Rotating by c, then by d, is the same as rotating by c+d.
Change-Id: I36df82261460ff80f7c6d39bcdf0e840cef1c91a
Reviewed-on: https://go-review.googlesource.com/c/go/+/424894
Reviewed-by: Wayne Zuo <wdvxdr@golangcn.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Ruinan Sun <Ruinan.Sun@arm.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Currently we use a full cmpstring to do the comparison for each
split in the binary search for a string switch.
Instead, split by comparing a single byte of the input string with a
constant. That will give us a much faster split (although it might be
not quite as good a split).
Fixes#53333
R=go1.20
Change-Id: I28c7209342314f367071e4aa1f2beb6ec9ff7123
Reviewed-on: https://go-review.googlesource.com/c/go/+/414894
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
The prove pass will mark some shifts bounded, and then we can use that
information to generate better code on riscv64.
Change-Id: Ia22f43d0598453c9417adac7017db28d7240948b
Reviewed-on: https://go-review.googlesource.com/c/go/+/422616
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joel Sing <joel@sing.id.au>
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
This is a follow up of CL 425101 on RISCV64.
According to RISCV Volume 1, Unprivileged Spec v. 20191213 Chapter 7.1:
If both the high and low bits of the same product are required, then the
recommended code sequence is: MULH[[S]U] rdh, rs1, rs2; MUL rdl, rs1, rs2
(source register specifiers must be in same order and rdh cannot be the
same as rs1 or rs2). Microarchitectures can then fuse these into a single
multiply operation instead of performing two separate multiplies.
So we should not split Muluhilo to separate instructions.
Updates #54607
Change-Id: If47461f3aaaf00e27cd583a9990e144fb8bcdb17
Reviewed-on: https://go-review.googlesource.com/c/go/+/425203
Auto-Submit: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Detect rotate instructions while still in architecture-independent form.
It's easier to do here, and we don't need to repeat it in each
architecture file.
Change-Id: I9396954b3f3b3bfb96c160d064a02002309935bb
Reviewed-on: https://go-review.googlesource.com/c/go/+/421195
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Eric Fang <eric.fang@arm.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: Ruinan Sun <Ruinan.Sun@arm.com>
Run-TryBot: Keith Randall <khr@golang.org>
Previously we convert $0 to the ZR register for some reasons, which causes
two problems:
1. Confusion, the special case of the ZR register needs to be considered
when dealing with constants. For encoding, some places we encode ZR, and
some places we encode $0, although we have converted $0 to ZR.
2. Unexpected instruction format. All instructions that support ZR register
operands can be replaced by $0.
This patch removes this conversion. Note that this patch may cause previously
unintendedly supported instruction formats to no longer be supported.
Change-Id: I3d8d2c06711b7614a38191397da7776417f1861c
Reviewed-on: https://go-review.googlesource.com/c/go/+/404316
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Eric Fang <eric.fang@arm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
On ARM64 we use two separate instructions to compute the hi and lo
results of a 64x64->128 multiplication. Lower to two separate ops
so if only one result is needed we can deadcode the other.
Fixes#54607.
Change-Id: Ib023e77eb2b2b0bcf467b45471cb8a294bce6f90
Reviewed-on: https://go-review.googlesource.com/c/go/+/425101
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
With the introduction of stack objects, VARKILL information is
no longer needed.
With stack objects, an object is dead when there are no more static
references to it, and the stack scanner can't find any live pointers
to it. VARKILL information isn't used to establish live ranges for
address-taken variables any more. In effect, the last static reference
*is* the VARKILL, and there's an additional dynamic liveness check
during stack scanning.
Next CL will actually rip out the VARKILL opcodes.
Change-Id: I030a2ab867445cf4e0e69397911f8a2e2f0ed07b
Reviewed-on: https://go-review.googlesource.com/c/go/+/419234
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Updated multiple tests in test/codegen: math.go, mathbits.go, shift.go
and slices.go to verify on ppc64/ppc64le as well
Change-Id: Id88dd41569b7097819fb4d451b615f69cf7f7a94
Reviewed-on: https://go-review.googlesource.com/c/go/+/412115
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Paul Murphy <murp@ibm.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
For signed comparisons, the following four optimization rules hold:
(CMPconst [0] z:(AND x y)) && z.Uses == 1 => (TST x y)
(CMPWconst [0] z:(AND x y)) && z.Uses == 1 => (TSTW x y)
(CMPconst [0] x:(ANDconst [c] y)) && x.Uses == 1 => (TSTconst [c] y)
(CMPWconst [0] x:(ANDconst [c] y)) && x.Uses == 1 => (TSTWconst [int32(c)] y)
But currently they only apply to jump instructions, not to conditional
instructions within a block, such as cset, csel, etc. This CL extends
the above rules into blocks so that conditional instructions can also be
optimized.
