Thanks Robert, I've added this to the tree as
Documentation/networking/fib_trie.txt
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
From: Eric Dumazet <[EMAIL PROTECTED]>
Date: Wed, 06 Jul 2005 00:34:14 +0200
> But please check the memset() is there, or else... panic for sure.
It got added by patch 2/3.
So if I had actually applied the "fixed" version of patch 1/3 which
contained the memset(), then patch 2/3 would not have a
Dave,
Below is an updated (by my colleagues) version of the notes I sent before.
it will help to understand the implementation and this even for us. :)
Cheers.
--ro
LC-trie implementation notes.
Node types
--
leaf
David S. Miller a écrit :
From: Eric Dumazet <[EMAIL PROTECTED]>
Date: Wed, 29 Jun 2005 10:30:23 +0200
[NET] hashed spinlocks in net/ipv4/route.c
- Locking abstraction
- Spinlocks moved out of rt hash table : Less memory (50%) used by rt hash
table. it's a win even on UP.
- Sizing of spinl
From: Thomas Graf <[EMAIL PROTECTED]>
Date: Wed, 6 Jul 2005 00:13:24 +0200
> Useful in combination with classful qdiscs to drop or
> temporary disable certain flows, e.g. one could block
> specific ds flows with dsmark.
>
> Unlike the noop qdisc it can be controlled by the user and
> statistic ac
From: "Qinghua(Kevin) Ye" <[EMAIL PROTECTED]>
Date: Tue, 5 Jul 2005 16:19:27 -0600
> Compared to NAPI, click polling will disable Nic interrupts during its
> operation, even there is no any packets in the rx buffer.
This destroys latency if you only recheck the RX buffer using
timer interrupts.
From: Eric Dumazet <[EMAIL PROTECTED]>
Date: Wed, 29 Jun 2005 10:30:23 +0200
> [NET] hashed spinlocks in net/ipv4/route.c
> - Locking abstraction
> - Spinlocks moved out of rt hash table : Less memory (50%) used by rt hash
> table. it's a win even on UP.
> - Sizing of spinlocks table depend
Applied, thanks Thomas.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Click is a modular software router developed by MIT. Since router is a
dedicated network traffice processor, polling can be introduced to reduce
the overhead of interrupt. Once click runs, it will continuously polling the
Nic interface, get packets and process it, send it out from selected
interfac
From: [EMAIL PROTECTED]
Date: Thu, 30 Jun 2005 23:59:40 +0200
> I suspect "#define __ARGS(x) ()" was deprecated before I was born.
Indeed :) Patch applied, thanks.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo inf
Useful in combination with classful qdiscs to drop or
temporary disable certain flows, e.g. one could block
specific ds flows with dsmark.
Unlike the noop qdisc it can be controlled by the user and
statistic accounting is done.
Signed-off-by: Thomas Graf <[EMAIL PROTECTED]>
Index: net-2.6/net/s
Also applied, thanks a lot.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
From: [EMAIL PROTECTED]
Date: Thu, 30 Jun 2005 23:59:39 +0200
> A trivial patch to improve the readability of dev_set_promiscuity()
> in net/core/dev.c. New code does exactly the same thing as original
> code.
Applied, thanks a lot.
-
To unsubscribe from this list: send the line "unsubscribe netd
From: Christoph Hellwig <[EMAIL PROTECTED]>
Date: Thu, 30 Jun 2005 22:12:51 +0200
> you were right and the sleeping locks in shaper were broken. Markus
> Kanet noticed this and also tested the patch below that switches
> locking to spinlocks.
>
> Signed-off-by: Christoph Hellwig <[EMAIL PROTECTED
From: Robert Olsson <[EMAIL PROTECTED]>
Date: Thu, 30 Jun 2005 11:23:12 +0200
> Below a patch to preallocate memory when doing resize of trie (inflate halve)
> If preallocations fails it just skips the resize of this tnode for this time.
>
> The oops we got when killing bgpd (with full routing) i
Cleanup the phy_lock deadlock because of relocking in the nway_reset path.
Also, don't need to do irqsave/restore for blink, just excluding bh is good
enough.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Index: sky2/drivers/net/skge.c
From: Eric Dumazet <[EMAIL PROTECTED]>
Date: Wed, 29 Jun 2005 10:50:28 +0200
>- rt hash table allocated using alloc_large_system_hash() function.
