Kernel Dump when accessing second drive

Dylan Reinhold dylan at ocnetworking.com
Sun May 16 23:07:47 PDT 2010


On 05/09/2010 12:15 AM, Dylan Reinhold wrote:
I added a second drive, to copy ~100GB's to my machine, but the kernel 
panics when I try and rsync the data.
The second drive is formated ext2 and mounted as readonly.

I walked the dump as much as I could (my thoughts are just a guess).

The dump is on leaf ~dylan/crash/*.11

Maybe the system is unable to get data from drive/controller.
Is the devstat structure mapped to the physical controller or the drive?
It seems that the kernel is unable to read from the memory mapped by 
that structure.

When its passed into getmicrouptime the memory seems to be gone.

(kgdb) f 9
#9  0xc02ff06c in getmicrouptime (tvp=0xa4) at 
/usr/src/sys/kern/kern_clock.c:972
972            tvp->tv_sec = gd->gd_time_seconds;
(kgdb) l 972
967    {
968        struct globaldata *gd = mycpu;
969        sysclock_t delta;
970
971        do {
972            tvp->tv_sec = gd->gd_time_seconds;
973            delta = gd->gd_hardclock.time - gd->gd_cpuclock_base;
974        } while (tvp->tv_sec != gd->gd_time_seconds);
975
976        if (delta >= sys_cputimer->freq) {
(kgdb) print tvp->tv_sec
Cannot access memory at address 0xa4
(kgdb) print &tvp->tv_sec
$1 = (time_t *) 0xa4

Dylan





More information about the Kernel mailing list