Re: 2.1.119 FourPort serial interrupt probing broken

Etienne Lorrain (lorrain@fb.sony.de)
Fri, 4 Sep 1998 13:58:55 +0001


Harald Koenig <koenig@tat.physik.uni-tuebingen.de> wrote:
> using 2.1.119 (haven't tried other 2.1.x versions) interrupt probing
> for AST compatible FourPort card is broken (using serial.o module,
> don't know if this is relevant). [...]
> used kernel config options are:
> CONFIG_SERIAL=m
> CONFIG_SERIAL_EXTENDED=y
> CONFIG_SERIAL_MANY_PORTS=y
> CONFIG_SERIAL_SHARE_IRQ=y
> CONFIG_SERIAL_DETECT_IRQ=y
> CONFIG_SERIAL_MULTIPORT=y

Hi,

You should not compile serial.c into a module *and* enable
DETECT_IRQ because then, you get what you asked for: each
time the module is inserted, it will try to detect the IRQ
of each initialised port - not safe.
Each time the module is inserted, you will also probe each
possible serial address to check if a port is there.

> no matter which oder I probe ports 4-7, cua4 and cua5 always
> get irq=0 instead of irq=5:
>
> cua0 at 0x03f8 (irq = 4) is a 16450
> cua1 at 0x02f8 (irq = 3) is a 16450
> cua2 at 0x03e8 (irq = 4) is a 16550A
> cua3 at 0x02e8 (irq = 3) is a 16550A
> ==> cua4 at 0x01a0 (irq = 0) is a 16550A (Fourport)
> ==> cua5 at 0x01a8 (irq = 0) is a 16550A (Fourport)
> cua6 at 0x01b0 (irq = 5) is a 16550A (Fourport)
> cua7 at 0x01b8 (irq = 5) is a 16550A (Fourport)
>
> (even if cua6 is probed first which will get irq=5 etc.).
> what's wrong here? I'm using setserial version 2.12 (SuSE 5.3).

Do not know, and it's time for me to go on holliday.
I do not think DETECT_IRQ is activated for more than
the first four serial ports.

Etienne.

--
-- The world belong to its organizer.

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