Hi Dave,
David Miller wrote on 10/10/2007 02:13:31 AM:
Hopefully that new qdisc will just use the TX rings of the hardware
directly. They are typically large enough these days. That might avoid
some locking in this critical path.
Indeed, I also realized last night that for the default
David Miller [EMAIL PROTECTED] writes:
From: [EMAIL PROTECTED] (Eric W. Biederman)
Date: Fri, 28 Sep 2007 18:59:08 -0600
Currently we have the call path:
macvlan_open - dev_unicast_add - __dev_set_rx_mode -
__dev_set_promiscuity - ASSERT_RTNL - mutex_trylock
When mutex debugging
On Thu, Oct 11, 2007 at 12:57:31AM -0600, Eric W. Biederman wrote:
There was a practical suggestion by Herbert that ASSERT_RTNL have a
might_sleep() added. That suggestion will currently result in
ASSERT_RTNL firing unnecessarily from the macvlan_open code path.
As I've already said we
Thank you.
I was thinking about something like that when I said not find
explicit checks, but I was confused because the sk_buff code is
plenty of explicit checks and I understood those checks as a try to
make the code self-protected against wrong uses: I, sk_buff, offer
you this interface. Don´t
Herbert Xu [EMAIL PROTECTED] writes:
On Thu, Oct 11, 2007 at 12:57:31AM -0600, Eric W. Biederman wrote:
There was a practical suggestion by Herbert that ASSERT_RTNL have a
might_sleep() added. That suggestion will currently result in
ASSERT_RTNL firing unnecessarily from the macvlan_open
On Thu, Oct 11, 2007 at 02:23:35AM -0600, Eric W. Biederman wrote:
So I would object to a patch that caused the RTNL_ASSERT to not
warn about being called in an atomic context.
ASSERT_RTNL does not warn about being called in an atomic context
today!
Well it did didn't it or we wouldn't
On Friday 14 September 2007 23:42:52 David Miller wrote:
From: Joakim Koskela [EMAIL PROTECTED]
Date: Thu, 6 Sep 2007 19:00:10 +0300
This patch addresses a couple of issues related to interfamily ipsec
modes. The problem is that the structure of the routing info changes
with the family
On Wed, 2007-10-10 at 13:51 -0600, Eric W. Biederman wrote:
Yes. Netlink sockets are per-namespace and you can use the namespace
of a netlink socket to look up a netdev.
Ok, thanks. I still haven't really looked into the wireless vs. net
namespaces problem but this will probably help.
On Thu, 11 Oct 2007, TAKANO Ryousei wrote:
From: Ilpo Järvinen [EMAIL PROTECTED]
Subject: [RFC PATCH] [TCP]: Fix lost_retrans loop vs fastpath problems
Date: Tue, 9 Oct 2007 15:20:01 +0300
Thanks Ilpo! I am trying to evaluate this patch.
There's a minor problem in this 2nd patch, it's
* David Miller | 2007-10-10 16:25:28 [-0700]:
From: Sebastian Siewior [EMAIL PROTECTED]
Date: Wed, 10 Oct 2007 21:53:37 +0200
* Oliver Hartkopp | 2007-10-10 19:53:53 [+0200]:
CC [M] crypto/authenc.o
crypto/authenc.c: In function ?crypto_authenc_hash?:
crypto/authenc.c:88: warning:
Detection implemented with lost_retrans must work also when
fastpath is taken, yet most of the queue is skipped including
(very likely) those retransmitted skb's we're interested in.
This problem appeared when the hints got added, which removed
a need to always walk over the whole write queue
This addition of lost_retrans_low to tcp_sock might be
unnecessary, it's not clear how often lost_retrans worker is
executed when there wasn't work to do.
Cc: TAKANO Ryousei [EMAIL PROTECTED]
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
---
include/linux/tcp.h |2 ++
net/ipv4/tcp_input.c
Some small and simple changes. Minor testing done, until I got
one LostRetrans counted, which turned out to be hard task than
I thought (without specifically arranged scenario, just some
basic netem delay drops).
I was reading tcp_enter_loss while looking for Cedric's bug and
noticed bytes_acked adjustment is missing from FRTO side.
