xmms segfaults on custom kernel based on kernel-sourcecode-2.6.7-1.476

[Lars]

hi

testing a kernel based on kernel-sourcecode-2.6.7-1.476
i get an segfault trying to run xmms-1.2.10-4
the kernel says:

Jul  8 18:00:49 bigbadmachine kernel: bad: xmms(4208) scheduling while
atomic (2)!
Jul  8 18:00:49 bigbadmachine kernel:  [<c02f84a6>] schedule+0x76/0x45d
Jul  8 18:00:49 bigbadmachine kernel:  [<c0113045>]
sys_sched_yield+0x3a/0x40
Jul  8 18:00:49 bigbadmachine kernel:  [<c014c4a7>] coredump_wait+0x2b/0x82
Jul  8 18:00:49 bigbadmachine kernel:  [<c014c5cd>] do_coredump+0xcf/0x1c3
Jul  8 18:00:49 bigbadmachine kernel:  [<c011c517>]
__dequeue_signal+0x132/0x13b
Jul  8 18:00:49 bigbadmachine kernel:  [<c011c535>] dequeue_signal+0x15/0x54
Jul  8 18:00:49 bigbadmachine kernel:  [<c011daba>]
get_signal_to_deliver+0x2cd/0x32b
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104dd2>] do_signal+0x4e/0xc6
Jul  8 18:00:49 bigbadmachine kernel:  [<c01384e3>] vma_link+0x45/0xd2
Jul  8 18:00:49 bigbadmachine kernel:  [<c0139047>]
do_mmap_pgoff+0x4b0/0x5d6
Jul  8 18:00:49 bigbadmachine kernel:  [<c0109c9a>] sys_mmap2+0x6e/0xa1
Jul  8 18:00:49 bigbadmachine kernel:  [<c0110b0c>] do_page_fault+0x0/0x47f
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104e71>]
do_notify_resume+0x27/0x36
Jul  8 18:00:49 bigbadmachine kernel:  [<c02fa482>] work_notifysig+0x13/0x15
Jul  8 18:00:49 bigbadmachine kernel: bad: xmms(4208) scheduling while
atomic (2)!
Jul  8 18:00:49 bigbadmachine kernel:  [<c02f84a6>] schedule+0x76/0x45d
Jul  8 18:00:49 bigbadmachine kernel:  [<c0111ff6>] activate_task+0x53/0x5e
Jul  8 18:00:49 bigbadmachine kernel:  [<c02f88d9>]
wait_for_completion+0x10/0xca
Jul  8 18:00:49 bigbadmachine kernel:  [<c011c58c>] signal_wake_up+0x18/0x19
Jul  8 18:00:49 bigbadmachine kernel:  [<c011c953>]
specific_send_sig_info+0x86/0x8d
Jul  8 18:00:49 bigbadmachine kernel:  [<c014c446>] zap_threads+0x68/0x9e
Jul  8 18:00:49 bigbadmachine kernel:  [<c014c4d7>] coredump_wait+0x5b/0x82
Jul  8 18:00:49 bigbadmachine kernel:  [<c014c5cd>] do_coredump+0xcf/0x1c3
Jul  8 18:00:49 bigbadmachine kernel:  [<c011c517>]
__dequeue_signal+0x132/0x13b
Jul  8 18:00:49 bigbadmachine kernel:  [<c011c535>] dequeue_signal+0x15/0x54
Jul  8 18:00:49 bigbadmachine kernel:  [<c011daba>]
get_signal_to_deliver+0x2cd/0x32b
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104dd2>] do_signal+0x4e/0xc6
Jul  8 18:00:49 bigbadmachine kernel:  [<c01384e3>] vma_link+0x45/0xd2
Jul  8 18:00:49 bigbadmachine kernel:  [<c0139047>]
do_mmap_pgoff+0x4b0/0x5d6
Jul  8 18:00:49 bigbadmachine kernel:  [<c0109c9a>] sys_mmap2+0x6e/0xa1
Jul  8 18:00:49 bigbadmachine kernel:  [<c0110b0c>] do_page_fault+0x0/0x47f
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104e71>]
do_notify_resume+0x27/0x36
Jul  8 18:00:49 bigbadmachine kernel:  [<c02fa482>] work_notifysig+0x13/0x15
Jul  8 18:00:49 bigbadmachine kernel: note: xmms[4208] exited with
preempt_count 1
Jul  8 18:00:49 bigbadmachine kernel: bad: xmms(4208) scheduling while
atomic (1)!
Jul  8 18:00:49 bigbadmachine kernel:  [<c02f84a6>] schedule+0x76/0x45d
Jul  8 18:00:49 bigbadmachine kernel:  [<c0115a91>]
release_console_sem+0x8b/0xc0
Jul  8 18:00:49 bigbadmachine kernel:  [<c0113926>] mm_release+0x47/0x7c
Jul  8 18:00:49 bigbadmachine kernel:  [<c0116f4e>] do_exit+0xcc/0x36d
