Thread View: comp.os.linux.misc
10 messages
10 total messages
Started by Tom Bird
Sat, 16 May 1998 00:00
Promise caching IDE
Author: Tom Bird
Date: Sat, 16 May 1998 00:00
Date: Sat, 16 May 1998 00:00
13 lines
1086 bytes
1086 bytes
Linux won't talk to my Promise caching IDE controller (VLB) that I just got to upgrade my main Linux box. I have put 4Mb of RAM on it and it works fine in Win95 with no modifications. Linux though refuses to mount the file system (I think the error was 'read beyond end of device' or something similar). There is not a specific kernel (I think). Slackware 3.something with 2.0.0 kernel. TB
Re: Promise caching IDE
Author: Mike Pellatt
Date: Sat, 16 May 1998 00:00
Date: Sat, 16 May 1998 00:00
27 lines
1764 bytes
1764 bytes
On Sat, 16 May 1998 00:34:50 +0100, Tom Bird <hamster@nospam.demon.co.uk> wrote: >Linux won't talk to my Promise caching IDE controller (VLB) that I just got >to upgrade my main Linux box. I have put 4Mb of RAM on it and it works >fine in Win95 with no modifications. > >Linux though refuses to mount the file system (I think the error was 'read >beyond end of device' or something similar). > >There is not a specific kernel (I think). > >Slackware 3.something with 2.0.0 kernel. What size is the hard disk ?? Is it over 1GB ?? If so, this may be something to do with LBA (Logical Block Addressing) used to map drives with >1024 blocks so that the BIOS can address all of the disk. IME, sometimes Linux [1] gets (very) confused over this if what the BIOS tells it and what the partition table tells it disagree. That's the first thing I'd investigate. But maybe there are known problems with the Promise controller - someone with that knowledge may be along soon. [1] Or me. Or both of us. -- Mike Pellatt
Re: Promise caching IDE
Author: Richard Lamont
Date: Sat, 16 May 1998 00:00
Date: Sat, 16 May 1998 00:00
28 lines
1833 bytes
1833 bytes
In article <01bd805a$0dc10f10$0100b51f@dbqa>, "Tom Bird" <hamster@nospam.demon.co.uk> writes: > Linux won't talk to my Promise caching IDE controller (VLB) that I just got > to upgrade my main Linux box. I have put 4Mb of RAM on it and it works > fine in Win95 with no modifications. If it is the DC4030 VLB, then: (a) the main IDE driver doesn't like it. You will need to compile a kernel with the "old" IDE driver instead. (b) there is little point in using a cacheing disk controller with Linux. It is much more cost effective to add memory to the motherboard, because the kernel uses spare memory for disc cacheing anyway. In the end I threw my DC4030 VLB out, and replaced it with a GBP 10 generic VLB controller. These comments all relate to the situation as it was until about a year ago, when I got a new box. -- Richard Lamont To hell with Echelon and NSA/GCHQ spooks! richard@stonix.demon.co.uk Uncripple your Netscape from 40 to 128 bit http://www.stonix.demon.co.uk/ encryption now. http://www.fortify.net/
Re: Promise caching IDE
Author: Tom Bird
Date: Sat, 16 May 1998 00:00
Date: Sat, 16 May 1998 00:00
47 lines
2871 bytes
2871 bytes
Mike Pellatt <mike@mpellatt.demon.co.uk> wrote in article <slrn6lqj94.oo4.mike@mpellatt.demon.co.uk>... > On Sat, 16 May 1998 00:34:50 +0100, Tom Bird <hamster@nospam.demon.co.uk> wrote: > >Linux won't talk to my Promise caching IDE controller (VLB) that I just got > >to upgrade my main Linux box. I have put 4Mb of RAM on it and it works > >fine in Win95 with no modifications. > > > >Linux though refuses to mount the file system (I think the error was 'read > >beyond end of device' or something similar). > > > >There is not a specific kernel (I think). > > > >Slackware 3.something with 2.0.0 kernel. > > What size is the hard disk ?? Is it over 1GB ?? If so, this may be > something to do with LBA (Logical Block Addressing) used to map > drives with >1024 blocks so that the BIOS can address all of the > disk. IME, sometimes Linux [1] gets (very) confused over this if what > the BIOS tells it and what the partition table tells it disagree. > > That's the first thing I'd investigate. But maybe there are known > problems with the Promise controller - someone with that knowledge > may be along soon. Pretty mundane stuff - 260Mb HDD (Seagate) and even when I cut the controller down to it's basics (it has an onboard BIOS that kicks in after the RAM count) by disabling all caching and the faster disk access code - basically making it like a normal VLB disk controller it still didn't start. It works no problem in 95 with a definite speed increase. The card (I'm just running the machine with a normal VLB ATM) I have it here - it says DC4030VL-2 on it and has loads of ROM chips (3) with (C) 1993 Promise tech. they are P3402-E, P3402-O & P3402-X. It has two IDE interfaces and two (!) floppy interfaces. My next course of action would be to try a different mobo. I'll do that when I can be arsed but I suspect that as the thing works perfectly with the existing controller that there is some incompatibility with Linux (like I said, 95 was fine) and I'll have to look into a new kernel (I'm going to go RedHat as soon as 5.1 arrives). TB
Re: Promise caching IDE
Author: Ian Stirling
Date: Sat, 16 May 1998 00:00
Date: Sat, 16 May 1998 00:00
25 lines
2293 bytes
2293 bytes
In demon.tech.unix Mike Pellatt <mike@mpellatt.demon.co.uk> wrote: : On Sat, 16 May 1998 00:34:50 +0100, Tom Bird <hamster@nospam.demon.co.uk> wrote: :>Linux won't talk to my Promise caching IDE controller (VLB) that I just got :>to upgrade my main Linux box. I have put 4Mb of RAM on it and it works :>fine in Win95 with no modifications. :> :>Linux though refuses to mount the file system (I think the error was 'read :>beyond end of device' or something similar). :> :>There is not a specific kernel (I think). :> :>Slackware 3.something with 2.0.0 kernel. : What size is the hard disk ?? Is it over 1GB ?? If so, this may be : something to do with LBA (Logical Block Addressing) used to map Disks over 8G are also much fun to work with. Some older systems get rather confused, when you put in 12G's -- Ian Stirling. Designing a linux PDA, see http://www.mauve.demon.co.uk/ ----- ******* If replying by email, check notices in header ******* ----- If God hadn't intended us to eat animals, He wouldn't have made them out of MEAT! -John Cleese
Re: Promise caching IDE
Author: Tom Bird
Date: Sat, 16 May 1998 00:00
Date: Sat, 16 May 1998 00:00
42 lines
2173 bytes
2173 bytes
Richard Lamont <richard@stonix.demon.co.uk> wrote in article <6jk2qr$21n@stonix.demon.co.uk>... > In article <01bd805a$0dc10f10$0100b51f@dbqa>, > "Tom Bird" <hamster@nospam.demon.co.uk> writes: > > > Linux won't talk to my Promise caching IDE controller (VLB) that I just got > > to upgrade my main Linux box. I have put 4Mb of RAM on it and it works > > fine in Win95 with no modifications. > > If it is the DC4030 VLB, then: That's it. > (a) the main IDE driver doesn't like it. You will need to compile > a kernel with the "old" IDE driver instead. I'm still a Linux newbie - how do I do this? > (b) there is little point in using a cacheing disk controller with > Linux. It is much more cost effective to add memory to the > motherboard, because the kernel uses spare memory for disc > cacheing anyway. The mobo's at its limit now (no free slots and I haven't got / or intend buying any new RAM for it. > In the end I threw my DC4030 VLB out, and replaced it with a > GBP 10 generic VLB controller. That's what it has now. I liked the caching thing. When you say 'thrown out', do you mean in the bin or relegated to the drawer? It seems like a quality card. > These comments all relate to the situation as it was until about > a year ago, when I got a new box. I've just found a driver for my SCSI card so now I will be able to run Linux nearly full time. TB
Re: Promise caching IDE
Author: Nix
Date: Sat, 16 May 1998 00:00
Date: Sat, 16 May 1998 00:00
24 lines
1545 bytes
1545 bytes
"Tom Bird" <hamster@nospam.demon.co.uk> writes: > Linux won't talk to my Promise caching IDE controller (VLB) that I just got > to upgrade my main Linux box. I have put 4Mb of RAM on it and it works > fine in Win95 with no modifications. > > Linux though refuses to mount the file system (I think the error was 'read > beyond end of device' or something similar). > > There is not a specific kernel (I think). > > Slackware 3.something with 2.0.0 kernel. If you go up to a late 2.0.x series kernel and ask to see experimental config features, there's experimental Promise controller support. I don't know whether it works, though; esperi's got an ancient Promise ISA/IDE controller, but it works fine with the normal drivers. And loki's not got [E]IDE at all. -- `cc:Mail is a wonderful application, as long as you don't want to read or send mail.'
Re: Promise caching IDE
Author: Richard Lamont
Date: Sun, 17 May 1998 00:00
Date: Sun, 17 May 1998 00:00
53 lines
2783 bytes
2783 bytes
In article <01bd8113$11c88e80$0100b51f@dbqa>, "Tom Bird" <hamster@nospam.demon.co.uk> writes: > Richard Lamont <richard@stonix.demon.co.uk> wrote in article > <6jk2qr$21n@stonix.demon.co.uk>... >> In article <01bd805a$0dc10f10$0100b51f@dbqa>, >> "Tom Bird" <hamster@nospam.demon.co.uk> writes: >> >> > Linux won't talk to my Promise caching IDE controller (VLB) that I just > got >> > to upgrade my main Linux box. I have put 4Mb of RAM on it and it works >> > fine in Win95 with no modifications. >> >> If it is the DC4030 VLB, then: > > That's it. > >> (a) the main IDE driver doesn't like it. You will need to compile >> a kernel with the "old" IDE driver instead. > > I'm still a Linux newbie - how do I do this? Assuming you have the appropriate stuff installed (basically the d-series and k-series Slackware disks), see /usr/src/linux/README. (The procedure may seem a bit intimidating if you've never done it before, but it's painless once you get the hang of it.) >> (b) there is little point in using a cacheing disk controller with >> Linux. It is much more cost effective to add memory to the >> motherboard, because the kernel uses spare memory for disc >> cacheing anyway. > > The mobo's at its limit now (no free slots and I haven't got / or intend > buying any new RAM for it. > >> In the end I threw my DC4030 VLB out, and replaced it with a >> GBP 10 generic VLB controller. > > That's what it has now. I liked the caching thing. When you say 'thrown > out', do you mean in the bin or relegated to the drawer? Relegated to the cardboard box full of redundant computer bits in the spare room. > It seems like a quality card. That's marketing for you! -- Richard Lamont To hell with Echelon and NSA/GCHQ spooks! richard@stonix.demon.co.uk Uncripple your Netscape from 40 to 128 bit http://www.stonix.demon.co.uk/ encryption now. http://www.fortify.net/
Re: Promise caching IDE
Author: larry suse user
Date: Sun, 17 May 1998 00:00
Date: Sun, 17 May 1998 00:00
53 lines
2486 bytes
2486 bytes
In article <6jk2qr$21n@stonix.demon.co.uk>,
richard@stonix.demon.co.uk (Richard Lamont) writes:
> In article <01bd805a$0dc10f10$0100b51f@dbqa>,
> "Tom Bird" <hamster@nospam.demon.co.uk> writes:
>
>> Linux won't talk to my Promise caching IDE controller (VLB) that I just got
>> to upgrade my main Linux box. I have put 4Mb of RAM on it and it works
>> fine in Win95 with no modifications.
>
> If it is the DC4030 VLB, then:
>
> (a) the main IDE driver doesn't like it. You will need to compile
> a kernel with the "old" IDE driver instead.
> (b) there is little point in using a cacheing disk controller with
> Linux. It is much more cost effective to add memory to the
> motherboard, because the kernel uses spare memory for disc
> cacheing anyway.
>
> In the end I threw my DC4030 VLB out, and replaced it with a
> GBP 10 generic VLB controller.
>
> These comments all relate to the situation as it was until about
> a year ago, when I got a new box.
> Hi Tom.
Around two years ago I was running 2 V.L.B. ,, M.B. I had all sorts
of problems with vlb I.O cards going awol under linux I cured it by
upgrading to pci M.B.s with built in ports.
As computer components are so cheap now it will pay u to source new mb
cpu pci vidio etc etc and more ram as 4 meg is rather stingy for a decent
linux install then install a more recent linux dist. that will support a
pentium mb. Careful sourcing will get you the lot for just over 100 beer
coupons and it will fly !! cheap intel clone aand cheapest pci card will be
much faster than what u have at the mo .. have fun
Larry pleasant suse 5.2 the dogs bollicks!!
--
|=================================================================
Internet larry@g4izc2.keme.co.uk g4izc@gb7eip G4IZC-5 44.131.185.196
The blurb said "Requires windows 3.11 or better" So I installed Linux-2.0.30
Re: Promise caching IDE
Author: mike
Date: Mon, 18 May 1998 00:00
Date: Mon, 18 May 1998 00:00
34 lines
2214 bytes
2214 bytes
Richard Lamont (richard@stonix.demon.co.uk) wrote: : In article <01bd805a$0dc10f10$0100b51f@dbqa>, : "Tom Bird" <hamster@nospam.demon.co.uk> writes: : > Linux won't talk to my Promise caching IDE controller (VLB) that I just got : > to upgrade my main Linux box. I have put 4Mb of RAM on it and it works : > fine in Win95 with no modifications. : If it is the DC4030 VLB, then: : (a) the main IDE driver doesn't like it. You will need to compile : a kernel with the "old" IDE driver instead. : (b) there is little point in using a cacheing disk controller with : Linux. It is much more cost effective to add memory to the : motherboard, because the kernel uses spare memory for disc : cacheing anyway. The latest 2.1.* kernels include support for the 4030 controllers, it's in drivers/block/pdc4030.c. There are apparently some problems with the secondary interface but it is supposed to work. You also need to enable it at runtime. It doesn't however support tape or cdrom only harddisc. However you'll have to upgrade quite a few other programs to work with the newer kernels. It's all documented in Documentation/Changes in the kernel source tree. Whether this is worth doing probably depends on how much you enjoy mucking about and getting things working. Cheers. -- Mike Cavanagh. http://www.emgee.demon.co.uk/ http://turnpike.net/metro/emgee/index.html
Thread Navigation
This is a paginated view of messages in the thread with full content displayed inline.
Messages are displayed in chronological order, with the original post highlighted in green.
Use pagination controls to navigate through all messages in large threads.
Back to All Threads