Re: help need for powerpc issue

242 views
Skip to first unread message

Jo

unread,
Dec 1, 2009, 10:09:35 AM12/1/09
to burg-devel
Hello,

> >
> > Bean wrote:
> >>
> >> Hi,
> >>
> >> Have you make any progress on this problem ? I've created a forked
> >> project BURG which is based on grub2, you can try if that works for
> >> you.
> >>
> > Yes I make 2 step beyond (better then the song isn't it ;<).
> > The first I mentioned in a patch proposal
> > (<http://lists.gnu.org/archive/html/grub-devel/2009-11/msg00053.html>) which
> > get me rid of annoying "cursor-on cursor-on word unknown".
>
> Hi,
>
> Yep, the cursor-on problem affects sparc64-ieee1275 as well and it's
> quite annoying, I've disabled it in BURG.
>
Btw any comment on my proposal?

> >
> > The second is about the immediate fallback to 'rescue mode' at thestart up:
> > I just need to add normal module to my grub-mkelfimage ;<)
> >
> > That said, I read well your announce of your BURG proj and I already grab
> > the branch but I did yet test it, sorry ;<(
> > And tbh I don't know yet how much time will I still have access to the
> > system (the company start a process of restructuration which impact my job
> > but when, it's not yet clear. I hope to be able to advise you soon)
> >
Ok here am I back unfortunately with sad news: it failed to boot :

DEFAULT CATCH!, exception-handler=fff00300
at %SRR0: 0000000000c3c25c %SRR1: 800000000000b002
Call History
------------
@ - c3c1f0
close-package - c58060
(poplocals) - c3a758
(init-program) - c7e298
boot - c7ec7c
evaluate - c4a638
invalid pointer - d5f95d
invalid pointer - f
invalid pointer - f
catch - c38fe8
bt-task-boot-on-this - d0fd90
(poplocals) - c3a758
catch - c38fe8
execute-device-method - c58bcc
(poplocals) - c3a758
(select-boot-seq) - c59ba4

Client's Fix Pt Regs:
00 00100000000001f4 ffffffffffffffff 00000000deadbeef
fffffffffffffffc
04 0000000000000000 0000000000000000 0000000000000000
0000000000000001
08 0000000000001000 0000030002001000 0000000000000003
0000000000007000
0c 0000000022800000 0000000000c17100 0000000000c18000
000000000009c4b0
10 0000000000d94fa8 0000000000d94c39 0000000000c57d80
0000000000c58060
14 0000000000000000 ffffffffffffffff 0000000000000000
0000000000000000
18 0000000000c13000 0000000000c38000 0000000000c14d40
0000000000c16ec0
1c 0000000000c20000 0000000000c3fdf0 0000000000c11ea0
0000000000c10fa8
Special Regs:
%IV: 00000300 %CR: 84800000 %XER: 20000008 %DSISR:
08000000
%SRR0: 0000000000c3c25c %SRR1: 800000000000b002
%LR: 0000000000c3c1f0 %CTR: 0000000000000000
%DAR: ffffffffffffffff
Virtual PID = 0
ofdbg

Just build as usual:
1/ autogen.sh
2/ mkdir -p build; cd build
3/
TARGET_CFLAGS="-O0" CFLAGS="-O0" /MySrcTree/burg1781/configure
CC=gcc-4.4 TARGET_CC=gcc-4.4 --build=powerpc-linux-gnu --prefix=/opt/
grub2/burg1781 --includedir="\${prefix}/include" --mandir="\${prefix}/
share/man" --infodir="\${prefix}/share/info" --sysconfdir=/etc --
localstatedir=/var --libexecdir="\${prefix}/lib/grub2" --srcdir=/
MySrcTree/burg1781 --disable-werror --enable-grub-emu --enable-grub-
fstest --enable-grub-mkfont

4/ make

Did I missed something: even thought '--enable-grub-emu' option, make
didn't build grub-emu?

> >>> From your debug output, it seems that it's able to read from disk,
> >
> > Yes 2 evidences:
> > 1/ it read well and so display the (grub.cfg) menu
> > 2/ it boot well the initrd/vmlinux into the 'fat' formated grub file
> > system
> > ;<)
> >
> >> I
> >> wonder why it doesn't show any device in ls, perhaps it's fs related,
> >> you can use lsmod to show if the file system module is loaded.
> >>
> > Well, last test I did yesterday was with an image containing normal, ext2,
> > fat, part_msdos modules.
> > And lsmod showed me well ext2 but that doesn't help yet.
> >
> > So if I have some more time to investigate further, I will so try to add
> > some more grub_printf() in disk.c to discover what is it reading and
> > compare with grub-emu results.
>
> It could be some hidden bug in ext2, can grub-fstest read files from
> that partition ?
>
# grub-fstest /boot/initrd.img-2.6.30-2-powerpc64 ls
(loop0) (host)

Should I test something else?

That said I also notice that for sparc, it is named 'sparc64' so I
guess it's build as a 64bit application?

On the other hand I read from my OFS device-tree (DT) (also accessible
with linux /proc/device-tree) that for / the #address-cells as well as
#size-cells are set to 2; as far as I understand the first would mean
that it uses 64bit width address space and the second that forth unit
cells is also of 64bit width, is it correct?

But what's about the interface between OFS and OS (here grub2) is it a
32bit or a 64bit interface. I am worry that if grub request something
(address or data) the answer is of 64bit width of which grub only read
the half part?


> >
> > That said, am I right to say that grub discover disk with
> > grub_ofdisk_iterate() i.e. grub_devalias_iterate()?
> > That would explain why it only discover 1 disk
> > (/vdevice/v-scsi@30000016/disk@8100000000000000) and not the second one
> > (/vdevice/v-scsi@30000015/disk@8100000000000000: 15 in place of 16) like
> > does linux, though.
>
> That's right, it find disk device by iterating the device alias, and
> compare the device type.
>
Well I tried to boot directly from ofs prompt and I can define new
aliases (check with devalias of course) but either they are just
volatile or they are cleaned up by ofs menu, anyway they didn't
persist after a reboot: I still have to learn a lot about this OFS
implementation ;<)

Thanks in advance for further help,
J.

Bean

unread,
Dec 1, 2009, 10:38:58 AM12/1/09
to burg-...@googlegroups.com
On Tue, Dec 1, 2009 at 11:09 PM, Jo <rubi...@scarlet.be> wrote:
> Hello,
>
>> >
>> > Bean wrote:
>> >>
>> >> Hi,
>> >>
>> >> Have you make any progress on this problem ? I've created a forked
>> >> project BURG which is based on grub2, you can try if that works for
>> >> you.
>> >>
>> > Yes I make 2 step beyond (better then the song isn't it ;<).
>> > The first I mentioned in a patch proposal
>> > (<http://lists.gnu.org/archive/html/grub-devel/2009-11/msg00053.html>) which
>> > get me rid of annoying "cursor-on cursor-on word unknown".
>>
>> Hi,
>>
>> Yep, the cursor-on problem affects sparc64-ieee1275 as well and it's
>> quite annoying, I've disabled it in BURG.
>>
> Btw any comment on my proposal?

Hi,

My method is similar, although instead of adding a new variable, it
set a flags GRUB_IEEE1275_FLAG_SET_CURSOR using
grub_ieee1275_set_flag. The flags are mostly used to enable some
feature in specific ieee1275 platform, I think the cursor-on problem
belongs to them.
r1871 is a merge with grub, there are some issue, it should be fixed
in r1872. If that doesn't work, please go back to r1870, it should be
more stable.

with r1871, qemu is a standalone target, it needs to be compile
separately. Although you can use grub-fstest, it's quite convenient to
debug fs related issue.

>
>> >>> From your debug output, it seems that it's able to read from disk,
>> >
>> > Yes 2 evidences:
>> >        1/ it read well and so display the (grub.cfg) menu
>> >        2/ it boot well the initrd/vmlinux into the 'fat' formated grub file
>> > system
>> > ;<)
>> >
>> >> I
>> >> wonder why it doesn't show any device in ls, perhaps it's fs related,
>> >> you can use lsmod to show if the file system module is loaded.
>> >>
>> > Well, last test I did yesterday was with an image containing normal, ext2,
>> > fat, part_msdos modules.
>> > And lsmod showed me well ext2 but that doesn't help yet.
>> >
>> > So if I have some more time to investigate further, I will so try to add
>> > some more grub_printf() in  disk.c to discover what  is it reading and
>> > compare with grub-emu results.
>>
>> It could be some hidden bug in ext2, can grub-fstest read files from
>> that partition ?
>>
> # grub-fstest /boot/initrd.img-2.6.30-2-powerpc64 ls
> (loop0) (host)
>
> Should I test something else?

Some usage:

grub-fstest /dev/sda ls -l
Shows partitions, for example (loop0,1), (loop0,2), etc.

grub-fstest /dev/sda ls "(loop0,1)/boot/"
List files

grub-fstest /dev/sda cat "(loop0,1)/boot/grub/grub.cfg"
Display file content.

>
> That said I also notice that for sparc, it is named 'sparc64' so I
> guess it's build as a 64bit application?
>
> On the other hand I read from my OFS device-tree (DT) (also accessible
> with linux /proc/device-tree) that for / the #address-cells as well as
> #size-cells are set to 2; as far as I understand the first would mean
> that it uses 64bit width address space and the second that forth unit
> cells is also of 64bit width, is it correct?
>
> But what's about the interface between OFS and OS (here grub2) is it a
> 32bit or a 64bit interface. I am worry that if grub request something
> (address or data) the answer is of 64bit width of which grub only read
> the half part?
>

The current powerpc-ieee1275 is 32-bit, I suspect your OFS is 64-bit.
Although, ppc32 is compatible with ppc64, it should work as well.

>
>> >
>> > That said, am I right to say that grub discover disk with
>> > grub_ofdisk_iterate() i.e. grub_devalias_iterate()?
>> > That would explain why it only discover 1 disk
>> > (/vdevice/v-scsi@30000016/disk@8100000000000000) and not the second one
>> > (/vdevice/v-scsi@30000015/disk@8100000000000000: 15 in place of 16) like
>> > does linux, though.
>>
>> That's right, it find disk device by iterating the device alias, and
>> compare the device type.
>>
> Well I tried to boot directly from ofs prompt and I can define new
> aliases (check with devalias of course) but either they are just
> volatile or they are cleaned up by ofs menu, anyway they didn't
> persist after a reboot: I still have to learn a lot about this OFS
> implementation ;<)

Oh, your disk is not in the device alias ?

--
Bean

Project: https://launchpad.net/burg
Document: https://help.ubuntu.com/community/Burg
Mailing list: http://groups.google.com/group/burg-devel

Bean

unread,
Dec 1, 2009, 10:59:59 AM12/1/09
to burg-...@googlegroups.com
On Tue, Dec 1, 2009 at 11:38 PM, Bean <bean...@gmail.com> wrote:
>>> > That said, am I right to say that grub discover disk with
>>> > grub_ofdisk_iterate() i.e. grub_devalias_iterate()?
>>> > That would explain why it only discover 1 disk
>>> > (/vdevice/v-scsi@30000016/disk@8100000000000000) and not the second one
>>> > (/vdevice/v-scsi@30000015/disk@8100000000000000: 15 in place of 16) like
>>> > does linux, though.
>>>
>>> That's right, it find disk device by iterating the device alias, and
>>> compare the device type.
>>>
>> Well I tried to boot directly from ofs prompt and I can define new
>> aliases (check with devalias of course) but either they are just
>> volatile or they are cleaned up by ofs menu, anyway they didn't
>> persist after a reboot: I still have to learn a lot about this OFS
>> implementation ;<)
>
> Oh, your disk is not in the device alias ?

Hi,

BTW, OFS command nvalias should define a persistent alias.

Jo

unread,
Dec 1, 2009, 12:40:57 PM12/1/09
to burg-devel
Hey,

On Dec 1, 4:59 pm, Bean <bean12...@gmail.com> wrote:
> On Tue, Dec 1, 2009 at 11:38 PM, Bean <bean12...@gmail.com> wrote:
> >>> > That said, am I right to say that grub discover disk with
> >>> > grub_ofdisk_iterate() i.e. grub_devalias_iterate()?
> >>> > That would explain why it only discover 1 disk
> >>> > (/vdevice/v-scsi@30000016/disk@8100000000000000) and not the second one
> >>> > (/vdevice/v-scsi@30000015/disk@8100000000000000: 15 in place of 16) like
> >>> > does linux, though.
>
> >>> That's right, it find disk device by iterating the device alias, and
> >>> compare the device type.
>
> >> Well I tried to boot directly from ofs prompt and I can define new
> >> aliases (check with devalias of course) but either they are just
> >> volatile or they are cleaned up by ofs menu, anyway they didn't
> >> persist after a reboot: I still have to learn a lot about this OFS
> >> implementation ;<)
>
> > Oh, your disk is not in the device alias ?
>
> Hi,
>
> BTW, OFS command nvalias should define a persistent alias.
>
Do you know what?
That makes the drill:
0> nvalias disk1 /vdevice/v-scsi@30000015/disk@8100000000000000 ok
0> nvalias disk2 /vdevice/v-scsi@30000016/disk@8100000000000000 ok

reboot to check:
0 > nvalias
ibm,sp /vdevice/IBM,sp@4000
disk2 /vdevice/v-scsi@30000016/disk@8100000000000000
disk1 /vdevice/v-scsi@30000015/disk@8100000000000000
network /vdevice/l-lan@30000002
net /vdevice/l-lan@30000002
network1 /vdevice/l-lan@30000003
scsi /vdevice/v-scsi@30000015
nvram /vdevice/nvram@4002
rtc /vdevice/rtc@4001
screen /vdevice/vty@30000000
ok

Now boot grub and go to the grub-shell (for my part I just [Escape]
and [ctrl]-c)
sh:grub> ls
(disk2) (disk2,7) (disk2,6) (disk2,5) (disk2,3) (disk2,1) (disk1)
(disk1,2) (disk1,1)

Now try to reboot with such menu entry:
menuentry "Debian GNU/Linux, with Linux 2.6.30-2-powerpc64" {
insmod ext2
set root=(hd1,3)
search --no-floppy --fs-uuid --set 16ef0754-c845-46e9-a948-
b9669ee68329
linux /vmlinux-2.6.30-2-powerpc64
root=UUID=3c51c43e-63a7-4ff1-9b1c-cf98addcb7ed ro sysrq=1
insmod=sym53c8xx insmod=ipr qu
iet
initrd /initrd.img-2.6.30-2-powerpc64
}

it boot fine (and to be absolutely sure I rename vmlinux and initrd
into grub fs with .sav extension so that the original default menu
entry
menuentry "Grub2 tst Debian GNU/Linux, with Linux 2.6.30-2-
powerpc64" {
linux /vmlinux-2.6.30-2-powerpc64
root=UUID=3c51c43e-63a7-4ff1-9b1c-cf98addcb7ed ro sysrq=1
insmod=sym53c8xx insmod=ipr
initrd /initrd.img-2.6.30-2-powerpc64
}

would failed and now I can remove it ;<)

Still have to check why following menu entry:
menuentry "openSUSE 11.1 (ppc) (on /dev/mapper/system-root)" {
insmod lvm
insmod ext2
set root=(system-root)
search --no-floppy --fs-uuid --set 40ed7be1-44a9-493c-866e-
ea835eb94524
linux /boot/vmlinux-2.6.27.7-9-ppc64 root=/dev/mapper/system-
root
initrd /boot/initrd-2.6.27.7-9-ppc64
}

still failed.... (may be lvm.mod wasn't the last release or try to
simply add it in grub boot img?)

Thanks a lot for your kind support,
J.

PS: this success was realized with latest deb src tree 1.97+20091115
to which I added small stuff (cursor patch proposed on grub ml, and
compile with -O0).
Any way I will try your next (or previous?) burg and will advise you
asap.

