2003-02-27 07:17:39

by _N3X_

[permalink] [raw]
Subject: serial_cs with devfs

in serial_cs.c there is a hardcoded device name in the code, stating
ttySx, that's not the right thing with devfs..
here's a patch of my own (should work also on the latest prepatch),
just to show the problem, i think that the defines around it could be
more carefully chosen, but what's the deal.. have a nice day.
_N3X_ <[email protected]>

diff -ru linux-2.4.19/drivers/char/pcmcia/serial_cs.c linux-2.4.19-fix/drivers/char/pcmcia/serial_cs.c
--- linux-2.4.19/drivers/char/pcmcia/serial_cs.c 2001-12-21 18:41:54.000000000 +0100
+++ linux-2.4.19-fix/drivers/char/pcmcia/serial_cs.c 2002-11-06 10:14:07.000000000 +0100
@@ -256,7 +256,11 @@
}

info->line[info->ndev] = line;
+#ifdef CONFIG_DEVFS_FS
+ sprintf(info->node[info->ndev].dev_name, "tts/%d", line);
+#else
sprintf(info->node[info->ndev].dev_name, "ttyS%d", line);
+#endif /* CONFIG_DEVFS_FS */
info->node[info->ndev].major = TTY_MAJOR;
info->node[info->ndev].minor = 0x40+line;
if (info->ndev > 0)


2003-02-27 07:34:28

by Zwane Mwaikambo

[permalink] [raw]
Subject: Re: serial_cs with devfs

On Thu, 27 Feb 2003, _N3X_ wrote:

> in serial_cs.c there is a hardcoded device name in the code, stating
> ttySx, that's not the right thing with devfs..
> here's a patch of my own (should work also on the latest prepatch),
> just to show the problem, i think that the defines around it could be
> more carefully chosen, but what's the deal.. have a nice day.
> _N3X_ <[email protected]>
>
> diff -ru linux-2.4.19/drivers/char/pcmcia/serial_cs.c linux-2.4.19-fix/drivers/char/pcmcia/serial_cs.c
> --- linux-2.4.19/drivers/char/pcmcia/serial_cs.c 2001-12-21 18:41:54.000000000 +0100
> +++ linux-2.4.19-fix/drivers/char/pcmcia/serial_cs.c 2002-11-06 10:14:07.000000000 +0100
> @@ -256,7 +256,11 @@
> }
>
> info->line[info->ndev] = line;
> +#ifdef CONFIG_DEVFS_FS
> + sprintf(info->node[info->ndev].dev_name, "tts/%d", line);
> +#else
> sprintf(info->node[info->ndev].dev_name, "ttyS%d", line);
> +#endif /* CONFIG_DEVFS_FS */
> info->node[info->ndev].major = TTY_MAJOR;
> info->node[info->ndev].minor = 0x40+line;
> if (info->ndev > 0)

*cough* All that patch does is accentuate is one thing...

Zwane
--
function.linuxpower.ca