|
From: Sven L. <lu...@dp...> - 2001-05-16 08:19:23
|
On Tue, May 15, 2001 at 07:18:52PM +0200, Giorgio Terzi wrote: > Hello Christian, hello all, > > ... > > Floppys on m68k are completely useless AFAIK. Since Amigalilo is not > > supported by Debian, who needs floppies anyways when everything comes on CD? > > I have "loaded" from floppies the images i have put on them only > for the sake of test! > > I have not bootstrapped with rescue floppy-disk! > > I also know that is impossible for Amiga to bootstrap with a MS-DOS floppy > disk. > But i have tested that, if someone is mad like me, :)) he can load and > configure > the kernel and the drivers by means of floppy disks , is it so bad ? > Debian dbootstrap program gives also this chance! > > I have tried to bootstrap APUS with a 1,76K Amiga floppy-disk and with > this format the problem is not the space, but in bootstrap program > (i use Warp-Up) that after the loading of the (of course compressed) > kernel crashes. what program are you using on the floppy disk ? I suppose that if it is the same as the one you use normally, it would be dependent on warp-up. But i guess warp-up is not in rom, so did you put it on the floppy also ? Did you try it with the standard powerup stuff ? Since you are booting from a floppy, this should be the standard setting, is it not ? Could you provide us with some logs (bootstrap & dmesg) of the crashed launch ? Post it to the linux-apus list please. Also the script you use for launching could be usefull. > I am trying to understand why... > For an 880k floppy the space problem makes impossible to do it. > But i wish to be contraddicted... :)) Well, we could split the kernel in 2 and join it in ram: before launching it. Or we could try for a less than 880Ko modular kernel, but i have not big hopes on this one. Michel Daenzer is the apus kernel package maintainer for debian. Both solution may be used, if there is time to try it out and such. > For Sven Luther: > I was very busy in the last days, and i think i have lost some steps.:)) Ok, i understand you. BTW, Upto now, you were able to bootstarp from the floppy disk set, and able to install and configure the kernel & modules, isn't it. All this with the 2.2.10 kernel. (btw did you solve the 2.2.19 requirement of install.sh ? did you try the little patch i sent to you yesterday or the day before ? > Please may you list me what do you need for the next tests ? > Thank you. Ok, now the next step would be : 1) download the base tarball of woody (if there is such a thing) and try the installation of it. If this works, then i would call the basic apus boot floppies useable. You could try the networked (trough ppp) install after that, but i guess this should be ok. 2) you could try it with an 2.4.x kernel, to see if it works also, and to see if much changes where needed for it, what do you think ? 3) coordinate with alan for the floppy disk booting thingy. That's all i see for now. Friendly, Sven Luther |
|
From: Alan B. <al...@ms...> - 2001-05-16 10:05:13
|
hi, > Well, we could split the kernel in 2 and join it in ram: before launching it. > Or we could try for a less than 880Ko modular kernel, but i have not big hopes > on this one. Michel Daenzer is the apus kernel package maintainer for debian. a small, but brief note: IS there not a very very big problem with having the kernel in ram: when launching? joining two files up into ram: is no problem (6 lines of code max.) but doesnt ram: just get completely trashed when the bootstrap loads up the kernel (and thus corrupting the kernel image?) i thought this was the case and people have to have the installer and kernel on media, rather than in volatile ram: > 3) coordinate with alan for the floppy disk booting thingy. working on disk-layout, tools and required libraries now. alan |
|
From: Geert U. <ge...@li...> - 2001-05-16 10:10:50
|
On Wed, 16 May 2001, Alan Buxey wrote:
> > Well, we could split the kernel in 2 and join it in ram: before launching it.
> > Or we could try for a less than 880Ko modular kernel, but i have not big hopes
> > on this one. Michel Daenzer is the apus kernel package maintainer for debian.
>
> a small, but brief note:
>
> IS there not a very very big problem with having the kernel in ram: when
> launching?
