On Sex, 2004-05-28 at 17:04 +0100, Nuno Ferreira wrote:
> OK, apparently now the cp never finishes and nothing gets written to
> disk, so i could not get the full output from sysrq-t.
> I wrote the call stack from the process, it that enough?
> The call stack from the cp process:
> __make_request
> schedule_timeout
> process_timeout
> io_schedule_timeout
> blk_congestion_wait
> autoremove_wake_function
> autoremove_wake_function
> get_dirty_limits
> balance_dirty_pages
> generic_file_aio_write_nolock
> __copy_to_user_ll
> file_read_actor
> __generic_file_aio_read
> file_read_actor
> generic_file_write_nolock
> do_sync_read
> do_IRQ
> common_interrupt
> generic_file_write
> cifs_write_wrapper
> do_sync_read
> vfs_write
> sys_write
> syscall_call
>
> Also the call from a vi process that was called after the cp reached
> that state and was also blocked
> schedule_timeout
> ext3_mark_inode_dirty
> process_timeout
> io_schedule_timeout
> blk_congestion_wait
> autoremove_wake_function
> get_dirty_limits
> balance_dirty_pages
> generic_file_aio_write_nolock
> buffered_rmqueue
> __alloc_pages
> generic_file_aio_write
> ext2_file_write
> do_sync_write
> do_lookup
> path_release
> sys_access
> vfs_write
> sys_write
> syscall_call
>
> If more detail is needed I can try to capture it but syslog not writing
> to disk I don't know how to capture the sysrq-t output
Hi, is the above information enough to try and find the problem or is
there a way that I can capture the full sysrq-t output?
--
Nuno Ferreira