| su17171 | Dear Friend
Dear Friend
I could not get connected to you on phone so I decided to send this
email to you.Presently, I am out of the country.Since, I could not
reach you before I travelled I lodged your money including the
interest with the correspondent Bank Cotonou-Benin, in respect of the
little assistance you rendered to me during the cost of transfering my
sum.
The Account Officer Name,Dr. Ken Ibru is informed on this contact him
to receive your fund. His email is:
( ken55@sify.com )
Ple...
| Apr 5, 4:37 pm 2008 |
| Anthony Liguori |
| Apr 5, 1:26 pm 2008 |
| Anthony Liguori |
| Apr 5, 1:18 pm 2008 |
| Anthony Liguori | Re: [PATCH RFC 1/5] vringfd syscall
I have a tough time seeing what you're demonstrating here. Perhaps some
Should select VIRTIO && VIRTIO_RING
It seems unlikely that a caller could place both in_iov/out_iov on the
stack since to do it safely, you would have to use vring.num which is
determined by userspace. Since you have to kmalloc() the buffers
anyway, why not just allocate a single data structure within this
Oh, this is when it's safe to use. You don't seem to be acquiring
current->mm->mmap_sem ...
| Apr 5, 1:13 pm 2008 |
| Dan Nicholson | [PATCH] sundance: Set carrier status on link change events
Check if the link is available when a changed interrupt has been
received and set the carrier status appropriately. The code is copied
nearly verbatim from the dl2k module. The link status could be used in
more places in the driver, but this is enough to get the carrier status
reported to userspace. Fixes kernel bug #7487:
http://bugzilla.kernel.org/show_bug.cgi?id=7487
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
---
drivers/net/sundance.c | 94 +++++++++++++++++++++++++++++++...
| Apr 5, 10:21 am 2008 |
| Rusty Russell | [PATCH RFC 1/5] vringfd syscall
For virtualization, we've developed virtio_ring for efficient communication.
This would also work well for userspace-kernel communication, particularly
for things like the tun device. By using the same ABI, we can join guests
to the host kernel trivially.
These patches are fairly alpha; I've seen some network stalls I have to
track down and there are some fixmes.
Comments welcome!
Rusty.
diff -r 99132ad16999 Documentation/test_vring.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/Docum...
| Apr 5, 8:02 am 2008 |
| Rusty Russell | [PATCH RFC 2/5] vringfd base/offset
It turns out the lguest (and possibly kvm) want the addresses in the
ring buffer to only cover a certain part of memory, and be offset.
It makes sense that this be an ioctl.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
diff -r 08fb00b8acab Documentation/ioctl-number.txt
--- a/Documentation/ioctl-number.txt Sat Apr 05 21:31:40 2008 +1100
+++ b/Documentation/ioctl-number.txt Sat Apr 05 22:00:10 2008 +1100
@@ -183,6 +183,7 @@ 0xAC 00-1F linux/raw.h
0xAC 00-1F linux/raw.h
0xAD 00 ...
| Apr 5, 8:04 am 2008 |
| Avi Kivity |
| Apr 5, 8:44 am 2008 |
| Rusty Russell | [PATCH RFC 3/5] tun: vringfd receive support.
This patch modifies tun to allow a vringfd to specify the receive
buffer. Because we can't copy to userspace in bh context, we queue
like normal then use the "pull" hook to actually do the copy.
More thought needs to be put into the possible races with ring
registration and a simultaneous close, for example (see FIXME).
We use struct virtio_net_hdr prepended to packets in the ring to allow
userspace to receive GSO packets in future (at the moment, the tun
driver doesn't tell the stack it can han...
| Apr 5, 8:05 am 2008 |
| Rusty Russell | [PATCH RFC 4/5] tun: vringfd xmit support.
This patch modifies tun to allow a vringfd to specify the send
buffer. The user does a write to push out packets from the buffer.
Again, more thought needs to be put into the possible races with ring
registration.
Again we use the 'struct virtio_net_hdr' to allow userspace to send
GSO packets. In this case, it can hint how much to copy, and the
other pages will be made into skb fragments.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
diff -r 8270b5fdf03f drivers/net/tun.c
---...
| Apr 5, 8:06 am 2008 |
| Rusty Russell | [PATCH RFC 5/5] lguest support
This is how lguest uses the vringfd tun support. It needs more cleanup,
but it seems to basically work.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
diff -r 6979348a6ece Documentation/lguest/lguest.c
--- a/Documentation/lguest/lguest.c Sat Apr 05 22:02:28 2008 +1100
+++ b/Documentation/lguest/lguest.c Sat Apr 05 22:12:25 2008 +1100
@@ -43,6 +43,7 @@
#include "linux/virtio_console.h"
#include "linux/virtio_rng.h"
#include "linux/virtio_ring.h"
+#include "linux/vring.h"
#incl...
| Apr 5, 8:09 am 2008 |
| Rusty Russell | [PATCH] net: add destructor for skb data.
If we want to notify something when an skb is truly finished (such as
for tun vringfd support), we need a destructor on the data. We don't
need to add other fields, since we can just allocate extra room at the
end.
(I wonder if we could *reduce* the shinfo allocation where no frags are needed?)
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
---
include/linux/skbuff.h | 29 ++++++++++++++++++++++++++---
net/core/skbuff.c | 12 +++++++++---
2 files changed, 35 insertions(+)...
| Apr 5, 7:56 am 2008 |
| Evgeniy Polyakov |
| Apr 5, 9:06 am 2008 |
| Rusty Russell | [PATCH 1/2] net: make struct tun_struct private to tun.c
There's no reason for this to be in the header, and it just hurts
recompile time.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
diff -r f657af5295ee drivers/net/tun.c
--- a/drivers/net/tun.c Tue Apr 01 21:19:09 2008 +1000
+++ b/drivers/net/tun.c Fri Apr 04 16:03:31 2008 +1100
@@ -67,9 +67,42 @@
#include <asm/system.h>
#include <asm/uaccess.h>
+/* Uncomment to enable debugging */
+/* #define TUN_DEBUG 1 */
+
#ifdef TUN_DEBUG
static int debug;
+
+#define DBG ...
| Apr 5, 7:53 am 2008 |
| Rusty Russell | [PATCH 2/2] net: check for underlength tap writes
If the user gives a packet under 14 bytes, we'll end up reading off the end
of the skb (not oopsing, just reading off the end).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
diff -r 99132ad16999 drivers/net/tun.c
--- a/drivers/net/tun.c Sat Apr 05 21:20:32 2008 +1100
+++ b/drivers/net/tun.c Sat Apr 05 22:47:20 2008 +1100
@@ -286,8 +286,11 @@ static __inline__ ssize_t tun_get_user(s
return -EFAULT;
}
- if ((tun->flags & TUN_TYPE_MASK) == TUN_TAP_DEV)
+ if ((tun->...
| Apr 5, 7:54 am 2008 |
| Eric Dumazet |
| Apr 5, 8:38 am 2008 |
| Simon Horman |
| Apr 4, 9:07 pm 2008 |
| Julius Volz |
| Apr 5, 6:17 am 2008 |
| Nick Piggin |
| Apr 4, 9:06 pm 2008 |
| Sangtae Ha | Re: A Linux TCP SACK Question
Can you run the attached script and run your testing again?
I think it might be the problem of your dual cores balance the
interrupts on your testing NIC.
As we do a lot of things with SACK, cache misses and etc. might affect
your performance.
In default setting, I disabled tcp segment offload and did a smp
affinity setting to CPU 0.
Please change "INF" to your interface name and let us know the results.
Sangtae
| Apr 5, 5:17 pm 2008 |
| Ilpo Järvinen |
| Apr 5, 1:42 pm 2008 |
| Soeren Sonnenburg | Re: 2.6.25-rc8: WARNING: at net/ipv4/tcp_input.c:2173 tcp_ma...
OK I am getting this now as the first spot:
P: 4 L: 2 vs 2 S: 0 vs 3 F: 0 vs 0 w: 4023500226-4023505874 (0)
skb 0 f495c180
skb 1 f480a180
skb 2 f4994600
head 3 f495c780
skb 4 f5b32480
TCP wq(s) LL <
WARNING: at net/ipv4/tcp_ipv4.c:240 tcp_verify_wq+0x319/0x3c0()
another one:
WARNING: at net/ipv4/tcp_output.c:1475 __tcp_push_pending_frames+0x70/0x830()
P: 4 L: 2 vs 2 S: 0 vs 3 F: 0 vs 0 w: 4023500226-4023505874 (0)
skb 0 f495c180
skb 1 f480a180
skb 2 f4994600
head 3 f495c780
skb...
| Apr 5, 12:40 pm 2008 |
| Ilpo Järvinen |
| Apr 5, 1:52 pm 2008 |
| Soeren Sonnenburg |
| Apr 5, 2:14 pm 2008 |
| Ilpo Järvinen |
| Apr 5, 2:41 pm 2008 |
| Rami Rosen | Re: [GIT PULL net-2.6.26] [IPV6] MROUTE: Support multicast r...
Hi,
It seems to me that there is an error in this patch, when building on
64 machines.
It seems to me that this error will not occur when building on 32 machines,
though I did not verified this. It seems to me that the reason is that
the tail member of sk_buff is declared as unsigned int when building
on 64 bit machines, and is declared as unsigned char* when building
on 32 bit machines.
More details:
I had git-cloned this tree and change to the mroute-20080403 git branch
and tied to build w...
| Apr 5, 6:48 am 2008 |
| YOSHIFUJI Hideaki / | Re: [GIT PULL net-2.6.26] [IPV6] MROUTE: Support multicast r...
Hello.
Good point, thanks.
I've pushed the following fix into the tree:
git://git.linux-ipv6.org/gitroot/yoshfuji/linux-2.6-dev.git net-2.6.26-mroute-20080405
on top of current net-2.6.26 tree.
Regards,
--yoshfuji
---------------
Subject: [IPV6] MROUTE: Use skb_tail_pointer(skb) instead of skb->tail.
This bug resulted in compilation error on 64bit machines.
Pointed out by Rami Rosen <roszenrami@gmail.com>.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
-...
| Apr 5, 9:46 am 2008 |
| Georgi Chorbadzhiyski |
| Apr 5, 7:09 am 2008 |
| Grant Grundler |
| Apr 5, 2:19 pm 2008 |
