summaryrefslogtreecommitdiff
path: root/src/math
Commit message (Collapse)AuthorAgeFilesLines
* math/bits: add examples for leading zero methodsDylan Waits2017-07-151-0/+38
| | | | | | | | Change-Id: Ib491d144387a7675af370f7b925fe6e62440d153 Reviewed-on: https://go-review.googlesource.com/48966 Run-TryBot: Kevin Burke <kev@inburke.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Kevin Burke <kev@inburke.com>
* math/rand: add concurrency warning to overview commentAditya Mukerjee2017-07-151-1/+2
| | | | | | Change-Id: I52efa7aa72a23256e5ca56470ffeba975ed8f739 Reviewed-on: https://go-review.googlesource.com/48760 Reviewed-by: Bryan Mills <bcmills@google.com>
* math: add a Sqrt exampleMartynas Budriūnas2017-07-151-0/+20
| | | | | | | | Change-Id: I259e25b9d0b069912053a250e9739e04fafca54d Reviewed-on: https://go-review.googlesource.com/48892 Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math: clarify comment about bit-identical results across architecturesIan Lance Taylor2017-06-061-2/+1
| | | | | | | | | Updates #18354. Change-Id: I76bc4a73d8dc99eeda14b395e451d75a65184191 Reviewed-on: https://go-review.googlesource.com/45013 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Rob Pike <r@golang.org>
* math: add doc note about floating point operationgulyasm2017-06-061-0/+3
| | | | | | | | | | | | Go doesn't guarantee that the result of floating point operations will be the same on different architectures. It was not stated in the documentation, that can lead to confusion. Fixes #18354 Change-Id: Idb1b4c256fb9a7158a74256136eca3b8ce44476f Reviewed-on: https://go-review.googlesource.com/34938 Reviewed-by: Ian Lance Taylor <iant@golang.org>
* math/big: fix Add, Sub when receiver aliases 2nd operandAlberto Donizetti2017-05-312-2/+39
| | | | | | | | | | Fixes #20490 Change-Id: I9cfa604f9ff94df779cb9b4cbbd706258fc473ac Reviewed-on: https://go-review.googlesource.com/44150 Run-TryBot: Alberto Donizetti <alb.donizetti@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
* internal/cpu: new package to detect cpu featuresMartin Möhrmann2017-05-102-16/+4
| | | | | | | | | | | | | | | | | Implements detection of x86 cpu features that are used in the go standard library. Changes all standard library packages to use the new cpu package instead of using runtime internal variables to check x86 cpu features. Updates: #15403 Change-Id: I2999a10cb4d9ec4863ffbed72f4e021a1dbc4bb9 Reviewed-on: https://go-review.googlesource.com/41476 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
* cmd/compile: ppc64x intrinsics for math/bitsLynn Boger2017-05-101-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | This adds math/bits intrinsics for OnesCount, Len, TrailingZeros on ppc64x. benchmark old ns/op new ns/op delta BenchmarkLeadingZeros-16 4.26 1.71 -59.86% BenchmarkLeadingZeros16-16 3.04 1.83 -39.80% BenchmarkLeadingZeros32-16 3.31 1.82 -45.02% BenchmarkLeadingZeros64-16 3.69 1.71 -53.66% BenchmarkTrailingZeros-16 2.55 1.62 -36.47% BenchmarkTrailingZeros32-16 2.55 1.77 -30.59% BenchmarkTrailingZeros64-16 2.78 1.62 -41.73% BenchmarkOnesCount-16 3.19 0.93 -70.85% BenchmarkOnesCount32-16 2.55 1.18 -53.73% BenchmarkOnesCount64-16 3.22 0.93 -71.12% Update #18616 I also made a change to bits_test.go because when debugging some failures the output was not quite providing the right argument information. Change-Id: Ia58d31d1777cf4582a4505f85b11a1202ca07d3e Reviewed-on: https://go-review.googlesource.com/41630 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> Reviewed-by: Keith Randall <khr@golang.org>
* math: use SIMD to accelerate additional scalar math functions on s390xBill O'Farrell2017-05-0834-65/+4720
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As necessary, math functions were structured to use stubs, so that they can be accelerated with assembly on any platform. Technique used was minimax polynomial approximation using tables of polynomial coefficients, with argument range reduction. Benchmark New Old Speedup BenchmarkAcos 12.2 47.5 3.89 BenchmarkAcosh 18.5 56.2 3.04 BenchmarkAsin 13.1 40.6 3.10 BenchmarkAsinh 19.4 62.8 3.24 BenchmarkAtan 10.1 23 2.28 BenchmarkAtanh 19.1 53.2 2.79 BenchmarkAtan2 16.5 33.9 2.05 BenchmarkCbrt 14.8 58 3.92 BenchmarkErf 10.8 20.1 1.86 BenchmarkErfc 11.2 23.5 2.10 BenchmarkExp 8.77 53.8 6.13 BenchmarkExpm1 10.1 38.3 3.79 BenchmarkLog 13.1 40.1 3.06 BenchmarkLog1p 12.7 38.3 3.02 BenchmarkPowInt 31.7 40.5 1.28 BenchmarkPowFrac 33.1 141 4.26 BenchmarkTan 11.5 30 2.61 Accuracy was tested against a high precision reference function to determine maximum error. Note: ulperr is error in "units in the last place" max ulperr Acos 1.15 Acosh 1.07 Asin 2.22 Asinh 1.72 Atan 1.41 Atanh 3.00 Atan2 1.45 Cbrt 1.18 Erf 1.29 Erfc 4.82 Exp 1.00 Expm1 2.26 Log 0.94 Log1p 2.39 Tan 3.14 Pow will have 99.99% correctly rounded results with reasonable inputs producing numeric (non Inf or NaN) results Change-Id: I850e8cf7b70426e8b54ec49d74acd4cddc8c6cb2 Reviewed-on: https://go-review.googlesource.com/38585 Reviewed-by: Michael Munday <munday@ca.ibm.com> Run-TryBot: Michael Munday <munday@ca.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math/big: improve performance for addVV/subVV for ppc64xCarlos Eduardo Seo2017-04-251-12/+37
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change adds a better asm implementation of addVV for ppc64x, with speedups up to nearly 3x in the best cases. benchmark old ns/op new ns/op delta BenchmarkAddVV/1-8 7.33 5.81 -20.74% BenchmarkAddVV/2-8 8.72 6.49 -25.57% BenchmarkAddVV/3-8 10.5 7.08 -32.57% BenchmarkAddVV/4-8 12.7 7.57 -40.39% BenchmarkAddVV/5-8 14.3 8.06 -43.64% BenchmarkAddVV/10-8 27.6 11.1 -59.78% BenchmarkAddVV/100-8 218 82.4 -62.20% BenchmarkAddVV/1000-8 2064 718 -65.21% BenchmarkAddVV/10000-8 20536 7153 -65.17% BenchmarkAddVV/100000-8 211004 72403 -65.69% benchmark old MB/s new MB/s speedup BenchmarkAddVV/1-8 8729.74 11006.26 1.26x BenchmarkAddVV/2-8 14683.65 19707.55 1.34x BenchmarkAddVV/3-8 18226.96 27103.63 1.49x BenchmarkAddVV/4-8 20204.50 33805.81 1.67x BenchmarkAddVV/5-8 22348.64 39694.06 1.78x BenchmarkAddVV/10-8 23212.74 57631.08 2.48x BenchmarkAddVV/100-8 29300.07 77629.53 2.65x BenchmarkAddVV/1000-8 31000.56 89094.54 2.87x BenchmarkAddVV/10000-8 31163.61 89469.16 2.87x BenchmarkAddVV/100000-8 30331.16 88393.73 2.91x It also adds the use of CTR for the loop counter in subVV, instead of manually updating the loop counter. This is slightly faster. Change-Id: Ic4b05cad384fd057972d46a5618ed5c3039d7460 Reviewed-on: https://go-review.googlesource.com/41010 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>
* math: remove asm version of sincos everywhere, except 386Ilya Tocar2017-04-2410-173/+16
| | | | | | | | | | | | | | | | | | | | We have dedicated asm implementation of sincos only on 386 and amd64, on everything else we are just jumping to generic version. However amd64 version is actually slower than generic one: Sincos-6 34.4ns ± 0% 24.8ns ± 0% -27.79% (p=0.000 n=8+10) So remove all sincos*.s and keep only generic and 386. Updates #19819 Change-Id: I7eefab35743729578264f52f6d23ee2c227c92a5 Reviewed-on: https://go-review.googlesource.com/41200 Run-TryBot: Ilya Tocar <ilya.tocar@intel.com> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
* cmd/asm, cmd/internal/obj/s390x, math: add LGDR and LDGR instructionsMichael Munday2017-04-175-19/+19
| | | | | | | | | | | The instructions allow moves between floating point and general purpose registers without any conversion taking place. Change-Id: I82c6f3ad9c841a83783b5be80dcf5cd538ff49e6 Reviewed-on: https://go-review.googlesource.com/38777 Run-TryBot: Michael Munday <munday@ca.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
* math/bits: support negative rotation count and remove RotateRightRobert Griesemer2017-04-112-157/+34
| | | | | | | | | | | | | | | | | | | | | | | | For details see the discussion on the issue below. RotateLeft functions can now be inlined because the don't panic anymore for negative rotation counts. name old time/op new time/op delta RotateLeft-8 6.72ns ± 2% 1.86ns ± 0% -72.33% (p=0.016 n=5+4) RotateLeft8-8 4.41ns ± 2% 1.67ns ± 1% -62.15% (p=0.008 n=5+5) RotateLeft16-8 4.46ns ± 6% 1.65ns ± 0% -63.06% (p=0.008 n=5+5) RotateLeft32-8 4.50ns ± 5% 1.67ns ± 1% -62.86% (p=0.008 n=5+5) RotateLeft64-8 4.54ns ± 1% 1.85ns ± 1% -59.32% (p=0.008 n=5+5) https://perf.golang.org/search?q=upload:20170411.4 (Measured on 2.3 GHz Intel Core i7 running macOS 10.12.3.) For #18616. Change-Id: I0828d80d54ec24f8d44954a57b3d6aeedb69c686 Reviewed-on: https://go-review.googlesource.com/40394 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* all: fix minor misspellingsEric Lagergren2017-04-031-1/+1
| | | | | | | | Change-Id: I1f1cfb161640eb8756fb1a283892d06b30b7a8fa Reviewed-on: https://go-review.googlesource.com/39356 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math/big: Unify divWW implementation for ppc64 and ppc64le.Carlos Eduardo Seo2017-03-313-64/+39
| | | | | | | | | | | | | | Starting in go1.9, the minimum processor requirement for ppc64 is POWER8. So it may now use the same divWW implementation as ppc64le. Updates #19074 Change-Id: If1a85f175cda89eee06a1024ccd468da6124c844 Reviewed-on: https://go-review.googlesource.com/39010 Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
* math: speed up Log on amd64Ilya Tocar2017-03-291-0/+1
| | | | | | | | | | | | | | | After https://golang.org/cl/31490 we break false output dependency for CVTS.. in compiler generated code. I've looked through asm code, which uses CVTS.. and added XOR to the only case where it affected performance. Log-6 21.6ns ± 0% 19.9ns ± 0% -7.87% (p=0.000 n=10+10) Change-Id: I25d9b405e3041a3839b40f9f9a52e708034bb347 Reviewed-on: https://go-review.googlesource.com/38771 Run-TryBot: Ilya Tocar <ilya.tocar@intel.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
* math/big: replace local versions of bitLen, nlz with math/bits versionsRobert Griesemer2017-03-2317-141/+17
| | | | | | | | | | | | Verified that BenchmarkBitLen time went down from 2.25 ns/op to 0.65 ns/op an a 2.3 GHz Intel Core i7, before removing that benchmark (now covered by math/bits benchmarks). Change-Id: I3890bb7d1889e95b9a94bd68f0bdf06f1885adeb Reviewed-on: https://go-review.googlesource.com/38464 Run-TryBot: Robert Griesemer <gri@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/big: fix TestFloatSetFloat64StringRobert Griesemer2017-03-231-11/+13
| | | | | | | | | | | | | A -0 constant is the same as 0. Use explicit negative zero for float64 -0.0. Also, fix two test cases that were wrong. Fixes #19673. Change-Id: Ic09775f29d9bc2ee7814172e59c4a693441ea730 Reviewed-on: https://go-review.googlesource.com/38463 Run-TryBot: Robert Griesemer <gri@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
* math/big: make nat.setUint64 vet-friendlyJosh Bleecher Snyder2017-03-111-15/+5
| | | | | | | | | | | | | | | | | | | nat.setUint64 is nicely generic. By assuming 32- or 64-bit words, however, we can write simpler code, and eliminate some shifts in dead code that vet complains about. Generated code for 64 bit systems is unaltered. Generated code for 32 bit systems is much better. For 386, the routine length drops from 325 bytes of code to 271 bytes of code, with fewer loops. Change-Id: I1bc14c06272dee37a7fcb48d33dd1e621eba945d Reviewed-on: https://go-review.googlesource.com/38070 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
* all: remove the the duplicate wordsEitan Adler2017-03-061-1/+1
| | | | | | | | | Change-Id: I6343c162e27e2e492547c96f1fc504909b1c03c0 Reviewed-on: https://go-review.googlesource.com/37793 Reviewed-by: Daniel Martí <mvdan@mvdan.cc> Reviewed-by: Ian Lance Taylor <iant@golang.org> Run-TryBot: Ian Lance Taylor <iant@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math/bits: move left-over functionality from bits_impl.go to bits.goRobert Griesemer2017-02-282-72/+59
| | | | | | | | | | Removes an extra function call for TrailingZeroes and thus may increase chances for inlining. Change-Id: Iefd8d4402dc89b64baf4e5c865eb3dadade623af Reviewed-on: https://go-review.googlesource.com/37613 Run-TryBot: Robert Griesemer <gri@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/bits: faster LeadingZeros and Len functionsRobert Griesemer2017-02-284-33/+75
| | | | | | | | | | | | | | | benchmark old ns/op new ns/op delta BenchmarkLeadingZeros-8 8.43 3.10 -63.23% BenchmarkLeadingZeros8-8 8.13 1.33 -83.64% BenchmarkLeadingZeros16-8 7.34 2.07 -71.80% BenchmarkLeadingZeros32-8 7.99 2.87 -64.08% BenchmarkLeadingZeros64-8 8.13 2.96 -63.59% Measured on 2.3 GHz Intel Core i7 running macOS 10.12.3. Change-Id: Id343531b408d42ac45f10c76f60e85bdb977f91e Reviewed-on: https://go-review.googlesource.com/37582 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/bits: faster TrailingZeroes8Robert Griesemer2017-02-284-9/+29
| | | | | | | | | | | | | For sizes > 8, the existing code is faster. benchmark old ns/op new ns/op delta BenchmarkTrailingZeros8-8 1.95 1.29 -33.85% Measured on 2.3 GHz Intel Core i7 running macOS 10.12.3. Change-Id: I6f3a33ec633a2c544ec29693c141f2f99335c745 Reviewed-on: https://go-review.googlesource.com/37581 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/bits: faster OnesCount using table lookups for sizes 8,16,32Robert Griesemer2017-02-283-18/+31
| | | | | | | | | | | | | | | For uint64, the existing algorithm is faster. benchmark old ns/op new ns/op delta BenchmarkOnesCount8-8 1.95 0.97 -50.26% BenchmarkOnesCount16-8 2.54 1.39 -45.28% BenchmarkOnesCount32-8 2.61 1.96 -24.90% Measured on 2.3 GHz Intel Core i7 running macOS 10.12.3. Change-Id: I6cc42882fef3d24694720464039161e339a9ae99 Reviewed-on: https://go-review.googlesource.com/37580 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/bits: faster Reverse8/16 functions using table lookupsRobert Griesemer2017-02-253-9/+96
| | | | | | | | | | | | | | | | Measured on 2.3 GHz Intel Core i7, running macOS 10.12.3: benchmark old ns/op new ns/op delta BenchmarkReverse8-8 1.70 0.99 -41.76% BenchmarkReverse16-8 2.24 1.32 -41.07% Fixes #19279. Change-Id: I398cf8a3513b7fa63c130efc7846a7c5353999d4 Reviewed-on: https://go-review.googlesource.com/37459 Run-TryBot: Robert Griesemer <gri@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/bits: fix incorrect doc strings for TrailingZeros functionsRobert Griesemer2017-02-251-5/+5
| | | | | | Change-Id: I3e40018ab1903d3b9ada7ad7812ba71ea2a428e7 Reviewed-on: https://go-review.googlesource.com/37456 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/big: use math/bits where appropriateRobert Griesemer2017-02-244-97/+18
| | | | | | | | | | | | | | This change adds math/bits as a new dependency of math/big. - use bits.LeadingZeroes instead of local implementation (they are identical, so there's no performance loss here) - leave other functionality local (ntz, bitLen) since there's faster implementations in math/big at the moment Change-Id: I1218aa8a1df0cc9783583b090a4bb5a8a145c4a2 Reviewed-on: https://go-review.googlesource.com/37141 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math: speed up and improve accuracy of Pow10Martin Möhrmann2017-02-222-33/+65
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Removes init function from the math package. Allows stripping of arrays with pre-computed values used for Pow10 from binaries if Pow10 is not used. cmd/go shrinks by 128 bytes. Fixed small values like 10**-323 being 0 instead of 1e-323. Overall precision is increased but still not as good as predefined constants for some inputs. Samples: Pow10(208) before: 1.0000000000000006662e+208 after: 1.0000000000000000959e+208 Pow10(202) before 1.0000000000000009895e+202 after 1.0000000000000001193e+202 Pow10(60) before 1.0000000000000001278e+60 after 0.9999999999999999494e+60 Pow10(-100) before 0.99999999999999938551e-100 after 0.99999999999999989309e-100 Pow10(-200) before 0.9999999999999988218e-200 after 1.0000000000000001271e-200 name old time/op new time/op delta Pow10Pos-4 44.6ns ± 2% 1.2ns ± 1% -97.39% (p=0.000 n=19+17) Pow10Neg-4 50.8ns ± 1% 4.1ns ± 2% -92.02% (p=0.000 n=17+19) Change-Id: If094034286b8ac64be3a95fd9e8ffa3d4ad39b31 Reviewed-on: https://go-review.googlesource.com/36331 Reviewed-by: Robert Griesemer <gri@golang.org> Run-TryBot: Martin Möhrmann <moehrmann@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math: add more tests for special cases of Bessel functions Y0, Y1, YnAlexander Döring2017-02-221-0/+8
| | | | | | | | | | | | Test finite negative x with Y0(-1), Y1(-1), Yn(2,-1), Yn(-3,-1). Also test the special case Yn(0,0). Fixes #19130. Change-Id: I95f05a72e1c455ed8ddf202c56f4266f03f370fd Reviewed-on: https://go-review.googlesource.com/37310 Reviewed-by: Robert Griesemer <gri@golang.org>
* math/big: define Word as uint instead of uintptrRobert Griesemer2017-02-211-1/+1
| | | | | | | | | | | | | | | | | | | | For compatibility with math/bits uint operations. When math/big was written originally, the Go compiler used 32bit int/uint values even on a 64bit machine. uintptr was the type that represented the machine register size. Now, the int/uint types are sized to the native machine register size, so they are the natural machine Word type. On most machines, the size of int/uint correspond to the size of uintptr. On platforms where uint and uintptr have different sizes, this change may lead to performance differences (e.g., amd64p32). Change-Id: Ief249c160b707b6441848f20041e32e9e9d8d8ca Reviewed-on: https://go-review.googlesource.com/37372 Run-TryBot: Robert Griesemer <gri@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/bits: faster OnesCountRobert Griesemer2017-02-192-47/+82
| | | | | | | | | | | | | | | | | | | | Using some additional suggestions per "Hacker's Delight". Added documentation and extra tests. Measured on 1.7 GHz Intel Core i7, running macOS 10.12.3. benchmark old ns/op new ns/op delta BenchmarkOnesCount-4 7.34 5.38 -26.70% BenchmarkOnesCount8-4 2.03 1.98 -2.46% BenchmarkOnesCount16-4 2.56 2.50 -2.34% BenchmarkOnesCount32-4 2.98 2.39 -19.80% BenchmarkOnesCount64-4 4.22 2.96 -29.86% Change-Id: I566b0ef766e55cf5776b1662b6016024ebe5d878 Reviewed-on: https://go-review.googlesource.com/37223 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math: protect benchmarked functions from being optimized awayMartin Möhrmann2017-02-181-71/+200
| | | | | | | | | | | | | Add exported global variables and store the results of benchmarked functions in them. This prevents the current compiler optimizations from removing the instructions that are needed to compute the return values of the benchmarked functions. Change-Id: If8b08424e85f3796bb6dd73e761c653abbabcc5e Reviewed-on: https://go-review.googlesource.com/37195 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math/bits: added benchmarks for Leading/TrailingZerosRobert Griesemer2017-02-171-34/+114
| | | | | | | | | | | | | | | | | | | BenchmarkLeadingZeros-8 200000000 8.80 ns/op BenchmarkLeadingZeros8-8 200000000 8.21 ns/op BenchmarkLeadingZeros16-8 200000000 7.49 ns/op BenchmarkLeadingZeros32-8 200000000 7.80 ns/op BenchmarkLeadingZeros64-8 200000000 8.67 ns/op BenchmarkTrailingZeros-8 1000000000 2.05 ns/op BenchmarkTrailingZeros8-8 2000000000 1.94 ns/op BenchmarkTrailingZeros16-8 2000000000 1.94 ns/op BenchmarkTrailingZeros32-8 2000000000 1.92 ns/op BenchmarkTrailingZeros64-8 2000000000 2.03 ns/op Change-Id: I45497bf2d6369ba6cfc88ded05aa735908af8908 Reviewed-on: https://go-review.googlesource.com/37220 Run-TryBot: Robert Griesemer <gri@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
* math/bits: faster Rotate functions, added respective benchmarksRobert Griesemer2017-02-173-26/+161
| | | | | | | | | | | | | | | | | | | | | Measured on 2.3 GHz Intel Core i7, running maxOS 10.12.3. benchmark old ns/op new ns/op delta BenchmarkRotateLeft-8 7.87 7.00 -11.05% BenchmarkRotateLeft8-8 8.41 4.52 -46.25% BenchmarkRotateLeft16-8 8.07 4.55 -43.62% BenchmarkRotateLeft32-8 8.36 4.73 -43.42% BenchmarkRotateLeft64-8 7.93 4.78 -39.72% BenchmarkRotateRight-8 8.23 6.72 -18.35% BenchmarkRotateRight8-8 8.76 4.39 -49.89% BenchmarkRotateRight16-8 9.07 4.44 -51.05% BenchmarkRotateRight32-8 8.85 4.46 -49.60% BenchmarkRotateRight64-8 8.11 4.43 -45.38% Change-Id: I79ea1e9e6fc65f95794a91f860a911efed3aa8a1 Reviewed-on: https://go-review.googlesource.com/37219 Reviewed-by: Matthew Dempsky <mdempsky@google.com>
