Linux-Hardware Digest #413, Volume #12 Mon, 6 Mar 00 13:13:11 EST
Contents:
Re: CHALLENGE: Port Linux to a Mountain Dew can! ("Robert W. Cunningham")
Aladin V Chipset Question (John Duffy)
need to change /dev/hdc to /dev/hde ([EMAIL PROTECTED])
Re: Looking for Win Lucent modem driver for LINUX ("Jason Byrne")
Trident Blade 3D & Mouse Trouble ([EMAIL PROTECTED])
Re: not sloooow, but sluggish linux modem (J.R. Lockwood)
Re: best graphics card? (Lien-Fei Alex Chu)
Laptop for Linux - what specifications? (Adrian Smith)
Re: Sound blaster pci 128 (Lien-Fei Alex Chu)
Re: 4 Celeron motherboard? (David C.)
Re: Binary compatibility (David C.)
Anyone with Delta 48X CD-ROM? (Jan Boshoff)
Trident 3d Blade, and Mouse Trouble, in Linux ([EMAIL PROTECTED])
Re: Binary compatibility (David C.)
Re: Binary compatibility (David C.)
Re: HELP - Install Linux on Last Partition of a 30 GB Hard Drive? ("Andrew Scutt")
Re: # of logical partition? (Henrik Carlqvist)
Re: SCSI? IDE? Opinions please (Mike Castle)
help with printk (Nils Wygant)
----------------------------------------------------------------------------
From: "Robert W. Cunningham" <[EMAIL PROTECTED]>
Subject: Re: CHALLENGE: Port Linux to a Mountain Dew can!
Date: Mon, 06 Mar 2000 16:10:12 GMT
Lew Pitcher wrote:
> This could be fun <vbg>
>
> On Sat, 04 Mar 2000 19:04:55 GMT, "Robert W. Cunningham"
> <[EMAIL PROTECTED]> wrote:
>
> >advisor wrote:
> >
> >> A friend of mine and I would like to post a public challenge to test the
> >> versatility of linux: we would like to see if someone can port linux to
> >> a mountain dew can - any necessary and reasonable modifications to the
> >> can (addition of transistors, etc.) are acceptable, as long as the end
> >> product can be displayed as something recognizable as a mountain dew can
> >> which runs linux. (consider the marketing potential, folks - the
> >> ultimate extreme soft drink powered by the ultimate extreme OS?)
> >>
> >> No prizes to be awarded, just the honor that comes with being the first
> >> person to ever port an OS to a soft drink.
> >>
> >> So, ladies and gentlemen, start your engines! Let the porting begin!
>
> >Does power have to be self-contained? A power cord would be ugly...
>
> How about an inductance powersource, instead. Although there _are_
> processors that can run off of battery power (even 9volt).
Actually, I'm thinking of one of those batteries from a Polaroid Spectra film
pack. They are thin, flexible, and have a good capacity.
Otherwise, one or two lithium CR123 batteries should do the job.
> >Linux runs on a Palm, and a naked Palm circuit board might just fit inside
> >a soda can... Just have a serial connector exposed for connecting a
> >console device...
>
> Although that may satisfy the letter of the challange, I believe that
> it violates the spirit of the challange. You might win on a
> technicality, but you wouldn't get any points for congeniality.
What seems to be needed is a high-integration low-power part that has most or
all of a Linux port completed for it. My favorite processor of this type is
the Hitachi 70xx series of embedded controller chips (lots of mips from a
trickle of battery power), but I doubt Linux has been ported to it.
Second would be the Motorola Dragonball EZ processor, the same one used in the
Palm.
Third would be one of the many x86 SOC (System On a Chip) designs from any of
several companies. The problem is I do not know if any of them that will
operate very long from batteries. The advantage of such systems is that some
of the x86 SOCs have VGA and sound on-chip.
In that case, what might be cool would be to place a connector for a laptop
port expansion unit on the side of the can (or on the end of a cable coming
from the can), and then just plug the can into a docking adapter for power and
I/O!
> >Get one of those new flexible plastic LCDs to wrap around the outside, and
> >you could even have a display. Just have it show the Mountain Dew can art
> >as a screen saver.
>
> Great Idea!
Unfortunately, they aren't yet available with controller interfaces that are
compatible with traditional LCD controllers. But the 4"x5" flexible displays
themselves are available is sample quantities for about $500 ($2500 if you get
the development kit).
> How about mounting a small piezoelectric button on the inside bottom
> of the can, and using it as a speaker. A _multimedia_ Linux Mountain
> Dew can ;-)
For usability and quality, I think I'd prefer a headphone jack.
But what about connecting a VR headset instead of a normal display? The
notion of a geek wearing a headset connected to a Mountain Dew can and dancing
about playing Quake seems like a ready-made commercial to me...
Frag The Dew!
-BobC
------------------------------
From: John Duffy <[EMAIL PROTECTED]>
Subject: Aladin V Chipset Question
Date: Mon, 6 Mar 2000 16:09:38 +0000
Hi.
Will motherboards using chipsets other than Intel work under Linux?
I am thinking about buying a Super 7 board, either the Gigabyte GA-5AX or the
ASUS P5A, for a second Linux box with an AMD K5 PR100 CPU.
Both of these boards use Aladin V chipsets.
Regards
John Duffy
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED]
Subject: need to change /dev/hdc to /dev/hde
Date: Mon, 06 Mar 2000 16:13:55 GMT
I've got an ABIT BP6 motherboard with four IDE channels.
IDE1 and IDE2 are UDMA33, IDE3 and IDE4 are UDMA66.
The hard drive is plugged into IDE2 and the CD-ROM is plugged into IDE1.
My hard drive device name is /dev/hdc.
However, I want to plug my drive into the UDMA66 IDE3 slot.
I do have a UDMA66 drive and a kernel patch to support the
UDMA66 drive.
When I plug it in there and start the new kernel, I notice it is
looking for a drive at a different device location, /dev/hde.
Is there a way to associate each kernel image with a specific drive
assignment via LILO? So if I type 'linux' at the LILO: prompt, it will
look for it's device at /dev/hdc, but if I type 'newkernel' at the LILO:
prompt, it will know to look for device /dev/hde?
If not, how can I change it so that on my next boot, it will know to
look for /dev/hde for the boot drive/image?
Thanks.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: "Jason Byrne" <[EMAIL PROTECTED]>
Subject: Re: Looking for Win Lucent modem driver for LINUX
Date: Mon, 6 Mar 2000 08:25:37 -0800
try http://www.o2.net/~gromitkc/winmodem.html (very good information)
or http://linmodems.org
money lady <[EMAIL PROTECTED]> wrote in message
news:KIKw4.1348$[EMAIL PROTECTED]...
> Hi
>
> Can someone please help me?? I am looking for an easy to under setup for
a
> driver for my Win LT modem to operate under Corel LINUX.
>
> I am reasonably new to LINUX, need this driver in order to use my modem in
> LINUX.
>
> If you know a site where i can get it, or if you have the driver, i would
> really be greatful to you if you could send the link or driver files to
this
> e-mail address.
>
> Thanks
>
>
>
> Carla
> [EMAIL PROTECTED]
>
>
------------------------------
From: [EMAIL PROTECTED]
Subject: Trident Blade 3D & Mouse Trouble
Date: Mon, 06 Mar 2000 16:20:43 GMT
Hi, I hope someone can help me. I am a Linux
Newbie, but learning the ropes the hard way :X
After much trouble, I got Red Hat 6.1 installed.
I downloaded and installed the alpha software for
my Trident 3d blade video card, and got XFree86 to
start up. I can only get it in 256 colors
though. I have tried forcing it to start with
more, with the command bpp16, it starts, and
looks a little better, but locks up quickly, and
will not even start up with bpp24.
Even at 256, I have a problem with my mouse. I
have an intellimouse with the intellieye.
Sometimes the mouse decides to just disappear, and
the whole system locks up. I have tried countless
different configurations, and although the one I
am running works (ie, the mouse does not get stuck
at the top of the screen) it is irritating to
have the mouse disappear.
I *think* it is a video card issue that I am
dealing with... does anybody know how I can go
about fixing this? After all the time I poured
into getting this far, I want this to work right!!
:)
Thank you for your help,
Danielle
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: J.R. Lockwood <[EMAIL PROTECTED]>
Subject: Re: not sloooow, but sluggish linux modem
Date: Mon, 06 Mar 2000 16:30:06 GMT
Thanks to all for the replies. I tried setting the speed of the serial
port to which the modem is connected to very high using setserial and it
didn't help. To those more experienced than myself -- what other kinds of
bottlenecks are plausible? It can't be my phone line as Windows achieves
the good throughput using exactly the same hardware. Is it possible that
the serial port is fast enough but that the CPU is ignoring it
periodically?
all the best,
J.R.
--
Posted via CNET Help.com
http://www.help.com/
------------------------------
From: Lien-Fei Alex Chu <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.x,uk.comp.os.linux
Subject: Re: best graphics card?
Date: Mon, 06 Mar 2000 11:54:22 -0500
Reply-To: [EMAIL PROTECTED]
Hi Dan:
I will suggest ATI All in WOnder 128. I am not sure if this fits your need.
But, it works great for me under Linux. This card is "hardware support" DVD.
This card work pretty good under Linux. You might have to do a little work to
get it up and running, but you will not have trouble finding the documentation
for that. However, I "do not " know if the driver under Linux for that card
suppport the "DVD" feature.
You can check up the follong website.
http://www4.ncsu.edu/~distclai/rage128-howto.html
Good luck.
Alex.
Dan Law wrote:
> Dirk Mueller wrote:
>
> > Dan Law wrote:
> > >
> > > Can anyone recommend a graphics card supported by Linux (well, XFree86)
> > > that would have a high quality mpeg decoder that can be used in DVD
> > > playback on a pc monitor?
------------------------------
From: Adrian Smith <[EMAIL PROTECTED]>
Subject: Laptop for Linux - what specifications?
Date: Mon, 06 Mar 2000 16:53:53 GMT
I'm hoping to acquire a SH or new low-end laptop to run Linux - can
anyone advise me if there are any hardware configurations to be coveted
(or avoided)? I'm going to have to do it here in France, and my French,
though passable, is not really up to detailed technical discussions, so
I want to be forearmed with as much knowledge as possible before I
start waving money at people with berets.
--
Adrian Smith
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Lien-Fei Alex Chu <[EMAIL PROTECTED]>
Subject: Re: Sound blaster pci 128
Date: Mon, 06 Mar 2000 12:01:52 -0500
Reply-To: [EMAIL PROTECTED]
I will say the card is supported. (RedHat 6.1). I am running RedHat6.0
and using SB 128. However, I did have to go through some trouble to get
it up and running. If you are looking for some solution, you have to
give more detail...
My card uses 1370 module... some use 1371.
Good luck.
Alex.
Felix wrote:
> Is Sound Blaster pci 128 supported ?
> I use the Red hat 6.1 and soundconfig doesn't work.
> My sound card doesn't sound.
------------------------------
From: [EMAIL PROTECTED] (David C.)
Subject: Re: 4 Celeron motherboard?
Date: 06 Mar 2000 12:02:06 -0500
Atle <[EMAIL PROTECTED]> writes:
>
> What I'm trying to say is that once a multiprocessor system becomes
> available, I have full confidence that Linux will support it well
> (imagine the fun in building support for a 8 Athlon system!!!) I just
> hope someone will build one ...
Multiple-CPU systems already exist, for more than just Celeron :-).
Linux already supports them. I've been running Linux on my dual-PPro
box (Micron Millennia Pro2 - Micronics W6Li motherboard) for several
months with no problems.
I installed RedHat 6.1 and didn't recompile anything. It auto-detects
and uses both CPUs. It hasn't crashed in the three months since I added
the second processor.
-- David
------------------------------
From: [EMAIL PROTECTED] (David C.)
Subject: Re: Binary compatibility
Date: 06 Mar 2000 12:07:56 -0500
"Barry Kiernan" <[EMAIL PROTECTED]> writes:
>
> Surely you don't mean that a Pascal program gets memory from the OS
> using one system call while a C program gets memory using another?
Probably correct. Some operating systems may provide more than one call
for a single function - and different compilers may use different ones,
but in general you are correct.
Your C runtime library and your Pascal runtime library both have to
(eventually) make the same system calls.
> The OS supplies one mechanism for allocating memory to user programs;
> each compiler on that platform maps to that system call, whatever form
> the syntax may take in the source language.
More or less. The details of how this works vary from language to
language and from compiler to compiler, but eventually, just about
everything (except raw number-crunching) ends up making system calls,
and these calls are OS-specific, not language- or compiler-specific.
-- David
------------------------------
From: Jan Boshoff <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.mandrake
Subject: Anyone with Delta 48X CD-ROM?
Date: Mon, 06 Mar 2000 12:00:58 -0500
Hi
I'm looking to buy a CD-ROM. Egghead has two hat a good price, the 40X
Asus CD-S400 and a Delta 48X. I'd prefer the Asus, but it's out of
stock.
