Hi Jeff,
I'm submitting three more patches for inclusion in netdev#upstream.
These patches are built over the series I resent yesterday night.
The patch numbering reflects the stacking.
Here is a brief description:
- avoid false positives in the xgmac hang workaround
- Properly set the
From: Divy Le Ray [EMAIL PROTECTED]
Qualify toggling of xgmac tx enable with not getting pause frames,
we might not make forward progress because the peer is sending
lots of pause frames.
Signed-off-by: Divy Le Ray [EMAIL PROTECTED]
---
drivers/net/cxgb3/common.h |1 +
From: Divy Le Ray [EMAIL PROTECTED]
The cxgb3 driver is incorrectly configuring the HW CQ context for CQ's
that use overflow-avoidance. Namely the RDMA control CQ. This results
in a bad DMA from the device to bus address 0. The solution is to set
the CQ_ERR bit in the context for these types
From: Divy Le Ray [EMAIL PROTECTED]
Currently, the driver only tries up to 5 times (5us) to get the results
of a CQ context operation. Testing has shown the chip can take as much
as 50us to return the response on SG_CONTEXT_CMD operations. So we up
the retry count to 100 to cover high loads.
On Wed, Aug 22, 2007 at 11:35:20PM -0700, Divy Le Ray wrote:
Hi Jeff,
I'm submitting three more patches for inclusion in netdev#upstream.
These patches are built over the series I resent yesterday night.
The patch numbering reflects the stacking.
Here is a brief description:
- avoid
On Thursday 23 August 2007 00:20, Andrew Theurer wrote:
David Miller wrote:
From: Andrew Theurer [EMAIL PROTECTED]
Date: Wed, 22 Aug 2007 16:55:03 -0500
Thanks for finally getting to test this, I thought nobody
would test this until it got merged into 2.6.24 :-/
Yes, sorry for the
Hello,
I noticed that it is possible that the kernel allocates the same UDP
port to an application that was used and closed immediately before the
new application got it. This means that applications that do not specify
an exact port and rely on the kernel to allocate a port for them might
see
From: Jan-Bernd Themann [EMAIL PROTECTED]
Date: Thu, 23 Aug 2007 08:55:29 +0200
We'd like to keep the possibility to switch back to a single queue
for now.
Please do not do this, we already have way too much configurability
out there.
If you have the physical hardware queues enabled, use
Hi David,
On Thursday 23 August 2007 10:17, David Miller wrote:
From: Jan-Bernd Themann [EMAIL PROTECTED]
Date: Thu, 23 Aug 2007 08:55:29 +0200
We'd like to keep the possibility to switch back to a single queue
for now.
Please do not do this, we already have way too much
sysctl(2) is supported but frozen.
Signed-off-by: Alexey Dobriyan [EMAIL PROTECTED]
---
drivers/net/wireless/ath5k_base.c | 21 ++---
1 file changed, 6 insertions(+), 15 deletions(-)
--- a/drivers/net/wireless/ath5k_base.c
+++ b/drivers/net/wireless/ath5k_base.c
@@ -2438,21
Andrew Morton pointed out that my changelog was unusable. Sorry!
Here is a second try with the changelog and kernel version changed.
Regards,
Jarek P.
(take 2)
Subject: request_irq() - fix DEBUG_SHIRQ handling
Mariusz Kozlowski reported lockdep's warning:
Doing napi_disable twice hangs ifdown of the device. e1000_down is the
common place to call napi_disable.
Signed-off-by: Krishna Kumar [EMAIL PROTECTED]
---
e1000_main.c |4
1 files changed, 4 deletions(-)
diff -ruNp org/drivers/net/e1000/e1000_main.c new/drivers/net/e1000/e1000_main.c
After applying patch1, I started getting waiting for count messages when
doing ifdown. Not sure if this is the right fix since the count was already
showing as -1 in that message, but this patch fixes the problem.
Signed-off-by: Krishna Kumar [EMAIL PROTECTED]
---
e1000_main.c |3 ++-
1
On Sat, Aug 18, 2007 at 11:44:12AM +0200, Jiri Slaby wrote:
define global BIT macro
move all local BIT defines to the new globally define macro.
Signed-off-by: Jiri Slaby [EMAIL PROTECTED]
Acked-by: Ralf Baechle [EMAIL PROTECTED]
for the MACE ethernet and MIPS bits.
