stringmap memory managed with malloc now

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35666 91177308-0d34-0410-b5e6-96231b3b80d8

Fix 2007-04-04-BadFoldBitcastIntoMalloc.ll

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35665 91177308-0d34-0410-b5e6-96231b3b80d8

Add new test.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35664 91177308-0d34-0410-b5e6-96231b3b80d8

Better still.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35663 91177308-0d34-0410-b5e6-96231b3b80d8

Fix test case.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35662 91177308-0d34-0410-b5e6-96231b3b80d8

These got better.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35661 91177308-0d34-0410-b5e6-96231b3b80d8

Re-materialize all loads from fixed stack slots.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35660 91177308-0d34-0410-b5e6-96231b3b80d8

Add isFixedObjectIndex. It returns true if the stack slot index is for a fixed stack object.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35659 91177308-0d34-0410-b5e6-96231b3b80d8

Trivially re-materializable instructions have spill weights that are half of what it would be otherwise.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35658 91177308-0d34-0410-b5e6-96231b3b80d8

Revert this until the Date::Parse module can be installed on the nightly
testers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35657 91177308-0d34-0410-b5e6-96231b3b80d8

For PR1302:
Use local variable names that match the function parameter name that it
is passed to so the code is more clear, to wit: is_bytecode -> is_native

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35656 91177308-0d34-0410-b5e6-96231b3b80d8

Fix comment.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35655 91177308-0d34-0410-b5e6-96231b3b80d8

For PR1302:
Rename LinkItems as NativeLinkItems since it is filled out by the Linker
library to contain only those items deemed to be native.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35654 91177308-0d34-0410-b5e6-96231b3b80d8

For PR1302:
Implement file tests for both LinkInLibrary and LinkInFile to determine if
the file is native. Don't generate warnings if the file is native.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35653 91177308-0d34-0410-b5e6-96231b3b80d8

Make the LinkInFile and LinkInLibrary both return an indication of whether
the file was found to be a native library (and thus not linked by these
functions).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35652 91177308-0d34-0410-b5e6-96231b3b80d8

For PR1302:
Adjust useage of sys::Path::FileType for new enumerator names.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35651 91177308-0d34-0410-b5e6-96231b3b80d8

For PR1302:
Implement recognition of COFF, ELF and Mach-O object/shared lib files.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35650 91177308-0d34-0410-b5e6-96231b3b80d8

For PR1302:
Make the FileType enumerators more readable and add COFF, ELF and Mach-O.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35649 91177308-0d34-0410-b5e6-96231b3b80d8

trivial optimization

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35648 91177308-0d34-0410-b5e6-96231b3b80d8

Appease Chris' pedantic streak.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35647 91177308-0d34-0410-b5e6-96231b3b80d8

Fix links to intrinsic functions outside LangRef.html

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35646 91177308-0d34-0410-b5e6-96231b3b80d8

Remove intrinsics that we will not be implementing.
Fix the anchor names of intrinsics to all be int_ so as to distinguish
them from the i_ prefix reserved for instructions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35645 91177308-0d34-0410-b5e6-96231b3b80d8

use calloc instead of new/memset, it is more efficient

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35644 91177308-0d34-0410-b5e6-96231b3b80d8

Initialize the symbol table to zero explicitly.  This ensures that the
symbol table does no allocations for prototypes or other lazily deserialized
functions, saving significant space and time.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35643 91177308-0d34-0410-b5e6-96231b3b80d8

Extend StringMap to support being initialized as completely empty.  When
initialized this way, they do not do a malloc to allocate their buckets.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35642 91177308-0d34-0410-b5e6-96231b3b80d8

New asm test case.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35641 91177308-0d34-0410-b5e6-96231b3b80d8

Implement inline asm modifier P.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35640 91177308-0d34-0410-b5e6-96231b3b80d8

Typo.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35639 91177308-0d34-0410-b5e6-96231b3b80d8

Adding more MMX instructions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35638 91177308-0d34-0410-b5e6-96231b3b80d8

make a new missing features section

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35637 91177308-0d34-0410-b5e6-96231b3b80d8

New test case.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35636 91177308-0d34-0410-b5e6-96231b3b80d8

Remove unused constant pool entries.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35635 91177308-0d34-0410-b5e6-96231b3b80d8

Updated

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35634 91177308-0d34-0410-b5e6-96231b3b80d8

Fix a bug I introduced with my patch yesterday which broke Qt (I converted
some constant exprs to apints).

Thanks to Anton for tracking down a small testcase that triggered this!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35633 91177308-0d34-0410-b5e6-96231b3b80d8

greatly reduce hte default size of stringmap.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35632 91177308-0d34-0410-b5e6-96231b3b80d8

move these to FrontendAda

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35631 91177308-0d34-0410-b5e6-96231b3b80d8

testcase for recent bugfix

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35630 91177308-0d34-0410-b5e6-96231b3b80d8

move ada tests

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35629 91177308-0d34-0410-b5e6-96231b3b80d8

New test case.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35628 91177308-0d34-0410-b5e6-96231b3b80d8

Fixed a bug that causes codegen of noop like add r0, r0, #0.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35627 91177308-0d34-0410-b5e6-96231b3b80d8

reinstate the previous two patches, with a bugfix :)

ldecod now passes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35626 91177308-0d34-0410-b5e6-96231b3b80d8

Starting implementation of the ELF32 ABI specification of varargs handling.
LowerVASTART emits the right code if the subtarget is ELF32, the other intrinsics
(VAARG, VACOPY and VAEND) are not yet implemented.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35625 91177308-0d34-0410-b5e6-96231b3b80d8

The PPC64 ELF ABI is "intended to use the same structure layout and calling convention rules
as the 64-bit PowerOpen ABI" (Reference http://www.linux-foundation.org/spec/ELF/ppc64/).
Change all ELF tests to ELF32.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35624 91177308-0d34-0410-b5e6-96231b3b80d8

Addition to the previous commit for getCalleeSavedRegClasses:

"The ELF ABI specifies F1-F8 registers as argument registers for double, not
F1-F10. This affects only ELF, not MachO."

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35623 91177308-0d34-0410-b5e6-96231b3b80d8

The ELF ABI specifies F1-F8 registers as argument registers for double, not
F1-F10. This affects only ELF, not MachO.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35622 91177308-0d34-0410-b5e6-96231b3b80d8

Prepare for Subversion migration by implementing a -usesvn to tell the
script to to check out llvm and llvm-test from Subversion instead of CVS.
Without this option the script will continue to check out from CVS. To
specify the Subversion URL, set the SVNURL environment variable or pass
-svnurl followed by the URL. For now, -svnurl will default to Reid's
temporary (read-only, daily snapshot) SVN server. Try it out if you like!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35621 91177308-0d34-0410-b5e6-96231b3b80d8

Reverting back to 1.723. The last two commits broke JM (and possibily others) on ARM.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35620 91177308-0d34-0410-b5e6-96231b3b80d8

Inverted logic.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35619 91177308-0d34-0410-b5e6-96231b3b80d8

Bad bad bug. findRegisterUseOperand() returns -1 if a use if not found.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35618 91177308-0d34-0410-b5e6-96231b3b80d8

Changed to new MMX_ recipes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35617 91177308-0d34-0410-b5e6-96231b3b80d8

Add FEMMS and ADDQ. Renamed MMX recipes to prepend the MMX_ to them.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35616 91177308-0d34-0410-b5e6-96231b3b80d8

split some code out into a helper function

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35615 91177308-0d34-0410-b5e6-96231b3b80d8

Split a whole ton of code out of visitICmpInst into visitICmpInstWithInstAndIntCst.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35614 91177308-0d34-0410-b5e6-96231b3b80d8

add missing operator

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35613 91177308-0d34-0410-b5e6-96231b3b80d8

Fix PR1253 and xor2.ll:test[01]

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35612 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase for PR1253

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35611 91177308-0d34-0410-b5e6-96231b3b80d8

add a helper function to constantint.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35610 91177308-0d34-0410-b5e6-96231b3b80d8

Arm supports negative strides as well, add them.  This lets us compile:
CodeGen/ARM/arm-negative-stride.ll to:

LBB1_2: @bb
        str r1, [r3, -r0, lsl #2]
        add r0, r0, #1
        cmp r0, r2
        bne LBB1_2      @bb

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35609 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase, where we should use a negative stride

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35608 91177308-0d34-0410-b5e6-96231b3b80d8

allow -1 strides to reuse "1" strides.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35607 91177308-0d34-0410-b5e6-96231b3b80d8

Add a regtest for cases we now recognize as rotates, thanks to Scott Michel's
recent dagcombine patch

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35606 91177308-0d34-0410-b5e6-96231b3b80d8

1. Insert custom lowering hooks for ISD::ROTR and ISD::ROTL.

2. Help DAGCombiner recognize zero/sign/any-extended versions of ROTR and ROTL
patterns. This was motivated by the X86/rotate.ll testcase, which should now
generate code for other platforms (and soon-to-come platforms.) Rewrote code
slightly to make it easier to read.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35605 91177308-0d34-0410-b5e6-96231b3b80d8

fix this testcase so it passes

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35604 91177308-0d34-0410-b5e6-96231b3b80d8

fix this testcase on ppc hosts

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35603 91177308-0d34-0410-b5e6-96231b3b80d8

fix off by 1 error in displacement computation

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35602 91177308-0d34-0410-b5e6-96231b3b80d8

fix the CodeGen/ARM/2007-03-13-InstrSched.ll regression: allow IV's with scales
to be folded into non-store instructions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35601 91177308-0d34-0410-b5e6-96231b3b80d8

Ugh. Copy coalescer does not update register numbers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35600 91177308-0d34-0410-b5e6-96231b3b80d8

New test case.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35599 91177308-0d34-0410-b5e6-96231b3b80d8

add support for the 'w' inline asm register class.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35598 91177308-0d34-0410-b5e6-96231b3b80d8

Update example notes to clarify system linker's role.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35597 91177308-0d34-0410-b5e6-96231b3b80d8

Regenerate.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35596 91177308-0d34-0410-b5e6-96231b3b80d8

Check for .svn directories too to determine if a debug build is appropriate.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35595 91177308-0d34-0410-b5e6-96231b3b80d8

1. Make use of APInt operation instead of using ConstantExpr::getXXX.
2. Use cheaper APInt methods.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35594 91177308-0d34-0410-b5e6-96231b3b80d8

Use uint32_t for bitwidth instead of unsigned.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35593 91177308-0d34-0410-b5e6-96231b3b80d8

Pass the type of the store access, not the type of the store, into the
target hook.  This allows us to codegen a loop as:

LBB1_1: @cond_next
        mov r2, #0
        str r2, [r0, +r3, lsl #2]
        add r3, r3, #1
        cmn r3, #1
        bne LBB1_1      @cond_next

instead of:

LBB1_1: @cond_next
        mov r2, #0
        str r2, [r0], #+4
        add r3, r3, #1
        cmn r3, #1
        bne LBB1_1      @cond_next

This looks the same, but has one fewer induction variable (and therefore,
one fewer register) live in the loop.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35592 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35591 91177308-0d34-0410-b5e6-96231b3b80d8

Fix a bug which caused us to never be able to use signed comparisons for
equality comparisons of a constant.  This allows us to codegen the 'sintzero'
loop in PR1288 as:

LBB1_1: ;cond_next
        li r4, 0
        addi r2, r2, 1
        stw r4, 0(r3)
        addi r3, r3, 4
        cmpwi cr0, r2, -1
        bne cr0, LBB1_1 ;cond_next

instead of:

