Since userspace can trivially trigger get_user_pages failure
by passing in an illegal virtual address/size pair, I suggest
removing the printk when this happends: I think kernel messages
should reflect kernel problems, not user level application bugs.
Signed-off-by: Michael S. Tsirkin <[EMAIL PROTECTED]>
Index: core/uverbs_mem.c
===================================================================
--- core/uverbs_mem.c (revision 1895)
+++ core/uverbs_mem.c (working copy)
@@ -69,11 +69,6 @@ int ib_umem_get(struct ib_device *dev, s
PAGE_SIZE / sizeof (struct page *)),
1, 0, page_list, NULL);
- if (ret < 0) {
- printk(KERN_ERR "get_user_pages: %d\n", ret);
- printk(KERN_ERR "failed at cur_base %lx\n", cur_base);
- }
-
if (ret < 0)
goto out;
--
MST - Michael S. Tsirkin
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
