mirror of
https://github.com/golang/go
synced 2024-11-25 23:28:00 -07:00
81c92352a7
491 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Lynn Boger
|
ebe49f98c8 |
cmd/compile: inline constant sized memclrNoHeapPointers calls on PPC64
Update the function isInlinableMemclr for ppc64 and ppc64le to enable inlining for the constant sized cases < 512. Larger cases can use dcbz which performs better but requires alignment checking so it is best to continue using memclrNoHeapPointers for those cases. Results on p10: MemclrKnownSize1 2.07ns ± 0% 0.57ns ± 0% -72.59% MemclrKnownSize2 2.56ns ± 5% 0.57ns ± 0% -77.82% MemclrKnownSize4 5.15ns ± 0% 0.57ns ± 0% -89.00% MemclrKnownSize8 2.23ns ± 0% 0.57ns ± 0% -74.57% MemclrKnownSize16 2.23ns ± 0% 0.50ns ± 0% -77.74% MemclrKnownSize32 2.28ns ± 0% 0.56ns ± 0% -75.28% MemclrKnownSize64 2.49ns ± 0% 0.72ns ± 0% -70.95% MemclrKnownSize112 2.97ns ± 2% 1.14ns ± 0% -61.72% MemclrKnownSize128 4.64ns ± 6% 2.45ns ± 1% -47.17% MemclrKnownSize192 5.45ns ± 5% 2.79ns ± 0% -48.87% MemclrKnownSize248 4.51ns ± 0% 2.83ns ± 0% -37.12% MemclrKnownSize256 6.34ns ± 1% 3.58ns ± 0% -43.53% MemclrKnownSize512 3.64ns ± 0% 3.64ns ± 0% -0.03% MemclrKnownSize1024 4.73ns ± 0% 4.73ns ± 0% +0.01% MemclrKnownSize4096 17.1ns ± 0% 17.1ns ± 0% +0.07% MemclrKnownSize512KiB 2.12µs ± 0% 2.12µs ± 0% ~ (all equal) Change-Id: If1abf5749f4802c64523a41fe0058bd144d0ea46 Reviewed-on: https://go-review.googlesource.com/c/go/+/464340 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> Reviewed-by: Jakub Ciolek <jakub@ciolek.dev> Reviewed-by: Archana Ravindar <aravind5@in.ibm.com> Reviewed-by: Ian Lance Taylor <iant@google.com> Reviewed-by: Carlos Eduardo Seo <carlos.seo@linaro.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Than McIntosh <thanm@google.com> |
||
Wayne Zuo
|
d7ac5d1480 |
cmd/compile: intrinsify math/bits/ReverseBytes{32|64} for 386
The BSWAPL instruction is supported in i486 and newer. https://github.com/golang/go/wiki/MinimumRequirements#386 says we support "All Pentium MMX or later". The Pentium is also referred to as i586, so that we are safe with these instructions. Change-Id: I6dea1f9d864a45bb07c8f8f35a81cfe16cca216c Reviewed-on: https://go-review.googlesource.com/c/go/+/465515 Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> Reviewed-by: Keith Randall <khr@golang.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com> Reviewed-by: Keith Randall <khr@google.com> |
||
Archana R
|
a432d89137 |
cmd/compile: add rules to emit SETBC/R instructions on power10
This CL adds rules that replaces instances of ISEL that produce a boolean result based on a condition register by SETBC/SETBCR operations. On Power10 these are convereted to SETBC/SETBCR instructions that use one register instead of 3 registers conventionally used by ISEL and hence reduces register pressure. On loops written specifically to exercise such instances of ISEL extensively, a performance improvement of 2.5% is seen on Power10. Also added verification tests to verify correct generation of SETBC/SETBCR instructions on Power10. Change-Id: Ib719897f09d893de40324440a43052dca026e8fa Reviewed-on: https://go-review.googlesource.com/c/go/+/449795 Reviewed-by: Michael Knyszek <mknyszek@google.com> Reviewed-by: Dmitri Shuralyov <dmitshur@google.com> Run-TryBot: Archana Ravindar <aravind5@in.ibm.com> Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gopher Robot <gobot@golang.org> |
||
|
Archana R
|
cd1fc87156 |
cmd/compile: intrinsify math/bits/ReverseBytes{16|32|64} for ppc64/power10
This change intrinsifies ReverseBytes{16|32|64} by generating the
corresponding new instructions in Power10: brh, brd and brw and
adds a verification test for the same.
On Power 9 and 8, the .go code performs optimally as it is.
Performance improvement seen on Power10:
ReverseBytes32 1.38ns ± 0% 1.18ns ± 0% -14.2
ReverseBytes64 1.52ns ± 0% 1.11ns ± 0% -26.87
ReverseBytes16 1.41ns ± 1% 1.18ns ± 0% -16.47
Change-Id: I88f127f3ab9ba24a772becc21ad90acfba324b37
Reviewed-on: https://go-review.googlesource.com/c/go/+/446675
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
|
||
Keith Randall
|
6224db9b4d |
cmd/compile: schedule values with no in-block uses later
When scheduling a block, deprioritize values whose results aren't used until subsequent blocks. For #58166, this has the effect of pushing the induction variable increment to the end of the block, past all the other uses of the pre-incremented value. Do this only with optimizations on. Debuggers have a preference for values in source code order, which this CL can degrade. Fixes #58166 Fixes #57976 Change-Id: I40d5885c661b142443c6d4702294c8abe8026c4f Reviewed-on: https://go-review.googlesource.com/c/go/+/463751 Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Keith Randall <khr@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> |
||
Jakub Ciolek
|
1fc585dc2f |
cmd/compile: inline known-size memclrNoHeapPointers calls
This patch rewrites known-size calls to memclrNoHeapPointers with an OpZero.
This significantly improves performance and lets some clears get DSE'd.
One of the cases where this applies is zeroing a known-size array, example:
var x [256]int8
...
for a := range x {
x[a] = 0
}
Other cases can be found in the runtime itself where memclrNoHeapPointers is sometimes directly invoked with a constant.
It seems that for some sized-clears on some architectures (AMD64, maybe others) the memcrlNoHeapPointers is more performant than OpZero.
See the issue #56997 for more details.
Benches ARM (M1 Pro):
name old time/op new time/op delta
MemclrKnownSize1-10 2.03ns ± 0% 0.31ns ± 0% -84.69% (p=0.000 n=18+19)
MemclrKnownSize2-10 1.97ns ± 0% 0.31ns ± 0% -84.19% (p=0.000 n=12+19)
MemclrKnownSize4-10 2.02ns ± 0% 0.31ns ± 0% -84.56% (p=0.000 n=12+20)
MemclrKnownSize8-10 2.02ns ± 0% 0.31ns ± 0% -84.59% (p=0.000 n=14+19)
MemclrKnownSize16-10 2.15ns ± 0% 0.31ns ± 0% -85.50% (p=0.000 n=18+19)
MemclrKnownSize32-10 2.48ns ± 0% 0.31ns ± 0% -87.48% (p=0.000 n=20+19)
MemclrKnownSize64-10 1.93ns ± 0% 0.62ns ± 0% -67.88% (p=0.000 n=20+19)
MemclrKnownSize112-10 2.48ns ± 0% 1.80ns ± 0% -27.74% (p=0.000 n=19+20)
MemclrKnownSize128-10 10.0ns ±112% 2.0ns ± 0% -79.76% (p=0.000 n=18+17)
MemclrKnownSize192-10 27.4ns ±103% 2.6ns ± 0% -90.38% (p=0.000 n=16+19)
MemclrKnownSize248-10 9.67ns ±43% 3.26ns ± 0% -66.29% (p=0.000 n=19+19)
MemclrKnownSize256-10 85.4ns ±148% 3.3ns ± 0% -96.18% (p=0.000 n=20+20)
MemclrKnownSize512-10 223ns ±54% 6ns ± 0% -97.42% (p=0.000 n=18+20)
MemclrKnownSize1024-10 216ns ±26% 11ns ± 0% -95.00% (p=0.000 n=18+15)
MemclrKnownSize4096-10 265ns ± 2% 88ns ± 0% -66.84% (p=0.000 n=19+17)
MemclrKnownSize512KiB-10 9.91µs ± 1% 10.23µs ± 2% +3.14% (p=0.000 n=19+19)
[Geo mean] 15.6ns 2.5ns -83.62%
name old speed new speed delta
MemclrKnownSize1-10 493MB/s ± 0% 3216MB/s ± 0% +553.04% (p=0.000 n=18+19)
MemclrKnownSize2-10 1.02GB/s ± 0% 6.43GB/s ± 0% +532.33% (p=0.000 n=16+19)
MemclrKnownSize4-10 1.99GB/s ± 0% 12.86GB/s ± 0% +547.67% (p=0.000 n=18+20)
MemclrKnownSize8-10 3.96GB/s ± 0% 25.72GB/s ± 0% +548.81% (p=0.000 n=19+19)
MemclrKnownSize16-10 7.46GB/s ± 0% 51.43GB/s ± 0% +589.72% (p=0.000 n=20+19)
MemclrKnownSize32-10 12.9GB/s ± 0% 102.9GB/s ± 0% +698.60% (p=0.000 n=20+18)
MemclrKnownSize64-10 33.1GB/s ± 0% 103.0GB/s ± 0% +211.34% (p=0.000 n=19+19)
MemclrKnownSize112-10 45.1GB/s ± 0% 62.4GB/s ± 0% +38.38% (p=0.000 n=19+20)
MemclrKnownSize128-10 13.3GB/s ±107% 63.5GB/s ± 0% +378.03% (p=0.000 n=19+18)
MemclrKnownSize192-10 6.97GB/s ±139% 72.72GB/s ± 0% +943.44% (p=0.000 n=19+19)
MemclrKnownSize248-10 25.9GB/s ±46% 76.1GB/s ± 0% +194.16% (p=0.000 n=20+17)
MemclrKnownSize256-10 8.64GB/s ±196% 78.51GB/s ± 0% +808.19% (p=0.000 n=20+20)
MemclrKnownSize512-10 2.33GB/s ±86% 89.13GB/s ± 0% +3719.50% (p=0.000 n=17+20)
MemclrKnownSize1024-10 4.85GB/s ±32% 94.93GB/s ± 0% +1856.74% (p=0.000 n=18+19)
MemclrKnownSize4096-10 15.4GB/s ± 2% 46.6GB/s ± 0% +201.55% (p=0.000 n=19+18)
MemclrKnownSize512KiB-10 52.9GB/s ± 1% 51.3GB/s ± 2% -3.04% (p=0.000 n=19+19)
[Geo mean] 7.54GB/s 42.86GB/s +468.76%
Intel Alder Lake 12600k:
name old time/op new time/op delta
MemclrKnownSize1-16 0.59ns ± 3% 0.38ns ± 6% -36.00% (p=0.000 n=19+18)
MemclrKnownSize2-16 0.57ns ± 1% 0.19ns ± 5% -66.27% (p=0.000 n=19+19)
MemclrKnownSize4-16 0.66ns ± 2% 0.36ns ±21% -45.12% (p=0.000 n=19+20)
MemclrKnownSize8-16 0.74ns ± 1% 0.30ns ±26% -59.81% (p=0.000 n=18+20)
MemclrKnownSize16-16 1.00ns ± 7% 0.21ns ± 8% -79.51% (p=0.000 n=20+19)
MemclrKnownSize32-16 0.95ns ± 1% 0.40ns ± 1% -57.61% (p=0.000 n=20+18)
MemclrKnownSize64-16 1.20ns ± 2% 0.41ns ± 0% -65.82% (p=0.000 n=20+18)
MemclrKnownSize112-16 1.27ns ± 2% 1.03ns ± 0% -19.35% (p=0.000 n=20+18)
MemclrKnownSize128-16 1.34ns ± 2% 1.03ns ± 0% -23.02% (p=0.000 n=20+20)
MemclrKnownSize192-16 1.92ns ± 2% 1.44ns ± 0% -24.89% (p=0.000 n=20+16)
MemclrKnownSize248-16 2.77ns ± 1% 3.29ns ± 0% +18.81% (p=0.000 n=20+16)
MemclrKnownSize256-16 1.92ns ± 1% 1.86ns ± 0% -3.49% (p=0.000 n=19+15)
MemclrKnownSize512-16 2.81ns ± 2% 3.49ns ± 0% +24.15% (p=0.000 n=20+17)
MemclrKnownSize1024-16 4.02ns ± 1% 6.78ns ± 0% +68.44% (p=0.000 n=20+18)
MemclrKnownSize4096-16 17.2ns ± 2% 14.4ns ± 0% -16.73% (p=0.000 n=20+17)
MemclrKnownSize512KiB-16 6.71µs ± 1% 6.52µs ± 0% -2.85% (p=0.000 n=20+18)
[Geo mean] 2.60ns 1.71ns -34.06%
name old speed new speed delta
MemclrKnownSize1-16 1.71GB/s ± 3% 2.67GB/s ± 6% +56.39% (p=0.000 n=19+18)
MemclrKnownSize2-16 3.52GB/s ± 2% 10.43GB/s ± 6% +196.04% (p=0.000 n=20+20)
MemclrKnownSize4-16 6.06GB/s ± 1% 10.83GB/s ±11% +78.63% (p=0.000 n=19+18)
MemclrKnownSize8-16 10.7GB/s ± 1% 27.0GB/s ±21% +151.49% (p=0.000 n=18+20)
MemclrKnownSize16-16 16.0GB/s ± 8% 78.1GB/s ± 7% +387.24% (p=0.000 n=20+19)
MemclrKnownSize32-16 33.6GB/s ± 1% 79.4GB/s ± 1% +135.89% (p=0.000 n=20+18)
MemclrKnownSize64-16 53.3GB/s ± 2% 155.9GB/s ± 0% +192.58% (p=0.000 n=20+18)
MemclrKnownSize112-16 88.0GB/s ± 2% 109.1GB/s ± 0% +23.97% (p=0.000 n=20+18)
MemclrKnownSize128-16 95.3GB/s ± 2% 123.8GB/s ± 0% +29.88% (p=0.000 n=20+20)
MemclrKnownSize192-16 100GB/s ± 2% 133GB/s ± 0% +33.12% (p=0.000 n=20+17)
MemclrKnownSize248-16 89.7GB/s ± 1% 75.5GB/s ± 0% -15.84% (p=0.000 n=20+19)
MemclrKnownSize256-16 133GB/s ± 1% 138GB/s ± 0% +3.61% (p=0.000 n=19+14)
MemclrKnownSize512-16 182GB/s ± 2% 147GB/s ± 0% -19.46% (p=0.000 n=20+17)
MemclrKnownSize1024-16 254GB/s ± 1% 151GB/s ± 0% -40.64% (p=0.000 n=20+18)
MemclrKnownSize4096-16 237GB/s ± 2% 285GB/s ± 0% +20.09% (p=0.000 n=20+17)
MemclrKnownSize512KiB-16 78.2GB/s ± 1% 80.4GB/s ± 0% +2.93% (p=0.000 n=20+18)
[Geo mean] 42.1GB/s 63.8GB/s +51.53%
compilecmp linux/amd64:
runtime
runtime.(*pallocData).allocAll 85 -> 45 (-47.06%)
runtime.(*pageAlloc).allocRange 942 -> 923 (-2.02%)
runtime.(*pageAlloc).free 798 -> 774 (-3.01%)
runtime.(*pageBits).clearAll 66 -> 20 (-69.70%)
runtime.startCheckmarks 255 -> 246 (-3.53%)
runtime.(*pallocData).freeAll 86 -> 46 (-46.51%)
runtime.(*pallocBits).freeAll 66 -> 20 (-69.70%)
runtime.(*consistentHeapStats).unsafeClear 66 -> 19 (-71.21%)
runtime.newproc1 965 -> 933 (-3.32%)
crypto/rc4
crypto/rc4.(*Cipher).Reset 78 -> 69 (-11.54%)
compress/bzip2
compress/bzip2.(*reader).readBlock 2973 -> 2941 (-1.08%)
image/jpeg
image/jpeg.(*decoder).processDHT 1179 -> 1166 (-1.10%)
index/suffixarray
index/suffixarray.bucketMax_8_32 394 -> 241 (-38.83%)
index/suffixarray.freq_8_32 317 -> 185 (-41.64%)
index/suffixarray.freq_8_64 317 -> 178 (-43.85%)
index/suffixarray.bucketMin_8_32 394 -> 243 (-38.32%)
index/suffixarray.bucketMin_8_64 398 -> 234 (-41.21%)
index/suffixarray.bucketMax_8_64 398 -> 234 (-41.21%)
compress/flate
compress/flate.(*huffmanBitWriter).generateCodegen 965 -> 838 (-13.16%)
compress/flate.(*compressor).reset 429 -> 409 (-4.66%)
cmd/vendor/golang.org/x/sys/unix
cmd/vendor/golang.org/x/sys/unix.(*FdSet).Zero 66 -> 60 (-9.09%)
cmd/vendor/golang.org/x/sys/unix.(*Ifreq).SetInet4Addr 211 -> 129 (-38.86%)
cmd/vendor/golang.org/x/sys/unix.(*Ifreq).SetUint32 98 -> 14 (-85.71%)
cmd/vendor/golang.org/x/sys/unix.(*Ifreq).clear 66 -> 11 (-83.33%)
cmd/vendor/golang.org/x/sys/unix.(*Ifreq).SetUint16 101 -> 15 (-85.15%)
cmd/vendor/golang.org/x/sys/unix.(*CPUSet).Zero 66 -> 60 (-9.09%)
internal/coverage/decodemeta
internal/coverage/decodemeta.(*CoverageMetaFileReader).rdUint64 325 -> 293 (-9.85%)
crypto/tls
crypto/tls.(*halfConn).setTrafficSecret 253 -> 247 (-2.37%)
crypto/tls.(*Conn).readRecordOrCCS 10315 -> 10283 (-0.31%)
crypto/tls.(*halfConn).changeCipherSpec 271 -> 261 (-3.69%)
crypto/tls.(*Conn).writeRecordLocked 1765 -> 1748 (-0.96%)
file before after Δ %
runtime.s 512467 512164 -303 -0.059%
crypto/rc4.s 955 946 -9 -0.942%
compress/bzip2.s 9586 9554 -32 -0.334%
image/jpeg.s 32122 32109 -13 -0.040%
index/suffixarray.s 38547 37644 -903 -2.343%
compress/flate.s 46668 46521 -147 -0.315%
cmd/vendor/golang.org/x/sys/unix.s 118620 118301 -319 -0.269%
internal/coverage/decodemeta.s 7224 7192 -32 -0.443%
crypto/tls.s 288762 288697 -65 -0.023%
cmd/compile/internal/ssa.s 3639799 3640727 +928 +0.025%
total 20790248 20789353 -895 -0.004%
src/runtime benchmarks (Linux Alder Lake 12600k):
name old time/op new time/op delta
MakeChan/Byte-16 26.2ns ± 2% 25.6ns ± 3% -2.05% (p=0.003 n=9+10)
MakeChan/Int-16 33.9ns ± 2% 33.3ns ± 4% -1.99% (p=0.015 n=10+10)
MakeChan/Ptr-16 54.2ns ± 2% 53.7ns ± 1% -0.90% (p=0.016 n=10+9)
MakeChan/Struct/0-16 23.8ns ± 3% 23.4ns ± 1% -1.72% (p=0.009 n=10+8)
MakeChan/Struct/32-16 55.9ns ± 2% 53.9ns ± 1% -3.48% (p=0.000 n=10+10)
MakeChan/Struct/40-16 63.5ns ± 1% 61.1ns ± 2% -3.79% (p=0.000 n=10+9)
ChanNonblocking-16 0.22ns ± 0% 0.22ns ± 0% +0.40% (p=0.011 n=9+8)
SelectUncontended-16 4.63ns ± 1% 4.62ns ± 0% -0.35% (p=0.001 n=10+8)
SelectSyncContended-16 1.58µs ± 2% 1.59µs ± 1% ~ (p=0.540 n=10+10)
SelectAsyncContended-16 290ns ± 0% 291ns ± 0% +0.14% (p=0.012 n=8+9)
SelectNonblock-16 0.95ns ± 1% 0.95ns ± 1% ~ (p=0.546 n=9+9)
ChanUncontended-16 239ns ± 3% 242ns ± 6% ~ (p=0.886 n=9+10)
ChanContended-16 17.7µs ± 1% 18.2µs ± 1% +2.87% (p=0.000 n=10+9)
ChanSync-16 109ns ± 2% 109ns ± 1% ~ (p=0.342 n=10+10)
ChanSyncWork-16 6.55µs ± 1% 6.53µs ± 1% ~ (p=0.101 n=10+10)
ChanProdCons0-16 502ns ± 1% 499ns ± 0% -0.55% (p=0.001 n=10+9)
ChanProdCons10-16 373ns ± 2% 377ns ± 1% ~ (p=0.095 n=10+9)
ChanProdCons100-16 224ns ± 2% 223ns ± 3% ~ (p=0.150 n=9+10)
ChanProdConsWork0-16 491ns ± 1% 484ns ± 0% -1.26% (p=0.000 n=10+9)
ChanProdConsWork10-16 451ns ± 2% 448ns ± 2% ~ (p=0.210 n=8+10)
ChanProdConsWork100-16 406ns ± 0% 407ns ± 1% ~ (p=0.138 n=8+8)
SelectProdCons-16 509ns ± 0% 509ns ± 0% ~ (p=0.917 n=9+9)
ReceiveDataFromClosedChan-16 12.1ns ± 0% 12.1ns ± 0% ~ (p=0.780 n=10+10)
ChanCreation-16 22.6ns ± 1% 22.4ns ± 0% -0.72% (p=0.001 n=10+8)
ChanSem-16 165ns ± 1% 166ns ± 1% +0.72% (p=0.002 n=10+10)
ChanPopular-16 500µs ± 2% 498µs ± 1% ~ (p=0.218 n=10+10)
ChanClosed-16 0.29ns ± 0% 0.29ns ± 0% +0.09% (p=0.019 n=9+8)
CallClosure-16 1.28ns ± 0% 1.27ns ± 0% -0.51% (p=0.000 n=9+9)
CallClosure1-16 1.50ns ± 0% 1.50ns ± 0% ~ (p=0.123 n=9+9)
CallClosure2-16 8.86ns ± 1% 8.86ns ± 3% ~ (p=0.590 n=9+10)
CallClosure3-16 8.75ns ± 2% 8.69ns ± 2% ~ (p=0.247 n=10+10)
CallClosure4-16 8.65ns ± 2% 8.56ns ± 2% ~ (p=0.105 n=10+10)
Complex128DivNormal-16 2.47ns ± 0% 2.47ns ± 0% ~ (p=0.790 n=10+9)
Complex128DivNisNaN-16 4.44ns ± 0% 4.43ns ± 0% ~ (p=0.564 n=10+10)
Complex128DivDisNaN-16 4.48ns ± 0% 4.48ns ± 0% ~ (p=0.101 n=10+10)
Complex128DivNisInf-16 2.58ns ± 0% 2.58ns ± 0% ~ (p=0.808 n=10+10)
Complex128DivDisInf-16 6.30ns ± 0% 6.31ns ± 0% ~ (p=0.305 n=10+10)
SetTypePtr-16 0.73ns ± 1% 0.73ns ± 3% ~ (p=0.644 n=10+10)
SetTypePtr8-16 4.12ns ± 0% 4.12ns ± 0% ~ (p=0.127 n=10+10)
SetTypePtr16-16 4.13ns ± 1% 4.12ns ± 0% ~ (p=0.109 n=10+10)
SetTypePtr32-16 4.12ns ± 0% 4.12ns ± 0% ~ (p=0.203 n=9+10)
SetTypePtr64-16 4.12ns ± 0% 4.12ns ± 0% ~ (p=0.696 n=10+10)
SetTypePtr126-16 6.91ns ± 0% 6.91ns ± 0% ~ (p=0.469 n=10+10)
SetTypePtr128-16 6.66ns ± 0% 6.67ns ± 0% ~ (p=0.246 n=9+10)
SetTypePtrSlice-16 54.1ns ± 1% 54.1ns ± 1% ~ (p=0.509 n=9+10)
SetTypeNode1-16 4.13ns ± 1% 4.12ns ± 0% ~ (p=0.342 n=10+10)
SetTypeNode1Slice-16 10.1ns ± 1% 10.0ns ± 1% -1.18% (p=0.000 n=10+10)
SetTypeNode8-16 4.12ns ± 0% 4.12ns ± 0% ~ (p=0.137 n=8+8)
SetTypeNode8Slice-16 22.6ns ± 0% 22.6ns ± 0% ~ (p=0.423 n=10+10)
SetTypeNode64-16 6.90ns ± 0% 6.91ns ± 0% ~ (p=0.275 n=10+10)
SetTypeNode64Slice-16 173ns ± 0% 173ns ± 0% ~ (p=0.610 n=9+10)
SetTypeNode64Dead-16 5.53ns ± 0% 5.52ns ± 0% ~ (p=0.123 n=10+6)
SetTypeNode64DeadSlice-16 150ns ± 0% 150ns ± 0% ~ (p=0.398 n=10+10)
SetTypeNode124-16 6.90ns ± 0% 6.90ns ± 0% ~ (p=0.779 n=10+10)
SetTypeNode124Slice-16 222ns ± 5% 217ns ± 0% ~ (p=0.302 n=10+10)
SetTypeNode126-16 6.66ns ± 0% 6.66ns ± 0% ~ (p=0.324 n=10+9)
SetTypeNode126Slice-16 218ns ± 0% 218ns ± 0% ~ (p=0.119 n=9+10)
SetTypeNode128-16 9.76ns ± 0% 9.73ns ± 0% -0.31% (p=0.003 n=9+10)
SetTypeNode128Slice-16 279ns ± 0% 278ns ± 0% ~ (p=0.112 n=10+9)
SetTypeNode130-16 9.77ns ± 0% 9.73ns ± 0% -0.33% (p=0.002 n=10+10)
SetTypeNode130Slice-16 284ns ± 0% 284ns ± 0% ~ (p=0.668 n=10+10)
SetTypeNode1024-16 51.2ns ± 0% 51.6ns ± 1% ~ (p=0.080 n=9+9)
SetTypeNode1024Slice-16 1.83µs ± 0% 1.82µs ± 0% ~ (p=0.115 n=10+10)
Allocation-16 4.64µs ± 1% 4.37µs ± 1% -5.69% (p=0.000 n=9+9)
ReadMemStats-16 5.62µs ± 2% 5.55µs ± 5% -1.36% (p=0.050 n=10+10)
WriteBarrier-16 4.95ns ± 3% 4.99ns ± 3% ~ (p=0.255 n=10+10)
BulkWriteBarrier-16 1.69ns ± 2% 1.63ns ± 4% -3.77% (p=0.001 n=10+10)
ScanStackNoLocals-16 12.8ms ± 2% 12.9ms ± 1% +0.72% (p=0.019 n=10+10)
MSpanCountAlloc/bits=64-16 1.65ns ± 0% 1.65ns ± 0% ~ (p=0.124 n=10+10)
MSpanCountAlloc/bits=128-16 2.08ns ± 1% 2.06ns ± 1% -0.87% (p=0.000 n=10+10)
MSpanCountAlloc/bits=256-16 2.71ns ± 1% 2.69ns ± 1% -0.74% (p=0.001 n=10+9)
MSpanCountAlloc/bits=512-16 4.15ns ± 0% 4.23ns ± 2% +2.15% (p=0.000 n=10+10)
MSpanCountAlloc/bits=1024-16 7.89ns ± 1% 7.89ns ± 1% ~ (p=0.867 n=10+10)
Hash5-16 1.93ns ± 1% 2.01ns ± 0% +3.99% (p=0.000 n=10+8)
Hash16-16 2.04ns ± 1% 2.21ns ± 1% +8.61% (p=0.000 n=10+10)
Hash64-16 2.67ns ± 0% 2.67ns ± 0% ~ (p=0.154 n=9+9)
Hash1024-16 16.4ns ± 0% 16.4ns ± 0% +0.17% (p=0.020 n=9+10)
Hash65536-16 886ns ± 0% 885ns ± 0% ~ (p=0.725 n=10+10)
AlignedLoad-16 0.96ns ± 2% 0.95ns ± 3% ~ (p=0.123 n=10+10)
UnalignedLoad-16 0.95ns ± 2% 1.01ns ± 2% +6.31% (p=0.000 n=10+10)
EqEfaceConcrete-16 0.31ns ± 3% 0.33ns ± 5% +8.10% (p=0.000 n=10+10)
EqIfaceConcrete-16 0.31ns ±13% 0.28ns ± 2% -9.23% (p=0.001 n=10+10)
NeEfaceConcrete-16 0.29ns ± 1% 0.31ns ± 7% +5.59% (p=0.010 n=8+8)
NeIfaceConcrete-16 0.28ns ± 2% 0.29ns ± 1% +4.49% (p=0.000 n=9+8)
ConvT2EByteSized/bool-16 0.53ns ± 1% 0.52ns ± 1% -2.18% (p=0.000 n=10+10)
ConvT2EByteSized/uint8-16 0.53ns ± 1% 0.53ns ± 0% +1.22% (p=0.000 n=10+10)
ConvT2ESmall-16 1.13ns ± 0% 1.13ns ± 0% ~ (p=0.774 n=9+9)
ConvT2EUintptr-16 1.03ns ± 0% 1.04ns ± 0% +0.50% (p=0.000 n=10+8)
ConvT2ELarge-16 14.4ns ± 2% 14.4ns ± 1% ~ (p=0.726 n=10+10)
ConvT2ISmall-16 1.13ns ± 0% 1.13ns ± 0% ~ (p=0.693 n=9+10)
ConvT2IUintptr-16 1.03ns ± 0% 1.03ns ± 0% +0.44% (p=0.000 n=10+10)
ConvT2ILarge-16 14.2ns ± 1% 14.4ns ± 1% +0.85% (p=0.007 n=9+10)
ConvI2E-16 0.54ns ± 1% 0.54ns ± 0% -0.39% (p=0.037 n=10+8)
ConvI2I-16 2.68ns ± 0% 2.70ns ± 1% +0.73% (p=0.000 n=9+10)
AssertE2T-16 0.28ns ± 1% 0.39ns ± 5% +37.38% (p=0.000 n=10+10)
AssertE2TLarge-16 0.42ns ± 2% 0.48ns ± 1% +14.92% (p=0.000 n=9+10)
AssertE2I-16 2.67ns ± 0% 2.67ns ± 0% ~ (p=0.352 n=9+9)
AssertI2T-16 0.37ns ± 3% 0.34ns ± 1% -6.16% (p=0.000 n=10+10)
AssertI2I-16 2.67ns ± 0% 2.67ns ± 0% ~ (p=0.286 n=10+10)
AssertI2E-16 0.54ns ± 1% 0.54ns ± 0% -0.94% (p=0.000 n=10+10)
AssertE2E-16 0.41ns ± 0% 0.41ns ± 1% ~ (p=0.880 n=9+9)
AssertE2T2-16 0.41ns ± 1% 0.41ns ± 1% ~ (p=0.725 n=10+10)
AssertE2T2Blank-16 0.24ns ± 5% 0.21ns ± 1% -14.79% (p=0.000 n=10+9)
AssertI2E2-16 0.69ns ± 0% 0.69ns ± 1% ~ (p=0.541 n=10+10)
AssertI2E2Blank-16 0.26ns ± 9% 0.21ns ± 1% -18.86% (p=0.000 n=10+9)
AssertE2E2-16 0.53ns ± 1% 0.53ns ± 1% +0.72% (p=0.004 n=10+10)
AssertE2E2Blank-16 0.23ns ± 4% 0.21ns ± 1% -8.42% (p=0.000 n=10+10)
ConvT2Ezero/zero/16-16 1.13ns ± 0% 1.14ns ± 1% ~ (p=0.583 n=9+10)
ConvT2Ezero/zero/32-16 1.13ns ± 0% 1.13ns ± 0% ~ (p=0.417 n=10+10)
ConvT2Ezero/zero/64-16 1.03ns ± 1% 1.03ns ± 0% ~ (p=0.051 n=10+10)
ConvT2Ezero/zero/str-16 1.03ns ± 0% 1.03ns ± 0% ~ (p=0.132 n=10+10)
ConvT2Ezero/zero/slice-16 1.14ns ± 0% 1.15ns ± 0% +0.49% (p=0.001 n=10+10)
ConvT2Ezero/zero/big-16 123ns ± 1% 123ns ± 1% ~ (p=0.171 n=10+10)
ConvT2Ezero/nonzero/str-16 19.4ns ± 1% 19.3ns ± 3% ~ (p=0.548 n=9+10)
ConvT2Ezero/nonzero/slice-16 22.2ns ± 2% 22.0ns ± 2% ~ (p=0.109 n=10+10)
ConvT2Ezero/nonzero/big-16 123ns ± 1% 123ns ± 1% ~ (p=0.446 n=10+8)
ConvT2Ezero/smallint/16-16 1.13ns ± 0% 1.14ns ± 1% ~ (p=0.362 n=10+10)
ConvT2Ezero/smallint/32-16 1.13ns ± 0% 1.13ns ± 0% ~ (p=0.907 n=10+9)
ConvT2Ezero/smallint/64-16 1.04ns ± 0% 1.03ns ± 0% -0.38% (p=0.002 n=10+10)
ConvT2Ezero/largeint/16-16 6.65ns ± 1% 6.65ns ± 2% ~ (p=0.618 n=10+9)
ConvT2Ezero/largeint/32-16 6.75ns ± 3% 6.63ns ± 2% -1.77% (p=0.015 n=10+10)
ConvT2Ezero/largeint/64-16 9.19ns ± 1% 9.26ns ± 2% ~ (p=0.123 n=10+10)
Malloc8-16 8.66ns ± 1% 8.89ns ± 2% +2.74% (p=0.000 n=10+10)
Malloc16-16 13.7ns ± 1% 13.8ns ± 1% +0.71% (p=0.022 n=10+8)
MallocTypeInfo8-16 11.7ns ± 3% 11.6ns ± 2% ~ (p=0.469 n=10+10)
MallocTypeInfo16-16 18.3ns ± 1% 18.2ns ± 2% ~ (p=0.251 n=9+10)
MallocLargeStruct-16 195ns ± 1% 198ns ± 1% +1.65% (p=0.000 n=9+10)
GoroutineSelect-16 1.10ms ± 1% 1.12ms ± 1% +1.36% (p=0.000 n=10+8)
GoroutineBlocking-16 986µs ± 1% 998µs ± 1% +1.23% (p=0.002 n=10+10)
GoroutineForRange-16 985µs ± 1% 1001µs ± 1% +1.68% (p=0.000 n=10+10)
GoroutineIdle-16 679µs ± 1% 691µs ± 0% +1.74% (p=0.000 n=10+9)
HashStringSpeed-16 5.33ns ± 5% 5.19ns ± 4% ~ (p=0.113 n=9+9)
HashBytesSpeed-16 8.20ns ± 3% 8.24ns ± 1% ~ (p=0.497 n=10+9)
HashInt32Speed-16 4.01ns ± 2% 3.90ns ± 4% -2.63% (p=0.011 n=9+10)
HashInt64Speed-16 3.94ns ± 4% 3.79ns ± 1% -3.74% (p=0.003 n=10+9)
HashStringArraySpeed-16 12.5ns ± 4% 12.3ns ± 1% ~ (p=0.055 n=10+10)
MegMap-16 3.72ns ± 1% 3.73ns ± 1% ~ (p=0.484 n=9+10)
MegOneMap-16 2.28ns ± 1% 2.27ns ± 1% ~ (p=0.287 n=10+10)
MegEqMap-16 22.0µs ± 3% 22.3µs ± 2% +1.48% (p=0.028 n=10+9)
MegEmptyMap-16 0.93ns ± 1% 0.92ns ± 1% -0.52% (p=0.030 n=10+10)
SmallStrMap-16 3.77ns ± 0% 3.77ns ± 0% ~ (p=0.324 n=10+10)
MapStringKeysEight_16-16 3.91ns ± 0% 3.91ns ± 0% ~ (p=0.088 n=9+9)
MapStringKeysEight_32-16 3.58ns ± 1% 3.50ns ± 0% -2.11% (p=0.000 n=10+10)
MapStringKeysEight_64-16 3.58ns ± 1% 3.50ns ± 0% -2.23% (p=0.000 n=10+10)
MapStringKeysEight_1M-16 3.57ns ± 1% 3.50ns ± 0% -1.92% (p=0.000 n=10+10)
IntMap-16 2.89ns ± 1% 2.89ns ± 0% ~ (p=0.381 n=10+10)
MapFirst/1-16 1.60ns ± 1% 1.59ns ± 2% -0.49% (p=0.020 n=10+9)
MapFirst/2-16 1.61ns ± 0% 1.59ns ± 1% -1.17% (p=0.001 n=10+10)
MapFirst/3-16 1.61ns ± 1% 1.59ns ± 1% -1.45% (p=0.000 n=10+10)
MapFirst/4-16 1.60ns ± 1% 1.59ns ± 1% -1.16% (p=0.000 n=10+10)
MapFirst/5-16 1.60ns ± 1% 1.58ns ± 1% -0.98% (p=0.000 n=10+10)
MapFirst/6-16 1.60ns ± 1% 1.59ns ± 1% -0.87% (p=0.001 n=10+10)
MapFirst/7-16 1.60ns ± 1% 1.59ns ± 1% -0.79% (p=0.002 n=10+10)
MapFirst/8-16 1.60ns ± 1% 1.59ns ± 1% -0.67% (p=0.017 n=9+10)
MapFirst/9-16 2.83ns ± 0% 2.83ns ± 0% ~ (p=0.492 n=10+10)
MapFirst/10-16 2.83ns ± 0% 2.84ns ± 0% +0.24% (p=0.017 n=10+10)
MapFirst/11-16 2.83ns ± 0% 2.83ns ± 0% ~ (p=0.445 n=10+10)
MapFirst/12-16 2.83ns ± 0% 2.83ns ± 0% ~ (p=0.564 n=10+10)
MapFirst/13-16 2.83ns ± 0% 2.84ns ± 0% ~ (p=0.175 n=9+10)
MapFirst/14-16 2.83ns ± 0% 2.83ns ± 0% ~ (p=0.322 n=10+9)
MapFirst/15-16 2.83ns ± 0% 2.84ns ± 1% ~ (p=0.209 n=10+10)
MapFirst/16-16 2.83ns ± 1% 2.84ns ± 0% ~ (p=0.238 n=10+10)
MapMid/1-16 1.64ns ± 0% 1.64ns ± 0% ~ (p=0.453 n=10+9)
MapMid/2-16 1.86ns ± 1% 1.86ns ± 0% ~ (p=0.764 n=10+9)
MapMid/3-16 1.86ns ± 0% 1.86ns ± 1% ~ (p=1.000 n=10+10)
MapMid/4-16 2.06ns ± 0% 2.06ns ± 0% -0.27% (p=0.014 n=10+9)
MapMid/5-16 2.06ns ± 0% 2.06ns ± 0% ~ (p=0.075 n=9+10)
MapMid/6-16 2.27ns ± 0% 2.27ns ± 1% ~ (p=0.898 n=10+10)
MapMid/7-16 2.27ns ± 1% 2.26ns ± 0% -0.23% (p=0.049 n=10+10)
MapMid/8-16 2.47ns ± 0% 2.47ns ± 1% ~ (p=0.840 n=10+10)
MapMid/9-16 4.21ns ± 7% 4.13ns ±19% ~ (p=0.315 n=10+10)
MapMid/10-16 4.17ns ± 7% 4.31ns ± 5% +3.37% (p=0.021 n=10+9)
MapMid/11-16 4.18ns ± 7% 4.32ns ± 6% +3.50% (p=0.015 n=10+10)
MapMid/12-16 4.34ns ± 7% 4.30ns ± 5% ~ (p=0.858 n=9+10)
MapMid/13-16 4.25ns ± 6% 4.28ns ± 6% ~ (p=0.489 n=9+9)
MapMid/14-16 3.75ns ±23% 3.90ns ±16% ~ (p=0.353 n=10+10)
MapMid/15-16 3.87ns ±25% 3.95ns ±26% ~ (p=0.315 n=10+10)
MapMid/16-16 4.06ns ±19% 3.94ns ±16% ~ (p=0.796 n=10+10)
MapLast/1-16 1.65ns ± 0% 1.65ns ± 0% ~ (p=0.607 n=10+10)
MapLast/2-16 1.86ns ± 0% 1.86ns ± 0% +0.26% (p=0.029 n=10+10)
MapLast/3-16 2.06ns ± 1% 2.06ns ± 0% ~ (p=0.689 n=8+9)
MapLast/4-16 2.27ns ± 1% 2.26ns ± 0% ~ (p=0.148 n=10+9)
MapLast/5-16 2.47ns ± 0% 2.47ns ± 0% ~ (p=0.385 n=9+10)
MapLast/6-16 2.67ns ± 0% 2.68ns ± 0% ~ (p=0.202 n=9+10)
MapLast/7-16 2.88ns ± 0% 2.88ns ± 0% ~ (p=0.751 n=10+10)
MapLast/8-16 3.08ns ± 0% 3.08ns ± 0% ~ (p=0.826 n=10+9)
MapLast/9-16 4.31ns ± 6% 4.54ns ± 5% ~ (p=0.070 n=9+8)
MapLast/10-16 4.25ns ± 5% 4.42ns ± 6% ~ (p=0.321 n=9+8)
MapLast/11-16 4.59ns ±16% 5.42ns ±44% +17.99% (p=0.019 n=10+10)
MapLast/12-16 5.04ns ±19% 6.11ns ±28% +21.35% (p=0.005 n=9+10)
MapLast/13-16 6.00ns ±35% 5.76ns ± 3% ~ (p=0.173 n=10+8)
MapLast/14-16 4.27ns ± 5% 4.53ns ± 6% +6.14% (p=0.007 n=10+10)
MapLast/15-16 4.41ns ± 1% 4.44ns ± 7% ~ (p=0.515 n=8+10)
MapLast/16-16 4.18ns ± 6% 4.99ns ±18% +19.48% (p=0.000 n=10+10)
MapCycle-16 7.48ns ± 2% 7.46ns ± 1% ~ (p=0.699 n=10+10)
RepeatedLookupStrMapKey32-16 6.98ns ± 3% 6.73ns ± 2% -3.63% (p=0.000 n=10+10)
RepeatedLookupStrMapKey1M-16 14.7µs ± 5% 14.7µs ± 4% ~ (p=0.604 n=9+10)
MakeMap/[Byte]Byte-16 58.5ns ± 1% 58.5ns ± 1% ~ (p=0.780 n=10+9)
MakeMap/[Int]Int-16 113ns ± 0% 113ns ± 1% ~ (p=0.100 n=8+10)
NewEmptyMap-16 2.47ns ± 0% 2.47ns ± 0% ~ (p=0.638 n=10+10)
NewSmallMap-16 11.5ns ± 1% 11.6ns ± 0% +1.18% (p=0.000 n=10+10)
MapIter-16 42.2ns ± 0% 42.8ns ± 1% +1.50% (p=0.000 n=10+10)
MapIterEmpty-16 1.85ns ± 0% 1.85ns ± 0% ~ (p=0.651 n=10+10)
SameLengthMap-16 1.85ns ± 1% 1.85ns ± 0% ~ (p=0.247 n=10+10)
BigKeyMap-16 7.18ns ± 1% 7.42ns ± 4% +3.33% (p=0.004 n=10+10)
BigValMap-16 7.03ns ± 2% 7.19ns ± 1% +2.33% (p=0.000 n=10+9)
SmallKeyMap-16 5.32ns ± 1% 5.24ns ± 1% -1.41% (p=0.000 n=10+10)
MapPopulate/1-16 6.30ns ± 0% 6.41ns ± 1% +1.81% (p=0.000 n=8+10)
MapPopulate/10-16 239ns ± 2% 234ns ± 2% -2.05% (p=0.001 n=9+10)
MapPopulate/100-16 4.19µs ± 2% 4.22µs ± 2% ~ (p=0.171 n=10+10)
MapPopulate/1000-16 52.3µs ± 1% 52.5µs ± 1% ~ (p=0.133 n=9+10)
MapPopulate/10000-16 459µs ± 1% 466µs ± 2% +1.45% (p=0.005 n=10+10)
MapPopulate/100000-16 4.22ms ± 2% 4.25ms ± 2% ~ (p=0.393 n=10+10)
ComplexAlgMap-16 12.5ns ± 1% 12.4ns ± 1% -0.95% (p=0.022 n=10+10)
GoMapClear/Reflexive/1-16 9.61ns ± 1% 9.58ns ± 0% -0.27% (p=0.027 n=10+10)
GoMapClear/Reflexive/10-16 10.0ns ± 1% 10.0ns ± 1% ~ (p=0.648 n=9+9)
GoMapClear/Reflexive/100-16 31.4ns ± 0% 31.4ns ± 1% ~ (p=0.305 n=9+10)
GoMapClear/Reflexive/1000-16 147ns ± 0% 149ns ± 2% +1.21% (p=0.000 n=10+10)
GoMapClear/Reflexive/10000-16 3.99µs ± 0% 4.00µs ± 0% +0.21% (p=0.018 n=9+10)
GoMapClear/NonReflexive/1-16 41.4ns ± 2% 41.7ns ± 1% +0.55% (p=0.043 n=9+10)
GoMapClear/NonReflexive/10-16 50.3ns ± 1% 50.9ns ± 1% +1.16% (p=0.000 n=10+10)
GoMapClear/NonReflexive/100-16 125ns ± 0% 126ns ± 0% +0.96% (p=0.000 n=8+10)
GoMapClear/NonReflexive/1000-16 1.08µs ± 0% 1.08µs ± 1% ~ (p=0.097 n=10+10)
GoMapClear/NonReflexive/10000-16 8.18µs ± 2% 8.10µs ± 0% -0.91% (p=0.019 n=10+8)
MapStringConversion/32/simple-16 4.66ns ± 1% 4.69ns ± 3% ~ (p=0.905 n=9+10)
MapStringConversion/32/struct-16 4.65ns ± 3% 4.94ns ± 2% +6.23% (p=0.000 n=10+10)
MapStringConversion/32/array-16 4.69ns ± 3% 4.72ns ± 3% ~ (p=0.631 n=10+10)
MapStringConversion/64/simple-16 4.14ns ± 0% 4.14ns ± 1% ~ (p=0.342 n=10+10)
MapStringConversion/64/struct-16 4.13ns ± 0% 4.13ns ± 0% ~ (p=0.809 n=10+10)
MapStringConversion/64/array-16 4.13ns ± 1% 4.13ns ± 1% ~ (p=0.752 n=10+10)
MapInterfaceString-16 7.90ns ±23% 8.51ns ±33% ~ (p=0.604 n=9+10)
MapInterfacePtr-16 7.68ns ±29% 7.10ns ±36% ~ (p=0.353 n=10+10)
NewEmptyMapHintLessThan8-16 3.70ns ± 0% 3.70ns ± 0% ~ (p=0.209 n=10+10)
NewEmptyMapHintGreaterThan8-16 270ns ± 1% 272ns ± 1% +0.71% (p=0.005 n=10+9)
MapPop100-16 6.45µs ± 0% 6.50µs ± 1% +0.77% (p=0.000 n=10+10)
MapPop1000-16 114µs ± 1% 114µs ± 1% ~ (p=0.190 n=10+10)
MapPop10000-16 2.28ms ± 2% 2.28ms ± 2% ~ (p=0.912 n=10+10)
MapAssign/Int32/256-16 4.75ns ± 2% 4.82ns ± 4% ~ (p=0.101 n=10+10)
MapAssign/Int32/65536-16 16.4ns ± 1% 16.7ns ± 0% +1.44% (p=0.000 n=10+9)
MapAssign/Int64/256-16 4.79ns ± 5% 4.79ns ± 1% ~ (p=0.616 n=10+8)
MapAssign/Int64/65536-16 17.1ns ± 1% 16.8ns ± 0% -1.28% (p=0.000 n=10+8)
MapAssign/Str/256-16 6.07ns ± 6% 6.24ns ± 2% +2.84% (p=0.035 n=10+9)
MapAssign/Str/65536-16 21.4ns ± 0% 21.4ns ± 3% ~ (p=0.300 n=7+10)
MapOperatorAssign/Int32/256-16 4.82ns ± 3% 4.81ns ± 3% ~ (p=0.684 n=10+10)
MapOperatorAssign/Int32/65536-16 16.8ns ± 1% 16.5ns ± 1% -1.68% (p=0.000 n=9+10)
MapOperatorAssign/Int64/256-16 4.74ns ± 1% 4.77ns ± 3% ~ (p=0.563 n=10+9)
MapOperatorAssign/Int64/65536-16 16.9ns ± 1% 17.2ns ± 1% +1.88% (p=0.000 n=10+10)
MapOperatorAssign/Str/256-16 1.09µs ± 1% 1.10µs ± 2% ~ (p=0.210 n=10+10)
MapOperatorAssign/Str/65536-16 184ns ± 9% 184ns ± 8% ~ (p=0.922 n=10+9)
MapAppendAssign/Int32/256-16 13.8ns ±10% 14.4ns ±11% ~ (p=0.190 n=10+10)
MapAppendAssign/Int32/65536-16 28.9ns ± 5% 30.7ns ± 6% +6.13% (p=0.001 n=9+10)
MapAppendAssign/Int64/256-16 14.5ns ±12% 13.8ns ± 8% -5.02% (p=0.037 n=10+10)
MapAppendAssign/Int64/65536-16 30.9ns ± 1% 30.4ns ± 2% -1.56% (p=0.001 n=10+10)
MapAppendAssign/Str/256-16 30.2ns ± 6% 30.0ns ±10% ~ (p=0.645 n=10+10)
MapAppendAssign/Str/65536-16 44.5ns ± 4% 46.8ns ± 3% +5.17% (p=0.001 n=8+9)
MapDelete/Int32/100-16 18.7ns ± 0% 18.7ns ± 0% -0.27% (p=0.017 n=10+10)
MapDelete/Int32/1000-16 17.6ns ± 1% 17.5ns ± 1% -0.85% (p=0.000 n=9+10)
MapDelete/Int32/10000-16 18.7ns ± 0% 18.3ns ± 1% -1.92% (p=0.000 n=10+10)
MapDelete/Int64/100-16 19.1ns ± 0% 19.2ns ± 0% +0.68% (p=0.000 n=10+9)
MapDelete/Int64/1000-16 17.7ns ± 2% 18.3ns ± 1% +3.00% (p=0.000 n=10+10)
MapDelete/Int64/10000-16 18.8ns ± 1% 19.2ns ± 0% +2.01% (p=0.000 n=10+9)
MapDelete/Str/100-16 26.5ns ± 0% 26.4ns ± 1% -0.73% (p=0.000 n=10+10)
MapDelete/Str/1000-16 23.5ns ± 2% 23.4ns ± 1% ~ (p=0.425 n=10+10)
MapDelete/Str/10000-16 25.1ns ± 0% 25.1ns ± 1% +0.28% (p=0.037 n=10+10)
MapDelete/Pointer/100-16 20.6ns ± 1% 20.6ns ± 0% ~ (p=0.117 n=10+10)
MapDelete/Pointer/1000-16 19.2ns ± 1% 19.4ns ± 1% +0.97% (p=0.004 n=10+10)
MapDelete/Pointer/10000-16 20.0ns ± 0% 20.1ns ± 1% +0.52% (p=0.022 n=10+10)
Memmove/0-16 0.21ns ± 2% 0.21ns ± 1% ~ (p=0.671 n=10+10)
Memmove/1-16 0.93ns ± 0% 0.93ns ± 0% +0.21% (p=0.034 n=10+10)
Memmove/2-16 0.93ns ± 0% 0.93ns ± 0% ~ (p=0.101 n=10+10)
Memmove/3-16 0.93ns ± 1% 0.93ns ± 1% +0.49% (p=0.004 n=10+10)
Memmove/4-16 1.03ns ± 0% 1.03ns ± 0% ~ (p=0.260 n=10+10)
Memmove/5-16 1.13ns ± 0% 1.13ns ± 0% +0.20% (p=0.034 n=10+10)
Memmove/6-16 1.13ns ± 0% 1.13ns ± 1% ~ (p=0.126 n=10+10)
Memmove/7-16 1.13ns ± 0% 1.13ns ± 1% +0.22% (p=0.028 n=10+10)
Memmove/8-16 1.13ns ± 0% 1.13ns ± 0% ~ (p=0.545 n=9+10)
Memmove/9-16 1.25ns ± 0% 1.35ns ± 0% +7.98% (p=0.000 n=10+10)
Memmove/10-16 1.25ns ± 0% 1.35ns ± 0% +7.96% (p=0.000 n=9+9)
Memmove/11-16 1.25ns ± 0% 1.35ns ± 0% +8.53% (p=0.000 n=10+9)
Memmove/12-16 1.25ns ± 0% 1.35ns ± 1% +8.24% (p=0.000 n=10+10)
Memmove/13-16 1.25ns ± 0% 1.34ns ± 0% +7.75% (p=0.000 n=10+10)
Memmove/14-16 1.25ns ± 0% 1.35ns ± 1% +8.28% (p=0.000 n=10+9)
Memmove/15-16 1.25ns ± 0% 1.35ns ± 0% +8.07% (p=0.000 n=10+9)
Memmove/16-16 1.25ns ± 0% 1.35ns ± 1% +8.35% (p=0.000 n=9+10)
Memmove/32-16 1.34ns ± 0% 1.36ns ± 1% +1.22% (p=0.000 n=10+10)
Memmove/64-16 1.45ns ± 0% 1.64ns ± 0% +13.07% (p=0.000 n=10+9)
Memmove/128-16 1.86ns ± 0% 2.02ns ± 0% +8.64% (p=0.000 n=10+10)
Memmove/256-16 2.47ns ± 0% 2.49ns ± 1% +1.14% (p=0.000 n=10+10)
Memmove/512-16 3.96ns ± 1% 3.96ns ± 0% ~ (p=0.182 n=10+10)
Memmove/1024-16 5.90ns ± 1% 5.87ns ± 1% ~ (p=0.258 n=9+9)
Memmove/2048-16 9.62ns ± 1% 9.62ns ± 2% ~ (p=0.963 n=8+9)
Memmove/4096-16 16.4ns ± 0% 17.1ns ± 4% +4.19% (p=0.003 n=8+9)
MemmoveOverlap/32-16 1.62ns ± 1% 1.68ns ± 1% +3.53% (p=0.000 n=10+10)
MemmoveOverlap/64-16 1.64ns ± 0% 1.65ns ± 0% +0.29% (p=0.002 n=9+9)
MemmoveOverlap/128-16 2.06ns ± 0% 2.06ns ± 0% ~ (p=0.070 n=10+10)
MemmoveOverlap/256-16 2.67ns ± 0% 2.67ns ± 0% +0.26% (p=0.012 n=10+10)
MemmoveOverlap/512-16 6.20ns ±18% 5.74ns ± 0% ~ (p=0.645 n=10+8)
MemmoveOverlap/1024-16 7.28ns ± 0% 7.30ns ± 0% +0.28% (p=0.006 n=8+10)
MemmoveOverlap/2048-16 11.9ns ± 0% 12.0ns ± 1% +0.37% (p=0.014 n=9+9)
MemmoveOverlap/4096-16 23.3ns ± 1% 23.1ns ± 1% -0.84% (p=0.000 n=8+10)
MemmoveUnalignedDst/0-16 1.03ns ± 0% 1.03ns ± 0% +0.19% (p=0.007 n=10+10)
MemmoveUnalignedDst/1-16 1.24ns ± 0% 1.25ns ± 1% +0.52% (p=0.022 n=10+10)
MemmoveUnalignedDst/2-16 1.23ns ± 0% 1.23ns ± 0% ~ (p=0.051 n=10+10)
MemmoveUnalignedDst/3-16 1.23ns ± 0% 1.23ns ± 0% +0.14% (p=0.006 n=9+9)
MemmoveUnalignedDst/4-16 1.23ns ± 0% 1.24ns ± 1% +0.37% (p=0.004 n=10+10)
MemmoveUnalignedDst/5-16 1.35ns ± 0% 1.35ns ± 0% ~ (p=0.075 n=10+10)
MemmoveUnalignedDst/6-16 1.34ns ± 0% 1.34ns ± 0% ~ (p=0.779 n=10+10)
MemmoveUnalignedDst/7-16 1.34ns ± 0% 1.34ns ± 0% ~ (p=1.000 n=10+10)
MemmoveUnalignedDst/8-16 1.34ns ± 0% 1.35ns ± 1% +0.39% (p=0.024 n=10+10)
MemmoveUnalignedDst/9-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.849 n=10+10)
MemmoveUnalignedDst/10-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.255 n=10+10)
MemmoveUnalignedDst/11-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.304 n=10+10)
MemmoveUnalignedDst/12-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.672 n=10+10)
MemmoveUnalignedDst/13-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.435 n=10+10)
MemmoveUnalignedDst/14-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.340 n=10+10)
MemmoveUnalignedDst/15-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.911 n=10+9)
MemmoveUnalignedDst/16-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.074 n=10+10)
MemmoveUnalignedDst/32-16 1.62ns ± 0% 1.63ns ± 0% ~ (p=0.059 n=10+10)
MemmoveUnalignedDst/64-16 1.65ns ± 0% 1.65ns ± 0% ~ (p=0.234 n=10+10)
MemmoveUnalignedDst/128-16 2.06ns ± 0% 2.06ns ± 0% ~ (p=0.709 n=10+9)
MemmoveUnalignedDst/256-16 3.69ns ± 0% 3.70ns ± 0% ~ (p=0.144 n=10+10)
MemmoveUnalignedDst/512-16 4.15ns ± 1% 4.14ns ± 0% ~ (p=0.778 n=10+8)
MemmoveUnalignedDst/1024-16 7.52ns ± 0% 7.53ns ± 1% ~ (p=0.650 n=9+9)
MemmoveUnalignedDst/2048-16 12.9ns ± 0% 12.9ns ± 1% ~ (p=0.548 n=8+8)
MemmoveUnalignedDst/4096-16 25.4ns ± 0% 25.4ns ± 0% ~ (p=0.947 n=9+9)
MemmoveUnalignedDstOverlap/32-16 4.08ns ± 0% 4.09ns ± 0% ~ (p=0.360 n=10+10)
MemmoveUnalignedDstOverlap/64-16 4.56ns ± 0% 4.56ns ± 0% ~ (p=0.705 n=10+9)
MemmoveUnalignedDstOverlap/128-16 4.67ns ± 0% 4.67ns ± 0% ~ (p=0.397 n=10+10)
MemmoveUnalignedDstOverlap/256-16 5.08ns ± 0% 5.08ns ± 0% ~ (p=0.159 n=10+9)
MemmoveUnalignedDstOverlap/512-16 8.45ns ± 5% 8.19ns ± 0% -3.10% (p=0.021 n=10+9)
MemmoveUnalignedDstOverlap/1024-16 9.55ns ± 0% 9.56ns ± 0% ~ (p=0.221 n=8+8)
MemmoveUnalignedDstOverlap/2048-16 14.0ns ± 0% 14.0ns ± 1% ~ (p=0.200 n=10+9)
MemmoveUnalignedDstOverlap/4096-16 26.5ns ± 0% 26.5ns ± 0% ~ (p=0.458 n=10+9)
MemmoveUnalignedSrc/0-16 1.02ns ± 1% 0.99ns ± 1% -2.67% (p=0.000 n=10+9)
MemmoveUnalignedSrc/1-16 1.13ns ± 0% 1.13ns ± 1% -0.25% (p=0.027 n=10+9)
MemmoveUnalignedSrc/2-16 1.13ns ± 1% 1.13ns ± 0% -0.28% (p=0.012 n=10+9)
MemmoveUnalignedSrc/3-16 1.24ns ± 1% 1.23ns ± 0% -0.25% (p=0.022 n=9+10)
MemmoveUnalignedSrc/4-16 1.24ns ± 0% 1.23ns ± 1% ~ (p=0.118 n=9+10)
MemmoveUnalignedSrc/5-16 1.34ns ± 0% 1.34ns ± 1% ~ (p=0.564 n=8+10)
MemmoveUnalignedSrc/6-16 1.34ns ± 0% 1.34ns ± 0% -0.39% (p=0.000 n=10+10)
MemmoveUnalignedSrc/7-16 1.34ns ± 0% 1.34ns ± 0% ~ (p=0.235 n=10+10)
MemmoveUnalignedSrc/8-16 1.34ns ± 0% 1.34ns ± 0% -0.37% (p=0.002 n=10+9)
MemmoveUnalignedSrc/9-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.579 n=10+9)
MemmoveUnalignedSrc/10-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.534 n=10+9)
MemmoveUnalignedSrc/11-16 1.44ns ± 0% 1.44ns ± 1% ~ (p=0.415 n=10+10)
MemmoveUnalignedSrc/12-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.218 n=10+10)
MemmoveUnalignedSrc/13-16 1.44ns ± 0% 1.44ns ± 1% ~ (p=0.693 n=10+10)
MemmoveUnalignedSrc/14-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.901 n=10+10)
MemmoveUnalignedSrc/15-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.379 n=10+10)
MemmoveUnalignedSrc/16-16 1.44ns ± 1% 1.44ns ± 0% ~ (p=0.538 n=10+10)
MemmoveUnalignedSrc/32-16 1.60ns ± 1% 1.60ns ± 0% ~ (p=0.491 n=10+10)
MemmoveUnalignedSrc/64-16 1.65ns ± 0% 1.65ns ± 0% ~ (p=0.564 n=10+10)
MemmoveUnalignedSrc/128-16 2.09ns ± 0% 2.09ns ± 0% ~ (p=0.497 n=10+9)
MemmoveUnalignedSrc/256-16 2.70ns ± 0% 2.78ns ± 1% +2.81% (p=0.000 n=10+10)
MemmoveUnalignedSrc/512-16 4.31ns ± 0% 4.30ns ± 0% -0.26% (p=0.031 n=8+9)
MemmoveUnalignedSrc/1024-16 7.28ns ± 0% 7.21ns ± 1% -1.05% (p=0.000 n=8+10)
MemmoveUnalignedSrc/2048-16 13.0ns ± 0% 13.0ns ± 0% ~ (p=0.180 n=9+8)
MemmoveUnalignedSrc/4096-16 25.4ns ± 0% 25.3ns ± 1% ~ (p=0.054 n=10+10)
MemmoveUnalignedSrcOverlap/32-16 4.04ns ± 0% 4.06ns ± 0% +0.62% (p=0.000 n=9+10)
MemmoveUnalignedSrcOverlap/64-16 4.12ns ± 0% 4.12ns ± 0% ~ (p=0.421 n=10+10)
MemmoveUnalignedSrcOverlap/128-16 4.53ns ± 0% 4.52ns ± 0% ~ (p=0.251 n=10+10)
MemmoveUnalignedSrcOverlap/256-16 6.17ns ± 0% 6.15ns ± 0% -0.35% (p=0.000 n=10+9)
MemmoveUnalignedSrcOverlap/512-16 7.43ns ± 0% 7.44ns ± 0% ~ (p=0.524 n=9+8)
MemmoveUnalignedSrcOverlap/1024-16 8.94ns ± 0% 8.94ns ± 0% ~ (p=0.419 n=8+8)
MemmoveUnalignedSrcOverlap/2048-16 13.2ns ± 0% 14.5ns ±21% ~ (p=0.107 n=8+10)
MemmoveUnalignedSrcOverlap/4096-16 25.6ns ± 0% 25.6ns ± 1% ~ (p=0.650 n=9+9)
Memclr/5-16 0.86ns ± 1% 0.86ns ± 2% ~ (p=0.531 n=9+9)
Memclr/16-16 1.04ns ± 0% 1.04ns ± 0% +0.32% (p=0.013 n=9+10)
Memclr/64-16 1.23ns ± 0% 1.26ns ± 0% +2.28% (p=0.000 n=10+10)
Memclr/256-16 2.27ns ± 0% 2.27ns ± 0% ~ (p=0.127 n=10+10)
Memclr/4096-16 17.1ns ± 1% 17.3ns ± 0% +0.88% (p=0.000 n=10+10)
Memclr/65536-16 821ns ± 0% 822ns ± 0% ~ (p=0.516 n=10+10)
Memclr/1M-16 14.1µs ± 1% 14.0µs ± 1% ~ (p=0.516 n=10+10)
Memclr/4M-16 86.1µs ± 1% 85.9µs ± 0% ~ (p=0.123 n=10+10)
Memclr/8M-16 174µs ± 2% 173µs ± 0% ~ (p=0.408 n=10+8)
Memclr/16M-16 385µs ± 4% 387µs ± 0% ~ (p=0.173 n=10+8)
Memclr/64M-16 2.18ms ± 0% 2.19ms ± 0% ~ (p=0.113 n=10+9)
GoMemclr/5-16 0.82ns ± 0% 0.82ns ± 0% ~ (p=0.346 n=9+10)
GoMemclr/16-16 1.02ns ± 0% 1.02ns ± 0% +0.22% (p=0.003 n=10+8)
GoMemclr/64-16 1.14ns ± 0% 1.14ns ± 0% ~ (p=0.948 n=10+9)
GoMemclr/256-16 2.06ns ± 0% 2.06ns ± 0% ~ (p=0.868 n=10+10)
MemclrRange/1K_2K-16 457ns ± 0% 428ns ± 1% -6.38% (p=0.000 n=10+10)
MemclrRange/2K_8K-16 1.46µs ± 0% 1.46µs ± 0% ~ (p=0.700 n=10+10)
MemclrRange/4K_16K-16 1.16µs ± 0% 1.16µs ± 0% ~ (p=0.567 n=9+10)
MemclrRange/160K_228K-16 20.7µs ± 0% 20.7µs ± 0% ~ (p=0.160 n=10+10)
ClearFat7-16 0.38ns ± 5% 0.21ns ± 1% -45.79% (p=0.000 n=9+10)
ClearFat8-16 0.21ns ± 3% 0.12ns ± 2% -44.16% (p=0.000 n=8+9)
ClearFat11-16 0.35ns ± 3% 0.21ns ± 1% -40.46% (p=0.000 n=9+9)
ClearFat12-16 0.23ns ± 9% 0.21ns ± 1% -10.23% (p=0.000 n=10+9)
ClearFat13-16 0.22ns ± 6% 0.21ns ± 2% -6.53% (p=0.000 n=10+10)
ClearFat14-16 0.22ns ± 4% 0.21ns ± 1% -5.97% (p=0.000 n=10+10)
ClearFat15-16 0.22ns ± 4% 0.21ns ± 1% -6.96% (p=0.000 n=10+9)
ClearFat16-16 0.19ns ± 9% 0.12ns ± 6% -34.89% (p=0.000 n=9+10)
ClearFat24-16 0.23ns ± 6% 0.21ns ± 1% -10.26% (p=0.000 n=10+9)
ClearFat32-16 0.22ns ± 5% 0.21ns ± 2% -5.31% (p=0.000 n=10+10)
ClearFat40-16 0.34ns ± 4% 0.62ns ± 1% +83.00% (p=0.000 n=10+10)
ClearFat48-16 0.33ns ± 2% 0.41ns ± 0% +26.71% (p=0.000 n=10+10)
ClearFat56-16 0.41ns ± 1% 0.41ns ± 0% ~ (p=0.838 n=10+10)
ClearFat64-16 0.41ns ± 0% 0.41ns ± 0% ~ (p=0.178 n=10+8)
ClearFat72-16 0.82ns ± 0% 0.82ns ± 0% ~ (p=0.669 n=10+10)
ClearFat128-16 1.04ns ± 0% 1.04ns ± 0% ~ (p=0.679 n=10+10)
ClearFat256-16 1.86ns ± 0% 1.86ns ± 0% ~ (p=0.066 n=9+10)
ClearFat512-16 3.50ns ± 0% 3.50ns ± 0% ~ (p=0.626 n=10+10)
ClearFat1024-16 6.79ns ± 0% 6.79ns ± 0% ~ (p=0.986 n=10+10)
ClearFat1032-16 13.6ns ± 0% 13.6ns ± 0% +0.13% (p=0.044 n=10+10)
ClearFat1040-16 10.3ns ± 0% 10.3ns ± 0% ~ (p=0.175 n=10+9)
CopyFat7-16 0.37ns ±13% 0.25ns ± 1% -31.74% (p=0.000 n=10+9)
CopyFat8-16 0.17ns ± 1% 0.17ns ± 2% +1.35% (p=0.004 n=9+9)
CopyFat11-16 0.26ns ± 1% 0.30ns ± 3% +12.58% (p=0.000 n=9+10)
CopyFat12-16 0.28ns ± 2% 0.26ns ± 1% -5.66% (p=0.000 n=9+9)
CopyFat13-16 0.26ns ± 0% 0.28ns ± 4% +7.35% (p=0.000 n=8+10)
CopyFat14-16 0.29ns ± 6% 0.26ns ± 2% -10.46% (p=0.000 n=10+9)
CopyFat15-16 0.26ns ± 1% 0.30ns ± 6% +14.12% (p=0.000 n=8+10)
CopyFat16-16 0.21ns ± 1% 0.21ns ± 0% ~ (p=0.426 n=8+8)
CopyFat24-16 0.29ns ± 3% 0.25ns ± 1% -12.27% (p=0.000 n=9+10)
CopyFat32-16 0.26ns ± 4% 0.29ns ± 4% +11.71% (p=0.000 n=10+10)
CopyFat64-16 0.46ns ± 8% 0.42ns ± 1% -8.37% (p=0.002 n=10+10)
CopyFat72-16 0.82ns ± 0% 0.82ns ± 0% ~ (p=0.563 n=10+10)
CopyFat128-16 1.53ns ± 0% 1.54ns ± 0% +0.62% (p=0.000 n=10+10)
CopyFat256-16 2.68ns ± 0% 2.65ns ± 1% -1.23% (p=0.000 n=10+10)
CopyFat512-16 4.93ns ± 1% 5.19ns ± 3% +5.16% (p=0.000 n=9+9)
CopyFat520-16 6.99ns ± 0% 6.99ns ± 0% ~ (p=0.539 n=10+10)
CopyFat1024-16 11.5ns ± 1% 9.8ns ± 1% -14.98% (p=0.000 n=9+10)
CopyFat1032-16 13.6ns ± 0% 13.6ns ± 0% ~ (p=0.728 n=10+10)
CopyFat1040-16 11.0ns ± 0% 11.1ns ± 0% +0.53% (p=0.000 n=10+10)
Issue18740/2byte-16 10.1µs ± 0% 10.1µs ± 0% ~ (p=0.342 n=10+10)
Issue18740/4byte-16 2.34µs ± 0% 2.35µs ± 0% +0.30% (p=0.002 n=10+8)
Issue18740/8byte-16 1.28µs ± 0% 1.28µs ± 0% +0.32% (p=0.000 n=9+10)
Finalizer-16 345µs ± 1% 336µs ± 0% -2.55% (p=0.000 n=10+9)
FinalizerRun-16 450ns ± 3% 420ns ± 1% -6.65% (p=0.000 n=10+10)
PallocBitsSummarize/Unpacked00-16 2.88ns ± 0% 2.88ns ± 0% ~ (p=0.358 n=10+10)
PallocBitsSummarize/UnpackedFFFFFFFFFFFFFFFF-16 15.2ns ± 0% 15.2ns ± 0% ~ (p=0.925 n=10+10)
PallocBitsSummarize/UnpackedAA-16 16.4ns ± 0% 16.3ns ± 0% ~ (p=0.113 n=9+9)
PallocBitsSummarize/UnpackedAAAAAAAAAAAAAAAA-16 16.5ns ± 0% 16.6ns ± 0% ~ (p=0.238 n=10+10)
PallocBitsSummarize/Unpacked80000000AAAAAAAA-16 37.8ns ± 1% 36.4ns ± 0% -3.70% (p=0.000 n=10+9)
PallocBitsSummarize/UnpackedAAAAAAAA00000001-16 41.8ns ± 1% 39.9ns ± 0% -4.68% (p=0.000 n=9+10)
PallocBitsSummarize/UnpackedBBBBBBBBBBBBBBBB-16 18.3ns ± 0% 18.3ns ± 0% ~ (p=0.781 n=10+10)
PallocBitsSummarize/Unpacked80000000BBBBBBBB-16 38.8ns ± 1% 38.1ns ± 0% -1.78% (p=0.000 n=9+10)
PallocBitsSummarize/UnpackedBBBBBBBB00000001-16 37.5ns ± 0% 36.1ns ± 1% -3.88% (p=0.000 n=8+10)
PallocBitsSummarize/UnpackedCCCCCCCCCCCCCCCC-16 21.8ns ± 0% 21.9ns ± 0% +0.20% (p=0.018 n=10+9)
PallocBitsSummarize/Unpacked4444444444444444-16 21.8ns ± 0% 21.9ns ± 0% +0.20% (p=0.029 n=10+9)
PallocBitsSummarize/Unpacked4040404040404040-16 26.5ns ± 0% 26.5ns ± 0% -0.24% (p=0.001 n=9+10)
PallocBitsSummarize/Unpacked4000400040004000-16 33.4ns ± 1% 31.3ns ± 0% -6.20% (p=0.000 n=9+10)
PallocBitsSummarize/Unpacked1000404044CCAAFF-16 36.4ns ± 1% 35.9ns ± 0% -1.50% (p=0.000 n=10+10)
FindBitRange64/Pattern00Size2-16 0.34ns ± 1% 0.35ns ± 1% +3.80% (p=0.000 n=10+9)
FindBitRange64/Pattern00Size8-16 0.70ns ± 1% 0.70ns ± 0% -0.68% (p=0.000 n=10+10)
FindBitRange64/Pattern00Size32-16 0.70ns ± 1% 0.69ns ± 0% -0.86% (p=0.001 n=10+8)
FindBitRange64/PatternFFFFFFFFFFFFFFFFSize2-16 0.34ns ± 1% 0.35ns ± 1% +4.45% (p=0.000 n=9+8)
FindBitRange64/PatternFFFFFFFFFFFFFFFFSize8-16 1.54ns ± 0% 1.54ns ± 1% ~ (p=0.914 n=9+9)
FindBitRange64/PatternFFFFFFFFFFFFFFFFSize32-16 2.78ns ± 0% 2.78ns ± 0% ~ (p=0.295 n=9+10)
FindBitRange64/PatternAASize2-16 0.34ns ± 2% 0.35ns ± 2% +4.61% (p=0.000 n=10+10)
FindBitRange64/PatternAASize8-16 0.70ns ± 1% 0.70ns ± 1% -0.82% (p=0.005 n=10+10)
FindBitRange64/PatternAASize32-16 0.70ns ± 1% 0.70ns ± 0% -0.73% (p=0.003 n=10+9)
FindBitRange64/PatternAAAAAAAAAAAAAAAASize2-16 0.34ns ± 2% 0.35ns ± 2% +3.94% (p=0.000 n=10+10)
FindBitRange64/PatternAAAAAAAAAAAAAAAASize8-16 0.70ns ± 1% 0.70ns ± 1% -0.67% (p=0.025 n=10+10)
FindBitRange64/PatternAAAAAAAAAAAAAAAASize32-16 0.70ns ± 1% 0.70ns ± 1% ~ (p=0.118 n=9+10)
FindBitRange64/Pattern80000000AAAAAAAASize2-16 0.34ns ± 1% 0.35ns ± 2% +3.72% (p=0.000 n=10+9)
FindBitRange64/Pattern80000000AAAAAAAASize8-16 0.70ns ± 1% 0.70ns ± 0% ~ (p=0.102 n=10+10)
FindBitRange64/Pattern80000000AAAAAAAASize32-16 0.70ns ± 1% 0.70ns ± 1% -0.55% (p=0.011 n=10+10)
FindBitRange64/PatternAAAAAAAA00000001Size2-16 0.34ns ± 2% 0.35ns ± 1% +3.83% (p=0.000 n=10+9)
FindBitRange64/PatternAAAAAAAA00000001Size8-16 0.70ns ± 1% 0.70ns ± 1% ~ (p=0.065 n=10+10)
FindBitRange64/PatternAAAAAAAA00000001Size32-16 0.70ns ± 1% 0.70ns ± 1% -0.95% (p=0.002 n=10+10)
FindBitRange64/PatternBBBBBBBBBBBBBBBBSize2-16 0.34ns ± 0% 0.35ns ± 1% +4.12% (p=0.000 n=8+10)
FindBitRange64/PatternBBBBBBBBBBBBBBBBSize8-16 1.24ns ± 0% 1.23ns ± 0% -0.30% (p=0.002 n=10+9)
FindBitRange64/PatternBBBBBBBBBBBBBBBBSize32-16 1.24ns ± 0% 1.24ns ± 0% -0.17% (p=0.023 n=9+10)
FindBitRange64/Pattern80000000BBBBBBBBSize2-16 0.34ns ± 1% 0.35ns ± 2% +4.82% (p=0.000 n=9+10)
FindBitRange64/Pattern80000000BBBBBBBBSize8-16 1.24ns ± 1% 1.24ns ± 0% ~ (p=0.063 n=10+10)
FindBitRange64/Pattern80000000BBBBBBBBSize32-16 1.24ns ± 0% 1.24ns ± 0% ~ (p=0.164 n=9+10)
FindBitRange64/PatternBBBBBBBB00000001Size2-16 0.34ns ± 1% 0.35ns ± 1% +4.38% (p=0.000 n=8+10)
FindBitRange64/PatternBBBBBBBB00000001Size8-16 1.24ns ± 1% 1.24ns ± 0% ~ (p=0.052 n=10+10)
FindBitRange64/PatternBBBBBBBB00000001Size32-16 1.24ns ± 0% 1.23ns ± 0% -0.40% (p=0.000 n=10+10)
FindBitRange64/PatternCCCCCCCCCCCCCCCCSize2-16 0.34ns ± 0% 0.35ns ± 2% +3.96% (p=0.000 n=9+10)
FindBitRange64/PatternCCCCCCCCCCCCCCCCSize8-16 1.24ns ± 0% 1.23ns ± 0% -0.30% (p=0.000 n=10+9)
FindBitRange64/PatternCCCCCCCCCCCCCCCCSize32-16 1.24ns ± 0% 1.24ns ± 1% ~ (p=0.284 n=10+10)
FindBitRange64/Pattern4444444444444444Size2-16 0.34ns ± 1% 0.35ns ± 1% +3.91% (p=0.000 n=9+9)
FindBitRange64/Pattern4444444444444444Size8-16 0.70ns ± 1% 0.70ns ± 1% ~ (p=0.617 n=10+10)
FindBitRange64/Pattern4444444444444444Size32-16 0.70ns ± 1% 0.70ns ± 1% -0.60% (p=0.006 n=10+10)
FindBitRange64/Pattern4040404040404040Size2-16 0.34ns ± 2% 0.35ns ± 2% +3.67% (p=0.000 n=10+10)
FindBitRange64/Pattern4040404040404040Size8-16 0.70ns ± 2% 0.70ns ± 1% -0.87% (p=0.014 n=10+10)
FindBitRange64/Pattern4040404040404040Size32-16 0.70ns ± 1% 0.70ns ± 1% ~ (p=0.256 n=10+10)
FindBitRange64/Pattern4000400040004000Size2-16 0.34ns ± 2% 0.35ns ± 3% +4.71% (p=0.000 n=10+10)
FindBitRange64/Pattern4000400040004000Size8-16 0.70ns ± 1% 0.70ns ± 1% ~ (p=0.393 n=10+10)
FindBitRange64/Pattern4000400040004000Size32-16 0.70ns ± 1% 0.70ns ± 1% -0.86% (p=0.014 n=10+10)
NetpollBreak-16 1.49µs ± 1% 1.50µs ± 3% ~ (p=0.181 n=8+10)
Syscall-16 3.68ns ± 1% 3.66ns ± 2% ~ (p=0.148 n=10+10)
SyscallWork-16 5.15ns ± 1% 5.13ns ± 0% ~ (p=0.188 n=10+9)
SyscallExcess-16 3.89ns ± 2% 3.83ns ± 1% -1.52% (p=0.001 n=10+10)
SyscallExcessWork-16 5.34ns ± 1% 5.31ns ± 0% -0.64% (p=0.000 n=10+9)
PingPongHog-16 397ns ± 7% 394ns ±11% ~ (p=0.912 n=10+10)
StackGrowth-16 67.9ns ± 0% 68.8ns ± 0% +1.28% (p=0.000 n=10+8)
StackGrowthDeep-16 7.70µs ± 1% 8.48µs ± 2% +10.06% (p=0.000 n=9+10)
CreateGoroutines-16 124ns ± 1% 124ns ± 1% ~ (p=0.254 n=10+10)
CreateGoroutinesParallel-16 25.7ns ± 1% 27.6ns ± 2% +7.51% (p=0.000 n=10+10)
CreateGoroutinesCapture-16 823ns ± 1% 821ns ± 2% ~ (p=0.699 n=10+10)
CreateGoroutinesSingle-16 175ns ± 3% 172ns ± 3% -1.90% (p=0.011 n=10+10)
ClosureCall-16 0.11ns ± 7% 0.12ns ± 3% ~ (p=0.842 n=9+10)
WakeupParallelSpinning/0s-16 11.4µs ± 0% 11.4µs ± 0% ~ (p=0.325 n=9+10)
WakeupParallelSpinning/1µs-16 15.4µs ± 0% 15.4µs ± 1% ~ (p=0.955 n=10+10)
WakeupParallelSpinning/2µs-16 18.7µs ± 2% 18.9µs ± 2% ~ (p=0.052 n=10+10)
WakeupParallelSpinning/5µs-16 30.7µs ± 0% 30.7µs ± 0% -0.03% (p=0.003 n=10+10)
WakeupParallelSpinning/10µs-16 48.8µs ± 0% 48.8µs ± 0% ~ (p=0.670 n=10+10)
WakeupParallelSpinning/20µs-16 90.8µs ± 0% 90.8µs ± 0% -0.02% (p=0.004 n=10+10)
WakeupParallelSpinning/50µs-16 211µs ± 0% 211µs ± 0% ~ (p=0.194 n=10+10)
WakeupParallelSpinning/100µs-16 323µs ± 0% 323µs ± 0% ~ (p=1.000 n=10+9)
WakeupParallelSyscall/0s-16 118µs ± 0% 118µs ± 0% ~ (p=0.447 n=10+9)
WakeupParallelSyscall/1µs-16 119µs ± 2% 119µs ± 1% ~ (p=0.604 n=10+9)
WakeupParallelSyscall/2µs-16 120µs ± 1% 121µs ± 3% ~ (p=0.263 n=8+10)
WakeupParallelSyscall/5µs-16 126µs ± 2% 126µs ± 2% ~ (p=0.510 n=10+9)
WakeupParallelSyscall/10µs-16 136µs ± 1% 137µs ± 1% ~ (p=0.095 n=9+10)
WakeupParallelSyscall/20µs-16 156µs ± 2% 157µs ± 3% ~ (p=0.604 n=10+9)
WakeupParallelSyscall/50µs-16 221µs ± 1% 220µs ± 1% ~ (p=0.063 n=10+10)
WakeupParallelSyscall/100µs-16 326µs ± 0% 325µs ± 0% -0.26% (p=0.003 n=9+10)
Matmult-16 0.67ns ± 2% 0.66ns ± 2% ~ (p=0.256 n=10+10)
Fastrand-16 0.08ns ±11% 0.08ns ±13% ~ (p=0.661 n=9+10)
Fastrand64-16 0.08ns ±11% 0.08ns ± 6% ~ (p=0.631 n=10+10)
FastrandHashiter-16 1.76ns ± 1% 1.76ns ± 1% ~ (p=0.854 n=8+8)
Fastrandn/2-16 0.86ns ± 1% 0.86ns ± 1% +1.09% (p=0.000 n=10+9)
Fastrandn/3-16 0.85ns ± 1% 0.86ns ± 1% +1.23% (p=0.001 n=10+10)
Fastrandn/4-16 0.85ns ± 1% 0.87ns ± 2% +1.60% (p=0.000 n=10+10)
Fastrandn/5-16 0.85ns ± 1% 0.86ns ± 1% +1.05% (p=0.000 n=10+10)
IfaceCmp100-16 46.6ns ± 0% 46.1ns ± 0% -1.18% (p=0.000 n=10+10)
IfaceCmpNil100-16 26.8ns ± 0% 26.8ns ± 0% ~ (p=0.777 n=10+8)
EfaceCmpDiff-16 132ns ± 0% 130ns ± 0% -0.95% (p=0.000 n=10+9)
EfaceCmpDiffIndirect-16 209ns ± 0% 211ns ± 0% +1.14% (p=0.000 n=10+9)
Defer-16 3.40ns ± 1% 3.04ns ± 0% -10.67% (p=0.000 n=10+10)
Defer10-16 29.4ns ± 2% 27.2ns ± 3% -7.26% (p=0.000 n=10+10)
DeferMany-16 110ns ± 6% 113ns ± 2% +3.45% (p=0.017 n=9+9)
PanicRecover-16 67.6ns ± 0% 67.7ns ± 2% ~ (p=0.436 n=9+9)
GoroutineProfile/small-nil/idle-16 3.90µs ± 4% 3.86µs ± 2% ~ (p=0.305 n=10+9)
GoroutineProfile/small-nil/loaded-16 4.82µs ± 6% 4.82µs ± 4% ~ (p=0.905 n=10+9)
GoroutineProfile/small/idle-16 103µs ± 3% 102µs ± 3% ~ (p=0.113 n=9+9)
GoroutineProfile/small/loaded-16 432µs ± 5% 440µs ±13% ~ (p=0.604 n=9+10)
GoroutineProfile/large-nil/idle-16 3.86µs ± 3% 3.82µs ± 3% ~ (p=0.210 n=10+10)
GoroutineProfile/large-nil/loaded-16 4.90µs ± 2% 4.90µs ± 5% ~ (p=0.780 n=10+9)
GoroutineProfile/large/idle-16 2.58ms ± 1% 2.52ms ± 1% -2.38% (p=0.000 n=10+10)
GoroutineProfile/large/loaded-16 8.62ms ± 9% 8.90ms ±11% ~ (p=0.400 n=9+10)
GoroutineProfile/sparse-nil/idle-16 3.85µs ± 4% 3.81µs ± 3% ~ (p=0.470 n=10+10)
GoroutineProfile/sparse-nil/loaded-16 4.82µs ± 4% 4.69µs ± 5% ~ (p=0.052 n=10+10)
GoroutineProfile/sparse/idle-16 102µs ± 4% 102µs ± 2% ~ (p=0.497 n=10+9)
GoroutineProfile/sparse/loaded-16 438µs ± 7% 437µs ± 6% ~ (p=0.796 n=10+10)
RWMutexUncontended-16 6.79ns ± 0% 6.78ns ± 0% ~ (p=0.228 n=10+8)
RWMutexWrite100-16 85.4ns ± 0% 87.1ns ± 0% +2.00% (p=0.000 n=10+8)
RWMutexWrite10-16 168ns ±25% 152ns ±11% ~ (p=0.063 n=10+10)
RWMutexWorkWrite100-16 106ns ± 0% 106ns ± 3% ~ (p=0.136 n=10+10)
RWMutexWorkWrite10-16 567ns ± 3% 571ns ± 1% ~ (p=0.326 n=10+9)
SemTable/OneAddrCollision/n=1000-16 15.9µs ± 1% 16.0µs ± 1% +0.50% (p=0.031 n=9+9)
SemTable/ManyAddrCollision/n=1000-16 56.2µs ± 1% 56.8µs ± 1% +1.06% (p=0.000 n=10+10)
SemTable/OneAddrCollision/n=2000-16 32.6µs ± 2% 32.9µs ± 4% ~ (p=0.156 n=10+9)
SemTable/ManyAddrCollision/n=2000-16 118µs ± 0% 119µs ± 0% +0.75% (p=0.000 n=9+10)
SemTable/OneAddrCollision/n=4000-16 65.3µs ± 1% 65.6µs ± 3% ~ (p=0.497 n=9+10)
SemTable/ManyAddrCollision/n=4000-16 245µs ± 0% 248µs ± 2% +1.36% (p=0.000 n=9+10)
SemTable/OneAddrCollision/n=8000-16 131µs ± 1% 130µs ± 1% -1.01% (p=0.002 n=9+10)
SemTable/ManyAddrCollision/n=8000-16 503µs ± 1% 508µs ± 0% +0.97% (p=0.000 n=10+10)
MakeSliceCopy/mallocmove/Byte-16 67.6ns ± 1% 64.1ns ± 2% -5.20% (p=0.000 n=10+10)
MakeSliceCopy/mallocmove/Int-16 65.0ns ± 7% 61.7ns ± 4% -5.08% (p=0.009 n=10+10)
MakeSliceCopy/mallocmove/Ptr-16 88.1ns ± 1% 79.9ns ± 1% -9.29% (p=0.000 n=10+10)
MakeSliceCopy/makecopy/Byte-16 65.2ns ± 6% 63.4ns ± 0% ~ (p=0.500 n=10+8)
MakeSliceCopy/makecopy/Int-16 63.2ns ± 1% 64.1ns ± 1% +1.34% (p=0.001 n=9+9)
MakeSliceCopy/makecopy/Ptr-16 88.1ns ± 1% 80.1ns ± 1% -9.09% (p=0.000 n=10+10)
MakeSliceCopy/nilappend/Byte-16 69.8ns ± 1% 65.7ns ± 3% -5.80% (p=0.000 n=10+10)
MakeSliceCopy/nilappend/Int-16 69.6ns ± 2% 67.2ns ± 1% -3.50% (p=0.000 n=10+9)
MakeSliceCopy/nilappend/Ptr-16 91.5ns ± 1% 83.8ns ± 1% -8.42% (p=0.000 n=9+10)
MakeSlice/Byte-16 6.64ns ± 3% 6.58ns ± 2% ~ (p=0.393 n=10+10)
MakeSlice/Int16-16 8.60ns ± 1% 8.38ns ± 3% -2.48% (p=0.001 n=9+10)
MakeSlice/Int-16 17.7ns ± 3% 16.9ns ± 1% -4.67% (p=0.000 n=10+9)
MakeSlice/Ptr-16 24.0ns ± 3% 23.3ns ± 2% -3.25% (p=0.000 n=10+9)
MakeSlice/Struct/24-16 34.1ns ± 1% 32.0ns ± 1% -6.11% (p=0.000 n=10+10)
MakeSlice/Struct/32-16 39.1ns ± 4% 38.2ns ± 1% ~ (p=0.829 n=10+8)
MakeSlice/Struct/40-16 47.0ns ± 5% 43.0ns ± 2% -8.55% (p=0.000 n=10+9)
GrowSlice/Byte-16 15.3ns ± 3% 15.0ns ± 2% -1.75% (p=0.005 n=9+9)
GrowSlice/Int16-16 18.9ns ± 2% 18.4ns ± 2% -2.71% (p=0.000 n=10+9)
GrowSlice/Int-16 33.9ns ± 1% 32.2ns ± 1% -4.89% (p=0.000 n=10+9)
GrowSlice/Ptr-16 45.3ns ± 2% 43.5ns ± 1% -4.12% (p=0.000 n=10+10)
GrowSlice/Struct/24-16 61.9ns ± 2% 60.0ns ± 4% -3.10% (p=0.002 n=10+10)
GrowSlice/Struct/32-16 79.9ns ± 2% 72.3ns ± 3% -9.58% (p=0.000 n=8+10)
GrowSlice/Struct/40-16 97.1ns ± 7% 88.8ns ± 5% -8.49% (p=0.000 n=10+10)
ExtendSlice/IntSlice-16 21.1ns ± 2% 20.3ns ± 2% -3.71% (p=0.000 n=10+10)
ExtendSlice/PointerSlice-16 26.8ns ± 2% 26.3ns ± 2% -1.86% (p=0.004 n=10+10)
ExtendSlice/NoGrow-16 1.23ns ± 0% 1.30ns ± 1% +5.03% (p=0.000 n=10+10)
Append-16 4.58ns ± 1% 4.53ns ± 0% -1.11% (p=0.000 n=10+10)
AppendGrowByte-16 1.46ms ± 8% 1.42ms ± 7% -3.24% (p=0.035 n=10+10)
AppendGrowString-16 27.8ms ± 4% 27.2ms ± 5% ~ (p=0.052 n=10+10)
AppendSlice/1Bytes-16 1.03ns ± 1% 1.04ns ± 1% ~ (p=0.303 n=10+10)
AppendSlice/4Bytes-16 1.04ns ± 0% 1.05ns ± 0% +0.79% (p=0.000 n=9+10)
AppendSlice/7Bytes-16 1.23ns ± 1% 1.24ns ± 0% +0.45% (p=0.001 n=10+10)
AppendSlice/8Bytes-16 1.24ns ± 0% 1.24ns ± 0% ~ (p=0.183 n=10+10)
AppendSlice/15Bytes-16 1.37ns ± 1% 1.43ns ± 1% +3.88% (p=0.000 n=10+10)
AppendSlice/16Bytes-16 1.37ns ± 1% 1.42ns ± 1% +3.63% (p=0.000 n=9+10)
AppendSlice/32Bytes-16 1.44ns ± 0% 1.47ns ± 1% +1.83% (p=0.000 n=10+10)
AppendSliceLarge/1024Bytes-16 257ns ± 2% 234ns ± 1% -8.96% (p=0.000 n=8+9)
AppendSliceLarge/4096Bytes-16 871ns ± 6% 812ns ± 1% -6.80% (p=0.000 n=10+10)
AppendSliceLarge/16384Bytes-16 3.15µs ± 6% 3.04µs ± 5% ~ (p=0.052 n=10+10)
AppendSliceLarge/65536Bytes-16 10.7µs ± 7% 10.8µs ± 2% ~ (p=0.278 n=10+9)
AppendSliceLarge/262144Bytes-16 42.9µs ± 1% 39.6µs ± 5% -7.75% (p=0.000 n=9+10)
AppendSliceLarge/1048576Bytes-16 147µs ± 4% 144µs ± 4% -2.21% (p=0.035 n=10+10)
AppendStr/1Bytes-16 1.20ns ± 0% 1.20ns ± 0% ~ (p=0.755 n=10+10)
AppendStr/4Bytes-16 1.13ns ± 0% 1.14ns ± 1% +1.20% (p=0.000 n=10+10)
AppendStr/8Bytes-16 1.24ns ± 0% 1.25ns ± 0% +0.93% (p=0.000 n=10+10)
AppendStr/16Bytes-16 1.40ns ± 0% 1.42ns ± 0% +2.10% (p=0.000 n=9+10)
AppendStr/32Bytes-16 1.44ns ± 0% 1.45ns ± 0% +0.99% (p=0.000 n=10+10)
AppendSpecialCase-16 8.64ns ± 1% 8.89ns ± 2% +2.90% (p=0.000 n=10+10)
Copy/1Byte-16 1.24ns ± 1% 1.24ns ± 0% -0.28% (p=0.000 n=10+6)
Copy/1String-16 1.24ns ± 0% 1.23ns ± 0% ~ (p=0.160 n=10+10)
Copy/2Byte-16 1.24ns ± 0% 1.24ns ± 0% ~ (p=0.115 n=10+10)
Copy/2String-16 1.24ns ± 0% 1.24ns ± 1% ~ (p=0.954 n=10+10)
Copy/4Byte-16 1.24ns ± 0% 1.24ns ± 0% -0.44% (p=0.001 n=10+10)
Copy/4String-16 1.23ns ± 0% 1.23ns ± 0% ~ (p=0.081 n=10+10)
Copy/8Byte-16 1.37ns ± 0% 1.34ns ± 0% -1.79% (p=0.000 n=9+9)
Copy/8String-16 1.34ns ± 0% 1.34ns ± 0% -0.58% (p=0.000 n=9+10)
Copy/12Byte-16 1.44ns ± 0% 1.44ns ± 0% ~ (p=0.149 n=9+9)
Copy/12String-16 1.44ns ± 0% 1.45ns ± 0% ~ (p=0.124 n=9+9)
Copy/16Byte-16 1.44ns ± 0% 1.44ns ± 0% -0.19% (p=0.004 n=10+9)
Copy/16String-16 1.44ns ± 0% 1.45ns ± 0% +0.30% (p=0.008 n=10+10)
Copy/32Byte-16 1.63ns ± 1% 1.62ns ± 1% -0.72% (p=0.002 n=10+10)
Copy/32String-16 1.60ns ± 1% 1.64ns ± 0% +2.23% (p=0.000 n=10+10)
Copy/128Byte-16 2.06ns ± 0% 2.06ns ± 0% ~ (p=0.757 n=9+10)
Copy/128String-16 2.07ns ± 0% 2.07ns ± 0% +0.36% (p=0.004 n=10+10)
Copy/1024Byte-16 6.07ns ± 2% 6.00ns ± 1% -1.20% (p=0.000 n=9+10)
Copy/1024String-16 6.05ns ± 0% 5.95ns ± 1% -1.54% (p=0.000 n=10+9)
AppendInPlace/NoGrow/Byte-16 288ns ± 1% 284ns ± 1% -1.58% (p=0.000 n=10+10)
AppendInPlace/NoGrow/1Ptr-16 844ns ± 1% 809ns ± 3% -4.13% (p=0.000 n=9+10)
AppendInPlace/NoGrow/2Ptr-16 1.47µs ± 1% 1.46µs ± 1% ~ (p=0.388 n=9+10)
AppendInPlace/NoGrow/3Ptr-16 1.87µs ± 7% 1.91µs ± 1% ~ (p=0.166 n=10+8)
AppendInPlace/NoGrow/4Ptr-16 2.66µs ± 1% 2.67µs ± 3% ~ (p=0.968 n=9+10)
AppendInPlace/Grow/Byte-16 126ns ± 2% 121ns ± 2% -4.06% (p=0.000 n=10+10)
AppendInPlace/Grow/1Ptr-16 132ns ± 2% 127ns ± 2% -4.28% (p=0.000 n=10+9)
AppendInPlace/Grow/2Ptr-16 196ns ± 2% 188ns ± 1% -4.20% (p=0.000 n=10+8)
AppendInPlace/Grow/3Ptr-16 264ns ± 1% 260ns ± 1% -1.51% (p=0.000 n=9+10)
AppendInPlace/Grow/4Ptr-16 297ns ± 2% 294ns ± 2% ~ (p=0.085 n=10+10)
StackCopyPtr-16 36.4ms ± 2% 36.7ms ± 2% ~ (p=0.481 n=10+10)
StackCopy-16 33.9ms ± 3% 32.6ms ± 1% -3.87% (p=0.000 n=10+8)
StackCopyNoCache-16 1.00ms ± 5% 1.01ms ± 5% ~ (p=0.143 n=10+10)
StackCopyWithStkobj-16 11.0ms ± 3% 10.9ms ± 4% ~ (p=0.579 n=10+10)
Issue18138-16 49.2µs ± 5% 49.0µs ± 4% ~ (p=1.000 n=10+9)
CompareStringEqual-16 1.39ns ± 1% 1.45ns ± 2% +3.80% (p=0.000 n=8+10)
CompareStringIdentical-16 0.55ns ± 1% 0.55ns ± 0% +0.42% (p=0.007 n=10+10)
CompareStringSameLength-16 1.03ns ± 0% 1.03ns ± 0% ~ (p=0.430 n=9+10)
CompareStringDifferentLength-16 0.11ns ± 2% 0.11ns ± 3% ~ (p=0.139 n=9+10)
CompareStringBigUnaligned-16 23.9µs ± 1% 24.0µs ± 1% ~ (p=0.370 n=9+8)
CompareStringBig-16 22.0µs ± 3% 22.2µs ± 3% ~ (p=0.243 n=9+10)
ConcatStringAndBytes-16 10.7ns ± 1% 10.0ns ± 2% -6.33% (p=0.000 n=10+10)
SliceByteToString/1-16 1.34ns ± 0% 1.34ns ± 0% ~ (p=0.057 n=10+10)
SliceByteToString/2-16 6.67ns ± 2% 6.60ns ± 3% ~ (p=0.101 n=10+10)
SliceByteToString/4-16 7.76ns ± 2% 7.56ns ± 3% -2.59% (p=0.001 n=10+10)
SliceByteToString/8-16 9.81ns ± 4% 9.57ns ± 2% -2.48% (p=0.005 n=10+10)
SliceByteToString/16-16 14.0ns ± 3% 13.7ns ± 2% -2.31% (p=0.009 n=10+10)
SliceByteToString/32-16 17.3ns ± 1% 16.7ns ± 2% -3.41% (p=0.000 n=10+10)
SliceByteToString/64-16 25.1ns ± 1% 24.1ns ± 2% -3.93% (p=0.000 n=9+10)
SliceByteToString/128-16 38.6ns ± 1% 36.5ns ± 1% -5.60% (p=0.000 n=10+10)
RuneCount/lenruneslice/ASCII-16 4.12ns ± 0% 4.11ns ± 0% ~ (p=0.382 n=10+10)
RuneCount/lenruneslice/Japanese-16 25.4ns ± 2% 25.6ns ± 2% ~ (p=0.138 n=9+10)
RuneCount/lenruneslice/MixedLength-16 17.1ns ± 0% 17.2ns ± 0% +0.59% (p=0.000 n=9+9)
RuneCount/rangeloop/ASCII-16 3.30ns ± 1% 3.29ns ± 0% ~ (p=0.267 n=10+10)
RuneCount/rangeloop/Japanese-16 20.1ns ± 1% 24.9ns ± 1% +24.31% (p=0.000 n=9+9)
RuneCount/rangeloop/MixedLength-16 16.5ns ± 1% 16.7ns ± 1% +1.34% (p=0.000 n=10+10)
RuneCount/utf8.RuneCountInString/ASCII-16 5.71ns ± 1% 5.73ns ± 2% ~ (p=0.579 n=10+10)
RuneCount/utf8.RuneCountInString/Japanese-16 22.0ns ± 6% 18.4ns ± 3% -16.41% (p=0.000 n=9+10)
RuneCount/utf8.RuneCountInString/MixedLength-16 15.0ns ± 1% 14.9ns ± 1% -1.01% (p=0.004 n=9+10)
RuneIterate/range/ASCII-16 2.69ns ± 1% 2.72ns ± 0% +0.94% (p=0.026 n=10+9)
RuneIterate/range/Japanese-16 24.5ns ± 2% 25.3ns ± 2% +3.23% (p=0.000 n=9+10)
RuneIterate/range/MixedLength-16 17.0ns ± 1% 17.1ns ± 1% +0.85% (p=0.000 n=10+10)
RuneIterate/range1/ASCII-16 2.70ns ± 1% 2.72ns ± 0% ~ (p=0.058 n=9+9)
RuneIterate/range1/Japanese-16 24.1ns ± 2% 25.2ns ± 3% +4.30% (p=0.000 n=10+10)
RuneIterate/range1/MixedLength-16 16.9ns ± 1% 17.7ns ± 0% +5.04% (p=0.000 n=10+8)
RuneIterate/range2/ASCII-16 2.84ns ± 8% 2.72ns ± 1% -4.28% (p=0.003 n=10+9)
RuneIterate/range2/Japanese-16 22.7ns ± 4% 25.2ns ± 3% +10.97% (p=0.000 n=10+10)
RuneIterate/range2/MixedLength-16 17.0ns ± 1% 17.2ns ± 0% +0.95% (p=0.000 n=10+10)
ArrayEqual-16 0.40ns ± 5% 0.35ns ± 2% -11.83% (p=0.000 n=10+10)
Func/Name-16 8.05ns ± 1% 8.09ns ± 1% +0.40% (p=0.025 n=8+10)
Func/Entry-16 1.73ns ± 1% 1.66ns ± 1% -3.93% (p=0.000 n=10+10)
Func/FileLine-16 27.5ns ± 2% 26.0ns ± 0% -5.50% (p=0.000 n=10+10)
[Geo mean] 16.7ns 15.7ns -6.08%
name old speed new speed delta
SetTypePtr-16 11.0GB/s ± 1% 11.0GB/s ± 3% ~ (p=0.684 n=10+10)
SetTypePtr8-16 15.5GB/s ± 0% 15.5GB/s ± 0% ~ (p=0.123 n=10+10)
SetTypePtr16-16 31.0GB/s ± 1% 31.1GB/s ± 0% ~ (p=0.123 n=10+10)
SetTypePtr32-16 62.1GB/s ± 0% 62.2GB/s ± 0% ~ (p=0.123 n=10+10)
SetTypePtr64-16 124GB/s ± 0% 124GB/s ± 0% ~ (p=0.684 n=10+10)
SetTypePtr126-16 146GB/s ± 0% 146GB/s ± 0% ~ (p=0.481 n=10+10)
SetTypePtr128-16 154GB/s ± 0% 154GB/s ± 0% ~ (p=0.243 n=9+10)
SetTypePtrSlice-16 151GB/s ± 1% 151GB/s ± 1% ~ (p=0.497 n=9+10)
SetTypeNode1-16 5.82GB/s ± 1% 5.82GB/s ± 0% ~ (p=0.353 n=10+10)
SetTypeNode1Slice-16 76.1GB/s ± 1% 77.0GB/s ± 1% +1.19% (p=0.000 n=10+10)
SetTypeNode8-16 19.4GB/s ± 0% 19.4GB/s ± 0% ~ (p=0.130 n=8+8)
SetTypeNode8Slice-16 113GB/s ± 0% 113GB/s ± 0% ~ (p=0.604 n=10+9)
SetTypeNode64-16 76.5GB/s ± 0% 76.5GB/s ± 0% ~ (p=0.190 n=10+10)
SetTypeNode64Slice-16 97.8GB/s ± 0% 97.7GB/s ± 0% ~ (p=0.549 n=9+10)
SetTypeNode64Dead-16 95.5GB/s ± 0% 95.7GB/s ± 0% ~ (p=0.118 n=10+6)
SetTypeNode64DeadSlice-16 112GB/s ± 0% 112GB/s ± 0% ~ (p=0.353 n=10+10)
SetTypeNode124-16 146GB/s ± 0% 146GB/s ± 0% ~ (p=0.853 n=10+10)
SetTypeNode124Slice-16 146GB/s ± 5% 149GB/s ± 0% ~ (p=0.315 n=10+10)
SetTypeNode126-16 154GB/s ± 0% 154GB/s ± 0% ~ (p=0.356 n=10+9)
SetTypeNode126Slice-16 150GB/s ± 0% 150GB/s ± 0% ~ (p=0.095 n=9+10)
SetTypeNode128-16 107GB/s ± 0% 107GB/s ± 0% +0.31% (p=0.003 n=9+10)
SetTypeNode128Slice-16 119GB/s ± 0% 120GB/s ± 0% ~ (p=0.156 n=10+9)
SetTypeNode130-16 108GB/s ± 0% 108GB/s ± 0% +0.33% (p=0.002 n=10+10)
SetTypeNode130Slice-16 119GB/s ± 0% 119GB/s ± 0% ~ (p=0.739 n=10+10)
SetTypeNode1024-16 160GB/s ± 0% 159GB/s ± 1% ~ (p=0.113 n=9+9)
SetTypeNode1024Slice-16 144GB/s ± 0% 144GB/s ± 0% ~ (p=0.063 n=10+10)
Hash5-16 2.59GB/s ± 1% 2.49GB/s ± 0% -3.90% (p=0.000 n=10+9)
Hash16-16 7.85GB/s ± 1% 7.23GB/s ± 1% -7.92% (p=0.000 n=10+10)
Hash64-16 24.0GB/s ± 0% 23.9GB/s ± 0% ~ (p=0.190 n=9+9)
Hash1024-16 62.4GB/s ± 0% 62.3GB/s ± 0% -0.16% (p=0.017 n=9+10)
Hash65536-16 74.0GB/s ± 0% 74.0GB/s ± 0% ~ (p=0.796 n=10+10)
Memmove/1-16 1.08GB/s ± 0% 1.08GB/s ± 0% -0.21% (p=0.035 n=10+10)
Memmove/2-16 2.16GB/s ± 0% 2.15GB/s ± 0% ~ (p=0.105 n=10+10)
Memmove/3-16 3.24GB/s ± 1% 3.22GB/s ± 1% -0.49% (p=0.004 n=10+10)
Memmove/4-16 3.89GB/s ± 0% 3.89GB/s ± 0% ~ (p=0.218 n=10+10)
Memmove/5-16 4.42GB/s ± 0% 4.42GB/s ± 0% ~ (p=0.075 n=10+10)
Memmove/6-16 5.31GB/s ± 0% 5.29GB/s ± 1% ~ (p=0.218 n=10+10)
Memmove/7-16 6.19GB/s ± 0% 6.18GB/s ± 0% -0.15% (p=0.035 n=10+9)
Memmove/8-16 7.07GB/s ± 0% 7.07GB/s ± 0% ~ (p=0.684 n=10+10)
Memmove/9-16 7.22GB/s ± 0% 6.68GB/s ± 0% -7.37% (p=0.000 n=10+10)
Memmove/10-16 8.02GB/s ± 0% 7.43GB/s ± 0% -7.38% (p=0.000 n=9+9)
Memmove/11-16 8.83GB/s ± 0% 8.13GB/s ± 0% -7.87% (p=0.000 n=10+9)
Memmove/12-16 9.62GB/s ± 0% 8.89GB/s ± 1% -7.61% (p=0.000 n=10+10)
Memmove/13-16 10.4GB/s ± 0% 9.7GB/s ± 0% -7.20% (p=0.000 n=10+10)
Memmove/14-16 11.2GB/s ± 0% 10.4GB/s ± 1% -7.64% (p=0.000 n=10+9)
Memmove/15-16 12.0GB/s ± 0% 11.1GB/s ± 0% -7.46% (p=0.000 n=10+9)
Memmove/16-16 12.8GB/s ± 0% 11.8GB/s ± 1% -7.67% (p=0.000 n=10+10)
Memmove/32-16 23.8GB/s ± 0% 23.5GB/s ± 1% -1.20% (p=0.000 n=10+10)
Memmove/64-16 44.2GB/s ± 0% 39.1GB/s ± 0% -11.56% (p=0.000 n=10+9)
Memmove/128-16 68.7GB/s ± 0% 63.2GB/s ± 0% -7.95% (p=0.000 n=10+10)
Memmove/256-16 104GB/s ± 0% 103GB/s ± 0% -1.13% (p=0.000 n=10+10)
Memmove/512-16 129GB/s ± 1% 129GB/s ± 0% ~ (p=0.165 n=10+10)
Memmove/1024-16 174GB/s ± 1% 174GB/s ± 1% ~ (p=0.258 n=9+9)
Memmove/2048-16 213GB/s ± 1% 213GB/s ± 2% ~ (p=0.963 n=8+9)
Memmove/4096-16 250GB/s ± 1% 240GB/s ± 4% -3.83% (p=0.006 n=9+9)
MemmoveOverlap/32-16 19.8GB/s ± 1% 19.1GB/s ± 1% -3.40% (p=0.000 n=10+10)
MemmoveOverlap/64-16 39.0GB/s ± 0% 38.8GB/s ± 0% -0.28% (p=0.001 n=9+9)
MemmoveOverlap/128-16 62.2GB/s ± 0% 62.1GB/s ± 0% ~ (p=0.063 n=10+10)
MemmoveOverlap/256-16 96.0GB/s ± 0% 95.8GB/s ± 0% -0.26% (p=0.009 n=10+10)
MemmoveOverlap/512-16 83.6GB/s ±16% 89.2GB/s ± 0% ~ (p=0.696 n=10+8)
MemmoveOverlap/1024-16 141GB/s ± 0% 140GB/s ± 0% -0.28% (p=0.006 n=8+10)
MemmoveOverlap/2048-16 172GB/s ± 0% 171GB/s ± 1% -0.38% (p=0.008 n=9+9)
MemmoveOverlap/4096-16 176GB/s ± 1% 177GB/s ± 1% +0.84% (p=0.001 n=8+10)
MemmoveUnalignedDst/1-16 806MB/s ± 0% 802MB/s ± 1% -0.52% (p=0.023 n=10+10)
MemmoveUnalignedDst/2-16 1.62GB/s ± 0% 1.62GB/s ± 0% -0.11% (p=0.041 n=10+10)
MemmoveUnalignedDst/3-16 2.43GB/s ± 0% 2.43GB/s ± 0% -0.14% (p=0.006 n=9+9)
MemmoveUnalignedDst/4-16 3.24GB/s ± 0% 3.23GB/s ± 1% -0.36% (p=0.007 n=10+10)
MemmoveUnalignedDst/5-16 3.71GB/s ± 0% 3.71GB/s ± 0% ~ (p=0.063 n=10+10)
MemmoveUnalignedDst/6-16 4.48GB/s ± 0% 4.47GB/s ± 0% ~ (p=0.912 n=10+10)
MemmoveUnalignedDst/7-16 5.22GB/s ± 0% 5.22GB/s ± 0% ~ (p=1.000 n=10+10)
MemmoveUnalignedDst/8-16 5.95GB/s ± 0% 5.93GB/s ± 1% -0.40% (p=0.023 n=10+10)
MemmoveUnalignedDst/9-16 6.24GB/s ± 0% 6.24GB/s ± 0% ~ (p=0.912 n=10+10)
MemmoveUnalignedDst/10-16 6.94GB/s ± 0% 6.94GB/s ± 0% ~ (p=0.353 n=10+10)
MemmoveUnalignedDst/11-16 7.64GB/s ± 0% 7.63GB/s ± 0% ~ (p=0.393 n=10+10)
MemmoveUnalignedDst/12-16 8.33GB/s ± 0% 8.33GB/s ± 0% ~ (p=0.971 n=10+10)
MemmoveUnalignedDst/13-16 9.02GB/s ± 0% 9.01GB/s ± 0% ~ (p=0.436 n=10+10)
MemmoveUnalignedDst/14-16 9.71GB/s ± 0% 9.71GB/s ± 0% ~ (p=0.280 n=10+10)
MemmoveUnalignedDst/15-16 10.4GB/s ± 0% 10.4GB/s ± 1% ~ (p=0.853 n=10+10)
MemmoveUnalignedDst/16-16 11.1GB/s ± 0% 11.1GB/s ± 0% ~ (p=0.089 n=10+10)
MemmoveUnalignedDst/32-16 19.7GB/s ± 1% 19.6GB/s ± 0% ~ (p=0.075 n=10+10)
MemmoveUnalignedDst/64-16 38.9GB/s ± 0% 38.8GB/s ± 0% ~ (p=0.218 n=10+10)
MemmoveUnalignedDst/128-16 62.1GB/s ± 0% 62.1GB/s ± 0% ~ (p=0.549 n=10+9)
MemmoveUnalignedDst/256-16 69.4GB/s ± 0% 69.3GB/s ± 0% ~ (p=0.105 n=10+10)
MemmoveUnalignedDst/512-16 124GB/s ± 1% 124GB/s ± 0% ~ (p=0.762 n=10+8)
MemmoveUnalignedDst/1024-16 136GB/s ± 0% 136GB/s ± 1% ~ (p=0.666 n=9+9)
MemmoveUnalignedDst/2048-16 159GB/s ± 0% 159GB/s ± 0% ~ (p=0.574 n=8+8)
MemmoveUnalignedDst/4096-16 161GB/s ± 0% 161GB/s ± 0% ~ (p=1.000 n=9+9)
MemmoveUnalignedDstOverlap/32-16 7.84GB/s ± 0% 7.83GB/s ± 0% ~ (p=0.353 n=10+10)
MemmoveUnalignedDstOverlap/64-16 14.0GB/s ± 0% 14.0GB/s ± 0% ~ (p=0.661 n=10+9)
MemmoveUnalignedDstOverlap/128-16 27.4GB/s ± 0% 27.4GB/s ± 0% ~ (p=0.353 n=10+10)
MemmoveUnalignedDstOverlap/256-16 50.4GB/s ± 0% 50.4GB/s ± 0% ~ (p=0.156 n=10+9)
MemmoveUnalignedDstOverlap/512-16 60.7GB/s ± 4% 62.5GB/s ± 0% +3.07% (p=0.022 n=10+9)
MemmoveUnalignedDstOverlap/1024-16 107GB/s ± 0% 107GB/s ± 0% ~ (p=0.234 n=8+8)
MemmoveUnalignedDstOverlap/2048-16 146GB/s ± 0% 146GB/s ± 1% ~ (p=0.182 n=10+9)
MemmoveUnalignedDstOverlap/4096-16 155GB/s ± 0% 155GB/s ± 0% ~ (p=0.400 n=10+9)
MemmoveUnalignedSrc/1-16 882MB/s ± 0% 884MB/s ± 1% +0.24% (p=0.033 n=10+9)
MemmoveUnalignedSrc/2-16 1.76GB/s ± 1% 1.77GB/s ± 0% +0.27% (p=0.028 n=10+9)
MemmoveUnalignedSrc/3-16 2.43GB/s ± 0% 2.43GB/s ± 0% +0.26% (p=0.027 n=9+10)
MemmoveUnalignedSrc/4-16 3.24GB/s ± 0% 3.24GB/s ± 1% ~ (p=0.079 n=9+10)
MemmoveUnalignedSrc/5-16 3.73GB/s ± 0% 3.73GB/s ± 1% ~ (p=0.829 n=8+10)
MemmoveUnalignedSrc/6-16 4.47GB/s ± 0% 4.49GB/s ± 0% +0.39% (p=0.000 n=10+10)
MemmoveUnalignedSrc/7-16 5.22GB/s ± 0% 5.23GB/s ± 0% ~ (p=0.280 n=10+10)
MemmoveUnalignedSrc/8-16 5.95GB/s ± 0% 5.98GB/s ± 0% +0.39% (p=0.001 n=10+9)
MemmoveUnalignedSrc/9-16 6.24GB/s ± 0% 6.25GB/s ± 0% ~ (p=0.549 n=10+9)
MemmoveUnalignedSrc/10-16 6.93GB/s ± 0% 6.94GB/s ± 0% ~ (p=0.604 n=10+9)
MemmoveUnalignedSrc/11-16 7.63GB/s ± 0% 7.63GB/s ± 1% ~ (p=0.353 n=10+10)
MemmoveUnalignedSrc/12-16 8.32GB/s ± 0% 8.32GB/s ± 0% ~ (p=0.218 n=10+10)
MemmoveUnalignedSrc/13-16 9.02GB/s ± 0% 9.00GB/s ± 1% ~ (p=0.684 n=10+10)
MemmoveUnalignedSrc/14-16 9.71GB/s ± 0% 9.71GB/s ± 0% ~ (p=0.739 n=10+10)
MemmoveUnalignedSrc/15-16 10.4GB/s ± 0% 10.4GB/s ± 0% ~ (p=0.353 n=10+10)
MemmoveUnalignedSrc/16-16 11.1GB/s ± 1% 11.1GB/s ± 0% ~ (p=0.579 n=10+10)
MemmoveUnalignedSrc/32-16 20.0GB/s ± 1% 20.0GB/s ± 0% ~ (p=0.631 n=10+10)
MemmoveUnalignedSrc/64-16 38.8GB/s ± 0% 38.8GB/s ± 0% ~ (p=0.579 n=10+10)
MemmoveUnalignedSrc/128-16 61.2GB/s ± 0% 61.2GB/s ± 0% ~ (p=0.780 n=10+9)
MemmoveUnalignedSrc/256-16 94.8GB/s ± 0% 92.2GB/s ± 1% -2.73% (p=0.000 n=10+10)
MemmoveUnalignedSrc/512-16 119GB/s ± 0% 119GB/s ± 0% +0.26% (p=0.027 n=8+9)
MemmoveUnalignedSrc/1024-16 141GB/s ± 0% 142GB/s ± 1% +1.07% (p=0.000 n=8+10)
MemmoveUnalignedSrc/2048-16 157GB/s ± 0% 157GB/s ± 0% ~ (p=0.167 n=9+8)
MemmoveUnalignedSrc/4096-16 161GB/s ± 0% 162GB/s ± 1% ~ (p=0.063 n=10+10)
MemmoveUnalignedSrcOverlap/32-16 7.93GB/s ± 0% 7.88GB/s ± 0% -0.63% (p=0.000 n=9+10)
MemmoveUnalignedSrcOverlap/64-16 15.5GB/s ± 0% 15.5GB/s ± 0% ~ (p=0.529 n=10+10)
MemmoveUnalignedSrcOverlap/128-16 28.3GB/s ± 0% 28.3GB/s ± 0% ~ (p=0.218 n=10+10)
MemmoveUnalignedSrcOverlap/256-16 41.5GB/s ± 0% 41.6GB/s ± 0% +0.35% (p=0.000 n=10+9)
MemmoveUnalignedSrcOverlap/512-16 68.9GB/s ± 0% 68.8GB/s ± 0% ~ (p=0.541 n=9+8)
MemmoveUnalignedSrcOverlap/1024-16 115GB/s ± 0% 115GB/s ± 0% ~ (p=0.382 n=8+8)
MemmoveUnalignedSrcOverlap/2048-16 155GB/s ± 0% 144GB/s ±18% ~ (p=0.101 n=8+10)
MemmoveUnalignedSrcOverlap/4096-16 160GB/s ± 0% 160GB/s ± 1% ~ (p=0.605 n=9+9)
Memclr/5-16 5.81GB/s ± 1% 5.80GB/s ± 2% ~ (p=0.546 n=9+9)
Memclr/16-16 15.5GB/s ± 0% 15.4GB/s ± 0% -0.32% (p=0.008 n=9+10)
Memclr/64-16 51.8GB/s ± 0% 50.7GB/s ± 0% -2.22% (p=0.000 n=10+10)
Memclr/256-16 113GB/s ± 0% 113GB/s ± 0% ~ (p=0.143 n=10+10)
Memclr/4096-16 239GB/s ± 1% 237GB/s ± 0% -0.87% (p=0.000 n=10+10)
Memclr/65536-16 79.8GB/s ± 0% 79.7GB/s ± 0% ~ (p=0.529 n=10+10)
Memclr/1M-16 74.6GB/s ± 1% 74.7GB/s ± 1% ~ (p=0.529 n=10+10)
Memclr/4M-16 48.7GB/s ± 1% 48.8GB/s ± 0% ~ (p=0.123 n=10+10)
Memclr/8M-16 48.2GB/s ± 2% 48.6GB/s ± 0% ~ (p=0.408 n=10+8)
Memclr/16M-16 43.6GB/s ± 4% 43.3GB/s ± 0% ~ (p=0.173 n=10+8)
Memclr/64M-16 30.7GB/s ± 0% 30.7GB/s ± 0% ~ (p=0.113 n=10+9)
GoMemclr/5-16 6.07GB/s ± 0% 6.08GB/s ± 0% ~ (p=0.367 n=9+10)
GoMemclr/16-16 15.6GB/s ± 0% 15.6GB/s ± 0% -0.22% (p=0.004 n=9+9)
GoMemclr/64-16 56.1GB/s ± 0% 56.1GB/s ± 0% ~ (p=0.968 n=10+9)
GoMemclr/256-16 125GB/s ± 0% 124GB/s ± 0% ~ (p=0.912 n=10+10)
MemclrRange/1K_2K-16 210GB/s ± 0% 224GB/s ± 1% +6.81% (p=0.000 n=10+10)
MemclrRange/2K_8K-16 228GB/s ± 0% 228GB/s ± 0% ~ (p=0.684 n=10+10)
MemclrRange/4K_16K-16 279GB/s ± 0% 279GB/s ± 0% ~ (p=0.780 n=9+10)
MemclrRange/160K_228K-16 80.3GB/s ± 0% 80.2GB/s ± 0% ~ (p=0.165 n=10+10)
Copy/1Byte-16 808MB/s ± 1% 810MB/s ± 0% +0.28% (p=0.000 n=10+8)
Copy/1String-16 810MB/s ± 0% 811MB/s ± 0% ~ (p=0.105 n=10+10)
Copy/2Byte-16 1.62GB/s ± 0% 1.62GB/s ± 0% ~ (p=0.182 n=10+9)
Copy/2String-16 1.62GB/s ± 1% 1.62GB/s ± 1% ~ (p=1.000 n=10+10)
Copy/4Byte-16 3.22GB/s ± 0% 3.24GB/s ± 0% +0.46% (p=0.000 n=10+10)
Copy/4String-16 3.24GB/s ± 0% 3.24GB/s ± 0% ~ (p=0.075 n=10+10)
Copy/8Byte-16 5.86GB/s ± 0% 5.96GB/s ± 0% +1.82% (p=0.000 n=9+9)
Copy/8String-16 5.95GB/s ± 0% 5.99GB/s ± 0% +0.59% (p=0.000 n=9+10)
Copy/12Byte-16 8.32GB/s ± 0% 8.32GB/s ± 0% ~ (p=0.190 n=9+9)
Copy/12String-16 8.31GB/s ± 0% 8.29GB/s ± 0% ~ (p=0.068 n=9+10)
Copy/16Byte-16 11.1GB/s ± 0% 11.1GB/s ± 0% +0.18% (p=0.003 n=10+9)
Copy/16String-16 11.1GB/s ± 0% 11.1GB/s ± 0% -0.31% (p=0.009 n=10+10)
Copy/32Byte-16 19.6GB/s ± 1% 19.8GB/s ± 1% +0.72% (p=0.002 n=10+10)
Copy/32String-16 20.0GB/s ± 0% 19.5GB/s ± 0% -2.19% (p=0.000 n=10+10)
Copy/128Byte-16 62.2GB/s ± 0% 62.1GB/s ± 0% ~ (p=0.661 n=9+10)
Copy/128String-16 61.9GB/s ± 0% 61.7GB/s ± 0% -0.35% (p=0.005 n=10+10)
Copy/1024Byte-16 169GB/s ± 2% 171GB/s ± 1% +1.21% (p=0.000 n=9+10)
Copy/1024String-16 169GB/s ± 0% 172GB/s ± 1% +1.57% (p=0.000 n=10+9)
CompareStringBigUnaligned-16 43.8GB/s ± 1% 43.7GB/s ± 1% ~ (p=0.370 n=9+8)
CompareStringBig-16 47.6GB/s ± 3% 47.3GB/s ± 3% ~ (p=0.243 n=9+10)
[Geo mean] 25.3GB/s 28.0GB/s +10.66%
name old p50-ns new p50-ns delta
ReadMemStatsLatency-16 98.4k ±37% 112.7k ±64% ~ (p=0.436 n=10+10)
ReadMetricsLatency-16 1.69k ± 3% 1.70k ± 2% ~ (p=0.646 n=9+10)
GoroutineProfile/small-nil/idle-16 3.75k ± 4% 3.72k ± 1% ~ (p=0.447 n=10+9)
GoroutineProfile/small-nil/loaded-16 4.33k ± 3% 4.31k ± 5% ~ (p=0.931 n=9+9)
GoroutineProfile/small/idle-16 102k ± 3% 101k ± 4% ~ (p=0.113 n=9+9)
GoroutineProfile/small/loaded-16 214k ± 3% 215k ± 6% ~ (p=0.842 n=9+10)
GoroutineProfile/large-nil/idle-16 3.70k ± 2% 3.65k ± 2% ~ (p=0.075 n=10+10)
GoroutineProfile/large-nil/loaded-16 4.36k ± 3% 4.31k ±10% ~ (p=0.631 n=10+10)
GoroutineProfile/large/idle-16 2.56M ± 1% 2.51M ± 1% -2.28% (p=0.000 n=10+10)
GoroutineProfile/large/loaded-16 6.77M ± 4% 6.85M ±19% ~ (p=0.536 n=7+10)
GoroutineProfile/sparse-nil/idle-16 3.66k ± 1% 3.64k ± 2% ~ (p=0.136 n=9+9)
GoroutineProfile/sparse-nil/loaded-16 4.25k ± 5% 4.15k ± 4% ~ (p=0.190 n=10+10)
GoroutineProfile/sparse/idle-16 102k ± 4% 101k ± 3% ~ (p=0.447 n=10+9)
GoroutineProfile/sparse/loaded-16 216k ± 4% 218k ± 4% ~ (p=0.549 n=9+10)
[Geo mean] 35.8k 35.9k +0.35%
name old p90-ns new p90-ns delta
ReadMemStatsLatency-16 983k ±310% 200k ±34% -79.62% (p=0.034 n=10+8)
ReadMetricsLatency-16 4.01k ±35% 3.75k ±17% ~ (p=0.315 n=10+10)
GoroutineProfile/small-nil/idle-16 4.21k ± 4% 4.27k ± 8% ~ (p=0.968 n=9+10)
GoroutineProfile/small-nil/loaded-16 5.58k ± 8% 5.35k ±12% ~ (p=0.190 n=10+10)
GoroutineProfile/small/idle-16 108k ± 6% 107k ± 7% ~ (p=0.497 n=9+10)
GoroutineProfile/small/loaded-16 450k ± 5% 432k ± 3% -3.92% (p=0.002 n=9+9)
GoroutineProfile/large-nil/idle-16 4.13k ± 7% 4.04k ± 2% ~ (p=0.181 n=10+8)
GoroutineProfile/large-nil/loaded-16 5.76k ± 4% 5.67k ± 5% ~ (p=0.190 n=10+10)
GoroutineProfile/large/idle-16 2.63M ± 2% 2.58M ± 1% -1.97% (p=0.000 n=10+10)
GoroutineProfile/large/loaded-16 16.9M ± 4% 17.0M ± 6% ~ (p=0.661 n=9+10)
GoroutineProfile/sparse-nil/idle-16 4.21k ±10% 4.07k ± 7% ~ (p=0.128 n=10+10)
GoroutineProfile/sparse-nil/loaded-16 5.55k ± 8% 5.38k ± 6% ~ (p=0.089 n=10+10)
GoroutineProfile/sparse/idle-16 106k ± 4% 106k ± 3% ~ (p=0.661 n=10+9)
GoroutineProfile/sparse/loaded-16 454k ± 6% 441k ± 6% -2.86% (p=0.043 n=10+10)
[Geo mean] 58.4k 51.0k -12.61%
name old p99-ns new p99-ns delta
ReadMemStatsLatency-16 983k ±310% 200k ±34% -79.62% (p=0.034 n=10+8)
ReadMetricsLatency-16 26.6k ±22% 26.6k ±17% ~ (p=0.971 n=10+10)
GoroutineProfile/small-nil/idle-16 5.27k ±12% 5.19k ±12% ~ (p=0.579 n=10+10)
GoroutineProfile/small-nil/loaded-16 7.28k ± 3% 7.04k ± 9% ~ (p=0.113 n=9+10)
GoroutineProfile/small/idle-16 114k ± 6% 113k ± 6% ~ (p=0.604 n=9+10)
GoroutineProfile/small/loaded-16 4.84M ±70% 6.06M ±131% ~ (p=0.842 n=9+10)
GoroutineProfile/large-nil/idle-16 5.38k ±19% 5.26k ±13% ~ (p=0.912 n=10+10)
GoroutineProfile/large-nil/loaded-16 7.38k ± 3% 7.23k ± 4% ~ (p=0.143 n=10+10)
GoroutineProfile/large/idle-16 2.79M ± 5% 2.72M ± 3% ~ (p=0.089 n=10+10)
GoroutineProfile/large/loaded-16 24.0M ±24% 24.9M ±29% ~ (p=0.684 n=10+10)
GoroutineProfile/sparse-nil/idle-16 5.32k ±17% 5.49k ±17% ~ (p=0.684 n=10+10)
GoroutineProfile/sparse-nil/loaded-16 7.25k ± 4% 6.97k ± 5% -3.90% (p=0.005 n=9+10)
GoroutineProfile/sparse/idle-16 113k ± 5% 112k ± 6% ~ (p=0.631 n=10+10)
GoroutineProfile/sparse/loaded-16 4.00M ±66% 4.26M ±65% ~ (p=0.489 n=9+9)
[Geo mean] 107k 97k -9.55%
name old alloc/op new alloc/op delta
NewEmptyMap-16 0.00B 0.00B ~ (all equal)
NewSmallMap-16 0.00B 0.00B ~ (all equal)
MapPopulate/1-16 0.00B 0.00B ~ (all equal)
MapPopulate/10-16 179B ± 0% 179B ± 0% ~ (all equal)
MapPopulate/100-16 3.35kB ± 0% 3.35kB ± 0% ~ (p=0.294 n=10+8)
MapPopulate/1000-16 53.3kB ± 0% 53.3kB ± 0% ~ (p=1.000 n=8+10)
MapPopulate/10000-16 428kB ± 0% 428kB ± 0% ~ (p=0.469 n=10+10)
MapPopulate/100000-16 3.62MB ± 0% 3.62MB ± 0% ~ (p=0.888 n=9+10)
MapStringConversion/32/simple-16 0.00B 0.00B ~ (all equal)
MapStringConversion/32/struct-16 0.00B 0.00B ~ (all equal)
MapStringConversion/32/array-16 0.00B 0.00B ~ (all equal)
MapStringConversion/64/simple-16 0.00B 0.00B ~ (all equal)
MapStringConversion/64/struct-16 0.00B 0.00B ~ (all equal)
MapStringConversion/64/array-16 0.00B 0.00B ~ (all equal)
NewEmptyMapHintLessThan8-16 0.00B 0.00B ~ (all equal)
NewEmptyMapHintGreaterThan8-16 1.15kB ± 0% 1.15kB ± 0% ~ (all equal)
MapAppendAssign/Int32/256-16 41.7B ±15% 44.3B ±12% ~ (p=0.106 n=10+10)
MapAppendAssign/Int32/65536-16 22.6B ± 6% 23.5B ± 6% +4.19% (p=0.025 n=9+10)
MapAppendAssign/Int64/256-16 43.5B ±10% 42.9B ± 7% ~ (p=0.757 n=10+10)
MapAppendAssign/Int64/65536-16 24.7B ± 7% 21.8B ± 6% -11.74% (p=0.000 n=10+10)
MapAppendAssign/Str/256-16 87.6B ±10% 89.2B ± 9% ~ (p=0.379 n=10+10)
MapAppendAssign/Str/65536-16 45.1B ±14% 47.2B ± 8% ~ (p=0.150 n=10+9)
CreateGoroutinesCapture-16 144B ± 0% 144B ± 0% ~ (all equal)
[Geo mean] 769B 770B +0.21%
name old allocs/op new allocs/op delta
NewEmptyMap-16 0.00 0.00 ~ (all equal)
NewSmallMap-16 0.00 0.00 ~ (all equal)
MapPopulate/1-16 0.00 0.00 ~ (all equal)
MapPopulate/10-16 1.00 ± 0% 1.00 ± 0% ~ (all equal)
MapPopulate/100-16 17.0 ± 0% 17.0 ± 0% ~ (all equal)
MapPopulate/1000-16 73.0 ± 0% 73.0 ± 0% ~ (all equal)
MapPopulate/10000-16 320 ± 0% 320 ± 0% ~ (p=1.000 n=10+10)
MapPopulate/100000-16 4.00k ± 0% 4.00k ± 0% ~ (p=0.753 n=10+10)
MapStringConversion/32/simple-16 0.00 0.00 ~ (all equal)
MapStringConversion/32/struct-16 0.00 0.00 ~ (all equal)
MapStringConversion/32/array-16 0.00 0.00 ~ (all equal)
MapStringConversion/64/simple-16 0.00 0.00 ~ (all equal)
MapStringConversion/64/struct-16 0.00 0.00 ~ (all equal)
MapStringConversion/64/array-16 0.00 0.00 ~ (all equal)
NewEmptyMapHintLessThan8-16 0.00 0.00 ~ (all equal)
NewEmptyMapHintGreaterThan8-16 1.00 ± 0% 1.00 ± 0% ~ (all equal)
MapAppendAssign/Int32/256-16 0.00 0.00 ~ (all equal)
MapAppendAssign/Int32/65536-16 0.00 0.00 ~ (all equal)
MapAppendAssign/Int64/256-16 0.00 0.00 ~ (all equal)
MapAppendAssign/Int64/65536-16 0.00 0.00 ~ (all equal)
MapAppendAssign/Str/256-16 0.00 0.00 ~ (all equal)
MapAppendAssign/Str/65536-16 0.00 0.00 ~ (all equal)
CreateGoroutinesCapture-16 5.00 ± 0% 5.00 ± 0% ~ (all equal)
[Geo mean] 26.0 26.0 +0.00%
Change-Id: I5fb03e93df8b380e04795afbdcd1c94aeeecacc6
Reviewed-on: https://go-review.googlesource.com/c/go/+/454255
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Jakub Ciolek <jakub@ciolek.dev>
TryBot-Result: Gopher Robot <gobot@golang.org>
|
||
Paul E. Murphy
|
1540531746 |
test/codegen: merge identical ppc64 and ppc64le tests
Manually consolidate the remaining ppc64/ppc64le test which are not so trivial to automatically merge. The remaining ppc64le tests are limited to cases where load/stores are merged (this only happens on ppc64le) and the race detector (only supported on ppc64le). Change-Id: I1f9c0f3d3ddbb7fbbd8c81fbbd6537394fba63ce Reviewed-on: https://go-review.googlesource.com/c/go/+/463217 Reviewed-by: Dmitri Shuralyov <dmitshur@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com> Run-TryBot: Paul Murphy <murp@ibm.com> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com> |
||
|
Paul E. Murphy
|
0301c6c351 |
test/codegen: combine trivial PPC64 tests into ppc64x
Use a small python script to consolidate duplicate
ppc64/ppc64le tests into a single ppc64x codegen test.
This makes small assumption that anytime two tests with
for different arch/variant combos exists, those tests
can be combined into a single ppc64x test.
E.x:
// ppc64le: foo
// ppc64le/power9: foo
into
// ppc64x: foo
or
// ppc64: foo
// ppc64le: foo
into
// ppc64x: foo
import glob
import re
files = glob.glob("codegen/*.go")
for file in files:
with open(file) as f:
text = [l for l in f]
i = 0
while i < len(text):
first = re.match("\s*// ?ppc64(le)?(/power[89])?:(.*)", text[i])
if first:
j = i+1
while j < len(text):
second = re.match("\s*// ?ppc64(le)?(/power[89])?:(.*)", text[j])
if not second:
break
if (not first.group(2) or first.group(2) == second.group(2)) and first.group(3) == second.group(3):
text[i] = re.sub(" ?ppc64(le|x)?"," ppc64x",text[i])
text=text[:j] + (text[j+1:])
else:
j += 1
i+=1
with open(file, 'w') as f:
f.write("".join(text))
Change-Id: Ic6b009b54eacaadc5a23db9c5a3bf7331b595821
Reviewed-on: https://go-review.googlesource.com/c/go/+/463220
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
|
||
|
Paul E. Murphy
|
a37672bb7b |
test/codegen: accept ppc64x as alias for ppc64le and ppc64 arches
This helps simplify the noise when adding ppc codegen tests. ppc64x is used in other places to indicate something which runs on either endian. This helps cleanup existing codegen tests which are mostly identical between endian variants. condmove tests are converted as an example. Change-Id: I2b2d98a9a1859015f62db38d62d9d5d7593435b4 Reviewed-on: https://go-review.googlesource.com/c/go/+/462895 Reviewed-by: Cherry Mui <cherryyz@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: Paul Murphy <murp@ibm.com> |
||
David Chase
|
e22bd2348c |
internal/abi,runtime: refactor map constants into one place
Previously TryBot-tested with bucket bits = 4. Also tested locally with bucket bits = 5. This makes it much easier to change the size of map buckets, and hopefully provides pointers to all the code that in some way depends on details of map layout. Change-Id: I9f6669d1eadd02f182d0bc3f959dc5f385fa1683 Reviewed-on: https://go-review.googlesource.com/c/go/+/462115 TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: David Chase <drchase@google.com> Reviewed-by: Austin Clements <austin@google.com> |
||
Jorropo
|
5c67ebbb31 |
cmd/compile: AMD64v3 remove unnecessary TEST comparision in isPowerOfTwo
With GOAMD64=V3 the canonical isPowerOfTwo function:
func isPowerOfTwo(x uintptr) bool {
return x&(x-1) == 0
}
Used to compile to:
temp := BLSR(x) // x&(x-1)
flags = TEST(temp, temp)
return flags.zf
However the blsr instruction already set ZF according to the result.
So we can remove the TEST instruction if we are just checking ZF.
Such as in multiple pieces of code around memory allocations.
This make the code smaller and faster.
Change-Id: Ia12d5a73aa3cb49188c0b647b1eff7b56c5a7b58
Reviewed-on: https://go-review.googlesource.com/c/go/+/448255
Run-TryBot: Jakub Ciolek <jakub@ciolek.dev>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
|
||
|
Jorropo
|
fc814056aa |
cmd/compile: rewrite empty makeslice to zerobase pointer
make\(\[\][a-zA-Z0-9]+, 0\) is seen 52 times in the go source. And at least 391 times on internet: https://grep.app/search?q=make%5C%28%5C%5B%5C%5D%5Ba-zA-Z0-9%5D%2B%2C%200%5C%29®exp=true This used to compile to calling runtime.makeslice. However we can copy what we do for []T{}, just use a zerobase pointer. On my machine this is 10x faster (from 3ns to 0.3ns). Note that an empty loop also runs in 0.3ns, so this really is free when you count superscallar execution. Change-Id: I1cfe7e69f5a7a4dabbc71912ce6a4f8a2d4a7f3c Reviewed-on: https://go-review.googlesource.com/c/go/+/454036 Reviewed-by: Cherry Mui <cherryyz@google.com> Reviewed-by: Keith Randall <khr@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Jakub Ciolek <jakub@ciolek.dev> |
||
|
Keith Randall
|
f959fb3872 |
cmd/compile: add anchored version of SP
The SPanchored opcode is identical to SP, except that it takes a memory argument so that it (and more importantly, anything that uses it) must be scheduled at or after that memory argument. This opcode ensures that a LEAQ of a variable gets scheduled after the corresponding VARDEF for that variable. This may lead to less CSE of LEAQ operations. The effect is very small. The go binary is only 80 bytes bigger after this CL. Usually LEAQs get folded into load/store operations, so the effect is only for pointerful types, large enough to need a duffzero, and have their address passed somewhere. Even then, usually the CSEd LEAQs will be un-CSEd because the two uses are on different sides of a function call and the LEAQ ends up being rematerialized at the second use anyway. Change-Id: Ib893562cd05369b91dd563b48fb83f5250950293 Reviewed-on: https://go-review.googlesource.com/c/go/+/452916 TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com> Reviewed-by: Martin Möhrmann <martin@golang.org> Reviewed-by: Keith Randall <khr@google.com> |
||
|
Keith Randall
|
1eb0465fa5 |
cmd/compile: turn off jump tables when spectre retpolines are on
Fixes #57097 Change-Id: I6ab659abbca1ae0ac8710674d39aec116fab0baa Reviewed-on: https://go-review.googlesource.com/c/go/+/455336 Reviewed-by: Keith Randall <khr@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: Keith Randall <khr@golang.org> |
||
|
Paul E. Murphy
|
dc6b7c86df |
cmd/compile: merge zero constant ISEL in PPC64 lateLower pass
Add a new SSA opcode ISELZ, similar to ISELB to represent a select of value or 0. Then, merge candidate ISEL opcodes inside the late lower pass. This avoids complicating rules within the the lower pass. Change-Id: I3b14c94b763863aadc834b0e910a85870c131313 Reviewed-on: https://go-review.googlesource.com/c/go/+/442596 TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com> Reviewed-by: Michael Knyszek <mknyszek@google.com> Run-TryBot: Paul Murphy <murp@ibm.com> Reviewed-by: Joedian Reid <joedian@golang.org> |
||
|
Wayne Zuo
|
268f4629df |
cmd/compile: enable brachelim pass on loong64
Change-Id: I4fd1c307901c265ab9865bf8a74460ddc15e5d14 Reviewed-on: https://go-review.googlesource.com/c/go/+/416735 Reviewed-by: Keith Randall <khr@google.com> Reviewed-by: xiaodong liu <teaofmoli@gmail.com> Auto-Submit: Wayne Zuo <wdvxdr@golangcn.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> |
||
|
Paul E. Murphy
|
390abbbbf1 |
codegen: check for PPC64 ISEL in condmove tests
ISEL is roughly equivalent to CMOV on PPC64. Verify ISEL generation in all reasonable cases. Note "ISEL test x y z" is the same as "ISEL !test y x z". test is always one of LT (0), GT (1), EQ (2), SO (3). Sometimes x and y are swapped if GE/LE/NE is desired. Change-Id: Ie1bf029224064e004d855099731fe5e8d05aa990 Reviewed-on: https://go-review.googlesource.com/c/go/+/445215 Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com> Reviewed-by: Bryan Mills <bcmills@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: Paul Murphy <murp@ibm.com> Reviewed-by: Than McIntosh <thanm@google.com> |
||
|
Paul E. Murphy
|
d031e9e07a |
cmd/compile/internal/ssa: re-adjust CarryChainTail scheduling priority
This needs to be as low as possible while not breaking priority assumptions of other scores to correctly schedule carry chains. Prior to the arm64 changes, it was set below ReadTuple. At the time, this prevented the MulHiLo implementation on PPC64 from occluding the scheduling of a full carry chain. Memory scores can also prevent better scheduling, as can be observed with crypto/internal/edwards25519/field.feMulGeneric. Fixes #56497 Change-Id: Ia4b54e6dffcce584faf46b1b8d7cea18a3913887 Reviewed-on: https://go-review.googlesource.com/c/go/+/447435 Reviewed-by: Cherry Mui <cherryyz@google.com> Reviewed-by: Keith Randall <khr@google.com> Run-TryBot: Paul Murphy <murp@ibm.com> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Bryan Mills <bcmills@google.com> Reviewed-by: Keith Randall <khr@golang.org> |
||
|
Keith Randall
|
9ce27feaeb |
cmd/compile: add rule for post-decomposed growslice optimization
The recently added rule only works before decomposing slices. Add a rule that works after decomposing slices. The reason we need the latter is because although the length may be a constant, it can be hidden inside a slice that is not constant (its pointer or capacity might be changing). By applying this optimization after decomposing slices, we can find more cases where it applies. Fixes #56440 Change-Id: I0094e59eee3065ab4d210defdda8227a6e897420 Reviewed-on: https://go-review.googlesource.com/c/go/+/446277 Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Keith Randall <khr@google.com> |
||
|
Keith Randall
|
0156b797e6 |
cmd/compile: recognize when the result of append has a constant length
Fixes a performance regression due to CL 418554. Fixes #56440 Change-Id: I6ff152e9b83084756363f49ee6b0844a7a284880 Reviewed-on: https://go-review.googlesource.com/c/go/+/445875 Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Keith Randall <khr@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com> |
||
|
Wayne Zuo
|
90a3527427 |
cmd/compile: intrinsify Sub64 on loong64
This is a follow up of CL 420095 on loong64. file before after Δ % compile/internal/ssa.a 35649482 35653274 +3792 +0.011% compile/internal/ssagen.a 4099858 4098728 -1130 -0.028% ecdh.a 227896 226896 -1000 -0.439% internal/nistec/fiat.a 1212254 1128184 -84070 -6.935% tls.a 3256800 3256802 +2 +0.000% big.a 1708518 1702496 -6022 -0.352% bits.a 106762 105734 -1028 -0.963% math.a 578762 577288 -1474 -0.255% netip.a 555922 555610 -312 -0.056% net.a 3286528 3286530 +2 +0.000% golang.org/x/crypto/internal/poly1305.a 109546 107686 -1860 -1.698% total 260392768 260299668 -93100 -0.036% Change-Id: Ieffca705aae5666501f284502d986ca179dde494 Reviewed-on: https://go-review.googlesource.com/c/go/+/428557 Reviewed-by: Carlos Amedee <carlos@golang.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> |
||
|
Wayne Zuo
|
97760ed651 |
cmd/compile: intrinsify Add64 on loong64
This is a follow up of CL 420094 on loong64.
Reduce go toolchain size slightly on linux/loong64.
compilecmp HEAD~1 -> HEAD
HEAD~1 (
|
||
|
Wayne Zuo
|
af668c689c |
cmd/compile: fold constant shift with extension on riscv64
For example: movb a0, a0 srai $1, a0, a0 the assembler will expand to: slli $56, a0, a0 srai $56, a0, a0 srai $1, a0, a0 this CL optimize to: slli $56, a0, a0 srai $57, a0, a0 Remove 270+ instructions from Go binary on linux/riscv64. Change-Id: I375e19f9d3bd54f2781791d8cbe5970191297dc8 Reviewed-on: https://go-review.googlesource.com/c/go/+/428496 Reviewed-by: Keith Randall <khr@google.com> Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> Reviewed-by: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> |
||
eric fang
|
ddc7d2a80c |
cmd/compile: add late lower pass for last rules to run
Usually optimization rules have corresponding priorities, some need to
be run first, some run next, and some run last, which produces the best
code. But currently our optimization rules have no priority, this CL
adds a late lower pass that runs those rules that need to be run at last,
such as split unreasonable constant folding. This pass can be seen as
the second round of the lower pass.
For example:
func foo(a, b uint64) uint64 {
d := a+0x1234568
d1 := b+0x1234568
return d&d1
}
The code generated by the master branch:
0x0004 00004 ADD $19088744, R0, R2 // movz+movk+add
0x0010 00016 ADD $19088744, R1, R1 // movz+movk+add
0x001c 00028 AND R1, R2, R0
This is because the current constant folding optimization rules do not
take into account the range of constants, causing the constant to be
loaded repeatedly. This CL splits these unreasonable constants folding
in the late lower pass. With this CL the generated code:
0x0004 00004 MOVD $19088744, R2 // movz+movk
0x000c 00012 ADD R0, R2, R3
0x0010 00016 ADD R1, R2, R1
0x0014 00020 AND R1, R3, R0
This CL also adds constant folding optimization for ADDS instruction.
In addition, in order not to introduce the codegen regression, an
optimization rule is added to change the addition of a negative number
into a subtraction of a positive number.
go1 benchmarks:
name old time/op new time/op delta
BinaryTree17-8 1.22s ± 1% 1.24s ± 0% +1.56% (p=0.008 n=5+5)
Fannkuch11-8 1.54s ± 0% 1.53s ± 0% -0.69% (p=0.016 n=4+5)
FmtFprintfEmpty-8 14.1ns ± 0% 14.1ns ± 0% ~ (p=0.079 n=4+5)
FmtFprintfString-8 26.0ns ± 0% 26.1ns ± 0% +0.23% (p=0.008 n=5+5)
FmtFprintfInt-8 32.3ns ± 0% 32.9ns ± 1% +1.72% (p=0.008 n=5+5)
FmtFprintfIntInt-8 54.5ns ± 0% 55.5ns ± 0% +1.83% (p=0.008 n=5+5)
FmtFprintfPrefixedInt-8 61.5ns ± 0% 62.0ns ± 0% +0.93% (p=0.008 n=5+5)
FmtFprintfFloat-8 72.0ns ± 0% 73.6ns ± 0% +2.24% (p=0.008 n=5+5)
FmtManyArgs-8 221ns ± 0% 224ns ± 0% +1.22% (p=0.008 n=5+5)
GobDecode-8 1.91ms ± 0% 1.93ms ± 0% +0.98% (p=0.008 n=5+5)
GobEncode-8 1.40ms ± 1% 1.39ms ± 0% -0.79% (p=0.032 n=5+5)
Gzip-8 115ms ± 0% 117ms ± 1% +1.17% (p=0.008 n=5+5)
Gunzip-8 19.4ms ± 1% 19.3ms ± 0% -0.71% (p=0.016 n=5+4)
HTTPClientServer-8 27.0µs ± 0% 27.3µs ± 0% +0.80% (p=0.008 n=5+5)
JSONEncode-8 3.36ms ± 1% 3.33ms ± 0% ~ (p=0.056 n=5+5)
JSONDecode-8 17.5ms ± 2% 17.8ms ± 0% +1.71% (p=0.016 n=5+4)
Mandelbrot200-8 2.29ms ± 0% 2.29ms ± 0% ~ (p=0.151 n=5+5)
GoParse-8 1.35ms ± 1% 1.36ms ± 1% ~ (p=0.056 n=5+5)
RegexpMatchEasy0_32-8 24.5ns ± 0% 24.5ns ± 0% ~ (p=0.444 n=4+5)
RegexpMatchEasy0_1K-8 131ns ±11% 118ns ± 6% ~ (p=0.056 n=5+5)
RegexpMatchEasy1_32-8 22.9ns ± 0% 22.9ns ± 0% ~ (p=0.905 n=4+5)
RegexpMatchEasy1_1K-8 126ns ± 0% 127ns ± 0% ~ (p=0.063 n=4+5)
RegexpMatchMedium_32-8 486ns ± 5% 483ns ± 0% ~ (p=0.381 n=5+4)
RegexpMatchMedium_1K-8 15.4µs ± 1% 15.5µs ± 0% ~ (p=0.151 n=5+5)
RegexpMatchHard_32-8 687ns ± 0% 686ns ± 0% ~ (p=0.103 n=5+5)
RegexpMatchHard_1K-8 20.7µs ± 0% 20.7µs ± 1% ~ (p=0.151 n=5+5)
Revcomp-8 175ms ± 2% 176ms ± 3% ~ (p=1.000 n=5+5)
Template-8 20.4ms ± 6% 20.1ms ± 2% ~ (p=0.151 n=5+5)
TimeParse-8 112ns ± 0% 113ns ± 0% +0.97% (p=0.016 n=5+4)
TimeFormat-8 156ns ± 0% 145ns ± 0% -7.14% (p=0.029 n=4+4)
Change-Id: I3ced26e89041f873ac989586514ccc5ee09f13da
Reviewed-on: https://go-review.googlesource.com/c/go/+/425134
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Eric Fang <eric.fang@arm.com>
|
||
|
Keith Randall
|
6485e8f503 |
cmd/compile: use stricter rule for possible partial overlap
Partial overlaps can only happen for strict sub-pieces of larger arrays. That's a much stronger condition than the current optimization rules. Update #54467 Change-Id: I11e539b71099e50175f37ee78fddf69283f83ee5 Reviewed-on: https://go-review.googlesource.com/c/go/+/433056 TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com> |
||
|
eric fang
|
cf53990b18 |
cmd/compile: Add some CMP and CMN optimization rules on arm64
This CL adds some optimizaion rules: 1, Converts CMP to CMN, or vice versa, when comparing with a negative number. 2, For equal and not equal comparisons, CMP can be converted to CMN in some cases. In theory we could do the same optimization for LT, LE, GT and GE, but need to account for overflow, this CL doesn't handle them. There are no noticeable performance changes. Change-Id: Ia49266c019ab7908ebc9510c2f02e121b1607869 Reviewed-on: https://go-review.googlesource.com/c/go/+/429795 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Keith Randall <khr@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: Eric Fang <eric.fang@arm.com> |
||
Joel Sing
|
a7bcc94719 |
cmd/compile: resolve known outcomes for SLTI/SLTIU on riscv64
When SLTI/SLTIU is used with ANDI/ORI, it may be possible to determine the outcome based on the values of the immediates. Resolve these cases. Improves code generation for various shift operations. While here, sort tests by architecture to improve readability and ease future maintenance. Change-Id: I87e71e016a0e396a928e7d6389a2df61583dfd8d Reviewed-on: https://go-review.googlesource.com/c/go/+/428217 Reviewed-by: Wayne Zuo <wdvxdr@golangcn.org> TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: Jenny Rakoczy <jenny@golang.org> Reviewed-by: Jenny Rakoczy <jenny@golang.org> Run-TryBot: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com> Auto-Submit: Jenny Rakoczy <jenny@golang.org> |
||
ruinan
|
454a058ffc |
cmd/compile: Add shiftIsBounded check for logic shifts of arm64
This CL adds shiftIsBounded checks for the Lsh* and Rsh* rules in arm64. There is no need to check the shift value again with CMP + CSEL when the shift value is valid. Change-Id: I54620de64f02a1b5a11089add237248ae2de01b4 Reviewed-on: https://go-review.googlesource.com/c/go/+/417714 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Keith Randall <khr@google.com> Reviewed-by: Heschi Kreinick <heschi@google.com> |
||
|
Keith Randall
|
5b1fbfba1c |
cmd/compile: rewrite >>c<<c to &^(1<<c-1)
Fixes #54496 Change-Id: I3c2ed8cd55836d5b07c8cdec00d3b584885aca79 Reviewed-on: https://go-review.googlesource.com/c/go/+/424856 Reviewed-by: Keith Randall <khr@google.com> Reviewed-by: Heschi Kreinick <heschi@google.com> Run-TryBot: Martin Möhrmann <martin@golang.org> TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <martin@golang.org> |
||
Matthew Dempsky
|
34f0029a85 |
cmd/compile/internal/noder: allow OCONVNOP for identical iface conversions
In go.dev/cl/421821, I included a hack to force OCONVNOP back to OCONVIFACE for conversions involving shape types and non-empty interfaces. The comment correctly noted that this was only needed for conversions between non-identical types, but the code was conservative and applied to even conversions between identical types. This CL adds an extra bool to record whether the conversion is between identical types, so we can keep OCONVNOP instead of forcing back to OCONVIFACE. This has a small improvement to generated code, because we no longer need a convI2I call (as demonstrated by codegen/ifaces.go). But more usefully, this is relevant to pruning unnecessary itab slots in runtime dictionaries (next CL). Change-Id: I94f89e961cd26629b925037fea58d283140766ff Reviewed-on: https://go-review.googlesource.com/c/go/+/427678 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Matthew Dempsky <mdempsky@google.com> Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com> Reviewed-by: Keith Randall <khr@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> |
||
Derek Parker
|
6605686e3b |
cmd/compile: new inline heuristic for struct compares
This CL changes the heuristic used to determine whether we can inline a
struct equality check or if we must generate a function and call that
function for equality.
The old method was to count struct fields, but this can lead to poor
in lining decisions. We should really be determining the cost of the
equality check and use that to determine if we should inline or generate
a function.
The new benchmark provided in this CL returns the following when compared
against tip:
```
name old time/op new time/op delta
EqStruct-32 2.46ns ± 4% 0.25ns ±10% -89.72% (p=0.000 n=39+39)
```
Fixes #38494
Change-Id: Ie06b80a2b2a03a3fd0978bcaf7715f9afb66e0ab
GitHub-Last-Rev:
|
||
|
ruinan
|
121344ac33 |
cmd/compile: optimize RotateLeft8/16 on arm64
This CL optimizes RotateLeft8/16 on arm64. For 16 bits, we form a 32 bits register by duplicating two 16 bits registers, then use RORW instruction to do the rotate shift. For 8 bits, we just use LSR and LSL instead of RORW because the code is simpler. Benchmark Old ThisCL delta RotateLeft8-46 2.16 ns/op 1.73 ns/op -19.70% RotateLeft16-46 2.16 ns/op 1.54 ns/op -28.53% Change-Id: I09cde4383d12e31876a57f8cdfd3bb4f324fadb0 Reviewed-on: https://go-review.googlesource.com/c/go/+/420976 Reviewed-by: Keith Randall <khr@google.com> Auto-Submit: Keith Randall <khr@golang.org> Reviewed-by: Keith Randall <khr@golang.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Heschi Kreinick <heschi@google.com> Run-TryBot: Keith Randall <khr@golang.org> |
||
|
ruinan
|
54c7bc9cff |
cmd/compile: optimize shift ops on arm64 when the shift value is v&63
For the following code case: var x uint64 x >> (shift & 63) We can directly genereta `x >> shift` on arm64, since the hardware will only use the bottom 6 bits of the shift amount. Benchmark old time/op new time/op delta ShiftArithmeticRight-8 0.40ns 0.31ns -21.7% Change-Id: Id58c8a5b2f6dd5c30c3876f4a36e11b4d81e2dc9 Reviewed-on: https://go-review.googlesource.com/c/go/+/425294 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Keith Randall <khr@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Auto-Submit: Keith Randall <khr@golang.org> Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Heschi Kreinick <heschi@google.com> |
||
|
Keith Randall
|
33a7e5a4b4 |
cmd/compile: combine multiple rotate instructions
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> |
||
|
Keith Randall
|
69aed4712d |
cmd/compile: use better splitting condition for string binary search
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> |
||
|
Wayne Zuo
|
da6556968f |
cmd/compile: simplify bounded shift on riscv64
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> |
||
|
Wayne Zuo
|
e8f0340fa4 |
cmd/compile: intrinsify RotateLeft{32,64} on loong64
Benchmark on crypto/sha256 (provided by Xiaodong Liu): name old time/op new time/op delta Hash8Bytes/New 1.19µs ± 0% 0.97µs ± 0% -18.75% (p=0.000 n=9+9) Hash8Bytes/Sum224 1.21µs ± 0% 0.97µs ± 0% -20.04% (p=0.000 n=9+10) Hash8Bytes/Sum256 1.21µs ± 0% 0.98µs ± 0% -19.16% (p=0.000 n=10+7) Hash1K/New 15.9µs ± 0% 12.4µs ± 0% -22.10% (p=0.000 n=10+10) Hash1K/Sum224 15.9µs ± 0% 12.4µs ± 0% -22.18% (p=0.000 n=8+10) Hash1K/Sum256 15.9µs ± 0% 12.4µs ± 0% -22.15% (p=0.000 n=10+9) Hash8K/New 119µs ± 0% 92µs ± 0% -22.40% (p=0.000 n=10+9) Hash8K/Sum224 119µs ± 0% 92µs ± 0% -22.41% (p=0.000 n=9+10) Hash8K/Sum256 119µs ± 0% 92µs ± 0% -22.40% (p=0.000 n=9+9) name old speed new speed delta Hash8Bytes/New 6.70MB/s ± 0% 8.25MB/s ± 0% +23.13% (p=0.000 n=10+10) Hash8Bytes/Sum224 6.60MB/s ± 0% 8.26MB/s ± 0% +25.06% (p=0.000 n=10+10) Hash8Bytes/Sum256 6.59MB/s ± 0% 8.15MB/s ± 0% +23.67% (p=0.000 n=10+7) Hash1K/New 64.3MB/s ± 0% 82.5MB/s ± 0% +28.36% (p=0.000 n=10+10) Hash1K/Sum224 64.3MB/s ± 0% 82.6MB/s ± 0% +28.51% (p=0.000 n=10+10) Hash1K/Sum256 64.3MB/s ± 0% 82.6MB/s ± 0% +28.46% (p=0.000 n=9+9) Hash8K/New 69.0MB/s ± 0% 89.0MB/s ± 0% +28.87% (p=0.000 n=10+8) Hash8K/Sum224 69.0MB/s ± 0% 89.0MB/s ± 0% +28.88% (p=0.000 n=9+10) Hash8K/Sum256 69.0MB/s ± 0% 88.9MB/s ± 0% +28.87% (p=0.000 n=8+9) Benchmark on crypto/sha512 (provided by Xiaodong Liu): name old time/op new time/op delta Hash8Bytes/New 1.55µs ± 0% 1.31µs ± 0% -15.67% (p=0.000 n=10+10) Hash8Bytes/Sum384 1.59µs ± 0% 1.35µs ± 0% -14.97% (p=0.000 n=10+10) Hash8Bytes/Sum512 1.62µs ± 0% 1.39µs ± 0% -14.02% (p=0.000 n=10+10) Hash1K/New 10.7µs ± 0% 8.6µs ± 0% -19.60% (p=0.000 n=8+8) Hash1K/Sum384 10.8µs ± 0% 8.7µs ± 0% -19.40% (p=0.000 n=9+9) Hash1K/Sum512 10.8µs ± 0% 8.7µs ± 0% -19.35% (p=0.000 n=9+10) Hash8K/New 74.6µs ± 0% 59.6µs ± 0% -20.08% (p=0.000 n=10+9) Hash8K/Sum384 74.7µs ± 0% 59.7µs ± 0% -20.04% (p=0.000 n=9+8) Hash8K/Sum512 74.7µs ± 0% 59.7µs ± 0% -20.01% (p=0.000 n=10+10) name old speed new speed delta Hash8Bytes/New 5.16MB/s ± 0% 6.12MB/s ± 0% +18.60% (p=0.000 n=10+8) Hash8Bytes/Sum384 5.02MB/s ± 0% 5.90MB/s ± 0% +17.56% (p=0.000 n=10+10) Hash8Bytes/Sum512 4.94MB/s ± 0% 5.74MB/s ± 0% +16.29% (p=0.000 n=10+9) Hash1K/New 95.4MB/s ± 0% 118.6MB/s ± 0% +24.38% (p=0.000 n=10+10) Hash1K/Sum384 95.0MB/s ± 0% 117.9MB/s ± 0% +24.06% (p=0.000 n=8+9) Hash1K/Sum512 94.8MB/s ± 0% 117.5MB/s ± 0% +23.99% (p=0.000 n=8+9) Hash8K/New 110MB/s ± 0% 137MB/s ± 0% +25.11% (p=0.000 n=9+6) Hash8K/Sum384 110MB/s ± 0% 137MB/s ± 0% +25.07% (p=0.000 n=9+8) Hash8K/Sum512 110MB/s ± 0% 137MB/s ± 0% +25.01% (p=0.000 n=10+10) Change-Id: I28ccfce634659305a336c8e0a3f8589f7361d661 Reviewed-on: https://go-review.googlesource.com/c/go/+/422317 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> Reviewed-by: Ian Lance Taylor <iant@google.com> Reviewed-by: David Chase <drchase@google.com> |
||
|
Wayne Zuo
|
a6219737e3 |
cmd/compile: intrinsify Sub64 on riscv64
After this CL, the performance difference in crypto/elliptic benchmarks on linux/riscv64 are: name old time/op new time/op delta ScalarBaseMult/P256 1.64ms ± 1% 1.60ms ± 1% -2.36% (p=0.008 n=5+5) ScalarBaseMult/P224 1.53ms ± 1% 1.47ms ± 2% -4.24% (p=0.008 n=5+5) ScalarBaseMult/P384 5.12ms ± 2% 5.03ms ± 2% ~ (p=0.095 n=5+5) ScalarBaseMult/P521 22.3ms ± 2% 13.8ms ± 1% -37.89% (p=0.008 n=5+5) ScalarMult/P256 4.49ms ± 2% 4.26ms ± 2% -5.13% (p=0.008 n=5+5) ScalarMult/P224 4.33ms ± 1% 4.09ms ± 1% -5.59% (p=0.008 n=5+5) ScalarMult/P384 16.3ms ± 1% 15.5ms ± 2% -4.78% (p=0.008 n=5+5) ScalarMult/P521 101ms ± 0% 47ms ± 2% -53.36% (p=0.008 n=5+5) Change-Id: I31cf0506e27f9d85f576af1813630a19c20dda8a Reviewed-on: https://go-review.googlesource.com/c/go/+/420095 Reviewed-by: Cherry Mui <cherryyz@google.com> Reviewed-by: Joel Sing <joel@sing.id.au> Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> TryBot-Result: Gopher Robot <gobot@golang.org> |
||
|
Wayne Zuo
|
969f48a3a2 |
cmd/compile: intrinsify Add64 on riscv64
According to RISCV instruction set manual v2.2 Sec 2.4, we can implement overflowing check for unsigned addition cheaply using SLTU instructions. After this CL, the performance difference in crypto/elliptic benchmarks on linux/riscv64 are: name old time/op new time/op delta ScalarBaseMult/P256 1.93ms ± 1% 1.64ms ± 1% -14.96% (p=0.008 n=5+5) ScalarBaseMult/P224 1.80ms ± 2% 1.53ms ± 1% -14.89% (p=0.008 n=5+5) ScalarBaseMult/P384 6.15ms ± 2% 5.12ms ± 2% -16.73% (p=0.008 n=5+5) ScalarBaseMult/P521 25.9ms ± 1% 22.3ms ± 2% -13.78% (p=0.008 n=5+5) ScalarMult/P256 5.59ms ± 1% 4.49ms ± 2% -19.79% (p=0.008 n=5+5) ScalarMult/P224 5.42ms ± 1% 4.33ms ± 1% -20.01% (p=0.008 n=5+5) ScalarMult/P384 19.9ms ± 2% 16.3ms ± 1% -18.15% (p=0.008 n=5+5) ScalarMult/P521 97.3ms ± 1% 100.7ms ± 0% +3.48% (p=0.008 n=5+5) Change-Id: Ic4c82ced4b072a4a6575343fa9f29dd09b0cabc4 Reviewed-on: https://go-review.googlesource.com/c/go/+/420094 Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com> Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> Reviewed-by: Joel Sing <joel@sing.id.au> TryBot-Result: Gopher Robot <gobot@golang.org> |
||
|
Wayne Zuo
|
b60432df14 |
cmd/compile: deadcode for LoweredMuluhilo on riscv64
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> |
||
|
Keith Randall
|
60ad3c48f5 |
cmd/compile: move SSA rotate instruction detection to arch-independent rules
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> |
||
|
eric fang
|
9f0f87c806 |
cmd/internal/obj/arm64: remove the transition from $0 to ZR
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> |
||
|
eric fang
|
0a52d80666 |
cmd/compile/internal/ssa: optimize memory moving on arm64
This CL optimizes memory moving with LDP and STP on arm64. Benchmarks: name old time/op new time/op delta ClearFat7-160 1.08ns ± 0% 0.95ns ± 0% -11.41% (p=0.008 n=5+5) ClearFat8-160 0.84ns ± 0% 0.84ns ± 0% -0.95% (p=0.008 n=5+5) ClearFat11-160 1.08ns ± 0% 0.95ns ± 0% -11.46% (p=0.008 n=5+5) ClearFat12-160 0.95ns ± 0% 0.95ns ± 0% ~ (p=0.063 n=4+5) ClearFat13-160 1.08ns ± 0% 0.95ns ± 0% -11.45% (p=0.008 n=5+5) ClearFat14-160 1.08ns ± 0% 0.95ns ± 0% -11.47% (p=0.008 n=5+5) ClearFat15-160 1.24ns ± 0% 0.95ns ± 0% -22.98% (p=0.029 n=4+4) ClearFat16-160 0.84ns ± 0% 0.83ns ± 0% -0.11% (p=0.008 n=5+5) ClearFat24-160 2.15ns ± 0% 2.15ns ± 0% ~ (all equal) ClearFat32-160 2.86ns ± 0% 2.86ns ± 0% ~ (p=0.333 n=5+4) ClearFat40-160 2.15ns ± 0% 2.15ns ± 0% ~ (all equal) ClearFat48-160 3.32ns ± 1% 3.31ns ± 1% ~ (p=0.690 n=5+5) ClearFat56-160 2.15ns ± 0% 2.15ns ± 0% ~ (all equal) ClearFat64-160 3.25ns ± 1% 3.26ns ± 1% ~ (p=0.841 n=5+5) ClearFat72-160 2.22ns ± 0% 2.22ns ± 0% ~ (p=0.444 n=5+5) ClearFat128-160 4.03ns ± 0% 4.04ns ± 0% +0.32% (p=0.008 n=5+5) ClearFat256-160 6.44ns ± 0% 6.44ns ± 0% +0.08% (p=0.016 n=4+5) ClearFat512-160 12.2ns ± 0% 12.2ns ± 0% +0.13% (p=0.008 n=5+5) ClearFat1024-160 24.3ns ± 0% 24.3ns ± 0% ~ (p=0.167 n=5+5) ClearFat1032-160 24.5ns ± 0% 24.5ns ± 0% ~ (p=0.238 n=4+5) ClearFat1040-160 29.2ns ± 0% 29.3ns ± 0% +0.34% (p=0.008 n=5+5) CopyFat7-160 1.43ns ± 0% 1.07ns ± 0% -24.97% (p=0.008 n=5+5) CopyFat8-160 0.89ns ± 0% 0.89ns ± 0% ~ (p=0.238 n=5+5) CopyFat11-160 1.43ns ± 0% 1.07ns ± 0% -24.97% (p=0.008 n=5+5) CopyFat12-160 1.07ns ± 0% 1.07ns ± 0% ~ (p=0.238 n=5+4) CopyFat13-160 1.43ns ± 0% 1.07ns ± 0% ~ (p=0.079 n=4+5) CopyFat14-160 1.43ns ± 0% 1.07ns ± 0% -24.95% (p=0.008 n=5+5) CopyFat15-160 1.79ns ± 0% 1.07ns ± 0% ~ (p=0.079 n=4+5) CopyFat16-160 1.07ns ± 0% 1.07ns ± 0% ~ (p=0.444 n=5+5) CopyFat24-160 1.84ns ± 2% 1.67ns ± 0% -9.28% (p=0.008 n=5+5) CopyFat32-160 3.22ns ± 0% 2.92ns ± 0% -9.40% (p=0.008 n=5+5) CopyFat64-160 3.64ns ± 0% 3.57ns ± 0% -1.96% (p=0.008 n=5+5) CopyFat72-160 3.56ns ± 0% 3.11ns ± 0% -12.89% (p=0.008 n=5+5) CopyFat128-160 5.06ns ± 0% 5.06ns ± 0% +0.04% (p=0.048 n=5+5) CopyFat256-160 9.13ns ± 0% 9.13ns ± 0% ~ (p=0.659 n=5+5) CopyFat512-160 17.4ns ± 0% 17.4ns ± 0% ~ (p=0.167 n=5+5) CopyFat520-160 17.2ns ± 0% 17.3ns ± 0% +0.37% (p=0.008 n=5+5) CopyFat1024-160 34.1ns ± 0% 34.0ns ± 0% ~ (p=0.127 n=5+5) CopyFat1032-160 80.9ns ± 0% 34.2ns ± 0% -57.74% (p=0.008 n=5+5) CopyFat1040-160 94.4ns ± 0% 41.7ns ± 0% -55.78% (p=0.016 n=5+4) Change-Id: I14186f9f82b0ecf8b6c02191dc5da566b9a21e6c Reviewed-on: https://go-review.googlesource.com/c/go/+/421654 Reviewed-by: Cherry Mui <cherryyz@google.com> Run-TryBot: Eric Fang <eric.fang@arm.com> Reviewed-by: Keith Randall <khr@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> |
||
Cherry Mui
|
8bf9e01473 |
cmd/compile: split Muluhilo op on ARM64
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> |
||
|
Keith Randall
|
908499adec |
cmd/compile: stop using VARKILL
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> |
||
|
Archana R
|
d09c6ac417 |
test/codegen: updated multiple tests to verify on ppc64,ppc64le
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> |
||
|
Wayne Zuo
|
09932f95f5 |
cmd/compile: combine more constant stores on amd64
Fixes #53324 Change-Id: I06149d860f858b082235e9d80bf0ea494679b386 Reviewed-on: https://go-review.googlesource.com/c/go/+/411614 Reviewed-by: Keith Randall <khr@google.com> Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> Reviewed-by: Keith Randall <khr@golang.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@google.com> |
||
|
eric fang
|
efe5929dbd |
cmd/compile/internal/ssa: optimize ARM64 code with TST
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> |
||
Cuong Manh Le
|
0f8dffd0aa |
all: use ":" for compiler generated symbols
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> |
||
|
Lynn Boger
|
c1bfefe9d1 |
cmd/compile: fix confusion with ANDCCconst in PPC64 rules
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> |
||
|
Keith Randall
|
c2a9c55823 |
cmd/compile: optimize unsafe.Slice generated code
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> |
||
cuiweixie
|
e7307034cc |
cmd/compile: store combine on amd64
Fixes #54120
Change-Id: I6915b6e8d459d9becfdef4fdcba95ee4dea6af05
GitHub-Last-Rev:
|
||
|
Matthew Dempsky
|
ab8d7dd75e |
cmd/compile: set LocalPkg.Path to -p flag
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> |
||
|
Cherry Mui
|
540f8c2b50 |
cmd/compile: use jump table on ARM64
Following CL 357330, use jump tables on ARM64. name old time/op new time/op delta Switch8Predictable-4 3.41ns ± 0% 3.21ns ± 0% ~ (p=0.079 n=4+5) Switch8Unpredictable-4 12.0ns ± 0% 9.5ns ± 0% -21.17% (p=0.000 n=5+4) Switch32Predictable-4 3.06ns ± 0% 2.82ns ± 0% -7.78% (p=0.008 n=5+5) Switch32Unpredictable-4 13.3ns ± 0% 9.5ns ± 0% -28.87% (p=0.016 n=4+5) SwitchStringPredictable-4 3.71ns ± 0% 3.21ns ± 0% -13.43% (p=0.000 n=5+4) SwitchStringUnpredictable-4 14.8ns ± 0% 15.1ns ± 0% +2.37% (p=0.008 n=5+5) Change-Id: Ia0b85df7ca9273cf70c05eb957225c6e61822fa6 Reviewed-on: https://go-review.googlesource.com/c/go/+/403979 TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Cherry Mui <cherryyz@google.com> Reviewed-by: David Chase <drchase@google.com> |
||
|
Paul E. Murphy
|
0c43878baa |
cmd/compile: lower Add64/Sub64 into ssa on PPC64
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> |
||
|
Jorropo
|
e1e056fa6a |
cmd/compile: fold constants found by prove
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:
|
||
|
Paul E. Murphy
|
c570f0eda2 |
cmd/compile: combine OR + NOT into ORN on PPC64
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> |
||
|
Cuong Manh Le
|
0668e3cb1a |
cmd/compile: support pointers to arrays in arrayClear
Fixes #52635 Change-Id: I85f182931e30292983ef86c55a0ab6e01282395c Reviewed-on: https://go-review.googlesource.com/c/go/+/403337 Reviewed-by: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gopher Robot <gobot@golang.org> Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com> Reviewed-by: Ian Lance Taylor <iant@google.com> |
||
|
Keith Randall
|
c4b2288755 |
cmd/compile: add jump table codegen test
Change-Id: Ic67f676f5ebe146166a0d3c1d78a802881320e49 Reviewed-on: https://go-review.googlesource.com/c/go/+/400375 Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com> Reviewed-by: Keith Randall <khr@google.com> |
||
|
Wayne Zuo
|
66f03f79da |
cmd/compile: add SHLX&SHRX without load
Change-Id: I79eb5e7d6bcb23f26d3a100e915efff6dae70391 Reviewed-on: https://go-review.googlesource.com/c/go/+/399061 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Keith Randall <khr@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com> |
||
|
Wayne Zuo
|
517781b391 |
cmd/compile: add SARXQload and SARXLload
Change-Id: I4e8dc5009a5b8af37d71b62f1322f94806d3e9d9 Reviewed-on: https://go-review.googlesource.com/c/go/+/399056 Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> Reviewed-by: Keith Randall <khr@golang.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Keith Randall <khr@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com> |
||
|
Wayne Zuo
|
d6320f1a58 |
cmd/compile: add SARX instruction for GOAMD64>=3
name old time/op new time/op delta ShiftArithmeticRight-8 0.68ns ± 5% 0.30ns ± 6% -56.14% (p=0.000 n=10+10) Change-Id: I052a0d7b9e6526d526276444e588b0cc288beff4 Reviewed-on: https://go-review.googlesource.com/c/go/+/399055 Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org> TryBot-Result: Gopher Robot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Auto-Submit: Keith Randall <khr@golang.org> Reviewed-by: Keith Randall <khr@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com> |
||
Wayne Zuo
|
32de2b0d1c |
cmd/compile: add MOVBE index load/store
Fixes #51724 Change-Id: I94e650a7482dc4c479d597f0162a6a89d779708d Reviewed-on: https://go-review.googlesource.com/c/go/+/395474 Reviewed-by: Keith Randall <khr@golang.org> Trust: Cherry Mui <cherryyz@google.com> |
||
|
Wayne Zuo
|
615d3c3040 |
test: adjust load and store test
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>
|
||
|
Wayne Zuo
|
7fbabe8d57 |
cmd/compile: use shlx&shrx instruction for GOAMD64>=v3
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> |
||
|
Wayne Zuo
|
a92ca51507 |
cmd/compile: use LZCNT instruction for GOAMD64>=3
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> |
||
|
Wayne Zuo
|
ba6df85c7c |
cmd/compile: add MOVBEWstore support for GOAMD64>=3
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> |
||
fanzha02
|
8ab42a945a |
cmd/compile: merge ANDconst and UBFX into UBFX on arm64
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> |
||
|
Keith Randall
|
aaa3d39f27 |
cmd/compile: include all entries in map literal hint size
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> |
||
|
Archana R
|
4056934e48 |
test/codegen: updated arithmetic tests to verify on ppc64,ppc64le
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> |
||
|
Archana R
|
8b9c0d1a79 |
test/codegen: updated comparison test to verify on ppc64,ppc64le
Updated test/codegen/comparison.go to verify memequal is inlined as implemented in CL 328291. Change-Id: If7824aed37ee1f8640e54fda0f9b7610582ba316 Reviewed-on: https://go-review.googlesource.com/c/go/+/357289 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> |
||
|
wdvxdr
|
fe7df4c4d0 |
cmd/compile: use MOVBE instruction for GOAMD64>=v3
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> |
||
Daniel Martí
|
b0351bfd7d |
Revert "cmd/compile: use MOVBE instruction for GOAMD64>=v3"
This reverts CL 354670. Reason for revert: broke make.bash with GOAMD64=v3. Fixes #49061. Change-Id: I7f2ed99b7c10100c4e0c1462ea91c4c9d8c609b2 Reviewed-on: https://go-review.googlesource.com/c/go/+/356790 Trust: Daniel Martí <mvdan@mvdan.cc> Run-TryBot: Daniel Martí <mvdan@mvdan.cc> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Koichi Shiraishi <zchee.io@gmail.com> Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com> |
||
|
Lynn Boger
|
33b3260c1e |
cmd/compile/internal/ssagen: set BitLen32 as intrinsic on PPC64
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> |
||
|
wdvxdr
|
3e5cc4d6f6 |
cmd/compile: use MOVBE instruction for GOAMD64>=v3
encoding/binary benchmark on my laptop: name old time/op new time/op delta ReadSlice1000Int32s-8 4.42µs ± 5% 4.20µs ± 1% -4.94% (p=0.046 n=9+8) ReadStruct-8 359ns ± 8% 368ns ± 5% +2.35% (p=0.041 n=9+10) WriteStruct-8 349ns ± 1% 357ns ± 1% +2.15% (p=0.000 n=8+10) ReadInts-8 235ns ± 1% 233ns ± 1% -1.01% (p=0.005 n=10+10) WriteInts-8 265ns ± 1% 274ns ± 1% +3.45% (p=0.000 n=10+10) WriteSlice1000Int32s-8 4.61µs ± 5% 4.59µs ± 5% ~ (p=0.986 n=10+10) PutUint16-8 0.56ns ± 4% 0.57ns ± 4% ~ (p=0.101 n=10+10) PutUint32-8 0.83ns ± 2% 0.56ns ± 6% -32.91% (p=0.000 n=10+10) PutUint64-8 0.81ns ± 3% 0.62ns ± 4% -23.82% (p=0.000 n=10+10) LittleEndianPutUint16-8 0.55ns ± 4% 0.55ns ± 3% ~ (p=0.926 n=10+10) LittleEndianPutUint32-8 0.41ns ± 4% 0.42ns ± 3% ~ (p=0.148 n=10+9) LittleEndianPutUint64-8 0.55ns ± 2% 0.56ns ± 6% ~ (p=0.897 n=10+10) ReadFloats-8 60.4ns ± 4% 59.0ns ± 1% -2.25% (p=0.007 n=10+10) WriteFloats-8 72.3ns ± 2% 71.5ns ± 7% ~ (p=0.089 n=10+10) ReadSlice1000Float32s-8 4.21µs ± 3% 4.18µs ± 2% ~ (p=0.197 n=10+10) WriteSlice1000Float32s-8 4.61µs ± 2% 4.68µs ± 7% ~ (p=1.000 n=8+10) ReadSlice1000Uint8s-8 250ns ± 4% 247ns ± 4% ~ (p=0.324 n=10+10) WriteSlice1000Uint8s-8 227ns ± 5% 229ns ± 2% ~ (p=0.193 n=10+7) PutUvarint32-8 15.3ns ± 2% 15.4ns ± 4% ~ (p=0.782 n=10+10) PutUvarint64-8 38.5ns ± 1% 38.6ns ± 5% ~ (p=0.396 n=8+10) name old speed new speed delta ReadSlice1000Int32s-8 890MB/s ±17% 953MB/s ± 1% +7.00% (p=0.027 n=10+8) ReadStruct-8 209MB/s ± 8% 204MB/s ± 5% -2.42% (p=0.043 n=9+10) WriteStruct-8 214MB/s ± 3% 210MB/s ± 1% -1.75% (p=0.003 n=9+10) ReadInts-8 127MB/s ± 1% 129MB/s ± 1% +1.01% (p=0.006 n=10+10) WriteInts-8 113MB/s ± 1% 109MB/s ± 1% -3.34% (p=0.000 n=10+10) WriteSlice1000Int32s-8 868MB/s ± 5% 872MB/s ± 5% ~ (p=1.000 n=10+10) PutUint16-8 3.55GB/s ± 4% 3.50GB/s ± 4% ~ (p=0.093 n=10+10) PutUint32-8 4.83GB/s ± 2% 7.21GB/s ± 6% +49.16% (p=0.000 n=10+10) PutUint64-8 9.89GB/s ± 3% 12.99GB/s ± 4% +31.26% (p=0.000 n=10+10) LittleEndianPutUint16-8 3.65GB/s ± 4% 3.65GB/s ± 4% ~ (p=0.912 n=10+10) LittleEndianPutUint32-8 9.74GB/s ± 3% 9.63GB/s ± 3% ~ (p=0.222 n=9+9) LittleEndianPutUint64-8 14.4GB/s ± 2% 14.3GB/s ± 5% ~ (p=0.912 n=10+10) ReadFloats-8 199MB/s ± 4% 203MB/s ± 1% +2.27% (p=0.007 n=10+10) WriteFloats-8 166MB/s ± 2% 168MB/s ± 7% ~ (p=0.089 n=10+10) ReadSlice1000Float32s-8 949MB/s ± 3% 958MB/s ± 2% ~ (p=0.218 n=10+10) WriteSlice1000Float32s-8 867MB/s ± 2% 857MB/s ± 6% ~ (p=1.000 n=8+10) ReadSlice1000Uint8s-8 4.00GB/s ± 4% 4.06GB/s ± 4% ~ (p=0.353 n=10+10) WriteSlice1000Uint8s-8 4.40GB/s ± 4% 4.36GB/s ± 2% ~ (p=0.193 n=10+7) PutUvarint32-8 262MB/s ± 2% 260MB/s ± 4% ~ (p=0.739 n=10+10) PutUvarint64-8 208MB/s ± 1% 207MB/s ± 5% ~ (p=0.408 n=8+10) Updates #45453 Change-Id: Ifda0d48d54665cef45d46d3aad974062633142c4 Reviewed-on: https://go-review.googlesource.com/c/go/+/354670 Run-TryBot: Alberto Donizetti <alb.donizetti@gmail.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Trust: Matthew Dempsky <mdempsky@google.com> |
||
|
Jake Ciolek
|
732f6fa9d5 |
cmd/compile: use ANDL for small immediates
We can rewrite ANDQ with an immediate fitting in 32bit with an ANDL, which is shorter to encode. Looking at Go binary itself, before the change there was: ANDL: 2337 ANDQ: 4476 After the change: ANDL: 3790 ANDQ: 3024 So we got rid of 1452 ANDQs This makes the Linux x86_64 binary 0.03% smaller. There seems to be an impact on performance. Intel Cascade Lake benchmarks (with perflock): name old time/op new time/op delta BinaryTree17-8 1.91s ± 1% 1.89s ± 1% -1.22% (p=0.000 n=21+18) Fannkuch11-8 2.34s ± 0% 2.34s ± 0% ~ (p=0.052 n=20+20) FmtFprintfEmpty-8 27.7ns ± 1% 27.4ns ± 3% ~ (p=0.497 n=21+21) FmtFprintfString-8 53.2ns ± 0% 51.5ns ± 0% -3.21% (p=0.000 n=20+19) FmtFprintfInt-8 57.3ns ± 0% 55.7ns ± 0% -2.89% (p=0.000 n=19+19) FmtFprintfIntInt-8 92.3ns ± 0% 88.4ns ± 1% -4.23% (p=0.000 n=20+21) FmtFprintfPrefixedInt-8 103ns ± 0% 103ns ± 0% +0.23% (p=0.000 n=20+21) FmtFprintfFloat-8 147ns ± 0% 148ns ± 0% +0.75% (p=0.000 n=20+21) FmtManyArgs-8 384ns ± 0% 381ns ± 0% -0.63% (p=0.000 n=21+21) GobDecode-8 3.86ms ± 1% 3.88ms ± 1% +0.52% (p=0.000 n=20+21) GobEncode-8 2.77ms ± 1% 2.77ms ± 0% ~ (p=0.078 n=21+21) Gzip-8 168ms ± 1% 168ms ± 0% +0.24% (p=0.000 n=20+20) Gunzip-8 25.1ms ± 0% 24.3ms ± 0% -3.03% (p=0.000 n=21+21) HTTPClientServer-8 61.4µs ± 8% 59.1µs ±10% ~ (p=0.088 n=20+21) JSONEncode-8 6.86ms ± 0% 6.70ms ± 0% -2.29% (p=0.000 n=20+19) JSONDecode-8 30.8ms ± 1% 30.6ms ± 1% -0.82% (p=0.000 n=20+20) Mandelbrot200-8 3.85ms ± 0% 3.85ms ± 0% ~ (p=0.191 n=16+17) GoParse-8 2.61ms ± 2% 2.60ms ± 1% ~ (p=0.561 n=21+20) RegexpMatchEasy0_32-8 48.5ns ± 2% 45.9ns ± 3% -5.26% (p=0.000 n=20+21) RegexpMatchEasy0_1K-8 139ns ± 0% 139ns ± 0% +0.27% (p=0.000 n=18+20) RegexpMatchEasy1_32-8 41.3ns ± 0% 42.1ns ± 4% +1.95% (p=0.000 n=17+21) RegexpMatchEasy1_1K-8 216ns ± 2% 216ns ± 0% +0.17% (p=0.020 n=21+19) RegexpMatchMedium_32-8 790ns ± 7% 803ns ± 8% ~ (p=0.178 n=21+21) RegexpMatchMedium_1K-8 23.5µs ± 5% 23.7µs ± 5% ~ (p=0.421 n=21+21) RegexpMatchHard_32-8 1.09µs ± 1% 1.09µs ± 1% -0.53% (p=0.000 n=19+18) RegexpMatchHard_1K-8 33.0µs ± 0% 33.0µs ± 0% ~ (p=0.610 n=21+20) Revcomp-8 348ms ± 0% 353ms ± 0% +1.38% (p=0.000 n=17+18) Template-8 42.0ms ± 1% 41.9ms ± 1% -0.30% (p=0.049 n=20+20) TimeParse-8 185ns ± 0% 185ns ± 0% ~ (p=0.387 n=20+18) TimeFormat-8 237ns ± 1% 241ns ± 1% +1.57% (p=0.000 n=21+21) [Geo mean] 35.4µs 35.2µs -0.66% name old speed new speed delta GobDecode-8 199MB/s ± 1% 198MB/s ± 1% -0.52% (p=0.000 n=20+21) GobEncode-8 277MB/s ± 1% 277MB/s ± 0% ~ (p=0.075 n=21+21) Gzip-8 116MB/s ± 1% 115MB/s ± 0% -0.25% (p=0.000 n=20+20) Gunzip-8 773MB/s ± 0% 797MB/s ± 0% +3.12% (p=0.000 n=21+21) JSONEncode-8 283MB/s ± 0% 290MB/s ± 0% +2.35% (p=0.000 n=20+19) JSONDecode-8 63.0MB/s ± 1% 63.5MB/s ± 1% +0.82% (p=0.000 n=20+20) GoParse-8 22.2MB/s ± 2% 22.3MB/s ± 1% ~ (p=0.539 n=21+20) RegexpMatchEasy0_32-8 660MB/s ± 2% 697MB/s ± 3% +5.57% (p=0.000 n=20+21) RegexpMatchEasy0_1K-8 7.36GB/s ± 0% 7.34GB/s ± 0% -0.26% (p=0.000 n=18+20) RegexpMatchEasy1_32-8 775MB/s ± 0% 761MB/s ± 4% -1.88% (p=0.000 n=17+21) RegexpMatchEasy1_1K-8 4.74GB/s ± 2% 4.74GB/s ± 0% -0.18% (p=0.020 n=21+19) RegexpMatchMedium_32-8 40.6MB/s ± 7% 39.9MB/s ± 9% ~ (p=0.191 n=21+21) RegexpMatchMedium_1K-8 43.7MB/s ± 5% 43.2MB/s ± 5% ~ (p=0.435 n=21+21) RegexpMatchHard_32-8 29.3MB/s ± 1% 29.4MB/s ± 1% +0.53% (p=0.000 n=19+18) RegexpMatchHard_1K-8 31.0MB/s ± 0% 31.0MB/s ± 0% ~ (p=0.572 n=21+20) Revcomp-8 730MB/s ± 0% 720MB/s ± 0% -1.36% (p=0.000 n=17+18) Template-8 46.2MB/s ± 1% 46.3MB/s ± 1% +0.30% (p=0.041 n=20+20) [Geo mean] 204MB/s 205MB/s +0.30% Change-Id: Iac75d0ec184a515ce0e65e19559d5fe2e9840514 Reviewed-on: https://go-review.googlesource.com/c/go/+/354970 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Trust: Josh Bleecher Snyder <josharian@gmail.com> Trust: Keith Randall <khr@golang.org> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Go Bot <gobot@golang.org> |
||
Alejandro García Montoro
|
e1c294a56d |
cmd/compile: eliminate successive swaps
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> |
||
Ruslan Andreev
|
810b08b8ec |
cmd/compile: inline memequal(x, const, sz) for small sizes
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> |
||
nimelehin
|
097a82f54d |
cmd/compile: don't emit unnecessary amd64 extension checks
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> |
||
|
wdvxdr
|
060cd73ab9 |
cmd/compile: use TZCNT instruction for GOAMD64>=v3
on my Intel CoffeeLake CPU: name old time/op new time/op delta TrailingZeros-8 0.68ns ± 1% 0.64ns ± 1% -6.26% (p=0.000 n=10+10) TrailingZeros8-8 0.70ns ± 1% 0.70ns ± 1% ~ (p=0.697 n=10+10) TrailingZeros16-8 0.70ns ± 1% 0.70ns ± 1% +0.57% (p=0.043 n=10+10) TrailingZeros32-8 0.66ns ± 1% 0.64ns ± 1% -3.35% (p=0.000 n=10+10) TrailingZeros64-8 0.68ns ± 1% 0.64ns ± 1% -5.84% (p=0.000 n=9+10) Updates #45453 Change-Id: I228ff2d51df24b1306136f061432f8a12bb1d6fd Reviewed-on: https://go-review.googlesource.com/c/go/+/353249 Trust: Michael Knyszek <mknyszek@google.com> Run-TryBot: Michael Knyszek <mknyszek@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> |
||
|
Lynn Boger
|
b8990ec932 |
test: update test/codegen/noextend.go to work with either ABI on ppc64x
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> |
||
|
Archana R
|
b35c668072 |
cmd/compile: add PPC64-specific inlining for runtime.memmove
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> |
||
|
Joel Sing
|
fe8347b61a |
cmd/compile: optimise immediate operands with constants on riscv64
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> |
||
|
Joel Sing
|
d413908320 |
test/codegen: add shift tests for RISCV64
Add tests for shift by constant, masked shifts and bounded shifts. While here, sort tests by architecture and keep order of tests consistent (lsh, rshU, rsh). Change-Id: I512d64196f34df9cb2884e8c0f6adcf9dd88b0fc Reviewed-on: https://go-review.googlesource.com/c/go/+/351289 Trust: Joel Sing <joel@sing.id.au> Reviewed-by: Michael Munday <mike.munday@lowrisc.org> Reviewed-by: Cherry Mui <cherryyz@google.com> Run-TryBot: Cherry Mui <cherryyz@google.com> |
||
|
Matthew Dempsky
|
04572fa29b |
cmd/compile: use BMI1 instructions for GOAMD64=v3 and higher
BMI1 includes four instructions (ANDN, BLSI, BLSMSK, BLSR) that are easy to peephole optimize, and which GCC always seems to favor using when available and applicable. Updates #45453. Change-Id: I0274184057058f5c579e5bc3ea9c414396d3cf46 Reviewed-on: https://go-review.googlesource.com/c/go/+/351130 Run-TryBot: Matthew Dempsky <mdempsky@google.com> Trust: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> |
||
|
Keith Randall
|
6f35430faa |
cmd/compile: allow rotates to be merged with logical ops on arm64
Fixes #48002 Change-Id: Ie3a157d55b291f5ac2ef4845e6ce4fefd84fc642 Reviewed-on: https://go-review.googlesource.com/c/go/+/350912 Trust: Keith Randall <khr@golang.org> Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com> |
||
|
Keith Randall
|
315dbd10c9 |
cmd/compile: fold double negate on arm64
Fixes #48467 Change-Id: I52305dbf561ee3eee6c1f053e555a3a6ec1ab892 Reviewed-on: https://go-review.googlesource.com/c/go/+/350910 Trust: Keith Randall <khr@golang.org> Trust: Josh Bleecher Snyder <josharian@gmail.com> Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> |
||
|
Keith Randall
|
83b36ffb10 |
cmd/compile: implement constant rotates on arm64
Explicit constant rotates work, but constant arguments to bits.RotateLeft* needed the additional rule. Fixes #48465 Change-Id: Ia7544f21d0e7587b6b6506f72421459cd769aea6 Reviewed-on: https://go-review.googlesource.com/c/go/+/350909 Trust: Keith Randall <khr@golang.org> Trust: Josh Bleecher Snyder <josharian@gmail.com> Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> |
||
Michael Munday
|
c69f5c0d76 |
cmd/compile: add support for Abs and Copysign intrinsics on riscv64
Also, add the FABSS and FABSD pseudo instructions to the assembler. The compiler could use FSGNJX[SD] directly but there doesn't seem to be much advantage to doing so and the pseudo instructions are easier to understand. Change-Id: Ie8825b8aa8773c69cc4f07a32ef04abf4061d80d Reviewed-on: https://go-review.googlesource.com/c/go/+/348989 Trust: Michael Munday <mike.munday@lowrisc.org> Run-TryBot: Michael Munday <mike.munday@lowrisc.org> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Joel Sing <joel@sing.id.au> |
||
|
fanzha02
|
2091bd3f26 |
cmd/compile: simiplify arm64 bitfield optimizations
In some rewrite rules for arm64 bitfield optimizations, the
bitfield lsb value and the bitfield width value are related
to datasize, some of them use datasize directly to check the
bitfield lsb value is valid, to get the bitfiled width value,
but some of them call isARM64BFMask() and arm64BFWidth()
functions. In order to be consistent, this patch changes them
all to use datasize.
Besides, this patch sorts the codegen test cases.
Run the "toolstash-check -all" command and find one inconsistent code
is as the following.
new: src/math/fma.go:104 BEQ 247
master: src/math/fma.go:104 BEQ 248
The above inconsistence is due to this patch changing the range of the
field lsb value in "UBFIZ" optimization rules from "lc+(32|16|8)<64" to
"lc<64", so that the following code is generated as "UBFIZ". The logical
of changed code is still correct.
The code of src/math/fma.go:160:
const uvinf = 0x7FF0000000000000
func FMA(a, b uint32) float64 {
ps := a+b
return Float64frombits(uint64(ps)<<63 | uvinf)
}
The new assembly code:
TEXT "".FMA(SB), LEAF|NOFRAME|ABIInternal, $0-16
MOVWU "".a(FP), R0
MOVWU "".b+4(FP), R1
ADD R1, R0, R0
UBFIZ $63, R0, $1, R0
ORR $9218868437227405312, R0, R0
MOVD R0, "".~r2+8(FP)
RET (R30)
The master assembly code:
TEXT "".FMA(SB), LEAF|NOFRAME|ABIInternal, $0-16
MOVWU "".a(FP), R0
MOVWU "".b+4(FP), R1
ADD R1, R0, R0
MOVWU R0, R0
LSL $63, R0, R0
ORR $9218868437227405312, R0, R0
MOVD R0, "".~r2+8(FP)
RET (R30)
Change-Id: I9061104adfdfd3384d0525327ae1e5c8b0df5c35
Reviewed-on: https://go-review.googlesource.com/c/go/+/265038
Trust: fannie zhang <Fannie.Zhang@arm.com>
Run-TryBot: fannie zhang <Fannie.Zhang@arm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
|
||
|
Michael Munday
|
8214257347 |
test/codegen: fix package name for test case
The codegen tests are currently skipped (see #48247). The test added in CL 346050 did not compile because it was in the main package but did not contain a main function. Changing the package to 'codegen' fixes the issue. Updates #48247. Change-Id: I0a0eaca8e6a7d7b335606d2c76a204ac0c12e6d5 Reviewed-on: https://go-review.googlesource.com/c/go/+/348392 Trust: Michael Munday <mike.munday@lowrisc.org> Run-TryBot: Michael Munday <mike.munday@lowrisc.org> Reviewed-by: Cherry Mui <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> |
||
|
Michael Munday
|
1da64686f8 |
test/codegen: fix compilation of bitfield tests
The codegen tests are currently skipped (see #48247) and the bitfield tests do not actually compile due to a duplicate function name (sbfiz5) added in CL 267602. Renaming the function fixes the issue. Updates #48247. Change-Id: I626fd5ef13732dc358e73ace9ddcc4cbb6ae5b21 Reviewed-on: https://go-review.googlesource.com/c/go/+/348391 Trust: Michael Munday <mike.munday@lowrisc.org> Run-TryBot: Michael Munday <mike.munday@lowrisc.org> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com> |
||
|
Michael Munday
|
fdc2072420 |
test/codegen: remove broken riscv64 test
This test is not executed by default (see #48247) and does not actually pass. It was added in CL 346689. The code generation changes made in that CL only change how instructions are assembled, they do not actually affect the output of the compiler. This test is unfortunately therefore invalid and will never pass. Updates #48247. Change-Id: I0c807e4a111336e5a097fe4e3af2805f9932a87f Reviewed-on: https://go-review.googlesource.com/c/go/+/348390 Trust: Michael Munday <mike.munday@lowrisc.org> Run-TryBot: Michael Munday <mike.munday@lowrisc.org> Reviewed-by: Cherry Mui <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> |
||
|
wdvxdr
|
21de6bc463 |
cmd/compile: simplify less with non-negative number and constant 0 or 1
The most common cases: len(s) > 0 len(s) < 1 and they can be simplified to: len(s) != 0 len(s) == 0 Fixes #48054 Change-Id: I16e5b0cffcfab62a4acc2a09977a6cd3543dd000 Reviewed-on: https://go-review.googlesource.com/c/go/+/346050 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com> |
||
|
fanzha02
|
7b69ddc171 |
cmd/compile: merge sign extension and shift into SBFIZ
This patch adds some rules to rewrite "(LeftShift (SignExtend x) lc)" expression as "SBFIZ". Add the test cases. Change-Id: I294c4ba09712eeb02c7a952447bb006780f1e60d Reviewed-on: https://go-review.googlesource.com/c/go/+/267602 Trust: fannie zhang <Fannie.Zhang@arm.com> Trust: Cherry Mui <cherryyz@google.com> Run-TryBot: fannie zhang <Fannie.Zhang@arm.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> |
||
|
fanzha02
|
43b05173a2 |
cmd/compile: merge zero/sign extensions with UBFX/SBFX on arm64
The UBFX and SBFX already zero/sign extend the result. Further zero/sign extensions are thus unnecessary as long as they leave the top bits unaltered. This patch absorbs zero/sign extensions into UBFX/SBFX. Add the related test cases. Change-Id: I7c4516c8b52d677f77bf3aaedab87c4a28056ec0 Reviewed-on: https://go-review.googlesource.com/c/go/+/265039 Trust: fannie zhang <Fannie.Zhang@arm.com> Trust: Keith Randall <khr@golang.org> Run-TryBot: fannie zhang <Fannie.Zhang@arm.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com> |
||
Ben Shi
|
37d4532867 |
cmd/internal/obj/riscv: simplify addition with constant
This CL simplifies riscv addition (add r, imm) to
(ADDI (ADDI r, imm/2), imm-imm/2) if imm is in specific ranges.
(-4096 <= imm <= -2049 or 2048 <= imm <= 4094)
There is little impact to the go1 benchmark, while the total
size of pkg/linux_riscv64 decreased by about 11KB.
Change-Id: I236eb8af3b83bb35ce9c0b318fc1d235e8ab9a4e
GitHub-Last-Rev:
|
||
|
Michael Munday
|
ea51e223c2 |
cmd/{asm,compile}: add fused multiply-add support on riscv64
Add support to the assembler for F[N]M{ADD,SUB}[SD] instructions.
Argument order is:
OP RS1, RS2, RS3, RD
Also, add support for the FMA intrinsic to the compiler. Automatic
FMA matching is left to a future CL.
Change-Id: I47166c7393b2ab6bfc2e42aa8c1a8997c3a071b3
Reviewed-on: https://go-review.googlesource.com/c/go/+/293030
Trust: Michael Munday <mike.munday@lowrisc.org>
Run-TryBot: Michael Munday <mike.munday@lowrisc.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Joel Sing <joel@sing.id.au>
|
||
|
Jake Ciolek
|
6cf1d5d0fa |
cmd/compile: generic SSA rules for simplifying 2 and 3 operand integer arithmetic expressions
This applies the following generic integer addition/subtraction transformations: x - (x + y) = -y (x - y) - x = -y y + (x - y) = x y + (z + (x - y) = x + z There's over 40 unique functions matching in Go. Hits 2 funcs in the runtime itself: runtime.stackfree() runtime.runqdrain() Go binary size reduced by 0.05% on Linux x86_64. StackCopy bench (perflocked Cascade Lake x86): name old time/op new time/op delta StackCopyPtr-8 87.3ms ± 1% 86.9ms ± 0% -0.45% (p=0.000 n=20+20) StackCopy-8 77.6ms ± 1% 77.0ms ± 0% -0.76% (p=0.000 n=20+20) StackCopyNoCache-8 2.28ms ± 2% 2.26ms ± 2% -0.93% (p=0.008 n=19+20) test/bench/go1 benchmarks (perflocked Cascade Lake x86): name old time/op new time/op delta BinaryTree17-8 1.88s ± 1% 1.88s ± 0% ~ (p=0.373 n=15+12) Fannkuch11-8 2.31s ± 0% 2.35s ± 0% +1.52% (p=0.000 n=15+14) FmtFprintfEmpty-8 26.6ns ± 0% 26.6ns ± 0% ~ (p=0.081 n=14+13) FmtFprintfString-8 48.6ns ± 0% 50.0ns ± 0% +2.86% (p=0.000 n=15+14) FmtFprintfInt-8 56.9ns ± 0% 54.8ns ± 0% -3.70% (p=0.000 n=15+15) FmtFprintfIntInt-8 90.4ns ± 0% 88.8ns ± 0% -1.78% (p=0.000 n=15+15) FmtFprintfPrefixedInt-8 104ns ± 0% 104ns ± 0% ~ (p=0.905 n=14+13) FmtFprintfFloat-8 148ns ± 0% 144ns ± 0% -2.19% (p=0.000 n=14+15) FmtManyArgs-8 389ns ± 0% 390ns ± 0% +0.35% (p=0.000 n=12+15) GobDecode-8 3.90ms ± 1% 3.88ms ± 0% -0.49% (p=0.000 n=15+14) GobEncode-8 2.73ms ± 0% 2.73ms ± 0% ~ (p=0.425 n=15+14) Gzip-8 169ms ± 0% 168ms ± 0% -0.52% (p=0.000 n=13+13) Gunzip-8 24.7ms ± 0% 24.8ms ± 0% +0.61% (p=0.000 n=15+15) HTTPClientServer-8 60.5µs ± 6% 60.4µs ± 7% ~ (p=0.595 n=15+15) JSONEncode-8 6.97ms ± 1% 6.93ms ± 0% -0.69% (p=0.000 n=14+14) JSONDecode-8 31.2ms ± 1% 30.8ms ± 1% -1.27% (p=0.000 n=14+14) Mandelbrot200-8 3.87ms ± 0% 3.87ms ± 0% ~ (p=0.652 n=15+14) GoParse-8 2.65ms ± 2% 2.64ms ± 1% ~ (p=0.202 n=15+15) RegexpMatchEasy0_32-8 45.1ns ± 0% 45.9ns ± 0% +1.68% (p=0.000 n=14+15) RegexpMatchEasy0_1K-8 140ns ± 0% 139ns ± 0% -0.44% (p=0.000 n=15+14) RegexpMatchEasy1_32-8 40.9ns ± 3% 40.5ns ± 0% -0.88% (p=0.000 n=15+13) RegexpMatchEasy1_1K-8 215ns ± 1% 220ns ± 1% +2.27% (p=0.000 n=15+15) RegexpMatchMedium_32-8 783ns ± 7% 738ns ± 0% ~ (p=0.361 n=15+15) RegexpMatchMedium_1K-8 24.1µs ± 6% 23.4µs ± 6% -2.94% (p=0.004 n=15+15) RegexpMatchHard_32-8 1.10µs ± 1% 1.09µs ± 1% -0.40% (p=0.006 n=15+14) RegexpMatchHard_1K-8 33.0µs ± 0% 33.0µs ± 0% ~ (p=0.535 n=12+14) Revcomp-8 354ms ± 0% 353ms ± 0% -0.23% (p=0.002 n=15+13) Template-8 42.0ms ± 1% 41.8ms ± 2% -0.37% (p=0.023 n=14+15) TimeParse-8 181ns ± 0% 180ns ± 1% -0.18% (p=0.014 n=12+13) TimeFormat-8 240ns ± 0% 242ns ± 1% +0.69% (p=0.000 n=12+15) [Geo mean] 35.2µs 35.1µs -0.43% name old speed new speed delta GobDecode-8 197MB/s ± 1% 198MB/s ± 0% +0.49% (p=0.000 n=15+14) GobEncode-8 281MB/s ± 0% 281MB/s ± 0% ~ (p=0.419 n=15+14) Gzip-8 115MB/s ± 0% 115MB/s ± 0% +0.52% (p=0.000 n=13+13) Gunzip-8 786MB/s ± 0% 781MB/s ± 0% -0.60% (p=0.000 n=15+15) JSONEncode-8 278MB/s ± 1% 280MB/s ± 0% +0.69% (p=0.000 n=14+14) JSONDecode-8 62.3MB/s ± 1% 63.1MB/s ± 1% +1.29% (p=0.000 n=14+14) GoParse-8 21.9MB/s ± 2% 22.0MB/s ± 1% ~ (p=0.205 n=15+15) RegexpMatchEasy0_32-8 709MB/s ± 0% 697MB/s ± 0% -1.65% (p=0.000 n=14+15) RegexpMatchEasy0_1K-8 7.34GB/s ± 0% 7.37GB/s ± 0% +0.43% (p=0.000 n=15+15) RegexpMatchEasy1_32-8 783MB/s ± 2% 790MB/s ± 0% +0.88% (p=0.000 n=15+13) RegexpMatchEasy1_1K-8 4.77GB/s ± 1% 4.66GB/s ± 1% -2.23% (p=0.000 n=15+15) RegexpMatchMedium_32-8 41.0MB/s ± 7% 43.3MB/s ± 0% ~ (p=0.360 n=15+15) RegexpMatchMedium_1K-8 42.5MB/s ± 6% 43.8MB/s ± 6% +3.07% (p=0.004 n=15+15) RegexpMatchHard_32-8 29.2MB/s ± 1% 29.3MB/s ± 1% +0.41% (p=0.006 n=15+14) RegexpMatchHard_1K-8 31.1MB/s ± 0% 31.1MB/s ± 0% ~ (p=0.495 n=12+14) Revcomp-8 718MB/s ± 0% 720MB/s ± 0% +0.23% (p=0.002 n=15+13) Template-8 46.3MB/s ± 1% 46.4MB/s ± 2% +0.38% (p=0.021 n=14+15) [Geo mean] 205MB/s 206MB/s +0.57% Change-Id: Ibd1afdf8b6c0b08087dcc3acd8f943637eb95ac0 Reviewed-on: https://go-review.googlesource.com/c/go/+/344930 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Go Bot <gobot@golang.org> Trust: Josh Bleecher Snyder <josharian@gmail.com> |
||
Meng Zhuo
|
efd206eb40 |
cmd/compile: intrinsify Mul64 on riscv64
According to RISCV instruction set manual v2.2 Sec 6.1 MULHU followed by MUL will be fused into one multiply by microarchitecture Benchstat on Hifive unmatched: name old time/op new time/op delta Hash8Bytes 245ns ± 3% 186ns ± 4% -23.99% (p=0.000 n=10+10) Hash320Bytes 1.94µs ± 1% 1.31µs ± 1% -32.38% (p=0.000 n=9+10) Hash1K 5.84µs ± 0% 3.84µs ± 0% -34.20% (p=0.000 n=10+9) Hash8K 45.3µs ± 0% 29.4µs ± 0% -35.04% (p=0.000 n=10+10) name old speed new speed delta Hash8Bytes 32.7MB/s ± 3% 43.0MB/s ± 4% +31.61% (p=0.000 n=10+10) Hash320Bytes 165MB/s ± 1% 244MB/s ± 1% +47.88% (p=0.000 n=9+10) Hash1K 175MB/s ± 0% 266MB/s ± 0% +51.98% (p=0.000 n=10+9) Hash8K 181MB/s ± 0% 279MB/s ± 0% +53.94% (p=0.000 n=10+10) Change-Id: I3561495d02a4a0ad8578e9b9819bf0a4eaca5d12 Reviewed-on: https://go-review.googlesource.com/c/go/+/329970 Reviewed-by: Joel Sing <joel@sing.id.au> Run-TryBot: Joel Sing <joel@sing.id.au> TryBot-Result: Go Bot <gobot@golang.org> Trust: Meng Zhuo <mzh@golangcn.org> |
||
|
Cherry Mui
|
6b1e4430bb |
[dev.typeparams] cmd/compile: implement clobberdead mode on ARM64
For debugging. Change-Id: I5875ccd2413b8ffd2ec97a0ace66b5cae7893b24 Reviewed-on: https://go-review.googlesource.com/c/go/+/324765 Trust: Cherry Mui <cherryyz@google.com> Run-TryBot: Cherry Mui <cherryyz@google.com> Reviewed-by: Than McIntosh <thanm@google.com> TryBot-Result: Go Bot <gobot@golang.org> |
||
|
Cherry Mui
|
165d39a1d4 |
[dev.typeparams] test: adjust codegen test for register ABI on ARM64
In codegen/arithmetic.go, previously there are MOVD's that match for loads of arguments. With register ABI there are no more such loads. Remove the MOVD matches. Change-Id: I920ee2629c8c04d454f13a0c08e283d3528d9a64 Reviewed-on: https://go-review.googlesource.com/c/go/+/324251 Trust: Cherry Mui <cherryyz@google.com> Run-TryBot: Cherry Mui <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Than McIntosh <thanm@google.com> |
||
Ruslan Andreev
|
3b321a9d12 |
cmd/compile: add arch-specific inlining for runtime.memmove
This CL add runtime.memmove inlining for AMD64 and ARM64. According to ssa dump from testcases generic rules can't inline memmomve properly due to one of the arguments is Phi operation. But this Phi op will be optimized out by later optimization stages. As a result memmove can be inlined during arch-specific rules. The commit add new optimization rules to arch-specific rules that can inline runtime.memmove if it possible during lowering stage. Optimization fires 5 times in Go source-code using regabi. Fixes #41662 Change-Id: Iaffaf4c482d068b5f0683d141863892202cc8824 Reviewed-on: https://go-review.googlesource.com/c/go/+/289151 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Go Bot <gobot@golang.org> Trust: David Chase <drchase@google.com> |
||
|
Keith Randall
|
b211fe0058 |
cmd/compile: remove bit operations that modify memory directly
These operations (BT{S,R,C}{Q,L}modify) are quite a bit slower than
other ways of doing the same thing.
