目前go程序1.3、1.4和1.5的编译性能差异很大，后者4x比前者慢一些。如何跟踪go编译器执行情况？
像这样的东西：
注意：valgrind+calgrind不起作用（教程）
valgrind --tool=callgrind /usr/bin/go build
==26982== Callgrind, a call-graph generating cache profiler
==26982== Copyright (C) 2002-2015, and GNU GPL'd, by Josef Weidendorfer et al.
==26982== Using Valgrind-3.11.0 and LibVEX; rerun with -h for copyright info
==26982== Command: /usr/bin/go build
==26982==
==26982== For interactive control, run 'callgrind_control -h'.
fatal error: rt_sigaction failure
runtime stack:
runtime.throw(0x9a2bf0, 0x14)
/usr/lib/go/src/runtime/panic.go:527 +0x90
runtime.setsig(0xc800000040, 0x4e40d0, 0x1)
/usr/lib/go/src/runtime/os1_linux.go:297 +0x197
runtime.initsig()
/usr/lib/go/src/runtime/signal1_unix.go:67 +0x13d
runtime.mstart1()
/usr/lib/go/src/runtime/proc1.go:717 +0xc9
runtime.mstart()
/usr/lib/go/src/runtime/proc1.go:691 +0x72
goroutine 1 [runnable]:
runtime.main()
/usr/lib/go/src/runtime/proc.go:28
runtime.goexit()
/usr/lib/go/src/runtime/asm_amd64.s:1696 +0x1
goroutine 17 [syscall, locked to thread]:
runtime.goexit()
/usr/lib/go/src/runtime/asm_amd64.s:1696 +0x1
==26982==
==26982== Events : Ir
==26982== Collected : 851285
==26982==
==26982== I refs: 851,285
答案1
没关系，找到了回答，但不够有用：
去1.5
perf record /usr/bin/go build
[ perf record: Woken up 9 times to write data ]
[ perf record: Captured and wrote 2.890 MB perf.data (73779 samples) ]
perf report --sort comm,dso
Samples: 73K of event 'cycles', Event count (approx.): 54540942397
Overhead Command Shared Object
82.19% compile compile
10.62% link link
5.07% compile [kernel.vmlinux]
0.76% link [kernel.vmlinux]
0.55% go go
0.52% go [kernel.vmlinux]
0.07% compile [ath9k]
0.05% compile [mac80211]
0.03% go [ext4]
0.03% compile [i8042]
0.03% compile [vdso]
0.01% go [vdso]
0.01% compile [libahci]
0.01% compile [ath9k_hw]
0.01% link [ext4]
0.01% go [scsi_mod]
0.01% link [mac80211]
0.01% go [jbd2]
0.01% go [libata]
0.00% compile [ext4]
0.00% link [libahci]
0.00% compile [psmouse]
0.00% compile [cfg80211]
0.00% compile [libata]
0.00% go [mac80211]
0.00% link [i8042]
0.00% link [scsi_mod]
0.00% compile [scsi_mod]
0.00% link [vdso]
0.00% go [ath9k]
0.00% go [cfg80211]
去1.4
perf record /usr/bin/go build
[ perf record: Woken up 2 times to write data ]
[ perf record: Captured and wrote 0.900 MB perf.data (22177 samples) ]
perf report --sort comm,dso
Samples: 22K of event 'cycles', Event count (approx.): 17031664028
Overhead Command Shared Object
66.28% 6g 6g
10.55% 6g [kernel.vmlinux]
9.88% 6l 6l
7.55% 6g libc-2.22.so
2.39% 6l libc-2.22.so
1.35% go go
1.09% 6l [kernel.vmlinux]
0.64% go [kernel.vmlinux]
0.07% 6g [ath9k]
0.06% 6g ld-2.22.so
0.05% 6g [mac80211]
0.03% 6g [ext4]
0.02% go [ext4]
0.01% go [jbd2]
0.01% 6g [ath9k_common]
0.00% go [mac80211]
0.00% 6g [ath9k_hw]
0.00% 6l ld-2.22.so
0.00% 6g [scsi_mod]
0.00% go [vdso]
去1.5
perf report
Samples: 72K of event 'cycles', Event count (approx.): 55631112443
Overhead Command Shared Object Symbol
6.59% compile compile [.] 0x000000000001bd6b
1.49% compile compile [.] 0x000000000001c2a4
1.00% compile compile [.] 0x000000000001bcf1
0.99% compile compile [.] 0x0000000000056bac
0.91% compile compile [.] 0x000000000001bcdd
0.80% compile compile [.] 0x0000000000010471
0.71% link link [.] 0x000000000001c51b
0.65% compile [kernel.vmlinux] [k] ioread32
0.54% compile compile [.] 0x000000000001bcad
0.49% compile [kernel.vmlinux] [k] page_fault
0.47% compile compile [.] 0x000000000001bc65
0.45% compile [kernel.vmlinux] [k] clear_page_c_e
0.43% compile compile [.] 0x000000000000f157
0.42% compile compile [.] 0x000000000001bcce
0.41% compile compile [.] 0x000000000001bc80
0.40% compile compile [.] 0x0000000000010462
0.39% compile compile [.] 0x000000000001c34c
0.38% compile compile [.] 0x000000000001bcd9
0.35% compile compile [.] 0x000000000001bcc2
0.33% compile compile [.] 0x000000000001bd5d
0.31% compile compile [.] 0x000000000001bcc7
0.31% compile compile [.] 0x0000000000056a74
0.30% compile compile [.] 0x000000000001bcb7
0.29% compile compile [.] 0x000000000001bc89
0.29% link link [.] 0x000000000001ca54
0.28% compile compile [.] 0x000000000001047e
0.28% compile compile [.] 0x000000000001bcfe
0.27% compile compile [.] 0x000000000001b834
0.27% link [kernel.vmlinux] [k] ioread32
0.26% compile compile [.] 0x000000000001bcb2
0.26% compile compile [.] 0x000000000001bc94
0.25% compile compile [.] 0x0000000000056b3b
0.25% compile compile [.] 0x000000000001bd4f
0.23% compile compile [.] 0x000000000001bd58
0.23% compile compile [.] 0x000000000001bc9c
0.21% compile compile [.] 0x000000000001bd54
0.21% compile compile [.] 0x00000000000b120f
0.20% compile compile [.] 0x000000000000f0e3
0.20% compile compile [.] 0x000000000001bd4a
0.19% compile compile [.] 0x000000000000b488
0.19% compile compile [.] 0x00000000000103e3
0.18% compile compile [.] 0x000000000001bd65
0.18% compile compile [.] 0x000000000001bc32
0.18% compile compile [.] 0x000000000002c835
0.18% compile compile [.] 0x000000000001bd76
0.18% compile compile [.] 0x00000000001740f2
0.18% compile compile [.] 0x0000000000056b12
0.17% compile compile [.] 0x0000000000057778
0.17% compile compile [.] 0x00000000001d5051
0.17% compile compile [.] 0x000000000001bcef
0.17% compile compile [.] 0x0000000000057282
0.16% compile [kernel.vmlinux] [k] __rmqueue
0.16% compile compile [.] 0x000000000001bc79
0.16% compile compile [.] 0x0000000000114d5a
0.16% compile compile [.] 0x000000000001be27
0.15% compile compile [.] 0x000000000001be36
0.15% compile compile [.] 0x000000000001bcbb
0.15% compile compile [.] 0x0000000000013c19
0.15% compile compile [.] 0x000000000001bf53
0.15% compile compile [.] 0x00000000000026cd
0.15% compile compile [.] 0x000000000005516e
0.15% compile compile [.] 0x000000000001bca6
0.14% compile compile [.] 0x000000000001bcf4
0.14% compile compile [.] 0x000000000001baa4
0.14% compile compile [.] 0x000000000002cff7
0.14% link link [.] 0x0000000000008171
0.14% compile compile [.] 0x000000000000f0b2