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

Bug#247101: Sarge d-i beta4 installation to Dual-Opteron PC

--- Martin Michlmayr <tbm@cyrius.com> wrote:
> Is SCSI limited to 15 partitions?  I thought SATA is usually done
> witth libata, and hence SCSI emulation.  What are the alternatives?

According to Linux ALLOCATED DEVICES mainteained by John Cagle now:
The descriptions about SCSI disk devices (0-15) states:  
   Partitions are handled in the same way as for IDE
   disks (see major number 3) except that the limit on
   partitions is 15.

> This doesn't matter for the installation; the installer will install a
> proper kernel with highmem support (at least it will if you use the
> full CD, not the netinst image).

> Same here.  It should recognize SMP and install the -smp kernel.

Above these two comments, I will take your words.
> PCMCIA should now not be installed anymore.

> Can you try a daily image from
> http://gluck.debian.org/cdimage/testing/sarge_d-i/ and see if any of
> the issues you've reported there have been fixed.  In particular, can
> you look how well SATA is supported (both with 24 and 2.6, ie boot
> with linux and linux26).  (Although it does use libata afaik).

I have installed two version 2.4 and 2.6. to /dev/hdb6, and 7 using image of
May 26. (Since I do a lot of testing I do not have enough partitions, and
needless to say 15 is not enough at all)    Currently linux24 fails at lilo
installation. (I hate grub)  But through chroot and fixing lilo.conf manually,
I can complete 1st stage installation even though partition step I do not like
at all.  I can see you guys have done a good job.  The only anoying messages
### 1. ###
Skipping Module usb-uhci. It's already loaded.
modprobe: Can't locate module ehci_hcd
Skipping ehci_hcd module; assuming it is compiled into the kernel.
### 2. ###
Starting hotplug subsystem:
 [failed]   net
** can't synthesize pci hotplug events
 [failed]   usb
** can't synthesize root hub events

I think the 1st one is caused by miscommunication between kernel configurator
and usb module people.  The 2nd one: kernel configurator is too ambicious
(everybody should be!); thus causing this message.

I also tried 2.6 image.  linux26 installation simply dies producing (Only last
5 line I produces here)

Using HPET for base-timer
Using HPET for gettimeofday
Detected 1395.976 MHz processor.
Using hpet for high-res timesource
Calibrating delay loop...

This looks a familiar death message which I have encounted developing my image
using 2.6.6, but I cannot recall how I overcame.  I have my config file which
do not produce this message though

Thank you Martin for your reply.  Use of libata and recognizing SATA disk as
ide are mutually exclusive kernel options.  And I know I am minority now
because a lot of people are for ATA SCSI  w/o regard of the benefit of one tiny
but important difference: the maximum # of partitions allowed.

I am currently working on debian Live CD image for AMD64.  And i386 development
is great help to me.  And since I do a lot of testing using many image w/o
regard to other partitions.  I am desparate for large disk and many partitions.
 Some people have suggested me to use LVM, but I doubt if I can do independent
OS installation w/o regard to other partitions.  My less than 510 bytes boot
manager which enables booting from any bios supported boot device is essential
for testing a lot of images you guys or other OS people produces.  If you are
interested in, you can take a look at

Toshikazu Aiyama

Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.

Reply to: