2.4.9-bk and X on 6500, not so good

3 messages, 2 authors, 2001-08-25 · open the first message on its own page

2.4.9-bk and X on 6500, not so good

From: Stefan Jeglinski <hidden>
Date: 2001-08-21 05:16:50

I have a stock 6500/300, on which I just compiled and installed the
ppc.bitkeeper linuxppc_2_4 (2.4.9) kernel. In console mode (RL3) it
works just fine.

But whenever I launch X (Gnome for example, or a plain window
manager), about 70% of the time it breaks into the mon debugger
before X has fully launched.

Problem is, there is no keyboard response at this point, and I can't
even restart with the 3-fingered approach (no response). Only thing
that works is power down, and restart.

Any suggestions? To create a backtrace appears impossible...


Stefan Jeglinski

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: 2.4.9-bk and X on 6500, not so good

From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Date: 2001-08-21 08:15:47

I have a stock 6500/300, on which I just compiled and installed the
ppc.bitkeeper linuxppc_2_4 (2.4.9) kernel. In console mode (RL3) it
works just fine.

But whenever I launch X (Gnome for example, or a plain window
manager), about 70% of the time it breaks into the mon debugger
before X has fully launched.

Problem is, there is no keyboard response at this point, and I can't
even restart with the 3-fingered approach (no response). Only thing
that works is power down, and restart.

Any suggestions? To create a backtrace appears impossible...
You can at least get the PC and LR values displayed by xmon and
lookup them in your System.map. Eventually, you can hack xmon
entry point to do a backtrace automatically.

Ben.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: 2.4.9-bk and X on 6500, not so good

From: Stefan Jeglinski <hidden>
Date: 2001-08-25 05:12:34

At 10:15 AM +0200 8/21/01, Benjamin Herrenschmidt wrote:
 >I have a stock 6500/300, on which I just compiled and installed the
quoted
ppc.bitkeeper linuxppc_2_4 (2.4.9) kernel. In console mode (RL3) it
works just fine.

But whenever I launch X (Gnome for example, or a plain window
manager), about 70% of the time it breaks into the mon debugger
before X has fully launched.

Problem is, there is no keyboard response at this point, and I can't
even restart with the 3-fingered approach (no response). Only thing
that works is power down, and restart.

Any suggestions? To create a backtrace appears impossible...
You can at least get the PC and LR values displayed by xmon and
lookup them in your System.map.
pc = C0182544
lr = C0182444

from System.map:

c01821ec t unix_stream_data_wait
c01822dc t unix_stream_recvmsg
c018279c t unix_shutdown
 Eventually, you can hack xmon
entry point to do a backtrace automatically.
I was looking at xmon.c to do this, but I'm afraid I'm unclear on
where to try to make this happen at the break. I tried naively to
place the backtrace(excp) command right after the mon> printf
statement, but this didn't work when the crash happened.


Stefan Jeglinski

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help