Patch for e100 on my z505je
Jeff,
I tried Marcelo's 2.4.21-pre2 and the driver fails the self-test, then dies in an unserviced interrupt storm. An obvious patch seems to fix the problem here.
Amazingly enough, the driver works in RH 2.4.18-14 (e100 2.1.6-k1), but if I take the same 2.1.6-k1 and implant it into Marcelo tree, everything dies as before. Something is fishy here.
-- Pete
drivers/net/e100/e100_main.c +4 -2
@@ -2288,6 +2288,8 @@/* initialize the nic state before running test */ e100_sw_reset (bdp,PORT_SOFTWARE_RESET ); +e100_dis_intr (bdp); +/* Setup the address of the self_test area */ selftest_cmd = bdp->selftest_phys;
@@ -2302,9 +2304,9 @@writel (selftest_cmd, &bdp->scb->scb_port);readw (&(bdp->scb->scb_status));/* flushes last write, read-safe */ -/* Wait at least 10 milliseconds for the self-test to complete */ +/* Wait at least 20 milliseconds for the self-test to complete */ set_current_state (TASK_UNINTERRUPTIBLE ); -schedule_timeout (HZ /100 +1 ); +schedule_timeout (HZ /50 +1 );/* disable interrupts since the're now enabled */ e100_dis_intr (bdp);