This patch defines an interleaved-load-combine pass. The pass searches
for ShuffleVector instructions that represent interleaved loads. Matches are
converted such that they will be captured by the InterleavedAccessPass.
The pass extends LLVMs capabilities to use target specific instruction
selection of interleaved load patterns (e.g.: ld4 on Aarch64
architectures).
Differential Revision: https://reviews.llvm.org/D52653
llvm-svn: 347208
Truncs are treated as sources if their produce a value of the same
type as the one we currently trying to promote. Truncs used to be
considered as a sink if their operand was the same value type.
We now allow smaller types in the search, so we should search through
truncs that produce a smaller value. These truncs can then be
converted to an AND mask.
This leaves sinks as being:
- points where the value in the register is being observed, such as
an icmp, switch or store.
- points where value types have to match, such as calls and returns.
- zext are included to ease the transformation and are generally
removed later on.
During this change, it also became apart from truncating sinks was
broken: if a sink used a source, its type information had already
been lost by the time the truncation happens. So I've changed the
method of caching the type information.
Differential Revision: https://reviews.llvm.org/D54515
llvm-svn: 347191
There is no variable-length shifts on MSP430. Therefore
"eat" 8 bits of shift via bswap & ext.
Path by Kristina Bessonova!
Differential Revision: https://reviews.llvm.org/D54623
llvm-svn: 347187
The shift requires a copy to avoid clobbering a register. Comparing with 0 uses an xor to produce 0 that will be overwritten with the compare results. So still requires 2 instructions, but should be one byte shorter since it doesn't need to encode an immediate.
llvm-svn: 347185
Previously we used an arithmetic shift right by 31, but that requires a copy to preserve the input. So we might as well materialize a zero and compare to it since the comparison will overwrite the register that contains the zeros. This should be one byte shorter.
llvm-svn: 347181
Leave just the v4i8->v4i64 and v8i8->v8i64, but only enable them on pre-sse4.1 targets when 64-bit mode is enabled. In those cases we end up creating sext loads that get scalarized to code that looks better than what we get from loading into a vector register and doing a multiple step sign extend using unpacks and shifts.
llvm-svn: 347180
Pre-SSE4.1 sext_invec for v2i64 is complicated because we don't have a v2i64 sra instruction. So instead we sign extend to i32 using unpack and sra, then copy the elements and do a v4i32 sra to fill with sign bits, then interleave the i32 sign extend and the sign bits. So really we're doing to two sign extends but only using half of the v4i32 intermediate result.
When the result is more than 128 bits, default type legalization would prefer to split the destination type all the way down to v2i64 with shuffles followed by v16i8/v8i16->v2i64 sext_inreg operations. This results in more instructions than necessary because we are only utilizing the lower 2 elements of the v4i32 intermediate result. Instead we can custom split a v4i8/v4i16->v4i64 sign_extend. Then we can sign extend v4i8/v4i16->v4i32 invec producing a full v4i32 result. Create the sign bit vector as a v4i32 then split and interleave with the sign bits using an punpackldq and punpackhdq.
llvm-svn: 347176
Some of these sequeces look pretty bad since we have to copy the sign bit from a 32 bit register to a 64 bit register to finish a sign extend.
llvm-svn: 347175
If we widen illegal types instead of promoting, we should be able to rely on the type legalizer to create the vector_inreg operations for us with some caveats.
This patch disables combineToExtendVectorInReg when we are using widening.
I've enabled custom legalization for v8i8->v8i64 extends under avx512f since the type legalizer would want to create a vector_inreg with a v64i8 input type which isn't legal without avx512bw. So we go to v16i8 with custom code using the relaxation of rules we get from D54346.
I've also enable custom legalization of v8i64 and v16i32 operations with with AVX. When the input type is 128 bits, the default splitting legalization would extend first 128->256, then do the a split to two 128 pieces. Extend each half to 256 and then concat the result. The custom legalization I've added instead uses a 128->256 bit vector_inreg extend that only reads the lower 64-bits for the low half of the split. Then shuffles the high 64-bits to the low 64-bits and does another vector_inreg extend.
llvm-svn: 347172
Summary: This is an improvement over the two pshufbs and punpcklqdq we'd get otherwise.
Reviewers: RKSimon, spatel
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D54671
llvm-svn: 347171
Sadly, this duplicates (twice) the logic from InstSimplify. There
might be some way to at least share the DAG versions of the code,
but copying the folds seems to be the standard method to ensure
that we don't miss these folds.
