Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

runtime/pprof: hangs when using pprof on a CPU-intensive program on OS X #5519

Closed
gopherbot opened this issue May 20, 2013 · 23 comments
Closed

runtime/pprof: hangs when using pprof on a CPU-intensive program on OS X #5519

gopherbot opened this issue May 20, 2013 · 23 comments
Labels
FrozenDueToAge
Milestone
Go1.2

Comments

@gopherbot
Copy link

by akalin:

What steps will reproduce the problem?
If possible, include a link to a program on play.golang.org.
1. Download aks_debug.go from http://play.golang.org/p/SUNgPMcfIM or
https://github.com/akalin/aks-go/blob/debug/aks_debug.go
2. build (tested with go 1.1):

go build aks_debug.go

3. the most reliable way to repro would be to run:

for i in $(seq 1 300); do (env GOGCTRACE=1 ./aks_debug >&/tmp/aks_debug.$i
&); done

You'd expect all programs to terminate, but chances are, when your computer settles
down, there will be one or two
aks_debug processes still hanging around. Check with ps. For example:

  PID TTY           TIME CMD
23780 ttys000    0:00.01 -bash
23826 ttys000    0:00.00 tmux att
26765 ttys001    0:02.11 -bash
27792 ttys001    0:02.04 ./aks_debug
27904 ttys001    0:02.09 ./aks_debug
17273 ttys002    0:00.04 -bash
 6770 ttys003    0:00.83 -bash

Check which output files correspond to those processes:
$ lsof /tmp/aks_debug.*
COMMAND     PID   USER   FD   TYPE DEVICE SIZE/OFF    NODE NAME
aks_debug 27792 akalin    1w   REG    1,4     2404 9553026
/private/tmp/aks_debug.5
aks_debug 27792 akalin    2w   REG    1,4     2404 9553026
/private/tmp/aks_debug.5
aks_debug 27904 akalin    1w   REG    1,4     2404 9553082
/private/tmp/aks_debug.61
aks_debug 27904 akalin    2w   REG    1,4     2404 9553082
/private/tmp/aks_debug.61

Then send a SIGQUIT to one of those to get the output:

kill -s SIGQUIT 27792
less /tmp/aks_debug.5

For example, I get the below log output. Looks like it's stuck on a
write syscall for the pprof output?

Let me know if I can do anything more.

-- START aks_debug.5 --

gc1(1): 0+0+0 ms, 0 -> 0 MB 20 -> 22 (23-1) objects, 0(0) handoff, 0(0) steal, 0
/0/0 yields
2013/05/14 02:33:47 0: multiplying...
2013/05/14 02:33:47 0: multiplying done; shifting...
2013/05/14 02:33:47 0: not shifting
2013/05/14 02:33:47 1: multiplying...
2013/05/14 02:33:47 1: multiplying done; shifting...
2013/05/14 02:33:47 1: not shifting
2013/05/14 02:33:47 2: multiplying...
2013/05/14 02:33:47 2: multiplying done; shifting...
2013/05/14 02:33:47 2: not shifting
2013/05/14 02:33:47 3: multiplying...
2013/05/14 02:33:47 3: multiplying done; shifting...
2013/05/14 02:33:47 3: not shifting
2013/05/14 02:33:47 4: multiplying...
2013/05/14 02:33:47 4: multiplying done; shifting...
2013/05/14 02:33:47 4: not shifting
2013/05/14 02:33:47 5: multiplying...
2013/05/14 02:33:47 5: multiplying done; shifting...
2013/05/14 02:33:47 5: not shifting
2013/05/14 02:33:47 6: multiplying...
2013/05/14 02:33:47 6: multiplying done; shifting...
2013/05/14 02:33:47 6: not shifting
2013/05/14 02:33:47 7: multiplying...
2013/05/14 02:33:47 7: multiplying done; shifting...
2013/05/14 02:33:47 7: not shifting
2013/05/14 02:33:47 8: multiplying...
2013/05/14 02:33:47 8: multiplying done; shifting...
2013/05/14 02:33:47 8: not shifting
2013/05/14 02:33:47 9: multiplying...
gc2(1): 0+0+0 ms, 0 -> 0 MB 206 -> 144 (229-85) objects, 0(0) handoff, 0(0) stea
l, 0/0/0 yields
2013/05/14 02:33:47 9: multiplying done; shifting...
2013/05/14 02:33:47 9: not shifting
2013/05/14 02:33:47 10: multiplying...
2013/05/14 02:33:47 10: multiplying done; shifting...
2013/05/14 02:33:47 10: not shifting
2013/05/14 02:33:47 11: multiplying...
gc3(1): 0+0+0 ms, 1 -> 1 MB 154 -> 144 (240-96) objects, 0(0) handoff, 0(0) stea
l, 0/0/0 yields
2013/05/14 02:33:47 11: multiplying done; shifting...
2013/05/14 02:33:47 11: not shifting
2013/05/14 02:33:47 12: multiplying...
gc4(1): 0+0+0 ms, 2 -> 2 MB 150 -> 145 (247-102) objects, 0(0) handoff, 0(0) ste
al, 0/0/0 yields
2013/05/14 02:33:48 12: multiplying done; shifting...
2013/05/14 02:33:48 12: not shifting
2013/05/14 02:33:48 13: multiplying...
gc5(1): 0+0+0 ms, 4 -> 3 MB 150 -> 145 (253-108) objects, 0(0)
handoff, 0(0) steal, 0/0/0 yields
2013/05/14 02:34:00 13: multiplying done; shifting...
2013/05/14 02:34:00 13: not shifting
2013/05/14 02:34:00 14: multiplying...
gc6(1): 0+0+0 ms, 7 -> 6 MB 149 -> 144 (258-114) objects, 0(0) handoff, 0(0) ste
al, 0/0/0 yields
SIGQUIT: quit
PC=0x229fb

