David Miller a écrit :
From: Eric Dumazet [EMAIL PROTECTED]
Date: Tue, 31 Oct 2006 19:48:48 +0100
We currently insert sockets/pipes dentries into the global dentry
hashtable. This is *useless* because there is currently no way
these entries can be used for a lookup(). (/proc/xxx/fd/xxx uses a
From: Eric Dumazet [EMAIL PROTECTED]
Date: Wed, 01 Nov 2006 09:21:06 +0100
No no, my patch takes care of that.
You still see the right link for pipes and sockets on /proc/$pid/fd/XXX
And (deleted) is correctly added to deleted files.
I see. Excellent :-)
-
To unsubscribe from this list:
On Wed, Nov 01, 2006 at 09:21:06AM +0100, Eric Dumazet wrote:
And (deleted) is correctly added to deleted files.
The hell it will.
touch a
touch b
exec 5a
mv b a
ls -l /proc/$$/fd/5
With your patch and without it, please.
PS: getting rid of socket dentries is a bad idea with the capital
On Wed, Nov 01, 2006 at 08:38:11AM +, Al Viro wrote:
On Wed, Nov 01, 2006 at 09:21:06AM +0100, Eric Dumazet wrote:
And (deleted) is correctly added to deleted files.
The hell it will.
touch a
touch b
exec 5a
mv b a
ls -l /proc/$$/fd/5
With your patch and without it, please.
On Wed, 25 Oct 2006 22:15:23 -0700, Simon Barber wrote:
Re: registering as a real protocol - yes this I have been going on about
for a while. This needs a few changes in how things work:
1. Register as a real protocol.
2. Change drivers to use netif_rx to receive frames (will also be more
poll/select() notifications.
This patch includes generic poll/select notifications.
kevent_poll works simialr to epoll and has the same issues (callback
is invoked not from internal state machine of the caller, but through
process awake, a lot of allocations and so on).
Signed-off-by: Evgeniy
Generic event handling mechanism.
Consider for inclusion.
Changes from 'take21' patchset:
* minor cleanups (different return values, removed unneded variables,
whitespaces and so on)
* fixed bug in kevent removal in case when kevent being removed
is the same as overflow_kevent (spotted by
Socket notifications.
This patch includes socket send/recv/accept notifications.
Using trivial web server based on kevent and this features
instead of epoll it's performance increased more than noticebly.
More details about various benchmarks and server itself
(evserver_kevent.c) can be found
Timer notifications.
Timer notifications can be used for fine grained per-process time
management, since interval timers are very inconvenient to use,
and they are limited.
This subsystem uses high-resolution timers.
id.raw[0] is used as number of seconds
id.raw[1] is used as number of
Core files.
This patch includes core kevent files:
* userspace controlling
* kernelspace interfaces
* initialization
* notification state machines
Some bits of documentation can be found on project's homepage (and links from
there):
Hi!
Generic event handling mechanism.
Consider for inclusion.
Changes from 'take21' patchset:
We are not interrested in how many times you spammed us, nor we want
to know what was wrong in previous versions. It would be nice to have
short summary of what this is good for, instead.
On Tue, Oct 31, 2006 at 10:00:47PM -0800, David Miller wrote:
From: Andy Gospodarek [EMAIL PROTECTED]
Date: Tue, 31 Oct 2006 17:06:00 -0500
I've got a kernel built where
CONFIG_DEBUG_SPINLOCK_SLEEP=y
is in the config and I've noticed some interesting behavior when
bringing up
Hi,
Even if the total lock time can be reduced, it's possible that interrupt
handler is executed while the interrupted code is still holding the
semaphore.
I think your method only decrease the frequency of this problem.
Why does reducing the lock time solve this problem?
there are several
From: Florian Westphal [EMAIL PROTECTED]
convert sprintf(a,b) to strcpy(a,b). Make tipc_bclink_name[] const.
Signed-off-by: Florian Westphal [EMAIL PROTECTED]
---
compile tested; diffed against davem/net-2.6.
--- a/net/tipc/bcast.c
+++ b/net/tipc/bcast.c
@@ -119,7 +119,7 @@ static struct
On Wed, Nov 01, 2006 at 11:28:05AM +0100, Jiri Benc wrote:
On Wed, 25 Oct 2006 22:15:23 -0700, Simon Barber wrote:
Re: registering as a real protocol - yes this I have been going on about
for a while. This needs a few changes in how things work:
1. Register as a real protocol.
2.
On Fri, 2006-27-10 at 11:10 +0200, Daniel Lezcano wrote:
No, it uses virtualization at layer 2 and I had already mention it
before (see the first email of the thread), but thank you for the email
thread pointer.
What would be really useful is someone takes the time and creates a
matrix of
Hi,
A new version of radvd has been released. There hasn't been any very
major changes, just regular maintenance and a one new experimental
feature (RDNSS option).
Get it at: http://www.litech.org/radvd/
Changes since 0.9.1:
* Fix AdvDefaultLifetime initalization, broken in 0.9.1.
*
On Wed, Nov 01, 2006 at 02:06:14PM +0100, Pavel Machek ([EMAIL PROTECTED])
wrote:
Hi!
Generic event handling mechanism.
Consider for inclusion.
Changes from 'take21' patchset:
We are not interrested in how many times you spammed us, nor we want
to know what was wrong in
David Miller a écrit :
It turns out that while procfs uses a different mechanism, those
procfs symlinks do point to the real socket dentry, so when you
readlink() on it you do d_path() on the real socket dentry.
Al Viro just suggested a way around this to me:
1) Just mark the dentry HASHED by
On Wednesday 01 November 2006 05:34, Larry Finger wrote:
John,
I had not responded to Michael's comments as I heard from another user with
thousands of these
assertions in his logs, and I have been waiting for his sprom values and
hoped to make a single
patch. It is good, however, that
On Tue, Oct 31, 2006 at 10:34:20PM -0600, Larry Finger wrote:
John W. Linville wrote:
On Wed, Oct 18, 2006 at 04:37:08PM +0200, Michael Buesch wrote:
@@ -257,7 +263,11 @@ void bcm43xx_leds_update(struct bcm43xx_
continue;
#endif /* CONFIG_BCM43XX_DEBUG */
On Wed, 1 Nov 2006, Pavel Machek wrote:
Hi!
Generic event handling mechanism.
Consider for inclusion.
Changes from 'take21' patchset:
We are not interrested in how many times you spammed us, nor we want
to know what was wrong in previous versions. It would be nice to have
Hi!
Generic event handling mechanism.
Consider for inclusion.
Changes from 'take21' patchset:
We are not interrested in how many times you spammed us, nor we want
to know what was wrong in previous versions. It would be nice to have
short summary of what this is good
What would be really useful is someone takes the time and creates a
matrix of the differences between the implementations.
It seems there are quiet a few differences but without such comparison
(to which all agree to) it is hard to form an opinion without a document
of some form.
If Dmitry is
net/bridge/netfilter/ebtables.c: In function 'ebt_dev_check':
net/bridge/netfilter/ebtables.c:89: warning: initialization discards qualifiers
from pointer target type
So make the char* a const char * and the warning is gone.
Signed-off-by: Meelis Roos [EMAIL PROTECTED]
diff --git
Ben Greear writes:
I've completed the first pass of my changes to pktgen in 2.6.18.
Many of these features are probably DOA based on previous conversations,
but perhaps this will help someone
Thanks. Well sometimes there is a need to capture and drop pkts and various
points, so it
Simplify two calls to skb_put by removing one call to it.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_main.c |7 +++
1 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index
Enable early receives on 82573 for jumbo frame performance. Jumbo's
are only supported on 82573L with ASPM disabled.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_main.c |6 ++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git
Remove debugging code disabling MULR (multiple reads). It's not usable
for a wide audience and there are no known problems with MULR right
now.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_main.c |8
1 files changed, 0 insertions(+), 8 deletions(-)
diff
Small whitespace changes, comment changes, typo fixes.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_ethtool.c | 27 +-
drivers/net/e1000/e1000_hw.c |6 +-
drivers/net/e1000/e1000_hw.h |5 +-
drivers/net/e1000/e1000_main.c| 101
This memsetting was added in a paranoid rage debugging TX hangs, but
are no longer of importance. We can beef up the performance quite a
bit removing them. Make sure to fill in next_to_watch to allow this.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
Signed-off-by: Jesse Brandeburg [EMAIL
On ich systems during PHY power down to D3, the voltage regulators
were left on.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
Signed-off-by: Jeff Kirsher [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_hw.c |9 +
drivers/net/e1000/e1000_hw.h |1 +
2 files changed, 6 insertions(+), 4
Order pci-e capability struct according to bus/pci bus width ordering
preserving the hard pci spec numbers.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
Signed-off-by: Jeff Kirsher [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_hw.h |8 +---
1 files changed, 5 insertions(+), 3 deletions(-)
ICH8 will soon be followed by newer chipsets bearing the same acronym,
thus we remove the '8' and make it independent of the version number in
the platform name.
Signed-off-by: Jeff Kirsher [EMAIL PROTECTED]
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_hw.c| 119
IA64 SMP systems were seeing TX issues with multiple cpu's attempting
to write tail registers unordered. This mmiowb() fixes the issue.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
Signed-off-by: Jesse Brandeburg [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_main.c |3 +++
1 files changed, 3
Enable TSO for IPV6. All e1000 hardware supports it. This reduces CPU
utilizations by 50% when transmitting IPv6 frames.
Fix symbol naming enabling ipv6 TSO. Turn off TSO6 for 10/100.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000.h |3 +++
This file needs some cleanups and reordering - logically order it
so that relevant defines and code are together with properly quoted
defaults.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_param.c | 78 ++-
1 files changed, 20
Spec fix: don't set IDE unless we are actually setting the tx
int delay time.
Signed-off-by: Auke Kok [EMAIL PROTECTED]
Signed-off-by: Jesse Brandeburg [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_main.c |7 +--
1 files changed, 5 insertions(+), 2 deletions(-)
diff --git
Add a netif_wake/start_queue counter to the ethtool statistics to indicated
to the user that their transmit ring could be too small for their workload.
Signed-off-by: Jesse brandeburg [EMAIL PROTECTED]
Cc: Jamal Hadi [EMAIL PROTECTED]
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
Hi,
Here are patches for e1000 for #upstream in netdev-2.6.git. Most of the
patches are maintenance and minor fixes only, but a few are significant
and more complex:
o Patch to enable IPv6 TSO for all e1000 hardware
o Add support for 3 newly packaged existing silicon devices
o Add a MSI
Add a new dynamic itr algorithm, with 2 modes, and make it the default
operation mode. This greatly reduces latency and increases small packet
performance, at the cost of some CPU utilization. Bulk traffic
throughput is unaffected.
The driver can limit the amount of interrupts per second that
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
Documentation/networking/e1000.txt | 451 +---
1 files changed, 266 insertions(+), 185 deletions(-)
diff --git a/Documentation/networking/e1000.txt
b/Documentation/networking/e1000.txt
index 5c0a5cc..61b171c 100644
Add a generic MSI interrupt routine that is IO read-free, speeding up
MSI interrupt handling.
Signed-off-by: Jesse Brandeburg [EMAIL PROTECTED]
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_main.c | 90 +++-
1 files changed, 88
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
drivers/net/e1000/e1000_main.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index 71524b1..276eeb8 100644
--- a/drivers/net/e1000/e1000_main.c
+++
John W. Linville wrote:
On Wed, Nov 01, 2006 at 11:28:05AM +0100, Jiri Benc wrote:
On Wed, 25 Oct 2006 22:15:23 -0700, Simon Barber wrote:
Re: registering as a real protocol - yes this I have been going on about
for a while. This needs a few changes in how things work:
1. Register as a real
On Wed, Nov 01, 2006 at 05:05:51PM +0100, Pavel Machek ([EMAIL PROTECTED])
wrote:
Hi!
Hi Pavel.
Kevent is a generic subsytem which allows to handle event notifications.
It supports both level and edge triggered events. It is similar to
poll/epoll in some cases, but it is more scalable,
Hallo, Evgeniy Polyakov.
On 2006-11-01, you wrote:
[]
Quantifying how much more scalable would be nice, as would be some
example where it is useful. (It makes my webserver twice as fast on
monster 64-cpu box).
Trivial kevent web-server can handle 3960+ req/sec on Xeon 2.4Ghz with
[...]
A deadlock was found in bridge netfilter code (2.4 only), when a device is
removed.
The device removal path causes a BPDU to be generated and ends up
self deadlocking on the BR lock.
Simple fix would be to avoid generating config bpdu's
immediately when becoming root bridge, and just let the
In kernel 2.6.19, there is a configuration option that enables the _must_check
logic. This patch adds a check and removes the warning for a pci_enable_device
call in the ipw2100 driver.
Signed-off-by: Larry Finger [EMAIL PROTECTED]
---
Index: wireless-2.6/drivers/net/wireless/ipw2100.c
Kok, Auke wrote:
Add a new dynamic itr algorithm, with 2 modes, and make it the default
operation mode. This greatly reduces latency and increases small
packet performance, at the cost of some CPU utilization. Bulk
traffic throughput is unaffected.
Thanks to the generous testing of Rick Jones
Libor Klepáč [EMAIL PROTECTED] :
[...]
i have build computer with msi MS-7235 motherboard, it has r8169 chip onboard
i'm using kernel version 2.6.18 from debian (and i also rebuild kernel
without
NAPI).
When i modprobe module , dmesg says
r8169 Gigabit Ethernet driver 2.2LK loaded
PCI:
Francois Romieu [EMAIL PROTECTED] :
[...]
Please send lspci -vvx, I need the hexadecimal ID.
Actually, no. This is a 8167, 2.6.19-rc4 includes the correct code for it.
--
Ueimor
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More
Dne středa 01 listopad 2006 23:00 Francois Romieu napsal(a):
Libor Klepáč [EMAIL PROTECTED] :
[...]
i have build computer with msi MS-7235 motherboard, it has r8169 chip
onboard i'm using kernel version 2.6.18 from debian (and i also rebuild
kernel without NAPI).
When i modprobe module
Dne středa 01 listopad 2006 23:13 Francois Romieu napsal(a):
Francois Romieu [EMAIL PROTECTED] :
[...]
Please send lspci -vvx, I need the hexadecimal ID.
Actually, no. This is a 8167, 2.6.19-rc4 includes the correct code for it.
ok, thanks, i'll try this version
libor
pgpEJVd7w5u2M.pgp
On Wed, Nov 01, 2006 at 06:20:43PM +, Oleg Verych ([EMAIL PROTECTED]) wrote:
Hallo, Evgeniy Polyakov.
Hello, Oleg.
On 2006-11-01, you wrote:
[]
Quantifying how much more scalable would be nice, as would be some
example where it is useful. (It makes my webserver twice as fast on
From: walter harms [EMAIL PROTECTED]
Date: Wed, 01 Nov 2006 22:38:26 +0100
These line
+ strcpy(bcbearer-media.name, tipc-multicast);
i gues that means tipc_bclink_name ?
Why? The original code used tipc-multicast which is a
different string than tipc_bclink_name which is
James Ketrenos wrote:
If people have issues with with specific components of d80211 prior to
its merging, stand up and state what they are and how not fixing them
would negatively impact people that aren't using the d80211 subsystem.
Don't take the above as me saying there aren't items that
My desktop machine (Fedora Core 6) crashed on during shutdown when
running latest
2.6.19-rc4 kernel. Machine is UP Amd64.
It looks like a null dereference in ip_conntrack. See screenshot.
http://picasaweb.google.com/stephen.hemminger/LinuxCrashes/photo#4992593950061821970
-
To unsubscribe
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Wed, 1 Nov 2006 11:45:50 -0800
A deadlock was found in bridge netfilter code (2.4 only), when a device is
removed.
The device removal path causes a BPDU to be generated and ends up
self deadlocking on the BR lock.
Simple fix would be to
Jeff Garzik wrote:
James Ketrenos wrote:
If people have issues with with specific components of d80211 prior to
its merging, stand up and state what they are and how not fixing them
would negatively impact people that aren't using the d80211 subsystem.
Don't take the above as me saying there
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Wed, 01 Nov 2006 16:56:35 -0800
My desktop machine (Fedora Core 6) crashed on during shutdown when
running latest
2.6.19-rc4 kernel. Machine is UP Amd64.
It looks like a null dereference in ip_conntrack. See screenshot.
From: Meelis Roos [EMAIL PROTECTED]
Date: Wed, 1 Nov 2006 19:01:14 +0200 (EET)
net/bridge/netfilter/ebtables.c: In function 'ebt_dev_check':
net/bridge/netfilter/ebtables.c:89: warning: initialization discards
qualifiers from pointer target type
So make the char* a const char * and the
On 11/1/06, Evgeniy Polyakov [EMAIL PROTECTED] wrote:
On Wed, Nov 01, 2006 at 06:20:43PM +, Oleg Verych ([EMAIL PROTECTED]) wrote:
Hallo, Evgeniy Polyakov.
Hello, Oleg.
On 2006-11-01, you wrote:
[]
Quantifying how much more scalable would be nice, as would be some
example where it
On 10/31/06, Eric Dumazet [EMAIL PROTECTED] wrote:
Arnaldo Carvalho de Melo a écrit :
Hi,
I've been working on some DWARF2 utilities and one of them,
pahole (Poke-a-Hole) can be used to find holes due to alignment rules
in structs, the full output of:
[EMAIL PROTECTED] net-2.6]$
hardware: ibm thinkpad t30
kernel: 2.6.18
problem:
airo(): Probing for PCI adapters
ACPI: PCI Interrupt Link [LNKC] enabled at IRQ 11
ACPI: PCI Interrupt :02:02.0[A] - Link [LNKC] - GSI 11 (level,
low) - IRQ 11
airo(eth1): Found an MPI350 card
airo(eth1): Max tries exceeded when issueing
When the periodic work function in bcm43xx was converted for voluntary
preemption
to reduce latency, a new function was created to estimate the badness of
each step, and this quantity was used to determine if preemption should be
enabled when periodic work was undertaken. This concept was quite
On Wed, 1 Nov 2006 18:48:56 -0800
Stephen Hemminger [EMAIL PROTECTED] wrote:
On Wed, 01 Nov 2006 17:49:31 -0800 (PST)
David Miller [EMAIL PROTECTED] wrote:
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Wed, 01 Nov 2006 16:56:35 -0800
My desktop machine (Fedora Core 6) crashed on
Hi,
I am trying to run PCI based Ethernet card (RTL 8139) on a big endian
host.
The pci controller that I have does not have any mechanism for swapping
the Big Endian data to Little Endian. While running the 8139too.c driver
I see that the data generated on the line is Endian swapped.
In
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Tue, 31 Oct 2006 21:58:48 -0800
Some basic stuff is on the parent page
http://developer.osdl.org/shemminger/tcp
Are you using tcptrace to generate those plots?
If so are you using xplot or gnuplot to build those
image files?
Thanks.
-
To
Arnaldo Carvalho de Melo a écrit :
On 10/31/06, Eric Dumazet [EMAIL PROTECTED] wrote:
Arnaldo Carvalho de Melo a écrit :
Hi,
I've been working on some DWARF2 utilities and one of them,
pahole (Poke-a-Hole) can be used to find holes due to alignment rules
in structs, the full output
Jesse Huang wrote:
Dear All:
How about current IP100A, sundance.c status? Should it be put into kernel or
not?
Is there any sentence should I need to modify?
It's in my queue. We are in a bug fix-only cycle right now, so it has
been a bit lower priority, but I will queue it for 2.6.20.
On Wed, Nov 01, 2006 at 06:12:41PM -0800, Nate Diller ([EMAIL PROTECTED]) wrote:
Indesiciveness has certainly been an issue here, but I remember akpm
and Ulrich both giving concrete suggestions. I was particularly
interested in Andrew's request to explain and justify the differences
between
On Wed, Oct 25, 2006 at 07:22:08PM -0700, David Brownell wrote:
On Wednesday 25 October 2006 4:58 pm, Randy Dunlap wrote:
On Wed, 25 Oct 2006 15:27:09 -0700 David Brownell wrote:
Instead, usbnet.c should #ifdef the relevant ethtool hooks
according to CONFIG_MII ... since it's
CC [M] net/sctp/sm_make_chunk.o
net/sctp/sm_make_chunk.c: In function 'sctp_unpack_cookie':
net/sctp/sm_make_chunk.c:1358: warning: initialization discards qualifiers from
pointer target type
The reason is that sctp_unpack_cookie() takes a const struct
sctp_endpoint and modifies the digest
David Miller wrote:
From: walter harms [EMAIL PROTECTED]
Date: Wed, 01 Nov 2006 22:38:26 +0100
These line
+strcpy(bcbearer-media.name, tipc-multicast);
i gues that means tipc_bclink_name ?
mea culpa, i should not write mail when tired.
Why? The original code used
Florian Westphal wrote:
walter harms [EMAIL PROTECTED] wrote:
These line
+ strcpy(bcbearer-media.name, tipc-multicast);
i gues that means tipc_bclink_name ?
The idea was to change how things are done, not _what_ is being done.
an even more secure version could be like this:
Dear Jeff:
Thanks for your help. We are happy to see new version of
our driver will be support in the new kernel.
Best Regards,
Jesse Huang
- Original Message -
From: Jeff Garzik [EMAIL PROTECTED]
To: Jesse Huang [EMAIL PROTECTED]
Cc: linux-kernel@vger.kernel.org;
Michael Buesch wrote:
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 fire after enabling
IRQs and would lead to an oops in the DMA TXstatus handling code.
Signed-off-by:
78 matches
Mail list logo