Unlike in IR, we don't run DAGCombiner to fixpoint, so there's no
way to ensure that we do these kinds of simplifications unless the
code is repeated at node creation time and during combines.
There were other tests that would become worthless with this
improvement that I changed as pre-commits:
rL347161
rL347164
rL347165
rL347166
rL347167
I'm not sure how to salvage the remaining tests (diffs in this patch).
So the x86 tests verify that the new code is working as intended.
The AMDGPU test is actually similar to my motivating case: we have
some undef value that has survived to machine IR in an x86 test, and
then it gets folded in some weird way, or we crash if we don't transfer
the undef flag. But we would have been better off never getting to that
point by doing these simplifications.
This will lead back to PR32023 someday...
https://bugs.llvm.org/show_bug.cgi?id=32023
llvm-svn: 347170
We were using the 'normalized' shuffle mask from resolveTargetShuffleInputs, which replaces zero/undef inputs with sentinel values. For SimplifyDemandedVectorElts we need the raw mask so we can correctly demand those 'zero' inputs that got normalized away, this requires an extra bit of logic to locally normalize undef inputs.
llvm-svn: 347158
The zero extend will require two stages of unpacks to implement. So its better to shrink the multiply using pmullw and then extend that result back to v4i32 using a single unpack.
llvm-svn: 347149
Comments in llc outputs are printed differently on different
platforms, some with '#', some with '##'. Removed non-essential
part of the checks.
llvm-svn: 347112
This tries to force the result type to vXi32 followed by a truncate. This can help avoid scalarization that would otherwise occur.
There's some annoying examples of an avx512 truncate instruction followed by a packus where we should really be able to just use one truncate. But overall this is still a net improvement.
llvm-svn: 347105
In one case probably you have be using it, in the other it
looks like it was redundant.
Differential Revision: https://reviews.llvm.org/D54644
llvm-svn: 347098
This NFC patch just adds test cases for conversions that currently
require scalarization of vectors. An updcoming patch will change
the legalization for these and it is more suitable on the review
to show the diferences in code gen rather than just the new code gen.
llvm-svn: 347090
When unwinding past a function that uses shadow call stack, we must
subtract 8 from the value of the x18 register. This patch causes us
to emit a call frame instruction that causes that to happen.
Differential Revision: https://reviews.llvm.org/D54609
llvm-svn: 347089
Summary:
As discussed in previous review, and noted in the FIXME, if `X` is actually an `lshr Y, Z` (logical!),
we can fold the `Z` into 'control`, and let the `BEXTR` do this too.
We could just insert those 8 bits of shift amount into control,
but it is better to instead zero-extend them, and 'or' them in place.
We can only do this for `lshr`, not `ashr`, because we do not know that the mask cover only the bits of `Y`,
and not any of the sign-extended bits.
The obvious question is, is this actually legal to do?
I believe it is. Relevant quotes, from `Intel® 64 and IA-32 Architectures Software Developer’s Manual`, `BEXTR — Bit Field Extract`:
* `Bit 7:0 of the second source operand specifies the starting bit position of bit extraction.`
* `A START value exceeding the operand size will not extract any bits from the second source operand.`
* `Only bit positions up to (OperandSize -1) of the first source operand are extracted.`
* `All higher order bits in the destination operand (starting at bit position LENGTH) are zeroed.`
* `The destination register is cleared if no bits are extracted.`
FIXME: if we can do this, i wonder if we should prefer `BEXTR` over `BZHI` in such cases.
Reviewers: RKSimon, craig.topper, spatel, andreadb
Reviewed By: RKSimon, craig.topper, andreadb
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D54095
llvm-svn: 347048
This commit introduces support for materialising 64-bit constants for RV64I,
making use of the RISCVMatInt::generateInstSeq helper in order to share logic
for immediate materialisation with the MC layer (where it's used for the li
pseudoinstruction).
test/CodeGen/RISCV/imm.ll is updated to test RV64, and gains new 64-bit
constant tests. It would be preferable if anyext constant returns were sign
rather than zero extended (see PR39092). This patch simply adds an explicit
signext to the returns in imm.ll.
Further optimisations for constant materialisation are possible, most notably
for mask-like values which can be generated my loading -1 and shifting right.
A future patch will standardise on the C++ codepath for immediate selection on
RV32 as well as RV64, and then add further such optimisations to
RISCVMatInt::generateInstSeq in order to benefit both RV32 and RV64 for
codegen and li expansion.
Differential Revision: https://reviews.llvm.org/D52962
llvm-svn: 347042
Sanjay Patel