>
> joining two files up into ram: is no problem (6 lines of code max.) but
> doesnt ram: just get completely trashed when the bootstrap loads up the
> kernel (and thus corrupting the kernel image?)
No, the bootstrap first allocates memory for the kernel (and the ramdisk), then
loads the image, and only then trashes the rest.
> i thought this was the case and people have to have the installer and
> kernel on media, rather than in volatile ram:
The only problem is that people might not have enough memory for both RAM: and
kernel/ramdisk.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@li...
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
|
|
From: Sven L. <lu...@dp...> - 2001-05-17 10:40:11
|
On Wed, May 16, 2001 at 12:09:02PM +0200, Geert Uytterhoeven wrote: > On Wed, 16 May 2001, Alan Buxey wrote: > > > Well, we could split the kernel in 2 and join it in ram: before launching it. > > > Or we could try for a less than 880Ko modular kernel, but i have not big hopes > > > on this one. Michel Daenzer is the apus kernel package maintainer for debian. > > > > a small, but brief note: > > > > IS there not a very very big problem with having the kernel in ram: when > > launching? > > > > joining two files up into ram: is no problem (6 lines of code max.) but > > doesnt ram: just get completely trashed when the bootstrap loads up the > > kernel (and thus corrupting the kernel image?) > > No, the bootstrap first allocates memory for the kernel (and the ramdisk), then > loads the image, and only then trashes the rest. > > > i thought this was the case and people have to have the installer and > > kernel on media, rather than in volatile ram: > > The only problem is that people might not have enough memory for both RAM: and > kernel/ramdisk. huh, how much is this ? i guess most apus users have at least 8MB by now, is that not enough ? Let's start a query on sourceforge to have an idea ? Or maybe we could clone the linux/m68k registry pages to say only apus and have them on sourceforge ? what do you think geert ? Friendly, Sven Luther |
|
From: Geert U. <ge...@li...> - 2001-05-17 11:13:19
|
On Thu, 17 May 2001, Sven LUTHER wrote:
> On Wed, May 16, 2001 at 12:09:02PM +0200, Geert Uytterhoeven wrote:
> > The only problem is that people might not have enough memory for both RAM: and
> > kernel/ramdisk.
>
> huh, how much is this ? i guess most apus users have at least 8MB by now, is
> that not enough ? Let's start a query on sourceforge to have an idea ?
>
> Or maybe we could clone the linux/m68k registry pages to say only apus and
> have them on sourceforge ? what do you think geert ?
Robert Ramiega took over maintenance of the Linux/m68k registry about 2 years
ago. However, the registry wasn't updated since Feb 2000.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@li...
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
|
|
From: Sven L. <lu...@dp...> - 2001-05-17 10:38:29
|
On Wed, May 16, 2001 at 11:04:51AM +0100, Alan Buxey wrote: > hi, > > > Well, we could split the kernel in 2 and join it in ram: before launching it. > > Or we could try for a less than 880Ko modular kernel, but i have not big hopes > > on this one. Michel Daenzer is the apus kernel package maintainer for debian. > > a small, but brief note: > > > IS there not a very very big problem with having the kernel in ram: when > launching? Non, i used to do it before, when there was the interrupt issue with the internal hardrive or the scsi one, don't remember. At that time you had to copy the stuff to ram:, wait for a bit that no more disk interrupts were pending, and then launch the stuff. Unless it changed since then, that is. > joining two files up into ram: is no problem (6 lines of code max.) but > doesnt ram: just get completely trashed when the bootstrap loads up the > kernel (and thus corrupting the kernel image?) Yes, but at that time, bootstrap will already have copied the kernel code to it's rigthfull place, so no problem. > i thought this was the case and people have to have the installer and > kernel on media, rather than in volatile ram: ram: in this case is just another media. > > 3) coordinate with alan for the floppy disk booting thingy. > > working on disk-layout, tools and required libraries now. :))) Friendly, sven Luther |
|
From: Giorgio T. <de...@ip...> - 2001-05-16 17:12:19
|
Hello Sven
...
> what program are you using on the floppy disk ? I suppose that if it is the
> same as the one you use normally, it would be dependent on warp-up. But i
> guess warp-up is not in rom, so did you put it on the floppy also ?
I am trying to use our classic kernel loader " `bootstrap` ".
> Did you try it with the standard powerup stuff ? Since you are booting from a
> floppy, this should be the standard setting, is it not ?
I shall test it...
> Could you provide us with some logs (bootstrap & dmesg) of the crashed launch
> ? Post it to the linux-apus list please. Also the script you use for
> launching could be usefull.
>> I am trying to understand why...
OK, i finish my tests and if i can't find solutions i will.
>> For an 880k floppy the space problem makes impossible to do it.
>> But i wish to be contraddicted... :))
>
> Well, we could split the kernel in 2 and join it in ram: before launching it.
> Or we could try for a less than 880Ko modular kernel, but i have not big
> hopes on this one. Michel Daenzer is the apus kernel package maintainer for
> debian.
As rightly written by Geert Uytterhoeven is possible to join in ram the kernel
but
as he also writes there maybe problems with Amigas with a small RAM amount.
> BTW,
>
> Upto now, you were able to bootstarp from the floppy disk set,
No... because as written on top this letter i am testing " `bootstrap` "
on floppy.
and able to
> install and configure the kernel & modules, isn't it. All this with the
> 2.2.10 kernel.
Yes, also using floppy-disks.
(btw did you solve the 2.2.19 requirement of install.sh ? did
> you try the little patch i sent to you yesterday or the day before ?
Right! , another thing i must put in the test's list.
> floppies useable. You could try the networked (trough ppp) install after
> that, but i guess this should be ok.
With ppp ? Nice...
But i was not able to find a ppp configuration mask in the
network section of dbootstrap program...
is it a new feature?
I haven't found the woody's base tarball, maybe is still early,
or maybe i have not found its right path,
but its path is so in ftp as in CD-ROM i think:
dists/woody/disks-powerpc/current/
is it right ?
Thank you again,
Regards
--
Giorgio Terzi
|
|
From: Duncan G. <DG...@Du...> - 2001-05-16 17:55:54
|
On 16-May-01, Giorgio Terzi wrote: >> Well, we could split the kernel in 2 and join it in ram: before >> launching it. Or we could try for a less than 880Ko modular kernel, but >> i have not big hopes on this one. Michel Daenzer is the apus kernel >> package maintainer for debian. GT> As rightly written by Geert Uytterhoeven is possible to join in GT> ram the kernel but as he also writes there maybe problems with GT> Amigas with a small RAM amount. Is this a relevant problem for APUS? How many users have PowerPC boards and less than the critical amount of RAM (8MB?)? Duncan -- ... "That's interesting," said Dougal, turning around in circles. |
|
From: Sven L. <lu...@dp...> - 2001-05-17 10:44:25
|
On Wed, May 16, 2001 at 07:01:33PM +0200, Giorgio Terzi wrote: > Hello Sven > > ... > > what program are you using on the floppy disk ? I suppose that if it is the > > same as the one you use normally, it would be dependent on warp-up. But i > > guess warp-up is not in rom, so did you put it on the floppy also ? > > I am trying to use our classic kernel loader " `bootstrap` ". > > > Did you try it with the standard powerup stuff ? Since you are booting from a > > floppy, this should be the standard setting, is it not ? > > I shall test it... > > > Could you provide us with some logs (bootstrap & dmesg) of the crashed launch > > ? Post it to the linux-apus list please. Also the script you use for > > launching could be usefull. > > >> I am trying to understand why... > > OK, i finish my tests and if i can't find solutions i will. > > >> For an 880k floppy the space problem makes impossible to do it. > >> But i wish to be contraddicted... :)) > > > > Well, we could split the kernel in 2 and join it in ram: before launching it. > > Or we could try for a less than 880Ko modular kernel, but i have not big > > hopes on this one. Michel Daenzer is the apus kernel package maintainer for > > debian. > > As rightly written by Geert Uytterhoeven is possible to join in ram the kernel > but > as he also writes there maybe problems with Amigas with a small RAM amount. > > > BTW, > > > > Upto now, you were able to bootstarp from the floppy disk set, > > No... because as written on top this letter i am testing " `bootstrap` " > on floppy. > > and able to > > install and configure the kernel & modules, isn't it. All this with the > > 2.2.10 kernel. > > Yes, also using floppy-disks. > > (btw did you solve the 2.2.19 requirement of install.sh ? did > > you try the little patch i sent to you yesterday or the day before ? > > Right! , another thing i must put in the test's list. > > > floppies useable. You could try the networked (trough ppp) install after > > that, but i guess this should be ok. > > With ppp ? Nice... > But i was not able to find a ppp configuration mask in the > network section of dbootstrap program... > is it a new feature? Just select the configure the network menu item, and it should ask you if youare going to use ppp or eth, if you choose ppp, i think it will ask you for all relevant info, and configure pppd for you (even fire up the connection before downloading base.attar.gz). Not sure this is still so though, but i guess it is. Nothing apus specific here thoguh.? > I haven't found the woody's base tarball, maybe is still early, > or maybe i have not found its right path, > but its path is so in ftp as in CD-ROM i think: > > dists/woody/disks-powerpc/current/ yes, i think, not sure though. Maybe there is not yet such a thing, is there a i386 one already ? BTW, normally you should be able to build it from the archive during the boot floppies build. It was so for potato. Friendly, Sven Luther |
|
From: Giorgio T. <de...@ip...> - 2001-05-18 16:36:22
Attachments:
powerpc.rules
|
Hello Sven,
here some test results plus other informations
questions & bugs
1)I have tested your patch, unfortunately it does not work,
but don't worry, i have found i think, a solution
for the 2.2.10 kernel version problem.
I attack the file powerpc.rules completely (is 800 bytes)
so you can sobstitute it in the boot-floppies/make
directory if this my modify will be accepted.
2) Is possible to load the root image with 2.4.3 kernel.
3) About the '"bootstrap --apus etc... program'" troubles,
seems i have found also a solution:
the problem was that Amiga setpatch program must be
present in the startup-sequence before "'bootstrap"'
here is a list of the files contained in the Amiga floppy
(1,76M) from which i boot Linux:
-------------------------------------------------------------
libs (dir)
68040.library 68040new.library
68060.library ppc.library
ppcdiss.library
s (dir)
startup-sequence
c (dir)
Assign bootstrap
Copy Execute
MakeDir ppcboot_pup
ppcboot_wup SetPatch
t (dir)
config.gz debian.txt
install.sh linux
readme.txt sys_map.gz
type.txt
-----------------------------------------------------------
and here the startup-sequence
----------------------------------------------------------
c:setpatch QUIET
makedir ram:env
makedir ram:t
assign env: ram:env
assign t: ram:t
copy sys:linux ram:vmlinux
copy c:ppc#? ram: >NIL:
copy c:bootstrap ram:
cd ram:
bootstrap --apus "root=/dev/hdb2 video=pm2fb:mode:1152x864-70"
---------------------------------------------------------
With WarpUp also works, with the little extra work of
to copy and setup its env vars from the floppy.
4) Do exists an utility like mkdosfs to Amigados format
floppy-disks under Linux ?
5) Lastly, but not least, a strange misterious bug...:
I think this is for Nicholai Benalal & | Michel Danzer
Compressed kernels:
If i gzip an APUS 2.2.10 kernel, more explicitly the one i have taken
from the Apus site it works with '"bootstrap'",
but if i use the 2.2.10 APUS compressed kernel extracted from the Debian
Package '"bootstrap'" complains that the kernel has a decompression
crc error
here is the output:
------------------------------------------------------------------
Linux/m68k Amiga Bootstrap version 5.6
Copyright 1993,1994 by Hamish Macdonald and Greg Harp
Amiga 4000 CPU: 68060 with internal FPU, AGA chipset
APUS: No BPPC+ SCSI interface detected.