Applied.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
From: "Qinghua(Kevin) Ye" <[EMAIL PROTECTED]>
Date: Tue, 5 Jul 2005 15:38:57 -0600
> I am thinking about extending the tg3 driver to support Click Polling.
What is Click Polling, and why would we want it?
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message
From: Robert Olsson <[EMAIL PROTECTED]>
Date: Wed, 29 Jun 2005 19:19:08 +0200
> Thanks Patrick!
>
> It looks straight forward and I don't see any problem in my test setup.
> I'm using a script to load full Internet routing.
I've applied Patrick's patch, thanks a lot everyone.
-
To unsubscribe fr
* David S. Miller <[EMAIL PROTECTED]> 2005-07-05 14:35
> From: Thomas Graf <[EMAIL PROTECTED]>
> Date: Tue, 5 Jul 2005 23:28:10 +0200
>
> > Can we agree on NET_XMIT_CN? The dev_xmit callers handle it the same
> > way as NET_XMIT_SUCCESS but we still have a chance to check for this
> > behaviour wi
From: "Michael Chan" <[EMAIL PROTECTED]>
Date: Wed, 29 Jun 2005 14:48:27 -0700
> [TG3]: support for ethtool -C
>
> Add support for ethtool -C with verification of user parameters.
I've added this patch to my tree, thanks Michael.
-
To unsubscribe from this list: send the line "unsubscribe netdev
From: Herbert Xu <[EMAIL PROTECTED]>
Date: Thu, 30 Jun 2005 12:38:28 +1000
> This patch makes IPv6 rcv registration happen last during initialisation.
Also applied, thanks Herbert.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
Mo
From: Herbert Xu <[EMAIL PROTECTED]>
Date: Thu, 30 Jun 2005 12:37:42 +1000
> Actually you never applied my patch :)
That's right. I was hoping someone would fix the
Acenic driver, oh well.
> Here they are again. This one makes IPv4 ip_rcv registration happen
> last in af_inet.
Applied, thanks
From: Thomas Graf <[EMAIL PROTECTED]>
Date: Tue, 5 Jul 2005 23:28:10 +0200
> Can we agree on NET_XMIT_CN? The dev_xmit callers handle it the same
> way as NET_XMIT_SUCCESS but we still have a chance to check for this
> behaviour within the parent qdisc.
It's not congestion, the packet went in fin
Hi, all,
I am thinking about extending the tg3 driver to support Click Polling. Click
has its patch on linux kernel, and adds some callback functions and polling
variable to the net_device struct(I did this work on 2.6.11 kernel, and on
AMD opteron):
netdev->polling = 0; // Check if click p
* David S. Miller <[EMAIL PROTECTED]> 2005-07-05 14:19
> From: jamal <[EMAIL PROTECTED]>
> Date: Tue, 05 Jul 2005 11:13:04 -0400
>
> > On Tue, 2005-05-07 at 17:03 +0200, Patrick McHardy wrote:
> >
> > > Sounds reasonable. As you said, blackhole implies NET_XMIT_SUCCESS,
> > > so I propose to use
From: Thomas Graf <[EMAIL PROTECTED]>
Date: Mon, 04 Jul 2005 13:48:11 +0200
> Current behaviour is to not report an error if a rate
> estimator is created together with a qdisc and the
> configuration of the rate estimator is bogus. This leads
> to unexpected behaviour because the user is not noti
Applied.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
From: jamal <[EMAIL PROTECTED]>
Date: Tue, 05 Jul 2005 11:13:04 -0400
> On Tue, 2005-05-07 at 17:03 +0200, Patrick McHardy wrote:
>
> > Sounds reasonable. As you said, blackhole implies NET_XMIT_SUCCESS,
> > so I propose to use this as default.
> >
>
> Actually i am not sure it makes sense to h
From: Thomas Graf <[EMAIL PROTECTED]>
Date: Mon, 04 Jul 2005 13:48:10 +0200
> Adds qdisc_alloc() to share code between qdisc_create()
> and qdisc_create_dflt(). Hides the qdisc alignment behind
> macros and makes use of them.
Applied.
-
To unsubscribe from this list: send the line "unsubscribe ne
From: Thomas Graf <[EMAIL PROTECTED]>
Date: Mon, 04 Jul 2005 13:29:25 +0200
> Reduce local_df to a bit field and ip_summed to a 2 bits
> field thus saving 13 bits. Move bit fields, packet type,
> and protocol into the spare area between the priority
> and the destructor. Saves 4 bytes on both, 32b
Applied.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Emanuil:
Could you send me 'ethtool -S' and 'ethtool -k' output after the failure?
ganesh.