Without the BTxmodify operations, there are two fallback ways the compiler
performs these operations: AND/OR/XOR operations directly on memory, or
load-BTx-write sequences. The compiler kinda chooses one arbitrarily
depending on rewrite rule application order. Currently, it uses
load-BTx-write for the Const benchmarks and AND/OR/XOR directly to memory
for the non-Const benchmarks. TBD, someone might investigate which of
the two fallback strategies is really better. For now, they are both
better than BTx ops.
name old time/op new time/op delta
BitSet-8 1.09µs ± 2% 0.64µs ± 5% -41.60% (p=0.000 n=9+10)
BitClear-8 1.15µs ± 3% 0.68µs ± 6% -41.00% (p=0.000 n=10+10)
BitToggle-8 1.18µs ± 4% 0.73µs ± 2% -38.36% (p=0.000 n=10+8)
BitSetConst-8 37.0ns ± 7% 25.8ns ± 2% -30.24% (p=0.000 n=10+10)
BitClearConst-8 30.7ns ± 2% 25.0ns ±12% -18.46% (p=0.000 n=10+10)
BitToggleConst-8 36.9ns ± 1% 23.8ns ± 3% -35.46% (p=0.000 n=9+10)
Fixes #45790
Update #45242
Change-Id: Ie33a72dc139f261af82db15d446cd0855afb4e59
Reviewed-on: https://go-review.googlesource.com/c/go/+/318149
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ben Shi <powerman1st@163.com>
|
||
|
Cherry Zhang
|
6b8e3e2d06 |
cmd/compile: reduce redundant register moves for regabi calls
Currently, if we have AX=a and BX=b, and we want to make a call
F(1, a, b), to move arguments into the desired registers it emits
MOVQ AX, CX
MOVL $1, AX // AX=1
MOVQ BX, DX
MOVQ CX, BX // BX=a
MOVQ DX, CX // CX=b
This has a few redundant moves.
This is because we process inputs in order. First, allocate 1 to
AX, which kicks out a (in AX) to CX (a free register at the
moment). Then, allocate a to BX, which kicks out b (in BX) to DX.
Finally, put b to CX.
Notice that if we start with allocating CX=b, then BX=a, AX=1,
we will not have redundant moves. This CL reduces redundant moves
by allocating them in different order: First, for inpouts that are
already in place, keep them there. Then allocate free registers.
Then everything else.
before after
cmd/compile binary size 23703888 23609680
text size 8565899 8533291
(with regabiargs enabled.)
Change-Id: I69e1bdf745f2c90bb791f6d7c45b37384af1e874
Reviewed-on: https://go-review.googlesource.com/c/go/+/311371
Trust: Cherry Zhang <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
|
||
|
David Chase
|
4df3d0e4df |
cmd/compile: rescue stmt boundaries from OpArgXXXReg and OpSelectN.
Fixes this failure:
go test cmd/compile/internal/ssa -run TestStmtLines -v
=== RUN TestStmtLines
stmtlines_test.go:115: Saw too many (amd64, > 1%) lines without
statement marks, total=88263, nostmt=1930
('-run TestStmtLines -v' lists failing lines)
The failure has two causes.
One is that the first-line adjuster in code generation was relocating
"first lines" to instructions that would either not have any code generated,
or would have the statment marker removed by a different believed-good heuristic.
The other was that statement boundaries were getting attached to register
values (that with the old ABI were loads from the stack, hence real instructions).
The register values disappear at code generation.
The fixes are to (1) note that certain instructions are not good choices for
"first value" and skip them, and (2) in an expandCalls post-pass, look for
register valued instructions and under appropriate conditions move their
statement marker to a compatible use.
Also updates TestStmtLines to always log the score, for easier comparison of
minor compiler changes.
Updates #40724.
Change-Id: I485573ce900e292d7c44574adb7629cdb4695c3f
Reviewed-on: https://go-review.googlesource.com/c/go/+/309649
Trust: David Chase <drchase@google.com>
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
|
||
|
Cherry Zhang
|
444d28295b |
test: make codegen/memops.go work with both ABIs
Following CL 309335, this fixes memops.go. Change-Id: Ia2458b5267deee9f906f76c50e70a021ea2fcb5b Reviewed-on: https://go-review.googlesource.com/c/go/+/309552 Trust: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Than McIntosh <thanm@google.com> |
||
|
Cherry Zhang
|
263e13d1f7 |
test: make codegen tests work with both ABIs
Some codegen tests were written with the assumption that arguments and results are in memory, and with a specific stack layout. With the register ABI, the assumption is no longer true. Adjust the tests to work with both cases. - For tests expecting in memory arguments/results, change to use global variables or memory-assigned argument/results. - Allow more registers. E.g. some tests expecting register names contain only letters (e.g. AX), but it can also contain numbers (e.g. R10). - Some instruction selection changes when operate on register vs. memory, e.g. ADDQ vs. LEAQ, MOVB vs. MOVL. Accept both. TODO: mathbits.go and memops.go still need fix. Change-Id: Ic5932b4b5dd3f5d30ed078d296476b641420c4c5 Reviewed-on: https://go-review.googlesource.com/c/go/+/309335 Trust: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> |
||
Pat Gavlin
|
359f44910f |
cmd/compile: fix long RMW bit operations on AMD64
Under certain circumstances, the existing rules for bit operations can
produce code that writes beyond its intended bounds. For example,
consider the following code:
func repro(b []byte, addr, bit int32) {
_ = b[3]
v := uint32(b[0]) | uint32(b[1])<<8 | uint32(b[2])<<16 | uint32(b[3])<<24 | 1<<(bit&31)
b[0] = byte(v)
b[1] = byte(v >> 8)
b[2] = byte(v >> 16)
b[3] = byte(v >> 24)
}
Roughly speaking:
1. The expression `1 << (bit & 31)` is rewritten into `(SHLL 1 bit)`
2. The expression `uint32(b[0]) | uint32(b[1])<<8 | uint32(b[2])<<16 |
uint32(b[3])<<24` is rewritten into `(MOVLload &b[0])`
3. The statements `b[0] = byte(v) ... b[3] = byte(v >> 24)` are
rewritten into `(MOVLstore &b[0], v)`
4. `(ORL (SHLL 1, bit) (MOVLload &b[0]))` is rewritten into
`(BTSL (MOVLload &b[0]) bit)`. This is a valid transformation because
the destination is a register: in this case, the bit offset is masked
by the number of bits in the destination register. This is identical
to the masking performed by `SHL`.
5. `(MOVLstore &b[0] (BTSL (MOVLload &b[0]) bit))` is rewritten into
`(BTSLmodify &b[0] bit)`. This is an invalid transformation because
the destination is memory: in this case, the bit offset is not
masked, and the chosen instruction may write outside its intended
32-bit location.
These changes fix the invalid rewrite performed in step (5) by
explicitly maksing the bit offset operand to `BT(S|R|C)(L|Q)modify`. In
the example above, the adjusted rules produce
`(BTSLmodify &b[0] (ANDLconst [31] bit))` in step (5).
These changes also add several new rules to rewrite bit sets, toggles,
and clears that are rooted at `(OR|XOR|AND)(L|Q)modify` operators into
appropriate `BT(S|R|C)(L|Q)modify` operators. These rules catch cases
where `MOV(L|Q)store ((OR|XOR|AND)(L|Q) ...)` is rewritten to
`(OR|XOR|AND)(L|Q)modify` before the `(OR|XOR|AND)(L|Q) ...` can be
rewritten to `BT(S|R|C)(L|Q) ...`.
Overall, compilecmp reports small improvements in code size on
darwin/amd64 when the changes to the compiler itself are exlcuded:
file before after Δ %
runtime.s 536464 536412 -52 -0.010%
bytes.s 32629 32593 -36 -0.110%
strings.s 44565 44529 -36 -0.081%
os/signal.s 7967 7959 -8 -0.100%
cmd/vendor/golang.org/x/sys/unix.s 81686 81678 -8 -0.010%
math/big.s 188235 188253 +18 +0.010%
cmd/link/internal/loader.s 89295 89056 -239 -0.268%
cmd/link/internal/ld.s 633551 633232 -319 -0.050%
cmd/link/internal/arm.s 18934 18928 -6 -0.032%
cmd/link/internal/arm64.s 31814 31801 -13 -0.041%
cmd/link/internal/riscv64.s 7347 7345 -2 -0.027%
cmd/compile/internal/ssa.s 4029173 4033066 +3893 +0.097%
total 21298280 21301472 +3192 +0.015%
Change-Id: I2e560548b515865129e1724e150e30540e9d29ce
GitHub-Last-Rev:
|
||
|
fanzha02
|
3a0061822e |
cmd/compile: add arm64 rules to optimize go codes to constant 0
Optimize the following codes to constant 0.
function shift (x uint32) uint64 {
return uint64(x) >> 32
}
Change-Id: Ida6b39d713cc119ad5a2f01fd54bfd252cf2c975
Reviewed-on: https://go-review.googlesource.com/c/go/+/303830
Trust: fannie zhang <Fannie.Zhang@arm.com>
Run-TryBot: fannie zhang <Fannie.Zhang@arm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
|
||
|
fanzha02
|
b182ba7fab |
cmd/compile: optimize codes with arm64 REV16 instruction
Optimize some patterns into rev16/rev16w instruction.
Pattern1:
(c & 0xff00ff00)>>8 | (c & 0x00ff00ff)<<8
To:
rev16w c
Pattern2:
(c & 0xff00ff00ff00ff00)>>8 | (c & 0x00ff00ff00ff00ff)<<8
To:
rev16 c
This patch is a copy of CL 239637, contributed by Alice Xu(dianhong.xu@arm.com).
Change-Id: I96936c1db87618bc1903c04221c7e9b2779455b3
Reviewed-on: https://go-review.googlesource.com/c/go/+/268377
Trust: fannie zhang <Fannie.Zhang@arm.com>
Run-TryBot: fannie zhang <Fannie.Zhang@arm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
|
||
|
Cherry Zhang
|
6ae3b70ef2 |
cmd/compile: add clobberdeadreg mode
When -clobberdeadreg flag is set, the compiler inserts code that clobbers integer registers at call sites. This may be helpful for debugging register ABI. Only implemented on AMD64 for now. Change-Id: Ia203d3f891c30fd95d0103489056fe01d63a2899 Reviewed-on: https://go-review.googlesource.com/c/go/+/302809 Trust: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> |
||
|
fanzha02
|
f5e6d3e879 |
cmd/compile: add rewrite rules for conditional instructions on arm64
This CL adds rewrite rules for CSETM, CSINC, CSINV, and CSNEG. By adding
these rules, we can save one instruction.
For example,
func test(cond bool, a int) int {
if cond {
a++
}
return a
}
Before:
MOVD "".a+8(RSP), R0
ADD $1, R0, R1
MOVBU "".cond(RSP), R2
CMPW $0, R2
CSEL NE, R1, R0, R0
After:
MOVBU "".cond(RSP), R0
CMPW $0, R0
MOVD "".a+8(RSP), R0
CSINC EQ, R0, R0, R0
This patch is a copy of CL 285694. Co-authored-by: JunchenLi
<junchen.li@arm.com>
Change-Id: Ic1a79e8b8ece409b533becfcb7950f11e7b76f24
Reviewed-on: https://go-review.googlesource.com/c/go/+/302231
Trust: fannie zhang <Fannie.Zhang@arm.com>
Run-TryBot: fannie zhang <Fannie.Zhang@arm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
|
||
|
Cherry Zhang
|
8628bf9a97 |
cmd/compile: resurrect clobberdead mode
This CL resurrects the clobberdead debugging mode (CL 23924). When -clobberdead flag is set (TODO: make it GOEXPERIMENT?), the compiler inserts code that clobbers all dead stack slots that contains pointers. Mark windows syscall functions cgo_unsafe_args, as the code actually does that, by taking the address of one argument and passing it to cgocall. Change-Id: Ie09a015f4bd14ae6053cc707866e30ae509b9d6f Reviewed-on: https://go-review.googlesource.com/c/go/+/301791 Trust: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Than McIntosh <thanm@google.com> |
||
|
Meng Zhuo
|
afe517590c |
cmd/compile: loads from readonly globals into const for mips64x
Ref: CL 141118 Update #26498 Change-Id: If4ea55c080b9aa10183eefe81fefbd4072deaf3a Reviewed-on: https://go-review.googlesource.com/c/go/+/280646 Trust: Meng Zhuo <mzh@golangcn.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com> |
||
|
erifan01
|
600259b099 |
cmd/compile: use depth first topological sort algorithm for layout
The current layout algorithm tries to put consecutive blocks together,
so the priority of the successor block is higher than the priority of
the zero indegree block. This algorithm is beneficial for subsequent
register allocation, but will result in more branch instructions.
The depth-first topological sorting algorithm is a well-known layout
algorithm, which has applications in many languages, and it helps to
reduce branch instructions. This CL applies it to the layout pass.
The test results show that it helps to reduce the code size.
This CL also includes the following changes:
1, Removed the primary predecessor mechanism. The new layout algorithm is
not very friendly to register allocator in some cases, in order to adapt
to the new layout algorithm, a new primary predecessor selection strategy
is introduced.
2, Since the new layout implementation may place non-loop blocks between
loop blocks, some adaptive modifications have also been made to looprotate
pass.
3, The layout also affects the results of codegen, so this CL also adjusted
several codegen tests accordingly.
It is inevitable that this CL will cause the code size or performance of a
few functions to decrease, but the number of cases it improves is much larger
than the number of cases it drops.
Statistical data from compilecmp on linux/amd64 is as follow:
name old time/op new time/op delta
Template 382ms ± 4% 382ms ± 4% ~ (p=0.497 n=49+50)
Unicode 170ms ± 9% 169ms ± 8% ~ (p=0.344 n=48+50)
GoTypes 2.01s ± 4% 2.01s ± 4% ~ (p=0.628 n=50+48)
Compiler 190ms ±10% 189ms ± 9% ~ (p=0.734 n=50+50)
SSA 11.8s ± 2% 11.8s ± 3% ~ (p=0.877 n=50+50)
Flate 241ms ± 9% 241ms ± 8% ~ (p=0.897 n=50+49)
GoParser 366ms ± 3% 361ms ± 4% -1.21% (p=0.004 n=47+50)
Reflect 835ms ± 3% 838ms ± 3% ~ (p=0.275 n=50+49)
Tar 336ms ± 4% 335ms ± 3% ~ (p=0.454 n=48+48)
XML 433ms ± 4% 431ms ± 3% ~ (p=0.071 n=49+48)
LinkCompiler 706ms ± 4% 705ms ± 4% ~ (p=0.608 n=50+49)
ExternalLinkCompiler 1.85s ± 3% 1.83s ± 2% -1.47% (p=0.000 n=49+48)
LinkWithoutDebugCompiler 437ms ± 5% 437ms ± 6% ~ (p=0.953 n=49+50)
[Geo mean] 615ms 613ms -0.37%
name old alloc/op new alloc/op delta
Template 38.7MB ± 1% 38.7MB ± 1% ~ (p=0.834 n=50+50)
Unicode 28.1MB ± 0% 28.1MB ± 0% -0.22% (p=0.000 n=49+50)
GoTypes 168MB ± 1% 168MB ± 1% ~ (p=0.054 n=47+47)
Compiler 23.0MB ± 1% 23.0MB ± 1% ~ (p=0.432 n=50+50)
SSA 1.54GB ± 0% 1.54GB ± 0% +0.21% (p=0.000 n=50+50)
Flate 23.6MB ± 1% 23.6MB ± 1% ~ (p=0.153 n=43+46)
GoParser 35.1MB ± 1% 35.1MB ± 2% ~ (p=0.202 n=50+50)
Reflect 84.7MB ± 1% 84.7MB ± 1% ~ (p=0.333 n=48+49)
Tar 34.5MB ± 1% 34.5MB ± 1% ~ (p=0.406 n=46+49)
XML 44.3MB ± 2% 44.2MB ± 3% ~ (p=0.981 n=50+50)
LinkCompiler 131MB ± 0% 128MB ± 0% -2.74% (p=0.000 n=50+50)
ExternalLinkCompiler 120MB ± 0% 120MB ± 0% +0.01% (p=0.007 n=50+50)
LinkWithoutDebugCompiler 77.3MB ± 0% 77.3MB ± 0% -0.02% (p=0.000 n=50+50)
[Geo mean] 69.3MB 69.1MB -0.22%
file before after Δ %
addr2line 4104220 4043684 -60536 -1.475%
api 5342502 5249678 -92824 -1.737%
asm 4973785 4858257 -115528 -2.323%
buildid 2667844 2625660 -42184 -1.581%
cgo 4686849 4616313 -70536 -1.505%
compile 23667431 23268406 -399025 -1.686%
cover 4959676 4874108 -85568 -1.725%
dist 3515934 3450422 -65512 -1.863%
doc 3995581 3925469 -70112 -1.755%
fix 3379202 3318522 -60680 -1.796%
link 6743249 6629913 -113336 -1.681%
nm 4047529 3991777 -55752 -1.377%
objdump 4456151 4388151 -68000 -1.526%
pack 2435040 2398072 -36968 -1.518%
pprof 13804080 13565808 -238272 -1.726%
test2json 2690043 2645987 -44056 -1.638%
trace 10418492 10232716 -185776 -1.783%
vet 7258259 7121259 -137000 -1.888%
total 113145867 111204202 -1941665 -1.716%
The situation on linux/arm64 is as follow:
name old time/op new time/op delta
Template 280ms ± 1% 282ms ± 1% +0.75% (p=0.000 n=46+48)
Unicode 124ms ± 2% 124ms ± 2% +0.37% (p=0.045 n=50+50)
GoTypes 1.69s ± 1% 1.70s ± 1% +0.56% (p=0.000 n=49+50)
Compiler 122ms ± 1% 123ms ± 1% +0.93% (p=0.000 n=50+50)
SSA 12.6s ± 1% 12.7s ± 0% +0.72% (p=0.000 n=50+50)
Flate 170ms ± 1% 172ms ± 1% +0.97% (p=0.000 n=49+49)
GoParser 262ms ± 1% 263ms ± 1% +0.39% (p=0.000 n=49+48)
Reflect 639ms ± 1% 650ms ± 1% +1.63% (p=0.000 n=49+49)
Tar 243ms ± 1% 245ms ± 1% +0.82% (p=0.000 n=50+50)
XML 324ms ± 1% 327ms ± 1% +0.72% (p=0.000 n=50+49)
LinkCompiler 597ms ± 1% 596ms ± 1% -0.27% (p=0.001 n=48+47)
ExternalLinkCompiler 1.90s ± 1% 1.88s ± 1% -1.00% (p=0.000 n=50+50)
LinkWithoutDebugCompiler 364ms ± 1% 363ms ± 1% ~ (p=0.220 n=49+50)
[Geo mean] 485ms 488ms +0.49%
name old alloc/op new alloc/op delta
Template 38.7MB ± 0% 38.8MB ± 1% ~ (p=0.093 n=43+49)
Unicode 28.4MB ± 0% 28.4MB ± 0% +0.03% (p=0.000 n=49+45)
GoTypes 169MB ± 1% 169MB ± 1% +0.23% (p=0.010 n=50+50)
Compiler 23.2MB ± 1% 23.2MB ± 1% +0.11% (p=0.000 n=40+44)
SSA 1.54GB ± 0% 1.55GB ± 0% +0.45% (p=0.000 n=47+49)
Flate 23.8MB ± 2% 23.8MB ± 1% ~ (p=0.543 n=50+50)
GoParser 35.3MB ± 1% 35.4MB ± 1% ~ (p=0.792 n=50+50)
Reflect 85.2MB ± 1% 85.2MB ± 0% ~ (p=0.055 n=50+47)
Tar 34.5MB ± 1% 34.5MB ± 1% +0.06% (p=0.015 n=50+50)
XML 43.8MB ± 2% 43.9MB ± 2% +0.19% (p=0.000 n=48+48)
LinkCompiler 137MB ± 0% 136MB ± 0% -0.92% (p=0.000 n=50+50)
ExternalLinkCompiler 127MB ± 0% 127MB ± 0% ~ (p=0.516 n=50+50)
LinkWithoutDebugCompiler 84.0MB ± 0% 84.0MB ± 0% ~ (p=0.057 n=50+50)
[Geo mean] 70.4MB 70.4MB +0.01%
file before after Δ %
addr2line 4021557 4002933 -18624 -0.463%
api 5127847 5028503 -99344 -1.937%
asm 5034716 4936836 -97880 -1.944%
buildid 2608118 2594094 -14024 -0.538%
cgo 4488592 4398320 -90272 -2.011%
compile 22501129 22213592 -287537 -1.278%
cover 4742301 4713573 -28728 -0.606%
dist 3388071 3365311 -22760 -0.672%
doc 3802250 3776082 -26168 -0.688%
fix 3306147 3216939 -89208 -2.698%
link 6404483 6363699 -40784 -0.637%
nm 3941026 3921930 -19096 -0.485%
objdump 4383330
|
||
Josh Bleecher Snyder
|
43d5f213e2 |
cmd/compile: optimize multi-register shifts on amd64
amd64 can shift in bits from another register instead of filling with 0/1. This pattern is helpful when implementing 128 bit shifts or arbitrary length shifts. In the standard library, it shows up in pure Go math/big. Benchmarks results on amd64 with -tags=math_big_pure_go. name old time/op new time/op delta NonZeroShifts/1/shrVU-8 4.45ns ± 3% 4.39ns ± 1% -1.28% (p=0.000 n=30+27) NonZeroShifts/1/shlVU-8 4.13ns ± 4% 4.10ns ± 2% ~ (p=0.254 n=29+28) NonZeroShifts/2/shrVU-8 5.55ns ± 1% 5.63ns ± 2% +1.42% (p=0.000 n=28+29) NonZeroShifts/2/shlVU-8 5.70ns ± 2% 5.14ns ± 1% -9.82% (p=0.000 n=29+28) NonZeroShifts/3/shrVU-8 6.79ns ± 2% 6.35ns ± 2% -6.46% (p=0.000 n=28+29) NonZeroShifts/3/shlVU-8 6.69ns ± 1% 6.25ns ± 1% -6.60% (p=0.000 n=28+27) NonZeroShifts/4/shrVU-8 7.79ns ± 2% 7.06ns ± 2% -9.48% (p=0.000 n=30+30) NonZeroShifts/4/shlVU-8 7.82ns ± 1% 7.24ns ± 1% -7.37% (p=0.000 n=28+29) NonZeroShifts/5/shrVU-8 8.90ns ± 3% 7.93ns ± 1% -10.84% (p=0.000 n=29+26) NonZeroShifts/5/shlVU-8 8.68ns ± 1% 7.92ns ± 1% -8.76% (p=0.000 n=29+29) NonZeroShifts/10/shrVU-8 14.4ns ± 1% 12.3ns ± 2% -14.79% (p=0.000 n=28+29) NonZeroShifts/10/shlVU-8 14.1ns ± 1% 11.9ns ± 2% -15.55% (p=0.000 n=28+27) NonZeroShifts/100/shrVU-8 118ns ± 1% 96ns ± 3% -18.82% (p=0.000 n=30+29) NonZeroShifts/100/shlVU-8 120ns ± 2% 98ns ± 2% -18.46% (p=0.000 n=29+28) NonZeroShifts/1000/shrVU-8 1.10µs ± 1% 0.88µs ± 2% -19.63% (p=0.000 n=29+30) NonZeroShifts/1000/shlVU-8 1.10µs ± 2% 0.88µs ± 2% -20.28% (p=0.000 n=29+28) NonZeroShifts/10000/shrVU-8 10.9µs ± 1% 8.7µs ± 1% -19.78% (p=0.000 n=28+27) NonZeroShifts/10000/shlVU-8 10.9µs ± 2% 8.7µs ± 1% -19.64% (p=0.000 n=29+27) NonZeroShifts/100000/shrVU-8 111µs ± 2% 90µs ± 2% -19.39% (p=0.000 n=28+29) NonZeroShifts/100000/shlVU-8 113µs ± 2% 90µs ± 2% -20.43% (p=0.000 n=30+27) The assembly version is still faster, unfortunately, but the gap is narrowing. Speedup from pure Go to assembly: name old time/op new time/op delta NonZeroShifts/1/shrVU-8 4.39ns ± 1% 3.45ns ± 2% -21.36% (p=0.000 n=27+29) NonZeroShifts/1/shlVU-8 4.10ns ± 2% 3.47ns ± 3% -15.42% (p=0.000 n=28+30) NonZeroShifts/2/shrVU-8 5.63ns ± 2% 3.97ns ± 0% -29.40% (p=0.000 n=29+25) NonZeroShifts/2/shlVU-8 5.14ns ± 1% 3.77ns ± 2% -26.65% (p=0.000 n=28+26) NonZeroShifts/3/shrVU-8 6.35ns ± 2% 4.79ns ± 2% -24.52% (p=0.000 n=29+29) NonZeroShifts/3/shlVU-8 6.25ns ± 1% 4.42ns ± 1% -29.29% (p=0.000 n=27+26) NonZeroShifts/4/shrVU-8 7.06ns ± 2% 5.64ns ± 1% -20.05% (p=0.000 n=30+29) NonZeroShifts/4/shlVU-8 7.24ns ± 1% 5.34ns ± 2% -26.23% (p=0.000 n=29+29) NonZeroShifts/5/shrVU-8 7.93ns ± 1% 6.56ns ± 2% -17.26% (p=0.000 n=26+30) NonZeroShifts/5/shlVU-8 7.92ns ± 1% 6.27ns ± 1% -20.79% (p=0.000 n=29+25) NonZeroShifts/10/shrVU-8 12.3ns ± 2% 10.2ns ± 2% -17.21% (p=0.000 n=29+29) NonZeroShifts/10/shlVU-8 11.9ns ± 2% 10.5ns ± 2% -12.45% (p=0.000 n=27+29) NonZeroShifts/100/shrVU-8 95.9ns ± 3% 77.7ns ± 1% -19.00% (p=0.000 n=29+30) NonZeroShifts/100/shlVU-8 97.5ns ± 2% 66.8ns ± 2% -31.47% (p=0.000 n=28+30) NonZeroShifts/1000/shrVU-8 884ns ± 2% 705ns ± 1% -20.17% (p=0.000 n=30+28) NonZeroShifts/1000/shlVU-8 880ns ± 2% 590ns ± 1% -32.96% (p=0.000 n=28+25) NonZeroShifts/10000/shrVU-8 8.74µs ± 1% 7.34µs ± 3% -15.94% (p=0.000 n=27+30) NonZeroShifts/10000/shlVU-8 8.73µs ± 1% 6.00µs ± 1% -31.25% (p=0.000 n=27+28) NonZeroShifts/100000/shrVU-8 89.6µs ± 2% 75.5µs ± 2% -15.80% (p=0.000 n=29+29) NonZeroShifts/100000/shlVU-8 89.6µs ± 2% 68.0µs ± 3% -24.09% (p=0.000 n=27+30) Change-Id: I18f58d8f5513d737d9cdf09b8f9d14011ffe3958 Reviewed-on: https://go-review.googlesource.com/c/go/+/297050 Trust: Josh Bleecher Snyder <josharian@gmail.com> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> |
||
|
Paul E. Murphy
|
48ddf70128 |
cmd/asm,cmd/compile: support 5 operand RLWNM/RLWMI on ppc64
These instructions are actually 5 argument opcodes as specified by the ISA. Prior to this patch, the MB and ME arguments were merged into a single bitmask operand to workaround the limitations of the ppc64 assembler backend. This limitation no longer exists. Thus, we can pass operands for these opcodes without having to merge the MB and ME arguments in the assembler frontend or compiler backend. Likewise, support for 4 operand variants is unchanged. Change-Id: Ib086774f3581edeaadfd2190d652aaaa8a90daeb Reviewed-on: https://go-review.googlesource.com/c/go/+/298750 Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com> Reviewed-by: Carlos Eduardo Seo <carlos.seo@linaro.org> Trust: Carlos Eduardo Seo <carlos.seo@linaro.org> |
||
|
fanzha02
|
2b50ab2aee |
cmd/compile: optimize single-precision floating point square root
Add generic rule to rewrite the single-precision square root expression with one single-precision instruction. The optimization will reduce two times of precision converting between double-precision and single-precision. On arm64 flatform. previous: FCVTSD F0, F0 FSQRTD F0, F0 FCVTDS F0, F0 optimized: FSQRTS S0, S0 And this patch adds the test case to check the correctness. This patch refers to CL 241877, contributed by Alice Xu (dianhong.xu@arm.com) Change-Id: I6de5d02281c693017ac4bd4c10963dd55989bd7e Reviewed-on: https://go-review.googlesource.com/c/go/+/276873 Trust: fannie zhang <Fannie.Zhang@arm.com> Run-TryBot: fannie zhang <Fannie.Zhang@arm.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> |
||
Egon Elbre
|
3ee32439b5 |
cmd/compile: ARM64 optimize []float64 and []float32 access
Optimize load and store to []float64 and []float32.
Previously it used LSL instead of shifted register indexed load/store.
Before:
LSL $3, R0, R0
FMOVD F0, (R1)(R0)
After:
FMOVD F0, (R1)(R0<<3)
Fixes #42798
Change-Id: I0c0912140c3dce5aa6abc27097c0eb93833cc589
Reviewed-on: https://go-review.googlesource.com/c/go/+/273706
Reviewed-by: Cherry Zhang <cherryyz@google.com>
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Giovanni Bajo <rasky@develer.com>
|
||
|
Alejandro García Montoro
|
bf48163e8f |
cmd/compile: add rule to coalesce writes
The code generated when storing eight bytes loaded from memory created a series of small writes instead of a single, large one. The specific pattern of instructions generated stored 1 byte, then 2 bytes, then 4 bytes, and finally 1 byte. The new rules match this specific pattern both for amd64 and for s390x, and convert it into a single instruction to store the 8 bytes. arm64 and ppc64le already generated the right code, but the new codegen test covers also those architectures. Fixes #41663 Change-Id: Ifb9b464be2d59c2ed5034acf7b9c3e473f344030 Reviewed-on: https://go-review.googlesource.com/c/go/+/280456 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Trust: Josh Bleecher Snyder <josharian@gmail.com> Trust: Jason A. Donenfeld <Jason@zx2c4.com> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Go Bot <gobot@golang.org> |
||
|
Keith Randall
|
42cd40ee74 |
cmd/compile: improve bit test code
Some bit test instruction generation stopped triggering after the change to addressing modes. I suspect this was just because ANDQload was being generated before the rewrite rules could discover the BTQ. Fix that by decomposing the ANDQload when it is surrounded by a TESTQ (thus re-enabling the BTQ rules). Fixes #44228 Change-Id: I489b4a5a7eb01c65fc8db0753f8cec54097cadb2 Reviewed-on: https://go-review.googlesource.com/c/go/+/291749 Trust: Keith Randall <khr@golang.org> Trust: Josh Bleecher Snyder <josharian@gmail.com> Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> |
||
|
Cherry Zhang
|
401d7e5a24 |
[dev.regabi] cmd/compile: reserve X15 as zero register on AMD64
In ABIInternal, reserve X15 as constant zero, and use it to zero memory. (Maybe there can be more use of it?) The register is zeroed when transition to ABIInternal from ABI0. Caveat: using X15 generates longer instructions than using X0. Maybe we want to use X0? Change-Id: I12d5ee92a01fc0b59dad4e5ab023ac71bc2a8b7d Reviewed-on: https://go-review.googlesource.com/c/go/+/288093 Trust: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> |
||
|
David Chase
|
9a19481acb |
[dev.regabi] cmd/compile: make ordering for InvertFlags more stable
Current many architectures use a rule along the lines of // Canonicalize the order of arguments to comparisons - helps with CSE. ((CMP|CMPW) x y) && x.ID > y.ID => (InvertFlags ((CMP|CMPW) y x)) to normalize comparisons as much as possible for CSE. Replace the ID comparison with something less variable across compiler changes. This helps avoid spurious failures in some of the codegen-comparison tests (though the current choice of comparison is sensitive to Op ordering). Two tests changed to accommodate modified instruction choice. Change-Id: Ib35f450bd2bae9d4f9f7838ceaf7ec682bcf1e1a Reviewed-on: https://go-review.googlesource.com/c/go/+/280155 Trust: David Chase <drchase@google.com> Run-TryBot: David Chase <drchase@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com> |
||
|
Lynn Boger
|
0ae3b7cb74 |
cmd/compile: fix rules regression with shifts on PPC64
Some rules for PPC64 were checking for a case where a shift followed by an 'and' of a mask could be lowered, depending on the format of the mask. The function to verify if the mask was valid for this purpose was not checking if the mask was 0 which we don't want to allow. This case can happen if previous optimizations resulted in that mask value. This fixes isPPC64ValidShiftMask to check for a mask of 0 and return false. This also adds a codegen testcase to verify it doesn't try to match the rules in the future. Fixes #42610 Change-Id: I565d94e88495f51321ab365d6388c01e791b4dbb Reviewed-on: https://go-review.googlesource.com/c/go/+/270358 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Paul Murphy <murp@ibm.com> Reviewed-by: Carlos Eduardo Seo <carlos.seo@linaro.org> Trust: Lynn Boger <laboger@linux.vnet.ibm.com> |
||
Alberto Donizetti
|
7307e86afd |
test/codegen: go fmt
Fixes #42445 Change-Id: I9653ef094dba2a1ac2e3daaa98279d10df17a2a1 Reviewed-on: https://go-review.googlesource.com/c/go/+/268257 Trust: Alberto Donizetti <alb.donizetti@gmail.com> Trust: Martin Möhrmann <moehrmann@google.com> Run-TryBot: Alberto Donizetti <alb.donizetti@gmail.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com> |
||
Michael Munday
|
854e892ce1 |
cmd/compile: optimize shift pairs and masks on s390x
Optimize combinations of left and right shifts by a constant value into a 'rotate then insert selected bits [into zero]' instruction. Use the same instruction for contiguous masks since it has some benefits over 'and immediate' (not restricted to 32-bits, does not overwrite source register). To keep the complexity of this change under control I've only implemented 64 bit operations for now. There are a lot more optimizations that can be done with this instruction family. However, since their function overlaps with other instructions we need to be somewhat careful not to break existing optimization rules by creating optimization dead ends. This is particularly true of the load/store merging rules which contain lots of zero extensions and shifts. This CL does interfere with the store merging rules when an operand is shifted left before it is stored: binary.BigEndian.PutUint64(b, x << 1) This is unfortunate but it's not critical and somewhat complex so I plan to fix that in a follow up CL. file before after Δ % addr2line 4117446 4117282 -164 -0.004% api 4945184 4942752 -2432 -0.049% asm 4998079 4991891 -6188 -0.124% buildid 2685158 2684074 -1084 -0.040% cgo 4553732 4553394 -338 -0.007% compile 19294446 19245070 -49376 -0.256% cover 4897105 4891319 -5786 -0.118% dist 3544389 3542785 -1604 -0.045% doc 3926795 3927617 +822 +0.021% fix 3302958 3293868 -9090 -0.275% link 6546274 6543456 -2818 -0.043% nm 4102021 4100825 -1196 -0.029% objdump 4542431 4548483 +6052 +0.133% pack 2482465 2416389 -66076 -2.662% pprof 13366541 13363915 -2626 -0.020% test2json 2829007 2761515 -67492 -2.386% trace 10216164 10219684 +3520 +0.034% vet 6773956 6773572 -384 -0.006% total 107124151 106917891 -206260 -0.193% Change-Id: I7591cce41e06867ba10a745daae9333513062746 Reviewed-on: https://go-review.googlesource.com/c/go/+/233317 Run-TryBot: Michael Munday <mike.munday@ibm.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Trust: Michael Munday <mike.munday@ibm.com> |
||
|
Cherry Zhang
|
0387bedadf |
cmd/compile: remove racefuncenterfp when it is not needed
We already remove racefuncenter and racefuncexit if they are not needed (i.e. the function doesn't have any other race calls). racefuncenterfp is like racefuncenter but used on LR machines. Remove unnecessary racefuncenterfp as well. Change-Id: I65edb00e19c6d9ab55a204cbbb93e9fb710559f1 Reviewed-on: https://go-review.googlesource.com/c/go/+/267099 Trust: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> |
||
|
Paul E. Murphy
|
c3c6fbf314 |
cmd/compile: combine more 32 bit shift and mask operations on ppc64
Combine (AND m (SRWconst x)) or (SRWconst (AND m x)) when mask m is and the shift value produce constant which can be encoded into an RLWINM instruction. Combine (CLRLSLDI (SRWconst x)) if the combining of the underling rotate masks produces a constant which can be encoded into RLWINM. Likewise for (SLDconst (SRWconst x)) and (CLRLSDI (RLWINM x)). Combine rotate word + and operations which can be encoded as a single RLWINM/RLWNM instruction. The most notable performance improvements arise from the crypto benchmarks below (GOARCH=power8 on a ppc64le/linux): pkg:golang.org/x/crypto/blowfish goos:linux goarch:ppc64le ExpandKeyWithSalt 52.2µs ± 0% 47.5µs ± 0% -8.88% ExpandKey 44.4µs ± 0% 40.3µs ± 0% -9.15% pkg:golang.org/x/crypto/ssh/internal/bcrypt_pbkdf goos:linux goarch:ppc64le Key 57.6ms ± 0% 52.3ms ± 0% -9.13% pkg:golang.org/x/crypto/bcrypt goos:linux goarch:ppc64le Equal 90.9ms ± 0% 82.6ms ± 0% -9.13% DefaultCost 91.0ms ± 0% 82.7ms ± 0% -9.12% Change-Id: I59a0ca29face38f4ab46e37124c32906f216c4ce Reviewed-on: https://go-review.googlesource.com/c/go/+/260798 Run-TryBot: Carlos Eduardo Seo <carlos.seo@linaro.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com> Reviewed-by: Carlos Eduardo Seo <carlos.seo@linaro.com> Trust: Lynn Boger <laboger@linux.vnet.ibm.com> |
||
|
Keith Randall
|
04b8a9fea5 |
all: implement GO386=softfloat
Backstop support for non-sse2 chips now that 387 is gone. RELNOTE=yes Change-Id: Ib10e69c4a3654c15a03568f93393437e1939e013 Reviewed-on: https://go-review.googlesource.com/c/go/+/260017 Trust: Keith Randall <khr@golang.org> Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org> |
||
|
Keith Randall
|
fe2cfb74ba |
all: drop 387 support
My last 387 CL. So sad ... ... ... ... not! Fixes #40255 Change-Id: I8d4ddb744b234b8adc735db2f7c3c7b6d8bbdfa4 Reviewed-on: https://go-review.googlesource.com/c/go/+/258957 Trust: Keith Randall <khr@golang.org> Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com> |
||
|
Lynn Boger
|
cc2a5cf4b8 |
cmd/compile,cmd/internal/obj/ppc64: fix some shift rules due to a regression
A recent change to improve shifts was generating some invalid cases when the rule was based on an AND. The extended mnemonics CLRLSLDI and CLRLSLWI only allow certain values for the operands and in the mask case those values were not being checked properly. This adds a check to those rules to verify that the 'b' and 'n' values used when an AND was part of the rule have correct values. There was a bug in some diag messages in asm9. The message expected 3 values but only provided 2. Those are corrected here also. The test/codegen/shift.go was updated to add a few more cases to check for the case mentioned here. Some of the comments that mention the order of operands in these extended mnemonics were wrong and those have been corrected. Fixes #41683. Change-Id: If5bb860acaa5051b9e0cd80784b2868b85898c31 Reviewed-on: https://go-review.googlesource.com/c/go/+/258138 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> Reviewed-by: Paul Murphy <murp@ibm.com> Reviewed-by: Carlos Eduardo Seo <carlos.seo@gmail.com> TryBot-Result: Go Bot <gobot@golang.org> Trust: Lynn Boger <laboger@linux.vnet.ibm.com> |
||
|
Lynn Boger
|
a424f6e45e |
cmd/asm,cmd/compile,cmd/internal/obj/ppc64: add extswsli support on power9
This adds support for the extswsli instruction which combines extsw followed by a shift. New benchmark demonstrates the improvement: name old time/op new time/op delta ExtShift 1.34µs ± 0% 1.30µs ± 0% -3.15% (p=0.057 n=4+3) Change-Id: I21b410676fdf15d20e0cbbaa75d7c6dcd3bbb7b0 Reviewed-on: https://go-review.googlesource.com/c/go/+/257017 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Carlos Eduardo Seo <carlos.seo@gmail.com> Trust: Lynn Boger <laboger@linux.vnet.ibm.com> |
||
|
Lynn Boger
|
967465da29 |
cmd/compile: use combined shifts to improve array addressing on ppc64x
This change adds rules to find pairs of instructions that can
be combined into a single shifts. These instruction sequences
are common in array addressing within loops. Improvements can
be seen in many crypto packages and the hash packages.
These are based on the extended mnemonics found in the ISA
sections C.8.1 and C.8.2.
Some rules in PPC64.rules were moved because the ordering prevented
some matching.
The following results were generated on power9.
hash/crc32:
CRC32/poly=Koopman/size=40/align=0 195ns ± 0% 163ns ± 0% -16.41%
CRC32/poly=Koopman/size=40/align=1 200ns ± 0% 163ns ± 0% -18.50%
CRC32/poly=Koopman/size=512/align=0 1.98µs ± 0% 1.67µs ± 0% -15.46%
CRC32/poly=Koopman/size=512/align=1 1.98µs ± 0% 1.69µs ± 0% -14.80%
CRC32/poly=Koopman/size=1kB/align=0 3.90µs ± 0% 3.31µs ± 0% -15.27%
CRC32/poly=Koopman/size=1kB/align=1 3.85µs ± 0% 3.31µs ± 0% -14.15%
CRC32/poly=Koopman/size=4kB/align=0 15.3µs ± 0% 13.1µs ± 0% -14.22%
CRC32/poly=Koopman/size=4kB/align=1 15.4µs ± 0% 13.1µs ± 0% -14.79%
CRC32/poly=Koopman/size=32kB/align=0 137µs ± 0% 105µs ± 0% -23.56%
CRC32/poly=Koopman/size=32kB/align=1 137µs ± 0% 105µs ± 0% -23.53%
crypto/rc4:
RC4_128 733ns ± 0% 650ns ± 0% -11.32% (p=1.000 n=1+1)
RC4_1K 5.80µs ± 0% 5.17µs ± 0% -10.89% (p=1.000 n=1+1)
RC4_8K 45.7µs ± 0% 40.8µs ± 0% -10.73% (p=1.000 n=1+1)
crypto/sha1:
Hash8Bytes 635ns ± 0% 613ns ± 0% -3.46% (p=1.000 n=1+1)
Hash320Bytes 2.30µs ± 0% 2.18µs ± 0% -5.38% (p=1.000 n=1+1)
Hash1K 5.88µs ± 0% 5.38µs ± 0% -8.62% (p=1.000 n=1+1)
Hash8K 42.0µs ± 0% 37.9µs ± 0% -9.75% (p=1.000 n=1+1)
There are other improvements found in golang.org/x/crypto which are all in the
range of 5-15%.
Change-Id: I193471fbcf674151ffe2edab212799d9b08dfb8c
Reviewed-on: https://go-review.googlesource.com/c/go/+/252097
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: Carlos Eduardo Seo <cseo@linux.vnet.ibm.com>
|
||
|
Cuong Manh Le
|
27a30186ab |
cmd/compile,runtime: skip zero'ing order array for select statements
The order array was zero initialized by the compiler, but ends up being overwritten by the runtime anyway. So let the runtime takes full responsibility for initializing, save us one instruction per select. Fixes #40399 Change-Id: Iec1eca27ad7180d4fcb3cc9ef97348206b7fe6b8 Reviewed-on: https://go-review.googlesource.com/c/go/+/251517 Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com> |
||
|
Paul E. Murphy
|
7615b20d06 |
cmd/compile: generate subfic on ppc64
This merges an lis + subf into subfic, and for 32b constants lwa + subf into oris + ori + subf. The carry bit is no longer used in code generation, therefore I think we can clobber it as needed. Note, lowered borrow/carry arithmetic is self-contained and thus is not affected. A few extra rules are added to ensure early transformations to SUBFCconst don't trip up earlier rules, fold constant operations, or otherwise simplify lowering. Likewise, tests are added to ensure all rules are hit. Generic constant folding catches trivial cases, however some lowering rules insert arithmetic which can introduce new opportunities (e.g BitLen or Slicemask). I couldn't find a specific benchmark to demonstrate noteworthy improvements, but this is generating subfic in many of the default bent test binaries, so we are at least saving a little code space. Change-Id: Iad7c6e5767eaa9dc24dc1c989bd1c8cfe1982012 Reviewed-on: https://go-review.googlesource.com/c/go/+/249461 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Carlos Eduardo Seo <cseo@linux.vnet.ibm.com> |
||
|
fanzha02
|
d556c251a1 |
cmd/compile: add more generic rewrite rules to reassociate (op (op y C) x|C)
With this patch, opt pass can expose more obvious constant-folding
opportunites.
Example:
func test(i int) int {return (i+8)-(i+4)}
The previous version:
MOVD "".i(FP), R0
ADD $8, R0, R1
ADD $4, R0, R0
SUB R0, R1, R0
MOVD R0, "".~r1+8(FP)
RET (R30)
The optimized version:
MOVD $4, R0
MOVD R0, "".~r1+8(FP)
RET (R30)
This patch removes some existing reassociation rules, such as "x+(z-C)",
because the current generic rewrite rules will canonicalize "x-const"
to "x+(-const)", making "x+(z-C)" equal to "x+(z+(-C))".
This patch also adds test cases.
Change-Id: I857108ba0b5fcc18a879eeab38e2551bc4277797
Reviewed-on: https://go-review.googlesource.com/c/go/+/237137
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
|
||
Agniva De Sarker
|
8acbe4c0b3 |
cmd/compile: optimize unsigned comparisons with 0/1 on wasm
Updates #21439 Change-Id: I0fbcde6e0c2fc368fe686b271670f9d8be4a7900 Reviewed-on: https://go-review.googlesource.com/c/go/+/247557 Run-TryBot: Agniva De Sarker <agniva.quicksilver@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Richard Musiol <neelance@gmail.com> |
||
diaxu01
|
01aad9ea93 |
cmd/compile: Optimize ARM64's code with EON
This patch fuses pattern '(MVN (XOR x y))' into '(EON x y)'. Change-Id: I269c98ce198d51a4945ce8bd0e1024acbd1b7609 Reviewed-on: https://go-review.googlesource.com/c/go/+/239638 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com> |
||
|
Paul E. Murphy
|
e7c7ce646f |
cmd/compile: combine multiply/add into maddld on ppc64le/power9
Add a new lowering rule to match and replace such instances with the MADDLD instruction available on power9 where possible. Likewise, this plumbs in a new ppc64 ssa opcode to house the newly generated MADDLD instructions. When testing ed25519, this reduced binary size by 936B. Similarly, MADDLD combination occcurs in a few other less obvious cases such as division by constant. Testing of golang.org/x/crypto/ed25519 shows non-trivial speedup during keygeneration: name old time/op new time/op delta KeyGeneration 65.2µs ± 0% 63.1µs ± 0% -3.19% Signing 64.3µs ± 0% 64.4µs ± 0% +0.16% Verification 147µs ± 0% 147µs ± 0% +0.11% Similarly, this test binary has shrunk by 66488B. Change-Id: I077aeda7943119b41f07e4e62e44a648f16e4ad0 Reviewed-on: https://go-review.googlesource.com/c/go/+/248723 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com> |
||
|
Michael Munday
|
6e876f1985 |
cmd/compile: clean up and optimize s390x multiplication rules
Some of the existing optimizations aren't triggered because they are handled by the generic rules so this CL removes them. Also some constraints were copied without much thought from the amd64 rules and they don't make sense on s390x, so we remove those constraints. Finally, add a 'multiply by the sum of two powers of two' optimization. This makes sense on s390x as shifts are low latency and can also sometimes be optimized further (especially if we add support for RISBG instructions). name old time/op new time/op delta IntMulByConst/3-8 1.70ns ±11% 1.10ns ± 5% -35.26% (p=0.000 n=10+10) IntMulByConst/5-8 1.64ns ± 7% 1.10ns ± 4% -32.94% (p=0.000 n=10+9) IntMulByConst/12-8 1.65ns ± 6% 1.20ns ± 4% -27.16% (p=0.000 n=10+9) IntMulByConst/120-8 1.66ns ± 4% 1.22ns ±13% -26.43% (p=0.000 n=10+10) IntMulByConst/-120-8 1.65ns ± 7% 1.19ns ± 4% -28.06% (p=0.000 n=9+10) IntMulByConst/65537-8 0.86ns ± 9% 1.12ns ±12% +30.41% (p=0.000 n=10+10) IntMulByConst/65538-8 1.65ns ± 5% 1.23ns ± 5% -25.11% (p=0.000 n=10+10) Change-Id: Ib196e6bff1e97febfd266134d0a2b2a62897989f Reviewed-on: https://go-review.googlesource.com/c/go/+/248937 Run-TryBot: Michael Munday <mike.munday@ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> |
||
Junchen Li
|
06337823ef |
cmd/compile: optimize unsigned comparisons to 0/1 on arm64
For an unsigned integer, it's useful to convert its order test with 0/1
to its equality test with 0. We can save a comparison instruction that
followed by a conditional branch on arm64 since it supports
compare-with-zero-and-branch instructions. For example,
if x > 0 { ... } else { ... }
the original version:
CMP $0, R0
BLS 9
the optimized version:
CBZ R0, 8
Updates #21439
Change-Id: Id1de6f865f6aa72c5d45b29f7894818857288425
Reviewed-on: https://go-review.googlesource.com/c/go/+/246857
Reviewed-by: Keith Randall <khr@golang.org>
|
||
|
Keith Randall
|
ef9c8a38ad |
cmd/compile: don't rewrite (CMP (AND x y) 0) to TEST if AND has other uses
If the AND has other uses, we end up saving an argument to the AND in another register, so we can use it for the TEST. No point in doing that. Change-Id: I73444a6aeddd6f55e2328ce04d77c3e6cf4a83e0 Reviewed-on: https://go-review.googlesource.com/c/go/+/241280 Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> |
||
|
Junchen Li
|
e30fbe3757 |
cmd/compile: optimize unsigned comparisons to 0
There are some architecture-independent rules in #21439, since an unsigned integer >= 0 is always true and < 0 is always false. This CL adds these optimizations to generic rules. Updates #21439 Change-Id: Iec7e3040b761ecb1e60908f764815fdd9bc62495 Reviewed-on: https://go-review.googlesource.com/c/go/+/246617 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> |
||
|
Keith Randall
|
c4fed25553 |
cmd/compile: add floating point load+op operations to addressing modes pass
They were missed as part of the refactoring to use a separate addressing modes pass. Fixes #40426 Change-Id: Ie0418b2fac4ba1ffe720644ac918f6d728d5e420 Reviewed-on: https://go-review.googlesource.com/c/go/+/244859 Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com> |
||
Xiangdong Ji
|
e031318ca6 |
cmd/compile: ARM comparisons with 0 incorrect on overflow
Some ARM rewriting rules convert 'comparing to zero' conditions of if statements to a simplified version utilizing CMN and CMP instructions to branch over condition flags, in order to save one Add or Sub caculation. Such optimizations lead to wrong branching in case an overflow/underflow occurs when executing CMN or CMP. Fix the issue by introducing new block opcodes that don't honor the overflow/underflow flag: Block-Op Meaning ARM condition codes 1. LTnoov less than MI 2. GEnoov greater than or equal PL 3. LEnoov less than or equal MI || EQ 4. GTnoov greater than NEQ & PL The patch also adds a few test cases to cover scenarios that are specific to ARM and fine-tunes the code generation tests for 'x-const'. For more details please refer to the previous fix on 64-bit ARM: https://go-review.googlesource.com/c/go/+/233097 Go1 perf, 'old' is the non-optimized version, that is removing all concerned rewriting rules. name old time/op new time/op delta BinaryTree17-8 7.73s ± 0% 7.81s ± 0% +0.97% (p=0.000 n=7+8) Fannkuch11-8 7.06s ± 0% 7.00s ± 0% -0.83% (p=0.000 n=8+8) FmtFprintfEmpty-8 181ns ± 1% 183ns ± 1% +1.31% (p=0.001 n=8+8) FmtFprintfString-8 319ns ± 1% 325ns ± 2% +1.71% (p=0.009 n=7+8) FmtFprintfInt-8 358ns ± 1% 359ns ± 1% ~ (p=0.293 n=7+7) FmtFprintfIntInt-8 459ns ± 3% 456ns ± 1% ~ (p=0.869 n=8+8) FmtFprintfPrefixedInt-8 535ns ± 4% 538ns ± 4% ~ (p=0.572 n=8+8) FmtFprintfFloat-8 1.01µs ± 2% 1.01µs ± 2% ~ (p=0.625 n=8+8) FmtManyArgs-8 1.93µs ± 2% 1.93µs ± 1% ~ (p=0.979 n=8+7) GobDecode-8 16.1ms ± 1% 16.5ms ± 1% +2.32% (p=0.000 n=8+8) GobEncode-8 15.9ms ± 0% 15.8ms ± 1% -1.00% (p=0.000 n=8+7) Gzip-8 690ms ± 1% 670ms ± 0% -2.90% (p=0.000 n=8+8) Gunzip-8 109ms ± 1% 109ms ± 1% ~ (p=0.694 n=7+8) HTTPClientServer-8 149µs ± 3% 146µs ± 2% -1.70% (p=0.028 n=8+8) JSONEncode-8 50.5ms ± 1% 49.2ms ± 0% -2.60% (p=0.001 n=7+7) JSONDecode-8 135ms ± 2% 137ms ± 1% ~ (p=0.054 n=8+7) Mandelbrot200-8 951ms ± 0% 952ms ± 0% ~ (p=0.852 n=6+8) GoParse-8 9.47ms ± 1% 9.66ms ± 1% +2.01% (p=0.000 n=8+8) RegexpMatchEasy0_32-8 288ns ± 2% 277ns ± 2% -3.61% (p=0.000 n=8+8) RegexpMatchEasy0_1K-8 1.66µs ± 1% 1.69µs ± 2% +2.21% (p=0.001 n=7+7) RegexpMatchEasy1_32-8 334ns ± 1% 305ns ± 2% -8.86% (p=0.000 n=8+8) RegexpMatchEasy1_1K-8 2.14µs ± 2% 2.15µs ± 0% ~ (p=0.099 n=8+8) RegexpMatchMedium_32-8 13.3ns ± 1% 13.3ns ± 0% ~ (p=1.000 n=7+7) RegexpMatchMedium_1K-8 81.1µs ± 3% 80.7µs ± 1% ~ (p=0.955 n=7+8) RegexpMatchHard_32-8 4.26µs ± 0% 4.26µs ± 0% ~ (p=0.933 n=7+8) RegexpMatchHard_1K-8 124µs ± 0% 124µs ± 0% +0.31% (p=0.000 n=8+8) Revcomp-8 14.7ms ± 2% 14.5ms ± 1% -1.66% (p=0.003 n=8+8) Template-8 197ms ± 2% 200ms ± 3% +1.62% (p=0.021 n=8+8) TimeParse-8 1.33µs ± 1% 1.30µs ± 1% -1.86% (p=0.002 n=8+8) TimeFormat-8 3.04µs ± 1% 3.02µs ± 0% -0.60% (p=0.000 n=8+8) name old speed new speed delta GobDecode-8 47.6MB/s ± 1% 46.5MB/s ± 1% -2.28% (p=0.000 n=8+8) GobEncode-8 48.1MB/s ± 0% 48.6MB/s ± 1% +1.02% (p=0.000 n=8+7) Gzip-8 28.1MB/s ± 1% 29.0MB/s ± 0% +2.97% (p=0.000 n=8+8) Gunzip-8 178MB/s ± 1% 179MB/s ± 2% ~ (p=0.694 n=7+8) JSONEncode-8 38.4MB/s ± 1% 39.4MB/s ± 0% +2.67% (p=0.001 n=7+7) JSONDecode-8 14.3MB/s ± 2% 14.2MB/s ± 1% -0.81% (p=0.043 n=8+7) GoParse-8 6.12MB/s ± 1% 5.99MB/s ± 1% -2.00% (p=0.000 n=8+8) RegexpMatchEasy0_32-8 111MB/s ± 2% 115MB/s ± 2% +3.77% (p=0.000 n=8+8) RegexpMatchEasy0_1K-8 618MB/s ± 1% 604MB/s ± 2% -2.16% (p=0.001 n=7+7) RegexpMatchEasy1_32-8 95.7MB/s ± 1% 105.1MB/s ± 2% +9.76% (p=0.000 n=8+8) RegexpMatchEasy1_1K-8 479MB/s ± 2% 477MB/s ± 0% ~ (p=0.105 n=8+8) RegexpMatchMedium_32-8 75.2MB/s ± 1% 75.2MB/s ± 0% ~ (p=0.247 n=7+7) RegexpMatchMedium_1K-8 12.6MB/s ± 3% 12.7MB/s ± 1% ~ (p=0.538 n=7+8) RegexpMatchHard_32-8 7.52MB/s ± 0% 7.52MB/s ± 0% ~ (p=0.968 n=7+8) RegexpMatchHard_1K-8 8.26MB/s ± 0% 8.24MB/s ± 0% -0.30% (p=0.001 n=8+8) Revcomp-8 173MB/s ± 2% 176MB/s ± 1% +1.68% (p=0.003 n=8+8) Template-8 9.85MB/s ± 2% 9.69MB/s ± 3% -1.59% (p=0.021 n=8+8) Fixes #39303 Updates #38740 Change-Id: I0a5f87bfda679f66414c0041ace2ca2e28363f36 Reviewed-on: https://go-review.googlesource.com/c/go/+/236637 Run-TryBot: Keith Randall <khr@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> |
||
|
Xiangdong Ji
|
e8f5a33191 |
cmd/compile: fix incorrect rewriting to if condition
Some ARM64 rewriting rules convert 'comparing to zero' conditions of if
statements to a simplified version utilizing CMN and CMP instructions to
branch over condition flags, in order to save one Add or Sub caculation.
Such optimizations lead to wrong branching in case an overflow/underflow
occurs when executing CMN or CMP.
Fix the issue by introducing new block opcodes that don't honor the
overflow/underflow flag, in the following categories:
Block-Op Meaning ARM condition codes
1. LTnoov less than MI
2. GEnoov greater than or equal PL
3. LEnoov less than or equal MI || EQ
4. GTnoov greater than NEQ & PL
The backend generates two consecutive branch instructions for 'LEnoov'
and 'GTnoov' to model their expected behavior. A slight change to 'gc'
and amd64/386 backends is made to unify the code generation.
Add a test 'TestCondRewrite' as justification, it covers 32 incorrect rules
identified on arm64, more might be needed on other arches, like 32-bit arm.
Add two benchmarks profiling the aforementioned category 1&2 and category
3&4 separetely, we expect the first two categories will show performance
improvement and the second will not result in visible regression compared with
the non-optimized version.
This change also updates TestFormats to support using %#x.
Examples exhibiting where does the issue come from:
1: 'if x + 3 < 0' might be converted to:
before:
CMN $3, R0
BGE <else branch> // wrong branch is taken if 'x+3' overflows
after:
CMN $3, R0
BPL <else branch>
2: 'if y - 3 > 0' might be converted to:
before:
CMP $3, R0
BLE <else branch> // wrong branch is taken if 'y-3' underflows
after:
CMP $3, R0
BMI <else branch>
BEQ <else branch>
Benchmark data from different kinds of arm64 servers, 'old' is the non-optimized
version (not the parent commit), generally the optimization version outperforms.
S1:
name old time/op new time/op delta
CondRewrite/SoloJump 13.6ns ± 0% 12.9ns ± 0% -5.15% (p=0.000 n=10+10)
CondRewrite/CombJump 13.8ns ± 1% 12.9ns ± 0% -6.32% (p=0.000 n=10+10)
S2:
name old time/op new time/op delta
CondRewrite/SoloJump 11.6ns ± 0% 10.9ns ± 0% -6.03% (p=0.000 n=10+10)
CondRewrite/CombJump 11.4ns ± 0% 10.8ns ± 1% -5.53% (p=0.000 n=10+10)
S3:
name old time/op new time/op delta
CondRewrite/SoloJump 7.36ns ± 0% 7.50ns ± 0% +1.79% (p=0.000 n=9+10)
CondRewrite/CombJump 7.35ns ± 0% 7.75ns ± 0% +5.51% (p=0.000 n=8+9)
S4:
name old time/op new time/op delta
CondRewrite/SoloJump-224 11.5ns ± 1% 10.9ns ± 0% -4.97% (p=0.000 n=10+10)
CondRewrite/CombJump-224 11.9ns ± 0% 11.5ns ± 0% -2.95% (p=0.000 n=10+10)
S5:
name old time/op new time/op delta
CondRewrite/SoloJump 10.0ns ± 0% 10.0ns ± 0% -0.45% (p=0.000 n=9+10)
CondRewrite/CombJump 9.93ns ± 0% 9.77ns ± 0% -1.53% (p=0.000 n=10+9)
Go1 perf. data:
name old time/op new time/op delta
BinaryTree17 6.29s ± 1% 6.30s ± 1% ~ (p=1.000 n=5+5)
Fannkuch11 5.40s ± 0% 5.40s ± 0% ~ (p=0.841 n=5+5)
FmtFprintfEmpty 97.9ns ± 0% 98.9ns ± 3% ~ (p=0.937 n=4+5)
FmtFprintfString 171ns ± 3% 171ns ± 2% ~ (p=0.754 n=5+5)
FmtFprintfInt 212ns ± 0% 217ns ± 6% +2.55% (p=0.008 n=5+5)
FmtFprintfIntInt 296ns ± 1% 297ns ± 2% ~ (p=0.516 n=5+5)
FmtFprintfPrefixedInt 371ns ± 2% 374ns ± 7% ~ (p=1.000 n=5+5)
FmtFprintfFloat 435ns ± 1% 439ns ± 2% ~ (p=0.056 n=5+5)
FmtManyArgs 1.37µs ± 1% 1.36µs ± 1% ~ (p=0.730 n=5+5)
GobDecode 14.6ms ± 4% 14.4ms ± 4% ~ (p=0.690 n=5+5)
GobEncode 11.8ms ±20% 11.6ms ±15% ~ (p=1.000 n=5+5)
Gzip 507ms ± 0% 491ms ± 0% -3.22% (p=0.008 n=5+5)
Gunzip 73.8ms ± 0% 73.9ms ± 0% ~ (p=0.690 n=5+5)
HTTPClientServer 116µs ± 0% 116µs ± 0% ~ (p=0.686 n=4+4)
JSONEncode 21.8ms ± 1% 21.6ms ± 2% ~ (p=0.151 n=5+5)
JSONDecode 104ms ± 1% 103ms ± 1% -1.08% (p=0.016 n=5+5)
Mandelbrot200 9.53ms ± 0% 9.53ms ± 0% ~ (p=0.421 n=5+5)
GoParse 7.55ms ± 1% 7.51ms ± 1% ~ (p=0.151 n=5+5)
RegexpMatchEasy0_32 158ns ± 0% 158ns ± 0% ~ (all equal)
RegexpMatchEasy0_1K 606ns ± 1% 608ns ± 3% ~ (p=0.937 n=5+5)
RegexpMatchEasy1_32 143ns ± 0% 144ns ± 1% ~ (p=0.095 n=5+4)
RegexpMatchEasy1_1K 927ns ± 2% 944ns ± 2% ~ (p=0.056 n=5+5)
RegexpMatchMedium_32 16.0ns ± 0% 16.0ns ± 0% ~ (all equal)
RegexpMatchMedium_1K 69.3µs ± 2% 69.7µs ± 0% ~ (p=0.690 n=5+5)
RegexpMatchHard_32 3.73µs ± 0% 3.73µs ± 1% ~ (p=0.984 n=5+5)
RegexpMatchHard_1K 111µs ± 1% 110µs ± 0% ~ (p=0.151 n=5+5)
Revcomp 1.91s ±47% 1.77s ±68% ~ (p=1.000 n=5+5)
Template 138ms ± 1% 138ms ± 1% ~ (p=1.000 n=5+5)
TimeParse 787ns ± 2% 785ns ± 1% ~ (p=0.540 n=5+5)
TimeFormat 729ns ± 1% 726ns ± 1% ~ (p=0.151 n=5+5)
Updates #38740
Change-Id: I06c604874acdc1e63e66452dadee5df053045222
Reviewed-on: https://go-review.googlesource.com/c/go/+/233097
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
|
||
|
Keith Randall
|
2cb10d42b7 |
cmd/compile: in prove, zero right shifts of positive int by #bits - 1
Taking over Zach's CL 212277. Just cleaned up and added a test.
For a positive, signed integer, an arithmetic right shift of count
(bit-width - 1) equals zero. e.g. int64(22) >> 63 -> 0. This CL makes
prove replace these right shifts with a zero-valued constant.
These shifts may arise in source code explicitly, but can also be
created by the generic rewrite of signed division by a power of 2.
// Signed divide by power of 2.
// n / c = n >> log(c) if n >= 0
// = (n+c-1) >> log(c) if n < 0
// We conditionally add c-1 by adding n>>63>>(64-log(c))
(first shift signed, second shift unsigned).
(Div64 <t> n (Const64 [c])) && isPowerOfTwo(c) ->
(Rsh64x64
(Add64 <t> n (Rsh64Ux64 <t>
(Rsh64x64 <t> n (Const64 <typ.UInt64> [63]))
(Const64 <typ.UInt64> [64-log2(c)])))
(Const64 <typ.UInt64> [log2(c)]))
If n is known to be positive, this rewrite includes an extra Add and 2
extra Rsh. This CL will allow prove to replace one of the extra Rsh with
a 0. That replacement then allows lateopt to remove all the unneccesary
fixups from the generic rewrite.
There is a rewrite rule to handle this case directly:
(Div64 n (Const64 [c])) && isNonNegative(n) && isPowerOfTwo(c) ->
(Rsh64Ux64 n (Const64 <typ.UInt64> [log2(c)]))
But this implementation of isNonNegative really only handles constants
and a few special operations like len/cap. The division could be
handled if the factsTable version of isNonNegative were available.
Unfortunately, the first opt pass happens before prove even has a
chance to deduce the numerator is non-negative, so the generic rewrite
has already fired and created the extra Ops discussed above.
Fixes #36159
By Printf count, this zeroes 137 right shifts when building std and cmd.
Change-Id: Iab486910ac9d7cfb86ace2835456002732b384a2
Reviewed-on: https://go-review.googlesource.com/c/go/+/232857
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
|
||
Martin Möhrmann
|
6ed4661807 |
cmd/compile: optimize make+copy pattern to avoid memclr
match: m = make([]T, x); copy(m, s) for pointer free T and x==len(s) rewrite to: m = mallocgc(x*elemsize(T), nil, false); memmove(&m, &s, x*elemsize(T)) otherwise rewrite to: m = makeslicecopy([]T, x, s) This avoids memclear and shading of pointers in the newly created slice before the copy. With this CL "s" is only be allowed to bev a variable and not a more complex expression. This restriction could be lifted in future versions of this optimization when it can be proven that "s" is not referencing "m". Triggers 450 times during make.bash.. Reduces go binary size by ~8 kbyte. name old time/op new time/op delta MakeSliceCopy/mallocmove/Byte 71.1ns ± 1% 65.8ns ± 0% -7.49% (p=0.000 n=10+9) MakeSliceCopy/mallocmove/Int 71.2ns ± 1% 66.0ns ± 0% -7.27% (p=0.000 n=10+8) MakeSliceCopy/mallocmove/Ptr 104ns ± 4% 99ns ± 1% -5.13% (p=0.000 n=10+10) MakeSliceCopy/makecopy/Byte 70.3ns ± 0% 68.0ns ± 0% -3.22% (p=0.000 n=10+9) MakeSliceCopy/makecopy/Int 70.3ns ± 0% 68.5ns ± 1% -2.59% (p=0.000 n=9+10) MakeSliceCopy/makecopy/Ptr 102ns ± 0% 99ns ± 1% -2.97% (p=0.000 n=9+9) MakeSliceCopy/nilappend/Byte 75.4ns ± 0% 74.9ns ± 2% -0.63% (p=0.015 n=9+9) MakeSliceCopy/nilappend/Int 75.6ns ± 0% 76.4ns ± 3% ~ (p=0.245 n=9+10) MakeSliceCopy/nilappend/Ptr 107ns ± 0% 108ns ± 1% +0.93% (p=0.005 n=9+10) Fixes #26252 Change-Id: Iec553dd1fef6ded16197216a472351c8799a8e71 Reviewed-on: https://go-review.googlesource.com/c/go/+/146719 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Martin Möhrmann <moehrmann@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> |
||
|
Keith Randall
|
9ed0fb42e3 |
cmd/compile: add indexed memory modification ops to amd64
name old time/op new time/op delta Modify-16 404ns ± 1% 365ns ± 1% -9.73% (p=0.000 n=10+10) ConstModify-16 407ns ± 0% 385ns ± 2% -5.56% (p=0.000 n=9+10) Seems to generally help generated code. Binary size change is in the noise. Change-Id: I57891bfaf0f7dfc5d143bb9f7ebafc7079d2614f Reviewed-on: https://go-review.googlesource.com/c/go/+/228098 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> |