Ralf
-
To
According to Neil's comments, I have tried to correct the mistakes of my first
sending
Thank you for these comments Neil.
This is a small part of missing pieces of IPv6 support for the server.
It deals with the ip_map caching code part.
It changes the ip_map structure to be able to store INET6
The two different wireless code bases both define macros to ease
printing MAC addresses:
printk(KERN_INFO MAC address is MAC_FMT \n, MAC_ARG(addr));
This patch moves those macros to if_ether.h and uses them all over the
tree.
Signed-off-by: Johannes Berg [EMAIL PROTECTED]
---
David Miller wrote:
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Wed, 22 Aug 2007 10:46:15 -0700
This patch is on hold since the netlink changes haven't made it upstream yet.
I don't have the kernel side in my queue either, perhaps
I lost it or I didn't see it when it was sent out.
Benjamin Thery wrote:
From: [EMAIL PROTECTED]
Subject: net/core: Fix crash in dev_mc_sync()/dev_mc_unsync()
This patch fixes a crash that may occur when the routine dev_mc_sync()
deletes an address from the list it is currently going through. It
saves the pointer to the next element before
This incorporates the new napi_struct changes into e1000e. Included
bugfix for ifdown hang from Krishna Kumar for e1000.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000e/e1000.h |2 ++
drivers/net/e1000e/netdev.c | 35 ---
2 files changed,
Hi Aurelien,
Aurélien Charbon wrote:
According to Neil's comments, I have tried to correct the mistakes of my
first sending
I have some more comments.
@@ -1559,6 +1560,7 @@ exp_addclient(struct nfsctl_client *ncp)
{
struct auth_domain*dom;
inti, err;
+struct
Hi Aurélien-
Aurélien Charbon wrote:
According to Neil's comments, I have tried to correct the mistakes of my first
sending
Thank you for these comments Neil.
This is a small part of missing pieces of IPv6 support for the server.
It deals with the ip_map caching code part.
It changes the
On the day of Thursday 23 August 2007 Greg KH hast written:
On Wed, Aug 22, 2007 at 10:42:25PM +0200, Willy Tarreau wrote:
On Wed, Aug 22, 2007 at 08:15:03PM +0200, Prakash Punnoor wrote:
Hi,
even if Greg is waiting for some special invitation
(http://lkml.org/lkml/2007/8/14/229), I
On Wed, 2007-08-22 at 20:46 +0200, Johannes Berg wrote:
The two different wireless code bases both define macros to ease
printing MAC addresses:
There are also several different uses of the equivalent of
printk(%02x,addr[0])
for (i=1; i6; i++)
Krishna Kumar wrote:
Doing napi_disable twice hangs ifdown of the device. e1000_down is the
common place to call napi_disable.
Signed-off-by: Krishna Kumar [EMAIL PROTECTED]
---
e1000_main.c |4
1 files changed, 4 deletions(-)
diff -ruNp org/drivers/net/e1000/e1000_main.c
On Thu, 2007-08-23 at 09:01 -0700, Joe Perches wrote:
There are also several different uses of the equivalent of
printk(%02x,addr[0])
for (i=1; i6; i++)
printk(:%02x,addr[i]);
to print an ethernet MAC address.
Hm. I didn't know that, I can go through in a later
Hi David,
when trying to get our driver working with the new interface,
I found the following issue where I'm not sure how to solve it best:
netif_rx_reschedule() does not work when called after netif_rx_complete().
The problem is that netif_rx_reschedule currently adds the napi struct once
Add v4mapped address inline to avoid calls to ipv6_addr_type().
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 9059e0e..c2b6c11 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -418,6 +418,12 @@ static inline int ipv6_addr_diff(const struct in6_addr *a1, const struct
Subject: shaper: mark for removal
This driver has been marked obsolete for a long time and
is superseded by traffic schedulers.
Signed-off-by: Stephen Hemminger [EMAIL PROTECTED]
--- a/Documentation/feature-removal-schedule.txt2007-08-23
09:36:24.0 -0700
+++
Tóth László Attila wrote:
Hello,
I noticed that it is possible that the kernel allocates the same UDP
_Which_ kernel - or rather which rev? There are lots of linux kernels
potentially out there...
port to an application that was used and closed immediately before the
new application got
On Wed, 2007-08-22 at 20:05 -0700, David Miller wrote:
I would suggest, at this point, to make purpose built situation
specific interfaces that pass specific objects (the ones being
operated upon) to the audit layer.
Let the audit layer pick out the bits it actually wants in the
format it
On Thu, Aug 23, 2007 at 06:12:00PM +0200, Johannes Berg wrote:
On Thu, 2007-08-23 at 09:01 -0700, Joe Perches wrote:
There are also several different uses of the equivalent of
printk(%02x,addr[0])
for (i=1; i6; i++)
printk(:%02x,addr[i]);
to print an ethernet
Hello.
In article [EMAIL PROTECTED] (at Thu, 23 Aug 2007 12:40:54 -0400), Brian
Haley [EMAIL PROTECTED] says:
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 9059e0e..c2b6c11 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -418,6 +418,12 @@ static inline int
On Thu, Aug 23, 2007 at 05:50:41PM +0200, Prakash Punnoor wrote:
On the day of Thursday 23 August 2007 Greg KH hast written:
On Wed, Aug 22, 2007 at 10:42:25PM +0200, Willy Tarreau wrote:
On Wed, Aug 22, 2007 at 08:15:03PM +0200, Prakash Punnoor wrote:
Hi,
even if Greg is waiting
YOSHIFUJI Hideaki / wrote:
Please put this just after ipv6_addr_any(), not after
ipv6_addr_diff().
Ok, updated patch attached.
-Brian
Add v4mapped address inline to avoid calls to ipv6_addr_type().
Signed-off-by: Brian Haley [EMAIL PROTECTED]
diff --git a/include/net/ipv6.h
This patch causes UDP port allocation to be randomized like TCP.
The earlier code would always choose same port (ie first empty list).
Signed-off-by: Stephen Hemminger [EMAIL PROTECTED]
--- a/net/ipv4/udp.c2007-08-23 09:44:22.0 -0700
+++ b/net/ipv4/udp.c2007-08-23
On Thu, Aug 23, 2007 at 10:31:03AM +1000, Stephen Rothwell wrote:
On Wed, 22 Aug 2007 09:12:48 -0500 Olof Johansson [EMAIL PROTECTED] wrote:
-static unsigned int read_iob_reg(struct pasemi_mac *mac, unsigned int reg)
+static inline unsigned int read_iob_reg(struct pasemi_mac *mac, unsigned
Move away from using the pci config access functions for simple register
access. Our device has all of the registers in the config space (hey,
from the hardware point of view it looks reasonable :-), so we need to
somehow get to it. Newer firmwares have it in the device tree such that
we can just
Hi Greg,
On Thu, Aug 23, 2007 at 09:55:13AM -0700, Greg KH wrote:
It might help if someone sends a real patch that can be applied :)
This is getting really silly now :-) We're all wasting more time
wondering who will send the patch than posting it. I've lost, I got
fed up first, so here it is.
This patch doesn't seem to have gone in yet
Steve.
David Miller wrote:
From: Divy Le Ray [EMAIL PROTECTED]
Date: Mon, 13 Aug 2007 12:33:04 -0700
From: Divy Le Ray [EMAIL PROTECTED]
cxgb3 used netdev_priv() and dev-priv for different purposes.
In 2.6.23, netdev_priv() == dev-priv, cxgb3
From: Joy Latten [EMAIL PROTECTED]
Date: Thu, 23 Aug 2007 12:15:10 -0500
For example, when auditing the addition of a policy, either
xfrm_user_audit_policy_add(xp, result, skb) or
pfkey_audit_policy_add(xp, result) will get called.
I need two because xfrm_user gets loginuid/secid from
From: Auke Kok [EMAIL PROTECTED]
Date: Thu, 23 Aug 2007 07:59:11 -0700
This incorporates the new napi_struct changes into e1000e. Included
bugfix for ifdown hang from Krishna Kumar for e1000.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
Acked-by: David S. Miller [EMAIL PROTECTED]
-
To
Alexey Dobriyan napsal(a):
sysctl(2) is supported but frozen.
I've posted similar patch yesterday:
http://marc.info/?l=linux-mm-commitsm=118782442602108w=2
Signed-off-by: Alexey Dobriyan [EMAIL PROTECTED]
---
drivers/net/wireless/ath5k_base.c | 21 ++---
1 file changed,
On Sun, Aug 12, 2007 at 05:33:16PM +0200, Jiri Slaby wrote:
add ath5k wireless driver
Signed-off-by: Jiri Slaby [EMAIL PROTECTED]
Review still pending, but I went ahead and added this on the 'ath5k'
branch of wireless-dev. It is available on 'everything' as well.
Thanks,
John
--
John W.
Roland/All,
Here is the first swipe at keeping iwarp connections on their own ip
addresses to avoid conflicts with the host stack.
- this is a request for comments
- it is not yet tested fully (tested a prototype of the initial concept)
- still needs serialization/locking
- stays in our RDMA
Hi Al,
Speaking of cxgb3, could you explain what the hell is
static int do_term(struct t3cdev *dev, struct sk_buff *skb)
{
unsigned int hwtid = ntohl(skb-priority) 8 0xf;
doing? AFAIK, skb-priority is not net-endian...
the RDMA connection id is saved in the skb's priority
On Wed, 2007-22-08 at 13:21 -0700, David Miller wrote:
From: Rick Jones [EMAIL PROTECTED]
Date: Wed, 22 Aug 2007 10:09:37 -0700
Should it be any more or less worrysome than small packet
performance (eg the TCP_RR stuff I posted recently) being rather
worse with TSO enabled than with it
From: Divy Le Ray [EMAIL PROTECTED]
cxgb3 used netdev_priv() and dev-priv for different purposes.
In 2.6.23, netdev_priv() == dev-priv, cxgb3 needs a fix.
This patch is a partial backport of Dave Miller's changes in the
net-2.6.24 git branch.
Without this fix, cxgb3 crashes on 2.6.23.
On Thu, 2007-23-08 at 18:04 -0400, jamal wrote:
The litmus test is the same as any change that is supposed to improve
net performance - it has to demonstrate it is not intrusive and that it
improves (consistently) performance. The standard metrics are
{throughput, cpu-utilization, latency}
-stable review patch. If anyone has any objections, please let us know.
--
[NET]: Share correct feature code between bridging and bonding
http://bugzilla.kernel.org/show_bug.cgi?id=8797 shows that the
bonding driver may produce bogus combinations of the checksum
flags and
From: jamal [EMAIL PROTECTED]
Date: Thu, 23 Aug 2007 18:04:10 -0400
Possibly a bug - but you really should turn off TSO if you are doing
huge interactive transactions (which is fair because there is a clear
demarcation).
I don't see how this can matter.
TSO only ever does anything if you
jamal wrote:
[TSO already passed - iirc, it has been
demostranted to really not add much to throughput (cant improve much
over closeness to wire speed) but improve CPU utilization].
In the one gig space sure, but in the 10 Gig space, TSO on/off does make a
difference for throughput.
rick
On Thu, 2007-23-08 at 15:35 -0700, Rick Jones wrote:
jamal wrote:
[TSO already passed - iirc, it has been
demostranted to really not add much to throughput (cant improve much
over closeness to wire speed) but improve CPU utilization].
In the one gig space sure, but in the 10 Gig space,
[Adding netdev to CC]
On 21/08/07, poison [EMAIL PROTECTED] wrote:
Hello,
after running a few instances of bittorent-curses on 2.6.22 - 2.6.22.3 it
takes about 15min to 2hrs for my System to hang. 2.6.21.7 is definately fine,
2.6.21 probably (ran for 4hrs without hanging).
If I'm lucky the
100% reproducible hang on xmit timeout.
Just do a make -j4 modules on an nfs mounted kernel source.
attached is the messages log
berkley
--
// E. F. Berkley Shands, MSc//
** Exegy Inc.**
349 Marshall Road, Suite 100
St. Louis , MO 63119
Direct: (314) 218-3600 X450
Cell: (314) 303-2546
kmalloc() and friends return void*, no need to cast it.
Signed-off-by: Jesper Juhl [EMAIL PROTECTED]
---
drivers/net/wireless/libertas/debugfs.c |2 +-
drivers/net/wireless/libertas/ethtool.c |3 +--
2 files changed, 2 insertions(+), 3 deletions(-)
diff --git
The general kernel memory allocation functions return void pointers
and there is no need to cast their return values.
Signed-off-by: Jesper Juhl [EMAIL PROTECTED]
---
drivers/net/bsd_comp.c |6 ++
1 files changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/net/bsd_comp.c
I'd welcome the views of those familiar with TCP_DEFER_ACCEPT on a
recent issue I've worked on where connections between a Juniper DX (aka
redline) load-balancer and Apache 2.2 cluster caused random connection
failures.
Today, after 2 weeks debugging the issue, we confirmed the problem was
o Support policy flag with string format.
Note that kernel defines only one name localok for the flag
and it has not had any effect currently.
o Support state flag value XFRM_STATE_NOPMTUDISC.
o Fix to show detailed flags value when -s option is used.
o Fix minor typo.
Signed-off-by: Masahide
Remove unused or redundant usage for xfrm_filter.
Signed-off-by: Masahide NAKAMURA [EMAIL PROTECTED]
---
ip/xfrm_policy.c | 17 -
ip/xfrm_state.c |2 --
2 files changed, 0 insertions(+), 19 deletions(-)
diff --git a/ip/xfrm_policy.c b/ip/xfrm_policy.c
index
Fix xfrm state or policy flush message.
And minor updates are included:
o Use static buffer to show unknown value as string.
o Show policy type (ptype) only when kernel specified it.
o Clean-up xfrm_monitor.
Signed-off-by: Masahide NAKAMURA [EMAIL PROTECTED]
---
ip/ipxfrm.c | 48
Hello,
There are updates for ip command. They are almost minor fixes
and are not changes about 2.6.23 new features.
Please apply if it is not too late for next release.
--
Masahide NAKAMURA
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL
Jiri Slaby wrote:
s2io, rename BIT macro
BIT macro will be global definiton of (1x)
Signed-off-by: Jiri Slaby [EMAIL PROTECTED]
---
[snip]
cnt++;
if (cnt == 5)
diff --git a/drivers/net/s2io.h b/drivers/net/s2io.h
index 92983ee..448f899
On Thu, 23 Aug 2007, Rick Jones wrote:
jamal wrote:
[TSO already passed - iirc, it has been
demostranted to really not add much to throughput (cant improve much
over closeness to wire speed) but improve CPU utilization].
In the one gig space sure, but in the 10 Gig space, TSO on/off
subscribe netdev
-
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
On Thu, 23 Aug 2007 18:38:22 -0400
jamal [EMAIL PROTECTED] wrote:
On Thu, 2007-23-08 at 15:30 -0700, David Miller wrote:
From: jamal [EMAIL PROTECTED]
Date: Thu, 23 Aug 2007 18:04:10 -0400
Possibly a bug - but you really should turn off TSO if you are doing
huge interactive
On Thu, 23 Aug 2007 13:13:10 -0500 Olof Johansson [EMAIL PROTECTED] wrote:
out:
- pci_dev_put(mac-iob_pdev);
-out_put_dma_pdev:
- pci_dev_put(mac-dma_pdev);
-out_free_netdev:
+ if (mac-iob_pdev)
+ pci_dev_put(mac-iob_pdev);
+ if (mac-dma_pdev)
+
TJ wrote:
client SYN server LISTENING
client SYN ACK server SYN_RECEIVED (time-out 3s)
server: inet_rsk(req)-acked = 1
client ACK server (discarded)
client SYN ACK (DUP) server (time-out 6s)
client ACK (DUP) server (discarded)
client SYN ACK (DUP) server (time-out 12s)
On Thu, Aug 23, 2007 at 06:48:23PM -0500, Mr. Berkley Shands wrote:
100% reproducible hang on xmit timeout.
Just do a make -j4 modules on an nfs mounted kernel source.
Most likely you also had the problem with 2.6.22.2 (maybe you have not
tested this one, though). There were bug fixes for
On Thu, Aug 23, 2007 at 10:44:30AM +0200, Jarek Poplawski wrote:
Andrew Morton pointed out that my changelog was unusable. Sorry!
Here is a second try with the changelog and kernel version changed.
...
(take 2)
Subject: request_irq() - fix DEBUG_SHIRQ handling
...
Signed-off-by:
69 matches
Mail list logo