name old time/op new time/op delta
DivisiblePow2constI64-160 1.04ns ± 0% 0.86ns ± 0% -17.30% (p=0.008 n=5+5)
DivisiblePow2constI32-160 1.04ns ± 0% 0.87ns ± 0% -16.16% (p=0.016 n=4+5)
DivisiblePow2constI16-160 1.04ns ± 0% 0.87ns ± 0% -16.03% (p=0.008 n=5+5)
DivisiblePow2constI8-160 1.04ns ± 0% 0.86ns ± 0% -17.15% (p=0.008 n=5+5)
Change-Id: I6bc34bff30862210e8dd001e0340b8fe502fe3de
Reviewed-on: https://go-review.googlesource.com/c/go/+/420434
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Eric Fang <eric.fang@arm.com>
As it can't appear in user package paths.
There is a hack for handling "go:buildid" and "type:*" on windows/386.
Previously, windows/386 requires underscore prefix on external symbols,
but that's only applied for SHOSTOBJ/SUNDEFEXT or cgo export symbols.
"go.buildid" is STEXT, "type.*" is STYPE, thus they are not prefixed
with underscore.
In external linking mode, the external linker can't resolve them as
external symbols. But we are lucky that they have "." in their name,
so the external linker see them as Forwarder RVA exports. See:
- https://docs.microsoft.com/en-us/windows/win32/debug/pe-format#export-address-table
- https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=ld/pe-dll.c;h=e7b82ba6ffadf74dc1b9ee71dc13d48336941e51;hb=HEAD#l972)
This CL changes "." to ":" in symbols name, so theses symbols can not be
found by external linker anymore. So a hacky way is adding the
underscore prefix for these 2 symbols. I don't have enough knowledge to
verify whether adding the underscore for all STEXT/STYPE symbols are
fine, even if it could be, that would be done in future CL.
Fixes#37762
Change-Id: I92eaaf24c0820926a36e0530fdb07b07af1fcc35
Reviewed-on: https://go-review.googlesource.com/c/go/+/317917
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently there is a an ANDconst and an ANDCCconst op in PPC64,
which is confusing since they map onto the same instruction.
One of these ops sets the result of the AND operation, and the
other sets the flag (condition register).
This converts ANDCCconst into an op with the 2 expected results:
the integer result of the AND and the flag setting. The ANDconst
op has been removed.
Note that in the PPC64 ISA the only variation of the 'and immediate'
is the one that sets the condition bit, which probably led to the
original (confusing) implementation.
This also adds a few rules to improve the use of ANDCCconst with
ISELB and some testcases to verify those improvements.
Change-Id: I523703fa4da2098eb995dc3ba744d36fa28e41d4
Reviewed-on: https://go-review.googlesource.com/c/go/+/422015
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Paul Murphy <murp@ibm.com>
We don't need a multiply when the element type is size 0 or 1.
The panic functions don't return, so we don't need any post-call
code (register restores, etc.).
Change-Id: I0dcea5df56d29d7be26554ddca966b3903c672e5
Reviewed-on: https://go-review.googlesource.com/c/go/+/419754
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Since CL 391014, cmd/compile now requires the -p flag to be set the
build system. This CL changes it to initialize LocalPkg.Path to the
provided path, rather than relying on writing out `"".` into object
files and expecting cmd/link to substitute them.
However, this actually involved a rather long tail of fixes. Many have
already been submitted, but a few notable ones that have to land
simultaneously with changing LocalPkg:
1. When compiling package runtime, there are really two "runtime"
packages: types.LocalPkg (the source package itself) and
ir.Pkgs.Runtime (the compiler's internal representation, for synthetic
references). Previously, these ended up creating separate link
symbols (`"".xxx` and `runtime.xxx`, respectively), but now they both
end up as `runtime.xxx`, which causes lsym collisions (notably
inittask and funcsyms).
2. test/codegen tests need to be updated to expect symbols to be named
`command-line-arguments.xxx` rather than `"".foo`.
3. The issue20014 test case is sensitive to the sort order of field
tracking symbols. In particular, the local package now sorts to its
natural place in the list, rather than to the front.
Thanks to David Chase for helping track down all of the fixes needed
for this CL.
Updates #51734.
Change-Id: Iba3041cf7ad967d18c6e17922fa06ba11798b565
Reviewed-on: https://go-review.googlesource.com/c/go/+/393715
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
math/bits.Add64 and math/bits.Sub64 now lower and optimize
directly in SSA form.
The optimization of carry chains focuses around eliding
XER<->GPR transfers of the CA bit when used exclusively as an
input to a single carry operations, or when the CA value is
known.
This also adds support for handling XER spills in the assembler
which could happen if carry chains contain inter-dependencies
on each other (which seems very unlikely with practical usage),
or a clobber happens (SRAW/SRAD/SUBFC operations clobber CA).
With PPC64 Add64/Sub64 lowering into SSA and this patch, the net
performance difference in crypto/elliptic benchmarks on P9/ppc64le
are:
name old time/op new time/op delta
ScalarBaseMult/P256 46.3µs ± 0% 46.9µs ± 0% +1.34%
ScalarBaseMult/P224 356µs ± 0% 209µs ± 0% -41.14%
ScalarBaseMult/P384 1.20ms ± 0% 0.57ms ± 0% -52.14%
ScalarBaseMult/P521 3.38ms ± 0% 1.44ms ± 0% -57.27%
ScalarMult/P256 199µs ± 0% 199µs ± 0% -0.17%
ScalarMult/P224 357µs ± 0% 212µs ± 0% -40.56%
ScalarMult/P384 1.20ms ± 0% 0.58ms ± 0% -51.86%
ScalarMult/P521 3.37ms ± 0% 1.44ms ± 0% -57.32%
MarshalUnmarshal/P256/Uncompressed 2.59µs ± 0% 2.52µs ± 0% -2.63%
MarshalUnmarshal/P256/Compressed 2.58µs ± 0% 2.52µs ± 0% -2.06%
MarshalUnmarshal/P224/Uncompressed 1.54µs ± 0% 1.40µs ± 0% -9.42%
MarshalUnmarshal/P224/Compressed 1.54µs ± 0% 1.39µs ± 0% -9.87%
MarshalUnmarshal/P384/Uncompressed 2.40µs ± 0% 1.80µs ± 0% -24.93%
MarshalUnmarshal/P384/Compressed 2.35µs ± 0% 1.81µs ± 0% -23.03%
MarshalUnmarshal/P521/Uncompressed 3.79µs ± 0% 2.58µs ± 0% -31.81%
MarshalUnmarshal/P521/Compressed 3.80µs ± 0% 2.60µs ± 0% -31.67%
Note, P256 uses an asm implementation, thus, little variation is expected.
Change-Id: I88a24f6bf0f4f285c649e40243b1ab69cc452b71
Reviewed-on: https://go-review.googlesource.com/c/go/+/346870
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
It is hit ~70k times building go.
This make the go binary, 0.04% smaller.
I didn't included benchmarks because this is just constant foldings
and is hard to mesure objectively.
For example, this enable rewriting things like:
if x == 20 {
return x + 30 + z
}
Into:
if x == 20 {
return 50 + z
}
It's not just fixing programer's code,
the ssa generator generate code like this sometimes.
Change-Id: I0861f342b27f7227b5f1c34d8267fa0057b1bbbc
GitHub-Last-Rev: 4c2f9b5216
GitHub-Pull-Request: golang/go#52669
Reviewed-on: https://go-review.googlesource.com/c/go/+/403735
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
This shows up in a few crypto functions, and other
assorted places.
Change-Id: I5a7f4c25ddd4a6499dc295ef693b9fe43d2448ab
Reviewed-on: https://go-review.googlesource.com/c/go/+/404057
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
In the load tests, we only want to test the assembly produced by
the load operations. If we use the global variable sink, it will produce
one load operation and one store operation(assign to sink).
For example:
func load_be64(b []byte) uint64 {
sink64 = binary.BigEndian.Uint64(b)
}
If we compile this function with GOAMD64=v3, it may produce MOVBEQload
and MOVQstore or MOVQload and MOVBEQstore, but we only want MOVBEQload.