APUS: PCI Bridge detected.
Command line is 'root=/dev/hdb2 video=pm2fb:mode:1152x864-70 video=amifb:vga
video=map:000011 single'
Vertical Blank Frequency: 50Hz
Power Supply Frequency: 50Hz
EClock Frequency: 709379Hz
Found 5 AutoConfig Devices
Device 0: addr = 0x40000000
Device 1: addr = 0x00e90000
Device 2: addr = 0x00ea0000
Device 3: addr = 0x00eb0000
Device 4: addr = 0x00f01060
Found 1 Block of Memory
Block 0: 0x08000000 to 0x0ff80000 (130560K)
2048K of CHIP memory
The kernel will be located at 0x08000000
Kernel is compressed
Uncompressing kernel image
................................................................
Compressed image too large! Aborting.
crc errorDecompression error -- aborting
----------------------------------------------------------------
It can be gunzipped without errors ( from the Amiga side i use ADE
gzip1.2.4) and also used uncompressed, but , the mistery is more
misterious ;-) , if i gzip the kernel another time, also without using the
-9 option the problem returns.
Regards
--
Giorgio Terzi
|
|
From: Michel <mic...@ii...> - 2001-05-19 17:22:47
|
Giorgio Terzi wrote: > Compressed image too large! Aborting. > > crc errorDecompression error -- aborting Looks like an old boothack with a kernel size limit. We should probably upgrade to the latest one where that limit is lifted. -- Earthling Michel Dänzer (MrCooper) \ Debian GNU/Linux (powerpc) developer CS student, Free Software enthusiast \ XFree86 and DRI project member |
|
From: Sven L. <lu...@dp...> - 2001-05-22 13:08:40
|
On Sat, May 19, 2001 at 07:22:18PM +0200, Michel D=E4nzer wrote: > Giorgio Terzi wrote: >=20 > > Compressed image too large! Aborting. > >=20 > > crc errorDecompression error -- aborting >=20 > Looks like an old boothack with a kernel size limit. We should probably > upgrade to the latest one where that limit is lifted. Giorgio, you are taking the bootstrap from the boot-floppies package, isn= 't it ? It is indeed old (from around the potato freeze). We just need to replace it with a more modern version. Which one is the appropriate one ? I will check it in the boot floppies C= VS tree. more on the other tsuff later on, i have not much time right now. Friendly, Sven Luther |
|
From: Giorgio T. <de...@ip...> - 2001-06-07 18:30:45
|
Hello Sven, only to say that i have uploaded the boot-floopies images as you have asked me weeks ago, on the site ftp.lambda.u-strasbg.fr since 28-5-2001. I had also sent to you a mail to inform you about. I am sending this in the case that, that mail was unfortunately lost :-) I have also uploaded there a readme file with some notes. Kind Regards, G. Terzi |
|
From: Sven L. <lu...@dp...> - 2001-06-08 12:40:30
|
On Thu, Jun 07, 2001 at 08:27:33PM +0200, Giorgio Terzi wrote: > Hello Sven, > > only to say that i have uploaded the boot-floopies images as you have > asked me weeks ago, on the site ftp.lambda.u-strasbg.fr since 28-5-2001. > I had also sent to you a mail to inform you about. > > I am sending this in the case that, that mail was unfortunately lost :-) > > I have also uploaded there a readme file with some notes. Sorry, i was really busy this past days, will look at it next week or maybe this weekend if i have time. I cannot test them though. BEst would be for them to go to the linux-apus sourceforge site. I will try to integrate any changes though, ... Friendly, Sven Luther |
|
From: Michel <mic...@ii...> - 2001-05-16 17:59:22
|
Sven LUTHER wrote: > > I am trying to understand why... > > For an 880k floppy the space problem makes impossible to do it. > > But i wish to be contraddicted... :)) > > Well, we could split the kernel in 2 and join it in ram: before launching > it. Or we could try for a less than 880Ko modular kernel, but i have not big > hopes on this one. I am experimenting, and it looks promising. I basically moved everything except video and disk drivers to modules and here's what it gives so far: michdaen@pismo> gzip -cv9 vmlinux > vmapus.gz ~/src/apus-cvs/2.4 vmlinux: 60.1% michdaen@pismo> ll vmapus.gz ~/src/apus-cvs/2.4 -rw-r--r-- 1 michdaen src 888321 May 16 19:49 vmapus.gz michdaen@pismo> du -k vmapus.gz ~/src/apus-cvs/2.4 868 vmapus.gz Is this small enough? I guess I could squeeze out a few more K... Do we need support for foreign partition maps? Do the fbcon packed pixels modules work? Do we need Minix fs support yet? The other side of the story will probably be the space needed for the modules... -- Earthling Michel Dänzer (MrCooper) \ Debian GNU/Linux (powerpc) developer CS student, Free Software enthusiast \ XFree86 and DRI project member |
|
From: Geert U. <ge...@li...> - 2001-05-16 19:29:58
|
On Wed, 16 May 2001, Michel D=E4nzer wrote:
> Sven LUTHER wrote:
> > > I am trying to understand why...
> > > For an 880k floppy the space problem makes impossible to do it.
> > > But i wish to be contraddicted... :))
> >=20
> > Well, we could split the kernel in 2 and join it in ram: before launc=
hing
> > it. Or we could try for a less than 880Ko modular kernel, but i have =
not big
> > hopes on this one.
>=20
> I am experimenting, and it looks promising. I basically moved everythin=
g
> except video and disk drivers to modules and here's what it gives so fa=
r:
>=20
> michdaen@pismo> gzip -cv9 vmlinux > vmapus.gz ~/src/apus-=
cvs/2.4
> vmlinux: 60.1%
> michdaen@pismo> ll vmapus.gz ~/src/apus-=
cvs/2.4
> -rw-r--r-- 1 michdaen src 888321 May 16 19:49 vmapus.gz
> michdaen@pismo> du -k vmapus.gz ~/src/apus-=
cvs/2.4
> 868 vmapus.gz
>=20
> Is this small enough? I guess I could squeeze out a few more K...
Probably not yet.
> Do we need support for foreign partition maps?
Not for installation.
> Do the fbcon packed pixels modules work?
Yes, but only for modular drivers.
I'd say include fbcon-afb (amifb), fbcon-cfb8 (gfx cards) and fbcon-cfb16
(gfx cards with amiboot -v).
You don't need fbcon-ilbm since amifb doesn't need it unless you use
video=3Damifb:ilbm.
You don't need fbcon-mfb since amifb will fallback to fbcon-afb/fbcon-ilb=
m for
depth 1 when fbcon-mfb is not available.
You don't need fbcon-cfb24 and fbcon-cf32 since no one wants to run an in=
stall
console in those depths anyway. I don't think you need it for half-suppor=
ted
gfx cards with amiboot -v, only depth 16.
> Do we need Minix fs support yet?
Not for installation.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m6=
8k.org
In personal conversations with technical people, I call myself a hacker. =
But
when I'm talking to journalists I just say "programmer" or something like=
that.
-- Linus Torvalds
|
|
From: Michel <mic...@ii...> - 2001-05-16 23:32:11
|
Geert Uytterhoeven wrote: > You don't need fbcon-cfb24 and fbcon-cf32 since no one wants to run an > install console in those depths anyway. I don't think you need it for > half-supported gfx cards with amiboot -v, only depth 16. Will this prevent X from running in those depths? Thanks for your feedback! michdaen@pismo> du -sk src/apus-cvs/2.4/vmapus.gz ~ 856 src/apus-cvs/2.4/vmapus.gz More suggestions welcome... -- Earthling Michel Dänzer (MrCooper) \ Debian GNU/Linux (powerpc) developer CS student, Free Software enthusiast \ XFree86 and DRI project member |
|
From: Geert U. <ge...@li...> - 2001-05-17 07:38:19
|
On Thu, 17 May 2001, Michel D=E4nzer wrote:
> Geert Uytterhoeven wrote:
> > You don't need fbcon-cfb24 and fbcon-cf32 since no one wants to run a=
n
> > install console in those depths anyway. I don't think you need it for
> > half-supported gfx cards with amiboot -v, only depth 16.
