Hi Marco and Alberto,

---->To Marco:
Well, I have disabled CACHE because I have an ancient 5282 evaluation board.
The first series of this processor had a bug in the CACHE which avoided the
kernel to run (it always crashed). AFAIK, CACHE is currently fixed, but I
cannot use it!

I have reviewed other errata on old MCF5282 devices and no one seems to be
related with FEC problems. Actually there are some FEC problems when it
works at 10BASE-T, but my network is set to 100Mbps.

---->To Alberto:
Very good dissertation about interrupt levels! In fact I think that FEC
issue can be fixed on this way (I mean considering interrupt levels).
Actually, fec_enet_tx(), which is called by the ISR, calls spin_lock(),
which does not save the interrupt level status!! Instead, I think it should
do spin_lock_irqsave(). By calling spin_lock(), when spin_unlock is called
then interrupt level is set to 0, losing old level. Do you agree? I'm gonna
try to use the original FEC driver but substituting the spin_lock by the
spin_lock_irqsave.

Regards
_______________________________________________
uClinux-dev mailing list
[email protected]
http://mailman.uclinux.org/mailman/listinfo/uclinux-dev
This message was resent by [email protected]
To unsubscribe see:
http://mailman.uclinux.org/mailman/options/uclinux-dev

Reply via email to