To be more precise here is gdb log generated by using Tanu's advise:
>gdb pulseaudio
>(gdb) run -vvv
><wait until log output stops>
>Ctrl-C
>(gdb) thread apply all bt
Thread 2 (Thread 0xb3455b40 (LWP 16627)):
#0 0xb7fdd416 in __kernel_vsyscall ()
#1 0xb7719db9 in ppoll () from /lib/libc.so.6
#2 0xb7f5a841 in pa_rtpoll_run (p=0x80ee800, wait_op=true)
at pulsecore/rtpoll.c:314
#3 0xb357d58e in thread_func (userdata=0x808dde0)
at modules/alsa/alsa-sink.c:1775
#4 0xb7d1e920 in internal_thread_func (userdata=0x80925e8)
at pulsecore/thread-posix.c:83
#5 0xb7832cd1 in start_thread () from /lib/libpthread.so.0
#6 0xb772307e in clone () from /lib/libc.so.6
Thread 1 (Thread 0xb75fe700 (LWP 16622)):
#0 0xb7fdd416 in __kernel_vsyscall ()
#1 0xb783a081 in pause () from /lib/libpthread.so.0
#2 0xb7b1f665 in sem_wait () from /usr/lib/libpthread-stubs.so.0
#3 0xb7d1f71d in pa_semaphore_wait (s=0x808dcc0)
at pulsecore/semaphore-posix.c:63
#4 0xb7f35954 in pa_asyncmsgq_send (a=0x80cdea8, object=0x80d5508, code=11,
userdata=0x1, offset=0, chunk=0x0) at pulsecore/asyncmsgq.c:167
#5 0xb7f754bb in sink_set_state (s=0x80d5508, state=PA_SINK_IDLE)
at pulsecore/sink.c:414
#6 0xb7f7686e in pa_sink_put (s=0x80d5508) at pulsecore/sink.c:646
#7 0xb357f948 in pa_alsa_sink_new (m=0x80b2dc0, ma=0x80b2d78,
driver=0xb35b55c8 "modules/alsa/module-alsa-card.c", card=0x80c2108,
mapping=0x808a0f0) at modules/alsa/alsa-sink.c:2340
#8 0xb35b3add in init_profile (u=0x808c070)
at modules/alsa/module-alsa-card.c:269
#9 0xb35b4ba7 in module_alsa_card_LTX_pa__init (m=0x80b2dc0)
at modules/alsa/module-alsa-card.c:534
#10 0xb7f4f4c9 in pa_module_load (c=0x807e150,
name=0xb35bc749 "module-alsa-card",
argument=0x80d53e0 "device_id=\"0\" name=\"platform-snd_dummy.0\"
card_name=\"alsa_card.platform-snd_dummy.0\" namereg_fail=false
tsched=yes fixed_latency_range=no ignore_dB=no deferred_volume=yes
card_properties=\"module-udev-"...)
at pulsecore/module.c:109
#11 0xb35ba7d6 in verify_access (u=0x8088098, d=0x80d3e28)
at modules/module-udev-detect.c:332
#12 0xb35bac90 in card_changed (u=0x8088098, dev=0x80b2170)
at modules/module-udev-detect.c:409
#13 0xb35bb06e in process_device (u=0x8088098, dev=0x80b2170)
at modules/module-udev-detect.c:451
#14 0xb35bb0f9 in process_path (u=0x8088098,
path=0x8089418 "/sys/devices/platform/snd_dummy.0/sound/card0")
at modules/module-udev-detect.c:468
#15 0xb35bc02f in module_udev_detect_LTX_pa__init (m=0x8086cd8)
at modules/module-udev-detect.c:761
#16 0xb7f4f4c9 in pa_module_load (c=0x807e150,
name=0x8086c38 "module-udev-detect", argument=0x0)
at pulsecore/module.c:109
#17 0xb7f39bf7 in pa_cli_command_load (c=0x807e150, t=0x8078dd0,
buf=0x8078be8, fail=0x807841d) at pulsecore/cli-command.c:435
#18 0xb7f40fd1 in pa_cli_command_execute_line_stateful (c=0x807e150,
s=0xbffff2fc "load-module module-udev-detect", buf=0x8078be8,
fail=0x807841d, ifstate=0xbffff2f8) at pulsecore/cli-command.c:1987
#19 0xb7f41251 in pa_cli_command_execute_file_stream (c=0x807e150,
f=0x8081008, buf=0x8078be8, fail=0x807841d) at pulsecore/cli-command.c:2027
#20 0x080559db in main (argc=2, argv=0xbffffd84) at daemon/main.c:1077
Kill the program being debugged? (y or n)
2012/11/22 Stefan Stefanov <selectany at gmail.com>:
> Even more details:
>
> void pa_semaphore_wait(pa_semaphore *s) {
> int ret;
> pa_assert(s);
>
> do {
> ret = sem_wait(&s->sem);
> } while (ret < 0 && errno == EINTR);
>
> pa_assert(ret == 0);
> }
>
> Actually it hangs in sem_wait(...) API function.
