[Fixed] Chromium segfaults with glibc 2.15?

DarkSniper · 2012-01-05 06:55:48

After yesterday's upgrade and a reboot, I found that chromium was refusing to start from my panel shortcut. Tried launching it from a terminal and it simply printed "Segmentation Fault". I then checked dmesg, and it spit this at me:

[  410.814178] chromium[22417]: segfault at 17 ip 00007fc4624263ae sp 00007fc468539f10 error 6 in libc-2.15.so[7fc462311000+199000]

After clicking on the panel shortcut a few times, it'll eventually start up and work normally.

Versions of Stuff Involved:
chromium 16.0.912.77-1
glibc 2.15-4

Edit: This bug is fixed in glibc 2.15-5

Last edited by DarkSniper (2012-02-07 18:23:14)

skunktrader · 2012-01-05 08:13:59

I have this exact problem using 16.0.912.63

[ 4403.599527] chromium[18352]: segfault at 17 ip 00007f0647df13ae sp 00007f0631c303e0 error 6 in libc-2.15.so[7f0647cdc000+199000]
[ 4460.760180] chromium[18817]: segfault at 17 ip 00007fd5a10b23ae sp 00007fd5951e73e0 error 6 in libc-2.15.so[7fd5a0f9d000+199000]
[ 6565.261919] chromium[27027]: segfault at 17 ip 00007fc500b303ae sp 00007fc4f40543e0 error 6 in libc-2.15.so[7fc500a1b000+199000]
[ 6595.510474] chromium[27382]: segfault at 17 ip 00007f4d49ba43ae sp 00007f4d4fab23e0 error 6 in libc-2.15.so[7f4d49a8f000+199000]
[ 6668.539094] chromium[27938]: segfault at 17 ip 00007f8db1eee3ae sp 00007f8da541a3e0 error 6 in libc-2.15.so[7f8db1dd9000+199000]
[ 6878.079884] chromium[28939]: segfault at 17 ip 00007f06620953ae sp 00007f06555c0e20 error 6 in libc-2.15.so[7f0661f80000+199000]
[ 6930.551082] chromium[29500]: segfault at 17 ip 00007f08ca9823ae sp 00007f08bdeae3e0 error 6 in libc-2.15.so[7f08ca86d000+199000]
[ 7477.692003] chromium[32015]: segfault at 17 ip 00007f3a82e283ae sp 00007f3a88d363e0 error 6 in libc-2.15.so[7f3a82d13000+199000]
[ 7614.235645] chromium[603]: segfault at 17 ip 00007f4805cd73ae sp 00007f47f039b3e0 error 6 in libc-2.15.so[7f4805bc2000+199000]
[20440.171190] chromium[18075]: segfault at 17 ip 00007f53e114c3ae sp 00007f53e705a3e0 error 6 in libc-2.15.so[7f53e1037000+199000]

Allan · 2012-01-05 08:25:19

Post gdb output and what architecture you are using.

DarkSniper · 2012-01-05 08:36:23

Allan wrote:

Post gdb output and what architecture you are using.

I'm on x86_64, though I've never had to use gdb before. Would I follow the steps on the "Debug - Getting Traces" page of the wiki and build chromium with the debug symbols? [Since gdb quit saying that no debug symbols were found]

Last edited by DarkSniper (2012-01-05 08:37:27)

Allan · 2012-01-05 09:06:58

I would not worry about that if glibcis causing the crash. Just run "gdb chromium", type "run" and then "bt" after it crashes.

DarkSniper · 2012-01-05 09:23:18

Oh, I see. I thought that gdb required the debug symbols to work.

Had to use /usr/lib/chromium/chromium since gdb didn't want to load the wrapper script that's in /usr/bin. I got gdb to load chromium but I can't seem to replicate the crash through gdb. Every time I've tried it's loaded up just fine and has no stack when I type bt after closing it. Though the segfault still occurs in the terminal and from the panel plugin, happening immediately before the UI can even load.

Allan · 2012-01-05 09:27:26

Can you try "strace chromium" (with the strace package...) and see if you get the crash?

DarkSniper · 2012-01-05 09:46:44

That one caught the segfault.
Here's the output http://pastebin.com/KEEiuqkM

Execute_Method · 2012-01-06 04:26:49

Yes, I'm having the same trouble with the newest official chromium. x86_64, but I'm not at home right now. I was going crazy cause this is on my brand new system, and I have it running memtest right now while at work. I'll try to get a backtrace tomorrow.

Glad I'm not the only one.

Cheesecake Charlie · 2012-01-06 08:25:25

Having the same issue ... Very annoying

samuvuo · 2012-01-06 10:36:45

skunktrader wrote:

I have this exact problem using 16.0.912.63

Just noticed that pacman offers update to chromium-16.0.912.75-1. Does it fix this issue?

Last edited by samuvuo (2012-01-06 10:37:21)

Cheesecake Charlie · 2012-01-06 11:07:54

samuvuo wrote:

skunktrader wrote:
I have this exact problem using 16.0.912.63
Just noticed that pacman offers update to chromium-16.0.912.75-1. Does it fix this issue?

I have chromium-16.0.912.75-1 installed and still have the same problem

Lennie · 2012-01-06 14:26:23

For me Chromium worked fine until yesterday. After todays upgrade to Chromium 16.0.912.75-1 (64-bit), I get segmentation fault.

Hm strange, pacman.log says: [2012-01-04 21:04] upgraded glibc (2.14.1-4 -> 2.15-3) That's two days ago, and I used Chromium yesterday. (Only very short, just to do one thing that i couldn't do in Firefox. But then Chromium worked.) I tried to downgrade Chromium to version 16.0.912.63, but it didn't help, still 'segmentation fault'.

I also have Arch (64-bit) in Virtualbox, and I didn't update it yesterday. So there it has
Glibc 2.15-3
Chromium 16.0.912.63
and Chromium works fine. But when I start it from terminal it says:

[1:1:144245892:ERROR:nacl_fork_delegate_linux.cc(78)] Bad NaCl helper startup ack (0 bytes)

Last edited by Lennie (2012-01-06 14:27:46)

Execute_Method · 2012-01-06 18:06:46

Here is my gdb backtrace:
http://pastebin.com/T47j8f2H

Starting program: /usr/lib/chromium/chromium 
[Thread debugging using libthread_db enabled]
[New Thread 0x7fffed39b700 (LWP 2066)]
[New Thread 0x7fffecb9a700 (LWP 2067)]
[New Thread 0x7fffec399700 (LWP 2068)]
[New Thread 0x7fffeb552700 (LWP 2069)]
[New Thread 0x7fffead51700 (LWP 2070)]
[New Thread 0x7fffea550700 (LWP 2071)]
[New Thread 0x7fffe9d4f700 (LWP 2072)]
[New Thread 0x7fffe954e700 (LWP 2073)]
[New Thread 0x7ffff7fb9700 (LWP 2074)]
[New Thread 0x7fffe8d4d700 (LWP 2075)]
[New Thread 0x7fffe854c700 (LWP 2079)]
[New Thread 0x7ffff7f98700 (LWP 2078)]
[New Thread 0x7ffff7f77700 (LWP 2080)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff7fb9700 (LWP 2074)]
0x00007ffff208a3ae in __nscd_get_mapping () from /lib/libc.so.6
(gdb) bt
#0  0x00007ffff208a3ae in __nscd_get_mapping () from /lib/libc.so.6
#1  0x00007ffff208a574 in __nscd_get_map_ref () from /lib/libc.so.6
#2  0x00007ffff2087b62 in nscd_gethst_r () from /lib/libc.so.6
#3  0x00007ffff2088417 in __nscd_gethostbyname2_r () from /lib/libc.so.6
#4  0x00007ffff206dfb1 in gethostbyname2_r@@GLIBC_2.2.5 ()
   from /lib/libc.so.6
#5  0x00007ffff204442d in gaih_inet () from /lib/libc.so.6
#6  0x00007ffff20458be in getaddrinfo () from /lib/libc.so.6
#7  0x000055555668437c in ?? ()
#8  0x000055555667e745 in ?? ()
#9  0x00005555564368ae in ?? ()
#10 0x0000555557c5ccb1 in ?? ()
#11 0x0000555556437142 in ?? ()
#12 0x00007ffff4214e7a in start_thread () from /lib/libpthread.so.0
#13 0x00007ffff2057bad in clone () from /lib/libc.so.6
#14 0x0000000000000000 in ?? ()

Execute_Method · 2012-01-06 18:27:33

Also, here is an strace:
http://pastebin.com/FWBihfSQ

Ypnose · 2012-01-07 11:26:49

Got the same issue. I can reproduce it when I clear cache. Or when I start my computer and I launch Chromium for the first time.

u-ra · 2012-01-07 12:53:59

Got the same issue with the same backtrace. It seems to stop crashing if nscd is running.
Make sure you have /var/run/nscd and /var/db/nscd created for it to actually start.

jwhendy · 2012-01-12 16:28:52

Mine segfaulted for the first time this morning, though I've been living in Windows for the past few days, so I don't know when the issue came about. I tried the gdb trace mentioned by Alan above but couldn't get it to work. gdb responded that /usr/bin/chromium wasn't a readable binary. Quite perplexing. I switched to AUR's chromium-browser-bin and it's fine. Curious what's doing this...

DarkSniper · 2012-01-12 18:18:32

@jwhendy: That error occurs because /usr/bin/chromium is just a wrapper script, the actual executable lives in /usr/lib/chromium/chromium.
I'm pretty sure chromium-browser-bin uses continuous releases, so it might be using a newer version that no longer has this bug.
I'm currently building chromium-dev since the flash bug that I mentioned in my original post was fixed with version 18.x for the dev channel. Maybe it won't segfault since chromium-browser-bin doesn't.

Edit: Nope, current Dev build has the same issue as the stable, however, I was able to get a backtrace off of this one:

Starting program: /usr/lib/chromium-dev/chromium 
[Thread debugging using libthread_db enabled]
[New Thread 0x7fffec36d700 (LWP 19010)]
[New Thread 0x7fffebb6c700 (LWP 19011)]
[New Thread 0x7fffeb36b700 (LWP 19012)]
[19001:19010:8213703611:ERROR:object_proxy.cc(239)] Failed to call method: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.NetworkManager was not provided by any .service files
[New Thread 0x7fffea414700 (LWP 19013)]
[New Thread 0x7fffe9c13700 (LWP 19014)]
[New Thread 0x7fffe9412700 (LWP 19015)]
[New Thread 0x7fffe8c11700 (LWP 19016)]
[New Thread 0x7fffe8410700 (LWP 19017)]
[New Thread 0x7fffe7c0f700 (LWP 19018)]
[New Thread 0x7fffe740e700 (LWP 19019)]
[New Thread 0x7ffff7f9d700 (LWP 19020)]
[New Thread 0x7fffe6c0d700 (LWP 19023)]
[New Thread 0x7ffff7f7c700 (LWP 19024)]
[New Thread 0x7ffff7f5b700 (LWP 19025)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff7f9d700 (LWP 19020)]
0x00007ffff12733ae in __nscd_get_mapping () from /lib/libc.so.6
(gdb) bt
#0  0x00007ffff12733ae in __nscd_get_mapping () from /lib/libc.so.6
#1  0x00007ffff1273574 in __nscd_get_map_ref () from /lib/libc.so.6
#2  0x00007ffff1270b62 in nscd_gethst_r () from /lib/libc.so.6
#3  0x00007ffff1271417 in __nscd_gethostbyname2_r () from /lib/libc.so.6
#4  0x00007ffff1256fb1 in gethostbyname2_r@@GLIBC_2.2.5 () from /lib/libc.so.6
#5  0x00007ffff122d42d in gaih_inet () from /lib/libc.so.6
#6  0x00007ffff122e8be in getaddrinfo () from /lib/libc.so.6
#7  0x00005555566fdedc in ?? ()
#8  0x00005555566f8668 in ?? ()
#9  0x0000555557e08dba in ?? ()
#10 0x000055555648c3f2 in ?? ()
#11 0x00007ffff4227e7a in start_thread () from /lib/libpthread.so.0
#12 0x00007ffff1240bad in clone () from /lib/libc.so.6
#13 0x0000000000000000 in ?? ()

Last edited by DarkSniper (2012-01-13 04:03:22)

pedepy · 2012-01-14 14:00:36

dunno if it helps, but i've experienced the same crashes (or similar) using chrome, not chromium.

[  566.630889] chrome[1915]: segfault at b ip b346ff95 sp ae26b850 error 6 in libc-2.15.so[b3352000+19c000]

seems to happen mostly when i type something in the search/url bar, but it's happened a couple times "out of the blue" from no direct apparent action.i used to run the -dev build and tried the regular one but it still crashes. also, problems started before i updated to 2.15, which was just about a week ago.

ljdelight · 2012-01-15 09:51:03

Not sure if it helps, but here's a symbolized backtrace of Chromium 16.0.912.75 debug build. Also glibc 2.15-3.
http://pastebin.com/a7gnC1eG

(gdb) run
Starting program: /usr/lib/chromium/chromium 
[Thread debugging using libthread_db enabled]
[0115/032715:ERROR:logging_chrome.cc(325)] Unable to initialize logging to /usr/lib/chromium/chrome_debug.log: Permission denied
[0115/032715:ERROR:logging_chrome.cc(325)] Unable to initialize logging to /usr/lib/chromium/chrome_debug.log: Permission denied
[New Thread 0x7fffec110700 (LWP 1368)]
[New Thread 0x7fffeb90f700 (LWP 1369)]
[New Thread 0x7fffeb10e700 (LWP 1370)]
[New Thread 0x7fffea1c7700 (LWP 1371)]
[New Thread 0x7fffe99c6700 (LWP 1372)]
[New Thread 0x7fffe91c5700 (LWP 1373)]
[New Thread 0x7fffe89c4700 (LWP 1374)]
[New Thread 0x7fffe81c3700 (LWP 1375)]
[New Thread 0x7fffe79c2700 (LWP 1376)]
[New Thread 0x7fffe79a1700 (LWP 1379)]
[New Thread 0x7fffe7098700 (LWP 1380)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7fffe7098700 (LWP 1380)]
0x00007ffff209b3ae in __nscd_get_mapping () from /lib/libc.so.6
(gdb) backtrace
#0  0x00007ffff209b3ae in __nscd_get_mapping () from /lib/libc.so.6
#1  0x00007ffff2099518 in __nscd_get_nl_timestamp () from /lib/libc.so.6
#2  0x00007ffff2089678 in __check_pf () from /lib/libc.so.6
#3  0x00007ffff2056e65 in getaddrinfo () from /lib/libc.so.6
#4  0x0000555557376ad2 in net::SystemHostResolverProc (host=..., 
    address_family=net::ADDRESS_FAMILY_IPV4, host_resolver_flags=4, 
    addrlist=0x7fffe7097a60, os_error=0x7fffe7097a5c)
    at net/base/host_resolver_proc.cc:190
#5  0x00005555573651fd in net::ResolveAddrInfo (resolver_proc=0x0, host=..., 
    address_family=net::ADDRESS_FAMILY_IPV4, host_resolver_flags=4, 
    out=0x7fffe7097a60, os_error=0x7fffe7097a5c)
    at net/base/host_resolver_impl.cc:161
#6  0x000055555736a532 in net::HostResolverImpl::Job::DoLookup (
    this=0x7fffea82e320, start_time=..., attempt_number=1)
    at net/base/host_resolver_impl.cc:529
#7  0x00005555573764d1 in DispatchToMethod<net::HostResolverImpl::Job, void (net::HostResolverImpl::Job::*)(base::TimeTicks const&, unsigned int), base::TimeTicks, unsigned int> (obj=0x7fffea82e320, method=
    (void (net::HostResolverImpl::Job::*)(net::HostResolverImpl::Job * const, const base::TimeTicks &, unsigned int)) 0x55555736a4bc <net::HostResolverImpl::Job::DoLookup(base::TimeTicks const&, unsigned int)>, arg=...)
    at ./base/tuple.h:554
#8  0x00005555573760c4 in RunnableMethod<net::HostResolverImpl::Job, void (net::HostResolverImpl::Job::*)(base::TimeTicks const&, unsigned int), Tuple2<base::TimeTicks, unsigned int> >::Run (this=0x7fffe7181d40) at ./base/task.h:374
#9  0x0000555556f7a113 in base::subtle::TaskClosureAdapter::Run (
    this=0x7fffe717e360) at base/task.cc:71
#10 0x0000555556f3874f in base::internal::Invoker1<false, base::internal::InvokerStorage1<void (base::subtle::TaskClosureAdapter::*)(), base::subtle::TaskClosureAdapter*>, void (base::subtle::TaskClosureAdapter::*)()>::DoInvoke (
    base=0x7fffe717e390) at ./base/bind_internal.h:685
#11 0x00005555568c786b in base::Callback<void ()>::Run() const (
    this=0x7fffe7097c20) at ./base/callback.h:269
#12 0x00005555593cdaa6 in base::(anonymous namespace)::WorkerThread::ThreadMain
    (this=0x7fffea819160) at base/threading/worker_pool_posix.cc:94
#13 0x0000555556f7b867 in base::(anonymous namespace)::ThreadFunc (
    params=0x7fffea827aa0) at base/threading/platform_thread_posix.cc:54
#14 0x00007ffff4226e7a in start_thread () from /lib/libpthread.so.0
#15 0x00007ffff2068bad in clone () from /lib/libc.so.6
#16 0x0000000000000000 in ?? ()

mar04 · 2012-01-17 17:10:41

same here
64bit arch

jryan · 2012-01-17 17:11:36

Me too, 64 bit.

unexist · 2012-01-17 17:52:11

Same problem here, chromium usually refuses to start after boot. When I try to start chromium several times it works after e.g. the eighth try. Reinstalling chromium with pacman works as well. 64bit here as well.

Is there any bugreport?

jryan · 2012-01-17 17:53:05

unexist wrote:

Same problem here, chromium usually refuses to start after boot. When I try to start chromium several times it works after e.g. the eighth try. Reinstalling chromium with pacman works as well. 64bit here as well.
Is there any bugreport?

I looked and I didn't see one in Arch or Chromium project.

Arch Linux

#1 2012-01-05 06:55:48

[Fixed] Chromium segfaults with glibc 2.15?

#2 2012-01-05 08:13:59

Re: [Fixed] Chromium segfaults with glibc 2.15?

#3 2012-01-05 08:25:19

Re: [Fixed] Chromium segfaults with glibc 2.15?

#4 2012-01-05 08:36:23

Re: [Fixed] Chromium segfaults with glibc 2.15?

#5 2012-01-05 09:06:58

Re: [Fixed] Chromium segfaults with glibc 2.15?

#6 2012-01-05 09:23:18

Re: [Fixed] Chromium segfaults with glibc 2.15?

#7 2012-01-05 09:27:26

Re: [Fixed] Chromium segfaults with glibc 2.15?

#8 2012-01-05 09:46:44

Re: [Fixed] Chromium segfaults with glibc 2.15?

#9 2012-01-06 04:26:49

Re: [Fixed] Chromium segfaults with glibc 2.15?

#10 2012-01-06 08:25:25

Re: [Fixed] Chromium segfaults with glibc 2.15?

#11 2012-01-06 10:36:45

Re: [Fixed] Chromium segfaults with glibc 2.15?

#12 2012-01-06 11:07:54

Re: [Fixed] Chromium segfaults with glibc 2.15?

#13 2012-01-06 14:26:23

Re: [Fixed] Chromium segfaults with glibc 2.15?

#14 2012-01-06 18:06:46

Re: [Fixed] Chromium segfaults with glibc 2.15?

#15 2012-01-06 18:27:33

Re: [Fixed] Chromium segfaults with glibc 2.15?

#16 2012-01-07 11:26:49

Re: [Fixed] Chromium segfaults with glibc 2.15?

#17 2012-01-07 12:53:59

Re: [Fixed] Chromium segfaults with glibc 2.15?

#18 2012-01-12 16:28:52

Re: [Fixed] Chromium segfaults with glibc 2.15?

#19 2012-01-12 18:18:32

Re: [Fixed] Chromium segfaults with glibc 2.15?

#20 2012-01-14 14:00:36

Re: [Fixed] Chromium segfaults with glibc 2.15?

#21 2012-01-15 09:51:03

Re: [Fixed] Chromium segfaults with glibc 2.15?

#22 2012-01-17 17:10:41

Re: [Fixed] Chromium segfaults with glibc 2.15?

#23 2012-01-17 17:11:36

Re: [Fixed] Chromium segfaults with glibc 2.15?

#24 2012-01-17 17:52:11

Re: [Fixed] Chromium segfaults with glibc 2.15?

#25 2012-01-17 17:53:05

Re: [Fixed] Chromium segfaults with glibc 2.15?

Board footer