Re: Bug#364637: any other way around pmac_zilog problem? WAS: Bug#364637: I see, change in 8250.c between 2.6.15 and 2.6.16
On Tue, Aug 01, 2006 at 10:46:17AM +0100, Mark Hymers wrote:
> On Thu, 22, Jun, 2006 at 06:11:35PM +0200, Sven Luther spoke thus..
> > If 2.6.17 doesn't fix this, then i will have a look.
>
> It still doesn't work on 2.6.17 for me. I've tried this patch and it
> works on the Xserve G5s at work but it needs review as I'm not a kernel
> coder and I'm worried it might break the serial ports on other machines.
> This was tested building against the debian kernel source 2.6.17-5.
>
> Cheers,
>
> Mark
>
> --
> Mark Hymers <mark at hymers dot org dot uk>
>
> "I told you I was ill"
> The epitaph of Spike Milligan (1918-2002)
> #! /bin/sh -e
> ##
> ## All lines beginning with `## DP:' are a description of the patch.
> ## DP: Description: Disables legacy serial driver on powermacs.
> ## DP: Patch author: Mark Hymers <mark@hymers.org.uk>
> ## DP: Patch author: adapted from Sven's earlier version
> ## DP: Patch author: Sven Luther <luther@debian.org>
> ## DP: Patch author: adapted from the SuSE kernel tree.
> ## DP: Upstream status: workaround hack waiting for a clean legacy device solution.
>
> diff -aurN a/drivers/serial/8250.c b/drivers/serial/8250.c
> --- a/drivers/serial/8250.c 2006-06-18 02:49:35.000000000 +0100
> +++ b/drivers/serial/8250.c 2006-08-01 00:31:40.000000000 +0100
> @@ -2307,9 +2312,14 @@
>
> static int __init serial8250_console_init(void)
> {
> +#ifdef CONFIG_PPC_PMAC
> + printk("%s: nothing to do on PowerMac\n",__FUNCTION__);
> + return -ENODEV;
> +#else
> serial8250_isa_init_ports();
> register_console(&serial8250_console);
> return 0;
> +#endif
I don't believe this will work, since we have hardware who has a pc-like 8250
serial interface, and is powerpc, and will run the same kernel that will also
work on powermacs. This is why a static built-time check like you do is not
valid, and why the previous patch used a dynamic checked subarch variable.
There must be some workaround for that one, but i also believe that people
like benh have been working on a better solution to this. I lost sight of the
current state of this though.
Friendly,
Sven Luther
Reply to: