a5abba65b6
Fix PR1729: watch out for val# with no def.
...
llvm-svn: 42996
2007-10-15 18:33:50 +00:00
d6f7d44eae
Move CreateStackTemporary out to SelectionDAG
...
llvm-svn: 42995
2007-10-15 17:48:57 +00:00
9eb7a829e6
add a new CreateStackTemporary helper method.
...
llvm-svn: 42994
2007-10-15 17:47:20 +00:00
9d5b131e70
implement promotion of BR_CC operands, fixing bisort on ppc.
...
llvm-svn: 42992
2007-10-15 17:16:12 +00:00
8555e69def
updates from duncan
...
llvm-svn: 42991
2007-10-15 16:46:29 +00:00
bff4aea328
Achieve same result but use fewer lines of code.
...
llvm-svn: 42985
2007-10-15 15:31:35 +00:00
9130551996
Fast-track obviously over-large and over-small exponents during decimal->
...
integer conversion. In some such cases this makes us one or two orders
of magnitude faster than NetBSD's libc. Glibc seems to have a similar
fast path.
Also, tighten up some upper bounds to save a bit of memory.
llvm-svn: 42984
2007-10-15 15:00:55 +00:00
f6977d9842
Fix some typos. Call getTypeToTransformTo rather than
...
getTypeToExpandTo. The difference is that
getTypeToExpandTo gives the final result of expansion
(eg: i128 -> i32 on a 32 bit machine) while
getTypeToTransformTo does just one step (i128 -> i64).
llvm-svn: 42982
2007-10-15 13:30:18 +00:00
3cfb56d489
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally*
...
take a deleted nodes vector, instead of requiring it.
One more significant change: Implement the start of a legalizer that
just works on types. This legalizer is designed to run before the
operation legalizer and ensure just that the input dag is transformed
into an output dag whose operand and result types are all legal, even
if the operations on those types are not.
This design/impl has the following advantages:
1. When finished, this will *significantly* reduce the amount of code in
LegalizeDAG.cpp. It will remove all the code related to promotion and
expansion as well as splitting and scalarizing vectors.
2. The new code is very simple, idiomatic, and modular: unlike
LegalizeDAG.cpp, it has no 3000 line long functions. :)
3. The implementation is completely iterative instead of recursive, good
for hacking on large dags without blowing out your stack.
4. The implementation updates nodes in place when possible instead of
deallocating and reallocating the entire graph that points to some
mutated node.
5. The code nicely separates out handling of operations with invalid
results from operations with invalid operands, making some cases
simpler and easier to understand.
6. The new -debug-only=legalize-types option is very very handy :),
allowing you to easily understand what legalize types is doing.
This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is
enabled, this does nothing. However, this code is sufficient to legalize
all of the code in 186.crafty, olden and freebench on an x86 machine. The
biggest issues are:
1. Vectors aren't implemented at all yet
2. SoftFP is a mess, I need to talk to Evan about it.
3. No lowering to libcalls is implemented yet.
4. Various operations are missing etc.
5. There are FIXME's for stuff I hax0r'd out, like softfp.
Hey, at least it is a step in the right direction :). If you'd like to help,
just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If
this explodes it will tell you what needs to be implemented. Help is
certainly appreciated.
Once this goes in, we can do three things:
1. Add a new pass of dag combine between the "type legalizer" and "operation
legalizer" passes. This will let us catch some long-standing isel issues
that we miss because operation legalization often obfuscates the dag with
target-specific nodes.
2. We can rip out all of the type legalization code from LegalizeDAG.cpp,
making it much smaller and simpler. When that happens we can then
reimplement the core functionality left in it in a much more efficient and
non-recursive way.
3. Once the whole legalizer is non-recursive, we can implement whole-function
selectiondags maybe...
llvm-svn: 42981
2007-10-15 06:10:22 +00:00
b193517eed
One xform performed by LegalizeDAG is transformation of "store of fp" to "store of int".
...
Make two changes:
1) only xform "store of f32" if i32 is a legal type for the target.
2) only xform "store of f64" if either i64 or i32 are legal for the target.
3) if i64 isn't legal, manually lower to 2 stores of i32 instead of letting a
later pass of legalize do it. This is ugly, but helps future changes I'm
about to commit.
llvm-svn: 42980
2007-10-15 05:46:06 +00:00
2b827fd70d
avoid an APFloat copy.
...
llvm-svn: 42979
2007-10-15 05:34:10 +00:00
90e0b271df
Add a (disabled by default) way to view the ID of a node.
...
llvm-svn: 42978
2007-10-15 05:32:43 +00:00
207bd4d90e
Handle PPC long double in CBackend.
...
llvm-svn: 42972
2007-10-15 01:05:37 +00:00
fbbe570994
remove misleading comment.
...
llvm-svn: 42970
2007-10-14 20:35:12 +00:00
ebe491ea9c
If a target doesn't have HasMULHU or HasUMUL_LOHI, ExpandOp would return
...
without lo/hi set. Fall through to making a libcall instead.
llvm-svn: 42969
2007-10-14 18:35:05 +00:00
5fe658b21d
Consolidate logic for creating NaNs. Silence compiler warning.
...
llvm-svn: 42966
2007-10-14 10:39:51 +00:00
06077e7c3c
Whether arithmetic is supported is a property of the semantics. Make it
...
so, and clean up the checks by putting them in an inline function.
llvm-svn: 42965
2007-10-14 10:29:28 +00:00
4ed401b898
Separate out parsing of decimal number. Use this to only allocate
...
memory for the significand once up-front. Also ignore insignificant
trailing zeroes; this saves unnecessary multiplications later.
llvm-svn: 42964
2007-10-14 10:16:12 +00:00
4099f4f91a
Unbreak x86-64.
...
llvm-svn: 42962
2007-10-14 10:09:39 +00:00
8d6da9142c
When coalescing an EXTRACT_SUBREG and the dst register is a physical register,
...
the source register will be coalesced to the super register of the LHS. Properly
merge in the live ranges of the resulting coalesced interval that were part of
the original source interval to the live interval of the super-register.
llvm-svn: 42961
2007-10-14 10:08:34 +00:00
cdf3609130
Revert 42908 for now.
...
llvm-svn: 42960
2007-10-14 05:57:21 +00:00
2f6b6d6fb0
Fix type mismatch error in PPC Altivec (only causes
...
a problem when asserts are on). From vecLib.
llvm-svn: 42959
2007-10-14 01:58:32 +00:00
19db093b35
Disable some compile-time optimizations on PPC
...
long double.
llvm-svn: 42958
2007-10-14 01:56:47 +00:00
29af26f147
Clarify that fastcc has a problem with nested function
...
trampolines, rather than with nested functions themselves.
llvm-svn: 42955
2007-10-13 07:38:37 +00:00
f47e30627a
Enhance the truncstore optimization code to handle shifted
...
values and propagate demanded bits through them in simple cases.
This allows this code:
void foo(char *P) {
strcpy(P, "abc");
}
to compile to:
_foo:
ldrb r3, [r1]
ldrb r2, [r1, #+1]
ldrb r12, [r1, #+2]!
ldrb r1, [r1, #+1]
strb r1, [r0, #+3]
strb r2, [r0, #+1]
strb r12, [r0, #+2]
strb r3, [r0]
bx lr
instead of:
_foo:
ldrb r3, [r1, #+3]
ldrb r2, [r1, #+2]
orr r3, r2, r3, lsl #8
ldrb r2, [r1, #+1]
ldrb r1, [r1]
orr r2, r1, r2, lsl #8
orr r3, r2, r3, lsl #16
strb r3, [r0]
mov r2, r3, lsr #24
strb r2, [r0, #+3]
mov r2, r3, lsr #16
strb r2, [r0, #+2]
mov r3, r3, lsr #8
strb r3, [r0, #+1]
bx lr
testcase here: test/CodeGen/ARM/truncstore-dag-combine.ll
This also helps occasionally for X86 and other cases not involving
unaligned load/stores.
llvm-svn: 42954
2007-10-13 06:58:48 +00:00
5e6fe054a2
Add a simple optimization to simplify the input to
...
truncate and truncstore instructions, based on the
knowledge that they don't demand the top bits.
llvm-svn: 42952
2007-10-13 06:35:54 +00:00
c799fe9ed9
If the power of 5 is exact, and the reciprocal exact, the error is zero not one half-ulps. This prevents an infinite loop in rare cases.
...
llvm-svn: 42950
2007-10-13 03:34:08 +00:00
b63076504e
Local spiller optimization:
...
Turn this:
movswl %ax, %eax
movl %eax, -36(%ebp)
xorl %edi, -36(%ebp)
into
movswl %ax, %eax
xorl %edi, %eax
movl %eax, -36(%ebp)
by unfolding the load / store xorl into an xorl and a store when we know the
value in the spill slot is available in a register. This doesn't change the
number of instructions but reduce the number of times memory is accessed.
Also unfold some load folding instructions and reuse the value when similar
situation presents itself.
llvm-svn: 42947
2007-10-13 02:50:24 +00:00
7082dcf605
Change unfoldMemoryOperand(). User is now responsible for passing in the
...
register used by the unfolded instructions. User can also specify whether to
unfold the load, the store, or both.
llvm-svn: 42946
2007-10-13 02:35:06 +00:00
9490e0d078
Optionally create a MachineInstr without default implicit operands.
...
llvm-svn: 42945
2007-10-13 02:23:01 +00:00
e8d0bf2669
Correcting the corrections. Bad bad baaad emacs!
...
llvm-svn: 42935
2007-10-12 21:53:12 +00:00
1f0da1fefb
Corrected many typing errors. And removed 'nest' parameter handling
...
for fastcc from X86CallingConv.td. This means that nested functions
are not supported for calling convention 'fastcc'.
llvm-svn: 42934
2007-10-12 21:30:57 +00:00
371e6ca690
Dest type is always i8 *. This allows some simplification.
...
Do not filter memmove.
llvm-svn: 42930
2007-10-12 20:10:21 +00:00
a6286bd502
Due to the new tail call optimization, trampolines can no
...
longer be created for fastcc functions.
llvm-svn: 42925
2007-10-12 19:37:31 +00:00
61c574fc51
ppc long double. Implement fabs and fneg.
...
llvm-svn: 42924
2007-10-12 19:02:17 +00:00
409fa443fc
Update.
...
llvm-svn: 42922
2007-10-12 18:22:55 +00:00
ad618f66e6
Fix a bug in my patch last night that broke InstCombine/2007-10-12-Crash.ll
...
llvm-svn: 42920
2007-10-12 18:05:47 +00:00
a1a4a9ebfa
Implement i64->ppcf128 conversions.
...
llvm-svn: 42919
2007-10-12 17:52:03 +00:00
1410b8512c
Did mean to leave this in. INSERT_SUBREG isn't being coalesced yet.
...
llvm-svn: 42916
2007-10-12 17:16:50 +00:00
d502a82092
Remove duplicate comment.
...
llvm-svn: 42913
2007-10-12 16:05:57 +00:00
b93d90e98c
Implement correctly-rounded decimal->binary conversion, i.e. conversion
...
from user input strings.
Such conversions are more intricate and subtle than they may appear;
it is unlikely I have got it completely right first time. I would
appreciate being informed of any bugs and incorrect roundings you
might discover.
llvm-svn: 42912
2007-10-12 16:02:31 +00:00
e9dbe094aa
Remove a field that was never used.
...
llvm-svn: 42911
2007-10-12 15:35:10 +00:00
146fdb3eeb
If we're trying to be arbitrary precision, unsigned char clearly won't cut it. Needed for dec->bin conversions.
...
llvm-svn: 42910
2007-10-12 15:33:27 +00:00
7e74b17ad2
Don't attempt to mask no bits
...
llvm-svn: 42909
2007-10-12 15:31:31 +00:00
dc35bd79ca
Change the names used for internal labels to use the current
...
function symbol name instead of a codegen-assigned function
number.
Thanks Evan! :-)
llvm-svn: 42908
2007-10-12 14:53:36 +00:00
e3583817ac
Fix some corner cases with vectors in copyToRegs and copyFromRegs.
...
llvm-svn: 42907
2007-10-12 14:33:11 +00:00
4f056f3c10
Add support to SplitVectorOp for powi, where the second operand
...
is a scalar integer.
llvm-svn: 42906
2007-10-12 14:13:46 +00:00
8d978da3b0
Mark vector ctpop, cttz, and ctlz as Expand on x86.
...
llvm-svn: 42905
2007-10-12 14:09:42 +00:00
9013eaff9a
Mark vector pow, ctpop, cttz, and ctlz as Expand on PowerPC.
...
llvm-svn: 42904
2007-10-12 14:08:57 +00:00
11330f7526
Restrict EXTRACT_SUBREG coalescing to avoid negative performance impact.
...
llvm-svn: 42903
2007-10-12 09:15:53 +00:00
aa2d6ef81d
EXTRACT_SUBREG coalescing support. The coalescer now treats EXTRACT_SUBREG like
...
(almost) a register copy. However, it always coalesced to the register of the
RHS (the super-register). All uses of the result of a EXTRACT_SUBREG are sub-
register uses which adds subtle complications to load folding, spiller rewrite,
etc.
llvm-svn: 42899
2007-10-12 08:50:34 +00:00
89d5916921
Some clean up.
...
llvm-svn: 42898
2007-10-12 08:45:27 +00:00
09c0fe0a7f
Fold load / store into MOV32to32_ and MOV16to16_.
...
llvm-svn: 42895
2007-10-12 08:38:01 +00:00
f8c23f074b
Flag MOV32to32_ with EXTRACT_SUBREG. They should not be scheduled apart.
...
llvm-svn: 42894
2007-10-12 07:55:53 +00:00
5d8f7e0cc7
eliminate warning
...
llvm-svn: 42892
2007-10-12 07:44:54 +00:00
d8675e4915
Fix some 80 column violations.
...
Fix DecomposeSimpleLinearExpr to handle simple constants better.
Don't nuke gep(bitcast(allocation)) if the bitcast(allocation) will
fold the allocation. This fixes PR1728 and Instcombine/malloc3.ll
llvm-svn: 42891
2007-10-12 05:30:59 +00:00
05ff9e8cda
PPC long double. Implement a couple more conversions.
...
llvm-svn: 42888
2007-10-12 01:37:08 +00:00
be37007e64
Add intrinsics for sin, cos, and pow. These use llvm_anyfloat_ty, and so
...
may be overloaded with vector types. And add a testcase for codegen for
these.
llvm-svn: 42885
2007-10-12 00:01:22 +00:00
2a7de41682
Codegen support for vector intrinsics.
...
Factor out the code that expands the "nasty scalar code" for unrolling
vectors into a separate routine, teach it how to handle mixed
vector/scalar operands, as seen in powi, and use it for several operators,
including sin, cos, powi, and pow.
Add support in SplitVectorOp for fpow, fpowi and for several unary
operators.
llvm-svn: 42884
2007-10-11 23:57:53 +00:00
6472eb63c2
Implement ppc long double->uint conversion.
...
Make ppc long double constants print.
llvm-svn: 42882
2007-10-11 23:32:15 +00:00
482732af9d
Set ISD::FPOW to Expand.
...
llvm-svn: 42881
2007-10-11 23:21:31 +00:00
fd66486950
Add runtime library names for pow.
...
llvm-svn: 42880
2007-10-11 23:09:10 +00:00
daee002438
Add an ISD::FPOW node type.
...
llvm-svn: 42879
2007-10-11 23:06:37 +00:00
62f65edc32
Add missing argument to PALIGNR
...
llvm-svn: 42874
2007-10-11 20:58:37 +00:00
cb6832efe0
Fix an assertion abort on sparc. malloc(0) is allowed to
...
return NULL.
llvm-svn: 42871
2007-10-11 19:40:35 +00:00
9ccea99165
Added tail call optimization to the x86 back end. It can be
...
enabled by passing -tailcallopt to llc. The optimization is
performed if the following conditions are satisfied:
* caller/callee are fastcc
* elf/pic is disabled OR
elf/pic enabled + callee is in module + callee has
visibility protected or hidden
llvm-svn: 42870
2007-10-11 19:40:01 +00:00
007aa378ad
Next PPC long double bits. First cut at constants.
...
No compile-time support for constant operations yet,
just format transformations. Make readers and
writers work. Split constants into 2 doubles in
Legalize.
llvm-svn: 42865
2007-10-11 18:07:22 +00:00
899cc56612
Lower memcpy if it makes sense.
...
llvm-svn: 42864
2007-10-11 17:21:57 +00:00
d9c9c919da
Fix CodeGen/Generic/BasicInstrs.llx on sparc by marking divrem
...
illegal. Thanks to gabor for pointing this out!
llvm-svn: 42832
2007-10-10 18:10:57 +00:00
56ab90d3ad
Correct swapped arguments to getConstant.
...
llvm-svn: 42824
2007-10-10 09:54:50 +00:00
666323eacd
Next PPC long double bits: ppcf128->i32 conversion.
...
Surprisingly complicated.
Adds getTargetNode for 2 outputs, no inputs (missing).
llvm-svn: 42822
2007-10-10 01:01:31 +00:00
a9830a04eb
Bad choice of variable name.
...
llvm-svn: 42821
2007-10-10 00:11:40 +00:00
ad55a6079a
Fix an extremely stupid bug that prevented first round of coalescing (physical registers only) from happening.
...
llvm-svn: 42820
2007-10-09 23:36:27 +00:00
f5ea386698
Add new MemoryBuffer::getMemBufferCopy method.
...
llvm-svn: 42815
2007-10-09 21:46:38 +00:00
2af23f976b
Do not walk invalid iterator.
...
llvm-svn: 42812
2007-10-09 21:31:36 +00:00
66966403ce
Pass argc by value, not by reference, since it isn't modified.
...
llvm-svn: 42788
2007-10-09 16:04:57 +00:00
e8c8ef5234
LowerIntegerDivOrRem no longer exists.
...
llvm-svn: 42787
2007-10-09 15:45:13 +00:00
51554bf30e
Fix grammar in a comment.
...
llvm-svn: 42786
2007-10-09 15:44:37 +00:00
6d28778bfd
This is done.
...
llvm-svn: 42785
2007-10-09 15:42:21 +00:00
82bc90ac60
Under 64-bit mode use LEA64_32r instead of LEA64r to save a byte.
...
llvm-svn: 42783
2007-10-09 07:14:53 +00:00
5c50f59690
Change a #include into a forward declaration
...
llvm-svn: 42781
2007-10-09 03:40:30 +00:00
66d9572ced
Position Independent Code (PIC) support [3]
...
llvm-svn: 42780
2007-10-09 03:15:11 +00:00
35d86e60b6
Position Independent Code (PIC) support [2]
...
- Added a function to hold the stack location
where GP must be stored during LowerCALL
- AsmPrinter now emits directives based on
relocation type
- PIC_ set to default relocation type (same as GCC)
llvm-svn: 42779
2007-10-09 03:01:19 +00:00
5cef9cfd09
Position Independent Code (PIC) support [1]
...
- Modified instruction format to handle pseudo instructions
- Added LoadAddr SDNode to load symbols.
llvm-svn: 42778
2007-10-09 02:55:31 +00:00
f5ec10b64c
Bug fix. X86 was emitting redundant setcc and test instructions before a conditional move.
...
llvm-svn: 42774
2007-10-08 22:16:29 +00:00
5942e5a5fb
Call getFunctionNumber() instead of referencing FunctionNumber directly,
...
for consistency.
llvm-svn: 42769
2007-10-08 21:27:12 +00:00
a160361c85
Migrate X86 and ARM from using X86ISD::{,I}DIV and ARMISD::MULHILO{U,S} to
...
use ISD::{S,U}DIVREM and ISD::{S,U}MUL_HIO. Move the lowering code
associated with these operators into target-independent in LegalizeDAG.cpp
and TargetLowering.cpp.
llvm-svn: 42762
2007-10-08 18:33:35 +00:00
18109c88c3
Allow x86 compare to be commutable by default.
...
llvm-svn: 42761
2007-10-08 18:27:46 +00:00
44dd8fbd26
C and Objective Caml bindings for PHINode::addIncoming etc.
...
llvm-svn: 42760
2007-10-08 18:14:39 +00:00
5c6d0c3b99
DAGCombiner support for UDIVREM/SDIVREM and UMUL_LOHI/SMUL_LOHI.
...
Check if one of the two results unneeded so see if a simpler operator
could bs used. Also check to see if each of the two computations could be
simplified if they were split into separate operators. Factor out the code
that calls visit() so that it can be used for this purpose.
llvm-svn: 42759
2007-10-08 17:57:15 +00:00
acc8b68e5f
Remove debugging printf to get build going on x86-64.
...
llvm-svn: 42757
2007-10-08 17:49:43 +00:00
71f0d7d76e
When we start enabling SMUL_LOHI/UMUL_LOHI or SDIVREM/UDIVREM in
...
target-indepenent lowering, don't use them on PowerPC.
llvm-svn: 42755
2007-10-08 17:28:24 +00:00
b08c8bfe41
Add convenience overloads of SelectionDAG::getNode that take a SDVTList
...
and individual SDOperand operands.
llvm-svn: 42753
2007-10-08 15:49:58 +00:00
2c6a821fd7
Move the space in overview output for commands out of each of the
...
commands and into the common code.
llvm-svn: 42752
2007-10-08 15:45:12 +00:00
afedc063ab
Simplify getIntPtrType, allowing it to work for arbitrary pointer sizes.
...
llvm-svn: 42751
2007-10-08 15:16:25 +00:00
5e79a6108f
Use correct parentheses with the '&& "..."' idiom in an assert.
...
llvm-svn: 42750
2007-10-08 15:13:30 +00:00
fadf40a655
In -debug mode, dump SelectionDAGs both before and after the
...
optimization passes.
llvm-svn: 42749
2007-10-08 15:12:17 +00:00
13ab93efde
Add explicit keywords.
...
llvm-svn: 42747
2007-10-08 15:08:41 +00:00
49c6aab7b8
Use APInt::tcExtract. It's cleaner, and works :)
...
llvm-svn: 42746
2007-10-08 14:39:42 +00:00
b618216e56
Add a new function tcExtract for extracting a bignum from an
...
arbitrary range of bits embedded in the middle of another bignum.
This kind of operation is desirable in many cases of software
floating point, e.g. converting bignum integers to floating point
numbers of fixed precision (you want to extract the precision most
significant bits).
Elsewhere, add an assertion, and exit the shift functions early if
the shift count is zero.
llvm-svn: 42745
2007-10-08 13:47:12 +00:00
74529b7d5f
Begin the process of allowing DomTree on MBB's. Step One: template DomTreeNode by making it a typedef of a templated DomTreeNodeBase.
...
llvm-svn: 42743
2007-10-08 07:44:39 +00:00
21a58a72c5
Kill cycle of an live range is always the last use index + 1.
...
llvm-svn: 42742
2007-10-08 06:59:30 +00:00
783f7bbabd
C and Objective Caml bindings for getFunction and getNamedGlobal. Also enhanced
...
the Objective Caml 'declare_*' functions to behave more or less like
getOrInsertFunction.
llvm-svn: 42740
2007-10-08 03:45:09 +00:00
b20757d578
disable this entirely: it is causing use of invalidated iterators and infinite looping.
...
llvm-svn: 42739
2007-10-07 22:00:31 +00:00
8dd66ab3b2
Fix many regressions on x86 by avoiding dereferencing the end iterator.
...
llvm-svn: 42738
2007-10-07 21:53:12 +00:00
751ebf7b9a
C and Objective Caml bindings for GlobalVariable::isConstant.
...
llvm-svn: 42736
2007-10-07 17:31:42 +00:00
03f58ab706
Add back convertFromSignExtendedInteger.
...
llvm-svn: 42735
2007-10-07 12:15:41 +00:00
ba205229e7
Now that convertFromUnsignedParts has a sane, constant interface,
...
convertFromZeroExtendedInteger can be simplified as it doesn't need
to make a copy of the source bignum.
llvm-svn: 42734
2007-10-07 12:10:57 +00:00
6c1c8588ae
Reimplement convertFromUnsignedInteger so it is passed a const bignum.
...
It used to modify its argument in-place.
This interface is saner and the implementation more efficient. It will
be needed for decimal->binary conversion.
llvm-svn: 42733
2007-10-07 12:07:53 +00:00
5f00973393
convertFromInteger, as originally written, expected sign-extended
...
input. APInt unfortunately zero-extends signed integers, so Dale
modified the function to expect zero-extended input. Make this
assumption explicit in the function name.
llvm-svn: 42732
2007-10-07 11:45:55 +00:00
d3985924f4
combineLostFractions does not need to be a member function
...
llvm-svn: 42729
2007-10-07 08:51:21 +00:00
ffb4876cfb
C and Objective Caml bindings for PATypeHolder.