Since bytes_acked will only be used in tcp_cong_avoid, I think
it's safe to assume RTO would be spurious. During FRTO cwnd
will be not controlled by tcp_cong_avoid and if FRTO
This condition (plain R) can arise at least in recovery that
is triggered after tcp_undo_loss. There isn't any reason why
they should not be marked as lost, not marking makes in_flight
estimator to return too large values.
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
---
net/ipv4/tcp_input.c |
This is necessary for upcoming DSACK bugfix. Reduces sacktag
length which is not very sad thing at all... :-)
Notice that there's a need to handle out-of-mem at caller's
place.
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
---
net/ipv4/tcp_input.c | 54
It's on the way for future cutting of that function.
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
---
net/ipv4/tcp_input.c |9 +++--
1 files changed, 3 insertions(+), 6 deletions(-)
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index c827285..5004704 100644
---
Both sacked_out and fackets_out are directly known from how
parameter. Since fackets_out is accurate, there's no need for
recounting (sacked_out was previously unnecessarily counted
in the loop anyway).
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
---
net/ipv4/tcp_input.c | 26
Dave ( others?),
Here's the rebased SACKtag rewrite series rebased for you,
applies on top of the seven patch series I sent earlier
today (or on top of net-2.6 as soon as Dave picks them up
and pushes out).
Some trivial helper patches first (two first ones from
tcp-2.6), the most interesting
It is going to replace the sack fastpath hint quite soon... :-)
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
---
include/linux/tcp.h |6 --
include/net/tcp.h | 13 +
net/ipv4/tcp_input.c | 11 ++-
net/ipv4/tcp_output.c | 19 ++-
4 files
From: David S. Miller [EMAIL PROTECTED]
It is difficult to break out the inner-logic of
tcp_sacktag_write_queue() into worker functions because
so many local variables get updated in-place.
Start to overcome this by creating a structure block
of state variables that can be passed around into
From: David S. Miller [EMAIL PROTECTED]
Worker function that implements the main logic of
the inner-most loop of tcp_sacktag_write_queue().
Signed-off-by: David S. Miller [EMAIL PROTECTED]
Acked-by: Ilpo Järvinen [EMAIL PROTECTED]
---
net/ipv4/tcp_input.c | 205
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
---
include/linux/tcp.h |2 +-
net/ipv4/tcp_input.c | 75 ++---
2 files changed, 47 insertions(+), 30 deletions(-)
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 0ec6bb6..3e412f2
Key points of this patch are:
- In case new SACK information is advance only type, no skb
processing below previously discovered highest point is done
- Optimize cases below highest point too since there's no need
to always go up to highest point (which is very likely still
This is not intented to go to mainline, provided just for those
who are interested enough about the algorithm internals during
a test.
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
---
include/linux/snmp.h | 20 +++
net/ipv4/proc.c | 20 +++
- uccf should be set to NULL to not double-free memory on
subsequent calls;
- ind_hash_q and group_hash_q lists should be initialized in the
probe() function, instead of struct_init() (called by open()),
otherwise there will be an oops if ucc_geth_driver removed
prior 'ifconfig ethX up';
-
This patch adds netpoll support for the QE UCC Gigabit Ethernet
driver. The approach is very similar to the gianfar driver.
Tested using netconsole.
Signed-off-by: Anton Vorontsov [EMAIL PROTECTED]
---
drivers/net/ucc_geth.c | 19 +++
1 files changed, 19 insertions(+), 0
Lately I've got this nice badness on mdio bus removal:
Device 'e0103120:06' does not have a release() function, it is broken and must
be fixed.
[ cut here ]
Badness at drivers/base/core.c:107
NIP: c015c1a8 LR: c015c1a8 CTR: c0157488
REGS: c34bdcf0 TRAP: 0700 Not tainted
Hi Stephen,
Thank you for your comment.
{
+.ctl_name = NET_UDP_MEM,
+.procname = udp_mem,
+.data = sysctl_udp_mem,
+.maxlen = sizeof(sysctl_udp_mem),
+.mode = 0644,
+
From: Ilpo Järvinen [EMAIL PROTECTED]
Subject: Re: [RFC PATCH] [TCP]: Fix lost_retrans loop vs fastpath problems
Date: Thu, 11 Oct 2007 13:12:49 +0300 (EEST)
But, I got
a kernel panic at net_rx_action() in our experimental setting.
I am using the net-2.6.24 kernel _without_ this patch.
Roland Dreier wrote:
I also ran a test for the code in the branch of 2.6.24 and found a problem.
I see that ifconfig down doesn't return (for IPoIB interfaces) and it's
stuck in napi_disable() in the kernel (any idea why?)
For what it's worth, I took the upstream 2.6.23 git tree and
Hi,
it seems to me that irda_usb_net_open() must set self-netopen
under spinlock or disconnect() may fail to kill all URBs.
Signed-off-by: Oliver Neukum [EMAIL PROTECTED]
Regards
Oliver
--- a/drivers/net/irda/irda-usb.c 2007-10-11 17:24:35.0 +0200
Herbert Xu wrote:
On Wed, Oct 10, 2007 at 08:36:38PM -0400, Dave Jones wrote:
The e1000 changes you reference above, is this the changeset you mean?
commit 416b5d10afdc797c21c457ade3714e8f2f75edd9
Author: Auke Kok [EMAIL PROTECTED]
Date: Fri Jun 1 10:22:39 2007 -0700
e1000: disable
Herbert Xu [EMAIL PROTECTED] writes:
Well thanks to that warning we're on our way of improving the
code that triggered it in such a way that this warning will soon
go silent.
That's precisely the reason why I object to having this warning
removed. Now you have a good point that this
Sebastian Siewior wrote:
* David Miller | 2007-10-10 16:25:28 [-0700]:
From: Sebastian Siewior [EMAIL PROTECTED]
Date: Wed, 10 Oct 2007 21:53:37 +0200
* Oliver Hartkopp | 2007-10-10 19:53:53 [+0200]:
CC [M] crypto/authenc.o
crypto/authenc.c: In function
On Thu, Oct 11, 2007 at 05:45:28PM +0200, Oliver Neukum wrote:
Hi,
Hi,
You need to cc Samuel Ortiz and the IrDA mailing list if you
want any meaningful response.
it seems to me that irda_usb_net_open() must set self-netopen
under spinlock or disconnect() may fail to kill
Hi Jeff,
Kumar Gala wrote:
Begin forwarded message:
From: Jochen Friedrich [EMAIL PROTECTED]
Date: September 24, 2007 12:15:35 PM CDT
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED], Marcelo Tosatti [EMAIL PROTECTED]
Subject: [PATCH#2 3/4] [PPC] Compile fix for 8xx CPM Ehernet driver
Add #include asm/cacheflush.h for flush_dcache_range
to make the driver compile again.
CC arch/ppc/8xx_io/enet.o
arch/ppc/8xx_io/enet.c: In function 'scc_enet_start_xmit':
arch/ppc/8xx_io/enet.c:240: error: implicit declaration of function
'flush_dcache_range'
make[1]: ***
[added Andy Whitcroft, who is listed as a CHECKPATCH maintainer]
Stephen Hemminger wrote:
Minor formatting nits reported by checkpatch.pl script:
Thanks, I Will fix them.
...
WARNING: no space between function name and open parenthesis '('
#777: FILE: drivers/net/bnx2x.c:722:
+ case
On Thu, Oct 11, 2007 at 09:10:34AM -0700, Kok, Auke wrote:
Herbert Xu wrote:
On Wed, Oct 10, 2007 at 08:36:38PM -0400, Dave Jones wrote:
The e1000 changes you reference above, is this the changeset you mean?
commit 416b5d10afdc797c21c457ade3714e8f2f75edd9
Author: Auke Kok [EMAIL
Hi,
From RFC 3493, Section 5.2:
IPV6_MULTICAST_IF
Set the interface to use for outgoing multicast packets. The
argument is the index of the interface to use. If the
interface index is specified as zero, the system selects the
interface (for example,
Acked-by: David L Stevens [EMAIL PROTECTED]
Signed-off-by: Brian Haley [EMAIL PROTECTED]
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index 532425d..1334fc1 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -539,12 +539,15 @@ done:
case
Johannes Berg [EMAIL PROTECTED] writes:
On Wed, 2007-10-10 at 13:51 -0600, Eric W. Biederman wrote:
Yes. Netlink sockets are per-namespace and you can use the namespace
of a netlink socket to look up a netdev.
