From: James Morris [EMAIL PROTECTED]
Date: Thu, 27 Apr 2006 00:58:41 -0400 (EDT)
On Thu, 27 Apr 2006, Rusty Russell wrote:
netfilter (similarly raw sockets, bonding, divert). Or, we could delay
LOCAL_IN hook processing until we get to socket receive.
This an idea proposed for skfilter
From: Linas Vepstas [EMAIL PROTECTED]
If a PCI bus error/fault triggers a PCI bus reset, attempts to get the
ethernet packet count statistics from the hardware will fail, returning
garbage data upstream. This patch skips statistics data collection if the
PCI device is not on the bus.
This
From: [EMAIL PROTECTED] (Linas Vepstas)
Various PCI bus errors can be signaled by newer PCI controllers. This
patch adds the PCI error recovery callbacks to the intel ethernet e100
device driver. The patch has been tested, and appears to work well.
Signed-off-by: Linas Vepstas [EMAIL
From: Thibaut VARENE [EMAIL PROTECTED]
Fix a problem with Tulip 21142 HP branded PCI cards (PN#: B5509-66001),
which feature a NatSemi DP83840A PHY.
Without that patch, it is impossible to properly initialize the card's PHY,
and it's thus impossible to monitor/configure it.
It's a
From: Bjorn Helgaas [EMAIL PROTECTED]
Apparently the Intel PRO/100 device enables interrupts on reset. Unless
firmware explicitly disables PRO/100 interrupts, we can get a flood of
interrupts when a driver attaches to an unrelated device that happens to
share the PRO/100 IRQ.
This should
From: Andrew Morton [EMAIL PROTECTED]
In file included from drivers/net/smc911x.c:84:
drivers/net/smc911x.h:46:9: warning: SMC_USE_16BIT is not defined
drivers/net/smc911x.h:60:9: warning: SMC_USE_32BIT is not defined
drivers/net/smc911x.h:73:10: warning: SMC_USE_PXA_DMA is not defined
From: Linas Vepstas [EMAIL PROTECTED]
Various PCI bus errors can be signaled by newer PCI controllers. This
patch adds the PCI error recovery callbacks to the intel gigabit ethernet
e1000 device driver. The patch has been tested, and appears to work well.
[EMAIL PROTECTED]: minor cleanups]
From: Ingo Oeser [EMAIL PROTECTED]
general:
- endian annotation of the ring descriptors
nv_getlen():
- use htons() instead of __constant_htons()
to improvde readability and let the compiler constant fold it.
nv_rx_process():
- use a real for() loop in
From: Mark Brown [EMAIL PROTECTED]
Provide a module option which configures the natsemi driver to use the
external MII port on the chip but ignore any PHYs that may be attached to it.
The link state will be left as it was when the driver started and can be
configured via ethtool. Any PHYs that
[EMAIL PROTECTED] wrote:
+ if (startup) {
+ int timeout = 10; /* max 1 ms */
for (i = 0; i reset_length; i++)
[EMAIL PROTECTED] wrote:
From: Mark Brown [EMAIL PROTECTED]
Provide a module option which configures the natsemi driver to use the
external MII port on the chip but ignore any PHYs that may be attached to it.
The link state will be left as it was when the driver started and can be
configured
[EMAIL PROTECTED] wrote:
From: Ingo Oeser [EMAIL PROTECTED]
general:
- endian annotation of the ring descriptors
nv_getlen():
- use htons() instead of __constant_htons()
to improvde readability and let the compiler constant fold it.
nv_rx_process():
- use a
[EMAIL PROTECTED] wrote:
From: Bjorn Helgaas [EMAIL PROTECTED]
Apparently the Intel PRO/100 device enables interrupts on reset. Unless
firmware explicitly disables PRO/100 interrupts, we can get a flood of
interrupts when a driver attaches to an unrelated device that happens to
share the
On 4/27/06, Jeff Garzik [EMAIL PROTECTED] wrote:
[EMAIL PROTECTED] wrote:
+ if (startup) {
+ int timeout = 10; /* max 1 ms */
for (i = 0; i reset_length; i++)
On Wed, Apr 26, 2006 at 11:25:01PM -0700, David S. Miller ([EMAIL PROTECTED])
wrote:
We approached this from the understanding that an intelligent NIC
will be able to transition directly to userspace, which is a major
win. 0 copies to userspace would be sweet. I think we can still
Hi Dave:
I was looking through the xfrm input/output code in order to abstract
out the address family specific encapsulation/decapsulation code. During
that process I found this bug in the IP ID selection code in xfrm4_output.c.