Which compiler are you using (5g, 6g, 8g, gccgo)?

Whatever go build on go 1.1 on x86 OS X uses.

Which operating system are you using?

OS X

Which version are you using?  (run 'go version')

go 1.1 (homebrew)
@gopherbot
Copy link
Author

Comment 1 by akalin:

How long does it usually take for a bug to be triaged?

@minux
Copy link
Member

minux commented May 28, 2013

Comment 2:

couldn't reproduce on Mac OS X 10.6 with Go tip.
after the process hangs, and you sent a SIGQUIT to it, what does it print?

@gopherbot
Copy link
Author

Comment 3 by akalin:

Sorry, forgot to specify OS X version. I'm running 10.8.3.
Also forgot to put the output after the SIGQUIT.  Here's an example one:
SIGQUIT: quit
PC=0x229fb
goroutine 1 [chan receive]:
main.main()
        /Users/akalin/src/aks/aks_debug.go:46 +0x141
goroutine 3 [syscall]:
runtime/pprof.profileWriter(0x2102391b0, 0x2101d8020)
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/runtime/pprof/pprof.go:600
+0x1c
created by runtime/pprof.StartCPUProfile
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/runtime/pprof/pprof.go:594
+0x16b
goroutine 4 [running]:
syscall.Syscall()
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/syscall/asm_darwin_amd64.s:15
+0x5
syscall.write(0x2, 0x21023a240, 0x27, 0x40, 0x27, ...)
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/syscall/zsyscall_darwin_amd64.go:1274
+0x70
syscall.Write(0x14001, 0x21023a240, 0x27, 0x40, 0x37a28, ...)
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/syscall/syscall_unix.go:143
+0x5a
os.(*File).write(0x13b06, 0x210248838, 0x8, 0x4ff1d, 0x2102360f0, ...)
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/os/file_unix.go:188
+0x79
os.(*File).Write(0x2102360f0, 0x21023a240, 0x27, 0x21049b000, 0x27, ...)
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/os/file.go:144 +0xdb
log.(*Logger).Output(0x3c004, 0x21049b000, 0x14001, 0x3c004, 0x0, ...)
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/log/log.go:155 +0x34f
runtime/pprof.writeHeap(0xa0b00000030, 0x0, 0x4000000, 0xd5a000166a6,
0x21025a000, ...)
        /Users/akalin/homebrew/Cellar/go/1.1rc3/src/pkg/runtime/pprof/pprof.go:456
+0xd24
created by main.main
        /Users/akalin/src/aks/aks_debug.go:44 +0x128
rax     0xe
rbx     0x1831d8
rcx     0x7fff5fbff770
rdx     0x0
rdi     0xc07
rsi     0x175e90
rbp     0xffffffffffffffff
rsp     0x7fff5fbff770
r8      0x33359d
r9      0x1fe
r10     0x3488dcb46
r11     0x286
r12     0x2c227d9be987d
r13     0x2c2f0a6a92d90
r14     0x12fdf96808336600
r15     0x0
rip     0x229fb
rflags  0x286
cs      0x7
fs      0x0
gs      0x0

@gopherbot
Copy link
Author

Comment 4 by akalin:

IIRC, pprof on go has problems on OS X > 10.6, right? Could be why it's not reproing on
10.6.

