Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
> LILO may be able to break the 1024 cylinder barrier, but it _cannot_ break > the 8G barrier,. This is bacause of how lilo stores the sector address. The > address is only 24bits + drive specifier (8bits, 1 used as LINEAR address > identifier)+ sector count (8 bits). Yah, the LILO 1st stage loader translates the 'linear' addresses into CHS numbers according th the bios's current CHS geometry before making the call to the bios. > I am already looking into this and maybe in two weeks I will have it ready. > (I already rewrote the primary loader to use both the old and the new int13 > interface). In the current state, I aim at being able to address sector > using 32bits (the Int13 extensios alllow upto 64bits), which would push the > barrier to 2TB. Have you looked into the version at: ftp://sd.dynhost.com/pub/linux/lilo/ That's a modified lilo which uses int13 extensions. Neil.
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
If I can offer a comment on this: The Mandrake 7.0 install scripts could explain the LINEAR option a little better --- the current prompt only says that this option is not recommended for floppy-disks; it should say this option is recommended for large hard disks. I'm only guessing, but I don't think many people would be using the CD install script with the intention of installing LILO on a floppy. -- Gary Lawrence Murphy <[EMAIL PROTECTED]> TeleDynamics Communications Inc Business Telecom Services : Internet Consulting : http://www.teledyn.com Linux/GNU Education Group: http://www.egroups.com/group/linux-education/ "Computers are useless. They can only give you answers."(Pablo Picasso)
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
On Sat, Jan 29, 2000 at 10:52:39AM -0500, Jeff Garzik wrote: > Pixel wrote: > > > > Jeff Garzik <[EMAIL PROTECTED]> writes: > > > > > If you use the linear option, you can bust with 1024 cyl barrier with > > > current production LILO too. > > > > are u sure? lilo does not use the extended bios function > > Haven't tested, but Andre Hedrick said this on lkml very recently... > LILO may be able to break the 1024 cylinder barrier, but it _cannot_ break the 8G barrier,. This is bacause of how lilo stores the sector address. The address is only 24bits + drive specifier (8bits, 1 used as LINEAR address identifier)+ sector count (8 bits). I am already looking into this and maybe in two weeks I will have it ready. (I already rewrote the primary loader to use both the old and the new int13 interface). In the current state, I aim at being able to address sector using 32bits (the Int13 extensios alllow upto 64bits), which would push the barrier to 2TB. -- -- From : Robert 'Niteshadow' Varga email : [EMAIL PROTECTED], [EMAIL PROTECTED] PGPkey: http://hq.alert.sk/~niteshadow/pgpkey.txt -- PGP signature
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
- Original Message - From: Chmouel Boudjnah <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, January 28, 2000 2:38 AM Subject: Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue > Neil Koozer <[EMAIL PROTECTED]> writes: > > > nuni is a boot loader for Linux, using ext2 file system and IDE drives. > > sound nice, pixel ? > > --Chmouel > Take a look at nuni 0.4 here www.linuxforum.com/plug Hoyt
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Pixel wrote: > > Jeff Garzik <[EMAIL PROTECTED]> writes: > > > If you use the linear option, you can bust with 1024 cyl barrier with > > current production LILO too. > > are u sure? lilo does not use the extended bios function Haven't tested, but Andre Hedrick said this on lkml very recently... -- Jeff Garzik | See you at Linux Expo in Paris! Building 1024 | MandrakeSoft, Inc. |
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Jeff Garzik <[EMAIL PROTECTED]> writes: > If you use the linear option, you can bust with 1024 cyl barrier with > current production LILO too. are u sure? lilo does not use the extended bios function
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Pixel wrote: > > Neil Koozer <[EMAIL PROTECTED]> writes: > > > Pixel wrote: > > > > > > Chmouel Boudjnah <[EMAIL PROTECTED]> writes: > > > > > > > Neil Koozer <[EMAIL PROTECTED]> writes: > > > > > > > > > nuni is a boot loader for Linux, using ext2 file system and IDE drives. > > > > > > > > sound nice, pixel ? > > > > > > > > > > grub does the same, and it *does* handle scsi :ppp > > > > I was unaware that grub would boot above 1023 on pre-existing machines > > (the ones that people try to install linux on). I thought it depended > > on the extended bios to do this. The latest version of lilo also > > detects and uses extended bios calls. > > ok, i should have shut up :) > > i saw the extended bios handling in grub, but not in lilo (not the latest > version maybe) It's lilo version 22, but it might not be released. The extended bios stuff was added by another programmer other than Werner, and the last I heard he was trying to get Werner to adopt it as version 22. I've seen postings referring to lilo22, but it's not on metalab, so I'm not sure. > you do have a point, i agree! isn't the extended bios call available on most > modern boxes? I don't know how widespread it is; I haven't figured out a way to find out. I'm just fed up with bios's. I have one PC whose bios can not write correctly to a floppy, so it is impossible to make a boot disk with rawrite (this was my only PC back when I first installed Linux). Neil. p.s. I'll be gone for the weekend, so if there's more discussion on this I won't be able to participate til later.
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Pixel wrote: > > Neil Koozer <[EMAIL PROTECTED]> writes: > > > Pixel wrote: > > > > > > Chmouel Boudjnah <[EMAIL PROTECTED]> writes: > > > > > > > Neil Koozer <[EMAIL PROTECTED]> writes: > > > > > > > > > nuni is a boot loader for Linux, using ext2 file system and IDE drives. > > > > > > > > sound nice, pixel ? > > > > > > > > > > grub does the same, and it *does* handle scsi :ppp > > > > I was unaware that grub would boot above 1023 on pre-existing machines > > (the ones that people try to install linux on). I thought it depended > > on the extended bios to do this. The latest version of lilo also > > detects and uses extended bios calls. > > ok, i should have shut up :) > > i saw the extended bios handling in grub, but not in lilo (not the latest > version maybe) > > you do have a point, i agree! isn't the extended bios call available on most > modern boxes? If you use the linear option, you can bust with 1024 cyl barrier with current production LILO too. Jeff -- Jeff Garzik | See you at Linux Expo in Paris! Building 1024 | MandrakeSoft, Inc. |
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
I located lilo-22 here: ftp://sd.dynhost.com/pub/linux/lilo/ Neil.
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Neil Koozer <[EMAIL PROTECTED]> writes: > Pixel wrote: > > > > Chmouel Boudjnah <[EMAIL PROTECTED]> writes: > > > > > Neil Koozer <[EMAIL PROTECTED]> writes: > > > > > > > nuni is a boot loader for Linux, using ext2 file system and IDE drives. > > > > > > sound nice, pixel ? > > > > > > > grub does the same, and it *does* handle scsi :ppp > > I was unaware that grub would boot above 1023 on pre-existing machines > (the ones that people try to install linux on). I thought it depended > on the extended bios to do this. The latest version of lilo also > detects and uses extended bios calls. ok, i should have shut up :) i saw the extended bios handling in grub, but not in lilo (not the latest version maybe) you do have a point, i agree! isn't the extended bios call available on most modern boxes?
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Pixel wrote: > > Chmouel Boudjnah <[EMAIL PROTECTED]> writes: > > > Neil Koozer <[EMAIL PROTECTED]> writes: > > > > > nuni is a boot loader for Linux, using ext2 file system and IDE drives. > > > > sound nice, pixel ? > > > > grub does the same, and it *does* handle scsi :ppp I was unaware that grub would boot above 1023 on pre-existing machines (the ones that people try to install linux on). I thought it depended on the extended bios to do this. The latest version of lilo also detects and uses extended bios calls. nuni doesn't do scsi :( I was prepared to do that; it just would take a different little code snippet for each type of card. Then we could boot from things like 1502's that don't have an on-board bios patch. But I discovered that those scsi cards are like winmodems; the programming model is not available. > i really don't think relying on the bios for just booting is a bad idea. You > won't ever handle every possible hardware (dma66,scsi,ls120,...) doing this, are > you? It wouldn't be necessary for one loader do handle all cases. if (IDE) then nuni else (other) The case of IDE+ext2 would cover a large fraction of those first-time installers. I would be happy if the bios were adequate, but you know we have tons of first-time installers that hit the bios roadblock. This is the reason for nuni; I don't need it myself. dma66 is a non-issue; those will answer ATA calls (I don't have one to check this though). I presume things like ls120 could be reasonably added; the ATAPI specs are published and there wouldn't be 50 varieties like scsi. Well enough rambling for now, Neil.
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Jeff Garzik wrote: > > On 28 Jan 2000, Chmouel Boudjnah wrote: > > Neil Koozer <[EMAIL PROTECTED]> writes: > > > nuni is a boot loader for Linux, using ext2 file system and IDE drives. > > Does it dual-boot, or is it Linux/ext2 only? > > Jeff It doesn't have a selector of its own yet. It can be selected by any boot manager, or you can put it on floppys (one floppy per os). Neil.
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Chmouel Boudjnah <[EMAIL PROTECTED]> writes: > Neil Koozer <[EMAIL PROTECTED]> writes: > > > nuni is a boot loader for Linux, using ext2 file system and IDE drives. > > sound nice, pixel ? > grub does the same, and it *does* handle scsi :ppp i really don't think relying on the bios for just booting is a bad idea. You won't ever handle every possible hardware (dma66,scsi,ls120,...) doing this, are you?
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
On 28 Jan 2000, Chmouel Boudjnah wrote: > Neil Koozer <[EMAIL PROTECTED]> writes: > > nuni is a boot loader for Linux, using ext2 file system and IDE drives. Does it dual-boot, or is it Linux/ext2 only? Jeff
Re: [Cooker] nuni-0.04 boot loader eliminates 1024 cyl issue
Neil Koozer <[EMAIL PROTECTED]> writes: > nuni is a boot loader for Linux, using ext2 file system and IDE drives. sound nice, pixel ? --Chmouel