> Alan Cox wrote:
> >> The TCP Stall is back in version 2.1.87. I don't know in which
> version
> >> it actually came back, but I just downloaded the kernel so I could
> >> upgrade one of my machines at home.
> >
> >We know. I dont expect it to vanish too near in the future either.
> >
>
> And you are absolutely sure, that this stall is not caused by loosing
> serial interrupts?
>
> I have to enable IDE irq unmasking (hdparm -u1 /dev/hda) to get a
> reasonable throughput
> while downloading files.
> Otherwise the symptom is:
> no data arriving, ping works (of course)
I've been observing the same problem lately. Unmasking the IRQs doesn't
completely solve it.
> PS: I have an Opti Viper chipset (OPTi 82C558 Viper-M ISA+IDE), which
> does not
> enable unmasking at bootup.
> PPS: I have modified the opti621 driver to support native PIO modes for
> this chipset (11% faster).
> Mailing me your interest will speed up the process of putting it
> together and sending it to the corresponding maintainer.
I have a motherboard with these chips, and I'd be interested in trying
your patches.
/proc/pci:
PCI devices found:
Bus 0, device 20, function 0:
IDE interface: OPTi 82C621 (rev 17).
Medium devsel. Fast back-to-back capable. Master Capable. No bursts.
I/O at 0x3000.
Bus 0, device 17, function 0:
Ethernet controller: Realtek 8029 (rev 0).
Medium devsel. IRQ 11.
I/O at 0x6000.
Bus 0, device 1, function 0:
ISA bridge: OPTi 82C558 (rev 17).
Medium devsel. Fast back-to-back capable. Master Capable. No bursts.
Bus 0, device 0, function 0:
Host bridge: OPTi 82C557 (rev 21).
Medium devsel. Fast back-to-back capable. Master Capable. No bursts.
___
Trevor Johnson
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu