2003-08-16 00:34:40

by jeff millar

[permalink] [raw]
Subject: 2.6.0-test3 _still_ cannot mount root fs

Redhat's 2.4.20 will boot with root=/dev/hda3 pointing to an ext3
filesystem, without an initrd, with the following message

ext2fs warning (device IDE0(3,3) ) ext2_read_super: mounting ext3
filesystem as ext2
VFS: mounted root (ext2 filesystem) readonly

So I assumed that 2.6.0-test3 will also work with ext3 compiled as a module.
I tried the various suggestions root=/dev/hda3, root=0303, root=03:03
without any luck.

I then tested a suggestion that ext3 should get compiled in but get the same
results

> > > ============================================
> > > Case 2: grub, no initrd, kernel=/vmlinux-2.6.0-test3 ro root=/dev/hda3
> > >
> > > ...stuff scrolled off the screen...
> > > check >hda3<
> > > check <hda3< (<31)
> > > try_name() I
> > > try_name() >hda3<,0
> > > open > /dys/block/hda3/dev< = -1
> > > fail
> > > strtoul >3< -> 3
> > > try_name() II
> > > try_name hda < 3
> > > open >/sys/block/hda/dev< = 0
> > > read 0[32]=>3:0
> > > 5^(4)
> > > buf: >3:0<
> > > mkdev(3,0) -> 768
> > > open2 >/sys/block/hda/range< = 0
> > > read2 0[32] => 764
> > > < (3)
> > > buf: >64<
> > > strtoul >64< -> 64
> > > name to dev_t() done
> > > VFS: cannot open root dev "hda3" or hda3

Herbert's debug patch produced the same messages with ext3 compiled in.

Any suggestions for what to try next?

jeff


2003-08-16 15:00:32

by Herbert Poetzl

[permalink] [raw]
Subject: Re: 2.6.0-test3 _still_ cannot mount root fs


Hi Jeff!

On Fri, Aug 15, 2003 at 08:34:37PM -0400, jeff millar wrote:
> Redhat's 2.4.20 will boot with root=/dev/hda3 pointing to an ext3
> filesystem, without an initrd, with the following message
>
> ext2fs warning (device IDE0(3,3) ) ext2_read_super: mounting ext3
> filesystem as ext2
> VFS: mounted root (ext2 filesystem) readonly
>
> So I assumed that 2.6.0-test3 will also work with ext3 compiled as a module.
> I tried the various suggestions root=/dev/hda3, root=0303, root=03:03
> without any luck.
>
> I then tested a suggestion that ext3 should get compiled in but get the same
> results
>
> Herbert's debug patch produced the same messages with ext3 compiled in.
>
> Any suggestions for what to try next?

appended a short patch, which goes ontop of my other
debug patch ... this time, please concentrate on the
%% lines and the lines immediately above and below ...

best,
Herbert

> jeff

--- linux-2.6.0-test3-debug/init/do_mounts.c 2003-08-13 15:15:23.000000000 +0200
+++ linux-2.6.0-test3-debug-P1/init/do_mounts.c 2003-08-16 16:57:02.000000000 +0200
@@ -277,7 +277,11 @@
get_fs_names(fs_names);
retry:
for (p = fs_names; *p; p += strlen(p)+1) {
- int err = do_mount_root(name, p, flags, root_mount_data);
+ int err;
+
+ printk("%%%% do_mount_root() >%s<,>%s<, %08x ...\n", name, p, flags);
+ err = do_mount_root(name, p, flags, root_mount_data);
+ printk("%%%% ret = %d\n", err);
switch (err) {
case 0:
goto out;