Speed regression when going from Fedora 37 from Fedora 36

109 views
Skip to first unread message

Anthony Starks

unread,
Nov 15, 2022, 6:26:16 PM11/15/22
to golang-nuts
I just updated my X1 Carbon 5th gen, (8Gb RAM, 256 Gb SSD) system from Fedora Linux 36 to 37 and the same Go binaries (pdfdeck [1]) are about 2x slower.  The code was built with Go 1.19.3 and has been stable.  Rebuilding with a fresh install of go  has no impact.  

Question: is there anything in this version of Linux (6.0.8) and Go that would cause an issue?

$ uname -a 
Linux slab 6.0.8-300.fc37.x86_64 #1 SMP PREEMPT_DYNAMIC Fri Nov 11 15:09:04 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux

$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/home/ajstarks/.cache/go-build"
GOENV="/home/ajstarks/.config/go/env"
GOEXE=""
GOEXPERIMENT=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOINSECURE=""
GOMODCACHE="/home/ajstarks/gowork/pkg/mod"
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/home/ajstarks/gowork"
GOPRIVATE=""
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/home/ajstarks/go"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/home/ajstarks/go/pkg/tool/linux_amd64"
GOVCS=""
GOVERSION="go1.19.3"
GCCGO="gccgo"
GOAMD64="v1"
AR="ar"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD="/dev/null"
GOWORK=""
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -Wl,--no-gc-sections -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build299185330=/tmp/go-build -gno-record-gcc-switches"





Ian Lance Taylor

unread,
Nov 15, 2022, 6:34:45 PM11/15/22
to Anthony Starks, golang-nuts
On Tue, Nov 15, 2022 at 3:26 PM Anthony Starks <ajst...@gmail.com> wrote:
>
> I just updated my X1 Carbon 5th gen, (8Gb RAM, 256 Gb SSD) system from Fedora Linux 36 to 37 and the same Go binaries (pdfdeck [1]) are about 2x slower. The code was built with Go 1.19.3 and has been stable. Rebuilding with a fresh install of go has no impact.
>
> Question: is there anything in this version of Linux (6.0.8) and Go that would cause an issue?

Certainly sounds like it. But I have no idea what it would be. The
first step might be to run a binary using "strace -f --relative-times"
on both systems and see if you can spot anything significantly slower.

Ian

Anthony Starks

unread,
Nov 15, 2022, 6:46:03 PM11/15/22
to golang-nuts
Oddly the benchmark times are back to normal.  (after a few reboots and selinux fiddling (turning it off and then back on).
Probably some local soaking was needed for the new system. Sorry for the noise (and thanks for the strace tip)

Reply all
Reply to author
Forward
0 new messages