Howdy,
>> /dev/hde5 63 204862 204800
>> /dev/hde6 467007 598078 131072
>> /dev/hde7 204863 467006 262144
> I assume that what the BIOS detects as hd0 is what Linux identifies as the
> /dev/hde (Is this right ?).
Yes, exactly!!!
> What I do not understand with your example is that you tried (hd0,4) that is
> /dev/hde5, and (hd0,6) that is /dev/hde7, both without any file system (see
> above).
Perhaps a little more clarification will help.
This Abit Kt7-Raid mother board has 4 IDE drive connectors
The *only* hard drive in this system is physically connected as
ide2: master and thus becomes BIOS drive 0
Linux FreeBSD Grub
UDMA66 ide0: master /dev/hda /dev/ad0
slave /dev/hdb /dev/ad1
UDMA66 ide1: master /dev/hdc /dev/ad2
slave /dev/hdd /dev/ad3
UDMA100 ide2: master /dev/hde /dev/ad4 (hd0)
slave /dev/hdf /dev/ad5
UDMA100 ide3: master /dev/hdg /dev/ad6
slave /dev/hdh /dev/ad7
So with only one physical drive in this system we have
Linux FreeBSD Grub
/dev/hde /dev/ad4 (hd0)
Now let's move on to how that is partitioned
The output of sfdisk -l -x -uS /dev/hde that was in my 1st
post could be more clearly explained as follows. Note
that all the sector bounds of hde5 thru hde8 are between
the start and end of hde1. Likewise for hde2, hde3, hde4
Start End #sectors Id System
/dev/hde1 63 5631695 5631633 a5 BSD/386
/dev/hde5 63 204862 204800
/dev/hde6 467007 598078 131072
/dev/hde7 204863 467006 262144
/dev/hde8 598079 5631695 5033617
/dev/hde2 5631696 11775455 6143760 5 Extended
/dev/hde9 5631759 5660927 29169 83 Linux
- 5660928 6160895 499968 5 Extended
- 5631696 5631695 0 0 Empty
- 5631696 5631695 0 0 Empty
/dev/hde10 5660991 6160895 499905 82 Linux swap
- 6160896 11775455 5614560 5 Extended
- 5660928 5660927 0 0 Empty
- 5660928 5660927 0 0 Empty
/dev/hde11 6160959 11775455 5614497 83 Linux
- 6160896 6160895 0 0 Empty
- 6160896 6160895 0 0 Empty
- 6160896 6160895 0 0 Empty
/dev/hde3 11775456 18955439 7179984 a5 BSD/386
/dev/hde12 11775456 12037599 262144
/dev/hde13 12037600 12299743 262144
/dev/hde14 12299744 18955439 6655696
/dev/hde4 * 18955440 60036479 41081040 a5 BSD/386
/dev/hde15 19217584 19479727 262144
/dev/hde16 19479728 20004015 524288
/dev/hde17 18955440 19217583 262144
/dev/hde18 20004016 60036479 40032464
Now that we are convinced of these relationships I'll drop the sector
numbers. The mappings related by LBA sectors are as follows:
Linux FreeBSD Grub Mount ID
/dev/hde1 /dev/ad4s1 (hd0,0) 0xa5 BSD/386
/dev/hde5 /dev/ad4s1a (hd0,0,a) /
/dev/hde6 /dev/ad4s1b (hd0,0,b) swap
/dev/hde7 /dev/ad4s1e (hd0,0,e) /var
/dev/hde8 /dev/ad4s1f (hd0,0,f) /usr
/dev/hde2 /dev/ad4s2 (hd0,1) 0x05 Extended
/dev/hde9 /dev/ad4s5 (hd0,4) /boot 0x83 Linux
/dev/hde10 /dev/ad4s6 (hd0,5) swap 0x82 Linux swap
/dev/hde11 /dev/ad4s7 (hd0,6) / 0x83
Linux
/dev/hde3 /dev/ad4s3 (hd0,2) 0xa5 BSD/386
/dev/hde12 /dev/ad4s3a (hd0,2,a) /
/dev/hde13 /dev/ad4s3e (hd0,2,e) /var
/dev/hde14 /dev/ad4s3f (hd0,2,f) /usr
/dev/hde4 /dev/ad4s4 (hd0,3) 0xa5 BSD/386
/dev/hde15 /dev/ad4s4a (hd0,3,a) /
/dev/hde16 /dev/ad4s4b (hd0,3,b) swap
/dev/hde17 /dev/ad4s4e (hd0,3,e) /var
/dev/hde18 /dev/ad4s4f (hd0,3,f) /usr
So in terms of LBA addressing I'm pretty sure that this is how
Linux,FreeBSD and Grub are all looking at the layout.
>From my 1st post...
>> grub> debug
>> grub> fstest
>> grub> root (hd0,4)
>> <2, 0, 1024><2, 0, 24><128, 0,204><16, 0, 204> Filesystem type unknown,
>> partition type 0x83
this indicates to me that grub is detecting the linux 0x83 identifier in
(hd0,4) also indicated by the sfdisk output but the crux of the matter
is:
What is the criteria that grub uses to find and determine the ext2 file
system?
Can anyone ascertain the value of this grub debug diagnostic
<2, 0, 1024><2, 0, 24><128, 0,204><16, 0, 204> from the command above?
I am still very concerned by the output of sfdisk
>> #sfdisk -l -x -uS /dev/hde
>> Disk /dev/hde: 59560 cylinders, 16 heads, 63 sectors/track
>> Units = sectors of 512 bytes, counting from 0
>>
>> Device Boot Start End #sectors Id System
>> /dev/hde1 63 5631695 5631633 a5 BSD/386
>> /dev/hde2 5631696 11775455 6143760 5 Extended
>> start: (c,h,s) expected (1023,15,63) found (1023,255,63)
>> /dev/hde3 11775456 18955439 7179984 a5 BSD/386
>> start: (c,h,s) expected (1023,15,63) found (1023,255,63)
>> /dev/hde4 * 18955440 60036479 41081040 a5 BSD/386
>> start: (c,h,s) expected (1023,15,63) found (1023,255,63)
I
The diagnostic especially for hde2 worries me. If grub is relying on CHS
info to access that partiion's contents then that could be the root of
my problem. Can anyone comment on that?
Perhaps I could gain some info from this. Could anyone running linux
preferably on a IDE hard drive >10GB post the output of these 3
commands?
where hdX is replaced with the correct syntax for your drive.
#dmesg | grep -e \[sh\]d\[a-h\]
and
#sfdisk -l -x -uS /dev/hdX
and
#dd if=/dev/hdX count=1|hexdump -s 0x1be -e '"%07.7_ax " 16/1 "%02x "
"\n"'
Any other constructive comments are welcome.
1 3 o 1 3
13o13
BoB
_______________________________________________
Bug-grub mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/bug-grub