Re: strange network performance degradation 2.6.8-rc3-mm1 -> 2.6.8-rc4-mm1

[Francois Romieu]

Jurriaan <thunder7@xxxxxxxxx> :
[8139too does not work after switch from 2.6.8-rc3-mm1 to 2.6.8-rc4-mm1]
> The dmesg-output between 2.6.8-rc3-mm1 and 2.6.8-rc4-mm1 differs like
> this:
[...]
> < ACPI: PCI interrupt 0000:02:07.0[A] -> GSI 23 (level, low) -> IRQ 23
> ---
> > ** PCI interrupts are no longer routed automatically.  If this
> > ** causes a device to stop working, it is probably because the
> > ** driver failed to call pci_enable_device().  As a temporary
> > ** workaround, the "pci=routeirq" argument restores the old
> > ** behavior.  If this argument makes the device work again,
> > ** please email the output of "lspci" to bjorn.helgaas@xxxxxx
> > ** so I can fix the driver.
[...]
> This behaviour happens both with /proc/sys/net/ipv4/tcp_window_scaling
> set to 0 or 1 (setting it to 0 or 1 doesn't matter with 2.6.8-rc3-mm1).
> 
> What could be the problem here?

The transition from 2.6.8-rc3-mm1 to 2.6.8-rc4-mm1 changes both the
point outlined above and the 8139too driver.

You may:
1 - try the "pci=routeirq" argument at startup on your 2.6.8-rc4-mm1 if you
    have not done so;
2 - apply to your 2.6.8-rc4-mm1 tree:
    http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.8-rc4-mm1/8139too-mm-revert.patch
3 - apply to your 2.6.8-rc4-mm1 tree:
    http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.8-rc4-mm1/8139too-mm-revert.patch
    http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.8-rc4-mm1/8139too-10.patch
    http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.8-rc4-mm1/8139too-20.patch

Feedback on points 1, 2 and 3 will be welcome.

Please add netdev@xxxxxxxxxxx to the Cc: list as it is probably/almost
surely related to the driver itself.

--
Ueimor
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/