Time for patches to Linux v1.1?

Byron Thomas Faber

unread,

Apr 5, 1994, 6:52:02 PM4/5/94

to

I know there are about 1/2 dozen patches that could be applied now (easily)
that could take us up to Linux v1.1 alpha.

To name a few:
cluster diffs (scsi)
inline patches
IDE patches
+ others recently released to the kernel channel including:
c++ addition
>4k kmalloc

Questionables:
umsdos
double support
?

Since v1.1 of linux was to be a 'hackers' paradies it would be nice to
release a kernel with these patches in it. It wouldn't take alot of time
& wouldn't need to be stable, and would avoid me having to repatch things
every time I need to patch relative to 1.0.

Anyway, it was just a thought. It seems these would be a good/quick place
to start.

Byron Faber

--
`Playing this disk at loud volume may permanently damage your speakers or
other sound components.' -LFO
b-f...@uiuc.edu

Byron Thomas Faber

unread,

Apr 5, 1994, 9:49:52 PM4/5/94

to

r...@eats.com (Rob Newberry) writes:

>>I know there are about 1/2 dozen patches that could be applied now (easily)
>>that could take us up to Linux v1.1 alpha.

>>To name a few:
>...
>>>4k kmalloc

>Which patch gives >4k kmalloc? Will this fix the problem with reads in Linux'
>NFS server?

>(I always get a "I refuse to malloc X..." error -- and no, I cannot
>set the parameters on the NFS client, because it's a Mac. Just to avoid
>that whole thread again...)

>Rob

It passed through the kernel channel. Its not out for public use yet.

Subscribe to the kernel channel on niksula (?? FAQ question) if you want
more info on whats going on.

Byron

Linus Torvalds

unread,

Apr 6, 1994, 3:43:02 AM4/6/94

to

In article <2nsq2i$k...@vixen.cso.uiuc.edu>,

Byron Thomas Faber <bf1...@ehsn8.cen.uiuc.edu> wrote:
>I know there are about 1/2 dozen patches that could be applied now (easily)
>that could take us up to Linux v1.1 alpha.
>
>To name a few:
>cluster diffs (scsi)
>inline patches
>IDE patches
>+ others recently released to the kernel channel including:
>c++ addition
>>4k kmalloc
>
>Questionables:
>umsdos
>double support
>?
>
>Since v1.1 of linux was to be a 'hackers' paradies it would be nice to
>release a kernel with these patches in it. It wouldn't take alot of time
>& wouldn't need to be stable, and would avoid me having to repatch things
>every time I need to patch relative to 1.0.

I'll start on the v1.1 development tree today. The first version will
just contain the >4kB memory management stuff (not the same patches that
have appeared on the mailing lists; I wrote my own version but with the
same basic buddy-system algorithm). I'll probably tweak it to allow DMA
allocations while I'm at it.

Linus

Paul Burry

unread,

Apr 6, 1994, 9:58:03 PM4/6/94

to

..
|I'll start on the v1.1 development tree today. ...

When could we expect to see a modularized dynamically loaded kernel
like SLS 1.05?

It seems to me that if this is already do-able then implementing it
as early as possible would accelerate the already furious development
rate. Linus' integration effort should also be simpler.

--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Paul Burry
Voice: (613)-991-7329 Internet: pbu...@cse.dnd.ca
Fax: (613)-991-7323

Message has been deleted

Andreas Zeidler

unread,

Apr 7, 1994, 10:48:28 AM4/7/94

to

Hi Linus,

[...]

>I'll start on the v1.1 development tree today. The first version will
>just contain the >4kB memory management stuff (not the same patches that
>have appeared on the mailing lists; I wrote my own version but with the
>same basic buddy-system algorithm). I'll probably tweak it to allow DMA
>allocations while I'm at it.
>
> Linus

Are you going to support the new structure suggestment announced by...
somebody recently? As I can remember, they've already implemented it.
They reduced the kernel to a minimum and made modules out of the
numerous drivers. IMHO that's the way, the development of Linux should
go. There are just too much drivers within the kernel at the moment --
I mean, it's fine to have a driver for all kinds of PC-hardware (and
there's a lot :-)), but those things do not belong to a Unix-Kernel,
but should be add-ons. Now that someone has done exactly those patches,
the time would be right to do it "officially"... :)
What do you think about it?

Andi

ps: Thank you for Linux! -- I just love it... :))

lilo

unread,

Apr 7, 1994, 10:55:00 PM4/7/94

to

Just to second the comment made by someone in this thread. I would love to
see the kernel modularized as much as possible (and I know that's a
non-trivial, time-consuming task). But I'd like to see the various
modularized drivers distributed in the same source tree as the kernel, in
their up-to-date, approved, "production" versions, as a way of ensuring that
a set of production drivers, of the quality we've come to be able to expect,
continues to be maintained with the kernel. Linus, your management of the
kernel complex has been wonderful, and we all really appreciate your
continuing effort! :)

lilo

Rob Newberry

unread,

Apr 6, 1994, 12:29:24 AM4/6/94

to

>I know there are about 1/2 dozen patches that could be applied now (easily)
>that could take us up to Linux v1.1 alpha.

>To name a few: