size to avoid MSG_TRUNC)
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c |3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 81dc1bb..f1ffb34 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink
On Wed, Sep 02, 2015 at 05:56:36PM +0200, Daniel Borkmann wrote:
> you suggest or not), for two reasons: I think (will start experimenting
> more with it tomorrow), you would get an out of bounds access here in
> case the skb->data is the last slot in the ring buffer and reaches
> exactly to the ri
Thank you for the reply.
On Wed, Sep 02, 2015 at 11:47:26AM +0200, Daniel Borkmann wrote:
> On 09/02/2015 02:04 AM, Ken-ichirou MATSUZAWA wrote:
> >Talking about skb_copy path, original skb's shared info is accessed
> >only in copy_skb_header, to get gso related field. As a
On Wed, Aug 19, 2015 at 04:29:32PM +0200, Daniel Borkmann wrote:
> On 08/17/2015 11:02 PM, David Miller wrote:
> >From: Daniel Borkmann
> >Date: Fri, 14 Aug 2015 12:38:21 +0200
> >
> >>diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
> >>index 67d2104..4307446 100644
> >>--- a/net/
urn immediately even though there are
VALID frames.
To avoid situations like above, I think we need to scan all frames
to find VALID frames at poll() like netlink_alloc_skb(),
netlink_forward_ring() finding an UNUSED frame at skb allocation.
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netl
On Tue, Aug 25, 2015 at 08:22:03PM -0700, David Miller wrote:
> From: Ken-ichirou MATSUZAWA
> > I don't know the intension of setting VALID status in the skb
> > destructor. But I think it need to be set UNUSED status in case of
>
> I think the idea is to have the user
the "ring->head != head" loop test fails.
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index a774985..39fa91f 100644
--- a/net/ne
Thank you for the reply.
On Tue, Aug 25, 2015 at 08:17:12PM -0700, David Miller wrote:
> So if netlink_forward_ring() _actually_ sees an entry that we should
> advance past, it will cycle through the whole ring, advancing ring->head
> until it equals the "ring->head != head" loop test fails.
>
I don't know the intension of setting VALID status in the skb
destructor. But I think it need to be set UNUSED status in case of
error then release skb, or rx ring might be filled with RESERVED
frames.
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c |2 +-
1 file ch
urn immediately even though there are
VALID frames.
To avoid situations like above, I think we need to scan all frames
to find a VALID frame at poll() like netlink_alloc_skb(),
netlink_forward_ring() finding an UNUSED frame at skb allocation.
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netl
single_range(WRITE, ...
iov_iter_init(1, WRITE, ...
call path, so that we need to check the type by iter_is_iovec()
to accept the WRITE.
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/netlink/
Hi,
Thank you for taking your time and trying to understand, even though
one of samples is wrong. correct one is:
rx only mmaped nflog sample:
https://gist.github.com/chamaken/dc0f80c14862e8061c06/raw/365c8a106840368f313a3791958da9be0f5fbed0/rxring-nflog.c
> >Currently, what happens is
Hi,
Thank you for taking your time.
Please let me explain these with code samples on gist.
I can not describe and arrange it well, sorry.
normal socket nflog sample:
https://gist.github.com/chamaken/dc0f80c14862e8061c06/raw/2d6da8fff31ef61af77e68713fdb1d71978746a6/nflog.c
set iptables
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c | 28
1 file changed, 16 insertions(+), 12 deletions(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index d8f5151..9f9e9ac 100644
--- a/net/netlink/af_netlink.c
+++ b/net
We need to accept msg_iter.type 1(WRITE) which is set in sendto/sendmsg.
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index c03fad0..d8f5151 100644
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 45c8502..c03fad0 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -864,7
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netfilter/nfnetlink_log.c|2 +-
net/netfilter/nfnetlink_queue_core.c |6 +++---
net/netlink/af_netlink.c | 26 +-
3 files changed, 17 insertions(+), 17 deletions(-)
diff --git a/net/netfilter
It seems that we need helper functions for skb which is allocated
at netlink_alloc_skb() since it does not have skb_shared_info.
Signed-off-by: Ken-ichirou MATSUZAWA
---
include/linux/netlink.h | 21 +-
net/netlink/af_netlink.c | 55
Hi,
It would be better if someone else is working on this.
Or could you help me out? It's a tough work for me.
Changes from RFC:
* remove netlink_skb_zerocopy()
I made a big mistake, and ``len <= skb_tailroom(to)'' for mmaped
skb is always true in nfqnl_build_packet_message()
* copy
We need to accept msg_iter.type 1(WRITE) which is set in sendto/sendmsg.
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index a6fba4c..7e1610e 100644
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c | 28
1 file changed, 16 insertions(+), 12 deletions(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 7e1610e..8901acd 100644
--- a/net/netlink/af_netlink.c
+++ b/net
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netlink/af_netlink.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 5632ad0..a6fba4c 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -847,7
Signed-off-by: Ken-ichirou MATSUZAWA
---
net/netfilter/nfnetlink_log.c|2 +-
net/netfilter/nfnetlink_queue_core.c |8
net/netlink/af_netlink.c | 26 +-
3 files changed, 18 insertions(+), 18 deletions(-)
diff --git a/net/netfilter
It seems that we need helper functions for skb which is allocated
at netlink_alloc_skb() since it does not have skb_shared_info.
