- FCMOV only supports a subset of X86 conditions. Skip boolean
simplification if X86 condition is not valid for FCMOV.
- add a minimal test case for PR13577.
llvm-svn: 161732
- if a boolean test (X86ISD::CMP or X86ISD:SUB) checks a boolean value
generated from X86ISD::SETCC, try to simplify the boolean value
generation and checking by reusing the original EFLAGS with proper
condition code
- add hooks to X86 specific SETCC/BRCOND/CMOV, the major 3 places
consuming EFLAGS
part of patches fixing PR12312
llvm-svn: 161687
We perform the following:
1> Use SUB instead of CMP for i8,i16,i32 and i64 in ISel lowering.
2> Modify MachineCSE to correctly handle implicit defs.
3> Convert SUB back to CMP if possible at peephole.
Removed pattern matching of (a>b) ? (a-b):0 and like, since they are handled
by peephole now.
rdar://11873276
llvm-svn: 161462
Fast isel doesn't currently have support for translating builtin function
calls to target instructions. For embedded environments where the library
functions are not available, this is a matter of correctness and not
just optimization. Most of this patch is just arranging to make the
TargetLibraryInfo available in fast isel. <rdar://problem/12008746>
llvm-svn: 161232
large immediates. Add dag combine logic to recover in case the large
immediates doesn't fit in cmp immediate operand field.
int foo(unsigned long l) {
return (l>> 47) == 1;
}
we produce
%shr.mask = and i64 %l, -140737488355328
%cmp = icmp eq i64 %shr.mask, 140737488355328
%conv = zext i1 %cmp to i32
ret i32 %conv
which codegens to
movq $0xffff800000000000,%rax
andq %rdi,%rax
movq $0x0000800000000000,%rcx
cmpq %rcx,%rax
sete %al
movzbl %al,%eax
ret
TargetLowering::SimplifySetCC would transform
(X & -256) == 256 -> (X >> 8) == 1
if the immediate fails the isLegalICmpImmediate() test. For x86,
that's immediates which are not a signed 32-bit immediate.
Based on a patch by Eli Friedman.
PR10328
rdar://9758774
llvm-svn: 160346
uint32_t hi(uint64_t res)
{
uint_32t hi = res >> 32;
return !hi;
}
llvm IR looks like this:
define i32 @hi(i64 %res) nounwind uwtable ssp {
entry:
%lnot = icmp ult i64 %res, 4294967296
%lnot.ext = zext i1 %lnot to i32
ret i32 %lnot.ext
}
The optimizer has optimize away the right shift and truncate but the resulting
constant is too large to fit in the 32-bit immediate field. The resulting x86
code is worse as a result:
movabsq $4294967296, %rax ## imm = 0x100000000
cmpq %rax, %rdi
sbbl %eax, %eax
andl $1, %eax
This patch teaches the x86 lowering code to handle ult against a large immediate
with trailing zeros. It will issue a right shift and a truncate followed by
a comparison against a shifted immediate.
shrq $32, %rdi
testl %edi, %edi
sete %al
movzbl %al, %eax
It also handles a ugt comparison against a large immediate with trailing bits
set. i.e. X > 0x0ffffffff -> (X >> 32) >= 1
rdar://11866926
llvm-svn: 160312
multiple scalars and insert them into a vector. Next, we shuffle the elements
into the correct places, as before.
Also fix a small dagcombine bug in SimplifyBinOpWithSameOpcodeHands, when the
migration of bitcasts happened too late in the SelectionDAG process.
llvm-svn: 159991
The CopyToReg nodes that set up the argument registers before a call
must be glued to the call instruction. Otherwise, the scheduler may emit
the physreg copies long before the call, causing long live ranges for
the fixed registers.
Besides disabling good register allocation, that can also expose
problems when EmitInstrWithCustomInserter() splits a basic block during
the live range of a physreg.
llvm-svn: 159721
Before this patch in pic 32 bit code we would add the global base register
and not load from that address. This is a really old bug, but before the
introduction of the tls attributes we would never select initial exec for
pic code.
llvm-svn: 159409
The function live-out registers must be live at all function returns,
and %RCX is only used by eh.return. When a function also has a normal
return, only %RAX holds a return value.
This fixes PR13188.
llvm-svn: 159116
TargetLoweringObjectFileELF. Use this to support it on X86. Unlike ARM,
on X86 it is not easy to find out if .init_array should be used or not, so
the decision is made via TargetOptions and defaults to off.
Add a command line option to llc that enables it.
llvm-svn: 158692
This patch will generate the following for integer ABS:
movl %edi, %eax
negl %eax
cmovll %edi, %eax
INSTEAD OF
movl %edi, %ecx
sarl $31, %ecx
leal (%rdi,%rcx), %eax
xorl %ecx, %eax
There exists a target-independent DAG combine for integer ABS, which converts
integer ABS to sar+add+xor. For X86, we match this pattern back to neg+cmov.
This is implemented in PerformXorCombine.
rdar://10695237
llvm-svn: 158175
This patch will optimize the following
movq %rdi, %rax
subq %rsi, %rax
cmovsq %rsi, %rdi
movq %rdi, %rax
to
cmpq %rsi, %rdi
cmovsq %rsi, %rdi
movq %rdi, %rax
Perform this optimization if the actual result of SUB is not used.
rdar: 11540023
llvm-svn: 158126
This implements codegen support for accesses to thread-local variables
using the local-dynamic model, and adds a clean-up pass so that the base
address for the TLS block can be re-used between local-dynamic access on
an execution path.
llvm-svn: 157818
to pass around a struct instead of a large set of individual values. This
cleans up the interface and allows more information to be added to the struct
for future targets without requiring changes to each and every target.
NV_CONTRIB
llvm-svn: 157479
This patch will optimize -(x != 0) on X86
FROM
cmpl $0x01,%edi
sbbl %eax,%eax
notl %eax
TO
negl %edi
sbbl %eax %eax
In order to generate negl, I added patterns in Target/X86/X86InstrCompiler.td:
def : Pat<(X86sub_flag 0, GR32:$src), (NEG32r GR32:$src)>;
rdar: 10961709
llvm-svn: 156312
This will be used to determine whether it's profitable to turn a select into a
branch when the branch is likely to be predicted.
Currently enabled for everything but Atom on X86 and Cortex-A9 devices on ARM.
I'm not entirely happy with the name of this flag, suggestions welcome ;)
llvm-svn: 156233
This patch will optimize the following cases on X86
(a > b) ? (a-b) : 0
(a >= b) ? (a-b) : 0
(b < a) ? (a-b) : 0
(b <= a) ? (a-b) : 0
FROM
movl %edi, %ecx
subl %esi, %ecx
cmpl %edi, %esi
movl $0, %eax
cmovll %ecx, %eax
TO
xorl %eax, %eax
subl %esi, %edi
cmovll %eax, %edi
movl %edi, %eax
rdar: 10734411
llvm-svn: 155919
x == -y --> x+y == 0
x != -y --> x+y != 0
On x86, the generated code goes from
negl %esi
cmpl %esi, %edi
je .LBB0_2
to
addl %esi, %edi
je .L4
This case is correctly handled for ARM with "cmn".
Patch by Manman Ren.
rdar://11245199
PR12545
llvm-svn: 155739
* Model FPSW (the FPU status word) as a register.
* Add ISel patterns for the FUCOM*, FNSTSW and SAHF instructions.
* During Legalize/Lowering, build a node sequence to transfer the comparison
result from FPSW into EFLAGS. If you're wondering about the right-shift: That's
an implicit sub-register extraction (%ax -> %ah) which is handled later on by
the instruction selector.
Fixes PR6679. Patch by Christoph Erhardt!
llvm-svn: 155704
immediate. We can't use it here because the shuffle code does not check that
the lower part of the word is identical to the upper part.
llvm-svn: 155440
Original message:
Modify the code that lowers shuffles to blends from using blendvXX to vblendXX.
blendV uses a register for the selection while Vblend uses an immediate.
On sandybridge they still have the same latency and execute on the same execution ports.
llvm-svn: 154483
blendv uses a register for the selection while vblend uses an immediate.
On sandybridge they still have the same latency and execute on the same execution ports.
llvm-svn: 154396
legalizer always use the DAG entry node. This is wrong when the libcall is
emitted as a tail call since it effectively folds the return node. If
the return node's input chain is not the entry (i.e. call, load, or store)
use that as the tail call input chain.
PR12419
rdar://9770785
rdar://11195178
llvm-svn: 154370
in TargetLowering. There was already a FIXME about this location being
odd. The interface is simplified as a consequence. This will also make
it easier to change TLS models when compiling with PIE.
llvm-svn: 154292
Previously we used three instructions to broadcast an immediate value into a
vector register.
On Sandybridge we continue to load the broadcasted value from the constant pool.
llvm-svn: 154284
This allows us to keep passing reduced masks to SimplifyDemandedBits, but
know about all the bits if SimplifyDemandedBits fails. This allows instcombine
to simplify cases like the one in the included testcase.
llvm-svn: 154011
Specifically, remove the magic number when checking to see if the copy has a
glue operand and simplify the checking logic.
rdar://10930395
llvm-svn: 152041
In this instance we are generating the tail-call during legalizeDAG. The 2nd
floor call can't be a tail call because it clobbers %xmm1, which is defined by
the first floor call. The first floor call can't be a tail-call because it's
not in the tail position. The only reasonable way I could think to fix this
in a target-independent manner was to check for glue logic on the copy reg.
rdar://10930395
llvm-svn: 151877
the processor keeps a return addresses stack (RAS) which stores the address
and the instruction execution state of the instruction after a function-call
type branch instruction.
Calling a "noreturn" function with normal call instructions (e.g. bl) can
corrupt RAS and causes 100% return misprediction so LLVM should use a
unconditional branch instead. i.e.
mov lr, pc
b _foo
The "mov lr, pc" is issued in order to get proper backtrace.
rdar://8979299
llvm-svn: 151623
[Joe Groff] Hi everyone. My previous patch applied as r151382 had a few problems:
Clang raised a warning, and X86 LowerOperation would assert out for
fptoui f64 to i32 because it improperly lowered to an illegal
BUILD_PAIR. Here's a patch that addresses these issues. Let me know if
any other changes are necessary. Thanks.
llvm-svn: 151432
Call instructions no longer have a list of 43 call-clobbered registers.
Instead, they get a single register mask operand with a bit vector of
call-preserved registers.
This saves a lot of memory, 42 x 32 bytes = 1344 bytes per call
instruction, and it speeds up building call instructions because those
43 imp-def operands no longer need to be added to use-def lists. (And
removed and shifted and re-added for every explicit call operand).
Passes like LiveVariables, LiveIntervals, RAGreedy, PEI, and
BranchFolding are significantly faster because they can deal with call
clobbers in bulk.
Overall, clang -O2 is between 0% and 8% faster, uniformly distributed
depending on call density in the compiled code. Debug builds using
clang -O0 are 0% - 3% faster.
I have verified that this patch doesn't change the assembly generated
for the LLVM nightly test suite when building with -disable-copyprop
and -disable-branch-fold.
Branch folding behaves slightly differently in a few cases because call
instructions have different hash values now.
Copy propagation flushes its data structures when it crosses a register
mask operand. This causes it to leave a few dead copies behind, on the
order of 20 instruction across the entire nightly test suite, including
SPEC. Fixing this properly would require the pass to use different data
structures.
llvm-svn: 150638
Adds an instruction itinerary to all x86 instructions, giving each a default latency of 1, using the InstrItinClass IIC_DEFAULT.
Sets specific latencies for Atom for the instructions in files X86InstrCMovSetCC.td, X86InstrArithmetic.td, X86InstrControl.td, and X86InstrShiftRotate.td. The Atom latencies for the remainder of the x86 instructions will be set in subsequent patches.
Adds a test to verify that the scheduler is working.
Also changes the scheduling preference to "Hybrid" for i386 Atom, while leaving x86_64 as ILP.
Patch by Preston Gurd!
llvm-svn: 149558
It adds register mask operands to x86 call instructions. Once all the
backend passes support register mask operands, this will be permanently
enabled.
llvm-svn: 148438
In CanXFormVExtractWithShuffleIntoLoad we assumed that EXTRACT_VECTOR_ELT can be later handled by the DAGCombiner.
However, in some cases on AVX, the EXTRACT_VECTOR_ELT is legalized to EXTRACT_SUBVECTOR + EXTRACT_VECTOR_ELT, which
currently is not handled by the DAGCombiner. In this patch I added a check that we only extract from the XMM part.
llvm-svn: 148298
We know that the blend instructions only use the MSB, so if the mask is
sign-extended then we can convert it into a SHL instruction. This is a
common pattern because the type-legalizer sign-extends the i1 type which
is used by the LLVM-IR for the condition.
Added a new optimization in SimplifyDemandedBits for SIGN_EXTEND_INREG -> SHL.
llvm-svn: 148225
lc: X86ISelLowering.cpp:6480: llvm::SDValue llvm::X86TargetLowering::LowerVECTOR_SHUFFLE(llvm::SDValue, llvm::SelectionDAG&) const: Assertion `V1.getOpcode() != ISD::UNDEF&& "Op 1 of shuffle should not be undef"' failed.
Added a test.
llvm-svn: 148044
As the comment around 7746 says, it's better to use the x87 extended precision
here than SSE. And the generic code doesn't know how to do that. It also regains
the speed lost for the uint64_to_float.c testcase.
<rdar://problem/10669858>
llvm-svn: 147869
Testing: passed 'make check' including LIT tests for all sequences being handled (both SSE and AVX)
Reviewers: Evan Cheng, David Blaikie, Bruno Lopes, Elena Demikhovsky, Chad Rosier, Anton Korobeynikov
llvm-svn: 147601
This small bit of ASM code is sufficient to do what the old algorithm did:
movq %rax, %xmm0
punpckldq (c0), %xmm0 // c0: (uint4){ 0x43300000U, 0x45300000U, 0U, 0U }
subpd (c1), %xmm0 // c1: (double2){ 0x1.0p52, 0x1.0p52 * 0x1.0p32 }
#ifdef __SSE3__
haddpd %xmm0, %xmm0
#else
pshufd $0x4e, %xmm0, %xmm1
addpd %xmm1, %xmm0
#endif
It's arguably faster. One caveat, the 'haddpd' instruction isn't very fast on
all processors.
<rdar://problem/7719814>
llvm-svn: 147593
(x > y) ? x : y
=>
(x >= y) ? x : y
So for something like
(x - y) > 0 : (x - y) ? 0
It will be
(x - y) >= 0 : (x - y) ? 0
This makes is possible to test sign-bit and eliminate a comparison against
zero. e.g.
subl %esi, %edi
testl %edi, %edi
movl $0, %eax
cmovgl %edi, %eax
=>
xorl %eax, %eax
subl %esi, $edi
cmovsl %eax, %edi
rdar://10633221
llvm-svn: 147512
LZCNT instructions are available. Force promotion to i32 to get
a smaller encoding since the fix-ups necessary are just as complex for
either promoted type
We can't do standard promotion for CTLZ when lowering through BSR
because it results in poor code surrounding the 'xor' at the end of this
instruction. Essentially, if we promote the entire CTLZ node to i32, we
end up doing the xor on a 32-bit CTLZ implementation, and then
subtracting appropriately to get back to an i8 value. Instead, our
custom logic just uses the knowledge of the incoming size to compute
a perfect xor. I'd love to know of a way to fix this, but so far I'm
drawing a blank. I suspect the legalizer could be more clever and/or it
could collude with the DAG combiner, but how... ;]
llvm-svn: 147251
'bsf' instructions here.
This one is actually debatable to my eyes. It's not clear that any chip
implementing 'tzcnt' would have a slow 'bsf' for any reason, and unless
EFLAGS or a zero input matters, 'tzcnt' is just a longer encoding.
Still, this restores the old behavior with 'tzcnt' enabled for now.
llvm-svn: 147246
X86ISelLowering C++ code. Because this is lowered via an xor wrapped
around a bsr, we want the dagcombine which runs after isel lowering to
have a chance to clean things up. In particular, it is very common to
see code which looks like:
(sizeof(x)*8 - 1) ^ __builtin_clz(x)
Which is trying to compute the most significant bit of 'x'. That's
actually the value computed directly by the 'bsr' instruction, but if we
match it too late, we'll get completely redundant xor instructions.
The more naive code for the above (subtracting rather than using an xor)
still isn't handled correctly due to the dagcombine getting confused.
Also, while here fix an issue spotted by inspection: we should have been
expanding the zero-undef variants to the normal variants when there is
an 'lzcnt' instruction. Do so, and test for this. We don't want to
generate unnecessary 'bsr' instructions.
These two changes fix some regressions in encoding and decoding
benchmarks. However, there is still a *lot* to be improve on in this
type of code.
llvm-svn: 147244
use the zero-undefined variants of CTTZ and CTLZ. These are just simple
patterns for now, there is more to be done to make real world code using
these constructs be optimized and codegen'ed properly on X86.
The existing tests are spiffed up to check that we no longer generate
unnecessary cmov instructions, and that we generate the very important
'xor' to transform bsr which counts the index of the most significant
one bit to the number of leading (most significant) zero bits. Also they
now check that when the variant with defined zero result is used, the
cmov is still produced.
llvm-svn: 146974
undefined result. This adds new ISD nodes for the new semantics,
selecting them when the LLVM intrinsic indicates that the undef behavior
is desired. The new nodes expand trivially to the old nodes, so targets
don't actually need to do anything to support these new nodes besides
indicating that they should be expanded. I've done this for all the
operand types that I could figure out for all the targets. Owners of
various targets, please review and let me know if any of these are
incorrect.
Note that the expand behavior is *conservatively correct*, and exactly
matches LLVM's current behavior with these operations. Ideally this
patch will not change behavior in any way. For example the regtest suite
finds the exact same instruction sequences coming out of the code
generator. That's why there are no new tests here -- all of this is
being exercised by the existing test suite.
Thanks to Duncan Sands for reviewing the various bits of this patch and
helping me get the wrinkles ironed out with expanding for each target.
Also thanks to Chris for clarifying through all the discussions that
this is indeed the approach he was looking for. That said, there are
likely still rough spots. Further review much appreciated.
llvm-svn: 146466
This was actually a bit of a mess. TLI.setPrefLoopAlignment was clearly
documented as taking log2(bytes) units, but the x86 target would still
set a preferred loop alignment of '16'.
CodePlacementOpt passed this number on to the basic block, and
AsmPrinter interpreted it as bytes.
Now both MachineFunction and MachineBasicBlock use logarithmic
alignments.
Obviously, MachineConstantPool still measures alignments in bytes, so we
can emulate the thrill of using as.
llvm-svn: 145889
change, now you need a TargetOptions object to create a TargetMachine. Clang
patch to follow.
One small functionality change in PTX. PTX had commented out the machine
verifier parts in their copy of printAndVerify. That now calls the version in
LLVMTargetMachine. Users of PTX who need verification disabled should rely on
not passing the command-line flag to enable it.
llvm-svn: 145714
Before:
movabsq $4294967296, %rax ## encoding: [0x48,0xb8,0x00,0x00,0x00,0x00,0x01,0x00,0x00,0x00]
testq %rax, %rdi ## encoding: [0x48,0x85,0xf8]
jne LBB0_2 ## encoding: [0x75,A]
After:
btq $32, %rdi ## encoding: [0x48,0x0f,0xba,0xe7,0x20]
jb LBB0_2 ## encoding: [0x72,A]
btq is usually slower than testq because it doesn't fuse with the jump, but here we're better off
saving one register and a giant movabsq.
llvm-svn: 145103
When this field is true it means that the load is from constant (runt-time or compile-time) and so can be hoisted from loops or moved around other memory accesses
llvm-svn: 144100
On x86: (shl V, 1) -> add V,V
Hardware support for vector-shift is sparse and in many cases we scalarize the
result. Additionally, on sandybridge padd is faster than shl.
llvm-svn: 143311
fixes: Use a separate register, instead of SP, as the
calling-convention resource, to avoid spurious conflicts with
actual uses of SP. Also, fix unscheduling of calling sequences,
which can be triggered by pseudo-two-address dependencies.
llvm-svn: 143206
it fixes the dragonegg self-host (it looks like gcc is miscompiled).
Original commit messages:
Eliminate LegalizeOps' LegalizedNodes map and have it just call RAUW
on every node as it legalizes them. This makes it easier to use
hasOneUse() heuristics, since unneeded nodes can be removed from the
DAG earlier.
Make LegalizeOps visit the DAG in an operands-last order. It previously
used operands-first, because LegalizeTypes has to go operands-first, and
LegalizeTypes used to be part of LegalizeOps, but they're now split.
The operands-last order is more natural for several legalization tasks.
For example, it allows lowering code for nodes with floating-point or
vector constants to see those constants directly instead of seeing the
lowered form (often constant-pool loads). This makes some things
somewhat more complicated today, though it ought to allow things to be
simpler in the future. It also fixes some bugs exposed by Legalizing
using RAUW aggressively.
Remove the part of LegalizeOps that attempted to patch up invalid chain
operands on libcalls generated by LegalizeTypes, since it doesn't work
with the new LegalizeOps traversal order. Instead, define what
LegalizeTypes is doing to be correct, and transfer the responsibility
of keeping calls from having overlapping calling sequences into the
scheduler.
Teach the scheduler to model callseq_begin/end pairs as having a
physical register definition/use to prevent calls from having
overlapping calling sequences. This is also somewhat complicated, though
there are ways it might be simplified in the future.
This addresses rdar://9816668, rdar://10043614, rdar://8434668, and others.
Please direct high-level questions about this patch to management.
Delete #if 0 code accidentally left in.
llvm-svn: 143188
on every node as it legalizes them. This makes it easier to use
hasOneUse() heuristics, since unneeded nodes can be removed from the
DAG earlier.
Make LegalizeOps visit the DAG in an operands-last order. It previously
used operands-first, because LegalizeTypes has to go operands-first, and
LegalizeTypes used to be part of LegalizeOps, but they're now split.
The operands-last order is more natural for several legalization tasks.
For example, it allows lowering code for nodes with floating-point or
vector constants to see those constants directly instead of seeing the
lowered form (often constant-pool loads). This makes some things
somewhat more complicated today, though it ought to allow things to be
simpler in the future. It also fixes some bugs exposed by Legalizing
using RAUW aggressively.
Remove the part of LegalizeOps that attempted to patch up invalid chain
operands on libcalls generated by LegalizeTypes, since it doesn't work
with the new LegalizeOps traversal order. Instead, define what
LegalizeTypes is doing to be correct, and transfer the responsibility
of keeping calls from having overlapping calling sequences into the
scheduler.
Teach the scheduler to model callseq_begin/end pairs as having a
physical register definition/use to prevent calls from having
overlapping calling sequences. This is also somewhat complicated, though
there are ways it might be simplified in the future.
This addresses rdar://9816668, rdar://10043614, rdar://8434668, and others.
Please direct high-level questions about this patch to management.
llvm-svn: 143177
http://lab.llvm.org:8011/builders/llvm-x86_64-linux/builds/101
--- Reverse-merging r141854 into '.':
U test/MC/Disassembler/X86/x86-32.txt
U test/MC/Disassembler/X86/simple-tests.txt
D test/CodeGen/X86/bmi.ll
U lib/Target/X86/X86InstrInfo.td
U lib/Target/X86/X86ISelLowering.cpp
U lib/Target/X86/X86.td
U lib/Target/X86/X86Subtarget.h
llvm-svn: 141857
- x87: no min or max.
- SSE1: min/max for single precision scalars and vectors.
- SSE2: min/max for single and double precision scalars and vectors.
- AVX: as SSE2, but also supports the wider ymm vectors. (this is covered by the isTypeLegal check)
llvm-svn: 140296
dag-combine optimization to implement the ext-load efficiently (using shuffles).
For example the type <4 x i8> is stored in memory as i32, but it needs to
find its way into a <4 x i32> register. Previously we scalarized the memory
access, now we use shuffles.
llvm-svn: 139995
maxps and maxpd). This broke the sse41-blend.ll testcase by causing
maxpd to be produced rather than a cmp+blend pair, which is the reason
I tweaked it. Gives a small speedup on doduc with dragonegg when the
GCC vectorizer is used.
llvm-svn: 139986
take into consideration the presence of AVX. This change, together with
the SSEDomainFix enabled for AVX, makes AVX codegen to always (hopefully)
emit the same code as SSE for 128-bit vector ops. I don't
have a testcase for this, but AVX now beats SSE in performance for
128-bit ops in the majority of programas in the llvm testsuite
llvm-svn: 139817
However with this fix it does now.
Basically the operand order for the x86 target specific node
is not the same as the instruction, but since the intrinsic need that
specific order at the instruction definition, just change the order
during legalization. Also, there were some wrong invertions of condition
codes, such as GE => LE, GT => LT, fix that too. Fix PR10907.
llvm-svn: 139528
assert("not implemented for target shuffle node");
to:
assert(0 && "not implemented for target shuffle node");
This causes a test failure in CodeGen/X86/palignr.ll which has
been marked as XFAIL for the time being.
Test failure filed at PR10901.
llvm-svn: 139454
in Nadav's r139285 and r139287 commits.
1) Rename vsel.ll to a more descriptive name
2) Change the order of BLEND operands to "Op1, Op2, Cond", this is
necessary because PBLENDVB is already used in different places with
this order, and it was being emitted in the wrong way for vselect
3) Add AVX patterns and tests for the same SSE41 instructions
llvm-svn: 139305
with a vector condition); such selects become VSELECT codegen nodes.
This patch also removes VSETCC codegen nodes, unifying them with SETCC
nodes (codegen was actually often using SETCC for vector SETCC already).
This ensures that various DAG combiner optimizations kick in for vector
comparisons. Passes dragonegg bootstrap with no testsuite regressions
(nightly testsuite as well as "make check-all"). Patch mostly by
Nadav Rotem.
llvm-svn: 139159
init.trampoline and adjust.trampoline intrinsics, into two intrinsics
like in GCC. While having one combined intrinsic is tempting, it is
not natural because typically the trampoline initialization needs to
be done in one function, and the result of adjust trampoline is needed
in a different (nested) function. To get around this llvm-gcc hacks the
nested function lowering code to insert an additional parent variable
holding the adjust.trampoline result that can be accessed from the child
function. Dragonegg doesn't have the luxury of tweaking GCC code, so it
stored the result of adjust.trampoline in the memory GCC set aside for
the trampoline itself (this is always available in the child function),
and set up some new memory (using an alloca) to hold the trampoline.
Unfortunately this breaks Go which allocates trampoline memory on the
heap and wants to use it even after the parent has exited (!). Rather
than doing even more hacks to get Go working, it seemed best to just use
two intrinsics like in GCC. Patch mostly by Sanjoy Das.
llvm-svn: 139140
- Duplicate some store patterns to their AVX forms!
- Catched a bug while restricting the patterns subtarget, fix it
and update a testcase to check it properly
llvm-svn: 138851
code is inserted to first check if the current stacklet has enough
space. If so, space is allocated by simply decrementing the stack
pointer. Otherwise a runtime routine (__morestack_allocate_stack_space
in libgcc) is called which allocates the required memory from the
heap.
Patch by Sanjoy Das.
llvm-svn: 138818
from DYNAMIC_STACKALLOC.
Two new pseudo instructions (SEG_ALLOCA_32 and SEG_ALLOCA_64) which
will match X86SegAlloca (based on word size) are also added. They
will be custom emitted to inject the actual stack handling code.
Patch by Sanjoy Das.
llvm-svn: 138814
X86. Modify the pass added in the previous patch to call this new
code.
This new prologues generated will call a libgcc routine (__morestack)
to allocate more stack space from the heap when required
Patch by Sanjoy Das.
llvm-svn: 138812
explicit about which subtarget they refer to, and add AVX versions of
the ones we currently don't. Make the mask check more strict, to be
clear it won't be used to match to 256-bit versions!
llvm-svn: 138514
match splats in the form (splat (scalar_to_vector (load ...))) whenever
the load can be folded. All the logic and instruction emission is
working but because of PR8156, there are no ways to match loads, cause
they can never be folded for splats. Thus, the tests are XFAILed, but
I've tested and exercised all the logic using a relaxed version for
checking the foldable loads, as if the bug was already fixed. This
should work out of the box once PR8156 gets fixed since MayFoldLoad will
work as expected.
llvm-svn: 137810
vinsertf128 $1 + vpermilps $0, remove the old code that used to first
do the splat in a 128-bit vector and then insert it into a larger one.
This is better because the handling code gets simpler and also makes a
better room for the upcoming vbroadcast!
llvm-svn: 137807
there is no support for native 256-bit shuffles, be more smart in some
cases, for example, when you can extract specific 128-bit parts and use
regular 128-bit shuffles for them. Example:
For this shuffle:
shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32>
<i32 1, i32 0, i32 7, i32 6>
This was expanded to:
vextractf128 $1, %ymm1, %xmm2
vpextrq $0, %xmm2, %rax
vmovd %rax, %xmm1
vpextrq $1, %xmm2, %rax
vmovd %rax, %xmm2
vpunpcklqdq %xmm1, %xmm2, %xmm1
vpextrq $0, %xmm0, %rax
vmovd %rax, %xmm2
vpextrq $1, %xmm0, %rax
vmovd %rax, %xmm0
vpunpcklqdq %xmm2, %xmm0, %xmm0
vinsertf128 $1, %xmm1, %ymm0, %ymm0
ret
Now we get:
vshufpd $1, %xmm0, %xmm0, %xmm0
vextractf128 $1, %ymm1, %xmm1
vshufpd $1, %xmm1, %xmm1, %xmm1
vinsertf128 $1, %xmm1, %ymm0, %ymm0
llvm-svn: 137733
vectors. It operates on 128-bit elements instead of regular scalar
types. Recognize shuffles that are suitable for VPERM2F128 and teach
the x86 legalizer how to handle them.
llvm-svn: 137519
(for example, after integer operation), do not pack the registers into a YMM
before saving. Its better to save as two XMM registers.
Before:
vinsertf128 $1, %xmm3, %ymm0, %ymm3
vinsertf128 $0, %xmm1, %ymm3, %ymm1
vmovaps %ymm1, 416(%rsp)
After:
vmovaps %xmm3, 416+16(%rsp)
vmovaps %xmm1, 416(%rsp)
llvm-svn: 137308
data in-register prior to saving to memory. When we reorder the data in memory
we prevent the need to save multiple scalars to memory, making a single regular
store.
llvm-svn: 137238
avoid returning early for v8i32 types, which would only be valid for
vector with all zeros. Also split the handling of zeros and ones into separate
checking logic since they are handled differently. This fixes PR10547
llvm-svn: 136642
working on x86 (at least for trivial testcases); other architectures will
need more work so that they actually emit the appropriate instructions for
orderings stricter than 'monotonic'. (As far as I can tell, the ARM, PPC,
Mips, and Alpha backends need such changes.)
llvm-svn: 136457
usage of the shuffle bitmask. Both work in 128-bit lanes without
crossing, but in the former the mask of the high part is the same
used by the low part while in the later both lanes have independent
masks. Handle this properly and and add support for vpermilpd.
llvm-svn: 136200
On x86 we can't encode an immediate LHS of a sub directly. If the RHS comes from a XOR with a constant we can
fold the negation into the xor and add one to the immediate of the sub. Then we can turn the sub into an add,
which can be commuted and encoded efficiently.
This code is generated for __builtin_clz and friends.
llvm-svn: 136167
shuffle before inserting on a 256-bit vector.
- Add AVX versions of movd/movq instructions
- Introduce a few COPY patterns to match insert_subvector instructions.
This turns a trivial insert_subvector instruction into a register copy,
coalescing the xmm into a ymm and avoid emiting on more instruction.
llvm-svn: 136002
the way to go. Doing this here will prevent several node matches later,
and would have to force looking all the way through several
VINSERTF128/VEXTRACTF128 chains to optimize simple things.
llvm-svn: 135730
and was actually very wrong, fix it and make it simpler. Also remove the
ConcatVectors function, which is unused now.
- Fix a introduction of useless nodes in r126664 and r126264. The
VUNPCKL* should never be introduced cause we don't want duplicate
nodes for 128 AVX and non-AVX modes, the actual instruction
difference only exists during isel, but not for target specific DAG
nodes. We only introduce V* target nodes when there is no 128-bit
version already there.
- Fix a fragile test and make it more useful.
llvm-svn: 135729
instruction introduced in AVX, which can operate on 128 and 256-bit vectors.
