Francois Romieu [mailto:[EMAIL PROTECTED] wrote:
Afaics your change may disable the Rx irq right after the poll routine
enabled it again. It will not always work either.
The (slow) timeout watchdog could grab the poll handler and hack the
irq mask depending on whether poll was scheduled
On 11/29/06, David Miller [EMAIL PROTECTED] wrote:
From: Benjamin Herrenschmidt [EMAIL PROTECTED]
Date: Wed, 29 Nov 2006 09:57:24 +1100
This looks mostly fine.
I was thinking about the lockless stuff, and I wonder if there
is a clever way you can get it back down to one PIO on the
On 11/28/06, David Miller [EMAIL PROTECTED] wrote:
From: Eric Lemoine [EMAIL PROTECTED]
Date: Tue, 14 Nov 2006 22:54:40 +0100
On 11/14/06, David Miller [EMAIL PROTECTED] wrote:
From: Eric Lemoine [EMAIL PROTECTED]
Date: Tue, 14 Nov 2006 08:28:42 +0100
because it makes it explicit that
Fixes a typo which caused fib_props[] to have the wrong size
and makes sure the value used to index the array which is
provided by userspace via netlink is checked to avoid out of
bound access.
Signed-off-by: Thomas Graf [EMAIL PROTECTED]
Index: net-2.6/net/ipv4/fib_semantics.c
I 've already ported ar5k to linux and it works with madwifi versions
before the bsd-head merge, you can see more infos here -
http://madwifi.org/wiki/OpenHAL
If i can help in any way feel free to mail ;-)
Nick
2006/11/28, Michael Buesch [EMAIL PROTECTED]:
On Tuesday 28 November 2006 21:39,
Fixes a typo which caused fib_props[] to have the wrong size
and makes sure the value used to index the array which is
provided by userspace via netlink is checked to avoid out of
bound access.
Signed-off-by: Thomas Graf [EMAIL PROTECTED]
Index: net-2.6/net/decnet/dn_fib.c
On Wednesday 29 November 2006 14:55, Nick Kossifidis wrote:
I 've already ported ar5k to linux and it works with madwifi versions
No, you misunderstood me.
Madwifi is not a native driver and will never be accepted into
mainline. My attempt is to write a native d80211 driver based
on the ar5k
On Tue, 28 Nov 2006 20:56:05 +0100, Ivo van Doorn wrote:
After a succesfull authentication and association the matching retry counter
must be reset to 0.
Failure to do so will result in failure to authenticate after the interface
has been deauthenticated. This does not always happen after the
* Horms [EMAIL PROTECTED] 2006-11-29 15:21
This seems to be a pretty clean solution to a real problem.
Ultimately I would like to see IPVS move into the forward chain.
This seems to be a nice way to explore that, without breaking
any existing setups.
--
Horms
H:
On 11/29/06, Jiri Benc [EMAIL PROTECTED] wrote:
On Tue, 28 Nov 2006 20:56:05 +0100, Ivo van Doorn wrote:
After a succesfull authentication and association the matching retry
counter
must be reset to 0.
Failure to do so will result in failure to authenticate after the
interface
has been
On Wed, 29 Nov 2006 15:27:06 +0100, Ivo Van Doorn wrote:
Shouldn't this last one be:
ieee80211_set_disassoc(dev, ifsta, 0)
This one is called from the IOCTL request to dissassociate,
so the interface should still be authenticated (with a valid
auth retry counter).
Yes, of course. Thanks
Good luck then ;-)
If anyone wants to help on making ar5k work with newer madwifi
versions and fix bugs etc (that 'll also help bsd ppl) plzz mail me.
We can make it better.
Nick
P.S. Why not work on dawifi ?
2006/11/29, Michael Buesch [EMAIL PROTECTED]:
On Wednesday 29 November 2006 14:55,
On Wed, Nov 29, 2006 at 03:15:23PM +0100, Thomas Graf wrote:
* Horms [EMAIL PROTECTED] 2006-11-29 15:21
This seems to be a pretty clean solution to a real problem.
Ultimately I would like to see IPVS move into the forward chain.
This seems to be a nice way to explore that, without
On Wednesday 29 November 2006 15:34, Nick Kossifidis wrote:
Good luck then ;-)
If anyone wants to help on making ar5k work with newer madwifi
versions and fix bugs etc (that 'll also help bsd ppl) plzz mail me.
We can make it better.
Nick
P.S. Why not work on dawifi ?
Because it won't
On Wed, 2006-11-29 at 16:12 +0100, Michael Buesch wrote:
On Wednesday 29 November 2006 15:34, Nick Kossifidis wrote:
Good luck then ;-)
If anyone wants to help on making ar5k work with newer madwifi
versions and fix bugs etc (that 'll also help bsd ppl) plzz mail me.
We can make it
On Wed, Nov 29, 2006 at 04:12:33PM +0100, Michael Buesch wrote:
On Wednesday 29 November 2006 15:34, Nick Kossifidis wrote:
Good luck then ;-)
If anyone wants to help on making ar5k work with newer madwifi
versions and fix bugs etc (that 'll also help bsd ppl) plzz mail me.
We can make
On Wed, Nov 29, 2006 at 10:21:09AM -0500, Dan Williams wrote:
On Wed, 2006-11-29 at 16:12 +0100, Michael Buesch wrote:
On Wednesday 29 November 2006 15:34, Nick Kossifidis wrote:
Good luck then ;-)
If anyone wants to help on making ar5k work with newer madwifi
versions and fix bugs
Hi Horms,
I see that this patch probably makes IPVS code a bit complicated and
packet traversing less efficiently.
If I remember correctly, policy-based routing can work with IPVS in
kernel 2.2 and 2.4 for transparent cache cluster for a long time. It
should work in kernel 2.6 too.
For
On Wednesday 29 November 2006 16:24, David Kimdon wrote:
On Wed, Nov 29, 2006 at 04:12:33PM +0100, Michael Buesch wrote:
On Wednesday 29 November 2006 15:34, Nick Kossifidis wrote:
Good luck then ;-)
If anyone wants to help on making ar5k work with newer madwifi
versions and fix
On Wednesday 29 November 2006 16:24, David Kimdon wrote:
On Wed, Nov 29, 2006 at 04:12:33PM +0100, Michael Buesch wrote:
On Wednesday 29 November 2006 15:34, Nick Kossifidis wrote:
Good luck then ;-)
If anyone wants to help on making ar5k work with newer madwifi
versions and fix
On Wed, 2006-11-29 at 15:55 +0200, Nick Kossifidis wrote:
I 've already ported ar5k to linux and it works with madwifi versions
before the bsd-head merge, you can see more infos here -
http://madwifi.org/wiki/OpenHAL
If i can help in any way feel free to mail ;-)
Thanks, I'm trying it out
Hi.
On Wednesday 29 November 2006 16:24, David Kimdon wrote:
There is absolutely no reason why dadwifi can't be merged into the
mainline once the hal issue is resolved.
Last time we talked about that stuff, it was decided that
we don't want a HAL... See archives.
IIRC Pavel already
On Wed, Nov 29, 2006 at 04:38:56PM +0100, Michael Buesch wrote:
On Wednesday 29 November 2006 16:24, David Kimdon wrote:
On Wed, Nov 29, 2006 at 04:12:33PM +0100, Michael Buesch wrote:
On Wednesday 29 November 2006 15:34, Nick Kossifidis wrote:
Why do you say that?
There is absolutely
On Wednesday 29 November 2006 16:58, Michael Renzmann wrote:
Hi.
On Wednesday 29 November 2006 16:24, David Kimdon wrote:
There is absolutely no reason why dadwifi can't be merged into the
mainline once the hal issue is resolved.
