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

Re: Bad Debian/Alpha potato rescue floppy image?

On Monday, 27 Mar, David Huggins-Daines wrote:
> In fact, the new rescue disks aren't bootable by MILO at all because
> there is no space for it (this is MILO's fault for being a pig, aboot
> and APB both fit on a floppy with the kernel just fine).

Nah, it's the kernel that has become too big.  :-)

> Adam Di Carlo <adam@onshore.com> writes:
> > >Oddly enough, I could read the image from Linux (the system currently runs
> > >Redhat 5.2), but an ls from MILO just gave me some % and /s, as well as
> > >failing to boot.  So I made a new image, copying the files from the
> > >original image (I also uncompressed vmzlinux.gz, to make sure the gzipped

The latest MILOs, either 2.0.35-c7b (binary build 000209), or 2.2-15 (or
later), should work.

> The rescue disks are not just plain ext2 filesystems.  They also
> contain SRM boot blocks.  It's possible that srmbootfat is giving
> problems.  Since we don't use it anymore anyway (booting MILO from SRM
> is just a bad idea), there won't be any more such problems.

Booting MILO from SRM is not always a bad idea.  MILO's memory footprint
is 1 Mb less than that of SRM, and there are still low memory systems
around where it is a noticeable advantage.  On the other hand, a wasted
megabyte doesn't really matter for the installation system, so people
shouldn't actually need MILO disks to be SRM bootable.  However, since it
costs us practically nothing (one run of srmbootfat), why take this option
away?  Srmbootfat is absolutely harmless.

> If MILO can't read them while Linux can, then this is a bug in MILO.

A sufficiently old Linux kernel won't read them either (nor will it run on
Alpha, though...).


Reply to: