Jean Tourrilhes <[EMAIL PROTECTED]> writes:
> +#include <linux/types.h> /* for __u8, __u32 & ... */
> +
> #ifndef KERNEL_IRDA_H
> #define KERNEL_IRDA_H
>
> struct sockaddr_irda {
> sa_family_t sir_family; /* AF_IRDA */
> - u_int8_t sir_lsap_sel; /* LSAP selector */
> - u_int32_t sir_addr; /* Device address */
> + __u8 sir_lsap_sel; /* LSAP selector */
> + __u32 sir_addr; /* Device address */
> char sir_name[25]; /* Usually <service>:IrDA:TinyTP */
> };
Currently, user-space programs needs to use <linux/irda.h>. But I see no
reason why you should force them to use <linux/types.h> as well. This may
result in a mess for application programmers. My plan was that applications
could do:
#include <sys/types.h>
#include <linux/irda.h>
and the kernel could do:
#include <linux/types.h>
#include <linux/irda.h>
Why do you want to change this? I cannot apply this patch just because you
had to make these changes in order to make your applications compile. You
must tell me what you are thinking!!! Do we need two files, one irda.h for
glibc and one for the kernel? I was just trying to delay that decision for
a while!
-- Dag
--
/ Dag Brattli | The Linux-IrDA Project /
// University of Tromsoe, Norway | Infrared communication for Linux //
/// http://www.cs.uit.no/~dagb | http://www.cs.uit.no/linux-irda/ ///
_______________________________________________
Linux-IrDA mailing list - [EMAIL PROTECTED]
http://www4.pasta.cs.UiT.No/mailman/listinfo/linux-irda
