(please cc me, I follow the list on a web page)
The printk()s which enumerate media types during boot will oops
for some eeproms. They use the media type as an index into an
array of strings without doing any bounds checking.
This patch fixes it. The leaf->type is the killer for my card, I
have a descriptor where that is 128, the block_name table only has 6
entries. (The card is a Lite-On Communications Inc LNE100TX, based
on a Kingston card. Freebie from SWBell on DSL install.)
The "leaf->media & 0x0f" range check is sufficient, but not as nice
as the explicit size check on leaf->type. There are 16 entires in
medianame, but that is not known to this file during compilation.
(leaf->media could be as high as 0x3f, so some sort of check is called for.
If anyone named linus asks me to make a patch that uses the real size
of medianame I will. :-)
--- linux/drivers/net/tulip/eeprom.c.orig Wed Nov 1 14:32:18 2000
+++ linux/drivers/net/tulip/eeprom.c Wed Nov 1 14:14:02 2000
@@ -236,8 +236,9 @@
}
printk(KERN_INFO "%s: Index #%d - Media %s (#%d) described "
"by a %s (%d) block.\n",
- dev->name, i, medianame[leaf->media], leaf->media,
- block_name[leaf->type], leaf->type);
+ dev->name, i, medianame[leaf->media & 0x0f], leaf->media,
+ ((leaf->type < sizeof(block_name)/sizeof(block_name[0])) ? block_name[leaf->type] : "unknown"),
+ leaf->type);
}
if (new_advertise)
tp->to_advertise = new_advertise;
--
Jim Studt, President
The Federated Software Group, Inc.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue Nov 07 2000 - 21:00:10 EST