It considers a 256-bit vector as two independent 128-bit lanes. It can permute
any 32 or 64 elements inside a lane, and restricts the second lane to
have the same permutation of the first one. With the improved splat support
introduced early today, adding codegen for this instruction enable more
efficient 256-bit code:
Instead of:
vextractf128 $0, %ymm0, %xmm0
punpcklbw %xmm0, %xmm0
punpckhbw %xmm0, %xmm0
vinsertf128 $0, %xmm0, %ymm0, %ymm1
vinsertf128 $1, %xmm0, %ymm1, %ymm0
vextractf128 $1, %ymm0, %xmm1
shufps $1, %xmm1, %xmm1
movss %xmm1, 28(%rsp)
movss %xmm1, 24(%rsp)
movss %xmm1, 20(%rsp)
movss %xmm1, 16(%rsp)
vextractf128 $0, %ymm0, %xmm0
shufps $1, %xmm0, %xmm0
movss %xmm0, 12(%rsp)
movss %xmm0, 8(%rsp)
movss %xmm0, 4(%rsp)
movss %xmm0, (%rsp)
vmovaps (%rsp), %ymm0
We get:
vextractf128 $0, %ymm0, %xmm0
punpcklbw %xmm0, %xmm0
punpckhbw %xmm0, %xmm0
vinsertf128 $0, %xmm0, %ymm0, %ymm1
vinsertf128 $1, %xmm0, %ymm1, %ymm0
vpermilps $85, %ymm0, %ymm0
llvm-svn: 135662
refactor the code and add a bunch of comments. The final shuffle
emitted by handling 256-bit types is suitable for the VPERM shuffle
instruction which is going to be introduced in a next commit (with
a testcase which cover this commit)
llvm-svn: 135661
to MCRegisterInfo. Also initialize the mapping at construction time.
This patch eliminate TargetRegisterInfo from TargetAsmInfo. It's another step
towards fixing the layering violation.
llvm-svn: 135424
1) Make non-legal 256-bit loads to be promoted to v4i64. This lets us
canonize the loads and handle things the same way we use to handle
for 128-bit registers. Despite of what one of the removed comments
explained, the load promotion would not mess with VPERM, it's only a
matter of doing the appropriate bitcasts when this instructions comes
to be introduced. Also make LOAD v8i32 legal.
2) Doing 1) exposed two bugs:
- v4i64 was being promoted to itself for several opcodes (introduced
in r124447 by David Greene) causing endless recursion and the stack to
explode.
- there was no support for allOnes BUILD_VECTORs and ANDNP would fail to
match because it was generating early target constant pools during
lowering.
3) The testcases are already checked-in, doing 1) exposed the
bugs in the current testcases.
4) Tidy up code to be more clear and explicit about AVX.
llvm-svn: 135313
when determining validity of matching constraint. Allow i1
types access to the GR8 reg class for x86.
Fixes PR10352 and rdar://9777108
llvm-svn: 135180
During type legalization we often use the SIGN_EXTEND_INREG SDNode.
When this SDNode is legalized during the LegalizeVector phase, it is
scalarized because non-simple types are automatically marked to be expanded.
In this patch we add support for lowering SIGN_EXTEND_INREG manually.
This fixes CodeGen/X86/vec_sext.ll when running with the '-promote-elements'
flag.
llvm-svn: 135144
Drop the FpMov instructions, use plain COPY instead.
Drop the FpSET/GET instruction for accessing fixed stack positions.
Instead use normal COPY to/from ST registers around inline assembly, and
provide a single new FpPOP_RETVAL instruction that can access the return
value(s) from a call. This is still necessary since you cannot tell from
the CALL instruction alone if it returns anything on the FP stack. Teach
fast isel to use this.
This provides a much more robust way of handling fixed stack registers -
we can tolerate arbitrary FP stack instructions inserted around calls
and inline assembly. Live range splitting could sometimes break x87 code
by inserting spill code in unfortunate places.
As a bonus we handle floating point inline assembly correctly now.
llvm-svn: 134018
optimizations when emitting calls to the function; instead those calls may
use faster relocations which require the function to be immediately resolved
upon loading the dynamic object featuring the call. This is useful when it
is known that the function will be called frequently and pervasively and
therefore there is no merit in delaying binding of the function.
Currently only implemented for x86-64, where it turns into a call through
the global offset table.
Patch by Dan Gohman, who assures me that he's going to add LangRef documentation
for this once it's committed.
llvm-svn: 133080
floating-point comparison, generate a mask of 0s or 1s, and generally
DTRT with NaNs. Only profitable when the user wants a materialized 0
or 1 at runtime. rdar://problem/5993888
llvm-svn: 132404
non-zero.
- Teach X86 cmov optimization to eliminate the cmov from ctlz, cttz extension
when the source of X86ISD::BSR / X86ISD::BSF is proven to be non-zero.
rdar://9490949
llvm-svn: 131948
to have single return block (at least getting there) for optimizations. This
is general goodness but it would prevent some tailcall optimizations.
One specific case is code like this:
int f1(void);
int f2(void);
int f3(void);
int f4(void);
int f5(void);
int f6(void);
int foo(int x) {
switch(x) {
case 1: return f1();
case 2: return f2();
case 3: return f3();
case 4: return f4();
case 5: return f5();
case 6: return f6();
}
}
=>
LBB0_2: ## %sw.bb
callq _f1
popq %rbp
ret
LBB0_3: ## %sw.bb1
callq _f2
popq %rbp
ret
LBB0_4: ## %sw.bb3
callq _f3
popq %rbp
ret
This patch teaches codegenprep to duplicate returns when the return value
is a phi and where the phi operands are produced by tail calls followed by
an unconditional branch:
sw.bb7: ; preds = %entry
%call8 = tail call i32 @f5() nounwind
br label %return
sw.bb9: ; preds = %entry
%call10 = tail call i32 @f6() nounwind
br label %return
return:
%retval.0 = phi i32 [ %call10, %sw.bb9 ], [ %call8, %sw.bb7 ], ... [ 0, %entry ]
ret i32 %retval.0
This allows codegen to generate better code like this:
LBB0_2: ## %sw.bb
jmp _f1 ## TAILCALL
LBB0_3: ## %sw.bb1
jmp _f2 ## TAILCALL
LBB0_4: ## %sw.bb3
jmp _f3 ## TAILCALL
rdar://9147433
llvm-svn: 127953
not have native support for this operation (such as X86).
The legalized code uses two vector INT_TO_FP operations and is faster
than scalarizing.
llvm-svn: 127951
comparisons on x86. Essentially, the way this works is that SUB+SBB sets
the relevant flags the same way a double-width CMP would.
This is a substantial improvement over the generic lowering in LLVM. The output
is also shorter than the gcc-generated output; I haven't done any detailed
benchmarking, though.
llvm-svn: 127852
rather than an int. Thankfully, this only causes LLVM to miss optimizations, not
generate incorrect code.
This just fixes the zext at the return. We still insert an i32 ZextAssert when
reading a function's arguments, but it is followed by a truncate and another i8
ZextAssert so it is not optimized.
llvm-svn: 127766
testcases accordingly. Some are currently xfailed and will be filed
as bugs to be fixed or understood.
Performance results:
roughly neutral on SPEC
some micro benchmarks in the llvm suite are up between 100 and 150%, only
a pair of regressions that are due to be investigated
john-the-ripper saw:
10% improvement in traditional DES
8% improvement in BSDI DES
59% improvement in FreeBSD MD5
67% improvement in OpenBSD Blowfish
14% improvement in LM DES
Small compile time impact.
llvm-svn: 127208
regs. This is the only change in this checkin that may affects the
default scheduler. With better register tracking and heuristics, it
doesn't make sense to artificially lower the register limit so much.
Added -sched-high-latency-cycles and X86InstrInfo::isHighLatencyDef to
give the scheduler a way to account for div and sqrt on targets that
don't have an itinerary. It is currently defaults to 10 (the actual
number doesn't matter much), but only takes effect on non-default
schedulers: list-hybrid and list-ilp.
Added several heuristics that can be individually disabled for the
non-default sched=list-ilp mode. This helps us determine how much
better we can do on a given benchmark than the default
scheduler. Certain compute intensive loops run much faster in this
mode with the right set of heuristics, and it doesn't seem to have
much negative impact elsewhere. Not all of the heuristics are needed,
but we still need to experiment to decide which should be disabled by
default for sched=list-ilp.
llvm-svn: 127067
and 256-bit forms. Because the number of elements in a vector
does not determine the vector type (4 elements could be v4f32 or
v4f64), pass the full type of the vector to decode routines.
llvm-svn: 126664
In other words, do not keep track of argument's location. The debugger (gdb) is not prepared to see line table entries for arguments. For the debugger, "second" line table entry marks beginning of function body.
This requires some coordination with debugger to get this working.
- The debugger needs to be aware of prolog_end attribute attached with line table entries.
- The compiler needs to accurately mark prolog_end in line table entries (at -O0 and at -O1+)
llvm-svn: 126155
(LLVMX86Utils.a) to break cyclic library dependencies between
LLVMX86CodeGen.a and LLVMX86AsmParser.a. Previously this code was in
a header file and marked static but AVX requires some additional
functionality here that won't be used by all clients. Since including
unused static functions causes a gcc compiler warning, keeping it as a
header would break builds that use -Werror. Putting this in its own
library solves both problems at once.
llvm-svn: 125765
have their low bits set to zero. This allows us to optimize
out explicit stack alignment code like in stack-align.ll:test4 when
it is redundant.
Doing this causes the code generator to start turning FI+cst into
FI|cst all over the place, which is general goodness (that is the
canonical form) except that various pieces of the code generator
don't handle OR aggressively. Fix this by introducing a new
SelectionDAG::isBaseWithConstantOffset predicate, and using it
in places that are looking for ADD(X,CST). The ARM backend in
particular was missing a lot of addressing mode folding opportunities
around OR.
llvm-svn: 125470
This allows us to easily support 256-bit operations that don't have
native 256-bit support. This applies to integer operations, certain
types of shuffles and various othher things.
llvm-svn: 124910
matching EXTRACT_SUBVECTOR to VEXTRACTF128 along with support routines
to examine and translate index values. VINSERTF128 comes next. With
these two in place we can begin supporting more AVX operations as
INSERT/EXTRACT can be used as a fallback when 256-bit support is not
available.
llvm-svn: 124797
Reversing the operands allows us to fold, but doesn't force us to. Also, at
this point the DAG is still being optimized, so the check for hasOneUse is not
very precise.
llvm-svn: 124773
default implementation for x86, going through the stack in a similr
fashion to how the codegen implements BUILD_VECTOR. Eventually this
will get matched to VINSERTF128 if AVX is available.
llvm-svn: 124307
implementation of EXTRACT_SUBVECTOR for x86, going through the stack
in a similr fashion to how the codegen implements BUILD_VECTOR.
Eventually this will get matched to VEXTRACTF128 if AVX is available.
llvm-svn: 124292
The theory is it's still faster than a pair of movq / a quad of movl. This
will probably hurt older chips like P4 but should run faster on current
and future Intel processors. rdar://8817010
llvm-svn: 122955
the same as setcc. Optimize ADDC(0,0,FLAGS) -> SET_CARRY(FLAGS). This is
a step towards finishing off PR5443. In the testcase in that bug we now get:
movq %rdi, %rax
addq %rsi, %rax
sbbq %rcx, %rcx
testb $1, %cl
setne %dl
ret
instead of:
movq %rdi, %rax
addq %rsi, %rax
movl $0, %ecx
adcq $0, %rcx
testq %rcx, %rcx
setne %dl
ret
llvm-svn: 122219
their carry depenedencies with MVT::Flag operands) and use clean and beautiful
EFLAGS dependences instead.
We do this by changing the modelling of SBB/ADC to have EFLAGS input and outputs
(which is what requires the previous scheduler change) and change X86 ISelLowering
to custom lower ADDC and friends down to X86ISD::ADD/ADC/SUB/SBB nodes.
With the previous series of changes, this causes no changes in the testsuite, woo.
llvm-svn: 122213
the output to the correct register. Fixes a hidden problem uncovered
by the last patch where we'd try to DAG combine our MVT::Other node
oddly.
llvm-svn: 121358
result. This allows us to compile:
void *test12(long count) {
return new int[count];
}
into:
test12:
movl $4, %ecx
movq %rdi, %rax
mulq %rcx
movq $-1, %rdi
cmovnoq %rax, %rdi
jmp __Znam ## TAILCALL
instead of:
test12:
movl $4, %ecx
movq %rdi, %rax
mulq %rcx
seto %cl
testb %cl, %cl
movq $-1, %rdi
cmoveq %rax, %rdi
jmp __Znam
Of course it would be even better if the regalloc inverted the cmov to 'cmovoq',
which would eliminate the need for the 'movq %rdi, %rax'.
llvm-svn: 120936
backend that they were all implemented except umul. This one fell back
to the default implementation that did a hi/lo multiply and compared the
top. Fix this to check the overflow flag that the 'mul' instruction
sets, so we can avoid an explicit test. Now we compile:
void *func(long count) {
return new int[count];
}
into:
__Z4funcl: ## @_Z4funcl
movl $4, %ecx ## encoding: [0xb9,0x04,0x00,0x00,0x00]
movq %rdi, %rax ## encoding: [0x48,0x89,0xf8]
mulq %rcx ## encoding: [0x48,0xf7,0xe1]
seto %cl ## encoding: [0x0f,0x90,0xc1]
testb %cl, %cl ## encoding: [0x84,0xc9]
movq $-1, %rdi ## encoding: [0x48,0xc7,0xc7,0xff,0xff,0xff,0xff]
cmoveq %rax, %rdi ## encoding: [0x48,0x0f,0x44,0xf8]
jmp __Znam ## TAILCALL
instead of:
__Z4funcl: ## @_Z4funcl
movl $4, %ecx ## encoding: [0xb9,0x04,0x00,0x00,0x00]
movq %rdi, %rax ## encoding: [0x48,0x89,0xf8]
mulq %rcx ## encoding: [0x48,0xf7,0xe1]
testq %rdx, %rdx ## encoding: [0x48,0x85,0xd2]
movq $-1, %rdi ## encoding: [0x48,0xc7,0xc7,0xff,0xff,0xff,0xff]
cmoveq %rax, %rdi ## encoding: [0x48,0x0f,0x44,0xf8]
jmp __Znam ## TAILCALL
Other than the silly seto+test, this is using the o bit directly, so it's going in the right
direction.
llvm-svn: 120935
The user (i.e. whoever generated a call to the intrinsic in the first place) is
essentially asking for a particular instruction to be placed in the assembler.
If that instruction won't execute on the target machine, that's their problem
not ours. Two buildbots with processors that don't support SSE3 were barfing
on the apm.ll test in CodeGen/X86 because of this assertion.
llvm-svn: 120574
legalization time. Since at legalization time there is no mapping from
SDNode back to the corresponding LLVM instruction and the return
SDNode is target specific, this requires a target hook to check for
eligibility. Only x86 and ARM support this form of sibcall optimization
right now.
rdar://8707777
llvm-svn: 120501
nodes to indicate when ha16/lo16 modifiers should be used. This lets
us pass PowerPC/indirectbr.ll.
The one annoying thing about this patch is that the MCSymbolExpr isn't
expressive enough to represent ha16(label1-label2) which we need on
PowerPC. I have a terrible hack in the meantime, but this will have
to be revisited at some point.
Last major conversion item left is global variable references.
llvm-svn: 119105
The x86_mmx type is used for MMX intrinsics, parameters and
return values where these use MMX registers, and is also
supported in load, store, and bitcast.
Only the above operations generate MMX instructions, and optimizations
do not operate on or produce MMX intrinsics.
MMX-sized vectors <2 x i32> etc. are lowered to XMM or split into
smaller pieces. Optimizations may occur on these forms and the
result casted back to x86_mmx, provided the result feeds into a
previous existing x86_mmx operation.
The point of all this is prevent optimizations from introducing
MMX operations, which is unsafe due to the EMMS problem.
llvm-svn: 115243
ARM cross-compiler on x86, because the MMO size did not match the type size.
This fixes the MMO size and also the size of the stack object to match the
type size.
llvm-svn: 114554
(sbbl x, x) sets the registers to 0 or ~0. Combined with two's complement arithmetic, we can fold
the intermediate AND and the ADD into a single SUB.
This fixes <rdar://problem/8449754>.
llvm-svn: 114460
"getFixedStack" on the MachinePointerInfo class. While
this isn't the problem I'm setting out to solve, it is the
right way to eliminate PseudoSourceValue, so lets go with it.
llvm-svn: 114406
nodes to emit shuffles and don't do isel mask matching anymore.
- Add the selection of the remaining shuffle opcode (movddup)
- Introduce two new functions to "recognize" where we may get
potential folds and add several comments to them explaining why
they are not yet in the desidered shape.
- Add more patterns to fallback the case where we select
a specific shuffle opcode as if it could fold a load, but it
can't, so remap to a valid instruction.
- Add a couple of FIXMEs to address in the following days once
there's a good solution to the current folding problem.
llvm-svn: 113369
checking each standalone condition and decide whether emit target
specific nodes or remove the condition if it's already matched before.
llvm-svn: 113031
"Use target specific nodes instead of relying in unpckl and
unpckh pattern fragments during isel time. Also place a
depth limit in getShuffleScalarElt.
llvm-svn: 113020
- Teach getShuffleScalarElt how to handle more target
specific nodes, so the DAGCombine can make use of it.
- Add another hack to avoid the node update problem
during legalization. More description on the comments
llvm-svn: 112934
when the top elements of a vector are undefined. This happens all
the time for X86-64 ABI stuff because only the low 2 elements of
a 4 element vector are defined. For example, on:
_Complex float f32(_Complex float A, _Complex float B) {
return A+B;
}
We used to produce (with SSE2, SSE4.1+ uses insertps):
_f32: ## @f32
movdqa %xmm0, %xmm2
addss %xmm1, %xmm2
pshufd $16, %xmm2, %xmm2
pshufd $1, %xmm1, %xmm1
pshufd $1, %xmm0, %xmm0
addss %xmm1, %xmm0
pshufd $16, %xmm0, %xmm1
movdqa %xmm2, %xmm0
unpcklps %xmm1, %xmm0
ret
We now produce:
_f32: ## @f32
movdqa %xmm0, %xmm2
addss %xmm1, %xmm2
pshufd $1, %xmm1, %xmm1
pshufd $1, %xmm0, %xmm3
addss %xmm1, %xmm3
movaps %xmm2, %xmm0
unpcklps %xmm3, %xmm0
ret
This implements rdar://8368414
llvm-svn: 112378
Also teach this logic how to handle target specific shuffles if
needed, this is necessary while searching recursively for zeroed
scalar elements in vector shuffle operands.
llvm-svn: 112348
Mark _alloca call as clobberring EFLAGS, otherwise some DCE might remove
other flags-clobberring stuff (e.g. cmp instructions) occuring after
_alloca call.
llvm-svn: 112034
general idea here is to have a group of x86 target specific nodes which are
going to be selected during lowering and then directly matched in isel.
The commit includes the addition of those specific nodes and a *bunch* of
patterns, and incrementally we're going to switch between them and what we
have right now. Both the patterns and target specific nodes can change as
we move forward with this work.
llvm-svn: 111691
- Do not clobber al during variadic calls, this is AMD64 ABI-only feature
- Emit wincall64, where necessary
Patch by Cameron Esfahani!
llvm-svn: 111289
term goal here is to be able to match enough of vector_shuffle and build_vector
so all avx intrinsics which aren't mapped to their own built-ins but to
shufflevector calls can be codegen'd. This is the first (baby) step, support
building zeroed vectors.
llvm-svn: 110897
avoids trouble if the return type of TD->getPointerSize() is
changed to something which doesn't promote to a signed type,
and is simpler anyway.
Also, use getCopyFromReg instead of getRegister to read a
physical register's value.
llvm-svn: 110835
Apply the same approach of SSE4.1 ptest intrinsics but
create a new x86 node "testp" since AVX introduces
vtest{ps}{pd} instructions which set ZF and CF depending
on sign bit AND and ANDN of packed floating-point sources.
This is slightly different from what the "ptest" does.
Tests comming with the other 256 intrinsics tests.
llvm-svn: 110744
declared during the addition of the assembler support, the additional
changes are:
- Add missing intrinsics
- Move all SSE conversion instructions in X86InstInfo64.td to the SSE.td file.
- Duplicate some patterns to AVX mode.
- Step into PCMPEST/PCMPIST custom inserter and add AVX versions.
llvm-svn: 109878
We do sometimes load from a too small stack slot when dealing with x86 arguments
(varargs and smaller-than-32-bit args). It looks like we know what we are doing
in those cases, so I am going to remove the assert instead of artifically
enlarging stack slot sizes.
The assert in storeRegToStackSlot stays in. We don't want to write beyond the
bounds of a stack slot.
llvm-svn: 109764
The size of this object isn't used for anything - technically it is of variable
size.
This avoids a false positive from the assert in
X86InstrInfo::loadRegFromStackSlot, and fixes PR7735.
llvm-svn: 109652
appropriate for targets without detailed instruction iterineries.
The scheduler schedules for increased instruction level parallelism in
low register pressure situation; it schedules to reduce register pressure
when the register pressure becomes high.
On x86_64, this is a win for all tests in CFP2000. It also sped up 256.bzip2
by 16%.
llvm-svn: 109300
SSE, so we can't return floating point values if this
is disabled. Detect this error for clang.
With SSE1 only, f64 is a problem; it can be done, but
neither llvm-gcc nor clang has ever generated correct
code for it. Since nobody noticed this I think it's
OK to treat it as an error for now.
This also handles SSE-sized vectors of floating point.
8207686, 8204109.
llvm-svn: 109201
1) all registers were spilled as xmm, regardless of actual size
2) win64 abi doesn't do the varargs-size-in-%al thing
Still to look into:
xmm6-15 are marked as clobbered by call instructions on win64 even though they aren't.
llvm-svn: 109035
-enable-no-nans-fp-math and -enable-no-infs-fp-math. All of the current codegen fp math optimizations only care whether the fp arithmetics arguments and results can never be NaN.
llvm-svn: 108465
address cannot be allocated a register is in 32-bit mode where the first
three arguments are marked inreg. In that case EAX, EDX, and ECX will be
used for argument passing.
This fixes PR7610.
llvm-svn: 108327
- Check getBytesToPopOnReturn().
- Eschew ST0 and ST1 for return values.
- Fix the PIC base register initialization so that it doesn't ever
fail to end up the top of the entry block.
llvm-svn: 108039
it is popped, even if it is ununsed. A CopyFromReg node is too weak to represent
the required sideeffect, so insert an FpGET_ST0 instruction directly instead.
This will matter when CopyFromReg gets lowered to a generic COPY instruction.
llvm-svn: 108037
U utils/TableGen/FastISelEmitter.cpp
--- Reverse-merging r107943 into '.':
U test/CodeGen/X86/fast-isel.ll
U test/CodeGen/X86/fast-isel-loads.ll
U include/llvm/Target/TargetLowering.h
U include/llvm/Support/PassNameParser.h
U include/llvm/CodeGen/FunctionLoweringInfo.h
U include/llvm/CodeGen/CallingConvLower.h
U include/llvm/CodeGen/FastISel.h
U include/llvm/CodeGen/SelectionDAGISel.h
U lib/CodeGen/LLVMTargetMachine.cpp
U lib/CodeGen/CallingConvLower.cpp
U lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
U lib/CodeGen/SelectionDAG/FunctionLoweringInfo.cpp
U lib/CodeGen/SelectionDAG/FastISel.cpp
U lib/CodeGen/SelectionDAG/SelectionDAGISel.cpp
U lib/CodeGen/SelectionDAG/ScheduleDAGSDNodes.cpp
U lib/CodeGen/SelectionDAG/InstrEmitter.cpp
U lib/CodeGen/SelectionDAG/TargetLowering.cpp
U lib/Target/XCore/XCoreISelLowering.cpp
U lib/Target/XCore/XCoreISelLowering.h
U lib/Target/X86/X86ISelLowering.cpp
U lib/Target/X86/X86FastISel.cpp
U lib/Target/X86/X86ISelLowering.h
llvm-svn: 107987
like all other instructions, even though a segment is not
allowed. This resolves a bunch of gross hacks in the
encoder and makes LEA more consistent with the rest of the
instruction set.
No functionality change.
llvm-svn: 107934
the example in the testcase, we now generate:
_test1: ## @test1
movss 4(%esp), %xmm0
addss 8(%esp), %xmm0
movl 12(%esp), %eax
movss %xmm0, (%eax)
ret
instead of:
_test1: ## @test1
subl $20, %esp
movl 24(%esp), %eax
movq %mm0, (%esp)
movq %mm0, 8(%esp)
movss (%esp), %xmm0
addss 12(%esp), %xmm0
movss %xmm0, (%eax)
addl $20, %esp
ret
v2f32 support did not work reliably because most of the X86
backend didn't know it was legal. It was apparently only added
to support returning source-level v2f32 values in MMX registers
in x86-32 mode. If ABI compatibility is important on this
GCC-extended-vector type for some reason, then the frontend
should generate IR that returns v2i32 instead of v2f32. However,
we generally don't try very hard to be abi compatible on gcc
extended vectors.
llvm-svn: 107601
v2f32 as legal in 32-bit mode. It is just as terrible there,
but I just care about x86-64 and noone claims it is valuable
in 64-bit mode.
llvm-svn: 107600
for an "i" constraint should get lowered; PR 6309. While
this argument was passed around a lot, this is the only
place it was used, so it goes away from a lot of other
places.
llvm-svn: 106893
address requires a register or secondary load to compute
(most PIC modes). This improves "g" constraint handling. 8015842.
The test from 2007 is attempting to test the fix for PR1761,
but since -relocation-model=static doesn't work on Darwin
x86-64, it was not testing what it was supposed to be testing
and was passing erroneously. Fixed to use Linux x86-64.
llvm-svn: 106779
will conflict with another live range. The place which creates this scenerio is
the code in X86 that lowers a select instruction by splitting the MBBs. This
eliminates the need to check from the bottom up in an MBB for live pregs.
llvm-svn: 106066
x86 backend currently doesn't know how to handle them.
This doesn't really fix anything because LegalizeTypes doesn't know how to
handle them either. We do get a better error message, though.
llvm-svn: 105305
<1xi64> -> i64 to work in MMX registers on hosts where -no-sse
is the default (not mine). The right thing is
to accept this and make i64->f64 conversions go through memory,
but I don't have time right now.
llvm-svn: 103914
The implementation in LegalizeIntegerTypes to handle this as
sint64->float + appropriate power of 2 is subject to double rounding,
considered incorrect by numerics people. Use this implementation only
when it is safe. This leads to using library calls in some cases
that produced inline code before, but it's correct now.
(EVTToAPFloatSemantics belongs somewhere else, any suggestions?)
Add a correctly rounding (though not particularly fast) conversion
that uses X87 80-bit computations for x86-32.
7885399, 5901940. This shows up in gcc.c-torture/execute/ieee/rbug.c
in the gcc testsuite on some platforms.
llvm-svn: 103883
the variable actually tracks.
N.B., several back-ends are using "HasCalls" as being synonymous for something
that adjusts the stack. This isn't 100% correct and should be looked into.
llvm-svn: 103802
Move EmitTargetCodeForMemcpy, EmitTargetCodeForMemset, and
EmitTargetCodeForMemmove out of TargetLowering and into
SelectionDAGInfo to exercise this.
llvm-svn: 103481
form of DEBUG_VALUE, as it doesn't have reasonable default
behavior for unsupported targets. Add a new hook instead.
No functional change.
llvm-svn: 102320
user-defined operations that use MMX register types, but
the compiler shouldn't generate them on its own. This adds
a Synthesizable abstraction to represent this, and changes
the vector widening computation so it won't produce MMX types.
(The motivation is to remove noise from the ABI compatibility
part of the gcc test suite, which has some breakage right now.)
llvm-svn: 101951
const_casts, and it reinforces the design of the Target classes being
immutable.
SelectionDAGISel::IsLegalToFold is now a static member function, because
PIC16 uses it in an unconventional way. There is more room for API
cleanup here.
And PIC16's AsmPrinter no longer uses TargetLowering.
llvm-svn: 101635
with a fix for self-hosting
rotate CallInst operands, i.e. move callee to the back
of the operand array
the motivation for this patch are laid out in my mail to llvm-commits:
more efficient access to operands and callee, faster callgraph-construction,
smaller compiler binary
llvm-svn: 101465
with a fix
rotate CallInst operands, i.e. move callee to the back
of the operand array
the motivation for this patch are laid out in my mail to llvm-commits:
more efficient access to operands and callee, faster callgraph-construction,
smaller compiler binary
llvm-svn: 101397
of the operand array
the motivation for this patch are laid out in my mail to llvm-commits:
more efficient access to operands and callee, faster callgraph-construction,
smaller compiler binary
llvm-svn: 101364
Added support for address spaces and added a isVolatile field to memcpy, memmove, and memset,
e.g., llvm.memcpy.i32(i8*, i8*, i32, i32) -> llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1)
llvm-svn: 100304
Added support for address spaces and added a isVolatile field to memcpy, memmove, and memset,
e.g., llvm.memcpy.i32(i8*, i8*, i32, i32) -> llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1)
llvm-svn: 100191
1. Makes it possible to lower with floating point loads and stores.
2. Avoid unaligned loads / stores unless it's fast.
3. Fix some memcpy lowering logic bug related to when to optimize a
load from constant string into a constant.
4. Adjust x86 memcpy lowering threshold to make it more sane.
5. Fix x86 target hook so it uses vector and floating point memory
ops more effectively.
rdar://7774704
llvm-svn: 100090
e.g., llvm.memcpy.i32(i8*, i8*, i32, i32) -> llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1)
A update of langref will occur in a subsequent checkin.
llvm-svn: 99928
create symbols. It is extremely error prone and a source of a lot
of the remaining integrated assembler bugs on x86-64.
This fixes rdar://7807601.
llvm-svn: 99902
makes calls a little bit more consistent and allows easy removal of the
specializations in the future. Convert all callers to the templated functions.
llvm-svn: 99838
- Although it would be nice to allow this decoupling, the assembler needs to be able to reason about MCSymbolRefExprs in too many places to make this viable. We can use a target specific encoding of the variant if this becomes an issue.
- This patch also extends llvm-mc to support parsing of the modifiers, as opposed to lumping them in with the symbol.
llvm-svn: 98592
32-bit indices. Instead of shuffling each element out of the index vector,
when all indices are needed, just store the input vector to the stack and
load the elements out.
llvm-svn: 98588
to is local to the translation unit, we need to place fill the value of that
symbol into the non-lazy pointer.
This should conclude all Darwin changes for placing the LSDA into the TEXT
section. There is some cleanup to do. I.e., there's no longer a special need for
target-specific code here. But that can come later.
llvm-svn: 98564
necessary to swap the operands to handle NaN and negative zero properly.
Also, reintroduce logic for checking for NaN conditions when forming
SSE min and max instructions, fixed to take into consideration NaNs and
negative zeros. This allows forming min and max instructions in more
cases.
llvm-svn: 97025
during a tail call. A parameter might overwrite this stack slot during the tail
call.
The sequence during a tail call is:
1.) load return address to temp reg
2.) move parameters (might involve storing to return address stack slot)
3.) store return address to new location from temp reg
If the stack location is marked immutable CodeGen can colocate load (1) with the
store (3).
This fixes bug 6225.
llvm-svn: 96783
SSE min and max instructions. The real thing this code needs to be
concerned about is negative zero.
Update the sse-minmax.ll test accordingly, and add tests for
-enable-unsafe-fp-math mode as well.
llvm-svn: 96775
create an X86ISD::Cmp node with result type i64 on the
CodeGen/X86/shift-i256.ll testcase and the new isel was assert on it
downstream.
llvm-svn: 96768
dragonegg self-host build. I reverted 96640 in order to revert
96556 (96640 goes on top of 96556), but it also looks like with
both of them applied the breakage happens even earlier. The
symptom of the 96556 miscompile is the following crash:
llvm[3]: Compiling AlphaISelLowering.cpp for Release build
cc1plus: /home/duncan/tmp/tmp/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp:4982: void llvm::SelectionDAG::ReplaceAllUsesWith(llvm::SDNode*, llvm::SDNode*, llvm::SelectionDAG::DAGUpdateListener*): Assertion `(!From->hasAnyUseOfValue(i) || From->getValueType(i) == To->getValueType(i)) && "Cannot use this version of ReplaceAllUsesWith!"' failed.
Stack dump:
0. Running pass 'X86 DAG->DAG Instruction Selection' on function '@_ZN4llvm19AlphaTargetLowering14LowerOperationENS_7SDValueERNS_12SelectionDAGE'
g++: Internal error: Aborted (program cc1plus)
This occurs when building LLVM using LLVM built by LLVM (via
dragonegg). Probably LLVM has miscompiled itself, though it
may have miscompiled GCC and/or dragonegg itself: at this point
of the self-host build, all of GCC, LLVM and dragonegg were built
using LLVM. Unfortunately this kind of thing is extremely hard
to debug, and while I did rummage around a bit I didn't find any
smoking guns, aka obviously miscompiled code.
Found by bisection.
r96556 | evancheng | 2010-02-18 03:13:50 +0100 (Thu, 18 Feb 2010) | 5 lines
Some dag combiner goodness:
Transform br (xor (x, y)) -> br (x != y)
Transform br (xor (xor (x,y), 1)) -> br (x == y)
Also normalize (and (X, 1) == / != 1 -> (and (X, 1)) != / == 0 to match to "test on x86" and "tst on arm"
r96640 | evancheng | 2010-02-19 01:34:39 +0100 (Fri, 19 Feb 2010) | 16 lines
Transform (xor (setcc), (setcc)) == / != 1 to
(xor (setcc), (setcc)) != / == 1.
e.g. On x86_64
%0 = icmp eq i32 %x, 0
%1 = icmp eq i32 %y, 0
%2 = xor i1 %1, %0
br i1 %2, label %bb, label %return
=>
testl %edi, %edi
sete %al
testl %esi, %esi
sete %cl
cmpb %al, %cl
je LBB1_2
llvm-svn: 96672
If there exists a use of a build_vector that's the bitwise complement of the mask,
then transform the node to
(and (xor x, (build_vector -1,-1,-1,-1)), (build_vector ~c1,~c2,~c3,~c4)).
Since this transformation is only useful when 1) the given build_vector will
become a load from constpool, and 2) (and (xor x -1), y) matches to a single
instruction, I decided this is appropriate as a x86 specific transformation.
rdar://7323335
llvm-svn: 96389
lowering and requires that certain types exist in ValueTypes.h. Modified widening to
check if an op can trap and if so, the widening algorithm will apply only the op on
the defined elements. It is safer to do this in widening because the optimizer can't
guarantee removing unused ops in some cases.
llvm-svn: 95823
where callee's arguments are already in the caller's own caller's stack and
they line up perfectly. e.g.
extern int foo(int a, int b, int c);
int bar(int a, int b, int c) {
return foo(a, b, c);
}
llvm-svn: 95053
even when -tailcallopt is not specified and it does not require changing ABI.
First case is the most trivial one. Perform tail call optimization when both
the caller and callee do not return values and when the callee does not take
any input arguments.
llvm-svn: 94664
Target independent isel should always pass along the "tail call" property. Change
target hook LowerCall's parameter "isTailCall" into a refernce. If the target
decides it's impossible to honor the tail call request, it should set isTailCall
to false to make target independent isel happy.
llvm-svn: 94626
which is more convenient, and change getPICJumpTableRelocBaseExpr
to take a MachineFunction to match.
Next, move the X86 code that create a PICBase symbol to
X86TargetLowering::getPICBaseSymbol from
X86MCInstLower::GetPICBaseSymbol, which was an asmprinter specific
library. This eliminates a 'gross hack', and allows us to
implement X86ISelLowering::getPICJumpTableRelocBaseExpr which now
calls it.
This in turn allows us to eliminate the
X86AsmPrinter::printPICJumpTableSetLabel method, which was the
only overload of printPICJumpTableSetLabel.
llvm-svn: 94526
jump table entry kind, instead of overloading
AsmPrinter::printPICJumpTableEntry.
This has a pretty horrible and inefficient FIXME around how @GOTOFF
is currently smashed into the mcsymbol name, but otherwise this is
much cleaner.
llvm-svn: 94516
- getToken is modeled after StringRef::split but it can split on multiple
separator chars and skips leading seperators.
- SplitString is a StringRef::split variant for more than 2 elements with the
same behaviour as getToken.
llvm-svn: 93161
R11, and then asserting that the target was in R9. Since R9 isn't reserved for
the target anymore, and is used as an argument, this patch changes the
assertion.
llvm-svn: 93065
(OP (trunc x), (trunc y)) -> (trunc (OP x, y))
Unfortunately this simple change causes dag combine to infinite looping. The problem is the shrink demanded ops optimization tend to canonicalize expressions in the opposite manner. That is badness. This patch disable those optimizations in dag combine but instead it is done as a late pass in sdisel.
This also exposes some deficiencies in dag combine and x86 setcc / brcond lowering. Teach them to look pass ISD::TRUNCATE in various places.
llvm-svn: 92849
(or (x << c) | (y >> (64 - c))) ==> (shld64 x, y, c)
The isel patterns may not catch all the cases if general dag combine has reduced width of source operands.
llvm-svn: 92513
return partial registers. This affected the back-end lowering code some.
Also patch up some places I missed before in the "get" functions.
llvm-svn: 91880
incrementing the simple value type of the 16-bit type, which would give the
wrong type if an intemediate MVT (such as i24) were introduced.
llvm-svn: 91602
Note that "hasDotLocAndDotFile"-style debug info was already broken;
people wanting this functionality should implement it in the
AsmPrinter/DwarfWriter code.
llvm-svn: 89711
The large code model is documented at
http://www.x86-64.org/documentation/abi.pdf and says that calls should
assume their target doesn't live within the 32-bit pc-relative offset
that fits in the call instruction.
To do this, we turn off the global-address->target-global-address
conversion in X86TargetLowering::LowerCall(). The first attempt at
this broke the lazy JIT because it can separate the movabs(imm->reg)
from the actual call instruction. The lazy JIT receives the address of
the movabs as a relocation and needs to record the return address from
the call; and then when that call happens, it needs to patch the
movabs with the newly-compiled target. We could thread the call
instruction into the relocation and record the movabs<->call mapping
explicitly, but that seems to require at least as much new
complication in the code generator as this change.
To fix this, we make lazy functions _always_ go through a call
stub. You'd think we'd only have to force lazy calls through a stub on
difficult platforms, but that turns out to break indirect calls
through a function pointer. The right fix for that is to distinguish
between calls and address-of operations on uncompiled functions, but
that's complex enough to leave for someone else to do.
Another attempt at this defined a new CALL64i pseudo-instruction,
which expanded to a 2-instruction sequence in the assembly output and
was special-cased in the X86CodeEmitter's emitInstruction()
function. That broke indirect calls in the same way as above.
This patch also removes a hack forcing Darwin to the small code model.
Without far-call-stubs, the small code model requires things of the
JITMemoryManager that the DefaultJITMemoryManager can't provide.
Thanks to echristo for lots of testing!
llvm-svn: 88984
slots. The AsmPrinter will use this information to determine whether to
print a spill/reload comment.
Remove default argument values. It's too easy to pass a wrong argument
value when multiple arguments have default values. Make everything
explicit to trap bugs early.
Update all targets to adhere to the new interfaces..
llvm-svn: 87022
1. rename the movhp patfrag to movlhps, since thats what it actually matches
2. eliminate the bogus movhps load and store patterns, they were incorrect. The load transforms are already handled (correctly) by shufps/unpack.
3. revert a recent test change to its correct form.
llvm-svn: 86415
encounters an OEQ or UNE comparison, and update its callers to check
for this return status and recover. This fixes a problem resulting from
the LowerOperation hooks being called from LegalizeVectorOps, because
LegalizeVectorOps only lowers vectors, so OEQ and UNE comparisons may
still be at large. This fixes PR5092.
llvm-svn: 84640
stack slots and giving them different PseudoSourceValue's did not fix the
problem of post-alloc scheduling miscompiling llvm itself.
- Apply Dan's conservative workaround by assuming any non fixed stack slots can
alias other memory locations. This means a load from spill slot #1 cannot
move above a store of spill slot #2.
- Enable post-alloc scheduling for x86 at optimization leverl Default and above.
llvm-svn: 84424
it to hold the address of an sret return value, for x86-64 ABI purposes.
Also, fix the test that was originally intended to test this to actually
test it, using FileCheck.
llvm-svn: 83853
- Allocate MachineMemOperands and MachineMemOperand lists in MachineFunctions.
This eliminates MachineInstr's std::list member and allows the data to be
created by isel and live for the remainder of codegen, avoiding a lot of
copying and unnecessary translation. This also shrinks MemSDNode.
- Delete MemOperandSDNode. Introduce MachineSDNode which has dedicated
fields for MachineMemOperands.
- Change MemSDNode to have a MachineMemOperand member instead of its own
fields with the same information. This introduces some redundancy, but
it's more consistent with what MachineInstr will eventually want.
- Ignore alignment when searching for redundant loads for CSE, but remember
the greatest alignment.
Target-specific code which previously used MemOperandSDNodes with generic
SDNodes now use MemIntrinsicSDNodes, with opcodes in a designated range
so that the SelectionDAG framework knows that MachineMemOperand information
is available.
llvm-svn: 82794
And fix a bug with the behavior of min/max instructions formed from
fcmp uge comparisons.
Also, use FiniteOnlyFPMath() for this code instead of UnsafeFPMath,
as it is more specific.
llvm-svn: 82466
getSymbolForDwarfGlobalReference is smart enough to know that it
needs to register the stub it references with MachineModuleInfoMachO,
so that it gets emitted at the end of the file.
Move stub emission from X86ATTAsmPrinter::doFinalization to the
new X86ATTAsmPrinter::EmitEndOfAsmFile asmprinter hook. The important
thing here is that EmitEndOfAsmFile is called *after* the ehframes are
emitted, so we get all the stubs.
This allows us to remove a gross hack from the asmprinter where it would
"just know" that it needed to output stubs for personality functions.
Now this is all driven from a consistent interface.
The testcase change is just reordering the expected output now that the
stubs come out after the ehframe instead of before.
This also unblocks other changes that Bill wants to make.
llvm-svn: 82269
Eliminate the PersonalityPrefix/Suffix & NeedsIndirectEncoding
fields from MAI: they aren't part of the asm syntax, they are
related to the structure of the object file.
To replace their functionality, add a new
TLOF::getSymbolForDwarfGlobalReference method which asks targets
to decide how to reference a global from EH in a pc-relative way.
The default implementation just returns the symbol. The default
darwin implementation references the symbol through an indirect
$non_lazy_ptr stub. The bizarro x86-64 darwin specialization
handles the weird "foo@GOTPCREL+4" hack.
DwarfException.cpp now uses this to emit the reference to the
symbol in the right way, and this also eliminates another
horrible hack from DwarfException.cpp:
- if (strcmp(MAI->getPersonalitySuffix(), "+4@GOTPCREL"))
- O << "-" << MAI->getPCSymbol();
llvm-svn: 81991
- when transforming a vector shift of a non-immediate scalar shift amount, zero
extend the i32 shift amount to i64 since the vector shift reads 64 bits
- when transforming i16 vectors to use a vector shift, zero extend i16 shift amount
- improve the code quality in some cases when transforming vectors to use a vector shift
llvm-svn: 80935
disabling the use of 16-bit operations on x86. This doesn't yet work for
inline asms with 16-bit constraints, vectors with 16-bit elements,
trampoline code, and perhaps other obscurities, but it's enough to try
some experiments.
llvm-svn: 80930
encodings.
- Make some of the values emitted by the FDEs dependent upon the pointer
size. This is in line with how GCC does things. And it has the benefit of
working for Darwin in 64-bit mode now.
llvm-svn: 80428
moves. This avoids the need to promote the operands (or implicitly
extend them, a partial register update condition), and can reduce
i8 register pressure. This substantially speeds up code such as
write_hex in lib/Support/raw_ostream.cpp.
subclass-coalesce.ll is too trivial and no longer tests what it was
originally intended to test.
llvm-svn: 80184
Add patterns and instruction encoding information.
Add custom lowering to deal with hardwired return register of
uncertain type (xmm0).
llvm-svn: 79377
support unaligned mem access only for certain types. (Should it be size
instead?)
ARM v7 supports unaligned access for i16 and i32, some v6 variants support it
as well.
llvm-svn: 79127
Instead of awkwardly encoding calling-convention information with ISD::CALL,
ISD::FORMAL_ARGUMENTS, ISD::RET, and ISD::ARG_FLAGS nodes, TargetLowering
provides three virtual functions for targets to override:
LowerFormalArguments, LowerCall, and LowerRet, which replace the custom
lowering done on the special nodes. They provide the same information, but
in a more immediately usable format.
This also reworks much of the target-independent tail call logic. The
decision of whether or not to perform a tail call is now cleanly split
between target-independent portions, and the target dependent portion
in IsEligibleForTailCallOptimization.
This also synchronizes all in-tree targets, to help enable future
refactoring and feature work.
llvm-svn: 78142
When the return value is not used (i.e. only care about the value in the memory), x86 does not have to use add to implement these. Instead, it can use add, sub, inc, dec instructions with the "lock" prefix.
This is currently implemented using a bit of instruction selection trick. The issue is the target independent pattern produces one output and a chain and we want to map it into one that just output a chain. The current trick is to select it into a merge_values with the first definition being an implicit_def. The proper solution is to add new ISD opcodes for the no-output variant. DAG combiner can then transform the node before it gets to target node selection.
Problem #2 is we are adding a whole bunch of x86 atomic instructions when in fact these instructions are identical to the non-lock versions. We need a way to add target specific information to target nodes and have this information carried over to machine instructions. Asm printer (or JIT) can use this information to add the "lock" prefix.
llvm-svn: 77582
it is highly specific to the object file that will be generated in the end,
this introduces a new TargetLoweringObjectFile interface that is implemented
for each of ELF/MachO/COFF/Alpha/PIC16 and XCore.
Though still is still a brutal and ugly refactoring, this is a major step
towards goodness.
This patch also:
1. fixes a bunch of dangling pointer problems in the PIC16 backend.
2. disables the TargetLowering copy ctor which PIC16 was accidentally using.
3. gets us closer to xcore having its own crazy target section flags and
pic16 not having to shadow sections with its own objects.
4. fixes wierdness where ELF targets would set CStringSection but not
CStringSection_. Factor the code better.
5. fixes some bugs in string lowering on ELF targets.
llvm-svn: 77294
be useful, and it's currently unused. (Some issues: it isn't actually
rich enough to capture the semantics on many architectures, and
semantics can vary depending on the type being shifted.)
llvm-svn: 76633
flags set properly. (hasMemory is clearly irrelevant
when matching 'i', I don't understand what this was
supposed to be doing.)
gcc.apple/asm-block-25.c (test passed before by
accident, but generated code was wrong)
llvm-svn: 76503
This adds location info for all llvm_unreachable calls (which is a macro now) in
!NDEBUG builds.
In NDEBUG builds location info and the message is off (it only prints
"UREACHABLE executed").
llvm-svn: 75640
Basically, using:
lea symbol(%rip), %rax
is not valid in -static mode, because the current RIP may not be
within 32-bits of "symbol" when an app is built partially pic and
partially static. The fix for this is to compile it to:
lea symbol, %rax
It would be better to codegen this as:
movq $symbol, %rax
but that will come next.
The hard part of fixing this bug was fixing abi-isel, which was actively
testing for the wrong behavior. Also, the RUN lines are completely impossible
to understand what they are testing. To help with this, convert the -static
x86-64 codegen tests to use filecheck. This is much more stable and makes it
more clear what the codegen is expected to be.
llvm-svn: 75382
Make llvm_unreachable take an optional string, thus moving the cerr<< out of
line.
LLVM_UNREACHABLE is now a simple wrapper that makes the message go away for
NDEBUG builds.
llvm-svn: 75379
With the SVR4 ABI on PowerPC, vector arguments for vararg calls are passed differently depending on whether they are a fixed or a variable argument. Variable vector arguments always go into memory, fixed vector arguments are put
into vector registers. If there are no free vector registers available, fixed vector arguments are put on the stack.
The NumFixedArgs attribute allows to decide for an argument in a vararg call whether it belongs to the fixed or variable portion of the parameter list.
llvm-svn: 74764
have the alignment be calculated up front, and have the back-ends obey whatever
alignment is decided upon.
This allows for future work that would allow for precise no-op placement and the
like.
llvm-svn: 74564
fence-atomic-fence down to just the atomic op. This is possible thanks to
X86's relatively strong memory model, which guarantees that locked instructions
(which are used to implement atomics) are implicit fences.
llvm-svn: 74435
implementation primarily differs from the former in that the asmprinter
doesn't make a zillion decisions about whether or not something will be
RIP relative or not. Instead, those decisions are made by isel lowering
and propagated through to the asm printer. To achieve this, we:
1. Represent RIP relative addresses by setting the base of the X86 addr
mode to X86::RIP.
2. When ISel Lowering decides that it is safe to use RIP, it lowers to
X86ISD::WrapperRIP. When it is unsafe to use RIP, it lowers to
X86ISD::Wrapper as before.
3. This removes isRIPRel from X86ISelAddressMode, representing it with
a basereg of RIP instead.
4. The addressing mode matching logic in isel is greatly simplified.
5. The asmprinter is greatly simplified, notably the "NotRIPRel" predicate
passed through various printoperand routines is gone now.
6. The various symbol printing routines in asmprinter now no longer infer
when to emit (%rip), they just print the symbol.
I think this is a big improvement over the previous situation. It does have
two small caveats though: 1. I implemented a horrible "no-rip" modifier for
the inline asm "P" constraint modifier. This is a short term hack, there is
a much better, but more involved, solution. 2. I had to xfail an
-aggressive-remat testcase because it isn't handling the use of RIP in the
constant-pool reading instruction. This specific test is easy to fix without
-aggressive-remat, which I intend to do next.
llvm-svn: 74372
out of sync with regular cc.
The only difference between the tail call cc and the normal
cc was that one parameter register - R9 - was reserved for
calling functions through a function pointer. After time the
tail call cc has gotten out of sync with the regular cc.
We can use R11 which is also caller saved but not used as
parameter register for potential function pointers and
remove the special tail call cc on x86-64.
llvm-svn: 73233
on x86 to handle more cases. Fix a bug in said code that would cause it
to read past the end of an object. Rewrite the code in
SelectionDAGLegalize::ExpandBUILD_VECTOR to be a bit more general.
Remove PerformBuildVectorCombine, which is no longer necessary with
these changes. In addition to simplifying the code, with this change,
we can now catch a few more cases of consecutive loads.
llvm-svn: 73012
nodes for vectors with an i16 element type. Add an optimization for
building a vector which is all zeros/undef except for the bottom
element, where the bottom element is an i8 or i16.
llvm-svn: 72988
Update code generator to use this attribute and remove NoImplicitFloat target option.
Update llc to set this attribute when -no-implicit-float command line option is used.
llvm-svn: 72959
build vectors with i64 elements will only appear on 32b x86 before legalize.
Since vector widening occurs during legalize, and produces i64 build_vector
elements, the dag combiner is never run on these before legalize splits them
into 32b elements.
Teach the build_vector dag combine in x86 back end to recognize consecutive
loads producing the low part of the vector.
Convert the two uses of TLI's consecutive load recognizer to pass LoadSDNodes
since that was required implicitly.
Add a testcase for the transform.
Old:
subl $28, %esp
movl 32(%esp), %eax
movl 4(%eax), %ecx
movl %ecx, 4(%esp)
movl (%eax), %eax
movl %eax, (%esp)
movaps (%esp), %xmm0
pmovzxwd %xmm0, %xmm0
movl 36(%esp), %eax
movaps %xmm0, (%eax)
addl $28, %esp
ret
New:
movl 4(%esp), %eax
pmovzxwd (%eax), %xmm0
movl 8(%esp), %eax
movaps %xmm0, (%eax)
ret
llvm-svn: 72957
ADDC/ADDE use MVT::i1 (later, whatever it gets legalized to)
instead of MVT::Flag. Remove CARRY_FALSE in favor of 0; adjust
all target-independent code to use this format.
Most targets will still produce a Flag-setting target-dependent
version when selection is done. X86 is converted to use i32
instead, which means TableGen needs to produce different code
in xxxGenDAGISel.inc. This keys off the new supportsHasI1 bit
in xxxInstrInfo, currently set only for X86; in principle this
is temporary and should go away when all other targets have
been converted. All relevant X86 instruction patterns are
modified to represent setting and using EFLAGS explicitly. The
same can be done on other targets.
The immediate behavior change is that an ADC/ADD pair are no
longer tightly coupled in the X86 scheduler; they can be
separated by instructions that don't clobber the flags (MOV).
I will soon add some peephole optimizations based on using
other instructions that set the flags to feed into ADC.
llvm-svn: 72707
e.g.
orl $65536, 8(%rax)
=>
orb $1, 10(%rax)
Since narrowing is not always a win, e.g. i32 -> i16 is a loss on x86, dag combiner consults with the target before performing the optimization.
llvm-svn: 72507
FP_TO_XINT. Necessary for some cleanups I'm working on. Updated
from the previous version (r72431) to fix a bug and make some things a
bit clearer.
llvm-svn: 72445
systems instead of attempting to promote them to a 64-bit SINT_TO_FP or
FP_TO_SINT. This is in preparation for removing the type legalization
code from LegalizeDAG: once type legalization is gone from LegalizeDAG,
it won't be able to handle the i64 operand/result correctly.
This isn't quite ideal, but I don't think any other operation for any
target ends up in this situation, so treating this case specially seems
reasonable.
llvm-svn: 72324
PR2957
ISD::VECTOR_SHUFFLE now stores an array of integers representing the shuffle
mask internal to the node, rather than taking a BUILD_VECTOR of ConstantSDNodes
as the shuffle mask. A value of -1 represents UNDEF.
In addition to eliminating the creation of illegal BUILD_VECTORS just to
represent shuffle masks, we are better about canonicalizing the shuffle mask,
resulting in substantially better code for some classes of shuffles.
llvm-svn: 70225
ISD::VECTOR_SHUFFLE now stores an array of integers representing the shuffle
mask internal to the node, rather than taking a BUILD_VECTOR of ConstantSDNodes
as the shuffle mask. A value of -1 represents UNDEF.
In addition to eliminating the creation of illegal BUILD_VECTORS just to
represent shuffle masks, we are better about canonicalizing the shuffle mask,
resulting in substantially better code for some classes of shuffles.
A clean up of x86 shuffle code, and some canonicalizing in DAGCombiner is next.
llvm-svn: 69952
in the MachineFunction class, renaming it to addLiveIn for consistency with
the same method in MachineBasicBlock. Thanks for Anton for suggesting this.
llvm-svn: 69615
leaq foo@TLSGD(%rip), %rdi
as part of the instruction sequence. Using a register other than %rdi and then
copying it to %rdi is not valid.
llvm-svn: 69350
with SUBREG_TO_REG, teach SimpleRegisterCoalescing to coalesce
SUBREG_TO_REG instructions (which are similar to INSERT_SUBREG
instructions), and teach the DAGCombiner to take advantage of this on
targets which support it. This eliminates many redundant
zero-extension operations on x86-64.
This adds a new TargetLowering hook, isZExtFree. It's similar to
isTruncateFree, except it only applies to actual definitions, and not
no-op truncates which may not zero the high bits.
Also, this adds a new optimization to SimplifyDemandedBits: transform
operations like x+y into (zext (add (trunc x), (trunc y))) on targets
where all the casts are no-ops. In contexts where the high part of the
add is explicitly masked off, this allows the mask operation to be
eliminated. Fix the DAGCombiner to avoid undoing these transformations
to eliminate casts on targets where the casts are no-ops.
Also, this adds a new two-address lowering heuristic. Since
two-address lowering runs before coalescing, it helps to be able to
look through copies when deciding whether commuting and/or
three-address conversion are profitable.
Also, fix a bug in LiveInterval::MergeInClobberRanges. It didn't handle
the case that a clobber range extended both before and beyond an
existing live range. In that case, multiple live ranges need to be
added. This was exposed by the new subreg coalescing code.
Remove 2008-05-06-SpillerBug.ll. It was bugpoint-reduced, and the
spiller behavior it was looking for no longer occurrs with the new
instruction selection.
llvm-svn: 68576
builds.
--- Reverse-merging (from foreign repository) r68552 into '.':
U test/CodeGen/X86/tls8.ll
U test/CodeGen/X86/tls10.ll
U test/CodeGen/X86/tls2.ll
U test/CodeGen/X86/tls6.ll
U lib/Target/X86/X86Instr64bit.td
U lib/Target/X86/X86InstrSSE.td
U lib/Target/X86/X86InstrInfo.td
U lib/Target/X86/X86RegisterInfo.cpp
U lib/Target/X86/X86ISelLowering.cpp
U lib/Target/X86/X86CodeEmitter.cpp
U lib/Target/X86/X86FastISel.cpp
U lib/Target/X86/X86InstrInfo.h
U lib/Target/X86/X86ISelDAGToDAG.cpp
U lib/Target/X86/AsmPrinter/X86ATTAsmPrinter.cpp
U lib/Target/X86/AsmPrinter/X86IntelAsmPrinter.cpp
U lib/Target/X86/AsmPrinter/X86ATTAsmPrinter.h
U lib/Target/X86/AsmPrinter/X86IntelAsmPrinter.h
U lib/Target/X86/X86ISelLowering.h
U lib/Target/X86/X86InstrInfo.cpp
U lib/Target/X86/X86InstrBuilder.h
U lib/Target/X86/X86RegisterInfo.td
llvm-svn: 68560
This introduces a small regression on the generated code
quality in the case we are just computing addresses, not
loading values.
Will work on it and on X86-64 support.
llvm-svn: 68552
x * 40
=>
shlq $3, %rdi
leaq (%rdi,%rdi,4), %rax
This has the added benefit of allowing more multiply to be folded into addressing mode. e.g.
a * 24 + b
=>
leaq (%rdi,%rdi,2), %rax
leaq (%rsi,%rax,8), %rax
llvm-svn: 67917
%a = ...
%b = and i32 %a, 2
%c = srl i32 %b, 1
%d = br i32 %c,
into
%a = ...
%b = and %a, 2
%c = X86ISD::CMP %b, 0
%d = X86ISD::BRCOND %c ...
This applies only when the AND constant value has one bit set and the SRL
constant is equal to the log2 of the AND constant. The back-end is smart enough
to convert the result into a TEST/JMP sequence.
llvm-svn: 67728
1. ConstantPoolSDNode alignment field is log2 value of the alignment requirement. This is not consistent with other SDNode variants.
2. MachineConstantPool alignment field is also a log2 value.
3. However, some places are creating ConstantPoolSDNode with alignment value rather than log2 values. This creates entries with artificially large alignments, e.g. 256 for SSE vector values.
4. Constant pool entry offsets are computed when they are created. However, asm printer group them by sections. That means the offsets are no longer valid. However, asm printer uses them to determine size of padding between entries.
5. Asm printer uses expensive data structure multimap to track constant pool entries by sections.
6. Asm printer iterate over SmallPtrSet when it's emitting constant pool entries. This is non-deterministic.
Solutions:
1. ConstantPoolSDNode alignment field is changed to keep non-log2 value.
2. MachineConstantPool alignment field is also changed to keep non-log2 value.
3. Functions that create ConstantPool nodes are passing in non-log2 alignments.
4. MachineConstantPoolEntry no longer keeps an offset field. It's replaced with an alignment field. Offsets are not computed when constant pool entries are created. They are computed on the fly in asm printer and JIT.
5. Asm printer uses cheaper data structure to group constant pool entries.
6. Asm printer compute entry offsets after grouping is done.
7. Change JIT code to compute entry offsets on the fly.
llvm-svn: 66875
for i32/i64 expressions (we could also do i16 on cpus where
i16 lea is fast, but I didn't add this). On the example, we now
generate:
_test:
movl 4(%esp), %eax
cmpl $42, (%eax)
setl %al
movzbl %al, %eax
leal 4(%eax,%eax,8), %eax
ret
instead of:
_test:
movl 4(%esp), %eax
cmpl $41, (%eax)
movl $4, %ecx
movl $13, %eax
cmovg %ecx, %eax
ret
llvm-svn: 66869
related transformations out of target-specific dag combine into the
ARM backend. These were added by Evan in r37685 with no testcases
and only seems to help ARM (e.g. test/CodeGen/ARM/select_xform.ll).
Add some simple X86-specific (for now) DAG combines that turn things
like cond ? 8 : 0 -> (zext(cond) << 3). This happens frequently
with the recently added cp constant select optimization, but is a
very general xform. For example, we now compile the second example
in const-select.ll to:
_test:
movsd LCPI2_0, %xmm0
ucomisd 8(%esp), %xmm0
seta %al
movzbl %al, %eax
movl 4(%esp), %ecx
movsbl (%ecx,%eax,4), %eax
ret
instead of:
_test:
movl 4(%esp), %eax
leal 4(%eax), %ecx
movsd LCPI2_0, %xmm0
ucomisd 8(%esp), %xmm0
cmovbe %eax, %ecx
movsbl (%ecx), %eax
ret
This passes multisource and dejagnu.
llvm-svn: 66779
the same say the "test" instruction does in overflow cases,
so eliminating the test is only safe when those bits aren't
needed, as is the case for COND_E and COND_NE, or if it
can be proven that no overflow will occur. For now, just
restrict the optimization to COND_E and COND_NE and don't
do any overflow analysis.
llvm-svn: 66318
instruction. The class also consolidates the code for detecting constant
splats that's shared across PowerPC and the CellSPU backends (and might be
useful for other backends.) Also introduces SelectionDAG::getBUID_VECTOR() for
generating new BUILD_VECTOR nodes.
llvm-svn: 65296
(Note: Eventually, commits like this will be handled via a pre-commit hook that
does this automagically, as well as expand tabs to spaces and look for 80-col
violations.)
llvm-svn: 64827
in inline asm as signed (what gcc does). Add partial support
for x86-specific "e" and "Z" constraints, with appropriate
signedness for printing.
llvm-svn: 64400
Many targets build placeholder nodes for special operands, e.g.
GlobalBaseReg on X86 and PPC for the PIC base. There's no
sensible way to associate debug info with these. I've left
them built with getNode calls with explicit DebugLoc::getUnknownLoc operands.
I'm not too happy about this but don't see a good improvement;
I considered adding a getPseudoOperand or something, but it
seems to me that'll just make it harder to read.
llvm-svn: 63992
crashes or wrong code with codegen of large integers:
eliminate the legacy getIntegerVTBitMask and
getIntegerVTSignBit methods, which returned their
value as a uint64_t, so couldn't handle huge types.
llvm-svn: 63494
dagcombines that help it match in several more cases. Add
several more cases to test/CodeGen/X86/bt.ll. This doesn't
yet include matching for BT with an immediate operand, it
just covers more register+register cases.
llvm-svn: 63266
tidy up SDUse and related code.
- Replace the operator= member functions with a set method, like
LLVM Use has, and variants setInitial and setNode, which take
care up updating use lists, like LLVM Use's does. This simplifies
code that calls these functions.
- getSDValue() is renamed to get(), as in LLVM Use, though most
places can either use the implicit conversion to SDValue or the
convenience functions instead.
- Fix some more node vs. value terminology issues.
Also, eliminate the one remaining use of SDOperandPtr, and
SDOperandPtr itself.
llvm-svn: 62995
X86. This code:
void f() {
uint32_t x;
float y = (float)x;
}
used to be:
movl %eax, -8(%ebp)
movl [2^52 double], -4(%ebp)
movsd -8(%ebp), %xmm0
subsd [2^52 double], %xmm0
cvtsd2ss %xmm0, %xmm0
Is now:
movsd [2^52 double], %xmm0
movsd %xmm0, %xmm1
movd %ecx, %xmm2
orps %xmm2, %xmm1
subsd %xmm0, %xmm1
cvtsd2ss %xmm1, %xmm0
This is faster on X86. Note that there's an extra load of %xmm0 into %xmm1. That
will be fixed in a later coalescer fix.
llvm-svn: 62404
promote from i1 all the way up to the canonical SetCC type.
In order to discover an appropriate type to use, pass
MVT::Other to getSetCCResultType. In order to be able to
do this, change getSetCCResultType to take a type as an
argument, not a value (this is also more logical).
llvm-svn: 61542
This removes all the _8, _16, _32, and _64 opcodes and replaces each
group with an unsuffixed opcode. The MemoryVT field of the AtomicSDNode
is now used to carry the size information. In tablegen, the size-specific
opcodes are replaced by size-independent opcodes that utilize the
ability to compose them with predicates.
This shrinks the per-opcode tables and makes the code that handles
atomics much more concise.
llvm-svn: 61389
which are identical to the original patterns.
- Change the multiply with overflow so that we distinguish between signed and
unsigned multiplication. Currently, unsigned multiplication with overflow
isn't working!
llvm-svn: 60963
ISD::ADD to emit an implicit EFLAGS. This was horribly broken. Instead, replace
the intrinsic with an ISD::SADDO node. Then custom lower that into an
X86ISD::ADD node with a associated SETCC that checks the correct condition code
(overflow or carry). Then that gets lowered into the correct X86::ADDOvf
instruction.
Similar for SUB and MUL instructions.
llvm-svn: 60915
loops when they can be subsumed into addressing modes.
Change X86 addressing mode check to realize that
some PIC references need an extra register.
(I believe this is correct for Linux, if not, I'm sure
someone will tell me.)
llvm-svn: 60608
- LowerXADDO lowers [SU]ADDO into an ADD with an implicit EFLAGS define. The
EFLAGS are fed into a SETCC node which has the conditional COND_O or COND_C,
depending on the type of ADDO requested.
- LowerBRCOND now recognizes if it's coming from a SETCC node with COND_O or
COND_C set.
llvm-svn: 60388
ReplaceNodeResults: rather than returning a node which
must have the same number of results as the original
node (which means mucking around with MERGE_VALUES,
and which is also easy to get wrong since SelectionDAG
folding may mean you don't get the node you expect),
return the results in a vector.
llvm-svn: 60348
the conditional for the BRCOND statement. For instance, it will generate:
addl %eax, %ecx
jo LOF
instead of
addl %eax, %ecx
; About 10 instructions to compare the signs of LHS, RHS, and sum.
jl LOF
llvm-svn: 60123
a memset using 16-byte XMM stores, but where the stack realignment code
didn't work. Until it does (PR2962) disable use of xmm regs in memcpy
and memset formation for linux and other targets with insufficiently
aligned stacks.
This is part of PR2888
llvm-svn: 58317
LHS is a foldable load, then LHS and RHS are swapped
and SetCCOpcode is changed to SETUGT. But the later
code is expecting operands to be the wrong way round
for SETUGT, but they are not in this case, resulting
in an inverted compare. The solution is to move the
load normalization before the correction for SETUGT.
This bug was tickled by LegalizeTypes which happened
to legalize the testcase slightly differently to
LegalizeDAG.
llvm-svn: 58092
assume that i64 has been turned into a BUILD_PAIR
node (when called from LegalizeTypes this hasn't
happened yet) and don't use a vector shuffle mask
with an illegal element type.
llvm-svn: 57972
The same one Apple gcc uses, faster. Also gets the
extreme case in gcc.c-torture/execute/ieee/rbug.c
correct which we weren't before; this is not
sufficient to get the test to pass though, there
is another bug.
llvm-svn: 57926
in the 32-bit signed offset field of addresses. Even though this
may be intended, some linkers refuse to relocate code where the
relocated address computation overflows.
Also, fix the sign-extension of constant offsets to use the
actual pointer size, rather than the size of the GlobalAddress
node, which may be different, for example on x86-64 where MVT::i32
is used when the address is being fit into the 32-bit displacement
field.
llvm-svn: 57885
Where previously LLVM might emit code like this:
ucomisd %xmm1, %xmm0
setne %al
setp %cl
orb %al, %cl
jne .LBB4_2
it now emits this:
ucomisd %xmm1, %xmm0
jne .LBB4_2
jp .LBB4_2
It has fewer instructions and uses fewer registers, but it does
have more branches. And in the case that this code is followed by
a non-fallthrough edge, it may be followed by a jmp instruction,
resulting in three branch instructions in sequence. Some effort
is made to avoid this situation.
To achieve this, X86ISelLowering.cpp now recognizes FCMP_OEQ and
FCMP_UNE in lowered form, and replace them with code that emits
two branches, except in the case where it would require converting
a fall-through edge to an explicit branch.
Also, X86InstrInfo.cpp's branch analysis and transform code now
knows now to handle blocks with multiple conditional branches. It
uses loops instead of having fixed checks for up to two
instructions. It can now analyze and transform code generated
from FCMP_OEQ and FCMP_UNE.
llvm-svn: 57873
LowerOperation if it doesn't know what else to do.
This methods should probably be factorized some,
but this is good enough for the moment. Have
LowerATOMIC_BINARY_64 use EXTRACT_ELEMENT rather
than assuming the operand is a BUILD_PAIR (if it
is then getNode will automagically simplify the
EXTRACT_ELEMENT). This way LowerATOMIC_BINARY_64
usable from LegalizeTypes.
llvm-svn: 57831
and add a TargetLowering hook for it to use to determine when this
is legal (i.e. not in PIC mode, etc.)
This allows instruction selection to emit folded constant offsets
in more cases, such as the included testcase, eliminating the need
for explicit arithmetic instructions.
This eliminates the need for the C++ code in X86ISelDAGToDAG.cpp
that attempted to achieve the same effect, but wasn't as effective.
Also, fix handling of offsets in GlobalAddressSDNodes in several
places, including changing GlobalAddressSDNode's offset from
int to int64_t.
The Mips, Alpha, Sparc, and CellSPU targets appear to be
unaware of GlobalAddress offsets currently, so set the hook to
false on those targets.
llvm-svn: 57748
in 32-bit mode instead of assigning a register pair. This has nothing to
do with PR2356, but I happened to notice it while working on it.
llvm-svn: 57704
i.e. conditions that cannot be checked with a single instruction. For example,
SETONE and SETUEQ on x86.
- Teach legalizer to implement *illegal* setcc as a and / or of a number of
legal setcc nodes. For now, only implement FP conditions. e.g. SETONE is
implemented as SETO & SETNE, SETUEQ is SETUO | SETEQ.
- Move x86 target over.
llvm-svn: 57542
- Move the EH landing-pad code and adjust it so that it works
with FastISel as well as with SDISel.
- Add FastISel support for @llvm.eh.exception and
@llvm.eh.selector.
llvm-svn: 57539
`-fno-builtin' flag. Currently, it's used to replace "memset" with "_bzero"
instead of "__bzero" on Darwin10+. This arguably violates the meaning of this
flag, but is currently sufficient. The meaning of this flag should become more
specific over time.
llvm-svn: 56885
its size). Adjust various lowering functions to
pass this info through from CallInst. Use it to
implement sseregparm returns on X86. Remove
X86_ssecall calling convention.
llvm-svn: 56677
s/ParamAttr/Attribute/g
s/PAList/AttrList/g
s/FnAttributeWithIndex/AttributeWithIndex/g
s/FnAttr/Attribute/g
This sets the stage
- to implement function notes as function attributes and
- to distinguish between function attributes and return value attributes.
This requires corresponding changes in llvm-gcc and clang.
llvm-svn: 56622
- Add linkage to SymbolSDNode (default to external).
- Change ISD::ExternalSymbol to ISD::Symbol.
- Change ISD::TargetExternalSymbol to ISD::TargetSymbol
These changes pave the way to allowing SymbolSDNodes with non-external linkage.
llvm-svn: 56249
isImmediate(), isRegister(), and friends, to avoid confusion
about having two different names with the same meaning. I'm
not attached to the longer names, and would be ok with
changing to the shorter names if others prefer it.
llvm-svn: 56189
Currently it just holds the calling convention and flags
for isVarArgs and isTailCall.
And it has several utility methods, which eliminate magic
5+2*i and similar index computations in several places.
CallSDNodes are not CSE'd. Teach UpdateNodeOperands to handle
nodes that are not CSE'd gracefully.
llvm-svn: 56183
cmp-and-swap reversed the Cmp and Swap arguments; comments
make it clear this is unintentional. Unfortunately, the
x86 BE had a compensating reversal, which is removed here.
PPC is OK.
From inspection of the Alpha code I think it is OK, but
if somebody has that platform please check it out. I
cannot test on that platform.
llvm-svn: 56091
HandlePHINodesInSuccessorBlocks that works FastISel-style. This
allows PHI nodes to be updated correctly while using FastISel.
This also involves some code reorganization; ValueMap and
MBBMap are now members of the FastISel class, so they needn't
be passed around explicitly anymore. Also, SelectInstructions
is changed to SelectInstruction, and only does one instruction
at a time.
llvm-svn: 55746
ATOMIC_LOAD_ADD_{8,16,32,64} instead of ATOMIC_LOAD_ADD.
Increased the Hardcoded Constant OpActionsCapacity to match.
Large but boring; no functional change.
This is to support partial-word atomics on ppc; i8 is
not a valid type there, so by the time we get to lowering, the
ATOMIC_LOAD nodes looks the same whether the type was i8 or i32.
The information can be added to the AtomicSDNode, but that is the
largest SDNode; I don't fully understand the SDNode allocation,
but it is sensitive to the largest node size, so increasing
that must be bad. This is the alternative.
llvm-svn: 55457
assign it to a version of the xmm register with the regclass that matches its
type. This fixes PR2715, a bug handling some crazy xpcom case in mozilla.
llvm-svn: 55358
1. x86-64 byval alignment should be max of 8 and alignment of type. Previously the code was not doing what the commit message was saying.
2. Do not use byte repeat move and store operations. These are slow.
llvm-svn: 55139
class hold a MachineRegisterInfo member, and make the
MachineBasicBlock be passed in to SelectInstructions rather
than the FastISel constructor.
llvm-svn: 55076
generic SDNode's (nodes with their own constructors
should do sanity checking in the constructor). Add
sanity checks for BUILD_VECTOR and fix all the places
that were producing bogus BUILD_VECTORs, as found by
"make check". My favorite is the BUILD_VECTOR with
only two operands that was being used to build a
vector with four elements!
llvm-svn: 53850
MachineMemOperands. The pools are owned by MachineFunctions.
This drastically reduces the number of calls to malloc/free made
during the "Emit" phase of scheduling, as well as later phases
in CodeGen. Combined with other changes, this speeds up the
"instruction selection" phase of CodeGen by 10% in some cases.
llvm-svn: 53212
hook for each way in which a result type can be
legalized (promotion, expansion, softening etc),
just use one: ReplaceNodeResults, which returns
a node with exactly the same result types as the
node passed to it, but presumably with a bunch of
custom code behind the scenes. No change if the
new LegalizeTypes infrastructure is not turned on.
llvm-svn: 53137
to be passed the list of value types, and use this
where appropriate. Inappropriate places are where
the value type list is already known and may be
long, in which case the existing method is more
efficient.
llvm-svn: 53035
the need for a flavor operand, and add a new SDNode subclass,
LabelSDNode, for use with them to eliminate the need for a label id
operand.
Change instruction selection to let these label nodes through
unmodified instead of creating copies of them. Teach the MachineInstr
emitter how to emit a MachineInstr directly from an ISD label node.
This avoids the need for allocating SDNodes for the label id and
flavor value, as well as SDNodes for each of the post-isel label,
label id, and label flavor.
llvm-svn: 52943
purpose, and give it a custom SDNode subclass so that it doesn't
need to have line number, column number, filename string, and
directory string, all existing as individual SDNodes to be the
operands.
This was the only user of ISD::STRING, StringSDNode, etc., so
remove those and some associated code.
This makes stop-points considerably easier to read in
-view-legalize-dags output, and reduces overhead (creating new
nodes and copying std::strings into them) on code containing
debugging information.
llvm-svn: 52924
it impossible to create a MERGE_VALUES node with
only one result: sometimes it is useful to be able
to create a node with only one result out of one of
the results of a node with more than one result, for
example because the new node will eventually be used
to replace a one-result node using ReplaceAllUsesWith,
cf X86TargetLowering::ExpandFP_TO_SINT. On the other
hand, most users of MERGE_VALUES don't need this and
for them the optimization was valuable. So add a new
utility method getMergeValues for creating MERGE_VALUES
nodes which by default performs the optimization.
Change almost everywhere to use getMergeValues (and
tidy some stuff up at the same time).
llvm-svn: 52893
Added abstract class MemSDNode for any Node that have an associated MemOperand
Changed atomic.lcs => atomic.cmp.swap, atomic.las => atomic.load.add, and
atomic.lss => atomic.load.sub
llvm-svn: 52706
shuffle could be skipped. The check is invalid because the loop index i
doesn't correspond to the element actually inserted. The correct check is
already done a few lines earlier, for whether the element is already in
the right spot, so this shouldn't have any effect on the codegen for
code that was already correct.
llvm-svn: 52486
of apint codegen failure is the DAG combiner doing
the wrong thing because it was comparing MVT's using
< rather than comparing the number of bits. Removing
the < method makes this mistake impossible to commit.
Instead, add helper methods for comparing bits and use
them.
llvm-svn: 52098
and better control the abstraction. Rename the type
to MVT. To update out-of-tree patches, the main
thing to do is to rename MVT::ValueType to MVT, and
rewrite expressions like MVT::getSizeInBits(VT) in
the form VT.getSizeInBits(). Use VT.getSimpleVT()
to extract a MVT::SimpleValueType for use in switch
statements (you will get an assert failure if VT is
an extended value type - these shouldn't exist after
type legalization).
This results in a small speedup of codegen and no
new testsuite failures (x86-64 linux).
llvm-svn: 52044
code generator would do something like this:
f64 = load f32 <anyext>, f32mem
v2f64 = insertelt undef, %0, 0
v2f64 = insertelt %1, 0.0, 1
into
v2f64 = vzext_load f32mem
which on x86 is movsd, when you really wanted a cvtss2sd/movsd pair.
llvm-svn: 51624
Move platform independent code (lowering of possibly overwritten
arguments, check for tail call optimization eligibility) from
target X86ISelectionLowering.cpp to TargetLowering.h and
SelectionDAGISel.cpp.
Initial PowerPC tail call implementation:
Support ppc32 implemented and tested (passes my tests and
test-suite llvm-test).
Support ppc64 implemented and half tested (passes my tests).
On ppc tail call optimization is performed if
caller and callee are fastcc
call is a tail call (in tail call position, call followed by ret)
no variable argument lists or byval arguments
option -tailcallopt is enabled
Supported:
* non pic tail calls on linux/darwin
* module-local tail calls on linux(PIC/GOT)/darwin(PIC)
* inter-module tail calls on darwin(PIC)
If constraints are not met a normal call will be emitted.
A test checking the argument lowering behaviour on x86-64 was added.
llvm-svn: 50477
- Make targetlowering.h fit in 80 cols.
- Make LowerAsmOperandForConstraint const.
- Make lowerXConstraint -> LowerXConstraint
- Make LowerXConstraint return a const char* instead of taking a string byref.
llvm-svn: 50312
On Darwin / Linux x86-32, v8i8, v4i16, v2i32 values are passed in MM[0-2].
On Darwin / Linux x86-32, v1i64 values are passed in memory.
On Darwin x86-64, v8i8, v4i16, v2i32 values are passed in XMM[0-7].
On Darwin x86-64, v1i64 values are passed in 64-bit GPRs.
llvm-svn: 50257
argument. The x86-64 ABI requires the incoming value of %rdi to
be copied to %rax on exit from a function that is returning a
large C struct.
Also, add a README-X86-64 entry detailing the missed optimization
opportunity and proposing an alternative approach.
llvm-svn: 50075
Rename SDOperandImpl back to SDOperand.
Introduce the SDUse class that represents a use of the SDNode referred by
an SDOperand. Now it is more similar to Use/Value classes.
Patch is approved by Dan Gohman.
llvm-svn: 49795
memcpy lowering code; this ensures that the size node has the desired
result type. This fixes a regression from r49572 with @llvm.memcpy.i64
on x86-32.
llvm-svn: 49761
optimized x86-64 (and x86) calls so that they work (... at least for
my test cases).
Should fix the following problems:
Problem 1: When i introduced the optimized handling of arguments for
tail called functions (using a sequence of copyto/copyfrom virtual
registers instead of always lowering to top of the stack) i did not
handle byval arguments correctly e.g they did not work at all :).
Problem 2: On x86-64 after the arguments of the tail called function
are moved to their registers (which include ESI/RSI etc), tail call
optimization performs byval lowering which causes xSI,xDI, xCX
registers to be overwritten. This is handled in this patch by moving
the arguments to virtual registers first and after the byval lowering
the arguments are moved from those virtual registers back to
RSI/RDI/RCX.
llvm-svn: 49584
on any current target and aren't optimized in DAGCombiner. Instead
of using intermediate nodes, expand the operations, choosing between
simple loads/stores, target-specific code, and library calls,
immediately.
Previously, the code to emit optimized code for these operations
was only used at initial SelectionDAG construction time; now it is
used at all times. This fixes some cases where rep;movs was being
used for small copies where simple loads/stores would be better.
This also cleans up code that checks for alignments less than 4;
let the targets make that decision instead of doing it in
target-independent code. This allows x86 to use rep;movs in
low-alignment cases.
Also, this fixes a bug that resulted in the use of rep;stos for
memsets of 0 with non-constant memory size when the alignment was
at least 4. It's better to use the library in this case, which
can be significantly faster when the size is large.
This also preserves more SourceValue information when memory
intrinsics are lowered into simple loads/stores.
llvm-svn: 49572
LLVM Value/Use does and MachineRegisterInfo/MachineOperand does.
This allows constant time for all uses list maintenance operations.
The idea was suggested by Chris. Reviewed by Evan and Dan.
Patch is tested and approved by Dan.
On normal use-cases compilation speed is not affected. On very big basic
blocks there are compilation speedups in the range of 15-20% or even better.
llvm-svn: 48822
flags. This is needed by the new legalize types
infrastructure which wants to expand the 64 bit
constants previously used to hold the flags on
32 bit machines. There are two functional changes:
(1) in LowerArguments, if a parameter has the zext
attribute set then that is marked in the flags;
before it was being ignored; (2) PPC had some bogus
code for handling two word arguments when using the
ELF 32 ABI, which was hard to convert because of
the bogusness. As suggested by the original author
(Nicolas Geoffray), I've disabled it for the moment.
Tested with "make check" and the Ada ACATS testsuite.
llvm-svn: 48640
1. There is now a "PAListPtr" class, which is a smart pointer around
the underlying uniqued parameter attribute list object, and manages
its refcount. It is now impossible to mess up the refcount.
2. PAListPtr is now the main interface to the underlying object, and
the underlying object is now completely opaque.
3. Implementation details like SmallVector and FoldingSet are now no
longer part of the interface.
4. You can create a PAListPtr with an arbitrary sequence of
ParamAttrsWithIndex's, no need to make a SmallVector of a specific
size (you can just use an array or scalar or vector if you wish).
5. All the client code that had to check for a null pointer before
dereferencing the pointer is simplified to just access the
PAListPtr directly.
6. The interfaces for adding attrs to a list and removing them is a
bit simpler.
Phase #2 will rename some stuff (e.g. PAListPtr) and do other less
invasive changes.
llvm-svn: 48289
field to 32 bits, thus enabling correct handling of ByVal
structs bigger than 0x1ffff. Abstract interface a bit.
Fixes gcc.c-torture/execute/pr23135.c and
gcc.c-torture/execute/pr28982b.c in gcc testsuite (were ICE'ing
on ppc32, quietly producing wrong code on x86-32.)
llvm-svn: 48122
into a vector of zeros or undef, and when the top part is obviously
zero, we can just use movd + shuffle. This allows us to compile
vec_set-B.ll into:
_test3:
movl $1234567, %eax
andl 4(%esp), %eax
movd %eax, %xmm0
ret
instead of:
_test3:
subl $28, %esp
movl $1234567, %eax
andl 32(%esp), %eax
movl %eax, (%esp)
movl $0, 4(%esp)
movq (%esp), %xmm0
addl $28, %esp
ret
llvm-svn: 48090
2) Don't try to insert an i64 value into the low part of a
vector with movq on an x86-32 target. This allows us to
compile:
__m128i doload64(short x) {return _mm_set_epi16(0,0,0,0,0,0,0,1);}
into:
_doload64:
movaps LCPI1_0, %xmm0
ret
instead of:
_doload64:
subl $28, %esp
movl $0, 4(%esp)
movl $1, (%esp)
movq (%esp), %xmm0
addl $28, %esp
ret
llvm-svn: 48057
stack slot and store if the SINT_TO_FP is actually legal. This allows
us to compile:
double a(double b) {return (unsigned)b;}
to:
_a:
cvttsd2siq %xmm0, %rax
movl %eax, %eax
cvtsi2sdq %rax, %xmm0
ret
instead of:
_a:
subq $8, %rsp
cvttsd2siq %xmm0, %rax
movl %eax, %eax
cvtsi2sdq %rax, %xmm0
addq $8, %rsp
ret
crazy.
llvm-svn: 47660
GOT-style position independent code. Before only tail calls to
protected/hidden functions within the same module were optimized.
Now all function calls are tail call optimized.
llvm-svn: 47594
calls. Before arguments that could overwrite each other were
explicitly lowered to a stack slot, not giving the register allocator
a chance to optimize. Now a sequence of copyto/copyfrom virtual
registers ensures that arguments are loaded in (virtual) registers
before they are lowered to the stack slot (and might overwrite each
other). Also parameter stack slots are marked mutable for
(potentially) tail calling functions.
llvm-svn: 47593
instead of with mmx registers. This horribleness is apparently
done by gcc to avoid having to insert emms in places that really
should have it. This is the second half of rdar://5741668.
llvm-svn: 47474
GCC apparently does this, and code depends on not having to do
emms when this happens. This is x86-64 only so far, second half
should handle x86-32.
rdar://5741668
llvm-svn: 47470
has plain one-result scalar integer multiplication instructions.
This avoids expanding such instructions into MUL_LOHI sequences that
must be special-cased at isel time, and avoids the problem with that
code that provented memory operands from being folded.
This fixes PR1874, addressesing the most common case. The uncommon
cases of optimizing multiply-high operations will require work
in DAGCombiner.
llvm-svn: 47277
the return value is zero-extended if it isn't
sign-extended. It may also be any-extended.
Also, if a floating point value was returned
in a larger floating point type, pass 1 as the
second operand to FP_ROUND, which tells it
that all the precision is in the original type.
I think this is right but I could be wrong.
Finally, when doing libcalls, set isZExt on
a parameter if it is "unsigned". Currently
isSExt is set when signed, and nothing is
set otherwise. This should be right for all
calls to standard library routines.
llvm-svn: 47122
1) ConstantFP is now expand by default
2) ConstantFP is not turned into TargetConstantFP during Legalize
if it is legal.
This allows ConstantFP to be handled like Constant, allowing for
targets that can encode FP immediates as MachineOperands.
As a bonus, fix up Itanium FP constants, which now correctly match,
and match more constants! Hooray.
llvm-svn: 47121
initializer problem, a minor tweak to the way the
DAGISelEmitter finds load/store nodes, and a renaming of the
new PseudoSourceValue objects.
llvm-svn: 46827
Added ISD::DECLARE node type to represent llvm.dbg.declare intrinsic. Now the intrinsic calls are lowered into a SDNode and lives on through out the codegen passes.
For now, since all the debugging information recording is done at isel time, when a ISD::DECLARE node is selected, it has the side effect of also recording the variable. This is a short term solution that should be fixed in time.
llvm-svn: 46659
in the backend. Introduce a new SDNode type, MemOperandSDNode, for
holding a MemOperand in the SelectionDAG IR, and add a MemOperand
list to MachineInstr, and code to manage them. Remove the offset
field from SrcValueSDNode; uses of SrcValueSDNode that were using
it are all all using MemOperandSDNode now.
Also, begin updating some getLoad and getStore calls to use the
PseudoSourceValue objects.
Most of this was written by Florian Brander, some
reorganization and updating to TOT by me.
llvm-svn: 46585
This case returns the value in ST(0) and then has to convert it to an SSE
register. This causes significant codegen ugliness in some cases. For
example in the trivial fp-stack-direct-ret.ll testcase we used to generate:
_bar:
subl $28, %esp
call L_foo$stub
fstpl 16(%esp)
movsd 16(%esp), %xmm0
movsd %xmm0, 8(%esp)
fldl 8(%esp)
addl $28, %esp
ret
because we move the result of foo() into an XMM register, then have to
move it back for the return of bar.
Instead of hacking ever-more special cases into the call result lowering code
we take a much simpler approach: on x86-32, fp return is modeled as always
returning into an f80 register which is then truncated to f32 or f64 as needed.
Similarly for a result, we model it as an extension to f80 + return.
This exposes the truncate and extensions to the dag combiner, allowing target
independent code to hack on them, eliminating them in this case. This gives
us this code for the example above:
_bar:
subl $12, %esp
call L_foo$stub
addl $12, %esp
ret
The nasty aspect of this is that these conversions are not legal, but we want
the second pass of dag combiner (post-legalize) to be able to hack on them.
To handle this, we lie to legalize and say they are legal, then custom expand
them on entry to the isel pass (PreprocessForFPConvert). This is gross, but
less gross than the code it is replacing :)
This also allows us to generate better code in several other cases. For
example on fp-stack-ret-conv.ll, we now generate:
_test:
subl $12, %esp
call L_foo$stub
fstps 8(%esp)
movl 16(%esp), %eax
cvtss2sd 8(%esp), %xmm0
movsd %xmm0, (%eax)
addl $12, %esp
ret
where before we produced (incidentally, the old bad code is identical to what
gcc produces):
_test:
subl $12, %esp
call L_foo$stub
fstpl (%esp)
cvtsd2ss (%esp), %xmm0
cvtss2sd %xmm0, %xmm0
movl 16(%esp), %eax
movsd %xmm0, (%eax)
addl $12, %esp
ret
Note that we generate slightly worse code on pr1505b.ll due to a scheduling
deficiency that is unrelated to this patch.
llvm-svn: 46307
precision integers. This won't actually work
(and most of the code is dead) unless the new
legalization machinery is turned on. While
there, I rationalized the handling of i1, and
removed some bogus (and unused) sextload patterns.
For i1, this could result in microscopically
better code for some architectures (not X86).
It might also result in worse code if annotating
with AssertZExt nodes turns out to be more harmful
than helpful.
llvm-svn: 46280
1. Legalize now always promotes truncstore of i1 to i8.
2. Remove patterns and gunk related to truncstore i1 from targets.
3. Rename the StoreXAction stuff to TruncStoreAction in TLI.
4. Make the TLI TruncStoreAction table a 2d table to handle from/to conversions.
5. Mark a wide variety of invalid truncstores as such in various targets, e.g.
X86 currently doesn't support truncstore of any of its integer types.
6. Add legalize support for truncstores with invalid value input types.
7. Add a dag combine transform to turn store(truncate) into truncstore when
safe.
The later allows us to compile CodeGen/X86/storetrunc-fp.ll to:
_foo:
fldt 20(%esp)
fldt 4(%esp)
faddp %st(1)
movl 36(%esp), %eax
fstps (%eax)
ret
instead of:
_foo:
subl $4, %esp
fldt 24(%esp)
fldt 8(%esp)
faddp %st(1)
fstps (%esp)
movl 40(%esp), %eax
movss (%esp), %xmm0
movss %xmm0, (%eax)
addl $4, %esp
ret
llvm-svn: 46140
and switch various codegen pieces and the X86 backend over
to using it.
* Add some comments to SelectionDAGNodes.h
* Introduce a second argument to FP_ROUND, which indicates
whether the FP_ROUND changes the value of its input. If
not it is safe to xform things like fp_extend(fp_round(x)) -> x.
llvm-svn: 46125
it should work, but I have no machine to test
it on. Committed because it will at least
cause no harm, and maybe someone can test it
for me!
llvm-svn: 46098
make the 'fp return in ST(0)' optimization smart enough to
look through token factor nodes. THis allows us to compile
testcases like CodeGen/X86/fp-stack-retcopy.ll into:
_carg:
subl $12, %esp
call L_foo$stub
fstpl (%esp)
fldl (%esp)
addl $12, %esp
ret
instead of:
_carg:
subl $28, %esp
call L_foo$stub
fstpl 16(%esp)
movsd 16(%esp), %xmm0
movsd %xmm0, 8(%esp)
fldl 8(%esp)
addl $28, %esp
ret
Still not optimal, but much better and this is a trivial patch. Fixing
the rest requires invasive surgery that is is not llvm 2.2 material.
llvm-svn: 46054
commit all arguments where moved to the stack slot where they would
reside on a normal function call before the lowering to the tail call
stack slot. This was done to prevent arguments overwriting each other.
Now only arguments sourcing from a FORMAL_ARGUMENTS node or a
CopyFromReg node with virtual register (could also be a caller's
argument) are lowered indirectly.
--This line, and those below, will be ignored--
M X86/X86ISelLowering.cpp
M X86/README.txt
llvm-svn: 45867
unifying the copied algorithms and saving over 500 LOC. There should
be no functionality change, but please test on your favorite x86
target.
llvm-svn: 45627
that "machine" classes are used to represent the current state of
the code being compiled. Given this expanded name, we can start
moving other stuff into it. For now, move the UsedPhysRegs and
LiveIn/LoveOuts vectors from MachineFunction into it.
Update all the clients to match.
This also reduces some needless #includes, such as MachineModuleInfo
from MachineFunction.
llvm-svn: 45467
e.g. MO.isMBB() instead of MO.isMachineBasicBlock(). I don't plan on
switching everything over, so new clients should just start using the
shorter names.
Remove old long accessors, switching everything over to use the short
accessor: getMachineBasicBlock() -> getMBB(),
getConstantPoolIndex() -> getIndex(), setMachineBasicBlock -> setMBB(), etc.
llvm-svn: 45464
SelectionDAG::getConstant, in the same way as vector floating-point
constants. This allows the legalize expansion code for @llvm.ctpop and
friends to be usable with vector types.
llvm-svn: 44954
possible before resorting to pextrw and pinsrw.
- Better codegen for v4i32 shuffles masquerading as v8i16 or v16i8 shuffles.
- Improves (i16 extract_vector_element 0) codegen by recognizing
(i32 extract_vector_element 0) does not require a pextrw.
llvm-svn: 44836
the function type, instead they belong to functions
and function calls. This is an updated and slightly
corrected version of Reid Spencer's original patch.
The only known problem is that auto-upgrading of
bitcode files doesn't seem to work properly (see
test/Bitcode/AutoUpgradeIntrinsics.ll). Hopefully
a bitcode guru (who might that be? :) ) will fix it.
llvm-svn: 44359
sometimes emit "zero" and "all one" vectors multiple times,
for example:
_test2:
pcmpeqd %mm0, %mm0
movq %mm0, _M1
pcmpeqd %mm0, %mm0
movq %mm0, _M2
ret
instead of:
_test2:
pcmpeqd %mm0, %mm0
movq %mm0, _M1
movq %mm0, _M2
ret
This patch fixes this by always arranging for zero/one vectors
to be defined as v4i32 or v2i32 (SSE/MMX) instead of letting them be
any random type. This ensures they get trivially CSE'd on the dag.
This fix is also important for LegalizeDAGTypes, as it gets unhappy
when the x86 backend wants BUILD_VECTOR(i64 0) to be legal even when
'i64' isn't legal.
This patch makes the following changes:
1) X86TargetLowering::LowerBUILD_VECTOR now lowers 0/1 vectors into
their canonical types.
2) The now-dead patterns are removed from the SSE/MMX .td files.
3) All the patterns in the .td file that referred to immAllOnesV or
immAllZerosV in the wrong form now use *_bc to match them with a
bitcast wrapped around them.
4) X86DAGToDAGISel::SelectScalarSSELoad is generalized to handle
bitcast'd zero vectors, which simplifies the code actually.
5) getShuffleVectorZeroOrUndef is updated to generate a shuffle that
is legal, instead of generating one that is illegal and expecting
a later legalize pass to clean it up.
6) isZeroShuffle is generalized to handle bitcast of zeros.
7) several other minor tweaks.
This patch is definite goodness, but has the potential to cause random
code quality regressions. Please be on the lookout for these and let
me know if they happen.
llvm-svn: 44310
1) Change the interface to TargetLowering::ExpandOperationResult to
take and return entire NODES that need a result expanded, not just
the value. This allows us to handle things like READCYCLECOUNTER,
which returns two values.
2) Implement (extremely limited) support in LegalizeDAG::ExpandOp for MERGE_VALUES.
3) Reimplement custom lowering in LegalizeDAGTypes in terms of the new
ExpandOperationResult. This makes the result simpler and fully
general.
4) Implement (fully general) expand support for MERGE_VALUES in LegalizeDAGTypes.
5) Implement ExpandOperationResult support for ARM f64->i64 bitconvert and ARM
i64 shifts, allowing them to work with LegalizeDAGTypes.
6) Implement ExpandOperationResult support for X86 READCYCLECOUNTER and FP_TO_SINT,
allowing them to work with LegalizeDAGTypes.
LegalizeDAGTypes now passes several more X86 codegen tests when enabled and when
type legalization in LegalizeDAG is ifdef'd out.
llvm-svn: 44300
adjustment fields, and an optional flag. If there is a "dynamic_stackalloc" in
the code, make sure that it's bracketed by CALLSEQ_START and CALLSEQ_END. If
not, then there is the potential for the stack to be changed while the stack's
being used by another instruction (like a call).
This can only result in tears...
llvm-svn: 44037
transformation. Previously, it's restricted by ensuring the number of load uses
is one. Now the restriction is loosened up by allowing setcc uses to be
"extended" (e.g. setcc x, c, eq -> setcc sext(x), sext(c), eq).
llvm-svn: 43465
To do this it is necessary to add a "always inline" argument to the
memcpy node. For completeness I have also added this node to memmove
and memset. I have also added getMem* functions, because the extra
argument makes it cumbersome to use getNode and because I get confused
by it :-)
llvm-svn: 43172
was stored to the acutal stack slot before the parameters were
lowered to their stack slot. This could cause arguments to be
overwritten by the return address if the called function had less
parameters than the caller function. The update should remove the
last failing test case of llc-beta: SPASS.
llvm-svn: 43027
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
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
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
both results with a single div or idiv instruction. This uses new X86ISD
nodes for DIV and IDIV which are introduced during the legalize phase
so that the SelectionDAG's CSE can automatically eliminate redundant
computations.
llvm-svn: 42308
asm to die with:
ScheduleDAG.cpp:269: failed assertion `false && "Couldn't find the register class"'
instead of:
failed assertion `RegMap->getRegClass(VReg) == RC && "Register class of operand and regclass of use don't agree!"'
yay.
llvm-svn: 42259
keep f32 in SSE registers and f64 in x87. This
is effectively a new codegen mode.
Change addLegalFPImmediate to permit float and
double variants to do different things.
Adjust callers.
llvm-svn: 42246
Craig Topper