I am running Debian Potato with kernel 2.4.27-rc3 on a Sparc Server 20, 128MB RAM machine. After few hours I get the following error in the syslog kernel file Jul 15 01:09:25 mudshark kernel: Assertion failure in journal_stop() at transaction.c:1418: "journal_current_handle() == handle" Jul 15 01:09:25 mudshark kernel: kernel BUG at transaction.c:1418! Jul 15 01:09:25 mudshark kernel: Unable to handle kernel NULL pointer dereference Jul 15 01:09:25 mudshark kernel: tsk->{mm,active_mm}->context = 000000dd Jul 15 01:09:25 mudshark kernel: tsk->{mm,active_mm}->pgd = fc027800 Is this a Kernel Bug or a Ext3 Bug or an application problem or anything else? -- Regards, Ilya Balashov AnimaTel, Inc.
On Fri, Jul 16, 2004 at 05:41:19AM +0400, Ilya Balashov wrote:> I am running Debian Potato with kernel 2.4.27-rc3 on a Sparc Server 20, > 128MB RAM machine. > > After few hours I get the following error in the syslog kernel file > Jul 15 01:09:25 mudshark kernel: Assertion failure in journal_stop() at > transaction.c:1418: "journal_current_handle() == handle"I can't say for sure, but the last couple of times I've had to track down errors of this sort, it was caused by a kernel stack overflow corrupting the journal_info pointer in the task structure. I'd look out for a misbehaving device driver that is using too much stack space..... - Ted