Thread (3 messages) 3 messages, 3 authors, 2004-05-01

Re: Re: [PATCH] - fix video/fbmem.c warning.

From: Geert Uytterhoeven <geert@linux-m68k.org>
Date: 2004-05-01 18:57:39

On Fri, 30 Apr 2004, Andrew Morton wrote:
"Luiz Fernando N. Capitulino" [off-list ref] wrote:
quoted
 The patch bellow fix this warning:

 drivers/video/fbmem.c: In function `fb_cursor':
 drivers/video/fbmem.c:922: warning: passing arg 1 of `copy_from_user' discards qualifiers from pointer target type

 It is only a cast (several other calls for copy_from_user() need that).

 drivers/video/fbmem.c |    2 +-
 1 files changed, 1 insertion(+), 1 deletion(-)


diff -X dontdiff -Nparu a/drivers/video/fbmem.c a~/drivers/video/fbmem.c
--- a/drivers/video/fbmem.c	2004-04-26 15:25:52.000000000 -0300
+++ a~/drivers/video/fbmem.c	2004-04-27 16:48:15.000000000 -0300
@@ -914,7 +914,7 @@ fb_cursor(struct fb_info *info, struct f
 			return -ENOMEM;
 		}

-		if (copy_from_user(cursor.image.data, sprite->image.data, size) ||
+		if (copy_from_user((void *) cursor.image.data, sprite->image.data, size) ||
 		    copy_from_user(cursor.mask, sprite->mask, size)) {
 			kfree(cursor.image.data);
 			kfree(cursor.mask);
Why is fb_image.data declared to be a pointer to `const' data?  This
function proves that this assertion is false.
Because the data is const, except for the initial creation of fb_image.

The patch below (untested, not even compile tested!) should fix the problem
without introducing casts.

BTW, shouldn't fb_cursor.mask be const too, and treated similarly?
--- linux-2.6.5/drivers/video/fbmem.c.orig	2004-04-28 15:49:59.000000000 +0200
+++ linux-2.6.5/drivers/video/fbmem.c	2004-05-01 20:56:18.000000000 +0200
@@ -897,26 +897,29 @@

 	if (cursor.set & FB_CUR_SETSHAPE) {
 		int size = ((cursor.image.width + 7) >> 3) * cursor.image.height;
+		char *data;
+
 		if ((cursor.image.height != info->cursor.image.height) ||
 		    (cursor.image.width != info->cursor.image.width))
 			cursor.set |= FB_CUR_SETSIZE;

-		cursor.image.data = kmalloc(size, GFP_KERNEL);
-		if (!cursor.image.data)
+		data = kmalloc(size, GFP_KERNEL);
+		if (!data)
 			return -ENOMEM;

 		cursor.mask = kmalloc(size, GFP_KERNEL);
 		if (!cursor.mask) {
-			kfree(cursor.image.data);
+			kfree(data);
 			return -ENOMEM;
 		}

-		if (copy_from_user(&cursor.image.data, sprite->image.data, size) ||
+		if (copy_from_user(&data, sprite->image.data, size) ||
 		    copy_from_user(cursor.mask, sprite->mask, size)) {
-			kfree(cursor.image.data);
+			kfree(data);
 			kfree(cursor.mask);
 			return -EFAULT;
 		}
+		cursor.image.data = data;
 	}
 	info->cursor.set = cursor.set;
 	info->cursor.rop = cursor.rop;

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. 
Take an Oracle 10g class now, and we'll give you the exam FREE. 
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help