At that point dst is still the xfrm_dst for the current SA which
This patch fixes hello messages sent when a node is a level 1 router. Slightly
contrary to the spec (maybe) VMS ignores hello messages that do not name
level2 routers that it also knows about.
So, here we simply name all the routers that the node knows about rather just
other level1 routers.
(I
On Wednesday 26 April 2006 21:39, John W. Linville wrote:
On Fri, Apr 21, 2006 at 10:52:10PM +0200, Jiri Benc wrote:
On Fri, 21 Apr 2006 22:52:08 +0200, Michael Buesch wrote:
Can you please send your hacky patch for the bcm43xx
to me, so I can come up with a clean one?
Sure, actually
Hi Herbert,
Is there any update of this issue?
Regards,
- Aubrey
On 4/24/06, Aubrey [EMAIL PROTECTED] wrote:
On 4/24/06, Herbert Xu [EMAIL PROTECTED] wrote:
On Mon, Apr 24, 2006 at 01:42:12PM +0800, Aubrey wrote:
dev-last_rx = jiffies;
skb-dev = dev;
skb-protocol =
On Thursday 27 April 2006 04:00, Jeff Garzik wrote:
[EMAIL PROTECTED] wrote:
From: Bjorn Helgaas [EMAIL PROTECTED]
Apparently the Intel PRO/100 device enables interrupts on reset. Unless
firmware explicitly disables PRO/100 interrupts, we can get a flood of
interrupts when a driver
but clearly I should be using netperf to get more accurate cpu numbers
and a more convincing aggregate table :-)
Well, I'll not stop you :)
It is a bit rough/messy as a writeup, but here is what I've seen wrt the
latency vs throughput tradeoffs:
Hello.
We eliminated rt6_dflt_lock (to protect default router pointer)
at 2.6.17-rc1, and introduced rt6_select() for general router selection.
The function is called in the context of rt6_lock read-lock held,
but this means, we have some race conditions when we do round-robin.
Am I correct?
If
Hi. On 2.6.15.7 and 2.6.16.11, I have seen panics under heavy NFS
write load on an x86_64 system with two onboard Broadcom gigabit NICs.
It's a Supermicro P8SCi motherboard with an EMT64 Intel CPU. The aoe
driver in use is the aoe6-26 driver from the Coraid website.
I haven't yet trimmed down
On Thu, 2006-04-27 at 12:52 -0400, Ed L. Cashin wrote:
-- [please bite here ] -
Kernel BUG at drivers/net/tg3.c:2917
invalid opcode: [1] SMP
CPU 0
Most likely caused by IO re-ordering. Try the test patch in this
discussion:
David S. Miller wrote:
From: John Heffner [EMAIL PROTECTED]
Given that RFC2681 is Experimental (and I'm not aware of any current
efforts in the IETF to push it to the standard track), IHMO it would not
be inappropriate to make this behavior controlled via sysctl.
I have to respectfully
Ethtool Maintainers,
WRT my RFC (http://www.spinics.net/lists/netdev/msg02806.html) regarding
enhancing
ethtool posted on 04/11/06, would such a patch be accepted by the
maintainers of
ethtool? I don't want to spend time on these changes just to have them
rejected.
I've not received any objections
David G??mez [EMAIL PROTECTED] :
[...]
Does anybody in this list know why the IP1000 driver is not
included in the kernel ?
Afaik the driver has never been submitted for inclusion.
At least not on netdev@vger.kernel.org (hint, hint).
[...]