Bean

unread,
Dec 1, 2009, 1:02:52 PM12/1/09
to burg-...@googlegroups.com
Hi,

You can run insmod lvm in grub console, and use ls to see if there is
new device, grub-fstest can also use to debug lvm/raid. For example,
if lvm has three disks:

grub-fstest -c 3 /dev/sda1 /dev/sdb1 /dev/sdc1 ls

Jo

unread,
Dec 2, 2009, 5:30:46 AM12/2/09
to burg-devel


On Dec 1, 7:02 pm, Bean <bean12...@gmail.com> wrote:
[snip]
well here is more test results:

I first figure out that some stuff where compiled with -O2 option I so
apply following hack:
=== modified file 'configure.ac'
--- configure.ac 2009-12-01 08:42:58 +0000
+++ configure.ac 2009-12-02 07:29:44 +0000
@@ -121,7 +121,7 @@
pc) machine_CFLAGS="-DGRUB_MACHINE_PCBIOS=1" ;;
emu) machine_CFLAGS="-DGRUB_MACHINE_EMU=1" ;;
esac
-CFLAGS="$CFLAGS $machine_CFLAGS -O2"
+CFLAGS="$CFLAGS $machine_CFLAGS"

AC_SUBST(host_cpu)
AC_SUBST(host_os)

=== <> ===

and burg r 1780 & 1782 boot again but fall back to [grub rescue] mode
without any chance to do anythink :<(
Welcome to GRUB!

Unknown command `menuentry'
Try `help' for usage
error: invalid object file
error: invalid object file
error: invalid object file
Unknown command `}'
Try `help' for usage
Unknown command `menuentry'
Try `help' for usage
error: invalid object file
error: invalid object file
Unknown command `'
Try `help' for usage
Unknown command `
'
Try `help' for usage
Unknown command `'
Try `help' for usage
Unknown command `4'
Try `help' for usage
Unknown command `V'
Try `help' for usage
Unknown command `Z'
Try `help' for usage
Unknown command ``'
Try `help' for usage
Unknown command `x'
Try `help' for usage
Unknown command `?'
Try `help' for usage
Unknown command `?
Try `help' for usage
Unknown command `?
Try `help' for usage
Entering rescue mode...

grub rescue> lsmod
error: invalid object file

That said, I noticed a big diff if grub image size 170Kb for burg
against 200Kb for grub-20091115?
But I also noticed those warning when I run autogen.sh:
# ./autogen.sh
configure.ac:40: installing `./config.guess'
configure.ac:174: required file `./config.rpath' not found
configure.ac:40: installing `./config.sub'
configure.ac:45: installing `./install-sh'
configure.ac:45: installing `./missing'
automake: no `Makefile.am' found for any configure output
WARNING: C file isn't a module: pubkey.c
WARNING: C file isn't a module: primegen.c
WARNING: C file isn't a module: ecc.c
WARNING: C file isn't a module: dsa.c
WARNING: C file isn't a module: hash-common.c
WARNING: C file isn't a module: md.c
WARNING: C file isn't a module: hmac-tests.c
WARNING: C file isn't a module: rsa.c
WARNING: C file isn't a module: elgamal.c
WARNING: C file isn't a module: cipher.c
WARNING: C file isn't a module: ac.c

Otoh, I have some recent issue to rebuild lastest debian src tree
20091130:
gcc-4.4 -Icommands -I/MySrcTree/grub2-1.97+20091130/commands -I. -I./
include -I/MySrcTree/grub2-1.97+2009113
0/gnulib -I/MySrcTree/grub2-1.97+20091130/include -Wall -W -
DGRUB_LIBDIR=\"/usr/lib/`echo grub/powerpc-emu | sed 's,x,x,
'`\" -DLOCALEDIR=\"\" -O0 -DGRUB_MACHINE_EMU=1 -DGRUB_UTIL=1 -MD -c -
o grub_emu-commands_halt.o /MySrcTree/grub2-1.97+20091130/commands/
halt.c
/MySrcTree/grub2-1.97+20091130/commands/halt.c:29:27: error: grub/cpu/
halt.h: No such file or directory
/MySrcTree/grub2-1.97+20091130/commands/halt.c: In function
'grub_cmd_halt':
/MySrcTree/grub2-1.97+20091130/commands/halt.c:37: warning: implicit
declaration of function 'grub_halt'
make[1]: *** [grub_emu-commands_halt.o] Error 1
make[1]: Leaving directory `/MySrcTree/grub2-1.97+20091130/build/grub-
emu'
make: *** [build/grub-emu] Error 2

So I am not yet able to compare (but your burg tree seems to have
fixed this issue so I will try to back port to grub experimental
tree,...)

Ok back to working grub (20091115), lvm module (either loaded at run-
time or included in the image) seems to have issue.

Well the disk with lvm structure is sda:
# cfdisk /dev/sda (i.e. ofs alias disk1)
GNU cfdisk 1.2.3

AIX VDASD
Disk: /dev/sda Disk type: msdos Size: 10737418239B,
10737MB
Heads: 255 Sectors per track: 63 Cylinders: 1305

Number Flags Part Type Filesystem
Label Size

------------------------------------------------------------------------------
1 Bootable Primary
fat16 24.6MB
2 Primary [Linux
LVM] 10709MB

And grub-fstest find well stuff:
# grub-fstest /dev/sda ls -l
Device system-swap: Unknown filesystem
Device system-root: Filesystem type ext2, Last modification time
2009-12-02 09:07:45 Wednesday, UUID 40ed7be1-44a9-493c-866e-
ea835eb94524
Device system-home: Filesystem type ext2, Last modification time
2009-12-02 09:07:45 Wednesday, UUID b6ab41ec-cd1d-4a6f-
aac9-043f511df9c6
Device loop0: Partition table
Partition loop0,2: Unknown filesystem
Partition loop0,1: Filesystem type fat, Label 'root', UUID 4ad5-
c9f7
Device host: Filesystem type hostfs

But:
sh:grub> insmod lvm
sh:grub> lsmod
Name Ref Count Dependencies
lvm 1
minicmd 1
sh 1 normal
part_msdos 1
fat 1
ext2 1 fshelp
fshelp 2
normal 4 boot
boot 5

sh:grub> search --no-floppy --fs-uuid --set 40ed7be1-44a9-493c-866e-
ea835eb9452
4
error: no such device: 40ed7be1-44a9-493c-866e-ea835eb94524

I run the same with debug=disk (too much useless ouput) and it scan
well (disk1,1) the boot fat slice as well as (disk1,2) the target vg
(well of format lvm2).

Well so you agree to let lvm issue aside for the moment and try to
find the root cause of burg boot failure?

Thanks,
J.

Jo

unread,
Dec 2, 2009, 5:51:02 AM12/2/09
to burg-devel
[snip]
> Well so you agree to let lvm issue aside for the moment and try to
s/so you agree/do you agree/

Sorry 2 kbd keys too close for fat finger ;<)

J.

Jo

unread,
Dec 2, 2009, 6:39:39 AM12/2/09
to burg-devel
[snip]
>
> Well so you agree to let lvm issue aside for the moment and try to
> find the root cause of burg boot failure?
>
Well I just go to deb src tree 20091130 and get rid of grub-emu build
and even though I also encountered menu issue:
Unknown command `menuentry'
Try `help' for usage
Unknown command `}'
Try `help' for usage
Unknown command `menuentry'
Try `help' for usage
error: invalid cluster 57849

I can continue manually and boot linux (with always same lvm issue,
obviously)
e.g.
rescue> lsmod
Name Ref Count Dependencies
minicmd 1
linux 2 boot,elf
elf 3 gzio
gzio 4
search 1 extcmd
extcmd 2
part_msdos 1
fat 1
ext2 3 fshelp
fshelp 4
lvm 1
normal 2 boot
boot 6

Hth,
j.

Bean

unread,
Dec 2, 2009, 9:11:51 AM12/2/09
to burg-...@googlegroups.com
On Wed, Dec 2, 2009 at 6:30 PM, Jo <rubi...@scarlet.be> wrote:
> [snip]
> well here is more test results:
>
> I first figure out that some stuff where compiled with -O2 option I so
> apply following hack:
> === modified file 'configure.ac'
> --- configure.ac        2009-12-01 08:42:58 +0000
> +++ configure.ac        2009-12-02 07:29:44 +0000
> @@ -121,7 +121,7 @@
>   pc)  machine_CFLAGS="-DGRUB_MACHINE_PCBIOS=1" ;;
>   emu)  machine_CFLAGS="-DGRUB_MACHINE_EMU=1" ;;
>  esac
> -CFLAGS="$CFLAGS $machine_CFLAGS -O2"
> +CFLAGS="$CFLAGS $machine_CFLAGS"
>
>  AC_SUBST(host_cpu)
>  AC_SUBST(host_os)
>
> === <> ===
>

Hi,

The -O2 is used to enable dead code elimination for gcc. The
GRUB_AS_LIST macro uses a trick that would call grub_assert_fail if
the input data structure is invalid, but if it's ok, it shouldn't
generate any runtime code. If dead code elimination is not enabled,
the grub_assert_fail is still present still in object file and you get
unresolved symbol error. Although if you don't see the error, it's
probably because dead code elimination is enabled by default.
BURG uses a new object format, it can't load modules compiled using
grub. You can either replaces all mod files from burg, or embed the
modules you need in the image, here is a list of modules you probably
wants:

minicmd sh normal nmenu reboot halt part_msdos ext2 fat ls linux

> That said, I noticed a big diff if grub image size 170Kb for burg
> against 200Kb for grub-20091115?

Yep, one of the advantage of burg is its small image size, due to the
use of new object format.

> But I also noticed those warning when I run autogen.sh:
> # ./autogen.sh
> configure.ac:40: installing `./config.guess'
> configure.ac:174: required file `./config.rpath' not found
> configure.ac:40: installing `./config.sub'
> configure.ac:45: installing `./install-sh'
> configure.ac:45: installing `./missing'
> automake: no `Makefile.am' found for any configure output
> WARNING: C file isn't a module: pubkey.c
> WARNING: C file isn't a module: primegen.c
> WARNING: C file isn't a module: ecc.c
> WARNING: C file isn't a module: dsa.c
> WARNING: C file isn't a module: hash-common.c
> WARNING: C file isn't a module: md.c
> WARNING: C file isn't a module: hmac-tests.c
> WARNING: C file isn't a module: rsa.c
> WARNING: C file isn't a module: elgamal.c
> WARNING: C file isn't a module: cipher.c
> WARNING: C file isn't a module: ac.c
>

Actually I have fixed some of the warning already, but it's harmless anyway.
Thanks, this looks like a lvm bug. Could you list the commands you use
to create the lvm volume along with the software version so that I can
try to recreate that environment here ?

Jo

unread,
Dec 3, 2009, 4:55:06 AM12/3/09
to burg-devel


On Dec 2, 3:11 pm, Bean <bean12...@gmail.com> wrote:
Hello,

well a simple 'strings grubofs' (my grub image) shows me a lot of
grub_ symbol but any _ass, most probably the last hypothesis is right.
(I will try to find more details)
Oops my bad: I missed some reading ;<)

> You can either replaces all mod files from burg, or embed the
> modules you need in the image, here is a list of modules you probably
> wants:
>
> minicmd sh normal nmenu reboot halt part_msdos ext2 fat ls linux
>
Based on burg bzr 1782 build, I use first this last image to which I
just add 'search' module.
As all working fine, I replaced grub's modules by burg's one and use a
smallest image (only normal ext2 fat part_msdos modules) and all is
working fine ;<)
(excepted lvm)

> > That said, I noticed a big diff if grub image size 170Kb for burg
> > against 200Kb for grub-20091115?
>
> Yep, one of the advantage of burg is its small image size, due to the
> use of new object format.
>
Cool ;<)

>
>
> > But I also noticed those warning when I run autogen.sh:
> > # ./autogen.sh
> > configure.ac:40: installing `./config.guess'
> > configure.ac:174: required file `./config.rpath' not found
> > configure.ac:40: installing `./config.sub'
> > configure.ac:45: installing `./install-sh'
> > configure.ac:45: installing `./missing'
> > automake: no `Makefile.am' found for any configure output
> > WARNING: C file isn't a module: pubkey.c
> > WARNING: C file isn't a module: primegen.c
> > WARNING: C file isn't a module: ecc.c
> > WARNING: C file isn't a module: dsa.c
> > WARNING: C file isn't a module: hash-common.c
> > WARNING: C file isn't a module: md.c
> > WARNING: C file isn't a module: hmac-tests.c
> > WARNING: C file isn't a module: rsa.c
> > WARNING: C file isn't a module: elgamal.c
> > WARNING: C file isn't a module: cipher.c
> > WARNING: C file isn't a module: ac.c
>
> Actually I have fixed some of the warning already, but it's harmless anyway.
>
It seems effectively harmless ;<)
mmm, not sure I will still be able to find info: it was build by an
OpenSuse 10.3 installation DVD during Mars 2008 (at the same time of
my first debian install on powerpc <http://lists.debian.org/debian-
powerpc/2008/03/msg00038.html>). OpenSuse was updated to 11.1 in the
mean time. But I will check if I find back DVD.

That said, now that Grub and Burg let me boot my Debian, I can get rid
of this OpenSuse install and rebuild another debian (e.g. debootstrap)
on a freshly rebuild lvm struct on the second disk (actualy sda) to
see if that help but I will wait your feedback before definitely
deleting everything.

Thanka agin for your support,
J.
Message has been deleted

Jo

unread,
Dec 3, 2009, 5:32:56 AM12/3/09
to burg-devel


On Dec 2, 3:11 pm, Bean <bean12...@gmail.com> wrote:
Hello,

well a simple 'strings grubofs' (my grub image) shows me a lot of
grub_ symbol but any _ass, most probably the last hypothesis is right.
(I will try to find more details)

>
>
Oops my bad: I missed some reading ;<)

> You can either replaces all mod files from burg, or embed the
> modules you need in the image, here is a list of modules you probably
> wants:
>
> minicmd sh normal nmenu reboot halt part_msdos ext2 fat ls linux
>
Based on burg bzr 1782 build, I use first this last image to which I
just add 'search' module.
As all working fine, I replaced grub's modules by burg's one and use a
smallest image (only normal ext2 fat part_msdos modules) and all is
working fine ;<)
(excepted lvm)

> > That said, I noticed a big diff if grub image size 170Kb for burg
> > against 200Kb for grub-20091115?
>
> Yep, one of the advantage of burg is its small image size, due to the
> use of new object format.
>
Cool ;<)

>
>
> > But I also noticed those warning when I run autogen.sh:
> > # ./autogen.sh
> > configure.ac:40: installing `./config.guess'
> > configure.ac:174: required file `./config.rpath' not found
> > configure.ac:40: installing `./config.sub'
> > configure.ac:45: installing `./install-sh'
> > configure.ac:45: installing `./missing'
> > automake: no `Makefile.am' found for any configure output
> > WARNING: C file isn't a module: pubkey.c
> > WARNING: C file isn't a module: primegen.c
> > WARNING: C file isn't a module: ecc.c
> > WARNING: C file isn't a module: dsa.c
> > WARNING: C file isn't a module: hash-common.c
> > WARNING: C file isn't a module: md.c
> > WARNING: C file isn't a module: hmac-tests.c
> > WARNING: C file isn't a module: rsa.c
> > WARNING: C file isn't a module: elgamal.c
> > WARNING: C file isn't a module: cipher.c
> > WARNING: C file isn't a module: ac.c
>
> Actually I have fixed some of the warning already, but it's harmless anyway.
>
It seems effectively harmless ;<)

>
>
mmm, not sure I will still be able to find info: it was build by an
OpenSuse 10.3 installation DVD during Mars 2008 (at the same time of
my first debian install on powerpc <http://lists.debian.org/debian-
powerpc/2008/03/msg00038.html>). OpenSuse was updated to 11.1 in the
mean time. But I will check if I find back DVD.

