Re: posix message queues

[Manfred Spraul <manfred@colorfullife.com>]

Andrew Morton wrote:

>Manfred, if/when this gets merged up the long-suffering arch maintainers
>might appreciate a little test app to exercise the syscalls which must be
>added.  Do you have something suitable at hand?
>  
>
I have a few separate apps, I'll merge them into one app and post it to 
lkml. Probably tomorrow or Friday.

>
>Also, before we go too far the 64-bit guys may like to comment on the
>syscall interface.   For example, this:
>
>struct mq_attr {
>	long	mq_flags;	/* message queue flags			*/
>	long	mq_maxmsg;	/* maximum number of messages		*/
>	long	mq_msgsize;	/* maximum message size			*/
>	long	mq_curmsgs;	/* number of messages currently queued	*/
>	long	__reserved[4];	/* ignored for input, zeroed for output */
>};
>
>looks like it will require emulation for 32-bit apps.  But if we were to
>make these __u32 perhaps that could be avoided?
>  
>
No. The unix spec mandates 'long' for the final C-library api. Using u32 
would mean that glibc would have to add an emulation layer in user space 
for 64-bit archs. In the long run, 64-bit kernels running 64-bit apps 
will be the common case. I don't want to add an emulation layer to the 
common case.

--
    Manfred