2.4.22-pre9, 8139too driver: eth0: Too much work at interrupt
Hi,
[I searched lkml archives for this, but nothing really useful came up, othewr than it did not happen on much older versions of 2.4.x]
When I transfer any large files via scp or other means on the local 100 Mbps network, I get the following error messages:
eth0: Too much work at interrupt: IntrStatus=0x0040
Are they just debug printk's? My NFS connections stall and sometimes cause errors in programs writing via NFS. (The machine exhibiting the symptoms is the NFS client.)
This is with 2.4.22-pre9, UP IO-APIC and ACPI enabled. The board is a Gigabyte GA7VXP with onboard 8139 ethernet.
This also happens with 2.4.21. I haven't tried any older kernels since the machine is new.
Does anyone have any ideas? thanks in advance.
-m
Re: UPDATE: 2.4.22-pre9, 8139too driver: eth0: Too much work at interrupt
hi.
turns out the problem was my fault. I had been tweaking the pci latency_timer setting for the sound card using "setpci", I had set it too high and the sound card was hogging the bus, so to speak.
Hope this is helpful for others who might be experiencing this.
Anyway, may I just say that 2.4.22-pre9+ is turning out to be one of the best kernels yet for me.
-m
Quoted 25 lines · Marc Heckmann
Hi,
[I searched lkml archives for this, but nothing really useful came up, othewr than it did not happen on much older versions of 2.4.x]
When I transfer any large files via scp or other means on the local 100 Mbps network, I get the following error messages:
eth0: Too much work at interrupt: IntrStatus=0x0040
Are they just debug printk's? My NFS connections stall and sometimes cause errors in programs writing via NFS. (The machine exhibiting the symptoms is the NFS client.)
This is with 2.4.22-pre9, UP IO-APIC and ACPI enabled. The board is a Gigabyte GA7VXP with onboard 8139 ethernet.
This also happens with 2.4.21. I haven't tried any older kernels since the machine is new.
Does anyone have any ideas? thanks in advance.
-m