Ok, thanks. I still haven't really looked into the wireless vs. net
namespaces
On Wed, 2007-10-10 at 18:54 -0700, Stephen Hemminger wrote:
Please fix these...
Thanks,
All that code is going to be rewritten, as per Dave's request,
but I will make sure to test the new code with sparse.
Eliezer
-
To unsubscribe from this list: send the line unsubscribe netdev in
the
On Wed, 2007-10-10 at 17:59 -0700, David Miller wrote:
...
I was going to add this to the tree for 2.6.24 but there is simply
too much super-ugly stuff in this driver for me to do so.
We will rewrite bnx2x_hsi.h bnx2x_init.h and bnx2x_init_vlaues.h.
Does bnx2x_asm.h look OK?
Look, there
Hello all!
Posting this set of patches again, since there where no reaction last time I
sent them.
Regards,
Andreas Henriksson
Hello Stephen Hemminger and the rest of the people on the nevdev list!
I'm posting a bunch of patches for iproute2. A few I've written myself
and have a
Eric/David, the Fedora 8 RPM version 2.6.23-6.fc8 will have this if you
want to give it a shot too. It'll be at
http://people.redhat.com/davej/kernels/Fedora/f7.92/ when it's done
building in an hour or so.
Dave
Thanks, I'll give it a whirl this evening. I put a new net card in that
On Thu, Oct 11, 2007 at 08:25:32PM +0200, Andreas Henriksson wrote:
Patch from debian iproute package.
diff -urNad iproute-20060323~/ip/iplink.c iproute-20060323/ip/iplink.c
--- iproute-20060323~/ip/iplink.c 2006-03-22 00:57:50.0 +0100
+++ iproute-20060323/ip/iplink.c
On Thu, 11 Oct 2007 21:51:14 +0900
Satoshi OSHIMA [EMAIL PROTECTED] wrote:
Hi Stephen,
Thank you for your comment.
{
+ .ctl_name = NET_UDP_MEM,
+ .procname = udp_mem,
+ .data = sysctl_udp_mem,
+ .maxlen =
On tor, 2007-10-11 at 15:25 -0400, Lennart Sorensen wrote:
On Thu, Oct 11, 2007 at 08:25:32PM +0200, Andreas Henriksson wrote:
Patch from debian iproute package.
diff -urNad iproute-20060323~/ip/iplink.c iproute-20060323/ip/iplink.c
--- iproute-20060323~/ip/iplink.c 2006-03-22
It happens only when ib interfaces are slaves of a bonding device.
I thought before that the stuck is in napi_disable() but it's almost right.
I put prints before and after call to napi_disable and see that it is called
twice.
I'll try to investigate in this direction.
ib0:
On Thu, 11 Oct 2007 20:25:32 +0200
Andreas Henriksson [EMAIL PROTECTED] wrote:
Hello all!
Posting this set of patches again, since there where no reaction last time I
sent them.
Regards,
Andreas Henriksson
Since there are so many. Please send them as git separate emails to me and
Andreas Henriksson wrote:
Abort flush after 10 seconds.
This should be optional IMO.
Add references to lartc, also drop bogus reference to tc-filters
There's a bad habit of reporting bugs on lartc that belong to netdev,
please also make it perfectly clear that lartc is not for bug-reports.
On tor, 2007-10-11 at 13:11 -0700, Stephen Hemminger wrote:
Since there are so many. Please send them as git separate emails to me and
make sure they are in the proper format for importing into git. If you really
want to make it easy, just put them in git yourself and then use 'git
Andreas Henriksson wrote:
Add mpath to ip manpage.
From: Norbert Buchmuller [EMAIL PROTECTED]
The 'mpath' parameter of 'ip route' is not documented in the manual page
nor in ip-cref.tex.
...huge part of the text in the patch was taken from the net/ipv4/Kconfig
file in the Linux kernel
On tor, 2007-10-11 at 22:30 +0200, Patrick McHardy wrote:
Andreas Henriksson wrote:
Abort flush after 10 seconds.