* math/bits: faster OnesCount, added respective benchmarksRobert Griesemer2017-02-173-50/+119
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Also: Changed Reverse/ReverseBytes implementations to use the same (smaller) masks as OnesCount. BenchmarkOnesCount-8 37.0 6.26 -83.08% BenchmarkOnesCount8-8 7.24 1.99 -72.51% BenchmarkOnesCount16-8 11.3 2.47 -78.14% BenchmarkOnesCount32-8 18.4 3.02 -83.59% BenchmarkOnesCount64-8 40.0 3.78 -90.55% BenchmarkReverse-8 6.69 6.22 -7.03% BenchmarkReverse8-8 1.64 1.64 +0.00% BenchmarkReverse16-8 2.26 2.18 -3.54% BenchmarkReverse32-8 2.88 2.87 -0.35% BenchmarkReverse64-8 5.64 4.34 -23.05% BenchmarkReverseBytes-8 2.48 2.17 -12.50% BenchmarkReverseBytes16-8 0.63 0.95 +50.79% BenchmarkReverseBytes32-8 1.13 1.24 +9.73% BenchmarkReverseBytes64-8 2.50 2.16 -13.60% OnesCount-8 37.0ns ± 0% 6.3ns ± 0% ~ (p=1.000 n=1+1) OnesCount8-8 7.24ns ± 0% 1.99ns ± 0% ~ (p=1.000 n=1+1) OnesCount16-8 11.3ns ± 0% 2.5ns ± 0% ~ (p=1.000 n=1+1) OnesCount32-8 18.4ns ± 0% 3.0ns ± 0% ~ (p=1.000 n=1+1) OnesCount64-8 40.0ns ± 0% 3.8ns ± 0% ~ (p=1.000 n=1+1) Reverse-8 6.69ns ± 0% 6.22ns ± 0% ~ (p=1.000 n=1+1) Reverse8-8 1.64ns ± 0% 1.64ns ± 0% ~ (all samples are equal) Reverse16-8 2.26ns ± 0% 2.18ns ± 0% ~ (p=1.000 n=1+1) Reverse32-8 2.88ns ± 0% 2.87ns ± 0% ~ (p=1.000 n=1+1) Reverse64-8 5.64ns ± 0% 4.34ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes-8 2.48ns ± 0% 2.17ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes16-8 0.63ns ± 0% 0.95ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes32-8 1.13ns ± 0% 1.24ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes64-8 2.50ns ± 0% 2.16ns ± 0% ~ (p=1.000 n=1+1) Change-Id: I591b0ffc83fc3a42828256b6e5030f32c64f9497 Reviewed-on: https://go-review.googlesource.com/37218 Reviewed-by: Matthew Dempsky <mdempsky@google.com>
* math/bits: faster Reverse, ReverseBytesRobert Griesemer2017-02-171-33/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - moved from: x&m>>k | x&^m<<k to: x&m>>k | x<<k&m This permits use of the same constant m twice (*) which may be better for machines that can't use large immediate constants directly with an AND instruction and have to load them explicitly. *) CPUs don't usually have a &^ instruction, so x&^m becomes x&(^m) - simplified returns This improves the generated code because the compiler recognizes x>>k | x<<k as ROT when k is the bitsize of x. The 8-bit versions of these instructions can be significantly faster still if they are replaced with table lookups, as long as the table is in cache. If the table is not in cache, table-lookup is probably slower, hence the choice of an explicit register-only implementation for now. BenchmarkReverse-8 8.50 6.86 -19.29% BenchmarkReverse8-8 2.17 1.74 -19.82% BenchmarkReverse16-8 2.89 2.34 -19.03% BenchmarkReverse32-8 3.55 2.95 -16.90% BenchmarkReverse64-8 6.81 5.57 -18.21% BenchmarkReverseBytes-8 3.49 2.48 -28.94% BenchmarkReverseBytes16-8 0.93 0.62 -33.33% BenchmarkReverseBytes32-8 1.55 1.13 -27.10% BenchmarkReverseBytes64-8 2.47 2.47 +0.00% Reverse-8 8.50ns ± 0% 6.86ns ± 0% ~ (p=1.000 n=1+1) Reverse8-8 2.17ns ± 0% 1.74ns ± 0% ~ (p=1.000 n=1+1) Reverse16-8 2.89ns ± 0% 2.34ns ± 0% ~ (p=1.000 n=1+1) Reverse32-8 3.55ns ± 0% 2.95ns ± 0% ~ (p=1.000 n=1+1) Reverse64-8 6.81ns ± 0% 5.57ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes-8 3.49ns ± 0% 2.48ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes16-8 0.93ns ± 0% 0.62ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes32-8 1.55ns ± 0% 1.13ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes64-8 2.47ns ± 0% 2.47ns ± 0% ~ (all samples are equal) Change-Id: I0064de8c7e0e568ca7885d6f7064344bef91a06d Reviewed-on: https://go-review.googlesource.com/37215 Run-TryBot: Robert Griesemer <gri@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math/bits: fix benchmarks (make sure calls don't get optimized away)Robert Griesemer2017-02-171-9/+32