Last time we talked about that stuff, it was decided that
On Wed, Nov 29, 2006 at 03:33:07PM +0100, Jiri Benc wrote:
On Wed, 29 Nov 2006 15:27:06 +0100, Ivo Van Doorn wrote:
Shouldn't this last one be:
ieee80211_set_disassoc(dev, ifsta, 0)
This one is called from the IOCTL request to dissassociate,
so the interface should still be
I will be sending NetXen: 1G/10G Ethernet Driver updates in subsequent emails.
Thanks,
--Amit
-
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
Signed-off-by: Amit S. Kale [EMAIL PROTECTED]
netxen_nic_main.c |3 ++-
1 files changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/net/netxen/netxen_nic_main.c
b/drivers/net/netxen/netxen_nic_main.c
index 145bf47..a055208 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++
NetXen: 1G/10G Ethernet Driver updates
- These fixes take care of driver on machines with 4G memory
- Driver cleanup
Signed-off-by: Amit S. Kale [EMAIL PROTECTED]
netxen_nic.h | 41 ++
netxen_nic_ethtool.c | 19 ++--
netxen_nic_hw.c | 10 +-
Confusingly NET_PCI is also set for for non-PCI EISA configurations where
building this driver will result in a build error due to a reference to
pci_release_regions.
While at it, remove the EXPERIMENTAL - in all its uglyness and despite
the sincerest attempts of the buggy hardware the driver is
This patch adds auditing to ipsec.
An audit message occurs when an ipsec SA
or ipsec policy is created/deleted.
Patch was built against linux kernel 2.6.19-rc6.
Please let me know if this is acceptable.
Regards,
Joy
Signed-off-by: Joy Latten [EMAIL PROTECTED]
This patchset consists of three patches that add support for the remaining two
tag types from the CIPSO draft specification, the enumerated and range tags.
The most significant part about adding these two tags is that NetLabel now has
the ability to represent more than 240 categories (limitation
From: Paul Moore [EMAIL PROTECTED]
Add support for the enumerated tag (tag type #2) to the CIPSOv4 protocol.
The enumerated tag allows for 15 categories to be specified in a CIPSO option,
where each category is an unsigned 16 bit field with a maximum value of 65534.
See
From: Paul Moore [EMAIL PROTECTED]
Add support for the ranged tag (tag type #5) to the CIPSOv4 protocol.
The ranged tag allows for seven, or eight if zero is the lowest category,
category ranges to be specified in a CIPSO option. Each range is specified by
two unsigned 16 bit fields, each with
--- [EMAIL PROTECTED] wrote:
This patchset consists of three patches that add
support for the remaining two
tag types from the CIPSO draft specification, the
enumerated and range tags.
The most significant part about adding these two
tags is that NetLabel now has
the ability to represent
This changes the type of variable i in
rtl8169_init_one() from unsigned int to int.
i is checked for 0 later, which can never happen
for unsigned. This results in broken error handling.
Signed-off-by: Michael Buesch [EMAIL PROTECTED]
Index: linux-2.6/drivers/net/r8169.c
On Tue, Nov 28, 2006 at 03:33:25PM -0800, David Miller wrote:
From: Alexey Dobriyan [EMAIL PROTECTED]
Date: Wed, 22 Nov 2006 00:22:51 +0300
[CCing netdev, bug in pktgen]
[build modular pktgen]
while true; do modprobe pktgen rmmod pktgen; done
BUG: warning at
On Wed, 29 Nov 2006 08:03:28 -0800
David Kimdon [EMAIL PROTECTED] wrote:
On Wed, Nov 29, 2006 at 04:38:56PM +0100, Michael Buesch wrote:
On Wednesday 29 November 2006 16:24, David Kimdon wrote:
On Wed, Nov 29, 2006 at 04:12:33PM +0100, Michael Buesch wrote:
On Wednesday 29 November 2006
Eric W. Biederman wrote:
I think for cases across network socket namespaces it should
be a matter for the rules, to decide if the connection should
happen and what error code to return if the connection does not
happen.