This should be optional IMO.
Copying Alexander Wirt here again, as I think this is his patch.
I'll skip this patch in my next resend...
Add references to lartc, also drop
With the existence of the mangle table, how useful is the filter table?
Other than requiring the REJECT target to be ported to the mangle table, is
the filter table faster than the mangle table?
If not, then shouldn't the filter table be obsoleted to avoid confusion?
Thanks!
--
Al
-
To
From: David Stevens [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 10:49:14 -0700
Acked-by: David L Stevens [EMAIL PROTECTED]
Applied, thanks everyone!
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
From: Eliezer Tamir [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 19:53:39 +0200
Will it be OK if we replace it with 4-5 headers so we can maintain them
better manually. (I'm asking what is the upper bound to the number of
header files you would consider reasonable for this driver.)
It's not the
From: Andreas Henriksson [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 23:22:12 +0200
Copying Alexander Wirt here again, as I think this is his patch.
It is important that whoever is the package maintainer for an
upstream piece of code in any distribution:
1) Completely understands the patches he
Roland Dreier [EMAIL PROTECTED] wrote:
[...]
Yes, two napi_disable()s in a row without a matching napi_enable()
will deadlock. I guess the question is why the ipoib interface is
being stopped twice.
If you just take the net-2.6.24 tree (without bonding patches), does
bonding for ethernet
On tor, 2007-10-11 at 14:48 -0700, David Miller wrote:
From: Andreas Henriksson [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 23:22:12 +0200
Copying Alexander Wirt here again, as I think this is his patch.
It is important that whoever is the package maintainer for an
upstream piece of code in
From: Andreas Henriksson [EMAIL PROTECTED]
Date: Fri, 12 Oct 2007 00:35:07 +0200
If you want me to STFU please just say so and I'll give you all my
sincere appologies and go back to the status quo of no fixes shared
with others.
I applaud your efforts, it's not what the problem is.
I just
I wanted to verify that the size of a multicast UDP message received
with recvmsg matches the size of the message the kernel thinks the
message is.
So I went about using the FIONREAD ioctl as follows:
res = ioctl (fd, FIONCREAD, value);
assert (res != -1);
bytes_received = recvmsg (fd, msg_recv,
On Thu, Oct 11, 2007 at 09:10:34AM -0700, Kok, Auke wrote:
commit 416b5d10afdc797c21c457ade3714e8f2f75edd9
Author: Auke Kok [EMAIL PROTECTED]
Date: Fri Jun 1 10:22:39 2007 -0700
e1000: disable polling before registering netdevice
this patch actually called napi_disable() in
I'm having an issue with TSO right vs. hardware that can't take the
maximum segment size sent from the stack. I've been told that the
maximum packet size that can be sent to the hardware today is 64k, but
my hardware can only take 32k in certain modes per queue due to hardware
limitations. I
From: Steven Dake [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 16:08:15 -0700
I wanted to verify that the size of a multicast UDP message received
with recvmsg matches the size of the message the kernel thinks the
message is.
So I went about using the FIONREAD ioctl as follows:
res = ioctl
Questions of this sort should normally be directed to linux-net mailing
list.
From the code you quoted, I see at least one case where it
will fail -- when the allocated buffer you pass to recvmsg is smaller than
value (ie. the datagram is too big for the read buffer).
If that's not the problem,
From: Waskiewicz Jr, Peter P [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 16:27:14 -0700
I'm having an issue with TSO right vs. hardware that can't take the
maximum segment size sent from the stack. I've been told that the
maximum packet size that can be sent to the hardware today is 64k, but
my
From: TAKANO Ryousei [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 22:51:46 +0900 (JST)
Modules linked in: 8021q tcp_bic netconsole evdev joydev sg st sr_mod
ohci_hcd i2c_amd756 i2c_amd8111 i2c_core ipv6 tg3 usbhid usbcore ff_memless
dm_mod ext3 jbd sata_sil libata sd_mod scsi_mod
...
RIP
Hello,
I have developed a small patch for the TCP code in 2.6.19 and it works
flawlessly. A couple of days ago I decided to make it compatible with
2.6.22.5 and have stumbled upon a problem I cannot solve.
In 2.6.19 it seems that all packets (at least the ones my patch work
with) are
On Thu, 11 Oct 2007 14:48:43 -0700 (PDT)
David Miller [EMAIL PROTECTED] wrote:
From: Andreas Henriksson [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 23:22:12 +0200
Copying Alexander Wirt here again, as I think this is his patch.
It is important that whoever is the package maintainer for an
David Miller wrote:
From: Waskiewicz Jr, Peter P [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 16:27:14 -0700
I'm having an issue with TSO right vs. hardware that can't take the
maximum segment size sent from the stack. I've been told that the
maximum packet size that can be sent to the hardware
On Thu, 11 Oct 2007 16:48:27 -0700 (PDT)
David Miller [EMAIL PROTECTED] wrote:
From: TAKANO Ryousei [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 22:51:46 +0900 (JST)
Modules linked in: 8021q tcp_bic netconsole evdev joydev sg st sr_mod
ohci_hcd i2c_amd756 i2c_amd8111 i2c_core ipv6 tg3 usbhid
From: Kristian Evensen [EMAIL PROTECTED]
Date: Fri, 12 Oct 2007 00:54:37 +0200
I have developed a small patch for the TCP code in 2.6.19 and it works
flawlessly. A couple of days ago I decided to make it compatible with
2.6.22.5 and have stumbled upon a problem I cannot solve.
In 2.6.19
From: Rick Jones [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 16:50:46 -0700
For just messing about, might it be possible to tweak the socket buffer sizes
and tcp_tso_win_divisor to kludge things for a short while? Couldn't ship
that
way certainly, but assuming Peter's going to get his broken
From: Rick Jones [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 16:50:46 -0700
For just messing about, might it be possible to tweak the socket
buffer sizes and tcp_tso_win_divisor to kludge things for a short
while? Couldn't ship that way certainly, but assuming
Peter's going
to get
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 16:55:48 -0700
On Thu, 11 Oct 2007 16:48:27 -0700 (PDT)
David Miller [EMAIL PROTECTED] wrote:
Alternatively we could loop in tg3_poll() until either budget
is exhausted or tg3_has_work() returns false. Actually, this sounds
Waskiewicz Jr, Peter P wrote:
From: Rick Jones [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 16:50:46 -0700
For just messing about, might it be possible to tweak the socket
buffer sizes and tcp_tso_win_divisor to kludge things for a short
while? Couldn't ship that way certainly, but assuming
From: Ben Greear [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 17:17:48 -0700
Was the auto-tuning put in after 2.6.20? If not, has this
been tested through a higher latency link? Or, am I confused
and you are talking about some other setsockopt?
It's been there for quite some time.
-
To
From: [EMAIL PROTECTED] (Eric W. Biederman)
Date: Thu, 11 Oct 2007 10:33:39 -0600
Having ASSERT_RTNL warn if you were sleeping does not seem
intuitive from the name.
This instance of convoluted locking seems like a complete
one off to me, and if it will warn about other constructs
From: Ilpo_Järvinen [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 14:41:01 +0300
I was reading tcp_enter_loss while looking for Cedric's bug and
noticed bytes_acked adjustment is missing from FRTO side.
Since bytes_acked will only be used in tcp_cong_avoid, I think
it's safe to assume RTO would
On Thu, 11 Oct 2007 17:17:43 -0700 (PDT)
David Miller [EMAIL PROTECTED] wrote:
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 16:55:48 -0700
On Thu, 11 Oct 2007 16:48:27 -0700 (PDT)
David Miller [EMAIL PROTECTED] wrote:
Alternatively we could loop in tg3_poll()
From: Ilpo_Järvinen [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 14:41:02 +0300
This condition (plain R) can arise at least in recovery that
is triggered after tcp_undo_loss. There isn't any reason why
they should not be marked as lost, not marking makes in_flight
estimator to return too large
From: Ilpo_Järvinen [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 14:41:03 +0300
It's on the way for future cutting of that function.
Signed-off-by: Ilpo Järvinen [EMAIL PROTECTED]
Applied.
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL
From: Ilpo_Järvinen [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 14:41:04 +0300
This is necessary for upcoming DSACK bugfix. Reduces sacktag
length which is not very sad thing at all... :-)
Notice that there's a need to handle out-of-mem at caller's
place.
Signed-off-by: Ilpo Järvinen [EMAIL
From: Ilpo_Järvinen [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 14:41:05 +0300
Both sacked_out and fackets_out are directly known from how
parameter. Since fackets_out is accurate, there's no need for
recounting (sacked_out was previously unnecessarily counted
in the loop anyway).
From: Ilpo_Järvinen [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 14:41:06 +0300
Detection implemented with lost_retrans must work also when
fastpath is taken, yet most of the queue is skipped including
(very likely) those retransmitted skb's we're interested in.
This problem appeared when the
From: Ilpo_Järvinen [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 14:41:07 +0300
This addition of lost_retrans_low to tcp_sock might be
unnecessary, it's not clear how often lost_retrans worker is
executed when there wasn't work to do.
Cc: TAKANO Ryousei [EMAIL PROTECTED]
Signed-off-by: Ilpo
From: Ilpo_Järvinen [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 14:41:00 +0300
Dave, please apply at will.
Done.
-
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: Stephen Hemminger [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 17:31:49 -0700
You don't need to re-read the status register and process the PHY irq's
inside loop.
Try this:
Are you sure? What if a PHY interrupt comes in during the loop?
I'm just preserving the semantics of the driver
On Thu, 11 Oct 2007 17:40:26 -0700 (PDT)
David Miller [EMAIL PROTECTED] wrote:
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 17:31:49 -0700
You don't need to re-read the status register and process the PHY irq's
inside loop.
Try this:
Are you sure? What if a PHY
On my onboard SunBlade1500 5703 chips, which lack firmware, the APE
changes introduce bus timeouts while bringing the chip up,
specifically I get crashes in tg3_write_sig_post_reset() such as:
[ 36.066603] eth1: Tigon3 [partno(BCM95703A30U) rev 1002 PHY(5703)]
(PCI:33MHz:64-bit)
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 17:50:59 -0700
On Thu, 11 Oct 2007 17:40:26 -0700 (PDT)
David Miller [EMAIL PROTECTED] wrote:
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 17:31:49 -0700
You don't need to re-read the status
From: David Miller [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 18:00:31 -0700 (PDT)
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Thu, 11 Oct 2007 17:50:59 -0700
On Thu, 11 Oct 2007 17:40:26 -0700 (PDT)
David Miller [EMAIL PROTECTED] wrote:
From: Stephen Hemminger [EMAIL PROTECTED]
Here is what I'm checking into net-2.6 for now:
commit 6f535763165331bb91277d7519b507fed22034e5
Author: David S. Miller [EMAIL PROTECTED]
Date: Thu Oct 11 18:08:29 2007 -0700
[NET]: Fix NAPI completion handling in some drivers.
In order for the list handling in net_rx_action() to
Handle the corner case where budget is exhausted correctly.
Signed-off-by: Stephen Hemminger [EMAIL PROTECTED]
--- a/drivers/net/sky2.c2007-10-11 18:11:44.0 -0700
+++ b/drivers/net/sky2.c2007-10-11 18:12:50.0 -0700
@@ -2245,15 +2245,13 @@ static inline void
Make sure PCI register for PHY power gets set correctly.
Signed-off-by: Stephen Hemminger [EMAIL PROTECTED]
--- a/drivers/net/sky2.c2007-10-11 18:12:56.0 -0700
+++ b/drivers/net/sky2.c2007-10-11 18:13:00.0 -0700
@@ -606,20 +606,19 @@ static void
Includes:
* fix for new NAPI status loop
* use internal net_stats
* fixes for fiber PHY power
--
Stephen Hemminger [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
Put initialization in sequential order (same as other constants).
Signed-off-by: Stephen Hemminger [EMAIL PROTECTED]
--- a/drivers/net/sky2.c2007-10-11 18:13:00.0 -0700
+++ b/drivers/net/sky2.c2007-10-11 18:13:01.0 -0700
@@ -296,10 +296,10 @@ static const u16
1 - 100 of 130 matches
Mail list logo
