This new version is much more aggressive about doing "full" reduction in
cases where it reduces register pressure, and also more aggressive about
rewriting induction variables to count down (or up) to zero when doing so
reduces register pressure.
It currently uses fairly simplistic algorithms for finding reuse
opportunities, but it introduces a new framework allows it to combine
multiple strategies at once to form hybrid solutions, instead of doing
all full-reduction or all base+index.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@94061 91177308-0d34-0410-b5e6-96231b3b80d8
of int initializers), change some methods to be static functions,
use raw_ostream::write_hex instead of a smallstring dance with
APValue::toStringUnsigned(S, 16).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93991 91177308-0d34-0410-b5e6-96231b3b80d8
doing global variable classification anymore) and hookized, sink almost
all target targets global variable emission code into AsmPrinter and out
of each target.
Some notes:
1. PIC16 does completely custom and crazy stuff, so it is not changed.
2. XCore has some custom handling for extra directives. I'll look at it next.
3. This switches linux/ppc to use .globl instead of .global. If .globl is
actually wrong, let me know and I'll fix it.
4. This makes linux/ppc get a lot of random cases right which were obviously
wrong before, it is probably now a bit healthier.
5. Blackfin will probably start getting .comm and other things that it didn't
before. If this is undesirable, it should explicitly opt out of these
things by clearing the relevant fields of MCAsmInfo.
This leads to a nice diffstat:
14 files changed, 127 insertions(+), 830 deletions(-)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93858 91177308-0d34-0410-b5e6-96231b3b80d8
GCC would put weak zero initialized mutable data in the .bss section,
we would put it into a crasy '.gnu.linkonce.b.test,"aw",@nobits'
section. Fixing this will allow simplifications next up.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93844 91177308-0d34-0410-b5e6-96231b3b80d8
Instcombine does this but apparently there are situations where this pattern will escape the optimizer and / or created by isel. Here is a case that's seen in JavaScriptCore:
%t1 = sub i32 0, %a
%t2 = add i32 %t1, -1
The dag combiner pattern: ((c1-A)+c2) -> (c1+c2)-A
will fold it to -1 - %a.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93773 91177308-0d34-0410-b5e6-96231b3b80d8
adding an "i" to the suffix, indicating that the elements are integers, is
accepted but not part of the standard syntax. This helps us pass a few more
of the Neon tests from gcc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93677 91177308-0d34-0410-b5e6-96231b3b80d8
different BlockAddress labels, but nothing semantically important.
Add a FIXME that BlockAddress codegen is broken if the LLVM BB has
an empty name (e.g. strip was run).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93303 91177308-0d34-0410-b5e6-96231b3b80d8
For now, this pass is fairly conservative. It only perform the replacement when both the pre- and post- extension values are used in the block. It will miss cases where the post-extension values are live, but not used.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93278 91177308-0d34-0410-b5e6-96231b3b80d8
has an immediate with at least 32 bits of leading zeros, to avoid needing to
materialize that immediate in a register first.
FileCheckize, tidy, and extend a testcase to cover this case.
This fixes rdar://7527390.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93160 91177308-0d34-0410-b5e6-96231b3b80d8
new AsmPrinter. This is perhaps less elegant than describing them
in terms of MOV32r0 and subreg operations, but it allows the
current register to rematerialize them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93158 91177308-0d34-0410-b5e6-96231b3b80d8
ignore alignment requirements for SIMD memory operands. This
is useful on architectures like the AMD 10h that do not trap on
unaligned references if a status bit is twiddled at startup time.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93151 91177308-0d34-0410-b5e6-96231b3b80d8
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93065 91177308-0d34-0410-b5e6-96231b3b80d8
really does need to be a vector type, because
TargetLowering::getOperationAction for SIGN_EXTEND_INREG uses that type,
and it needs to be able to distinguish between vectors and scalars.
Also, fix some more issues with legalization of vector casts.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93043 91177308-0d34-0410-b5e6-96231b3b80d8
When folding a and(any_ext(load)) both the any_ext and the
load have to have only a single use.
This removes the anyext-uses.ll testcase which started failing
because it is unreduced and unclear what it is testing.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@92950 91177308-0d34-0410-b5e6-96231b3b80d8
(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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@92849 91177308-0d34-0410-b5e6-96231b3b80d8
(X != null) | (Y != null) --> (X|Y) != 0
(X == null) & (Y == null) --> (X|Y) == 0
so that instcombine can stop doing this for pointers. This is part of PR3351,
which is a case where instcombine doing this for pointers (inserting ptrtoint)
is pessimizing code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@92406 91177308-0d34-0410-b5e6-96231b3b80d8
multiply sequence when the power is a constant integer. Before, our
codegen for std::pow(.., int) always turned into a libcall, which was
really inefficient.
This should also make many gfortran programs happier I'd imagine.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@92388 91177308-0d34-0410-b5e6-96231b3b80d8
compare. On other targets we end up with a call to memcmp because we don't
want 16 individual byte loads. We should be able to use movups as well, but
we're failing to select the generated icmp.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@92107 91177308-0d34-0410-b5e6-96231b3b80d8
SDISel. This optimization was causing simplifylibcalls to
introduce type-unsafe nastiness. This is the first step, I'll be
expanding the memcmp optimizations shortly, covering things that
we really really wouldn't want simplifylibcalls to do.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@92098 91177308-0d34-0410-b5e6-96231b3b80d8
be non-optimal. To be precise, we should avoid folding loads if the instructions
only update part of the destination register, and the non-updated part is not
needed. e.g. cvtss2sd, sqrtss. Unfolding the load from these instructions breaks
the partial register dependency and it can improve performance. e.g.
movss (%rdi), %xmm0
cvtss2sd %xmm0, %xmm0
instead of
cvtss2sd (%rdi), %xmm0
An alternative method to break dependency is to clear the register first. e.g.
xorps %xmm0, %xmm0
cvtss2sd (%rdi), %xmm0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91672 91177308-0d34-0410-b5e6-96231b3b80d8
The change in SelectionDAGBuilder is needed to allow using bitcasts to convert
between f64 (the default type for ARM "d" registers) and 64-bit Neon vector
types. Radar 7457110.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91649 91177308-0d34-0410-b5e6-96231b3b80d8
This protects this test from depending on codegen not performing the
tail call optimization by default.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91648 91177308-0d34-0410-b5e6-96231b3b80d8
Fold (zext (and x, cst)) -> (and (zext x), cst)
DAG combiner likes to optimize expression in the other way so this would end up cause an infinite looping.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91574 91177308-0d34-0410-b5e6-96231b3b80d8
in local register allocator. If a reg-reg copy has a phys reg
input and a virt reg output, and this is the last use of the phys
reg, assign the phys reg to the virt reg. If a reg-reg copy has
a phys reg output and we need to reload its spilled input, reload
it directly into the phys reg than passing it through another reg.
Following 76208, there is sometimes no dependency between the def of
a phys reg and its use; this creates a window where that phys reg
can be used for spilling (this is true in linear scan also). This
is bad and needs to be fixed a better way, although 76208 works too
well in practice to be reverted. However, there should normally be
no spilling within inline asm blocks. The patch here goes a long way
towards making this actually be true.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91485 91177308-0d34-0410-b5e6-96231b3b80d8
1. Only perform (zext (shl (zext x), y)) -> (shl (zext x), y) when y is a constant. This makes sure it remove at least one zest.
2. If the shift is a left shift, make sure the original shift cannot shift out bits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91399 91177308-0d34-0410-b5e6-96231b3b80d8
The coalescer is supposed to clean these up, but when setting up parameters
for a function call, there may be copies to physregs. If the defining
instruction has been LICM'ed far away, the coalescer won't touch it.
The register allocation hint does not always work - when the register
allocator is backtracking, it clears the hints.
This patch takes care of a few more cases that r90163 missed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90502 91177308-0d34-0410-b5e6-96231b3b80d8
Add a testcase for the above transformation.
Fix a bogus use of APInt noticed while tracking this down.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90423 91177308-0d34-0410-b5e6-96231b3b80d8
both source operands. In the canonical form, the 2nd operand is changed to an
undef and the shuffle mask is adjusted to only reference elements from the 1st
operand. Radar 7434842.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90417 91177308-0d34-0410-b5e6-96231b3b80d8
- A valno should be set HasRedefByEC if there is an early clobber def in the middle of its live ranges. It should not be set if the def of the valno is defined by an early clobber.
- If a physical register def is tied to an use and it's an early clobber, it just means the HasRedefByEC is set since it's still one continuous live range.
- Add a couple of missing checks for HasRedefByEC in the coalescer. In general, it should not coalesce a vr with a physical register if the physical register has a early clobber def somewhere. This is overly conservative but that's the price for using such a nasty inline asm "feature".
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90269 91177308-0d34-0410-b5e6-96231b3b80d8
This means that well connected blocks are copy coalesced before the less connected blocks. Connected blocks are more difficult to
coalesce because intervals are more complicated, so handling them first gives a greater chance of success.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90194 91177308-0d34-0410-b5e6-96231b3b80d8
This helps us avoid silly copies when rematting values that are copied to a physical register:
leaq _.str44(%rip), %rcx
movq %rcx, %rsi
call _strcmp
becomes:
leaq _.str44(%rip), %rsi
call _strcmp
The coalescer will not touch the movq because that would tie down the physical register.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90163 91177308-0d34-0410-b5e6-96231b3b80d8
the problem only shows for msp430 and pic16 which is why it specifies
them using -march. But it is wrong to put such tests in CodeGen/Generic,
since not everyone builds these targets. Put a copy of the test in each
of the target test directories.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90005 91177308-0d34-0410-b5e6-96231b3b80d8
This violates the ABI (that area is "reserved"), and
while it is safe if all code is generated with current
compilers, there is some very old code around that uses
that slot for something else, and breaks if it is stored
into. Adjust testcases looking for current behavior.
I've verified that the stack frame size is right in all
testcases, whether it changed or not. 7311323.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@89811 91177308-0d34-0410-b5e6-96231b3b80d8
than doing the same via constpool:
1. Load from constpool costs 3 cycles on A9, movt/movw pair - just 2.
2. Load from constpool might stall up to 300 cycles due to cache miss.
3. Movt/movw does not use load/store unit.
4. Less constpool entries => better compiler performance.
This is only enabled on ELF systems, since darwin does not have needed
relocations (yet).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@89720 91177308-0d34-0410-b5e6-96231b3b80d8