Does anyone have any experience with a Delta 48X CD-ROM drive? Also, is
there a place on the web where this question of mine would easily be
answered (hardware reviews for cheap stuff - not the newest and
brightest).
Thanks!
Jan Boshoff
------------------------------
From: [EMAIL PROTECTED]
Subject: Trident 3d Blade, and Mouse Trouble, in Linux
Date: Mon, 06 Mar 2000 17:17:13 GMT
Hi, I hope someone can help me. I am a linux Newbie, but learning the
ropes the hard way. After much trouble, I got RH 6.1 installed. I d/l
and installed the alpha software for my Trident 3D Blade video card, and
got xfree86 ot start up. I can only get it in 256 colors though. I
have tried forcing it to start with more, with the command bpp16, it
starts, and looks a little better, but locks up quickly, and will not
even start up with bpp24.
Even at 256, I have a problem with my mouse. I have an intellimouse
with the intellieye. Sometimes the mouse decides to just disappear, and
the whole system locks up. I have tried countless different
configurations, and although the one I am running works (ie. the mouse
does not get stuck at the top of the screen anymore) it is irritating
to have the mouse disappear at random.
I *think* it is a video card issue that I am dealing with, does anybody
know how I can go about fixing this? After all the time I poured into
getting this far, I want this to work right!
Thank you in advance for any help you can offer.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] (David C.)
Subject: Re: Binary compatibility
Date: 06 Mar 2000 12:30:07 -0500
Barry Kiernan writes:
> David C. wrote in message
>> Barry Kiernan writes:
>>>
>>> 2. Even on the same platform (e.g win95) the object code generated
>>> from one source (C say) doesn't work with object code from another
>>> source (pascal for instance). Why not?
>>
>> While the instruction to add two 32-bit memory locations together may
>> be identical, the system calls used are completely different.
>
> Surely you don't mean that a Pascal program gets memory from the OS
> using one system call while a C program gets memory using another? The
> OS supplies one mechanism for allocating memory to user programs; each
> compiler on that platform maps to that system call, whatever form the
> syntax may take in the source language.
I should learn to read questions more thoroughly. After posting one
reply to this, I read some other replies to your message and found that
I misunderstood what you were asking.
You may or may not be able to link code from two languages on the same
operating system together.
The reason is similar to why different operating systems aren't
compatible.
Different languages (and sometimes, different compilers for the same
language) will have different conventions for calling functions.
For instance, when your function calls another function, what really
happens? Well, you know you have to execute an assembly subroutine-jump
call, but there's more than that. You need to pass parameters to the
function, and you need to get return values back. Where do you put
these? You could put them in registers (and if you do, which ones go
where?). You could put them on the stack (in what order?) If you use
the stack, who pops the values off again? (The calling function or the
called function?) etc.
With many compilers, you can specify the calling conventions. It may be
possible, this way, to tell your C compiler that a specific external
function it to be called using Pascal conventions, or something
similar. This can allow the code to be linked together.
If your compiler doesn't have the support, you could still make the
different functions work together by means of some kind of "glue" code -
an assembly-language routine which accepts parameters in one convention
and passes them to another function with a different convention.
If you try and link object modules from different languages where the
calling conventions don't match and you have no glue code to go between
them, you will wind up with code that doesn't work, and may even crash.
Consider all the possible different ways of passing parameters to
functions, and think of what will happen if the function called is using
a different scheme from the function that did the calling.
With modern object-oriented languages (like C++), the conventions become
far more complicated. Now, you've got to deal with virtual methods
(usually involving table-lookups prior to each function call), multiple
inheritcance (more complicated table-lookups), exception handlers (more
funky stack manipulation and alternate ways of returning from
functions), and other complicated things.
With C++, for instance, the calling conventions are so complex that it's
virtually unheard of for two different compilers to have a compatible
convention. As a matter of fact, it is often the case that two
different versions of the same compiler may have incompatible
conventions! (Much to the woe of vendors who want to ship C++ libraries
in binary form - they need to compile separate releases for every
version of every supported compiler.)
-- David
------------------------------
From: [EMAIL PROTECTED] (David C.)
Subject: Re: Binary compatibility
Date: 06 Mar 2000 12:44:25 -0500
[EMAIL PROTECTED] (David C.) writes:
>
> You may or may not be able to link code from two languages on the same
> operating system together.
>
> Different languages (and sometimes, different compilers for the same
> language) will have different conventions for calling functions.
I should also add that, in addition to calling conventions, there will
be run-time library incompatibilities.