...
llvm-svn: 42713
2007-10-07 00:13:35 +00:00
c3d661a0ee
Bindings for the verifier.
...
llvm-svn: 42707
2007-10-06 21:00:36 +00:00
67ac2de8bf
Oops, I really wanted to commit this part also :)
...
llvm-svn: 42700
2007-10-06 16:39:43 +00:00
c59496f737
Move merge code into new helper function.
...
llvm-svn: 42699
2007-10-06 16:17:49 +00:00
ed7beaa3fa
Adopting a uniform naming convention for type constructors in bindings.
...
llvm-svn: 42698
2007-10-06 16:05:20 +00:00
1046c73d12
Adopting a uniform naming convention for constant constructors in bindings.
...
llvm-svn: 42697
2007-10-06 15:11:06 +00:00
7ce3176696
Adding C and Ocaml bindings for ConstantExpr.
...
llvm-svn: 42696
2007-10-06 14:29:36 +00:00
0de312dd7d
Reapply 42677.
...
llvm-svn: 42692
2007-10-06 08:19:55 +00:00
32897f5eb0
Cleaner, more general exponent output.
...
llvm-svn: 42690
2007-10-06 07:29:25 +00:00
2a9bcb9ea7
silence warnings in no-assert build.
...
llvm-svn: 42687
2007-10-06 06:13:42 +00:00
82217bd155
revert evan's patch until the header is committed
...
llvm-svn: 42686
2007-10-06 06:08:17 +00:00
f4b5d491df
Added DAG xforms. e.g.
...
(vextract (v4f32 s2v (f32 load $addr)), 0) -> (f32 load $addr)
(vextract (v4i32 bc (v4f32 s2v (f32 load $addr))), 0) -> (i32 load $addr)
Remove x86 specific patterns.
llvm-svn: 42677
2007-10-06 02:46:29 +00:00
f864ac96d8
Next powerpc long double bits. Comparisons work,
...
although not well, and shortening FP converts.
llvm-svn: 42672
2007-10-06 01:24:11 +00:00
c8b650ab14
Fix and clarify some comments.
...
llvm-svn: 42670
2007-10-06 00:43:45 +00:00
0ea72a9a91
Generalize tcFullMultiply so that the operands can be of differing
...
part widths. Also, return the number of parts actually required to
hold the result's value.
Remove an over-cautious condition from rounding of float->hex conversion.
llvm-svn: 42669
2007-10-06 00:24:48 +00:00
1d0d24c4fd
Wrapping Value::dump.
...
llvm-svn: 42668
2007-10-06 00:08:49 +00:00
7330acdec7
Removing the silly CHelpers header by rolling wrap and unwrap into the C
...
bindings headers themselves, hidden behind #ifdef __cplusplus.
llvm-svn: 42666
2007-10-05 23:59:36 +00:00
1151ffde70
Commute x86 cmove instructions by swapping the operands and change the condition
...
to its inverse.
Testing this as llcbeta
llvm-svn: 42661
2007-10-05 23:13:21 +00:00
42a13757de
This is done.
...
llvm-svn: 42656
2007-10-05 22:34:59 +00:00
484cab7a2f
Enable convertToThreeAddress for X86 by default.
...
llvm-svn: 42655
2007-10-05 22:31:10 +00:00
a69f987b66
Fix bug in updating dominance frontier after loop
...
unswitch when frontier includes basic blocks that
are not inside loop.
llvm-svn: 42654
2007-10-05 22:29:34 +00:00
d3ccf00870
INC64_32r -> LEA64_32r is better than INC64_32r -> LEA32r, but it still can
...
cause performance degradation.
llvm-svn: 42653
2007-10-05 21:55:32 +00:00
fa2c828687
In 64-bit mode, avoid using leal with 32-bit 32-bit address size, e.g.
...
leal 1(%ecx), %edi, which requires 67H prefix.
llvm-svn: 42647
2007-10-05 20:34:26 +00:00
c0154c06d6
First round of ppc long double. call/return and
...
basic arithmetic works.
Rename RTLIB long double functions to distinguish
different flavors of long double; the lib functions
have different names, alas.
llvm-svn: 42644
2007-10-05 20:04:43 +00:00
aac0f8e351
Add support to convert more 64-bit instructions to 3-address instructions.
...
llvm-svn: 42642
2007-10-05 18:20:36 +00:00
97eba74a52
ADC and SBB uses EFLAGS.
...
llvm-svn: 42640
2007-10-05 17:59:57 +00:00
43c29dce18
Change a few more spaces to tabs in assembly output.
...
llvm-svn: 42638
2007-10-05 15:58:41 +00:00
b074f23dff
Change a space to a tab in the assembly output of a .globl directive
...
for consistency.
llvm-svn: 42637
2007-10-05 15:54:58 +00:00
12334acbfb
Legalize support for MUL_LOHI and DIVREM.
...
llvm-svn: 42636
2007-10-05 14:17:22 +00:00
2682bb6df2
Fix a typo in a comment.
...
llvm-svn: 42635
2007-10-05 14:11:58 +00:00
1a77dfba15
Provide names for MUL_LOHI and DIVREM operators.
...
llvm-svn: 42634
2007-10-05 14:11:04 +00:00
a8a9c15e30
Testing convertToThreeeAddress as X86 llcbeta.
...
llvm-svn: 42630
2007-10-05 08:04:01 +00:00
84d0ebc10a
Chain producing nodes cannot be moved, not chain reading nodes.
...
llvm-svn: 42627
2007-10-05 01:42:35 +00:00
991cf47221
Oops. Didn't mean to leave this in.
...
llvm-svn: 42626
2007-10-05 01:39:40 +00:00
79e9713b11
If a node that defines a physical register that is expensive to copy. The
...
scheduler will try a number of tricks in order to avoid generating the
copies. This may not be possible in case the node produces a chain value
that prevent movement. Try unfolding the load from the node before to allow
it to be moved / cloned.
llvm-svn: 42625
2007-10-05 01:39:18 +00:00
a851e2b92e
Added storeRegToAddr, loadRegFromAddr, and unfoldMemoryOperand's.
...
llvm-svn: 42624
2007-10-05 01:34:55 +00:00
6912b50958
Not needed any more.
...
llvm-svn: 42623
2007-10-05 01:34:14 +00:00
1ef9edce5f
Forgot these.
...
llvm-svn: 42622
2007-10-05 01:33:45 +00:00
f4cf5dcdd2
- Added a few target hooks to generate load / store instructions from / to any
...
address (not just from / to frameindexes).
- Added target hooks to unfold load / store instructions / SDNodes into separate
load, data processing, store instructions / SDNodes.
llvm-svn: 42621
2007-10-05 01:32:41 +00:00
4852303bdb
Add a variant of getTargetNode() that takes a vector of MVT::ValueType.
...
llvm-svn: 42620
2007-10-05 01:10:49 +00:00
fd11ef4665
Silence a warning.
...
llvm-svn: 42619
2007-10-05 01:09:32 +00:00
1f2b5f0e13
add a note.
...
llvm-svn: 42607
2007-10-04 15:47:27 +00:00
8f1946fa22
Add APFloat -> hexadecimal string conversion, as per %a and %A in C99.
...
Useful for diagnostics and debugging.
llvm-svn: 42598
2007-10-03 22:26:02 +00:00
8313e75ea7
Completely merge the implementation details of DomTree and PostDomTree.
...
Also, add a FIXME for a bug in PostDomTree calculation I noticed while writing this,
llvm-svn: 42593
2007-10-03 21:25:45 +00:00
3574759d85
Fix 80 col violation.
...
llvm-svn: 42591
2007-10-03 21:17:43 +00:00
e192e32577
Refactor code in a separate method.
...
llvm-svn: 42590
2007-10-03 21:16:08 +00:00
99f6ab7e4c
Add initial iterator support for folding set.
...
llvm-svn: 42589
2007-10-03 21:12:09 +00:00
8c41ed61d1
Simplify implementation of the FoldingSet circular list, a necessary step
...
to giving it iterators.
llvm-svn: 42586
2007-10-03 20:45:43 +00:00
c731c97fac
Use empty() member functions when that's what's being tested for instead
...
of comparing begin() and end().
llvm-svn: 42585
2007-10-03 19:26:29 +00:00
8c43e41d9b
Fix a using namespace llvm; in a header file.
...
llvm-svn: 42584
2007-10-03 19:04:09 +00:00
9d559cfff5
Tone down an overzealous optimization.
...
llvm-svn: 42582
2007-10-03 17:45:27 +00:00
4bdb84fe53
add a note
...
llvm-svn: 42579
2007-10-03 17:10:03 +00:00
30f65fe4a7
Fix PR1719, by not marking llvm.global.annotations internal.
...
llvm-svn: 42578
2007-10-03 17:05:40 +00:00
1ca1f80b38
Tweak RoundAwayFromZero the bit number below which is truncated, and make
...
it const.
Preparation for APFloat -> hexadecimal string conversion.
llvm-svn: 42576
2007-10-03 15:16:41 +00:00
4d4e9cdb62
add a note
...
llvm-svn: 42573
2007-10-03 06:10:59 +00:00
d66e0cd6c0
Fix PR1719, by not marking llvm.noinline internal.
...
llvm-svn: 42565
2007-10-03 03:59:15 +00:00
21ba176c4b
Bill's example is still not enough to repro this, but it has other issues that
...
seem significant as well.
llvm-svn: 42564
2007-10-03 03:40:24 +00:00
b60f254975
Factor some code from the DomTree and PostDomTree calculate methods up into
...
each one's runOnFunction method.
llvm-svn: 42563
2007-10-03 03:20:17 +00:00
3efc0758ae
Another micro-opt.
...
llvm-svn: 42554
2007-10-02 21:49:31 +00:00
f214ff8701
Another missed optimization with LICM.
...
llvm-svn: 42552
2007-10-02 21:43:06 +00:00
855011e5c6
Small label changes.
...
llvm-svn: 42549
2007-10-02 21:02:53 +00:00
4eb7ca4b4c
Now with source code.
...
llvm-svn: 42548
2007-10-02 21:01:16 +00:00
96ed3bb2d4
Now with LL code!
...
llvm-svn: 42547
2007-10-02 20:54:32 +00:00
9c4d61b523
Another missed optimization.
...
llvm-svn: 42546
2007-10-02 20:42:59 +00:00
88ea107fdb
Micro-optimization -- missed LICM opportunity.
...
llvm-svn: 42542
2007-10-02 19:55:05 +00:00
4d4e77af8e
Rewrite sqrt and powi to use anyfloat. By popular demand.
...
llvm-svn: 42537
2007-10-02 17:43:59 +00:00
1f79ba6fe6
Refactor code to add load / store folded instructions -> register only
...
instructions reverse map.
llvm-svn: 42509
2007-10-01 23:44:33 +00:00
b6c05b1f90
Fix stride computations for long double arrays.
...
llvm-svn: 42508
2007-10-01 23:08:35 +00:00
9765cc3bbb
Move the code that emits the .file directives so that it runs after the
...
SourceFiles list is fully filled in so that it sees all of the files.
llvm-svn: 42506
2007-10-01 22:40:20 +00:00
a3a67596f6
Remove simple scheduler.
...
llvm-svn: 42499
2007-10-01 20:44:07 +00:00
2a60ff1aeb
Relax unsafe use check. If there is one unconditional use inside the loop then it is safe to promote value even if there is another conditional use inside the loop.
...
llvm-svn: 42493
2007-10-01 18:12:58 +00:00
a1b7e95039
Typo. X86comi doesn't read / write chain's.
...
llvm-svn: 42492
2007-10-01 18:12:48 +00:00
28e19a48ce
Add getABITypeSize, getABITypeSizeInBits
...
llvm-svn: 42488
2007-10-01 16:03:14 +00:00
c0855f8a88
remove dup comment
...
llvm-svn: 42486
2007-09-30 19:08:12 +00:00
9150652b21
Constant fold int-to-long-double conversions;
...
use APFloat for int-to-float/double; use
round-to-nearest for these (implementation-defined,
seems to match gcc).
llvm-svn: 42484
2007-09-30 18:19:03 +00:00
28a2c4a9c4
Simplify and fix signed int -> FP conversions.
