[tulip] Interrupt problem (?) with OvisLink Fast Ethernet PCMCIA

David Ford david@kalifornia.com
Wed, 26 Jul 2000 13:25:08 -0700

This is a multi-part message in MIME format.
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Donald Becker wrote:

> You have correctly identified the problem.
> The important detail for the subject line is "2.4.0test5-pre3" -- this is a
> problem with the new kernel PCMCIA interface, not the driver itself.
> You would be better off reporting this to the linux-kernel list.
> Linus seems to want all traffic relating to any OS to be sent there :-<.

Unfortunately unless things have changed, depending on your hardware, no tulip
cards get interrupts delivered if you use the kernel pcmcia stuff.  I have
switched back to the dhinds pcmcia package and it not only is working properly
for tulip cards, but it doesn't crash four times a day; corresponding to
insertion/removal/suspend/resume cycles.

> > tulip-diag.c:v2.00 4/19/2000 Donald Becker (becker@scyld.com)
> > Index #1: Found a Digital DS21143 Tulip adapter at 0x1000.
> ...
> >  Interrupt sources are pending!  CSR5 is f0678046.
> >    Tx complete indication.
> >    Tx out of buffers indication.
> >    Rx Done indication.
> Everything is working fine, except for interrupts.

With my particular laptop, I came to the conclusion that the slots were
misnumbered in the kernel driver. 0,1 was shifted one place which resulted in
slot 0 not being usable period and slot 1 could only do my ray_cs card.  Unless
something has changed, I still suspect this.  In the kernel code, the
tulip_interrupt() (sp?) was never even called.

> The problem is not interrupt sharing.  The problem is the kernel handling
> *any* interrupt.  The TI1225 generates an interrupt for each event
> (e.g. insert or eject), and even those are not being reported.

Agreed.  I registered precisely one interrupt in /proc/interrupts, and that
corresponded to the insertion of the card.


"The difference between 'involvement' and 'commitment' is like an
eggs-and-ham breakfast: the chicken was 'involved' - the pig was

Content-Type: text/x-vcard; charset=us-ascii;
Content-Transfer-Encoding: 7bit
Content-Description: Card for David Ford
Content-Disposition: attachment;

org:<img src="http://www.kalifornia.com/images/paradise.jpg">
title:Blue Labs Developer
fn:David Ford