Re: Big-disk woes
From: Jules (julesrichardsonuk_at_remove.this.yahoo.co.uk)
Date: 01/28/05
- Next message: John Hasler: "Re: Big-disk woes"
- Previous message: Jules: "Re: Big-disk woes"
- In reply to: Bill Davidsen: "Re: Big-disk woes"
- Next in thread: Mark South: "Re: Big-disk woes"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Fri, 28 Jan 2005 14:29:19 +0000
On Fri, 28 Jan 2005 00:05:25 +0000, Bill Davidsen wrote:
> Jules wrote:
>> On Thu, 27 Jan 2005 15:57:08 +0100, Mark South wrote:
>>
>>
>>>On Thu, 27 Jan 2005 14:09:40 +0100, Frederic LIRZIN wrote:
>>>
>>>
>>>>there is a limit at 137 Gb ( so at 128 GB) with old bios systems.
>>>>if it use a 28 bit LBA mode your system is only able to see 137 Gb of
>>>>your disk 2^28 * 512 = 137438953472 b or 137 Gb or 128 GB.
>>>>to see the entire disk it must use 48 bit LBA mode which allow your
>>>>system to adress 2^48 * 512 = 1.44115e+17 b or 1.44115e+8 Gb or
>>>>134217728 GB.
>>>
>>>All true, as it applies to the BIOS.
>>>
>>>
>>>>so you must use the newiest bios with 48 LBA support for your
>>>>motherboard or see about your IDE motherbooard drivers.
>>>
>>>Not necessarily true: I am running Mandrake 10.1 on an Intel mobo with a
>>>BIOS that sees my 160Gb Maxtor as 137Gb. It still boots fine, and Linux
>>>can see and use the entire disk.
>>
>>
>> In the olden days the system still needed the bios to load any boot loader
>> and necessary files from /boot in order to bootstrap the kernel though;
>> it was only when the kernel was loaded the control could be passed over
>> to Linux (which would talk to the hardware directly and not need the BIOS
>> interface to the hard disk)
>>
>> Afraid I've been using SCSI disks for the last *mumble* years though, so
>> I'm not up to speed with what's happening in the IDE world these days.
>
> What olden days were those? What do you think loads the MBR now? There
> is nothing in the machine capable of doing anything when it boots, BUT
> the BIOS.
Sure, but I seem to remember days where the kernel would be in / and not
/boot, so the MBR (under BIOS control) would load the boot loader from
/boot, and that would load the kernel (or boot a different OS). I
presume the switch from BIOS control to direct hardware access only
happened when the kernel was loaded, so both everything in /boot and the
root partition had to be accessible by the BIOS.
Of course a) my memory's hazy as hell these days and b) it may be that
having the kernel in / was just convention at the time and putting it in
/boot would have worked equally well.
Olden days in terms of Linux I always think of as being early to mid
1990's :)
I think I still have SLS Linux on tape somewhere; for giggles
I'll have to try and get it running again one day.
> If you think the MBR is being loaded by anything something
> else, please tell us what it is. I haven't looked at the MBR code
> lately, but 512 bytes is too small for a whole bunch of device drivers,
> so I'm about 95% certain that the stage two loader comes from the BIOS
> i/o, and quite possibly the kernel load as well.
Yep, that would be my take on it too; I don't see how it could have
changed either - I just haven't looked at any of the code or used the
motherboard BIOS to do bootstrap for a long time (relying instead on
the on-board BIOS of the SCSI controller for boot)
cheers
Jules
- Next message: John Hasler: "Re: Big-disk woes"
- Previous message: Jules: "Re: Big-disk woes"
- In reply to: Bill Davidsen: "Re: Big-disk woes"
- Next in thread: Mark South: "Re: Big-disk woes"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
