[ On Sunday, July 27, 2003 at 11:03:40 (+0100), Charles Blundell wrote: ]
> Subject: Re: "panic: chgproccnt: lost user" on dual-CPU SS20 with 1.6.1_STABLE
>
> You have modified seteuid(), I presume. seteuid() in netbsd calls
> do_setresuid(), which in turn calls chgproccnt() only if
> r != -1 and != ruid. r is explicitly set to -1 by seteuid() in
> netbsd.
> i.e., this code path does not occur in netbsd, as far as I can tell.

Yeah, well as you can imagine by now I was wrong to say that the panic()
check wasn't in the code path because obviously it is.

What I think I meant to say was that I didn't see how the value being
checked could have gone out of range.

Sorry for the confusion.

In any case I'm still having no luck getting a host-built kernel to
boot.  This time with the only difference from the working kernel being:

    include	"arch/sparc/conf/GENERIC"
    options 	DDB
    options 	DDB_HISTORY_SIZE=100
    options 	DDB_ONPANIC=1

it says:

    Rebooting with command:                                               
    Boot device: /iommu/sbus/espdma@f,400000/esp@f,800000/sd@1,0  File and args: netbsd
    >> NetBSD/sparc Secondary Boot, Revision 1.12
    >> (woods@proven, Tue May 27 17:35:26 EDT 2003)
    Booting netbsd
    2921768+106924+263080 [187056+142961Memory Address not Aligned
    Type  help  for more information
    <#0> ok 

I'm beginning to suspect something's wrong with the host toolchain so
I'm now building a kernel on the cross-build machine where the original
working kernel was also built....

-- 
						Greg A. Woods

+1 416 218-0098                  VE3TCP            RoboHack <woods@robohack.ca>
Planix, Inc. <woods@planix.com>          Secrets of the Weird <woods@weird.com>