Signed-off-by: Ken-ichirou MATSUZAWA
---
include/linux/netlink.h | 22 -
net/netlink/af_netlink.c | 81
Hello,
I got a kernel panic below when I dumped using mmaped netlink socket
while monitoring it by nlmon tap device. I realized it is because
mmaped netlink skb does not have skb_shared_info but don't know how
to fix it in sane. This patch series seems to work fine for me but
I'm not sure it's ri
Hello, I'm sorry I make a fuss.
Only FYI, one of my box has 3Com NIC
00:0c.0 Ethernet controller: 3Com Corporation 3c905C-TX/TX-M [Tornado] (rev
78)
Subsystem: 3Com Corporation 3C905C-TX Fast Etherlink for PC Management
NIC
Flags: bus master, medium devsel, latency 32, IRQ 11
Hello,
From: Ken-ichirou MATSUZAWA <[EMAIL PROTECTED]>
Subject: Oops in xfrm_bundle_ok
Date: Thu, 05 Jul 2007 02:47:10 +0900 (JST)
> I got Oops like below. I glanced xfrm_bundle_ok() in
> xfrm_policy.c and __xfrm4.bundle_create() in xfrm4_policy.c.
It seems this was fixed by bel
Hello,
From: Ken-ichirou MATSUZAWA <[EMAIL PROTECTED]>
Subject: checksum of lsrr
Date: Wed, 04 Jul 2007 22:28:24 +0900 (JST)
> if we change destination address, we need recompute not only ip
> checksum but also tcp's, don't we?
I was ashamed of myself for much confus
Hello,
I got Oops like below. I glanced xfrm_bundle_ok() in
xfrm_policy.c and __xfrm4.bundle_create() in xfrm4_policy.c.
In __xfrm4.bundle_create(), xfrm_dst->next may be null but
in xfrm_bundle_ok(), later loop does not check null, only break
`if (last == first)'.
I tried to solve with only che
Hello,
I tried source routing with netcat just for my interest but it did
not work. tcpdump output follows:
00:07:40:53:bd:55 > 00:01:8e:b3:82:1c, ethertype IPv4 (0x0800),
length 86: (tos 0x0, ttl 64, id 50437, offset 0, flags [DF],
proto: TCP (6), length: 72, options (LSRR 172.27.129.2 172.2
Hello,
ip6ip6_lock is not unlocked in error path.
--- net/ipv6/ip6_tunnel.c.orig 2005-11-09 22:01:31.0 +0900
+++ net/ipv6/ip6_tunnel.c 2005-11-09 22:04:06.0 +0900
@@ -525,6 +525,7 @@
if ((t = ip6ip6_tnl_lookup(&ipv6h->saddr, &ipv6h->daddr)) != NULL) {
Hello,
It seems that because ipcomp_alloc_tfms() calls smp_processor_id(),
but a comment says not need locking. Is this right?
BUG: using smp_processor_id() in preemptible [0001] code: ipcomp/2801
caller is ipcomp_init_state+0x170/0x490
Call Trace:{debug_smp_processor_id+182}
{ipcomp
Hello,
I set up IPSec tunnel and iptables as follows:
iptables -t nat -A POSTROUTING -o eth0 -j MASQUERADE
spdadd 172.27.210.0/24 172.27.86.0/24 any -P in ipsec
esp/tunnel/172.27.1.3-172.27.1.10/require
ah/tunnel/172.27.1.3-172.27.1.10/require;
spdadd 172.27.86.
I am sorry for my sending many times with a trifle.
From: Ken-ichirou MATSUZAWA <[EMAIL PROTECTED]>
Subject: Re: ALIGN at crypt/cipher.c
Date: Thu, 14 Jul 2005 22:59:35 +0900 (JST)
> Is ``notl 32bit reg'' right? Is this a problem of gcc?
No, I think I can understand. align
Hello,
I can't understand asm so would you help to see things going well?
I changed crypto_init_cipher_ops() in crypt/cipher.c
# thanks to ``HOWTO find oops location''
ops->cit_ivsize = crypto_tfm_alg_blocksize(tfm);
align = crypto_tfm_alg_alignmask(tfm) + 1;
asm("#1");
addr = (
Hello, I was sorry for my less information.
From: Andi Kleen <[EMAIL PROTECTED]>
Subject: Re: ALIGN at crypt/cipher.c
Date: Mon, 11 Jul 2005 18:35:31 +0200
> > $ uname -a
> > Linux toptor 2.6.13-rc2-git1 #8 SMP Tue Jul 12 00:10:46 JST 2005 x86_64
> > GNU/Linux
> > I got oops at esp_output in ne
36 matches
Mail list logo
