> cc  -O2 -Werror -mc68020 -msoft-float -I. -I../../../../arch -I../../../.. -DLEV6_DEFER -DSERIBUF_SIZE="0x4000" -DTIMEZONE="0x12c" -DDST="1" -DM68030 -DFPCOPROC -DSWAPPAGER -DVNODEPAGER -DDEVPAGER -DINET -DQUOTA -DNFSSERVER -DNFSCLIENT -DFFS -DMFS -DPROCFS -DKERNFS -DFDESC -DNULLFS -DFIFO -DADOSFS -DUNION -DUMAPFS -DCOMPAT_10 -DCOMPAT_09 -DCOMPAT_43 -DCOMPAT_SUNOS -DTCP_COMPAT_42 -DCOMPAT_NOMID -DSYSVSHM -DSYSVMSG -DSYSVSEM -DUCONSOLE -DGENERIC -DLKM -DKTRACE -DDIAGNOSTIC -DNKMEMCLUSTERS=256 -DPANICWAIT -DDDB -DMACHINE_NONCONTIG -DGRF_ECS -DGRF_NTSC -D_KERNEL -Dmc68020 -Damiga  -c ../../../../nfs/nfs_vnops.c
> ../../../../nfs/nfs_vnops.c: In function `nfsspec_close':
> ../../../../nfs/nfs_vnops.c:3254: internal error--insn does not satisfy its constraints:
> (insn 135 50 52 (set (reg:DI 10 a2)
>         (mem/s:DI (plus:SI (reg:SI 11 a3)
>                 (const_int 168)))) 39 {movdi} (nil)
>     (nil))
> cc: Internal compiler error: program cc1 got fatal signal 6
> *** Error code 1

> This is on an Amiga with binaries from the 1.1 distribution...  Do I
> need a new gcc to compile a -current kernel now?

I got something very similar on one of my machines, probably the -sun3
machine, when building a kernel.  I just cut-and-pasted the command,
edited out the -O2, and ran it by hand, and it worked.

I'm hoping that once I get the world rebuilt and reinstalled the new
compiler binaries will build the kernel without needing human
intervention.

					der Mouse

			    mouse@collatz.mcrcim.mcgill.edu