2002-09-02 16:55:32

by Gerd Knorr

[permalink] [raw]
Subject: 2.5.33: modular ide breaks lilo ...

Hi,

I've tried building the ide driver modular and insmod it using an
initrd. The kernel boots just fine, but lilo complains:

bogomips root ~# lilo
Device 0x0300: Invalid partition table, 2nd entry
3D address: 1/0/262 (264096)
Linear address: 1/10/4175 (4209030)

I've also noticed that the fdisk output looks different depending on
modular vs. static ide, I suspect this is related. With a modular IDE
driver it looks like this:

bogomips root ~# fdisk -l /dev/hda

Disk /dev/hda: 16 heads, 63 sectors, 79780 cylinders
Units = cylinders of 1008 * 512 bytes

Device Boot Start End Blocks Id System
/dev/hda1 1 4176 2104483+ b Win95 FAT32
Partition 1 does not end on cylinder boundary:
phys=(261, 254, 63) should be (261, 15, 63)
/dev/hda2 * 4176 68659 32499495 5 Extended
Partition 2 does not end on cylinder boundary:
phys=(1023, 254, 63) should be (1023, 15, 63)
/dev/hda4 68659 79768 5598652+ a5 FreeBSD
Partition 4 does not end on cylinder boundary:
phys=(1023, 254, 63) should be (1023, 15, 63)
/dev/hda5 4176 6264 1052226 82 Linux swap
/dev/hda6 6264 18759 6297448+ 83 Linux
/dev/hda7 18759 68659 25149726 83 Linux

With ide built-in statically fdisk prints this:

bogomips root ~# fdisk -l

Disk /dev/hda: 255 heads, 63 sectors, 5005 cylinders
Units = cylinders of 16065 * 512 bytes

Device Boot Start End Blocks Id System
/dev/hda1 1 262 2104483+ b Win95 FAT32
/dev/hda2 * 263 4308 32499495 5 Extended
/dev/hda4 4309 5005 5598652+ a5 FreeBSD
/dev/hda5 263 393 1052226 82 Linux swap
/dev/hda6 394 1177 6297448+ 83 Linux
/dev/hda7 1178 4308 25149726 83 Linux

Any idea?

Gerd


2002-09-02 17:33:20

by Andries Brouwer

[permalink] [raw]
Subject: Re: 2.5.33: modular ide breaks lilo ...

On Mon, Sep 02, 2002 at 06:27:07PM +0200, Gerd Knorr wrote:

> I've tried building the ide driver modular and insmod it using an
> initrd. The kernel boots just fine, but lilo complains:
>
> bogomips root ~# lilo
> Device 0x0300: Invalid partition table, 2nd entry
> 3D address: 1/0/262 (264096)
> Linear address: 1/10/4175 (4209030)

What LILO version?
For many versions it will suffice to give LILO the linear or lba32 option.

> I've also noticed that the fdisk output looks different depending on
> modular vs. static ide, I suspect this is related. With a modular IDE
> driver it looks like this:
>
> bogomips root ~# fdisk -l /dev/hda
>
> Disk /dev/hda: 16 heads, 63 sectors, 79780 cylinders
> Units = cylinders of 1008 * 512 bytes
>
> With ide built-in statically fdisk prints this:
>
> bogomips root ~# fdisk -l
>
> Disk /dev/hda: 255 heads, 63 sectors, 5005 cylinders
> Units = cylinders of 16065 * 512 bytes

What fdisk version? Make sure you have a recent one.
Clearly, the rest of the fdisk output is a consequence of the different
geometries. The kernel boot messages will probably tell what happened.
I must still read this part of the kernel source again to see what the
current status is.

Andries

2002-09-03 10:55:35

by Gerd Knorr

[permalink] [raw]
Subject: Re: 2.5.33: modular ide breaks lilo ...

On Mon, Sep 02, 2002 at 07:37:48PM +0200, Andries Brouwer wrote:
> On Mon, Sep 02, 2002 at 06:27:07PM +0200, Gerd Knorr wrote:
>
> > I've tried building the ide driver modular and insmod it using an
> > initrd. The kernel boots just fine, but lilo complains:
> >
> > bogomips root ~# lilo
> > Device 0x0300: Invalid partition table, 2nd entry
> > 3D address: 1/0/262 (264096)
> > Linear address: 1/10/4175 (4209030)
>
> What LILO version?

22.1

> For many versions it will suffice to give LILO the linear or lba32 option.

There already is a lba32 option in lilo.conf ...

> > Disk /dev/hda: 16 heads, 63 sectors, 79780 cylinders
> > Units = cylinders of 1008 * 512 bytes

> > Disk /dev/hda: 255 heads, 63 sectors, 5005 cylinders
> > Units = cylinders of 16065 * 512 bytes
>
> What fdisk version? Make sure you have a recent one.

2.11n

> Clearly, the rest of the fdisk output is a consequence of the different
> geometries. The kernel boot messages will probably tell what happened.

modular:
[ ... ]
hda: IBM-DTLA-305040, ATA DISK drive
[ ... ]
hda: host protected area => 1
hda: 80418240 sectors (41174 MB) w/380KiB Cache, CHS=79780/16/63, UDMA(33)
hda: hda1 hda2 < hda5 hda6 hda7 > hda4
hda4: <bsd: hda8 hda9 hda10 hda11 >
[ ... ]

builtin:
[ ... ]
hda: IBM-DTLA-305040, ATA DISK drive
[ ... ]
hda: host protected area => 1
hda: 80418240 sectors (41174 MB) w/380KiB Cache, CHS=5005/255/63, UDMA(33)
hda: hda1 hda2 < hda5 hda6 hda7 > hda4
hda4: <bsd: hda8 hda9 hda10 hda11 >
[ ... ]

The boot messages and the CHS geometry displayed by fdisk match ...

Gerd

--
You can't please everybody. And usually if you _try_ to please
everybody, the end result is one big mess.
-- Linus Torvalds, 2002-04-20