Discovered when developing CL 395474.
Same for the store tests.
Change-Id: I65c3c742f1eff657c3a0d2dd103f51140ae8079e
Reviewed-on: https://go-review.googlesource.com/c/go/+/397875
Reviewed-by: Keith Randall <khr@golang.org>
Trust: Cherry Mui <cherryyz@google.com>
The SHRX/SHLX instruction can take any general register as the shift count operand, and can read source from memory. This CL introduces some operators to combine load and shift to one instruction.
For #47120
Change-Id: I13b48f53c7d30067a72eb2c8382242045dead36a
Reviewed-on: https://go-review.googlesource.com/c/go/+/385174
Reviewed-by: Keith Randall <khr@golang.org>
Trust: Cherry Mui <cherryyz@google.com>
LZCNT is similar to BSR, but BSR(x) is undefined when x == 0, so using
LZCNT can avoid a special case for zero input. Except that case,
LZCNTQ(x) == 63-BSRQ(x) and LZCNTL(x) == 31-BSRL(x).
And according to https://www.agner.org/optimize/instruction_tables.pdf,
LZCNT instructions are much faster than BSR on AMD CPU.
name old time/op new time/op delta
LeadingZeros-8 0.91ns ± 1% 0.80ns ± 7% -11.68% (p=0.000 n=9+9)
LeadingZeros8-8 0.98ns ±15% 0.91ns ± 1% -7.34% (p=0.000 n=9+9)
LeadingZeros16-8 0.94ns ± 3% 0.92ns ± 2% -2.36% (p=0.001 n=10+10)
LeadingZeros32-8 0.89ns ± 1% 0.78ns ± 2% -12.49% (p=0.000 n=10+10)
LeadingZeros64-8 0.92ns ± 1% 0.78ns ± 1% -14.48% (p=0.000 n=10+10)
Change-Id: I125147fe3d6994a4cfe558432780408e9a27557a
Reviewed-on: https://go-review.googlesource.com/c/go/+/396794
Reviewed-by: Keith Randall <khr@golang.org>
Trust: Emmanuel Odeke <emmanuel@orijtech.com>
Run-TryBot: Emmanuel Odeke <emmanuel@orijtech.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL add MOVBE support for 16-bit version, but MOVBEWload is
excluded because it does not satisfy zero extented.
For #51724
Change-Id: I3fadf20bcbb9b423f6355e6a1e340107e8e621ac
Reviewed-on: https://go-review.googlesource.com/c/go/+/396617
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Trust: Emmanuel Odeke <emmanuel@orijtech.com>
Add a new rewrite rule to merge ANDconst and UBFX into
UBFX.
Add test cases.
Change-Id: I24d6442d0c956d7ce092c3a3858d4a3a41771670
Reviewed-on: https://go-review.googlesource.com/c/go/+/377054
Trust: Fannie Zhang <Fannie.Zhang@arm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently we only include static entries in the hint for sizing
the map when allocating a map for a map literal. Change that to
include all entries.
This will be an overallocation if the dynamic entries in the map have
equal keys, but equal keys in map literals are rare, and at worst we
waste a bit of space.
Fixes#43020
Change-Id: I232f82f15316bdf4ea6d657d25a0b094b77884ce
Reviewed-on: https://go-review.googlesource.com/c/go/+/383634
Run-TryBot: Keith Randall <khr@golang.org>
Trust: Keith Randall <khr@golang.org>
Trust: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Updated multiple tests in test/codegen/arithmetic.go to verify
on ppc64/ppc64le as well
Change-Id: I79ca9f87017ea31147a4ba16f5d42ba0fcae64e1
Reviewed-on: https://go-review.googlesource.com/c/go/+/358546
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
In CL 354670, I copied some existing rules for convenience but forgot
to update the last rule which broke `GOAMD64=v3 ./make.bat`
Revive CL 354670
Change-Id: Ic1e2047c603f0122482a4b293ce1ef74d806c019
Reviewed-on: https://go-review.googlesource.com/c/go/+/356810
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Keith Randall <khr@golang.org>
Trust: Daniel Martí <mvdan@mvdan.cc>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Go Bot <gobot@golang.org>
It was noticed through some other investigation that BitLen32
was not generating the best code and found that it wasn't recognized
as an intrinsic. This corrects that and enables the test for PPC64.
