On Sat, 21 Apr 2007, Jan Engelhardt wrote:
On Apr 21 2007 20:33, Mark Fortescue wrote:
I will check that if I ever get far enough. The system locks up in
kernel_exec of the initail process /sbin/bash.static
(init=/sbin/bash.static).
Perhaps try sash if it is installed.
Linux-2.6.14.3 works. Linux-2.6.15.6 fails.
sash starts up with Linux-2.6.15.6 but ls fails:
BUG: soft lockup detected on CPU#0!
PSR: 004010c0 PC: f00fba04 NPC: f00fba08 Y: a4000000 Not tainted
PC: <__memcpy+0x208/0x640>
%G: 00000000 00000040 00000000 00000000 00000000 00000000 ff000000 00000040
%O: ff001de0 f0256f38 00000040 0000008c 00000060 f00fb7a0 ff001d18 f0062204
RPC: <__handle_mm_fault+0x64/0x8ac>
%L: ff001da0 000010a1 88b00e6a f0256ef8 00000010 00000000 00000000 0000000a
%I: f0257dc8 f0242698 efb9f6f0 00000000 00000030 f00fb7d0 ff001de8 f001ddb8
This looks like there may be an issue the memory allocation systems.
-more /proc/meminfo:
<< /proc/meminfo >>
MemTotal: 37996 kB
MemFree: 36512 kB
Buffers: 0 kB
Cached: 544 kB
SwapCached: 0 kB
Active: 60 kB
Inactive: 540 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 37996 kB
LowFree: 36512 kB
SwapTotal: 0 kB
SwapFree: 0 kB
Dirty: 0 kB
Writeback: 0 kB
Mapped: 252 kB
Slab: 652 kB
CommitLimit: 18996 kB
Committed_AS: 256 kB
PageTables: 12 kB
VmallocTotal: 22528 kB
VmallocUsed: 0 kB
VmallocChunk: 22528 kB
--More--
BUG: soft lockup detected on CPU#0!
PSR: 004010c6 PC: f0012258 NPC: f001225c Y: 7d000000 Not tainted
PC: <ret_trap_continue+0x0/0x44>
%G: 00000000 004010a6 00000008 00000000 f0257dfc 00000000 ff000000 00000040
%O: f0252698 00002000 e0004400 00000000 00000000 00000050 ff001f50 f0012450
RPC: <ret_trap_user_stack_is_bolixed+0x14/0x24>
%L: 00401080 00045374 00045378 00000002 00000004 e0004000 00000000 f0066078
%I: 00000000 00002000 e0004400 00000000 00000000 00000050 efa8a558 00033d70
Is there a setting to make L1-A enabled by default?
Not that I know of any boot-time option, but there is always the source code.
BTW, stop_a is enabled by default (which bugs me a bit since it is disabled
by default when aurora is up and running unless you set the sysctl somewhere.)
21:33 ichi:../sparc64/kernel > grep -r stop_a .
./setup.c:int stop_a_enabled = 1;
(same for sparc32)
stop_a_enabled is set in arch/sparc/kernel/setup.c as per the sparc64
setup.
I see no code that clears it however I have had problems with
drivers/serial/sunzilog.c in 2.6.20.7 (it was enabling interrupts before
setting up the interrupt handler which caused my Sparcstation to panic).
I will add in some debug statements to the driver to see if I am getting a
BREAK condition and the check that interrupts are now working incase there
still errors to be fixed.
I will send a patch in for the sunzilog driver when I get a 2.6.20.7
kernel that gets to a shell and has a working L1-A and serial line BREAK
condition.
Jan
--
-
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
