Re: mac80211 truesize bugs

3 messages, 2 authors, 2008-05-04 · open the first message on its own page

Re: mac80211 truesize bugs

From: Herbert Xu <hidden>
Date: 2008-05-04 09:44:44

On Sun, May 04, 2008 at 11:14:49AM +0200, Johannes Berg wrote:
I just re-did the test, and I definitely need 29 more bytes on, for
example, the IPv6 autodiscovery packets and ICMP packets generated with
ping(8). Some of them even need additional tailroom.
Could you find out the code path that created the skb in the
kernel so we can see why there isn't enough head room?

Thanks,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} [off-list ref]
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: mac80211 truesize bugs

From: Johannes Berg <johannes@sipsolutions.net>
Date: 2008-05-04 09:52:48

On Sun, 2008-05-04 at 17:44 +0800, Herbert Xu wrote:
On Sun, May 04, 2008 at 11:14:49AM +0200, Johannes Berg wrote:
quoted
I just re-did the test, and I definitely need 29 more bytes on, for
example, the IPv6 autodiscovery packets and ICMP packets generated with
ping(8). Some of them even need additional tailroom.
Could you find out the code path that created the skb in the
kernel so we can see why there isn't enough head room?
Yeah I was just considering putting a stacktrace into the skb when it's
allocated. Haven't done, lunchtime now.

johannes

Re: mac80211 truesize bugs

From: Johannes Berg <johannes@sipsolutions.net>
Date: 2008-05-04 11:26:09

On Sun, 2008-05-04 at 11:52 +0200, Johannes Berg wrote:
On Sun, 2008-05-04 at 17:44 +0800, Herbert Xu wrote:
quoted
On Sun, May 04, 2008 at 11:14:49AM +0200, Johannes Berg wrote:
quoted
I just re-did the test, and I definitely need 29 more bytes on, for
example, the IPv6 autodiscovery packets and ICMP packets generated with
ping(8). Some of them even need additional tailroom.
Could you find out the code path that created the skb in the
kernel so we can see why there isn't enough head room?
Yeah I was just considering putting a stacktrace into the skb when it's
allocated. Haven't done, lunchtime now.
http://johannes.sipsolutions.net/patches/kernel/all/2008-05-04-11%3a23/027-skb-alloc-stackdump.patch

will report results after kernel rebuild.

johannes
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help