That said, now that Grub and Burg let me boot my Debian, I can get rid
of this OpenSuse install and rebuild another debian (e.g. debootstrap)
on a freshly rebuild lvm struct on the second disk (actualy sda) to
see if that help but I will wait your feedback before definitely
deleting everything.

Thanka agin for your support,
J.

>

Jo

unread,
Dec 3, 2009, 7:50:37 AM12/3/09
to burg-devel
[snip]

> > Thanks, this looks like a lvm bug. Could you list the commands you use
> > to create the lvm volume along with the software version so that I can
> > try to recreate that environment here ?
>
> mmm, not sure I will still be able to find info: it was build by an
> OpenSuse 10.3 installation DVD during Mars 2008 (at the same time of
> my first debian install on powerpc <http://lists.debian.org/debian-
> powerpc/2008/03/msg00038.html>). OpenSuse was updated to 11.1 in the
> mean time. But I will check if I find back DVD.
>
> That said, now that Grub and Burg let me boot my Debian, I can get rid
> of this OpenSuse install and rebuild another debian (e.g. debootstrap)
> on a freshly rebuild lvm struct on the second disk (actualy sda) to
> see if that help but I will wait your feedback before definitely
> deleting everything.
>
Well in the deep of my archive I find back the dvd I used and the
installed packages lvm was lvm2-2.02.25-28.ppc.rpm with kernel-
ppc64-2.6.22.5-31.pp.rpm (not sure the installed sw was the one used
for the installation but that's the best I can find)?

Hth,
J.

Bean

unread,
Dec 3, 2009, 11:00:56 AM12/3/09
to burg-...@googlegroups.com
Hi,

I just create a lvm partition in my ppc machine which installs debian
lenny, and it seems all right. BTW, what's the result of "ls -l"
inside grub console, does it list partitions properly ?

Jo

unread,
Dec 4, 2009, 4:18:00 AM12/4/09
to burg-devel
[snip]
>
> Hi,
>
> I just create a lvm partition in my ppc machine which installs debian
> lenny, and it seems all right. BTW, what's the result of "ls -l"
> inside grub console, does it list partitions properly ?
>
Well it find well devices and even lvm srtuctures but not fs on those
structures; here is the info:

As lvm module doesn't part of my grub image:

grub> insmod lvm

grub> lsmod
Name Ref Count Dependencies
lvm 1
minicmd 1
affs 0 fshelp
afs 0 fshelp
afs_be 0 fshelp
befs 0 fshelp
befs_be 0 fshelp
cpio 0
hfs 0
hfsplus 0 fshelp
iso9660 0 fshelp
jfs 0
minix 0
ntfs 0 fshelp
reiserfs 0 fshelp
sfs 0 fshelp
tar 0
udf 0 fshelp
ufs1 0
ufs2 0
xfs 0 fshelp
ls 1 lib,extcmd
extcmd 2
nmenu 1 lib,boot
boot 2
sh 1 lib
part_msdos 1
fat 1
ext2 1 fshelp
fshelp 14
normal 2 lib
lib 9
kernel 1

grub> ls -l
Device system-swap: Unknown filesystem
Device system-root: Unknown filesystem
Device system-home: Unknown filesystem
Device p5tst001vg-Sources: Unknown filesystem
Device p5tst001vg-UsrSrc: Unknown filesystem
Device p5tst001vg-home: Unknown filesystem
Device p5tst001vg-tmp: Unknown filesystem
Device p5tst001vg-var: Unknown filesystem
Device disk2: Partition table
Partition disk2,7: Unknown filesystem
Partition disk2,6: Filesystem type ext2, Last modification
time 2009-12-03 14:29:59 Thursda
y, UUID 3c51c43e-63a7-4ff1-9b1c-cf98addcb7ed
Partition disk2,5: Unknown filesystem
Partition disk2,3: Filesystem type ext2, Last modification
time 2009-12-03 14:30:03 Thursda
y, UUID 16ef0754-c845-46e9-a948-b9669ee68329
Partition disk2,1: Filesystem type fat, Label grub2, UUID
4adc-336e
Device disk1: Partition table
Partition disk1,2: Unknown filesystem
Partition disk1,1: Filesystem type fat, Label 'root', UUID
4ad5-c9f7

some explamation:
disk1 (as I defined by previous nvalias) contains the opensuse
installation and contains well 2 slices:
1/ the small fat fs containing yaboot loader
2/ a biger slice containing the 'system' volume group which itself
contains well 3 logical volumes swap, root, home

disk2 contains my debian installation which is still of more classical
silces implementation:
1/ a small fat fs containing now burg image and modules
2/ a a biger slice of about 2Gb for the root (/) fs
3/ the last slice with the rest of the disk containing a vg
'p5tst001vg' for the lv tmp, var, home, 2 more stuff for my usage.

And iirc the disk1 where well formated with opensuse tools and disk2
with debian tools (iirc the first release of 5.0 netinst cd).

That said, this is a small system with all 4 disk bay occupied by disk
usefull for other purpose; more over there isn't external enclosure
available to build a new installation. Any way on disk2 there is still
a very small free space (about 7Mb) where I can create a new vg and a
lv with more recent debian tools (latest unstable updated last
Monday):
(p5_sid)root@p5tst001:~# vgcreate burgvg /dev/sdb2
No physical volume label read from /dev/sdb2
Physical volume "/dev/sdb2" successfully created
Volume group "burgvg" successfully created
(p5_sid)root@p5tst001:~# lvcreate -n tstlv -L 2M burgvg
Rounding up size to full physical extent 4.00 MiB
Logical volume "tstlv" created

(p5_sid)root@p5tst001:~# mke2fs -t ext3 -m 0 /dev/mapper/burgvg-tstlv
mke2fs 1.41.9 (22-Aug-2009)
Filesystem label=
OS type: Linux
Block size=1024 (log=0)
Fragment size=1024 (log=0)
1024 inodes, 4096 blocks
0 blocks (0.00%) reserved for the super user
First data block=1
Maximum filesystem blocks=4194304
1 block group
8192 blocks per group, 8192 fragments per group
1024 inodes per group

Writing inode tables: done
Creating journal (1024 blocks): done
Writing superblocks and filesystem accounting information: done

This filesystem will be automatically checked every 29 mounts or
180 days, whichever comes first. Use tune2fs -c or -i to override.

(p5_sid)root@p5tst001:~# mkdir /mnt/bt
(p5_sid)root@p5tst001:~# mount /dev/mapper/burgvg-tstlv /mnt/bt
(p5_sid)root@p5tst001:~# bdf
Filesystem 1K-blocks Used Available Use% Mounted on
[snip]
/dev/mapper/p5tst001vg-home
507748 26966 480782 6% /home
/dev/mapper/p5tst001vg-UsrSrc
2064208 825664 1238544 40% /usr/src
/dev/mapper/p5tst001vg-Sources
2567984 1758764 809220 69% /Sources
/dev/mapper/burgvg-tstlv
3963 1058 2905 27% /mnt/bt

At the reboot, grub see well the new lv but still failed to see the
underlying fs ;<(
grub> ls -l
Device system-swap: Unknown filesystem
Device system-root: Unknown filesystem
Device system-home: Unknown filesystem
Device burgvg-tstlv: Unknown filesystem
Device p5tst001vg-Sources: Unknown filesystem
Device p5tst001vg-UsrSrc: Unknown filesystem
Device p5tst001vg-home: Unknown filesystem
Device p5tst001vg-tmp: Unknown filesystem
Device p5tst001vg-var: Unknown filesystem
Device disk2: Partition table
Partition disk2,7: Unknown filesystem
Partition disk2,6: Filesystem type ext2, Last modification
time 2009-12-04 08:43:51 Friday,
UUID 3c51c43e-63a7-4ff1-9b1c-cf98addcb7ed
Partition disk2,5: Unknown filesystem
Partition disk2,3: Filesystem type ext2, Last modification
time 2009-12-04 08:43:50 Friday,
UUID 16ef0754-c845-46e9-a948-b9669ee68329
Partition disk2,2: Unknown filesystem
Partition disk2,1: Filesystem type fat, Label grub2, UUID
4adc-336e
Device disk1: Partition table
Partition disk1,2: Unknown filesystem
Partition disk1,1: Filesystem type fat, Label 'root', UUID
4ad5-c9f7

But for you it works so I will go back to the log of my build to see
if I can read some relevant warning?

Tx a lot,
J.
[snip]

Bean

unread,
Dec 4, 2009, 9:12:16 AM12/4/09
to burg-...@googlegroups.com
Hi,

Oh, I only create one partition inside lvm, perhaps I can try multiple
partition as well.

Bean

unread,
Dec 5, 2009, 11:28:53 AM12/5/09
to burg-...@googlegroups.com
On Fri, Dec 4, 2009 at 10:12 PM, Bean <bean...@gmail.com> wrote:
> Oh, I only create one partition inside lvm, perhaps I can try multiple
> partition as well.

Hi,

I've tried multiple volume and multiple partition, but still can't
reproduce your result, all seems ok.

However, I only have one disk, you can try removing alias disk1 and
let burg sees disk2 only, see if this makes any difference in
detecting lvm.

BTW, I recommend using stable release such as debian lenny. Last time
I try squeeze, there seems to be some strange issue and rarpd doesn't
work ! I think the powerpc port is not widely tested as i386 so there
could be hidden issue in testing/unstable release.

rubisher

unread,
Dec 5, 2009, 4:54:06 PM12/5/09
to burg-...@googlegroups.com
Hello,
Bean wrote:
> On Fri, Dec 4, 2009 at 10:12 PM, Bean <bean...@gmail.com> wrote:
>> Oh, I only create one partition inside lvm, perhaps I can try multiple
>> partition as well.
>
mmm yes that's what I did too when I had this small lv (only one lv in a small vg)

> Hi,
>
> I've tried multiple volume and multiple partition, but still can't
> reproduce your result, all seems ok.
>
> However, I only have one disk, you can try removing alias disk1 and
> let burg sees disk2 only, see if this makes any difference in
> detecting lvm.
>
Ok I can do that but I don't know how? (may be simply 'nvalias disk1'?)

> BTW, I recommend using stable release such as debian lenny. Last time
> I try squeeze, there seems to be some strange issue and rarpd doesn't
> work ! I think the powerpc port is not widely tested as i386 so there
> could be hidden issue in testing/unstable release.
>
Agree, do you still have to sent me gcc-4.4 and binutils release installed right know on your lenny?

Tx again,
J.

Bean

unread,
Dec 6, 2009, 1:48:51 AM12/6/09
to burg-...@googlegroups.com
On Sun, Dec 6, 2009 at 5:54 AM, rubisher <rubi...@scarlet.be> wrote:
>> However, I only have one disk, you can try removing alias disk1 and
>> let burg sees disk2 only, see if this makes any difference in
>> detecting lvm.
>>
> Ok I can do that but I don't know how? (may be simply 'nvalias disk1'?)

Hi,

It should be nvunalias disk1.

>> BTW, I recommend using stable release such as debian lenny. Last time
>> I try squeeze, there seems to be some strange issue and rarpd doesn't
>> work ! I think the powerpc port is not widely tested as i386 so there
>> could be hidden issue in testing/unstable release.
>>
> Agree, do you still have to sent me gcc-4.4 and binutils release installed
> right know on your lenny?

Debian lenny uses gcc-4.3, which works OOTB (no need to modify -O2). I
suspect the problem could be related to gcc-4.4, perhaps I can install
debian unstable on a testing partition and if it works.

Jo

unread,
Dec 7, 2009, 5:51:43 AM12/7/09
to burg-devel
Hello,

On Dec 6, 7:48 am, Bean <bean12...@gmail.com> wrote:
> On Sun, Dec 6, 2009 at 5:54 AM, rubisher <rubis...@scarlet.be> wrote:
> >> However, I only have one disk, you can try removing alias disk1 and
> >> let burg sees disk2 only, see if this makes any difference in
> >> detecting lvm.
>
> > Ok I can do that but I don't know how? (may be simply 'nvalias disk1'?)
>
> Hi,
>
> It should be nvunalias disk1.
>
Ok
0 > devalias
ibm,sp /vdevice/IBM,sp@4000
disk2 /vdevice/v-scsi@30000016/disk@8100000000000000
network /vdevice/l-lan@30000002
net /vdevice/l-lan@30000002
network1 /vdevice/l-lan@30000003
scsi /vdevice/v-scsi@30000015
nvram /vdevice/nvram@4002
rtc /vdevice/rtc@4001
screen /vdevice/vty@30000000
ok

but it doesn't help:
grub> ls -l
Device burgvg-tstlv: Unknown filesystem
Device p5tst001vg-Sources: Unknown filesystem
Device p5tst001vg-UsrSrc: Unknown filesystem
Device p5tst001vg-home: Unknown filesystem
Device p5tst001vg-tmp: Unknown filesystem
Device p5tst001vg-var: Unknown filesystem
Device disk2: Partition table
Partition disk2,7: Unknown filesystem
Partition disk2,6: Filesystem type ext2, Last modification
time 2009-12-07 06:37:35 Monday,
UUID 3c51c43e-63a7-4ff1-9b1c-cf98addcb7ed
Partition disk2,5: Unknown filesystem
Partition disk2,3: Filesystem type ext2, Last modification
time 2009-12-07 06:37:35 Monday,
UUID 16ef0754-c845-46e9-a948-b9669ee68329
Partition disk2,2: Unknown filesystem
Partition disk2,1: Filesystem type fat, Label grub2, UUID
4adc-336e


> >> BTW, I recommend using stable release such as debian lenny. Last time
> >> I try squeeze, there seems to be some strange issue and rarpd doesn't
> >> work ! I think the powerpc port is not widely tested as i386 so there
> >> could be hidden issue in testing/unstable release.
>
> > Agree, do you still have to sent me gcc-4.4 and binutils release installed
> > right know on your lenny?
>
> Debian lenny uses gcc-4.3, which works OOTB (no need to modify -O2). I
> suspect the problem could be related to gcc-4.4, perhaps I can install
> debian unstable on a testing partition and if it works.
>
Well so, I rebuild burg (always bzr 1782) with gcc-4.3 and various
optimization.
In general the build with gcc-4.3 work better then gcc-4.4 (I don't
encounter any more any failure to boot burg itself ;<) )

Never the less even after the down-grade from gcc-4.3 (unstable
4.3.4-6 to lenny 4.3.2-1.1) and binutils (unstable 2.20-4 to stable
2.18.1~cvs20080103) it doesn't help more:
GNU BURG version bzr.1782

[ Minimal BASH-like line editing is supported. For the first word,
TAB
lists possible command completions. Anywhere else TAB lists
possible
device/file completions. ESC at any time exits. ]

grub> lsmod
Name Ref Count Dependencies
lvm 1
linux 1 elf,boot
elf 2 gzio
gzio 3
ls 1 lib,extcmd
fat 1
ext2 1 fshelp
fshelp 2
part_msdos 1
halt 1
reboot 1
nmenu 1 lib,boot
boot 3
normal 2 lib
sh 1 lib
lib 6
minicmd 1
search 1 extcmd
extcmd 3
kernel 1
grub> ls -l
Device burgvg-tstlv: Unknown filesystem
Device p5tst001vg-Sources: Unknown filesystem
Device p5tst001vg-UsrSrc: Unknown filesystem
Device p5tst001vg-home: Unknown filesystem
Device p5tst001vg-tmp: Unknown filesystem
Device p5tst001vg-var: Unknown filesystem
Device disk2: Partition table
Partition disk2,7: Unknown filesystem
Partition disk2,6: Filesystem type ext2, Last modification
time 2009-12-07 09:58:03 Monday,
UUID 3c51c43e-63a7-4ff1-9b1c-cf98addcb7ed
Partition disk2,5: Unknown filesystem
Partition disk2,3: Filesystem type ext2, Last modification
time 2009-12-07 09:58:02 Monday,
UUID 16ef0754-c845-46e9-a948-b9669ee68329
Partition disk2,2: Unknown filesystem
Partition disk2,1: Filesystem type fat, Label grub2, UUID
4adc-336e

(I just change the 'GNU GRUB' to 'GNU BURG' to be sure I am well
booting the latest burg image)

I am a bit stuck right now :<(

Thanks any way,
J.

Bean

unread,
Dec 7, 2009, 10:04:04 AM12/7/09
to burg-...@googlegroups.com
Hi,

Well, this is quite tricky, perhaps you need to compare debug output
to find possible issue.

You can first create a image to include only the necessary modules, for example:

ls ext2 lvm reboot

and move the modules directory temporary so that it can't find *.lst
and *.mod files (avoid autoload).

In burg:

set debug=disk,fs
set pager=1
ls -l (burgvg-tstlv)

In linux:
grub-fstest -d disk,fs /dev/sda ls -l "(burgvg-tstlv)"

And compare the results.

Bean

unread,
Dec 7, 2009, 3:11:21 PM12/7/09
to burg-...@googlegroups.com
Hi,

Oh BTW, with r1784, gcc-4.4 should work OOTB.

Jo

unread,
Dec 8, 2009, 4:34:10 AM12/8/09
to burg-devel


On Dec 7, 9:11 pm, Bean <bean12...@gmail.com> wrote:
> Hi,
>
> Oh BTW, with r1784, gcc-4.4 should work OOTB.
>
Awesome job, it works well OOTB (i.e. with default -Os with gcc-4.4
debian release 4.4.2-3):
GNU BURG version bzr1784

[ Minimal BASH-like line editing is supported. For the first word,
TAB
lists possible command completions. Anywhere else TAB lists
possible
device/file completions. ESC at any time exits. ]

grub> ls
(burgvg-tstlv) (p5tst001vg-Sources) (p5tst001vg-UsrSrc) (p5tst001vg-
home) (p5tst001vg-tmp) (p5tst00
1vg-var) (disk2) (disk2,7) (disk2,6) (disk2,5) (disk2,3) (disk2,2)
(disk2,1)

But doesn't yet fix my vg issue, so here is what I got with only fs
debug (disk only add that it's reading disk, thought)

grub> set debug=fs
grub> ls -l (burgvg-tstlv)
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting affs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: affs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting afs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: afs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting afs_be...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: afs_be detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting befs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: befs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting befs_be...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: befs_be detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting cpiofs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: cpiofs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting hfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: hfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting hfsplus...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: hfsplus detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting iso9660...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: iso9660 detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting jfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: jfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting minix...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: minix detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting ntfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: ntfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting
reiserfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: reiserfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting sfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: sfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting tarfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: tarfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting udf...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: udf detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting ufs1...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: ufs1 detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting ufs2...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: ufs2 detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting xfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: xfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting fat...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: fat detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting ext2...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: ext2 detection
failed.
Device burgvg-tstlv: /Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58:
Detecting affs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: affs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting afs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: afs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting afs_be...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: afs_be detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting befs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: befs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting befs_be...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: befs_be detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting cpiofs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: cpiofs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting hfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: hfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting hfsplus...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: hfsplus detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting iso9660...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: iso9660 detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting jfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: jfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting minix...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: minix detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting ntfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: ntfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting
reiserfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: reiserfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting sfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: sfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting tarfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: tarfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting udf...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: udf detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting ufs1...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: ufs1 detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting ufs2...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: ufs2 detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting xfs...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: xfs detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting fat...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: fat detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:58: Detecting ext2...
/Sources/jso/Deb.grub2/burg-bzr1784/kern/fs.c:64: ext2 detection
failed.
Unknown filesystem

BTW, may be can you show me from your install what should I better
have as results?

Tx again,
J.

Jo

unread,
Dec 8, 2009, 5:32:21 AM12/8/09
to burg-devel


On Dec 7, 9:11 pm, Bean <bean12...@gmail.com> wrote:
> Hi,
>
> Oh BTW, with r1784, gcc-4.4 should work OOTB.
>
Awesome job, it works well OOTB (i.e. with default -Os with gcc-4.4
debian release 4.4.2-3):
GNU BURG version bzr1784

[ Minimal BASH-like line editing is supported. For the first word,
TAB
lists possible command completions. Anywhere else TAB lists
possible
device/file completions. ESC at any time exits. ]

Bean

unread,
Dec 8, 2009, 9:56:03 AM12/8/09
to burg-...@googlegroups.com
On Tue, Dec 8, 2009 at 5:34 PM, Jo <rubi...@scarlet.be> wrote:
> BTW, may be can you show me from your install what should I better
> have as results?

Hi,

It's not much output:

../kern/fs.c:58: Detecting ext2...
Device burgvg-tstlv: ../kern/fs.c:58L Detecting ext2...
Filesystem type ext2, Last modification time ..., UUID ...

The detection begins at line 58 in kern/fs.c:

grub_dprintf ("fs", "Detecting %s...\n", p->name);
(p->dir) (device, "/", dummy_func);
if (grub_errno == GRUB_ERR_NONE)
return p;

grub_error_push ();
grub_dprintf ("fs", "%s detection failed.\n", p->name);
grub_error_pop ();

First, you can clear grub_errno before running p->dir, this would
avoid error occur previously to affect the result:

grub_dprintf ("fs", "Detecting %s...\n", p->name);
grub_errno = 0;
(p->dir) (device, "/", dummy_func);
if (grub_errno == GRUB_ERR_NONE)
return p;

grub_error_push ();
grub_dprintf ("fs", "%s detection failed.\n", p->name);
grub_error_pop ();

The detecting code is in function grub_ext2_dir in fs/ext2.c, you can
add enough grub_printf to see when does it goes wrong, grub_ext2_dir
uses the following subroutine:

grub_ext2_mount detect ext2 fs, a likely place for error
grub_fshelp_find_file is called next, this is a helper function, which
call back to grub_ext2_iterate_dir.
grub_ext2_iterate_dir

Jo

unread,
Dec 10, 2009, 8:47:44 AM12/10/09
to burg-devel
Hello,

On Dec 8, 3:56 pm, Bean <bean12...@gmail.com> wrote:
Just intermediate results:
looks like some info seems to be clobbered somewhere, thought:
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:561: Ext data-
>sblock.magic = 0x00000000
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:562: grub_le_to_cpu16
(data->sblock.magic) = 0x000000
00
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:565: data->sblock:
0x1c1d0b0
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:566: size: 336
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:572:
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000

while with grub-fstest I got for the same fs on same lv:
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:561: Ext data-
>sblock.magic = 0x000053ef
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:566: size: 336
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:572:
000400000010000000000000590b0000f503000001000000
0000000000000000002000000020000000040000abcb184b
c6cb184b01001d0053ef0100010000007dcb184b004eed00
0000000001000000000000000b000000800000003c000000
02000000010000003a687767567b48b9baaa661835b760a5
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000
00000000000000000000000000000f000000000000000000
0000000000000000080000000000000000000000d78e5171
fd40210b7189f7aa3548a0eb010100000000000000000000
7dcb184ba2000000a3000000a4000000a5000000a6000000
a7000000a8000000a9000000aa000000ab000000ac000000
ad000000ae000000af010000000000000000000000001000

