example, I use the following for my machine:

$HOME/llvm/utils/NightlyTest.pl -parallel -release -enable-llcbeta \
     -spec2000path /Volumes/ProjectsDisk/cvs/benchmarks/speccpu2000-llvm/benchspec/ \
     -povraypath /Volumes/ProjectsDisk/cvs/benchmarks/povray31 \
     -namdpath /Volumes/ProjectsDisk/cvs/benchmarks/namd

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24136 91177308-0d34-0410-b5e6-96231b3b80d8

heh, scheduling was easy?

need to send chris, jim and sampo a box of fish each

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24135 91177308-0d34-0410-b5e6-96231b3b80d8

FORTRAN!!! :( and other similarly unfortunate things mean that on ia64
one sometimes needs to pass FP args in both FP *and* integer registers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24134 91177308-0d34-0410-b5e6-96231b3b80d8

so tablegen was thinking I might want to convert FPs to predicates.
clever little tablegen!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24133 91177308-0d34-0410-b5e6-96231b3b80d8

add support for int->FP and FP->int ops, and add ia64 patterns for these

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24132 91177308-0d34-0410-b5e6-96231b3b80d8

add zeroextend predicate->integer

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24131 91177308-0d34-0410-b5e6-96231b3b80d8

Add a flag to enable a darwin linker optimization

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24130 91177308-0d34-0410-b5e6-96231b3b80d8

Make constant pool entries use private labels.  This is important when you're
not compiling a whole program at a time :)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24129 91177308-0d34-0410-b5e6-96231b3b80d8

Updated alpha known problems.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24126 91177308-0d34-0410-b5e6-96231b3b80d8

Apparently these do pass on some alphas

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24125 91177308-0d34-0410-b5e6-96231b3b80d8

Fix an iterator invalidation problem in code used by the -strip pass

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24124 91177308-0d34-0410-b5e6-96231b3b80d8

Limit the search depth of MaskedValueIsZero to 6 instructions, to avoid
bad cases.  This fixes Markus's second testcase in PR639, and should
seal it for good.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24123 91177308-0d34-0410-b5e6-96231b3b80d8

Emit itinerary class in instruction info.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24122 91177308-0d34-0410-b5e6-96231b3b80d8

Generate cpu to itinerary map.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24121 91177308-0d34-0410-b5e6-96231b3b80d8

these tests fail on alpha for some reason

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24120 91177308-0d34-0410-b5e6-96231b3b80d8

1. Embed and not inherit vector for NodeGroup.

2. Iterate operands and not uses (performance.)

3. Some long pending comment changes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24119 91177308-0d34-0410-b5e6-96231b3b80d8

add FP compares and implicit register defs to the dag isel

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24118 91177308-0d34-0410-b5e6-96231b3b80d8

Update Getting Started for Visual Studio page.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24117 91177308-0d34-0410-b5e6-96231b3b80d8

Significantly simplify this code and make it more aggressive.  Instead of having
a special case hack for X86, make the hack more general: if an incoming argument
register is not used in any block other than the entry block, don't copy it to
a vreg.  This helps us compile code like this:

%struct.foo = type { int, int, [0 x ubyte] }
int %test(%struct.foo* %X) {
        %tmp1 = getelementptr %struct.foo* %X, int 0, uint 2, int 100
        %tmp = load ubyte* %tmp1                ; <ubyte> [#uses=1]
        %tmp2 = cast ubyte %tmp to int          ; <int> [#uses=1]
        ret int %tmp2
}

to:

_test:
        lbz r3, 108(r3)
        blr

instead of:

_test:
        lbz r2, 108(r3)
        or r3, r2, r2
        blr

The (dead) copy emitted to copy r3 into a vreg for extra-block uses was
increasing the live range of r3 past the load, preventing the coallescing.

This implements CodeGen/PowerPC/reg-coallesce-simple.ll

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24115 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24112 91177308-0d34-0410-b5e6-96231b3b80d8

Reduce the number of copies emitted as machine instructions by
generating results in vregs that will need them.  In the case of something
like this:  CopyToReg((add X, Y), reg1024), we no longer emit code like
this:

   reg1025 = add X, Y
   reg1024 = reg 1025

Instead, we emit:

   reg1024 = add X, Y

Whoa! :)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24111 91177308-0d34-0410-b5e6-96231b3b80d8

If the module has no t-t and the host is an alpha, default to using the Alpha BE

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24110 91177308-0d34-0410-b5e6-96231b3b80d8

fix some broken comparisons, this affected the Pattern isel too.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24109 91177308-0d34-0410-b5e6-96231b3b80d8

The pass was removed, but the spirit lives on

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24108 91177308-0d34-0410-b5e6-96231b3b80d8

This is implemented

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24107 91177308-0d34-0410-b5e6-96231b3b80d8

Codegen mul by negative power of two with a shift and negate.
This implements test/Regression/CodeGen/PowerPC/mul-neg-power-2.ll,
producing:

_foo:
        slwi r2, r3, 1
        subfic r3, r2, 63
        blr

instead of:

_foo:
        mulli r2, r3, -2
        addi r3, r2, 63
        blr

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24106 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24103 91177308-0d34-0410-b5e6-96231b3b80d8

Fix a problem that Nate noticed with LSR:

When inserting code for an addrec expression with a non-unit stride, be
more careful where we insert the multiply.  In particular, insert the multiply
in the outermost loop we can, instead of the requested insertion point.

This allows LSR to notice the mul in the right loop, reducing it when it gets
to it.  This allows it to reduce the multiply, where before it missed it.

This happens quite a bit in the test suite, for example, eliminating 2
multiplies in art, 3 in ammp, 4 in apsi, reducing from 1050 multiplies to
910 muls in galgel (!), from 877 to 859 in applu, and 36 to 30 in bzip2.

This speeds up galgel from 16.45s to 16.01s, applu from 14.21 to 13.94s and
fourinarow from 66.67s to 63.48s.

This implements Transforms/LoopStrengthReduce/nested-reduce.ll

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24102 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24099 91177308-0d34-0410-b5e6-96231b3b80d8

Revert an accidental commit.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24098 91177308-0d34-0410-b5e6-96231b3b80d8

More files deleted.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24097 91177308-0d34-0410-b5e6-96231b3b80d8

Make -time-passes output prettier

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24096 91177308-0d34-0410-b5e6-96231b3b80d8

Make negative immediates in patterns work correctly, silence some warnings
building the itanium backend.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24095 91177308-0d34-0410-b5e6-96231b3b80d8

add some FP stuff, some mix.* stuff, and constant pool support to the
DAG instruction selector, which should be destroyed one day (in the pattern
isel also) since ia64 can pack any constant in the instruction stream

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24094 91177308-0d34-0410-b5e6-96231b3b80d8

Fix <> in html

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24093 91177308-0d34-0410-b5e6-96231b3b80d8

First hack-and-slash on the release notes.  Comments and improvements welcome

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24092 91177308-0d34-0410-b5e6-96231b3b80d8

Remove the lowerconstantexprs pass

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24091 91177308-0d34-0410-b5e6-96231b3b80d8

This pass is now obsolete since all targets have moved to the SelectionDAG
infrastructure and the simple isels have been removed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24090 91177308-0d34-0410-b5e6-96231b3b80d8

Remove the LowerConstantExpressionsPass pass

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24089 91177308-0d34-0410-b5e6-96231b3b80d8

remove reference to this pass

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24088 91177308-0d34-0410-b5e6-96231b3b80d8