Hi Godmar, I had the same symptom last week. It isn't always related to VDSO: http://marc.info/?l=user-mode-linux-user&m=118162823019693&w=2
That patch resolved my issue, but I don't want to send you barking up the wrong tree. This specific issue is supposed to be resolved by 2.6.21.5 (guest kernel) - your kernel is newer. Plus, my backtrace was different than yours. As a side note, while users reported the issue on FC6 and FC7, I experienced it on RHE5 - 2.6.18. Scott. -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Godmar Back Sent: Wednesday, September 05, 2007 9:13 AM To: user-mode-linux-user@lists.sourceforge.net Subject: Re: [uml-user] hang after VFS: Mounted root (ext3 filesystem) readonly. I'm not intending to swamp this list, but I'd like to share some additional information about this hanging issue. At the time when UML hangs, there are four threads running in it. Below, I'm providing their stack traces as shown by gdb: #0 0x081963eb in __nanosleep_nocancel () at include/net/sock.h:280 #1 0x08061a8d in idle_sleep (secs=10) at arch/um/os-Linux/time.c:100 #2 0x080528ac in default_idle () at arch/um/kernel/process.c:209 #3 0x08054fee in init_idle_skas () at arch/um/kernel/skas/process.c:143 #4 0x080528b7 in cpu_idle () at arch/um/kernel/process.c:215 #5 0x081dc277 in rest_init () at init/main.c:446 #6 0x080486a3 in start_kernel () at init/main.c:636 #7 0x0805501e in start_kernel_proc (unused=0x0) at arch/um/kernel/skas/process.c:160 #8 0x0806085d in run_kernel_thread (fn=0x8054ff2 <start_kernel_proc>, arg=0x0, jmp_ptr=0x825842c) at arch/um/os-Linux/process.c:295 #9 0x08054e25 in new_thread_handler () at arch/um/kernel/skas/process.c:64 #10 0x00000000 in ?? () -- #0 0x08198f52 in syscall () at include/net/sock.h:280 #1 0x0805e981 in io_getevents (ctx_id=9781248, min_nr=1, nr=1, events=0xbba8fac, timeout=0x0) at arch/um/os-Linux/aio.c:49 #2 0x0805ea97 in aio_thread (arg=0x0) at arch/um/os-Linux/aio.c:123 #3 0x08199a4e in clone () at include/net/sock.h:280 -- #0 0x081973ce in __read_nocancel () at include/net/sock.h:280 #1 0x0805f4dd in os_read_file (fd=5, buf=0xbba2fd0, len=4) at arch/um/os-Linux/file.c:295 #2 0x0805d076 in io_thread (arg=0x0) at arch/um/drivers/ubd_kern.c:1457 #3 0x08199a4e in clone () at include/net/sock.h:280 fd=5 is a Unix socket here. I assume this is the socket on which the ubd driver thread is stuck. It appears to use a request/reply scheme here and is waiting for the next request. -- #0 0x081978ab in poll () at include/net/sock.h:280 #1 0x0806088f in write_sigio_thread (unused=0x0) at arch/um/os-Linux/sigio.c:62 #2 0x08199a4e in clone () at include/net/sock.h:280 This thread is polling fd=13, which is also a Unix socket --- I should point out that this is with a statically linked linux. Here's its maps to show the vdso location (in case this has any bearing). 00110000-00111000 rwxp 00110000 00:00 0 001be000-001c1000 rwxp 001be000 00:00 0 002fb000-002fc000 r-xp 002fb000 00:00 0 [vdso] 08048000-08270000 rwxp 00000000 00:18 20759172 /home/staff/gback/cs5204proj1/linux/linux 08270000-08283000 rwxp 08270000 00:00 0 08ade000-08b00000 rwxp 08ade000 00:00 0 08b00000-0a8c5000 rwxs 00ab8000 00:13 1045481 /dev/shm/vm_file-6e6Cpy (deleted) 0b000000-0b001000 rwxs 00e65000 00:13 1045481 /dev/shm/vm_file-6e6Cpy (deleted) 0b001000-0b002000 rwxs 00e69000 00:13 1045481 /dev/shm/vm_file-6e6Cpy (deleted) and then many deleted regions where vm_file-6e6Cpy was mapped. Overall, this doesn't look related to VDSO to me. Could it be a bug in the IPC logic in UML, possibly related to the host's handling of SIGIO? Could somebody more knowledgeable with UML's implementation details shed a light on this? Which thread would be the one to send something into the Unix socket? Is there a debugging flag or mode to trace this issue down? Thanks, - Godmar ------------------------------------------------------------------------ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ User-mode-linux-user mailing list User-mode-linux-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ User-mode-linux-user mailing list User-mode-linux-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user