...
llvm-svn: 42483
2007-09-30 18:17:01 +00:00
0b7cf862bc
AsmPrinters overriding getAnalysisUsage should call super.
...
And not super's super, either.
llvm-svn: 42482
2007-09-30 13:39:29 +00:00
46ade4eb9b
Template-ize more of the DomTree internal implementation details. Only the calculate() methods for DomTree and PostDomTree remain to be merged/template-ized.
...
llvm-svn: 42476
2007-09-30 04:17:16 +00:00
18a9273e61
Add a new DenseSet abstraction.
...
llvm-svn: 42474
2007-09-30 00:47:20 +00:00
f72ff0c379
Check that the second parameter of init_trampoline
...
can be resolved to a function. This is required
for codegen and used by instcombine.
llvm-svn: 42469
2007-09-29 16:25:54 +00:00
f5aa229ede
This is done already.
...
llvm-svn: 42467
2007-09-29 02:23:08 +00:00
37ca83d4e9
Collector is the base class for garbage collection code generators.
...
This version enhances the previous patch to add root initialization
as discussed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070910/053455.html
Collector gives its subclasses control over generic algorithms:
unsigned NeededSafePoints; //< Bitmask of required safe points.
bool CustomReadBarriers; //< Default is to insert loads.
bool CustomWriteBarriers; //< Default is to insert stores.
bool CustomRoots; //< Default is to pass through to backend.
bool InitRoots; //< If set, roots are nulled during lowering.
It also has callbacks which collectors can hook:
/// If any of the actions are set to Custom, this is expected to
/// be overriden to create a transform to lower those actions to
/// LLVM IR.
virtual Pass *createCustomLoweringPass() const;
/// beginAssembly/finishAssembly - Emit module metadata as
/// assembly code.
virtual void beginAssembly(Module &M, std::ostream &OS,
AsmPrinter &AP,
const TargetAsmInfo &TAI) const;
virtual void finishAssembly(Module &M,
CollectorModuleMetadata &CMM,
std::ostream &OS, AsmPrinter &AP,
const TargetAsmInfo &TAI) const;
Various other independent algorithms could be implemented, but were
not necessary for the initial two collectors. Some examples are
listed here:
http://llvm.org/docs/GarbageCollection.html#collector-algos
llvm-svn: 42466
2007-09-29 02:13:43 +00:00
6547e5a2bc
Demoting CHelpers.h to include/llvm/Support.
...
llvm-svn: 42465
2007-09-29 01:38:42 +00:00
342181c51d
Switch to densemap rather than std::set
...
llvm-svn: 42462
2007-09-29 00:50:40 +00:00
5fb5a1f389
Enabling new condition code modeling scheme.
...
llvm-svn: 42459
2007-09-29 00:00:36 +00:00
a90183e7d1
Teach SplitVectorOp how to split INSERT_VECTOR_ELT.
...
llvm-svn: 42457
2007-09-28 23:53:40 +00:00
a5e595d23a
If two instructions are both two-address code, favors (schedule closer to
...
terminator) the one that has a CopyToReg use. This fixes
2006-05-11-InstrSched.ll with -new-cc-modeling-scheme.
llvm-svn: 42453
2007-09-28 22:32:30 +00:00
8861abe6f2
really fix PR1581, thanks to Daniel Dunbar for pointing
...
this out.
llvm-svn: 42448
2007-09-28 20:50:50 +00:00
2be9ec5c91
Make Value::getNameLen not crash on an empty name, fixing PR1712.
...
llvm-svn: 42444
2007-09-28 20:09:40 +00:00
f72693f36e
Remove a poor scheduling heuristic.
...
llvm-svn: 42443
2007-09-28 19:37:35 +00:00
038dcc5136
Trim some unneeded fields.
...
llvm-svn: 42442
2007-09-28 19:24:24 +00:00
789b5a505b
Fix long double -> uint64 conversion.
...
llvm-svn: 42440
2007-09-28 18:44:17 +00:00
6bf69ed3cc
minor long double related changes
...
llvm-svn: 42439
2007-09-28 18:06:58 +00:00
25d506c41b
Make the checks for DW_FORM_data4 consistent with the others, and
...
add more such code for DIEDwarfLabel::SizeOf and DIEObjectLabel::SizeOf.
llvm-svn: 42435
2007-09-28 16:50:28 +00:00
0d23d63b9e
Use 32-bit data directives for DW_FORM_data4 format data, even on
...
targets with 64-bit addresses.
llvm-svn: 42434
2007-09-28 15:43:33 +00:00
6c04ac1db0
Refactor the memcpy lowering for the x86 target.
...
The only generated code difference is that now we call memcpy when
the size of the array is unknown. This matches GCC behavior and is
better since the run time value can be arbitrarily large.
llvm-svn: 42433
2007-09-28 12:53:01 +00:00
1f516560d1
Stop inventing new words. :-)
...
llvm-svn: 42429
2007-09-28 01:35:02 +00:00
d0d5c11d18
Remove unneeded #include.
...
llvm-svn: 42428
2007-09-28 01:25:07 +00:00
0fbef94cd0
Have PostDomTree use the newly templated DFSPass.
...
llvm-svn: 42427
2007-09-28 01:23:47 +00:00
edfc5b2204
Pessimisively assume ADJCALLSTACKDOWN / ADJCALLSTACKUP (which becomes sub / add) clobbers EFLAGS.
...
llvm-svn: 42426
2007-09-28 01:19:48 +00:00
25a00a63eb
Add sqrt and powi intrinsics for long double.
...
llvm-svn: 42423
2007-09-28 01:08:20 +00:00
1d1d0e7735
Don't do SRA for unions with long double fields.
...
Fixes a SWB crash.
llvm-svn: 42422
2007-09-28 00:21:38 +00:00
ee73e0b8cb
Convert DFSPass into a templated friend function, in preparation for making it common to DomTree and PostDomTree.
...
llvm-svn: 42420
2007-09-27 23:23:00 +00:00
a1d46c7d0a
TargetAsmInfo::getAddressSize() was incorrect for x86-64 and 64-bit targets
...
other than PPC64. Instead of fixing it, just remove it and fix all the
places that use it to use TargetData::getPointerSize() instead, as there
aren't very many. Most of the references were in DwarfWriter.cpp.
llvm-svn: 42419
2007-09-27 23:12:31 +00:00
613afce430
CollectorMetadata abstractly describes stack maps for a function.
...
It includes:
- location and of each safe point in machine code (identified by a
label)
- location of each root within the stack frame (identified by an
offset), including the metadata tag provided to llvm.gcroot in
the user program
- size of the stack frame (for collectors which want to cheat on
stack crawling :)
- and eventually will include liveness
It is to be populated by back-ends during code-generation.
CollectorModuleMetadata aggregates this information across the
entire module.
llvm-svn: 42418
2007-09-27 22:18:46 +00:00
99dc695da5
Use GR64 in 64-bit mode.
...
llvm-svn: 42417
2007-09-27 21:50:05 +00:00
658c57eb66
My previous Registry.h header, as well as Collectors.h, which is the
...
registry for dynamically-loaded garbage collection compiler plugins.
llvm-svn: 42415
2007-09-27 19:34:27 +00:00
5a71402be6
Doh. Calls clobber EFLAGS.
...
llvm-svn: 42413
2007-09-27 19:01:55 +00:00
e6f92253f5
Avoid inserting a live register more than once.
...
llvm-svn: 42410
2007-09-27 18:46:06 +00:00
a447de7065
Ignore redundant constraints
...
llvm-svn: 42397
2007-09-27 15:42:23 +00:00
3934961878
Build the correct range for loops with unusual bounds. Fix from Jay Foad.
...
llvm-svn: 42394
2007-09-27 14:12:54 +00:00
75439b3b78
Silence a compiler warning.
...
llvm-svn: 42389
2007-09-27 07:35:39 +00:00
bde499be60
Boogs.
...
llvm-svn: 42388
2007-09-27 07:29:27 +00:00
1ec79b41db
Be smarter about which node to force schedule. Reduce # of duplications + copies; Added statistics.
...
llvm-svn: 42387
2007-09-27 07:09:03 +00:00
cfd5f82890
Backtracking only when it won't create a cycle.
...
llvm-svn: 42384
2007-09-27 00:25:29 +00:00
34aa41c475
Make temporaries explicit to avoid premature
...
destruction of compiler-created ones.
llvm-svn: 42383
2007-09-26 23:20:33 +00:00
8e136a9dc4
- Move getPhysicalRegisterRegClass() from ScheduleDAG to MRegisterInfo.
...
- Added ability to emit cross class register copies to the BBRU scheduler.
- More aggressive backtracking.
llvm-svn: 42375
2007-09-26 21:36:17 +00:00
9acbf5aaf9
Whitespace and compiler warning cleanup.
...
llvm-svn: 42373
2007-09-26 21:33:42 +00:00
8728c3376a
- Added MRegisterInfo::getCrossCopyRegClass() hook. For register classes where reg to reg copies are not possible, this returns another register class which registers in the specified register class can be copied to (and copy back from).
...
- X86 copyRegToReg() now supports copying between EFLAGS and GR32 / GR64 registers.
llvm-svn: 42372
2007-09-26 21:31:07 +00:00
b93de587cb
Some assemblers do not recognize aliases pushfd, pushfq, popfd, and popfq. Just emit them as pushf and popf.
...
llvm-svn: 42371
2007-09-26 21:28:00 +00:00
b6d56401aa
Enable codegen for long double abs, sin, cos
...
llvm-svn: 42368
2007-09-26 21:10:55 +00:00
c23b66c58b
Added C and Ocaml bindings for functions, basic blocks, and
...
instruction creation. No support yet for instruction introspection.
Also eliminated allocas from the Ocaml bindings for portability,
and avoided unnecessary casts.
llvm-svn: 42367
2007-09-26 20:56:12 +00:00
f04d37d3a9
Fix f80 UNDEF.
...
llvm-svn: 42359
2007-09-26 17:26:49 +00:00
b4b352656a
Typos: POPQ -> POPFQ, POPD -> POPFD.
...
llvm-svn: 42348
2007-09-26 06:38:29 +00:00
c9e7b8ec50
move PR1160 here.
...
llvm-svn: 42347
2007-09-26 06:29:31 +00:00
c1e4e3743b
Allow copyRegToReg to emit cross register classes copies.
...
Tested with "make check"!
llvm-svn: 42346
2007-09-26 06:25:56 +00:00
fef69f5b4a
move PR1264 here.
...
llvm-svn: 42345
2007-09-26 06:15:48 +00:00
0a6f47cff9
Add pushf{d|q}, popf{d|q} to push and pop EFLAGS register.
...
llvm-svn: 42335
2007-09-26 01:29:06 +00:00
9b7f0e6eb4
translateX86CC updates the last two operands.
...
llvm-svn: 42333
2007-09-26 00:45:55 +00:00
e291f727e3
Correctly restore stack pointer after realignment in main() on Cygwin/Mingw32
...
llvm-svn: 42332
2007-09-26 00:13:34 +00:00
17663f40b9
Remove no-longer-used variable.
...
llvm-svn: 42329
2007-09-25 23:32:20 +00:00
e7439eff5a
Make APFloat->int conversions deterministic even in
...
cases with undefined behavior.
llvm-svn: 42328
2007-09-25 23:07:07 +00:00
5321fa44f4
Missing load / store folding entries.
...
llvm-svn: 42323
2007-09-25 22:10:43 +00:00
90910745bb
Partly revert invalid r41774
...
llvm-svn: 42322
2007-09-25 21:52:30 +00:00
57211c5550
More explicit keywords.
...
llvm-svn: 42316
2007-09-25 20:27:06 +00:00
384f4bd878
Add a newline to the end of this file.
...
llvm-svn: 42314
2007-09-25 20:08:48 +00:00
06919e8ef2
Fix a typo in a comment.
...
llvm-svn: 42313
2007-09-25 19:37:26 +00:00
8ee1ecfc50
New style x87 cmp instructions.
...
llvm-svn: 42312
2007-09-25 19:08:02 +00:00
7bba386f72
Handle multiple induction variables.
...
This fixes PR714.
llvm-svn: 42309
2007-09-25 18:24:48 +00:00
Evan Cheng