[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Missing nic drivers in gnumach compilation?



On Wed, 2001-10-17 at 20:24, Marcus Brinkmann wrote:
> On Wed, Oct 17, 2001 at 02:46:35PM +0100, Bob Ham wrote:
> > /proj/gnumach-20011013-1$ MIG=i386-gnu-mig CC=i386-gnu-gcc ./configure
> > --enable-floppy --enable-ide --enable-ne2000 --enable-via-rhine
>
> Mmmh.  In my i386/README.Drivers it wants --enable-viarhine (no second dash).
> Is this a bug in the README?
>
> > node@insanity:~/proj/gnumach-20011013-1$ grep VIA_RHINE i386/linux/device-drivers.h
> > #define CONFIG_VIA_RHINE 1
>
> This (and that it was compiled) seems to indicate the enable option was accepted,
> though.

Using --enable-viarhine produces the same value in the device-drivers.h
and makes the via-rhine.o.  It's probably autoconf's option resolution
that makes it not matter; it obviously only compares enough of the
--enable-via[-]rhine to disambiguate it from other options.

> If you want to be extra sure, use objdump --syms kernel and objdump -d kernel
> to verify that via_rhine_probe is included and called by ethif_probe.  I
> have no reason to believe otherwise, though.

Indeed, the code is there:

node@insanity:~/proj/hurd/gnumach$ objdump -d kernel |grep -50 via
[ ... ]
0014ca10 <ethif_probe>:
  14ca10:	55                   	push   %ebp
  14ca11:	89 e5                	mov    %esp,%ebp
  14ca13:	83 ec 14             	sub    $0x14,%esp
  14ca16:	53                   	push   %ebx
[ ... ]
  14ca3d:	e8 56 20 01 00       	call   15ea98 <via_rhine_probe>
  14ca42:	83 c4 10             	add    $0x10,%esp
  14ca45:	85 c0                	test   %eax,%eax
[ ... ]
0015ea98 <via_rhine_probe>:
  15ea98:	55                   	push   %ebp
  15ea99:	89 e5                	mov    %esp,%ebp


node@insanity:~/proj/hurd/gnumach$ objdump --syms kernel | grep via
00000000 l    df *ABS*	00000000 via-rhine.c
0015eab0 l     F .text	00000315 via_probe1
0015ea98 g     F .text	00000018 via_rhine_probe


> The boot message shows that no eth0 device could be found by the
> autoprobing.  If you want to debug this, you will have to go through the
> probing code in via-rhine.c and check where it fails (and compare it with a
> working Linux kernel source code if it differs there).

Not enough time for to get into deep debugging atm :/

> The linux drivers in GNU Mach are quite out of date.  We are working on
> OSKit-Mach, and OSKit has newer drivers.  This might be a future solution
> for you if hacking gnumach doesn't lead to any result.

CVS oskit-mach co'd, off to compile now..

Bob

--
Bob Ham:  bob@ham.org  http://pkl.net/~node/
IRC: 'node' on irc.openprojects.net: #slashdot  ICQ: 4396425 'node'

"The GNU philosophy is about freedom. To be free one must have
personal power. Personal power is an individual thing, difficult to
obtain and quick to perish." --Krisno Pryosusilo



Reply to: