2003-08-31 13:13:58

by M G Berberich

[permalink] [raw]
Subject: 2.6.0-test4 experiences

Hello,

i tried 2.6.0-test on a laptop (Sony Vaio FX502, AMD Mobile Duron
1GHz)

With a PCMCIA-Modem (ELSA Microlink 56k MC Internet, a V.90 Modem) I
have only 1/3 throughput with 2.6.0-test4 compared to a 2.4.X kernel
(which gives the normal throughput for 56k modem).

Yenta: CardBus bridge found at 0000:00:0a.0 [104d:80f6]
Yenta IRQ list 0008, PCI irq11
Socket status: 30000010
Yenta: CardBus bridge found at 0000:00:0a.1 [104d:80f6]
Yenta IRQ list 0008, PCI irq10
Socket status: 30000006
cs: memory probe 0x0c0000-0x0fffff: excluding 0xc0000-0xd3fff 0xdc000-0xdffff 0xe4000-0xfffff
serial_cs: RequestIRQ: Unsupported mode
ttyS1 at I/O 0x2f8 (irq = 0) is a 16550A
...
ttyS1: 1 input overrun(s)

pulling the card out I get

Trying to free nonexistent resource <000002f8-000002ff>

pushing it in again I get:

serial_cs: RequestIRQ: Unsupported mode
ttyS1 at I/O 0x2f8 (irq = 0) is a 16550A

lspci says:
00:0a.0 CardBus bridge: Texas Instruments PCI1420
00:0a.1 CardBus bridge: Texas Instruments PCI1420

So it seems the card is not getting an interrupt with 2.6.0-test4 for
some reason I don't understand.

BTW: 'serial_cs' has been renamed to '8250-cs' but there are still
references to 'serial_cs' in source, messages, Kconfig ("The module
will be called serial_cs.") and Documentation.

I'm not on kernel-mailing-list.

MfG
bmg

--
"Des is v?llig wurscht, was heut beschlos- | M G Berberich
sen wird: I bin sowieso dagegn!" | [email protected]
(SPD-Stadtrat Kurt Schindler; Regensburg) |


Attachments:
(No filename) (1.56 kB)
(No filename) (189.00 B)
Download all attachments

2003-08-31 13:33:31

by Russell King

[permalink] [raw]
Subject: Re: 2.6.0-test4 experiences

On Sun, Aug 31, 2003 at 03:13:13PM +0200, M G Berberich wrote:
> i tried 2.6.0-test on a laptop (Sony Vaio FX502, AMD Mobile Duron
> 1GHz)
>
> With a PCMCIA-Modem (ELSA Microlink 56k MC Internet, a V.90 Modem) I
> have only 1/3 throughput with 2.6.0-test4 compared to a 2.4.X kernel
> (which gives the normal throughput for 56k modem).
>
> Yenta: CardBus bridge found at 0000:00:0a.0 [104d:80f6]
> Yenta IRQ list 0008, PCI irq11
> Socket status: 30000010
> Yenta: CardBus bridge found at 0000:00:0a.1 [104d:80f6]
> Yenta IRQ list 0008, PCI irq10
> Socket status: 30000006
> cs: memory probe 0x0c0000-0x0fffff: excluding 0xc0000-0xd3fff 0xdc000-0xdffff 0xe4000-0xfffff
> serial_cs: RequestIRQ: Unsupported mode

Make sure you have CONFIG_ISA=y in your kernel configuration.
However, I'm thinking that maybe if CONFIG_ISA is not set, we
shouldn't probe ISA IRQs in yenta_socket...

> pulling the card out I get
>
> Trying to free nonexistent resource <000002f8-000002ff>

Proper fix for this is behind a large pile of other interdependent patches.

--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html