The following reply was made to PR kern/35102; it has been noted by GNATS.

From: Henry Nelson <netb@yuba.ne.jp>
To: Quentin Garnier <cube@cubidou.net>
Cc: gnats-bugs@NetBSD.org
Subject: Re: kern/35102: internal compiler error: Segmentation fault when building kernel
Date: Wed, 29 Nov 2006 06:51:55 +0900

 On Thu, Nov 23, 2006 at 11:53:49AM +0100, Quentin Garnier wrote:
 > On Thu, Nov 23, 2006 at 10:10:01AM +0000, henry nelson wrote:
 > > >Number:         35102
 > > >Category:       kern
 > > >Synopsis:       "sched.h:129: internal compiler error: Segmentation faul=
 [...]
 > > System: NetBSD note 3.0.0_STABLE NetBSD 3.0.0_STABLE (NXKERN) #0: Sat Apr=
 >  29 12:02:23 JST 2006  hennx@haus:/usr/src/sys/arch/i386/compile/NXKERN i386
 > > Architecture: i386
 > > Machine: i386
 > > >Description:
 > > #   compile  kern/pmatch.o
 > > cc -ffreestanding   -O2 -Werror -Wall -Wno-main -Wno-format-zero-length -=
 > Wpointer-arith -Wmissing-prototypes -Wstrict-prototypes -Wno-sign-compare -=
 > fno-zero-initialized-in-bss -I../../../../../../lib/libkern/arch/i386 -Di38=
 > 6 -I../../. -I../../../../../../arch -I../../../../../.. -nostdinc -DIPFILT=
 > ER_DEFAULT_BLOCK -DMAXUSERS=3D32 -D_KERNEL -D_KERNEL_OPT -I../../../../../.=
 > ./dist/ipf     -c    ../../../../../../lib/libkern/pmatch.c -o pmatch.o.tmp
 > > In file included from ../../../../../../sys/cpu_data.h:49,
 > >                  from ../../machine/cpu.h:59,
 > >                  from ../../machine/param.h:48,
 > >                  from ../../../../../../sys/param.h:145,
 > >                  from ../../../../../../lib/libkern/pmatch.c:32:
 > > ../../../../../../sys/sched.h:129: internal compiler error: Segmentation =
 > fault
 > > Please submit a full bug report,
 [...]
 > I'd start running memory tests...
 
 I ran "memtest86+ v1.65" for 62h30m with no errors (pass=66, errors=0).
 Also, the largest portion of memory was purchased less than a year ago.
 
 More information is that I succeeded in building the kernel on this
 machine.  The only difference was that the network card was pulled
 from the machine, and I had stopped the Apache server.
 
 Could swap be bad?  The hard disk is about 6 years old, although it
 hasn't had hard use, nor have I ever had trouble with it.  (If I should
 test swap, please give me some pointers as I have no idea how to do that.)
 
 Could it be the network card?  It has never given trouble before, either,
 but it's only been in newer Thinkpads running WindowsXP.  (Also, if this
 needs testing, I need pointers on how to do that.)
 
 Thanks. 
 
 -- 
 henry n.