***** Email Reply to soul@netrom.com *****
On Sat, 8 Jun 1996 jerijian@seas.ucla.edu wrote:
> Do you see what you get if you don't RTFM before upgrading
> from 1.2.13 to the latest pre2.0? More to come below...
>
> James M. Cassidy wrote:
> >
> >
> >
> > I got pre2.0.14, got the kernel compile and installed fine. When I
> > went to go to install it I got a few errors on boot-up that resemble:
> >
> >
> > Going multiuser...
> > SIOCSIFADDR: No such device
> > SIOCSIFBRDADDR: No such device
> > SIOCSIFNETMASK: No such device
> > Starting daemons: syslogd klogd portmap inetd lpd
> > fcntl_setlk() called by process 47 (lpd) with broken flock() emulation
> > fcntl_setlk() called by process 49 (lpd) with broken flock() emulation
> > Starting sendmail daemon (/usr/sbin/sendmail -bd -q 15m)...
> > fcntl_setlk() called by process 57 (agetty) with broken flock() emulation
> > fcntl_setlk() called by process 56 (agetty) with broken flock() emulation
> > fcntl_setlk() called by process 60 (agetty) with broken flock() emulation
> >
> > I belive the SIOCIFADDR is caused by my bad configurattion of networking. So
> > please just ignore that for now.
> >
> > The fcntl_setlk is caused by the new kernel. I did not compile the new
> > kernel support flock system. I hope I don't have to compile that option
> > to get many of my programs running cause I kind of don't like it and would
> > prefer not to have it on myy system.
>
> FAQ #1: To get rid of those messages, please upgrade to libc-5.2.18
> for ELF and libc-4.7.6 for a.out and recompile all of your statically
> linked binaries. Slackware 3.0 is old stuff you know (that's why I use
> Red Hat 3.0.3, which has a good user base and developer support, as
> well as readily-available contrib RPMs for Red Hat to work with the
> newest kernels. He he.).
>
> > Another error I get is when I do a 'ps'. And looks like the following:
> >
> > USER PID %CPU %MEM SIZE RSS TTY STAT START TIME COMMAND
> > Floating exception
> >
> > I have no 387 processor but I did compile the kernel with the Math Emulator
> > option. I even tried passing 'no387' to the kernel on startup to see if
> > it was somehow detecting a 387 and I still got this error.
>
> FAQ #2: This is old, old, old. Please upgrade to procps 0.99a!
>
> > The third error I got was from trying to mount cdrom. I have a old style
> > Mitsumi CD-ROM and I compiled that into the kernel. At startup it's
> > detecting my CD-ROM at irq11 but I'm pretty sure it's on irq10.
> >
> > Anyways when I try to mount /dev/mcdx1 with a cd-rom that has computer
> > stuff on track 1 and CD sound on the other tracks I get an opps error. I
> > know mcdx1 is probably track 2 a CD sound track and it should not be mountable
> > but I should be getting this nice big huge opps error as follows either:
> >
> > mice:/% mount -t msdos /dev/mcdx1 cdrom
> > Unable to handle kernel NULL pointer dereference at virtual address c0000088
> > current->tss.cr3 = 00615000, %cr3 = 00615000
> > *pde = 00102067
> > *pte = 00000027
> > Oops: 0000
> > CPU: 0
> > EIP: 0010:[<0018d49b>]
> > EFLAGS: 00010206
> > eax: 00000001 ebx: 00000000 ecx: 00000000 edx: 00716f70
> > esi: 001b1401 edi: 00716fb0 ebp: c0ed0000 esp: 00716d38
> > ds: 0018 es: 0018 fs: 002b gs: 002b ss: 0018
> > Process mount (pid: 115, process nr: 24, stackpage=00716000)
> > Stack: 001bbf7c 001b1401 00716fb0 fffffff8 00716e6c 00000000 00000000 00000000
> > 00000007 0000002b 00000001 00000004 08005be8 08005be8 08000000 0009e618
> > 0008c280 00000003 08006014 08005be8 001f9be8 08000000 40085000 00555790
> > Call Trace: [<00110955>] [<00115f03>] [<0010a36b>] [<001552c1>] [<001552db>] [<00155306>] [<001552c1>]
> > [<001552db>] [<00155306>] [<00129941>] [<001299a0>] [<00129b71>] [<00129757>] [<00129c17>] [<00121d37>]
> > [<00126bd5>] [<0010a3e2>]
> > Code: 83 bb 88 00 00 00 00 75 10 b8 fa ff ff ff 5b 5e 5f 81 c4 00
> > Segmentation fault
> >
> > I don't know if these errors are because of bad configuration on my part or
> > errors in the kernel. But I sure would like to know how to fix them.
>
> Okay, I admit this one isn't a FAQ. I have no solution for this one, I'm
> afraid. You might want to look at Documentation/cdrom/mcd* for some
> pointers on command-line parameters to get your interrupt settings right.
> Otherwise, this is a kernel bug...
>
> I'm sorry I was harsh on you. You should take the time to read the
> files in the Documentation/ subdirectory before posting old, known
> changes. And believe me, a LOT has changed in the 10 months between
> Linux 1.2.13 and Linux pre2.0.14.
>
> --
> Arthur D. Jerijian | "Who on earth can blame them? Ah, no wonder the
> jerijian@seas.ucla.edu | men of Troy and Argives under arms have suffered
> a540adj@pic.ucla.edu | years of agony all for her, for such a woman.
> adj@ucla.edu | Beauty, terrible beauty!"
> celestra@ix.netcom.com | Homer, the king of all poets
>
>