Change-Id: Iab496a8830c8552f507b7292649b1b660f3848b5
Reviewed-on: https://go-review.googlesource.com/c/go/+/355872
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
The code generated when storing eight bytes loaded from memory in big
endian introduced two successive byte swaps that did not actually
modified the data.
The new rules match this specific pattern both for amd64 and for arm64,
eliminating the double swap.
Fixes#41684
Change-Id: Icb6dc20b68e4393cef4fe6a07b33aba0d18c3ff3
Reviewed-on: https://go-review.googlesource.com/c/go/+/320073
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Daniel Martí <mvdan@mvdan.cc>
Trust: Dmitri Shuralyov <dmitshur@golang.org>
This CL adds late expanded memequal(x, const, sz) inlining for 2, 4, 8
bytes size. This PoC is using the same method as CL 248404.
This optimization fires about 100 times in Go compiler (1675 occurrences
reduced to 1574, so -6%).
Also, added unit-tests to codegen/comparisions.go file.
Updates #37275
Change-Id: Ia52808d573cb706d1da8166c5746ede26f46c5da
Reviewed-on: https://go-review.googlesource.com/c/go/+/328291
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Trust: David Chase <drchase@google.com>
In case of amd64 the compiler issues checks if extensions are
available on a platform. With GOAMD64 microarchitecture levels
provided, some of the checks could be eliminated.
Change-Id: If15c178bcae273b2ce7d3673415cb8849292e087
Reviewed-on: https://go-review.googlesource.com/c/go/+/352010
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
This updates the codegen tests in noextend.go so they are not
dependent on the ABI.
Change-Id: I8433bea9dc78830c143290a7e0cf901b2397d38a
Reviewed-on: https://go-review.googlesource.com/c/go/+/353070
Trust: Lynn Boger <laboger@linux.vnet.ibm.com>
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Add rule to PPC64.rules to inline runtime.memmove in more cases, as is
done for other target architectures
Updated tests in codegen/copy.go to verify changes are done on
ppc64/ppc64le
Updates #41662
Change-Id: Id937ce21f9b4f4047b3e66dfa3c960128ee16a2a
Reviewed-on: https://go-review.googlesource.com/c/go/+/352054
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Trust: Lynn Boger <laboger@linux.vnet.ibm.com>
Instructions with immediates can be precomputed when operating on a
constant - do so for SLTI/SLTIU, SLLI/SRLI/SRAI, NEG/NEGW, ANDI, ORI
and ADDI. Additionally, optimise ANDI and ORI when the immediate is
all ones or all zeroes.
In particular, the RISCV64 logical left and right shift rules
(Lsh*x*/Rsh*Ux*) produce sequences that check if the shift amount
exceeds 64 and if so returns zero. When the shift amount is a
constant we can precompute and eliminate the filter entirely.
Likewise the arithmetic right shift rules produce sequences that
check if the shift amount exceeds 64 and if so, ensures that the
lower six bits of the shift are all ones. When the shift amount
is a constant we can precompute the shift value.
Arithmetic right shift sequences like:
117fc: 00100513 li a0,1
11800: 04053593 sltiu a1,a0,64
11804: fff58593 addi a1,a1,-1
11808: 0015e593 ori a1,a1,1
1180c: 40b45433 sra s0,s0,a1
Are now a single srai instruction:
117fc: 40145413 srai s0,s0,0x1
Likewise for logical left shift (and logical right shift):
1d560: 01100413 li s0,17
1d564: 04043413 sltiu s0,s0,64
1d568: 40800433 neg s0,s0
1d56c: 01131493 slli s1,t1,0x11
1d570: 0084f433 and s0,s1,s0
Which are now a single slli (or srli) instruction:
1d120: 01131413 slli s0,t1,0x11
This removes more than 30,000 instructions from the Go binary and
should improve performance in a variety of areas - of note
runtime.makemap_small drops from 48 to 36 instructions. Similar
gains exist in at least other parts of runtime and math/bits.
Change-Id: I33f6f3d1fd36d9ff1bda706997162bfe4bb859b6
Reviewed-on: https://go-review.googlesource.com/c/go/+/350689
Trust: Joel Sing <joel@sing.id.au>
Reviewed-by: Michael Munday <mike.munday@lowrisc.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
