On Wednesday 14 April 2004 03:54 am, Richard Harke wrote:
> On Tuesday 13 April 2004 03:52 pm, Richard Harke wrote:
> > On Sunday 11 April 2004 10:52 pm, Richard Harke wrote:
> > > On Sunday 11 April 2004 03:40 pm, Matthew Wilcox wrote:
> > > > On Fri, Apr 09, 2004 at 06:05:55PM -0700, Richard Harke wrote:
> > > > > The error is a longish harangue from the menuconfig script based on
> > > > > getting a value of 139 back from lxdialog
> > > >
> > > > Return values above 128 are 128 + signal number. ie 139 - 128 = 11.
> > > > lxdialog received signal 11 which is SIGSEGV.
> > >
> > > Thanks, I'll see if I can catch it with GDB
> >
> > I installed libncurses5-dbg (54.-3) and libc6.1-dbg (2.3.2.ds1-11)
Also s is NULL This would be OK if the width was there but
> together these lead to the seg fault
> Its very late, I hope this is clear.
It was very and that wasn't it.
The problem is related to threads. The code leading up to
the seg fault is:
ld8 r14=[r14];;
add r14=r14,r13
nop.i 0x0
[MMI] ld8 r14=[r14];;
ld8 r14=[r14]
nop.i 0x0
It seg faults on the last ld8 shown because r14=0x0
But notice r13. It is apparently addressing an element of the
thread control block
Does this suggest an external cause related to configuration,
versions or ??
Richard
