Skip to contentSkip to conversation

e1000 (on-board CSA) lockup with 2.6.3 on ifconfig

DISCUSSION Alex Scheele 2 participants 2 messages Feb 19, 2004

lore
Share & tools
lore
#0 lore raw

Hi,

The lockup is probably caused by a patch that made it in the tree which adds an interrupt disable/enable to keep interrupt assertion state synced between 82547 and APIC. This patch was in the tree before but was reverted a while back because it locked up 82547 CSA-based LOMs. Apparently it made it back in. Below I have inserted a patch (against 2.6.3) which
removes it again. Feel free to test if it stops the lockups.

Regards,

Alex Scheele

e1000_main_new.c 2004-02-19 12:28:15.000000000 +0100 renamed from e1000_main.c 2004-02-18 04:57:26.000000000 +0100 +0 -16
@@ -2124,26 +2124,10 @@
                __netif_rx_schedule(netdev);
        }
 #else
-        /* Writing IMC and IMS is needed for 82547.
-          Due to Hub Link bus being occupied, an interrupt
-          de-assertion message is not able to be sent.
-          When an interrupt assertion message is generated later,
-          two messages are re-ordered and sent out.
-          That causes APIC to think 82547 is in de-assertion
-          state, while 82547 is in assertion state, resulting
-          in dead lock. Writing IMC forces 82547 into
-          de-assertion state.
-        */
-       if(hw->mac_type == e1000_82547 || hw->mac_type == e1000_82547_rev_2)
-               e1000_irq_disable(adapter);
-
        for(i = 0; i < E1000_MAX_INTR; i++)
                if(!e1000_clean_rx_irq(adapter) &
                   !e1000_clean_tx_irq(adapter))
                        break;
-
-       if(hw->mac_type == e1000_82547 || hw->mac_type == e1000_82547_rev_2)
-               e1000_irq_enable(adapter);
 #endif
 
        return IRQ_HANDLED;
#1 lore raw

tor, 2004-02-19 kl. 12:38 skrev Alex Scheele:

Quoted 8 lines

Hi,

The lockup is probably caused by a patch that made it in the tree which adds an interrupt disable/enable to keep interrupt assertion state synced between 82547 and APIC. This patch was in the tree before but was reverted a while back because it locked up 82547 CSA-based LOMs. Apparently it made it back in. Below I have inserted a patch (against 2.6.3) which removes it again. Feel free to test if it stops the lockups.

It works like a charm. My box now boots with kernel 2.6.3.

Thanks a lot!

Quoted 3 lines

Regards,

Alex Scheele

[patch snipped]