The card in question is:
Sundance Technology Inc
From: Evgeniy Polyakov [EMAIL PROTECTED]
Date: Thu, 27 Apr 2006 15:51:26 +0400
There are some caveats here found while developing zero-copy sniffer
[1]. Project's goal was to remap skbs into userspace in real-time.
While absolute numbers (posted to netdev@) were really high, it is only
From: John Heffner [EMAIL PROTECTED]
Date: Thu, 27 Apr 2006 13:47:33 -0400
(Most OS's don't do 2861, and it is not a standard.)
Are you so sure? Doing cwnd timeout largely predates the congestion
window validation work, in fact by several years.
In RFC 2581, it mentions Van Jacobson's
On Tue, Apr 25, 2006 at 11:16:29PM -0700, Kok, Auke wrote:
Update skb with the real packet size.
Signed-off-by: Jesse Brandeburg [EMAIL PROTECTED]
Signed-off-by: Auke Kok [EMAIL PROTECTED]
Signed-off-by: John Ronciak [EMAIL PROTECTED]
queued to -stable.
thanks,
greg k-h
-
To
On Thu, Apr 27, 2006 at 09:07:36AM -0700, Rick Jones wrote:
There should be three basic measures there - one is the single-instance
request-response test. The idea is to see minimum latency. That test
likes to see the interrupt throttle rate made very high, or disabled
completely.
The
having looked now at both 2861 and the 99 paper it references I see lots
of may's mights and belief but nothing real world.
the CWV vs non CWV was done against a TCP that did indeed reset cwnd
after an RTT of idle, so it wasn't showing reset at idle versus no reset
at idle. just CWV's less
On Thu, Apr 27, 2006 at 08:45:24AM -0700, Michael Chan wrote:
On Thu, 2006-04-27 at 12:52 -0400, Ed L. Cashin wrote:
-- [please bite here ] -
Kernel BUG at drivers/net/tg3.c:2917
invalid opcode: [1] SMP
CPU 0
Most likely caused by IO re-ordering. Try the test patch in
[EMAIL PROTECTED] wrote:
From: Evgeniy Polyakov [EMAIL PROTECTED]
Date: Thu, 27 Apr 2006 15:51:26 +0400
There are some caveats here found while developing zero-copy sniffer
[1]. Project's goal was to remap skbs into userspace in real-time.
While absolute numbers (posted to netdev@) were
On Thu, Apr 27, 2006 at 08:57:33PM +0800, Aubrey wrote:
Is there any update of this issue?
Assuming that the CHECKSUM_UNNECESSARY line wasn't there, then the
problem is simply that your packet has the wrong UDP checksum.
So I suggest that you print the packet out and compare it with
the
From: Ivo van Doorn [EMAIL PROTECTED]
Add flag handlers to set the state and capabilities of
the driver.
Signed-off-by: Ivo van Doorn [EMAIL PROTECTED]
diff -uprN wireless-dev-rt2x00/drivers/net/wireless/d80211/rt2x00/rt2400pci.c
From: Ivo van Doorn [EMAIL PROTECTED]
Add the eeprom_multiread function and clean up the code
a bit by using it as well. ;)
Signed-off-by: Ivo van Doorn [EMAIL PROTECTED]
diff -uprN wireless-dev-rt2x00/drivers/net/wireless/d80211/rt2x00/rt2400pci.c
From: Ivo van Doorn [EMAIL PROTECTED]
Add the *_set_state functions which makes sure
the device is switching state to awake or sleep.
Fix bad behaviour in the suspend routine,
and disable the radio before suspending.
Signed-off-by: Ivo van Doorn [EMAIL PROTECTED]
diff -uprN
From: Ivo van Doorn [EMAIL PROTECTED]
Coding style fix for all rt2x00 drivers.
This change was requested on the netdev list some time ago,
but the patch send then didn't contain all requested changes.
Signed-off-by: Ivo van Doorn [EMAIL PROTECTED]
Available on server:
From: Ivo van Doorn [EMAIL PROTECTED]
Add linux/dma-mapping.h header to allow compilation
on some architectures. Instead of dma_*_coherent
use pci_*consistent functions.
Signed-off-by: Ivo van Doorn [EMAIL PROTECTED]
diff -uprN wireless-dev-rt2x00/drivers/net/wireless/d80211/rt2x00/rt2400pci.c
From: Ivo van Doorn [EMAIL PROTECTED]
Now that rt2x00_pci and rt2x00_usb structures in the various
headers are generic enough, add 2 header files rt2x00pci and rt2x00usb
and make them contain all common information of the PCI or USB modules.
Signed-off-by: Ivo van Doorn [EMAIL PROTECTED]
From: Ivo van Doorn [EMAIL PROTECTED]
Fix panics when the interrupt handlers are being
run while the ring is empty.
During the interrupt handling break the loop
correctly when an error has been detected,
more work is being done after the loop.
Signed-off-by: Ivo van Doorn [EMAIL PROTECTED]
diff
From: Ivo van Doorn [EMAIL PROTECTED]
Store rx_params seperately outside the ring structure.
This is more safer and required because of some changes
in the way the rings are stored in the structure in the following
patches.
Signed-off-by: Ivo van Doorn [EMAIL PROTECTED]
diff -uprN
On Fri, Apr 28, 2006 at 12:03:12AM +0200, Ivo van Doorn wrote:
From: Ivo van Doorn [EMAIL PROTECTED]
Fix various little/big endian conversions.
rt2500pci should use cpu_to_le32 and rt2500usb should not.
While you're at it can you add __be* annotations to the hardware
datastructures so the
On Fri, Apr 28, 2006 at 12:02:52AM +0200, Ivo van Doorn wrote:
From: Ivo van Doorn [EMAIL PROTECTED]
Instead of dma_*_coherent
use pci_*consistent functions.
No point in doing that, quite reverse as you use the gfp_mask argument
which is a (small) pessimation here.
-
To unsubscribe from
Francois Romieu wrote:
David Gómez [EMAIL PROTECTED] :
[...]
Does anybody in this list know why the IP1000 driver is not
included in the kernel ?
Afaik the driver has never been submitted for inclusion.
At least not on netdev@vger.kernel.org (hint, hint).
[...]
The card in question is:
Netperf2 TOT now accesses the buffer that was just recv()'d rather than
the one that is about to be recv()'d.
We've posted netperf2 results with I/OAT enabled/disabled and the data
access option on/off at
http://kernel.org/pub/linux/kernel/people/grover/ioat/netperf-icb-1.5-postscaling-both.pdf
Chris Leech wrote:
Netperf2 TOT now accesses the buffer that was just recv()'d rather than
the one that is about to be recv()'d.
We've posted netperf2 results with I/OAT enabled/disabled and the data
access option on/off at
With 2.6.17-rc3, my E1000 won't get a dhcp lease.
Looking at tcpdump and ifconfig output, it's easy to see why.
It's recieving packets, but the packets transmitted field
of ifconfig never increases.
The last version I have built that worked ok was 2.6.17rc2-git3
NIC is ..
03:0e.0 Ethernet
-stable review patch. If anyone has any objections, please let us know.
--
Update skb with the real packet size.
Signed-off-by: Jesse Brandeburg [EMAIL PROTECTED]
Signed-off-by: Auke Kok [EMAIL PROTECTED]
Signed-off-by: John Ronciak [EMAIL PROTECTED]
Signed-off-by: Greg
And if its not possible to change the probability, is there another method I
can use instead?
Hi,
I am using netem to add loss and then adding another qdisc within netem
according to the wiki. Then i want to change the netem drop probability
without having to delete the qdisc and
Dave Jones wrote:
With 2.6.17-rc3, my E1000 won't get a dhcp lease.
Looking at tcpdump and ifconfig output, it's easy to see why.
It's recieving packets, but the packets transmitted field
of ifconfig never increases.
The last version I have built that worked ok was 2.6.17rc2-git3
*puzzled*
52 matches
Mail list logo