There is a potential in this to have an ambiguous case where two
On Wed, 29 Nov 2006 14:20:31 +0530
Basheer, Mansoor Ahamed [EMAIL PROTECTED] wrote:
Francois Romieu [mailto:[EMAIL PROTECTED] wrote:
Afaics your change may disable the Rx irq right after the poll routine
enabled it again. It will not always work either.
The (slow) timeout watchdog
On Wed, 29 Nov 2006, [EMAIL PROTECTED] wrote:
+{
+ /* The constant '16' is not random, it is the maximum number of
+ * high/low category range pairs as permitted by the CIPSO draft based
+ * on a maximum IPv4 header length of 60 bytes - the BUG_ON() assertion
+ * does a
From: Randy Dunlap [EMAIL PROTECTED]
Use NULL instead of 0 for pointers (cures sparse warnings).
drivers/net/sundance.c:1106:16: warning: Using plain integer as NULL pointer
drivers/net/sundance.c:1652:16: warning: Using plain integer as NULL pointer
Signed-off-by: Randy Dunlap [EMAIL
This changes the type of variable i in rtl8169_init_one()
from unsigned int to int. i is checked for 0 later,
which can never happen for unsigned. This results in broken
error handling.
Signed-off-by: Michael Buesch [EMAIL PROTECTED]
Signed-off-by: Francois Romieu [EMAIL PROTECTED]
diff
Stephen Hemminger wrote:
That motherboard has dual lan, are you using both of them?
I don't have that chip version, so hard to tell if it is using dual port
with a single chip
or not. There is a hack for the dual port PCI-X version already in the
driver,
that turns off receive checksums if
Brian Haley wrote:
Eric W. Biederman wrote:
I think for cases across network socket namespaces it should
be a matter for the rules, to decide if the connection should
happen and what error code to return if the connection does not
happen.
There is a potential in this to have an ambiguous case
All applied to:
git://git.infradead.org/~jmorris/selinux-net-2.6.20
Thanks,
- James
--
James Morris
[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
James Morris wrote:
All applied to:
git://git.infradead.org/~jmorris/selinux-net-2.6.20
Thanks.
Did you mean your kernel.org git tree?
--
paul moore
linux security @ hp
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More
On Monday 27 November 2006 14:11, Joy Latten wrote:
Please let me know if this is acceptable.
From an audit perspective, it looks good.
-Steve
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
Stephen Hemminger [EMAIL PROTECTED] :
[...]
@@ -1682,12 +1685,11 @@ static void rtl8139_tx_timeout_task (voi
rtl8139_tx_clear (tp);
spin_unlock_irq(tp-lock);
+ netif_poll_enable();
^ - dev
+
/* ...and finally, reset everything */
- if
On Wed, 29 Nov 2006 23:44:00 +0100
Francois Romieu [EMAIL PROTECTED] wrote:
Stephen Hemminger [EMAIL PROTECTED] :
[...]
@@ -1682,12 +1685,11 @@ static void rtl8139_tx_timeout_task (voi
rtl8139_tx_clear (tp);
spin_unlock_irq(tp-lock);
+ netif_poll_enable();
Stephen Hemminger [EMAIL PROTECTED] :
Francois Romieu [EMAIL PROTECTED] wrote:
Stephen Hemminger [EMAIL PROTECTED] :
[...]
@@ -1682,12 +1685,11 @@ static void rtl8139_tx_timeout_task (voi
rtl8139_tx_clear (tp);
spin_unlock_irq(tp-lock);
+ netif_poll_enable();
On Wed, 29 Nov 2006, Paul Moore wrote:
James Morris wrote:
All applied to:
git://git.infradead.org/~jmorris/selinux-net-2.6.20
Thanks.
Did you mean your kernel.org git tree?
There's a copy at infradead (which may have still been cloning if you
checked it immediately).