LBB1_1: ;cond_next
        addi r2, r2, 1
        li r4, 0
        xoris r5, r2, 65535
        stw r4, 0(r3)
        addi r3, r3, 4
        cmplwi cr0, r5, 65535
        bne cr0, LBB1_1 ;cond_next

This implements CodeGen/PowerPC/compare-simm.ll, and also cuts 74
instructions out of kc++.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35590 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35589 91177308-0d34-0410-b5e6-96231b3b80d8

Wrap long line

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35588 91177308-0d34-0410-b5e6-96231b3b80d8

use more obvious function name.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35587 91177308-0d34-0410-b5e6-96231b3b80d8

Treat xor of signbit like an add.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35586 91177308-0d34-0410-b5e6-96231b3b80d8

add a helper function.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35585 91177308-0d34-0410-b5e6-96231b3b80d8

simplify (x+c)^signbit as (x+c+signbit), pointed out by PR1288.  This implements
test/Transforms/InstCombine/xor.ll:test28

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35584 91177308-0d34-0410-b5e6-96231b3b80d8

creative way to add one.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35583 91177308-0d34-0410-b5e6-96231b3b80d8

my patch fixed the fixme.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35582 91177308-0d34-0410-b5e6-96231b3b80d8

Fix illegal assembly syntax.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35581 91177308-0d34-0410-b5e6-96231b3b80d8

bwsap -> bswap

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35580 91177308-0d34-0410-b5e6-96231b3b80d8

Regenerate

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35579 91177308-0d34-0410-b5e6-96231b3b80d8

Upgrade the bit count intrinsics to have an i32 result.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35578 91177308-0d34-0410-b5e6-96231b3b80d8

Add a test case to make sure that constant folding of the bit counting
intrinsics works.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35577 91177308-0d34-0410-b5e6-96231b3b80d8

reduce use of std::set

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35576 91177308-0d34-0410-b5e6-96231b3b80d8

Various passes before isel split edges and do other CFG-restructuring changes.

isel has its own particular features that it wants in the CFG, in order to
reduce the number of times a constant is computed, etc.  Make sure that we
clean up the CFG before doing any other things for isel.  Doing so can
dramatically reduce the number of split edges and reduce the number of
places that constants get computed.  For example, this shrinks
CodeGen/Generic/phi-immediate-factoring.ll from 44 to 37 instructions on X86,
and from 21 to 17 MBB's in the output.  This is primarily a code size win,
not a performance win.

This implements CodeGen/Generic/phi-immediate-factoring.ll and PR1296.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35575 91177308-0d34-0410-b5e6-96231b3b80d8

new testcase for PR1296

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35574 91177308-0d34-0410-b5e6-96231b3b80d8

- Divides the comparisons in two types: comparisons that only use N and Z
flags (ARMISD::CMPNZ) and comparisons that use all flags (ARMISD::CMP).
- Defines the instructions: TST, TEQ (ARM) and TST (Thumb).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35573 91177308-0d34-0410-b5e6-96231b3b80d8

add range version of insert()

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35572 91177308-0d34-0410-b5e6-96231b3b80d8

Regenerate

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35571 91177308-0d34-0410-b5e6-96231b3b80d8

Don't upgrade functions that look like a bswap intrinsic but aren't.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35570 91177308-0d34-0410-b5e6-96231b3b80d8

Let llvm-upgrade upgrade the bswap intrinsic.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35569 91177308-0d34-0410-b5e6-96231b3b80d8

Let llvm-upgrade upgrade bswap intrinsic.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35568 91177308-0d34-0410-b5e6-96231b3b80d8

For PR1297:
Make sure that the CTPOP result is casted to i32 as the bit counting
intrinsics all return i32 now (this affects CTLZ and CTTZ as well).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35567 91177308-0d34-0410-b5e6-96231b3b80d8