Date: Sun, 27 Jun 1999 21:23:57 +0800 From: Peter Wemm <peter@netplex.com.au> To: alpha@freebsd.org Subject: Re: Hmm!! Message-ID: <19990627132357.BAB6281@overcee.netplex.com.au> In-Reply-To: Your message of "Sun, 27 Jun 1999 21:10:16 %2B0800." <19990627131016.53F3781@overcee.netplex.com.au>
next in thread | previous in thread | raw e-mail | index | archive | help
Peter Wemm wrote:
[..]
> ashburton# csh
> ashburton# sysctl -n kern.bootfile
> /kernel
>
> ashburton# sh
> # sysctl -n kern.bootfile
> Segmentation fault - core dumped
[..]
> Some VM, pmap or fork/exec interaction perhaps?
>
> This is the same problem that showed up when I compiled in softupdates and
> went away when it was compiled out. With some other drivers present now
> it has come back again so perhaps it's a kernel size issue somehow? (I've
> had them before - see rev 1.21 in kern_fork.c back in 1996 that was kernel
> size related)
It gets stranger.. If I give the kernel config a major diet, it works with
SOFTUPDATES, compiled in and running on /home...
ashburton# ls -l /kernel*
-r-xr-xr-x 1 root wheel 2759622 Jun 27 21:03 /kernel
-rwxr-xr-x 1 root wheel 3810116 Mar 27 10:23 /kernel.GENERIC
-r-xr-xr-x 1 root wheel 3690516 Jun 27 20:19 /kernel.old
-r-xr-xr-x 1 root wheel 3583502 Jun 26 16:21 /kernel.sane
ashburton# size /kernel*
text data bss dec hex filename
1568138 177520 161256 1906914 1d18e2 /kernel
2218779 184776 170040 2573595 27451b /kernel.GENERIC
2259744 213600 175176 2648520 2869c8 /kernel.old
2014113 189088 173512 2376713 244409 /kernel.sane
ashburton# size -x /kernel*
text data bss dec hex filename
0x17ed8a 0x2b570 0x275e8 1906914 1d18e2 /kernel
0x21db1b 0x2d1c8 0x29838 2573595 27451b /kernel.GENERIC
0x227b20 0x34260 0x2ac48 2648520 2869c8 /kernel.old
0x1ebba1 0x2e2a0 0x2a5c8 2376713 244409 /kernel.sane
kernel.old was showing the segfaults for sysctl, kernel is fine - both are
exactly the same code but with a handful of PCI drivers, platform drivers,
EV4 compiled out, no NFS or CD9660, but with softupdates added.
kernel.sane was my last known good kernel prior to the buffer changes, and
it works reliably but fails if I add softupdates.
Just a datapoint...
Anybody got any bright ideas about this? (apart from run the memory tester
in SRM that is..)
Cheers,
-Peter
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990627132357.BAB6281>