lkml.org 
[lkml]   [1996]   [Apr]   [24]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: As 2.0 looms
Date

>
> Firstly we need to categorise each existing "Alpha", "experimental" or whatever
> driver appropriately. Some drivers marked ALPHA are most definitely not now
> ALPHA. Other stuff like the masquerade proxy agents and multicast routing
> need to be ALPHA clearly.

Yes. And the 'Do you want to be offered ALPHA test drivers' needs to be
moved to the top of ./drivers/net/Config.in so that it can be applied to
*all* drivers, and not just ether drivers as it is at present.

Also the 'EtherExpress support' should read 'EtherExpress-16 support'
as lots of people seem to get confused on this one. Speaking of which,
I did some short tests on a couple of ee-16 cards at 1.3.90, involving
simultaneous Rx and Tx ftp of a 4MB file (kernel source) cached in RAM,
and nothing broke. So this might be one candidate to "de-alpha" if
others using them continuously have also not encountered problems.

The ac3200 driver has been around for three years now and can probably
be moved off the alpha list as well. It is an EISA shared mem 8390
implementation and the only reason it is alpha is because there
have only been one or two linux users with these cards. The 8390 core
certainly isn't alpha, and shared mem "wrappers" around the 8390 core
are pretty simple.

The 3c507 was pretty much useless the last time I tried it, and I don't
think it has changed at all (not counting generic changes like the
multicast handling, etc.) in 1.3.x, so it should remain alpha. Not
surprising, as it is another dain-bramaged i82586 card.

A lot of the other cards listed as "alpha" are there simply due to
the fact that they are relatively uncommon, and thus feedback is low
to nonexistent. Not sure what you want to do about those...

>
> Secondly a load of error/debug messages need fixing - the giant barf from
> a floppy driver timeout, the routing cache debug, the arp cache debug etc

And many need to be tagged with KERN_DEBUG or at least KERN_INFO. For
example just starting "xcdplayer" and clicking on the buttons randomly
can cause the kernel to flood the screen with:

sr.c: "CDROM not ready. Make sure you have a disc in the drive."
sr_ioctl.c: "CDROM (ioctl) reports ILLEGAL REQUEST."

Another one that should be KERN_DEBUG is

keyboard.c: "keyboard error"

I'm sure others have some more to add to the list.
Paul.


\
 
 \ /
  Last update: 2005-03-22 13:37    [W:0.296 / U:2.444 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site