--
This patch disables auditing in ipsec when CONFIG_AUDITSYSCALL is
disabled in the kernel.
This patch also includes a bug fix for xfrm_state.c as a result of
original ipsec audit patch.
Let me know if it looks ok.
My mail gateway has been acting crazy so I apologize for any
replicas being sent
On Thu, 30 Nov 2006 00:32:19 +0100
Francois Romieu [EMAIL PROTECTED] wrote:
Stephen Hemminger [EMAIL PROTECTED] :
Francois Romieu [EMAIL PROTECTED] wrote:
Stephen Hemminger [EMAIL PROTECTED] :
[...]
@@ -1682,12 +1685,11 @@ static void rtl8139_tx_timeout_task (voi
Stephen Hemminger [EMAIL PROTECTED] :
[...]
Move the poll_enable to after hw_start() or put it inside hw_start.
after probably The order would be the opposite of the one used
in rtl8139_poll (which does __netif_rx_complete then irq_unlock)
and it's past 1 AM. It starts to be a bit foggy.
--
On Wed, 29 Nov 2006, Joy Latten wrote:
This patch disables auditing in ipsec when CONFIG_AUDITSYSCALL is
disabled in the kernel.
This patch also includes a bug fix for xfrm_state.c as a result of
original ipsec audit patch.
Let me know if it looks ok.
Also, the last patch contains no
On Wed, 29 Nov 2006, James Morris wrote:
On Wed, 29 Nov 2006, Joy Latten wrote:
This patch disables auditing in ipsec when CONFIG_AUDITSYSCALL is
disabled in the kernel.
This patch also includes a bug fix for xfrm_state.c as a result of
original ipsec audit patch.
Let me know
From: Patrick McHardy [EMAIL PROTECTED]
Date: Mon, 20 Nov 2006 14:08:38 +0100 (MET)
[NET_SCHED]: Set parent classid in default qdiscs
Set parent classids in default qdiscs to allow walking up the tree
from outside the qdiscs. This is needed by the next patch.
Signed-off-by: Patrick
From: Patrick McHardy [EMAIL PROTECTED]
Date: Mon, 20 Nov 2006 14:08:37 +0100 (MET)
[NET_SCHED]: sch_htb: perform qlen adjustment immediately in -delete
qlen adjustment should happen immediately in -delete and not in the
class destroy function because the reference count will not hit zero in
From: Patrick McHardy [EMAIL PROTECTED]
Date: Mon, 20 Nov 2006 14:08:44 +0100 (MET)
[NET_SCHED]: Fix endless loops (part 2): simple qdiscs
Convert the simple qdiscs to use qdisc_tree_decrease_qlen() where
necessary:
- all graft operations
- destruction of old child qdiscs in prio, red
From: Patrick McHardy [EMAIL PROTECTED]
Date: Mon, 20 Nov 2006 14:08:41 +0100 (MET)
[NET_SCHED]: Fix endless loops caused by inaccurate qlen counters (part 1)
There are multiple problems related to qlen adjustment that can lead
to an upper qdisc getting out of sync with the real number of
From: Patrick McHardy [EMAIL PROTECTED]
Date: Mon, 20 Nov 2006 14:08:48 +0100 (MET)
[NET_SCHED]: Fix endless loops (part 4): HTB
Convert HTB to use qdisc_tree_decrease_len() and add a callback
for deactivating a class when its child queue becomes empty.
Signed-off-by: Patrick McHardy
From: Patrick McHardy [EMAIL PROTECTED]
Date: Mon, 20 Nov 2006 14:08:46 +0100 (MET)
[NET_SCHED]: Fix endless loops (part 3): HFSC
Convert HFSC to use qdisc_tree_decrease_len() and add a callback
for deactivating a class when its child queue becomes empty.
All queue purging goes through
From: Patrick McHardy [EMAIL PROTECTED]
Date: Mon, 20 Nov 2006 16:01:03 +0100
I forgot to fix one (AFAICT purely theoretical) case ..