On 7/5/05, Emanuil Christov <[EMAIL PROTECTED]> wrote:
> Hi
> i have got troubles with Intel 82541 NIC with kernel 2.6.11.2, 2.6.12.2,
> 2.6.13-rc1-mm1,
> 2.4.27, 2.4.30, 2.4.31 both with acpi enabled or
Did anyone had any luck getting MSI-X working on Opteron platforms?
The newer 8132-based systems support MSI-X, but we could not get it working
with our card there.
It works fine on Xeon systems, I wonder if the implementation is
Xeon-centric...
-
To unsubscribe from this list: send the line "unsu
One of the devices listed was a Yukon2 based board, so the skge driver wouldn't
work for that type.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Index: sky2/drivers/net/skge.c
===
--- sky2.orig/drivers/net/skge.c
+++ sky2/dr
Here is a fix for a typo, thanks Eliot. Since transmitter not active when
device is
down, it wasn't really noticed.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Index: sky2/drivers/net/skge.c
===
--- sky2.orig/drivers/net/sk
On Tue, 2005-05-07 at 17:03 +0200, Patrick McHardy wrote:
> Sounds reasonable. As you said, blackhole implies NET_XMIT_SUCCESS,
> so I propose to use this as default.
>
Actually i am not sure it makes sense to have anything but
NET_XMIT_SUCCESS. Its a damn blackhole ;-> What else could it be doi
Thomas Graf wrote:
* Patrick McHardy <[EMAIL PROTECTED]> 2005-07-05 16:06
It has side-effects on locally originated TCP connections. You should
probably return NET_XMIT_SUCCESS.
I don't see the notification as a negative side-effect, it is intented.
It sure depends on the actual use of this b
Hi
i have got troubles with Intel 82541 NIC with kernel 2.6.11.2, 2.6.12.2,
2.6.13-rc1-mm1,
2.4.27, 2.4.30, 2.4.31 both with acpi enabled or disabled
when running rsync with 400-450mbit/s i've got an errors
and server become unaccesible from network.
lspci -v -s 07:04.0
:07:04.0 Ethernet co
* Patrick McHardy <[EMAIL PROTECTED]> 2005-07-05 16:06
> Thomas Graf wrote:
> >* jamal <[EMAIL PROTECTED]> 2005-07-05 09:04
> >
> >>BTW, enqueue returning the code by calling the qdisc::drop() - is that
> >>gonna cause some issues?
> >
> >It returns NET_XMIT_DROP which is not really black hole beha
Thomas Graf wrote:
* jamal <[EMAIL PROTECTED]> 2005-07-05 09:04
BTW, enqueue returning the code by calling the qdisc::drop() - is that
gonna cause some issues?
It returns NET_XMIT_DROP which is not really black hole behaviour
but should never produce side effects.
It has side-effects on loc
* jamal <[EMAIL PROTECTED]> 2005-07-05 09:04
> On Tue, 2005-05-07 at 14:30 +0200, Thomas Graf wrote:
>
> >
> > It is really argueable, my point is that it adds quite some flexibility
> > at no cost and I don't want to miss it anymore. ;-> Having a way to
> > temporary or permantely "disable" a le
On Tue, 2005-05-07 at 14:30 +0200, Thomas Graf wrote:
>
> It is really argueable, my point is that it adds quite some flexibility
> at no cost and I don't want to miss it anymore. ;-> Having a way to
> temporary or permantely "disable" a leaf qdisc seems right to me.
Ok, but not for the example
* jamal <[EMAIL PROTECTED]> 2005-07-05 08:09
> On Mon, 2005-04-07 at 15:49 +0200, Thomas Graf wrote:
> > Assume you have a classful setup with per client classes and you
> > want to disable one such class for a moment (maybe the customer
> > didn't pay the bills ;->). A simple qdisc replace to a bl
On Mon, 2005-04-07 at 15:49 +0200, Thomas Graf wrote:
> * jamal <[EMAIL PROTECTED]> 2005-07-04 09:16
> >
> > Why not just "fix" noop qdisc to have those features? Note attaching a
> > filter/drop action should give you the same functionality.
>
> I like noop as-is, i.e. one copy shared by all use
45 matches
Mail list logo
