[PATCH][RFC][REPOST] __init_msg(x) and friends macro

[PATCH][RFC][REPOST] __init_msg(x) and friends macro

[pazke]

[-- Attachment #1.1: Type: text/plain, Size: 723 bytes --]

Hi all,

soryy for such ugly subject line, but I already sent this patch 
to LKML and didn't get any reply.

Patch adds __init_msg (and friends) macro that places its argument 
(string constant) into corresponding .data.init section and returns
pointer to it. The goal of this patch is to allow constructions like this:

        static void __init foo(void)
	{
		printk(__init_msg(KERN_INFO "Some random long message "
				  "going to .data.init and then to bit bucket\n"));
	}

I hope this patch can save some memory and will be usefull :))

Best regards.

-- 
Andrey Panin            | Embedded systems software engineer
pazke@orbita1.ru        | PGP key: http://www.orbita1.ru/~pazke/AndreyPanin.asc

[-- Attachment #1.2: patch-__init_msg --]
[-- Type: text/plain, Size: 590 bytes --]

diff -ur linux.vanilla/include/linux/init.h linux/include/linux/init.h
--- linux.vanilla/include/linux/init.h	Mon May 14 15:51:20 2001
+++ linux/include/linux/init.h	Mon May 14 15:54:05 2001
@@ -155,4 +155,9 @@
 #define __devexitdata __exitdata
 #endif
 
+#define __init_msg(x) ({ static char msg[] __initdata = (x); msg; })
+#define __exit_msg(x) ({ static char msg[] __exitdata = (x); msg; })
+#define __devinit_msg(x) ({ static char msg[] __devinitdata = (x); msg; })
+#define __devexit_msg(x) ({ static char msg[] __devexitdata = (x); msg; })
+
 #endif /* _LINUX_INIT_H */

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [PATCH][RFC][REPOST] __init_msg(x) and friends macro

[Martin Mares]

Hi!

> Patch adds __init_msg (and friends) macro that places its argument 
> (string constant) into corresponding .data.init section and returns
> pointer to it. The goal of this patch is to allow constructions like this:
> 
>         static void __init foo(void)
> 	{
> 		printk(__init_msg(KERN_INFO "Some random long message "
> 				  "going to .data.init and then to bit bucket\n"));
> 	}
> 
> I hope this patch can save some memory and will be usefull :))

I think it's better to teach GCC understanding another function attribute
defining the default section of all read-only data defined in the function.

				Have a nice fortnight
-- 
Martin `MJ' Mares <mj@ucw.cz> <mj@suse.cz> http://atrey.karlin.mff.cuni.cz/~mj/
Even nostalgia isn't what it used to be.