Jul  8 18:00:49 bigbadmachine kernel:  [<c01172b7>] sys_exit_group+0x0/0xd
Jul  8 18:00:49 bigbadmachine kernel:  [<c011daf3>]
get_signal_to_deliver+0x306/0x32b
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104dd2>] do_signal+0x4e/0xc6
Jul  8 18:00:49 bigbadmachine kernel:  [<c01384e3>] vma_link+0x45/0xd2
Jul  8 18:00:49 bigbadmachine kernel:  [<c0139047>]
do_mmap_pgoff+0x4b0/0x5d6
Jul  8 18:00:49 bigbadmachine kernel:  [<c0109c9a>] sys_mmap2+0x6e/0xa1
Jul  8 18:00:49 bigbadmachine kernel:  [<c0110b0c>] do_page_fault+0x0/0x47f
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104e71>]
do_notify_resume+0x27/0x36
Jul  8 18:00:49 bigbadmachine kernel:  [<c02fa482>] work_notifysig+0x13/0x15
Jul  8 18:00:49 bigbadmachine kernel: bad: xmms(4208) scheduling while
atomic (1)!
Jul  8 18:00:49 bigbadmachine kernel:  [<c02f84a6>] schedule+0x76/0x45d
Jul  8 18:00:49 bigbadmachine kernel:  [<c0143c3f>] __fput+0xd1/0xf1
Jul  8 18:00:49 bigbadmachine kernel:  [<c011659c>]
put_files_struct+0x5b/0xae
Jul  8 18:00:49 bigbadmachine kernel:  [<c011705b>] do_exit+0x1d9/0x36d
Jul  8 18:00:49 bigbadmachine kernel:  [<c01172b7>] sys_exit_group+0x0/0xd
Jul  8 18:00:49 bigbadmachine kernel:  [<c011daf3>]
get_signal_to_deliver+0x306/0x32b
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104dd2>] do_signal+0x4e/0xc6
Jul  8 18:00:49 bigbadmachine kernel:  [<c01384e3>] vma_link+0x45/0xd2
Jul  8 18:00:49 bigbadmachine kernel:  [<c0139047>]
do_mmap_pgoff+0x4b0/0x5d6
Jul  8 18:00:49 bigbadmachine kernel:  [<c0109c9a>] sys_mmap2+0x6e/0xa1
Jul  8 18:00:49 bigbadmachine kernel:  [<c0110b0c>] do_page_fault+0x0/0x47f
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104e71>]
do_notify_resume+0x27/0x36
Jul  8 18:00:49 bigbadmachine kernel:  [<c02fa482>] work_notifysig+0x13/0x15
Jul  8 18:00:49 bigbadmachine kernel: bad: xmms(4208) scheduling while
atomic (1)!
Jul  8 18:00:49 bigbadmachine kernel:  [<c02f84a6>] schedule+0x76/0x45d
Jul  8 18:00:49 bigbadmachine kernel:  [<c0143c3f>] __fput+0xd1/0xf1
Jul  8 18:00:49 bigbadmachine kernel:  [<c011659c>]
put_files_struct+0x5b/0xae
Jul  8 18:00:49 bigbadmachine kernel:  [<c011705b>] do_exit+0x1d9/0x36d
Jul  8 18:00:49 bigbadmachine kernel:  [<c01172b7>] sys_exit_group+0x0/0xd
Jul  8 18:00:49 bigbadmachine kernel:  [<c011daf3>]
get_signal_to_deliver+0x306/0x32b
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104dd2>] do_signal+0x4e/0xc6
Jul  8 18:00:49 bigbadmachine kernel:  [<c01384e3>] vma_link+0x45/0xd2
Jul  8 18:00:49 bigbadmachine kernel:  [<c0139047>]
do_mmap_pgoff+0x4b0/0x5d6
Jul  8 18:00:49 bigbadmachine kernel:  [<c0109c9a>] sys_mmap2+0x6e/0xa1
Jul  8 18:00:49 bigbadmachine kernel:  [<c0110b0c>] do_page_fault+0x0/0x47f
Jul  8 18:00:49 bigbadmachine kernel:  [<c0104e71>]
do_notify_resume+0x27/0x36
Jul  8 18:00:49 bigbadmachine kernel:  [<c02fa482>] work_notifysig+0x13/0x15


guess it has something todo with the voluntary preemption patch?
the precompiled kernel works ok with xmms, and i have seen that
preempt is disabled and preemp_voluntary is enabled in this one.
i have both enabled in my custom config, and everything plays nice and
speedy so far, with the exception of xmms.
haven't found any info about preempt_voluntary on the web, so my question
is, can one use preempt and preempt_voluntary together, or
is it a silly idea?


best
lars