For instance, most of the time, your library calls (like C's malloc())
do not merely map onto system calls. They typically do much more.
For example, operating systems don't always have memory-allocation
functions that are convenient for application use.
On OS/2, for instance, you must allocate memory in increments of a page
(4K), and memory usage is very inefficient unless you allocate in
multiples of 64K.
On UNIX, on the other hand, you don't actually allocate memory.
Instead, user-accessible memory is a contiguous block from a fixed
location up to a "break" address. The system call used moves the
location of the break address.
The result is that your runtime library (like the one in your C or
Pascal environment) will not map your explicit allocation calls to
system calls. Instead, they will allocate a large block from the
system, and will do its own management of the memory within that block -
allocating more if necessary.
I think you can easily see how this would lead to a crash if you
allocate memory from a C module and free it from a Pascal module, or
vice versa.
In other words, cross-language compatibility requires more than the mere
ability to make fuction calls between the languages. It's actually a
very hard problem that an operating system can not completely solve.
-- David
------------------------------
From: "Andrew Scutt" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,comp.os.linux.misc,linux.redhat.list
Subject: Re: HELP - Install Linux on Last Partition of a 30 GB Hard Drive?
Date: Mon, 6 Mar 2000 17:41:32 -0000
Actually I've just seen a new release of Lilo that claims to get round the
1024 limit in one of the other Linux groups.....
Scutty
Karel Venken <[EMAIL PROTECTED]> wrote in message
news:89ugsg$634$[EMAIL PROTECTED]...
> If you have Windows 98, you can place a menu in the config.sys referring
to
> eg. windows and another entry to linux and boot in this case with loadlin
> before you load any drivers and so on. Then you do not have to bother with
> this 1024 cylinder limit. I use this for quite some time now. You can even
> place a copy command in your config.sys so that you allways boot
> automatically the last system you ran. Of course, if you have other
> bootloaders this might interfere because then you might first need to
select
> Windows 98. (Eg. as far as I know, you can not boot Windows NT from within
> Windows 98 as you can Linux)
>
>
------------------------------
From: Henrik Carlqvist <[EMAIL PROTECTED]>
Subject: Re: # of logical partition?
Date: Sun, 05 Mar 2000 16:32:51 +0100
chien-yu chen wrote:
> no matter what I try to mount, hde9 or hde10, it'll mount hde9. so I
> was wondering if the partition table has beeen mess up or
> what.
What is the output if you type:
/sbin/fdisk -l /dev/hde
What is the output if you type:
ls -l /dev/hde*
The first command should show you where different partitions live on
disk. If they are overlapping something is very wrong! The second
command should show you major and minor numbers for your devices. The
major number should be 33 and the minor number should be the number of
the partition. If hde9 and hde10 have the same minor number it could
also explain your problems.
regards Henrik
--
spammer strikeback:
root@localhost
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Mike Castle)
Subject: Re: SCSI? IDE? Opinions please
Date: Mon, 06 Mar 2000 17:55:35 GMT
In article <[EMAIL PROTECTED]>,
J Bland <[EMAIL PROTECTED]> wrote:
>My personal preference is to have a main hard disc on scsi, doesn't have to
>be too big but for those things that are loaded often or are going to have lots
Another good option is putting SWAP on SCSI. If you can find some old
40Meg SCSI drives for cheap, stripe SWAP across them.
mrc
--
Mike Castle Life is like a clock: You can work constantly
[EMAIL PROTECTED] and be right all the time, or not work at all
www.netcom.com/~dalgoda/ and be right at least twice a day. -- mrc
We are all of us living in the shadow of Manhattan. -- Watchmen
------------------------------
From: [EMAIL PROTECTED] (Nils Wygant)
Subject: help with printk
Date: Mon, 06 Mar 2000 17:54:10 GMT
I need some help.......
I'm writing a device driver and I can't see any of the debugging
printk's that are made from within the module code. Printk is the kernel
equivalent to printf but it sends it's output to the console. I've read
that when using printk's from within X (i'm using RedHat 6.0 and Gnome)
you won't see the output because ofX's use of "Virtual"
terminals......???
I did a man on xterm and saw a -C option but this had no effect.
It's suppose to redirect all Console output to the active "virtual"
terminal.
It also says that it's not guaranteed to work. o-well.... this was the
case it seems.
Anyway, does anyone have any suggestions?
Thanks,
-Nils
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.hardware) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Hardware Digest
******************************
