Re: [patch] hid: picolcd: testing the wrong variable

Thanks for spotting!

But while fixing this one also fix the ordering (which was never
reached due to this wrong check) a few lines later.

Patch (combined) at the end.

Thanks,
Bruno


On Fri, 6 Aug 2010 08:28:49 Dan Carpenter [off-list ref] wrote:

quoted

"ref_cnt" is a point to the reference count and it's non-null.  We really
want to test the reference count itself.

Signed-off-by: Dan Carpenter <redacted>

diff --git a/drivers/hid/hid-picolcd.c b/drivers/hid/hid-picolcd.c
index 346f0e3..c0bdeba 100644
--- a/drivers/hid/hid-picolcd.c
+++ b/drivers/hid/hid-picolcd.c

@@ -547,7 +547,7 @@ static void picolcd_fb_destroy(struct fb_info *info)
 	ref_cnt--;
 	mutex_lock(&info->lock);
 	(*ref_cnt)--;
-	may_release = !ref_cnt;
+	may_release = !*ref_cnt;
 	mutex_unlock(&info->lock);
 	if (may_release) {
 		framebuffer_release(info);

--- a/drivers/hid/hid-picolcd.c
+++ b/drivers/hid/hid-picolcd.c

@@ -547,11 +547,11 @@ static void picolcd_fb_destroy(struct fb
 	ref_cnt--;
 	mutex_lock(&info->lock);
 	(*ref_cnt)--;
-	may_release = !ref_cnt;
+	may_release = !*ref_cnt;
 	mutex_unlock(&info->lock);
 	if (may_release) {
-		framebuffer_release(info);
 		vfree((u8 *)info->fix.smem_start);
+		framebuffer_release(info);
 	}
 }