@davecheney
Copy link
Contributor

Comment 5:

To the best of my knowledge, runtime/pprof returns unreliable trace data on all
supported versions of OS X, although there is some suggestion that 10.8.3 has fixed the
problem. Save this, there are no other reports of pprof hanging Go programs on OS X.

@davecheney
Copy link
Contributor

Comment 6:

Could you please try this (slightly simplified) example
http://play.golang.org/p/zrrv9wEuRj on your machine using tip.

Labels changed: added os-macosx.

Status changed to WaitingForReply.

@gopherbot
Copy link
Author

Comment 7 by akalin:

Okay, I downloaded the go source using the instructions on
http://golang.org/doc/install/source, moved to the default branch, and built go.
I then compiled and followed the repro steps above with my program as well as your
program with go built from tip.
The problem reproduces with my program, but not with your program. This is consistent
with my earlier experience of trying to find a minimal test case -- I found that
removing the channel made the program go away.
Let me know if I can do anything else.

@davecheney
Copy link
Contributor

Comment 8:

Could you please try again with my version so we get the logging when the profile dump
starts and finishes, but add back your channel bits. As an experiment (because I don't
know how else to capture this information), also export GOGCTRACE=1. 
If the program hangs with the channel added, just for arguments sake, please hit it with
a SIGQUIT and post the results.
My working theory at the moment is gc is blocking progress of the profile dump. It isn't
a great theory.

@gopherbot
Copy link
Author

Comment 9 by akalin:

Hunh, I managed to get it to repro with your version!  Here's the output with
GOGCTRACE=1 and me killing it with SIGQUIT:
gc1(1): 0+0+0 ms, 0 -> 0 MB 1170 -> 1171 (1173-2) objects, 0(0) handoff, 0(0) steal,
0/0/0 yields
2013/05/29 03:53:24 0: multiplying...
2013/05/29 03:53:24 0: multiplying done; shifting...
2013/05/29 03:53:24 0: not shifting
2013/05/29 03:53:24 1: multiplying...
2013/05/29 03:53:24 1: multiplying done; shifting...
2013/05/29 03:53:24 1: not shifting
2013/05/29 03:53:24 2: multiplying...
2013/05/29 03:53:24 2: multiplying done; shifting...
2013/05/29 03:53:24 2: not shifting
2013/05/29 03:53:24 3: multiplying...
2013/05/29 03:53:24 3: multiplying done; shifting...
2013/05/29 03:53:24 3: not shifting
2013/05/29 03:53:24 4: multiplying...
2013/05/29 03:53:24 4: multiplying done; shifting...
2013/05/29 03:53:24 4: not shifting
2013/05/29 03:53:24 5: multiplying...
2013/05/29 03:53:24 5: multiplying done; shifting...
2013/05/29 03:53:24 5: not shifting
2013/05/29 03:53:24 6: multiplying...
2013/05/29 03:53:24 6: multiplying done; shifting...
2013/05/29 03:53:24 6: not shifting
2013/05/29 03:53:24 7: multiplying...
2013/05/29 03:53:24 7: multiplying done; shifting...
2013/05/29 03:53:24 7: not shifting
2013/05/29 03:53:24 8: multiplying...
2013/05/29 03:53:24 8: multiplying done; shifting...
2013/05/29 03:53:24 8: not shifting
2013/05/29 03:53:24 9: multiplying...
gc2(1): 0+0+0 ms, 0 -> 0 MB 1354 -> 1292 (1378-86) objects, 0(0) handoff, 0(0) steal,
0/0/0 yields
2013/05/29 03:53:24 9: multiplying done; shifting...
2013/05/29 03:53:24 9: not shifting
2013/05/29 03:53:24 10: multiplying...
2013/05/29 03:53:24 10: multiplying done; shifting...
2013/05/29 03:53:24 10: not shifting
2013/05/29 03:53:24 11: multiplying...
gc3(1): 0+0+0 ms, 1 -> 1 MB 1302 -> 1292 (1389-97) objects, 0(0) handoff, 0(0) steal,
0/0/0 yields
2013/05/29 03:53:26 11: multiplying done; shifting...
2013/05/29 03:53:26 11: not shifting
2013/05/29 03:53:26 12: multiplying...
2013/05/29 03:53:30 12: multiplying done; shifting...
2013/05/29 03:53:30 12: not shifting
2013/05/29 03:53:30 13: multiplying...
gc4(1): 0+0+0 ms, 4 -> 3 MB 1302 -> 1293 (1400-107) objects, 0(0) handoff, 0(0) steal,
0/0/0 yields
2013/05/29 03:53:42 13: multiplying done; shifting...
2013/05/29 03:53:42 13: not shifting
2013/05/29 03:53:42 14: multiplying...
gc5(1): 0+0+0 ms, 7 -> 7 MB 1298 -> 1293 (1406-113) objects, 0(0) handoff, 0(0) steal,
0/0/0 yields
SIGQUIT: quit
PC=0x22b0b
runtime/pprof.profileWriter(0x21025c1e0, 0x21023a408)
    /Users/akalin/src/go.repo/src/pkg/runtime/pprof/pprof.go:600 +0x1c
created by runtime/pprof.StartCPUProfile
    /Users/akalin/src/go.repo/src/pkg/runtime/pprof/pprof.go:594 +0x16b
goroutine 1 [running]:
syscall.Syscall()
    /Users/akalin/src/go.repo/src/pkg/syscall/asm_darwin_amd64.s:15 +0x5
syscall.write(0x2, 0x21025d200, 0x27, 0x40, 0x27, ...)
    /Users/akalin/src/go.repo/src/pkg/syscall/zsyscall_darwin_amd64.go:1274 +0x70
syscall.Write(0x14001, 0x21025d200, 0x27, 0x40, 0x37ae8, ...)
    /Users/akalin/src/go.repo/src/pkg/syscall/syscall_unix.go:143 +0x5a
os.(*File).write(0x13b66, 0x21026b038, 0x8, 0x5024d, 0x2102590f0, ...)
    /Users/akalin/src/go.repo/src/pkg/os/file_unix.go:188 +0x79
os.(*File).Write(0x2102590f0, 0x21025d200, 0x27, 0x2104b9000, 0x27, ...)
    /Users/akalin/src/go.repo/src/pkg/os/file.go:144 +0xdb
log.(*Logger).Output(0x3c004, 0x2104b9000, 0x14001, 0x3c004, 0x0, ...)
    /Users/akalin/src/go.repo/src/pkg/log/log.go:155 +0x33a
rax     0xe
rbx     0x221034f008
rcx     0x221024ce88
rdx     0x0
rdi     0xb07
rsi     0x18cef0
rbp     0xffffffffffffffff
rsp     0x221024ce88
r8      0x1000000
r9      0x0
r10     0x2101ef900
r11     0x286
r12     0x62681980af85
r13     0x626bdafe1f70
r14     0x1302986693559e00
r15     0x109b70
rip     0x22b0b
rflags  0x286
cs      0x7
fs      0x0
gs      0x0

@davecheney
Copy link
Contributor

Comment 10:

That isn't my version, where is the "Starting profile dump..." message ?

@gopherbot
Copy link
Author

Comment 11 by akalin:

It's after doWork(), but doWork() never finishes. The hang always happens in doWork().

@gopherbot
Copy link
Author

Comment 12 by akalin:

Note that the loop in doWork is from 0 to 14, and it's hanging in the last iteration
when doing the multiply.

@davecheney
Copy link
Contributor

Comment 13:

Can you please post the current test code on play.golang.org so we're both working from
the same page.

@davecheney
Copy link
Contributor

Comment 14:

Also, try not running your program under tmux (just a hunch)

@gopherbot
Copy link
Author

Comment 15 by akalin:

It's exactly the link you posted above: http://play.golang.org/p/zrrv9wEuRj

@gopherbot
Copy link
Author

Comment 16 by akalin:

Also, hangs seem to happen even without running from within tmux.

@davecheney
Copy link
Contributor

Comment 17:

I don't understand. That SIGQUIT stack trace is completely screwed. Where is the
doWork() goroutine ?
Can you please rerun it with GOTRACEBACK=2

@gopherbot
Copy link
Author

Comment 18 by akalin:

Unfortunately, your version seems to repro the hang quite rarely compared to mine.
I combined the versions in #8 to get http://play.golang.org/p/GOq8WP1cxN
Here's an example log for that one with a hang and a SIGQUIT:
gc1(1): 0+0+0 ms, 0 -> 0 MB 1170 -> 1171 (1173-2) objects, 0(0) handoff, 0(0) steal,
0/0/0 yields
2013/05/29 04:43:11 0: multiplying...
2013/05/29 04:43:11 0: multiplying done; shifting...
2013/05/29 04:43:11 0: not shifting
2013/05/29 04:43:11 1: multiplying...
2013/05/29 04:43:11 1: multiplying done; shifting...
2013/05/29 04:43:11 1: not shifting
2013/05/29 04:43:11 2: multiplying...
2013/05/29 04:43:11 2: multiplying done; shifting...
2013/05/29 04:43:11 2: not shifting
2013/05/29 04:43:11 3: multiplying...
2013/05/29 04:43:11 3: multiplying done; shifting...
2013/05/29 04:43:11 3: not shifting
2013/05/29 04:43:11 4: multiplying...
2013/05/29 04:43:11 4: multiplying done; shifting...
2013/05/29 04:43:11 4: not shifting
2013/05/29 04:43:11 5: multiplying...
2013/05/29 04:43:11 5: multiplying done; shifting...
2013/05/29 04:43:11 5: not shifting
2013/05/29 04:43:11 6: multiplying...
2013/05/29 04:43:11 6: multiplying done; shifting...
2013/05/29 04:43:11 6: not shifting
2013/05/29 04:43:11 7: multiplying...
2013/05/29 04:43:11 7: multiplying done; shifting...
2013/05/29 04:43:11 7: not shifting
2013/05/29 04:43:11 8: multiplying...
2013/05/29 04:43:11 8: multiplying done; shifting...
2013/05/29 04:43:11 8: not shifting
2013/05/29 04:43:11 9: multiplying...
gc2(1): 0+0+0 ms, 0 -> 0 MB 1358 -> 1297 (1379-82) objects, 0(0) handoff, 0(0) steal,
0/0/0 yields
2013/05/29 04:43:11 9: multiplying done; shifting...
2013/05/29 04:43:11 9: not shifting
2013/05/29 04:43:11 10: multiplying...
2013/05/29 04:43:11 10: multiplying done; shifting...
2013/05/29 04:43:11 10: not shifting
2013/05/29 04:43:11 11: multiplying...
gc3(1): 0+0+0 ms, 1 -> 1 MB 1307 -> 1296 (1390-94) objects, 0(0) handoff, 0(0) steal,
0/0/0 yields
2013/05/29 04:43:12 11: multiplying done; shifting...
2013/05/29 04:43:12 11: not shifting
2013/05/29 04:43:12 12: multiplying...
SIGQUIT: quit
PC=0x22b6b
runtime.mach_semaphore_wait()
    /Users/akalin/src/go.repo/src/pkg/runtime/sys_darwin_amd64.s:391 +0xb
runtime.mach_semacquire(0x2200000a0b, 0xffffffffffffffff)
    /Users/akalin/src/go.repo/src/pkg/runtime/os_darwin.c:426 +0xbe
runtime.semasleep(0xffffffffffffffff)
    /Users/akalin/src/go.repo/src/pkg/runtime/os_darwin.c:32 +0x58
runtime.notesleep(0x198398)
    /Users/akalin/src/go.repo/src/pkg/runtime/lock_sema.c:159 +0xba
stopm()
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:799 +0xdf
exitsyscall0(0x2101ef400)
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:1423 +0xc8
runtime.mcall()
    /Users/akalin/src/go.repo/src/pkg/runtime/asm_amd64.s:195 +0x49
goroutine 1 [chan receive]:
runtime.park(0xb6b0, 0x21026e0b0, 0x191c35)
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:1179 +0x64
runtime.chanrecv(0xbd5a0, 0x21026e060, 0x221023eed0, 0x0, 0x0, ...)
    /Users/akalin/src/go.repo/src/pkg/runtime/chan.c:366 +0x544
runtime.chanrecv1()
    /Users/akalin/src/go.repo/src/pkg/runtime/chan.c:458 +0x38
main.main()
    /tmp/aks_debug-3.go:45 +0x150
runtime.main()
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:181 +0x92
runtime.goexit()
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:1227
goroutine 2 [runnable]:
runtime.exitsyscall()
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:1395 +0x13f
runtime.MHeap_Scavenger()
    /Users/akalin/src/go.repo/src/pkg/runtime/mheap.c:475 +0xf0
runtime.goexit()
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:1227
created by runtime.main
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:164
goroutine 3 [syscall]:
runtime.entersyscallblock()
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:1338 +0x16e
getprofile(0x221024cf68, 0x221034f000)
    /Users/akalin/src/go.repo/src/pkg/runtime/cpuprof.c:362 +0x367
runtime.CPUProfile(0x0, 0x0, 0x0)
    /Users/akalin/src/go.repo/src/pkg/runtime/cpuprof.c:436 +0x32
runtime/pprof.profileWriter(0x21025c1e0, 0x21023a408)
    /Users/akalin/src/go.repo/src/pkg/runtime/pprof/pprof.go:600 +0x1c
runtime.goexit()
    /Users/akalin/src/go.repo/src/pkg/runtime/proc.c:1227
created by runtime/pprof.StartCPUProfile
    /Users/akalin/src/go.repo/src/pkg/runtime/pprof/pprof.go:594 +0x16b
goroutine 4 [running]:
syscall.Syscall()
    /Users/akalin/src/go.repo/src/pkg/syscall/asm_darwin_amd64.s:15 +0x5
syscall.write(0x2, 0x21025d200, 0x27, 0x40, 0x27, ...)
    /Users/akalin/src/go.repo/src/pkg/syscall/zsyscall_darwin_amd64.go:1274 +0x70
syscall.Write(0x5001, 0x21025d200, 0x27, 0x40, 0x37b48, ...)
    /Users/akalin/src/go.repo/src/pkg/syscall/syscall_unix.go:143 +0x5a
os.(*File).write(0x13bc6, 0x21026b838, 0x8, 0x502ad, 0x2102590f0, ...)
    /Users/akalin/src/go.repo/src/pkg/os/file_unix.go:188 +0x79
os.(*File).Write(0x2102590f0, 0x21025d200, 0x27, 0x210332000, 0x27, ...)
    /Users/akalin/src/go.repo/src/pkg/os/file.go:144 +0xdb
log.(*Logger).Output(0xf004, 0x210332000, 0x5001, 0xf004, 0x0, ...)
    /Users/akalin/src/go.repo/src/pkg/log/log.go:155 +0x33a
runtime.nilintercopy()
    /Users/akalin/src/go.repo/src/pkg/runtime/alg.c:409 +0x11
handoff(0x2170)
    /Users/akalin/src/go.repo/src/pkg/runtime/mgc0.c:1361 +0x74
created by main.main
    /tmp/aks_debug-3.go:43 +0x137
rax     0xe
rbx     0x198398
rcx     0x7fff5fbff840
rdx     0x0
rdi     0xa0b
rsi     0x18cef0
rbp     0xffffffffffffffff
rsp     0x7fff5fbff840
r8      0x2fd2a
r9      0x37
r10     0x39ea8cec
r11     0x286
r12     0x62681980af85
r13     0x65316ef11058
r14     0x13029b2c3e4aae00
r15     0x0
rip     0x22b6b
rflags  0x286
cs      0x7
fs      0x0
gs      0x0

@gopherbot
Copy link
Author

Comment 19 by akalin:

Forgot to mention that I did run it with GOTRACEBACK=2

@rsc
Copy link
Contributor

rsc commented Jul 30, 2013

Comment 20:

It's unclear what to do here. Eventually I think we're going to have to stop supporting
profiling on OS X. Apple has shown no interest in fixing the kernel bug, and working
around it gets harder and harder.

Labels changed: added priority-later, go1.2, removed priority-triage.

Status changed to Accepted.

@rsc
Copy link
Contributor

rsc commented Jul 30, 2013

Comment 21:

Labels changed: added feature.

@davecheney
Copy link
Contributor

Comment 22:

As regrettable as that sounds, I think it is the only sensible option. Explaining to
folks that OS X profiling is subtly (or grossly) broken is a constant support cost. 
I don't have any experience the Xcode profilers, but ill see if a go.test binary can be
run under Instruments

@rsc
Copy link
Contributor

rsc commented Aug 5, 2013

Comment 23:

This issue was closed by revision d3066e4.

Status changed to Fixed.

@rsc rsc added this to the Go1.2 milestone Apr 14, 2015
@rsc rsc removed the go1.2 label Apr 14, 2015
@golang golang locked and limited conversation to collaborators Jun 24, 2016
This issue was closed.
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees
No one assigned
Labels
FrozenDueToAge
Projects
None yet
Milestone
Go1.2
Development

No branches or pull requests
4 participants
@davecheney @rsc @minux @gopherbot