SplitKit will soon need two copies of these data structures, and the
algorithms will also be useful when LiveIntervalAnalysis becomes
independent of LiveVariables.
llvm-svn: 139572
Splitting a landing pad takes considerable care because of PHIs and other
nasties. The problem is that the jump table needs to jump to the landing pad
block. However, the landing pad block can be jumped to only by an invoke
instruction. So we clone the landingpad instruction into its own basic block,
have the invoke jump to there. The landingpad instruction's basic block's
successor is now the target for the jump table.
But because of PHI nodes, we need to create another basic block for the jump
table to jump to. This is definitely a hack, because the values for the PHI
nodes may not be defined on the edge from the jump table. But that's okay,
because the jump table is simply a construct to mimic what is happening in the
CFG. So the values are mysteriously there, even though there is no value for the
PHI from the jump table's edge (hence calling this a hack).
llvm-svn: 139545
No tests; these changes aren't really interesting in the sense that the logic is the same for volatile and atomic.
I believe this completes all of the changes necessary for the optimizer to handle loads and stores correctly. I'm going to try and come up with some additional testing, though.
llvm-svn: 139533
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
SplitKit always computes a complement live range to cover the places
where the original live range was live, but no explicit region has been
allocated.
Currently, the complement live range is created to be as small as
possible - it never overlaps any of the regions. This minimizes
register pressure, but if the complement is going to be spilled anyway,
that is not very important. The spiller will eliminate redundant
spills, and hoist others by making the spill slot live range overlap
some of the regions created by splitting. Stack slots are cheap.
This patch adds the interface to enable spill modes in SplitKit. In
spill mode, SplitKit will assume that the complement is going to spill,
so it will allow it to overlap regions in order to avoid back-copies.
By doing some of the spiller's work early, the complement live range
becomes simpler. In some cases, it can become much simpler because no
extra PHI-defs are required. This will speed up both splitting and
spilling.
This is only the interface to enable spill modes, no implementation yet.
llvm-svn: 139500
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
better.
Don't immediately give up when an add operation can't be trivially
sign/zero-extended within a loop. If it has NSW/NUW flags, generate a
new expression with sign extended (non-recurrent) operand. As before,
if SCEV says that all sign extends are loop invariant, then we can
widen the operation.
llvm-svn: 139453
such.
I'm doing this now for completeness because I can't think of/remember
any reason that it was left out. I'm not sure it will help anything,
but if we don't do it we need to explain why in comments.
llvm-svn: 139450
In some cases such as interpreters using indirectbr, the CFG can be very
complicated, and live range splitting may be forced to insert a large
number of phi-defs. When that happens, traceSiblingValue can spend a
lot of time zipping around in the CFG looking for defs and reloads.
This patch causes more information to be cached in SibValues, and the
cached values are used to terminate searches early. This speeds up
spilling by 20x in one interpreter test case. For more typical code,
this is just a 10% speedup of spilling.
The previous version had bugs that caused miscompilations. They have
been fixed.
llvm-svn: 139378
single field (Flags), which is a bitwise OR of items from the TB_*
enum. This makes it easier to add new information in the future.
* Gives every static array an equivalent layout: { RegOp, MemOp, Flags }
* Adds a helper function, AddTableEntry, to avoid duplication of the
insertion code.
* Renames TB_NOT_REVERSABLE to TB_NO_REVERSE.
* Adds TB_NO_FORWARD, which is analogous to TB_NO_REVERSE, except that
it prevents addition of the Reg->Mem entry. (This is going to be used
by Native Client, in the next CL).
Patch by David Meyer
llvm-svn: 139311
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
The immediate offset of the non-writeback i8 form (encoding T4) allows
negative offsets only. The positive offset form of the encoding is the
LDRT instruction. Immediate offsets in the range [0,255] use encoding T3
instead.
llvm-svn: 139254
In some cases such as interpreters using indirectbr, the CFG can be very
complicated, and live range splitting may be forced to insert a large
number of phi-defs. When that happens, traceSiblingValue can spend a
lot of time zipping around in the CFG looking for defs and reloads.
This patch causes more information to be cached in SibValues, and the
cached values are used to terminate searches early. This speeds up
spilling by 20x in one interpreter test case. For more typical code,
this is just a 10% speedup of spilling.
llvm-svn: 139247
There is no 16-bit wide encoding, so the .w suffix isn't needed (indeed, isn't
documented as allowed). Also add the missing '!' token on the _UPD
variant.
llvm-svn: 139243
Choose 32-bit vs. 16-bit encoding when there's no .w suffix in post-processing
as match classes are insufficient to handle the context-sensitiveness of
the writeback operand's legality for the 16-bit encodings.
llvm-svn: 139242
(The fix for the related failures on x86 is going to be nastier because we actually need Acquire memoperands attached to the atomic load instrs, etc.)
llvm-svn: 139221
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
Now the 'S' instructions, e.g. ADDS, treat S bit as optional operand as well.