Don't know if related but it becomes even worse when some additional
modules are loaded:
# 1st ls -l (burgvg-tstlv)
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:561: Ext data-
>sblock.magic = 0x0000883b
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:566: size: 336
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:572:
4166006f006e0074002e000f00646d006f0064000000ffff
ffff0000ffffffff464f4e54202020204d4f44200000ea48
883b883b0000ea48883b370e2f20000041660073005f0066
0069000f00216c0065002e006d006f00640000000000ffff
46535f46494c45204d4f44200000ea48883b883b0000ea48
883b490e7e0600004166007300680065006c000f00b47000
2e006d006f00640000000000ffffffff465348454c502020
4d4f44200000ea48883b883b0000ea48883b4d0e9b0b0000
41660073005f00750075000f002b690064002e006d006f00
640000000000ffff46535f55554944204d4f44200000ea48
883b883b0000ea48883b530e8d0600004167006500740074
0065000f0049780074002e006d006f00640000000000ffff
47455454455854204d4f44200000ea48883b883b0000ea48
883b570e950e000041670066007800720067000f007d6e00

# 2d ls -l (burgvg-tstlv)
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:561: Ext data-
>sblock.magic = 0x0000380038
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:566: size: 336
/Sources/jso/Deb.grub2/burg-bzr1784/fs/ext2.c:572:
5400d97e500bf8005520421e5529c42e7d4a03787d605b78
5300f0427d494b78901f0010913f00144e80042138000001
2f830000419e000838000000812100002f80000191390000
7f21cb78409e008ca01f000c5409442e5400c23e7d200378
7fbd0214813e0578801e05747c390b787fa5eb7838c00008
7ea7ab78552a421e5528c00e552b021e7d085378556bc23e
540a463e7d085b785529463e540bc00e7d2943787d4b5b78
540a421e2b0900007d4b5b785400c42e388000007fc3f378
2f8900007d6003784199fe98409e000c7f80e840419dfe8c
7fc3f378480000013d20000080690000397f015048000000
7c2b0b789421fff03c8000007c0802a63884000038840014
90010014480000017c7f1b783c6000003863000048000001
3d2000003961001093e9000048000000386000004bffffbc
9421fff03c8000007c0802a63c6000003884000038630000

I will so now check how the 'device' structure passed to grub_fs_probe
is filled up.

Cheers,
J.

PS: any chance that making structure 'static' could help a bit (to
avoid issue with gcc optimization?)

Bean

unread,
Dec 11, 2009, 5:32:09 AM12/11/09
to burg-...@googlegroups.com
Hi,

I think this is related to disk read, the first line to read from disk is

grub_disk_read (disk, 1 * 2, 0, sizeof (struct grub_ext2_sblock),
&data->sblock);

from grub_ext2_mount, add some marker around this line, and enable
disk debug. The call should end up in grub_ofdisk_read function of
disk/ieee1275/ofdisk.c.

For grub-fstest, the read should endup in grub_loopback_read in
disk/loopback.c, there is no grub_dprint line here, add one similar to
grub_ofdisk_read.

Now compare the results, see if they read from the same sectors.

Bean

unread,
Dec 11, 2009, 1:27:38 PM12/11/09
to burg-...@googlegroups.com
Hi,

I finally track down a bug in ieee1275 which I believe to be the
cause, please try r1787.

Jo

unread,
Dec 14, 2009, 5:19:42 AM12/14/09
to burg-devel
Hello,


On Dec 11, 7:27 pm, Bean <bean12...@gmail.com> wrote:
> Hi,
>
> I finally track down a bug in ieee1275 which I believe to be the
> cause, please try r1787.
>
> --
> Bean
>
Very sorry but doesn't actually help (even r1788):
GNU BURG version 1.97-b

[ Minimal BASH-like line editing is supported. For the first word,
TAB
lists possible command completions. Anywhere else TAB lists
possible
device/file completions. ESC at any time exits. ]

grub>
grub> ls -l
Device burgvg-tstlv: Unknown filesystem
Device p5tst001vg-Sources: Unknown filesystem
Device p5tst001vg-UsrSrc: Unknown filesystem
Device p5tst001vg-home: Unknown filesystem
Device p5tst001vg-tmp: Unknown filesystem
Device p5tst001vg-var: Unknown filesystem
Device disk2: Partition table
Partition disk2,7: Unknown filesystem
Partition disk2,6: Filesystem type ext2, Last modification
time 2009-12-14 08:47:38 Monday,
UUID 3c51c43e-63a7-4ff1-9b1c-cf98addcb7ed
Partition disk2,5: Unknown filesystem
Partition disk2,3: Filesystem type ext2, Last modification
time 2009-12-14 08:47:38 Monday,
UUID 16ef0754-c845-46e9-a948-b9669ee68329
Partition disk2,2: Unknown filesystem
Partition disk2,1: Filesystem type fat, Label grub2, UUID
4adc-336e

grub> ls -l set debuf g=fs,disk
grub> set debug=fs,disk ls -l ls -l (burgvg-
tstlv)
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:260: Opening `burgvg-
tstlv'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:58: Detecting fat...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading `burgvg-
tstlv'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`disk2,2'...
/Sources/jso/Deb.grub2/burg-bzr1788/disk/ieee1275/ofdisk.c:239:
Reading handle 0x18fb600: sector 0x
b980, size 0x8, buf 0x1c16bf0.
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:64: fat detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:58: Detecting ext2...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading `burgvg-
tstlv'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:64: ext2 detection
failed.
Device burgvg-tstlv: /Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:
260: Opening `burgvg-tstlv'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:58: Detecting fat...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading `burgvg-
tstlv'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:64: fat detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:58: Detecting ext2...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading `burgvg-
tstlv'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:64: ext2 detection
failed.
Unknown filesystem/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:348:
Closing `burgvg-tstlv'.

/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:348: Closing `burgvg-
tstlv'.

I also try another lv but not more success ;<(

grub> ls -l (p5tst001vg-var)
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:260: Opening
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:58: Detecting fat...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`disk2,7'...
/Sources/jso/Deb.grub2/burg-bzr1788/disk/ieee1275/ofdisk.c:239:
Reading handle 0x18fb680: sector 0x
a4f9a0, size 0x8, buf 0x1c1bcd0.
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:64: fat detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:58: Detecting ext2...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`disk2,7'...
/Sources/jso/Deb.grub2/burg-bzr1788/disk/ieee1275/ofdisk.c:239:
Reading handle 0x18fb680: sector 0x
a4fba8, size 0x8, buf 0x1c21f40.
Device p5tst001vg-var: /Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:
260: Opening `p5tst001vg-var
'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:58: Detecting fat...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:64: fat detection
failed.
/Sources/jso/Deb.grub2/burg-bzr1788/kern/fs.c:58: Detecting ext2...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
Filesystem type ext2/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:
404: Reading `p5tst001vg-var'..
.
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
, Last modification time 2009-12-14 08:47:38 Monday/Sources/jso/
Deb.grub2/burg-bzr1788/kern/disk.c:
404: Reading `p5tst001vg-var'...

=== > Here above I was even curious of this msg ("Last modification
time") but that's today and I am absolutely sure to haven't mount any
more since at least 4 days ago.

/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:404: Reading
`p5tst001vg-var'...
, UUID 16ef0754-c845-46e9-a948-b9669ee68329/Sources/jso/Deb.grub2/burg-
bzr1788/kern/disk.c:348: Clo
sing `p5tst001vg-var'.

/Sources/jso/Deb.grub2/burg-bzr1788/kern/disk.c:348: Closing
`p5tst001vg-var'.
grub>

And when I re-launch this same cmdl, I don't it any more
(and I tried compilation OOTB (gcc -Os) and previous tips (gcc -O0)
but nothing help yet.)

I will so continue to insert grub_print() according to your previous
advises ;<)

Tx again,
J.

PS: I do not yet reinstall modules and still work with a grub kernel
with the following built-in matter: search minicmd sh normal nmenu
reboot halt
part_msdos ext2 fat ls linux lvm

Bean

unread,
Dec 14, 2009, 7:32:54 AM12/14/09
to burg-...@googlegroups.com
Hi,

If I guess right, the problem is caused by ofdisk and not lvm. Perhaps
you can try formating your small partition as ext3 and see if burg can
access files in it.

Jo

unread,
Dec 14, 2009, 10:11:51 AM12/14/09