>=20
> Will this prevent X from running in those depths?
No, fbcon-* files are used for the console only.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m6=
8k.org
In personal conversations with technical people, I call myself a hacker. =
But
when I'm talking to journalists I just say "programmer" or something like=
that.
-- Linus Torvalds
|
|
From: Alan B. <al...@ms...> - 2001-05-17 10:02:48
|
hi, > > Do we need Minix fs support yet? > > Not for installation. the ramdisk image no longer uses MiniFS? if so minifs could be dropped ages ago. I guess we also dont need OCS gfx support...as PowerPC cards only work on ECS and AGA systems. On this line of thought...sound isnt essential, neither is parallel port?, joystick.. alan |
|
From: Geert U. <ge...@li...> - 2001-05-17 10:18:52
|
On Thu, 17 May 2001, Alan Buxey wrote:
> > > Do we need Minix fs support yet?
> >
> > Not for installation.
>
> the ramdisk image no longer uses MiniFS? if so minifs could be dropped
> ages ago.
Hmm... I don't know. Someone should check this.
> I guess we also dont need OCS gfx support...as PowerPC cards only work on
> ECS and AGA systems.
Yep, you can remove OCS/ECS support, but it'll save only a few bytes, I guess.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@li...
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
|
|
From: Alan B. <al...@ms...> - 2001-05-17 10:37:44
|
hi, > Yep, you can remove OCS/ECS support, but it'll save only a few bytes, I guess. dont A3k PowerUP owners have to have ECS - if they dont have GFX card? (are there any such people?) alan |
|
From: Ken T. <ke...@we...> - 2001-05-17 10:57:15
|
On Thu, 17 May 2001, Alan Buxey wrote: [Cc: list deleted] > On this line of thought...sound isnt essential, neither is parallel port?, > joystick.. Don't think joystick compiles in 2.4 Ken. |
|
From: Geert U. <ge...@li...> - 2001-05-17 11:13:38
|
On Thu, 17 May 2001, Ken Tyler wrote:
> On Thu, 17 May 2001, Alan Buxey wrote:
> [Cc: list deleted]
>
> > On this line of thought...sound isnt essential, neither is parallel port?,
> > joystick..
>
> Don't think joystick compiles in 2.4
I thought I fixed that in late 2.3.x. Damned, someone broke it again?
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@li...
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
|
|
From: Ken T. <ke...@we...> - 2001-05-17 15:50:50
|
On Thu, 17 May 2001, Geert Uytterhoeven wrote: > On Thu, 17 May 2001, Ken Tyler wrote: > > Don't think joystick compiles in 2.4 > > I thought I fixed that in late 2.3.x. Damned, someone broke it again? Has not compiled since 2.2 AFAIK, unless it was fixed in m68k and the changes didn't make it to APUS. Ken. |
|
From: Geert U. <ge...@li...> - 2001-05-17 16:00:38
|
On Fri, 18 May 2001, Ken Tyler wrote:
> On Thu, 17 May 2001, Geert Uytterhoeven wrote:
> > On Thu, 17 May 2001, Ken Tyler wrote:
> > > Don't think joystick compiles in 2.4
> >
> > I thought I fixed that in late 2.3.x. Damned, someone broke it again?
>
> Has not compiled since 2.2 AFAIK, unless it was fixed in m68k and the
> changes didn't make it to APUS.
I remember mailing them to Vojtech, and they even made it to Linus, IIRC.
Of course this doesn't prevent anybody from breaking the stuff again later...
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@li...
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
|