Also fix isel hook to correctly set the optional operand.
rdar://10073745
llvm-svn: 139157
Even if there's no mode switch performed, the .code directive should still
be sent to the output streamer. Otherwise, for example, an output asm stream
is not equivalent to the input stream which generated it (a dependency on
the input target triple arm vs. thumb is introduced which was not originally
there).
llvm-svn: 139155
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
instructions are more aligned than the CPU requires, and adds some additional
directives, to follow in future patches. Patch by David Meyer!
llvm-svn: 139125
- Drop support for X >u 0, it's equivalent to X != 0 and should be canonicalized into the latter.
- Add X < 1 -> unlikely, which is what instcombine canonicalizes X <= 0 into.
- Add X > -1 -> likely, which is what instcombine canonicalizes X >= 0 into.
llvm-svn: 139110
call. The call may be in the same BB as the landingpad instruction. If that's
the case, then inserting the loads after the landingpad inst, but before the
extractvalues, causes undefined behavior.
llvm-svn: 139088
If we have a chain of zext -> assert_zext -> zext -> use, the first zext would get simplified away because of the later zext, and then the later zext would get simplified away because of the assert. The solution is to teach SimplifyDemandedBits that assert_zext demands all of the high bits of its input, rather than only those demanded by its users. No testcase because the only example I have manifests as llvm-gcc miscompiling LLVM, and I haven't found a smaller case that reproduces this problem.
Fixes <rdar://problem/10063365>.
llvm-svn: 139059
The explanation about a 0 argument being materialized as xor is no
longer valid. Rematerialization will check if EFLAGS is live before
clobbering it.
The code produced by X86TargetLowering::EmitLoweredSelect does not
clobber EFLAGS.
This causes one less testb instruction to be generated in the cmov.ll
test case.
llvm-svn: 139057
slots. This fixes a bug where the number of nodes coming into the PHI node may
not equal the number of predecessors. E.g., two or more landingpad instructions
may require a PHI before reaching the eh.exception and eh.selector instructions.
llvm-svn: 139035
This changes loop unrolling to use the same mechanism for trip count
computation as indvars. This is a stronger check that tends to unroll
more loops. A very common side-effect is that many single iteration
loops will be removed sooner. The real goal was simply to remove
dependence on canonical IVs.
x86 is break even.
ARM performance changes to expect (+ is good):
External/SPEC/CFP2000/183.equake/183.equake +13%
SingleSource/Benchmarks/Dhrystone/fldry +21%
MultiSource/Applications/spiff/spiff +3%
SingleSource/Benchmarks/Stanford/Puzzle -14%
The Puzzle regression is actually an improvement in loop optimization
that defeats GVN: rdar://problem/10065079.
llvm-svn: 139009
Perform the upgrading in steps.
* First, create a map of the invokes to the EH intrinsics.
* Next, take that mapping and determine if the invoke's unwind destination has a
single predecessor. If not, then create a new empty block to hold the new
landingpad instruction.
* Create a landingpad instruction into the uwnind destination. Fill it with the
values from the old selector. Map the old intrinsic calls to the new
landingpad values (there may be multiple landingpad instructions per instrinic
call pairs).
* Go through the old intrinsic calls, create a PHI node when necessary, and then
replace their values with the new values from the landingpad instructions.
* Delete all dead instructions.
* ???
* Profit!
llvm-svn: 138990
to be unreliable on platforms which require memcpy calls, and it is
complicating broader legalize cleanups. It is hoped that these cleanups
will make memcpy byval easier to implement in the future.
llvm-svn: 138977
- On COFF the .lcomm directive has an alignment argument.
- On ELF we fall back to .local + .comm
Based on a patch by NAKAMURA Takumi.
Fixes PR9337, PR9483 and PR10128.
llvm-svn: 138976
An instruction may define part of a register where the other bits are
undefined. In that case, it is safe to rematerialize the instruction.
For example:
%vreg2:ssub_0<def> = VLDRS <cp#0>, 0, pred:14, pred:%noreg, %vreg2<imp-def>
The extra <imp-def> operand indicates that the instruction does not read
the other parts of the virtual register, so a remat is safe.
This patch simply allows multiple def operands for the virtual register.
It is MI->readsVirtualRegister() that determines if we depend on a
previous value so remat is impossible.
llvm-svn: 138953
