3c59x with linuxppc

[Bogdan Costescu]

On Fri, 5 May 2000, Andreas Tobler wrote:

> Do you mean only in vortex_get_stats and in vortex_open there are locks
> necessary?
> Or where exactly does the media selection happen? (sorry newbie Q)

No, in vortex_open no locks are necessary; vortex_get_stats has locks
already.
The media selection happens mainly in vortex_probe1/vortex_open and to
some extent in vortex_timer.

> I tried to replace the spin_locks with the original cli/sti, but I
> didn't manage to get it working so far. Limited time.

Why would you want to do that? You have the original driver (0.99H from
kernels up to 2.2.15) which uses cli/sti. If you want to track the change
that broke your setup, you might want to take the other way around: find
out the exact changes (with diff and by taking a look at the list from
the beginning of the source) and apply them gradually to 0.99H.

As far as I saw from your logs, the driver finishes well the
initialization (vortex_open), set_rx_mode is called several times and then
nothing. The driver has printk's in boomerang_start_xmit and
vortex_interrupt/boomerang_rx (so both Tx and Rx paths which should be
touched if there is some net activity) but they do not appear...
Do you also get these (and only these) messages on the console? (I'm
asking this because some messages may be lost on system crash).

Sincerely,

Bogdan Costescu

IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
E-mail: Bogdan.Costescu@IWR.Uni-Heidelberg.De

-------------------------------------------------------------------
To unsubscribe send a message body containing "unsubscribe"
to linux-vortex-request@beowulf.org