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Sum up function results and store them in an exported (global) variable. This prevents the compiler from optimizing away the otherwise side-effect free function calls. We now have more realistic set of benchmark numbers... Measured on 2.3 GHz Intel Core i7, running maxOS 10.12.3. Note: These measurements are based on the same "old" implementation as the prior measurements (commit 7d5c003). benchmark old ns/op new ns/op delta BenchmarkReverse-8 72.9 8.50 -88.34% BenchmarkReverse8-8 13.2 2.17 -83.56% BenchmarkReverse16-8 21.2 2.89 -86.37% BenchmarkReverse32-8 36.3 3.55 -90.22% BenchmarkReverse64-8 71.3 6.81 -90.45% BenchmarkReverseBytes-8 11.2 3.49 -68.84% BenchmarkReverseBytes16-8 6.24 0.93 -85.10% BenchmarkReverseBytes32-8 7.40 1.55 -79.05% BenchmarkReverseBytes64-8 10.5 2.47 -76.48% Reverse-8 72.9ns ± 0% 8.5ns ± 0% ~ (p=1.000 n=1+1) Reverse8-8 13.2ns ± 0% 2.2ns ± 0% ~ (p=1.000 n=1+1) Reverse16-8 21.2ns ± 0% 2.9ns ± 0% ~ (p=1.000 n=1+1) Reverse32-8 36.3ns ± 0% 3.5ns ± 0% ~ (p=1.000 n=1+1) Reverse64-8 71.3ns ± 0% 6.8ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes-8 11.2ns ± 0% 3.5ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes16-8 6.24ns ± 0% 0.93ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes32-8 7.40ns ± 0% 1.55ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes64-8 10.5ns ± 0% 2.5ns ± 0% ~ (p=1.000 n=1+1) Change-Id: I8aef1334b84f6cafd25edccad7e6868b37969efb Reviewed-on: https://go-review.googlesource.com/37213 Reviewed-by: Matthew Dempsky <mdempsky@google.com>
* math/bits: much faster ReverseBytes, added respective benchmarksRobert Griesemer2017-02-173-16/+52
| | | | | | | | | | | | | | | | | | | | | Measured on 2.3 GHz Intel Core i7, running maxOS 10.12.3. benchmark old ns/op new ns/op delta BenchmarkReverseBytes-8 11.4 3.51 -69.21% BenchmarkReverseBytes16-8 6.87 0.64 -90.68% BenchmarkReverseBytes32-8 7.79 0.65 -91.66% BenchmarkReverseBytes64-8 11.6 0.64 -94.48% name old time/op new time/op delta ReverseBytes-8 11.4ns ± 0% 3.5ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes16-8 6.87ns ± 0% 0.64ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes32-8 7.79ns ± 0% 0.65ns ± 0% ~ (p=1.000 n=1+1) ReverseBytes64-8 11.6ns ± 0% 0.6ns ± 0% ~ (p=1.000 n=1+1) Change-Id: I67b529652b3b613c61687e9e185e8d4ee40c51a2 Reviewed-on: https://go-review.googlesource.com/37211 Run-TryBot: Robert Griesemer <gri@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
* math/bits: much faster Reverse, added respective benchmarksRobert Griesemer2017-02-173-13/+77
| | | | | | | | | | | | | | | | | | | | | | | Measured on 2.3 GHz Intel Core i7, running maxOS 10.12.3. name old time/op new time/op delta Reverse-8 76.6ns ± 0% 8.1ns ± 0% ~ (p=1.000 n=1+1) Reverse8-8 12.6ns ± 0% 0.6ns ± 0% ~ (p=1.000 n=1+1) Reverse16-8 20.8ns ± 0% 0.6ns ± 0% ~ (p=1.000 n=1+1) Reverse32-8 36.5ns ± 0% 0.6ns ± 0% ~ (p=1.000 n=1+1) Reverse64-8 74.0ns ± 0% 6.4ns ± 0% ~ (p=1.000 n=1+1) benchmark old ns/op new ns/op delta BenchmarkReverse-8 76.6 8.07 -89.46% BenchmarkReverse8-8 12.6 0.64 -94.92% BenchmarkReverse16-8 20.8 0.64 -96.92% BenchmarkReverse32-8 36.5 0.64 -98.25% BenchmarkReverse64-8 74.0 6.38 -91.38% Change-Id: I6b99b10cee2f2babfe79342b50ee36a45a34da30 Reviewed-on: https://go-review.googlesource.com/37149 Run-TryBot: Robert Griesemer <gri@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
* math/bits: expand doc strings for all functionsRobert Griesemer2017-02-171-35/+101
| | | | | | | | | | | Follow-up on https://go-review.googlesource.com/36315. No functionality change. For #18616. Change-Id: Id4df34dd7d0381be06eea483a11bf92f4a01f604 Reviewed-on: https://go-review.googlesource.com/37140 Reviewed-by: Matthew Dempsky <mdempsky@google.com>
* math: fix typos in Bessel function docsShenghou Ma2017-02-161-3/+3
| | | | | | | | | | | | While we're at it, also document Yn(0, 0) = -Inf for completeness. Fixes #18823. Change-Id: Ib6db68f76d29cc2373c12ebdf3fab129cac8c167 Reviewed-on: https://go-review.googlesource.com/35970 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math/bits: added package for bit-level counting and manipulationRobert Griesemer2017-02-163-0/+727
| | | | | | | | | | | Initial platform-independent implementation. For #18616. Change-Id: I4585c55b963101af9059c06c1b8a866cb384754c Reviewed-on: https://go-review.googlesource.com/36315 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Russ Cox <rsc@golang.org>
* math/big: simplify bool expressionDaniel Martí2017-02-141-1/+1
| | | | | | | | Change-Id: I280c53be455f2fe0474ad577c0f7b7908a4eccb2 Reviewed-on: https://go-review.googlesource.com/36993 Reviewed-by: Ian Lance Taylor <iant@golang.org> Run-TryBot: Ian Lance Taylor <iant@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math/big: fix s390x test build tagsMichael Munday2017-02-141-1/+1
| | | | | | | | | | | The tests failed to compile when using the math_big_pure_go tag on s390x. Change-Id: I2a09f53ff6562ab9bc9b886cffc0f6205bbfcfbb Reviewed-on: https://go-review.googlesource.com/36956 Run-TryBot: Michael Munday <munday@ca.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* all: fix some printf format stringsJosh Bleecher Snyder2017-02-141-1/+1
| | | | | | | | | | Appease vet. Change-Id: Ie88de08b91041990c0eaf2e15628cdb98d40c660 Reviewed-on: https://go-review.googlesource.com/36938 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
* cmd/{asm,internal/obj/s390x}, math: remove emulated float instructionsMichael Munday2017-02-105-22/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | | The s390x port was based on the ppc64 port and, because of the way the port was done, inherited some instructions from it. ppc64 supports 3-operand (4-operand for FMADD etc.) floating point instructions but s390x doesn't (the destination register is always an input) and so these were emulated. There is a bug in the emulation of FMADD whereby if the destination register is also a source for the multiplication it will be clobbered. This doesn't break any assembly code in the std lib but could affect future work. To fix this I have gone through the floating point instructions and removed all unnecessary 3-/4-operand emulation. The compiler doesn't need it and assembly writers don't need it, it's just a source of bugs. I've also deleted the FNMADD family of emulated instructions. They aren't used anywhere. Change-Id: Ic07cedcf141a6a3b43a0c84895460f6cfbf56c04 Reviewed-on: https://go-review.googlesource.com/33350 Run-TryBot: Michael Munday <munday@ca.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
* math: check overflow in amd64 Exp implementationAlberto Donizetti2017-02-102-3/+34
| | | | | | | | | | | | | | | | | | | | Unlike the pure go implementation used by every other architecture, the amd64 asm implementation of Exp does not fail early if the argument is known to overflow. Make it fail early. Cost of the check is < 1ns (on an old Sandy Bridge machine): name old time/op new time/op delta Exp-4 18.3ns ± 1% 18.7ns ± 1% +2.08% (p=0.000 n=18+20) Fixes #14932 Fixes #18912 Change-Id: I04b3f9b4ee853822cbdc97feade726fbe2907289 Reviewed-on: https://go-review.googlesource.com/36271 Run-TryBot: Alberto Donizetti <alb.donizetti@gmail.com> Reviewed-by: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
* math/big: add IsInt64/IsUint64 predicatesRobert Griesemer2017-02-072-47/+110
| | | | | | Change-Id: Ia5ed3919cb492009ac8f66d175b47a69f83ee4f1 Reviewed-on: https://go-review.googlesource.com/36487 Reviewed-by: Alan Donovan <adonovan@google.com>
* crypto/*: document use or non-use of constant-time algorithmsRuss Cox2016-12-071-1/+4
| | | | | | | | | | Fixes #16821. Change-Id: I63d5f3d7cfba1c76259912d754025c5f3cbe4a56 Reviewed-on: https://go-review.googlesource.com/31573 Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
* math/rand: export Source64, mainly for documentation valueRuss Cox2016-11-231-8/+11
| | | | | | | | | | | | There is some code value too: types intending to implement Source64 can write a conversion confirming that. For #4254 and the Go 1.8 release notes. Change-Id: I7fc350a84f3a963e4dab317ad228fa340dda5c66 Reviewed-on: https://go-review.googlesource.com/33456 Run-TryBot: Russ Cox <rsc@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
View Lorry Controller Status