Also applied to net-2.6.20, thanks a lot Patrick.
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
From: Alexey Dobriyan [EMAIL PROTECTED]
Date: Wed, 29 Nov 2006 23:04:37 +0300
Looks like worker thread strategically clears it if scheduled at wrong
moment.
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -3292,7 +3292,6 @@ static void pktgen_thread_worker(struct
hi, Wensong. Thanks for your appraise.
I see that this patch probably makes IPVS code a bit complicated and
packet traversing less efficiently.
In my opinion, worry about the side-effect to the packet throughput is not
necessary. First, normal packets with mark rarely appear in the
Hello,
I found a bug in my previous patch for af_key.
The patch breaks transport mode.
This is a fixed version.
Signed-off-by: Miika Komu [EMAIL PROTECTED]
Signed-off-by: Diego Beltrami [EMAIL PROTECTED]
Signed-off-by: Kazunori Miyazawa [EMAIL PROTECTED]
diff --git a/net/key/af_key.c
From: Phil Oester [EMAIL PROTECTED]
Date: Wed, 29 Nov 2006 17:49:04 -0800
Getting an oops on boot here, caused by commit
e81c73596704793e73e6dbb478f41686f15a4b34 titled
[NET]: Fix MAX_HEADER setting.
Reverting that patch fixes things up for me. Dave?
I suspect that it might be because I
On Wed, Nov 29, 2006 at 06:15:37PM -0800, David Miller wrote:
In fact it does, the NDISC code is using MAX_HEADER incorrectly. It
needs to explicitly allocate space for the struct ipv6hdr in 'len'.
Luckily the TCP ipv6 code was doing it right.
What a horrible bug, this patch should fix it.
On 29-11-2006 08:49, Herbert Xu wrote:
Peter Zijlstra [EMAIL PROTECTED] wrote:
=
[ INFO: possible irq lock inversion dependency detected ]
2.6.19-rc6 #4
-
nc/1854 just changed the
Jesse Brandeburg wrote:
John wrote:
Here is some context for those who have been added to the CC list:
http://groups.google.com/group/linux.kernel/browse_frm/thread/bdc8fd08fb601c26
As far as I understand, some consider the eepro100 driver to be
obsolete, and it has been considered for
On Wed, 2006-11-29 at 18:49 +1100, Herbert Xu wrote:
Peter Zijlstra [EMAIL PROTECTED] wrote:
=
[ INFO: possible irq lock inversion dependency detected ]
2.6.19-rc6 #4
-
On Wed, Nov 29, 2006 at 12:42:24PM +0100, Peter Zijlstra wrote:
However I'm not quite sure yet how to teach lockdep about this. The
proposed patch will shut it up though.
As a rule I think we should never make semantic changes to shut up
lockdep.
Cheers,
--
Visit Openswan at
Jarek Poplawski [EMAIL PROTECTED] writes:
And if we talk about names:
+ Spotted by Krzysztof Halasa.
I wound't care less btw.
--
Krzysztof Halasa
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
From: Herbert Xu [EMAIL PROTECTED]
Date: Wed, 29 Nov 2006 23:07:09 +1100
On Wed, Nov 29, 2006 at 12:42:24PM +0100, Peter Zijlstra wrote:
However I'm not quite sure yet how to teach lockdep about this. The
proposed patch will shut it up though.
As a rule I think we should never make
On Wed, 29 Nov 2006 21:08:00 +0100
Rafael J. Wysocki [EMAIL PROTECTED] wrote:
On Wednesday, 29 November 2006 20:54, Rafael J. Wysocki wrote:
On Tuesday, 28 November 2006 11:02, Andrew Morton wrote:
Temporarily at
http://userweb.kernel.org/~akpm/2.6.19-rc6-mm2/
Will appear
On Wednesday, 29 November 2006 22:30, Andrew Morton wrote:
On Wed, 29 Nov 2006 21:08:00 +0100
Rafael J. Wysocki [EMAIL PROTECTED] wrote:
On Wednesday, 29 November 2006 20:54, Rafael J. Wysocki wrote:
On Tuesday, 28 November 2006 11:02, Andrew Morton wrote:
Temporarily at
-stable review patch. If anyone has any objections, please let us know.
--
From: Michael Buesch [EMAIL PROTECTED]
Drain the Microcode TX-status-FIFO before we enable IRQs.
This is required, because the FIFO may still have entries left
from a previous run. Those would immediately
On Wednesday, 29 November 2006 22:31, Rafael J. Wysocki wrote:
On Wednesday, 29 November 2006 22:30, Andrew Morton wrote:
On Wed, 29 Nov 2006 21:08:00 +0100
Rafael J. Wysocki [EMAIL PROTECTED] wrote:
On Wednesday, 29 November 2006 20:54, Rafael J. Wysocki wrote:
On Tuesday, 28
From: Wenji Wu [EMAIL PROTECTED]
Greetings,
For Linux TCP, when the network applcaiton make system call to move data
from
socket's receive buffer to user space by calling tcp_recvmsg(). The socket
will
be locked. During the period, all the incoming packet for the TCP socket
will go
to the
From: Wenji Wu [EMAIL PROTECTED]
Greetings,
For Linux TCP, when the network applcaiton make system call to move data
from
socket's receive buffer to user space by calling tcp_recvmsg(). The socket
will
be locked. During the period, all the incoming packet for the TCP socket
will go
to the
From: Wenji Wu [EMAIL PROTECTED]
Greetings,
For Linux TCP, when the network applcaiton make system call to move data
from
socket's receive buffer to user space by calling tcp_recvmsg(). The socket
will
be locked. During the period, all the incoming packet for the TCP socket
will go
to the
From: Wenji Wu [EMAIL PROTECTED]
Greetings,
For Linux TCP, when the network applcaiton make system call to move data
from
socket's receive buffer to user space by calling tcp_recvmsg(). The socket
will
be locked. During the period, all the incoming packet for the TCP socket
will go
to the
From: Wenji Wu [EMAIL PROTECTED]
Greetings,
For Linux TCP, when the network applcaiton make system call to move data
from
socket's receive buffer to user space by calling tcp_recvmsg(). The socket
will
be locked. During the period, all the incoming packet for the TCP socket
will go
to the
On Thu, 30 Nov 2006 00:08:21 +0100
Rafael J. Wysocki [EMAIL PROTECTED] wrote:
On Wednesday, 29 November 2006 22:31, Rafael J. Wysocki wrote:
On Wednesday, 29 November 2006 22:30, Andrew Morton wrote:
On Wed, 29 Nov 2006 21:08:00 +0100
Rafael J. Wysocki [EMAIL PROTECTED] wrote:
On
On Wed, 29 Nov 2006 17:22:10 -0600
Wenji Wu [EMAIL PROTECTED] wrote:
From: Wenji Wu [EMAIL PROTECTED]
Greetings,
For Linux TCP, when the network applcaiton make system call to move data
from
socket's receive buffer to user space by calling tcp_recvmsg(). The socket
will
be locked.
Wenji Wu wrote:
From: Wenji Wu [EMAIL PROTECTED]
Greetings,
For Linux TCP, when the network applcaiton make system call to move data
from
socket's receive buffer to user space by calling tcp_recvmsg(). The socket
will
be locked. During the period, all the incoming packet for the TCP socket
Please, it is very difficult to review your work the way you have
submitted this patch as a set of 4 patches. These patches have not
been split up logically, but rather they have been split up per
file with the same exact changelog message in each patch posting.
This is very clumsy, and
On Wed, 29 Nov 2006 16:53:11 -0800 (PST)
David Miller [EMAIL PROTECTED] wrote:
Please, it is very difficult to review your work the way you have
submitted this patch as a set of 4 patches. These patches have not
been split up logically, but rather they have been split up per
file with the
On Thursday, 30 November 2006 00:26, Andrew Morton wrote:
On Thu, 30 Nov 2006 00:08:21 +0100
Rafael J. Wysocki [EMAIL PROTECTED] wrote:
On Wednesday, 29 November 2006 22:31, Rafael J. Wysocki wrote:
On Wednesday, 29 November 2006 22:30, Andrew Morton wrote:
On Wed, 29 Nov 2006
From: Andrew Morton [EMAIL PROTECTED]
Date: Wed, 29 Nov 2006 17:08:35 -0800
On Wed, 29 Nov 2006 16:53:11 -0800 (PST)
David Miller [EMAIL PROTECTED] wrote:
Please, it is very difficult to review your work the way you have
submitted this patch as a set of 4 patches. These patches have
Yes, when CONFIG_PREEMPT is disabled, the problem won't happen. That is why I
put for 2.6 desktop, low-latency desktop in the uploaded paper. This
problem happens in the 2.6 Desktop and Low-latency Desktop.
We could also pepper tcp_recvmsg() with some very carefully placed preemption
That yield() will need to be removed - yield()'s behaviour is truly
awfulif the system is otherwise busy. What is it there for?
Please read the uploaded paper, which has detailed description.
thanks,
wenji
- Original Message -
From: Andrew Morton [EMAIL PROTECTED]
Date: Wednesday,
From: Wenji Wu [EMAIL PROTECTED]
Date: Wed, 29 Nov 2006 19:56:58 -0600
We could also pepper tcp_recvmsg() with some very carefully placed
preemption disable/enable calls to deal with this even with
CONFIG_PREEMPT enabled.
I also think about this approach. But since the problem happens in
On Thu, 30 Nov 2006 02:04:15 +0100
Rafael J. Wysocki [EMAIL PROTECTED] wrote:
git-netdev-all.patch
git-netdev-all-fixup.patch
libphy-dont-do-that.patch
Are you able to eliminate libphy-dont-do-that.patch?
Is a broken-out version of git-netdev-all.patch available from
Hi.
Michael Buesch wrote:
IIRC Pavel already explained that getting rid of the HAL per se should be
no problem - it could easily be dissolved into the driver, if that is one
of the requirements to be fulfilled before the driver (MadWifi or DadWifi)
is considered for mainline inclusion. As soon
On 11/30/06, David Miller [EMAIL PROTECTED] wrote:
From: Alexey Dobriyan [EMAIL PROTECTED]
Date: Wed, 29 Nov 2006 23:04:37 +0300
Looks like worker thread strategically clears it if scheduled at wrong
moment.
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -3292,7 +3292,6 @@ static void
On Wed, 2006-11-29 at 17:08 -0800, Andrew Morton wrote:
+ if (p-backlog_flag == 0) {
+ if (!TASK_INTERACTIVE(p) || expired_starving(rq)) {
+ enqueue_task(p, rq-expired);
+ if (p-static_prio
* David Miller [EMAIL PROTECTED] wrote:
We can make explicitl preemption checks in the main loop of
tcp_recvmsg(), and release the socket and run the backlog if
need_resched() is TRUE.
This is the simplest and most elegant solution to this problem.
yeah, i like this one. If the problem
* Wenji Wu [EMAIL PROTECTED] wrote:
That yield() will need to be removed - yield()'s behaviour is truly
awfulif the system is otherwise busy. What is it there for?
Please read the uploaded paper, which has detailed description.
do you have any URL for that?
Ingo
-
To
From: Ingo Molnar [EMAIL PROTECTED]
Date: Thu, 30 Nov 2006 07:17:58 +0100
* David Miller [EMAIL PROTECTED] wrote:
We can make explicitl preemption checks in the main loop of
tcp_recvmsg(), and release the socket and run the backlog if
need_resched() is TRUE.
This is the simplest
1 - 100 of 104 matches
Mail list logo