the RX errors I'm seeing every once in a while are still there in
2.6.16rc5. Maybe they happen if there is some IO load on the IDE disks,
just upgraded openoffice.rpm
Mar 3 08:56:35 nectarine kernel: eth0: RX MAC fifo overflow smac[03810400].
Mar 3 08:56:35 nectarine kernel: eth0: RX MAC fifo ov
Hi,
while testing the rt2x00 driver, I found the following inconsistency
(don't know wether this is considered a bug).
If you already have a netdevice named wlan0, but no corresponding
wlan0.11 or wlan0ap device, you'll get the following devices while using
the dscape 802.11 stack:
wlan1, wlan0ap
> >otoh, it's likely that SLES10 and RH-whatever will end up shipping with
> >those patches as a backport, so we might as well merge it for them?
We're not awfully keen on seeing these in 2.6.16. We're deep into beta
and can only keep tracking mainline rcs right now as long as they're
bugfix only.
On Fri, 03 Mar 2006 09:54:10 +0100, Matthias Mueller wrote:
> while testing the rt2x00 driver, I found the following inconsistency
> (don't know wether this is considered a bug).
> If you already have a netdevice named wlan0, but no corresponding
> wlan0.11 or wlan0ap device, you'll get the followi
Jiri Benc wrote:
> On Fri, 03 Mar 2006 09:54:10 +0100, Matthias Mueller wrote:
>
>>while testing the rt2x00 driver, I found the following inconsistency
>>(don't know wether this is considered a bug).
>>If you already have a netdevice named wlan0, but no corresponding
>>wlan0.11 or wlan0ap device,
net/xfrm/xfrm_user.c: In function `verify_sec_ctx_len':
net/xfrm/xfrm_user.c:104: warning: comparison is always false due to limited
range of data type
if (sec_ctx->sadb_x_ctx_len > PAGE_SIZE)
PAGE_SIZE is 64k on powerpc, and that's a 16-bit variable.
-
To unsubscribe from this list: sen
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hello!
I'm using the entries in "/proc/sys/net/ipv4/conf/{all,}" to
enable or disable certain additional features of the kernel's network
code.
In "include/linux/inetdevice.h" some entries are defined with logical ANDs
and some are defined with logic
ATTN:Managing Director
I am a financial consultant based in Burkina Faso. I have a client (a widow)
she has USD15.5M with a private equity investment trust company for safe
keeping only. She wishes to invest in a stable economy.
Her interest is in companies with potentials for rapid growth in long
On 3/3/06, Herbert Xu <[EMAIL PROTECTED]> wrote:
> Stephen Hemminger <[EMAIL PROTECTED]> wrote:
> >
> > Also, you should use 2 rather than NET_IP_ALIGN here.
> > The point of NET_IP_ALIGN is to allow architectures where unaligned
> > DMA is expensive to redefine NET_IP_ALIGN to 0. But in this case
On Thursday 02 March 2006 01:21, Stephen Hemminger wrote:
> This patch has several fixes for sky2 hangs based on IRQ
> problems and oversize frames. It is very new, so really not
> ready for 2.6.16. It works form me on Yukon EC and XL.
Works for me, too.
Tested it with my rsync scenario, no hangs
Please queue the following patches for post 2.6.16 release.
01 Remove duplicate includes of linux/in.h and linux/ip.h
02 Fix misplaced parenthesis in mv643xx_eth_port_disable_rx
03 Rename "channels" to "queues"
04 Select CONFIG_MII on CONFIG_MV643XX_ETH
05 Refactor tx command queuing cod
From: Dale Farnsworth <[EMAIL PROTECTED]>
Use better terminology for HW queues. No functional changes.
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/mv643xx_eth.c | 44 ++--
1 file changed, 22 insertions(+), 22 deletions(-)
Index: linux-
From: Dale Farnsworth <[EMAIL PROTECTED]>
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
These includes were added twice:
in commit 78a5e534758349fd3effc90ce1152b55368f52ee by Olaf Hering and
in commit b6298c22c5e9f698812e2520003ee178aad50c10 by Al Viro.
This patch reverts 78a5e53475834
From: Dale Farnsworth <[EMAIL PROTECTED]>
This bug could result in a system hang.
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/mv643xx_eth.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
Index: linux-2.6-mv643xx_enet/drivers/net/mv643xx_eth.c
=
Jeff Kirsher wrote:
On 3/2/06, Jeff Garzik <[EMAIL PROTECTED]> wrote:
So, I'll ask you guys to fix up your tree to apply without conflicts to
the latest git, and then resubmit the push. You don't have to resend
all 19 patches, I've reviewed and approved those already.
Jeff
Patch 19/
>From : Dale Farnsworth <[EMAIL PROTECTED]>
Recent patches for the mv643xx_eth driver now use the MII interface
library. Select MII so it gets built when that driver is selected.
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/Kconfig |1 +
1 file changed, 1 insertion(+
From: Dale Farnsworth <[EMAIL PROTECTED]>
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/mv643xx_eth.c | 66 +---
drivers/net/mv643xx_eth.h | 47 ++---
2 files changed, 51 insertions(+), 62 deletions(-)
Index: linux-2.
From: Dale Farnsworth <[EMAIL PROTECTED]>
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/mv643xx_eth.c | 259 +++-
drivers/net/mv643xx_eth.h |4
2 files changed, 85 insertions(+), 178 deletions(-)
Index: linux-2.6-mv643xx_enet/drivers/n
From: Dale Farnsworth <[EMAIL PROTECTED]>
The task level rx queue refill feature hasn't ever worked
(at least in 2.6) and is of dubious value. Remove it.
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/mv643xx_eth.c | 59 +++-
drivers/net/m
From: Dale Farnsworth <[EMAIL PROTECTED]>
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/mv643xx_eth.c | 75
drivers/net/mv643xx_eth.h |5 --
2 files changed, 25 insertions(+), 55 deletions(-)
Index: linux-2.6-mv643xx_enet/drivers/n
From: Dale Farnsworth <[EMAIL PROTECTED]>
Simplify and remove redundant code for filling transmit descriptors.
No changes in features; it's just a code reorganization/cleanup.
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/mv643xx_eth.c | 551 +-
From: Dale Farnsworth <[EMAIL PROTECTED]>
Now that the BIT0-BIT31 defines are no longer used by mv643xx_eth.c,
remove them from mv643xx_eth.h.
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
drivers/net/mv643xx_eth.h | 152 +++-
1 file changed, 49 insert
Jeff Kirsher wrote:
+#define DEFAULT_DEBUG_LEVEL_SHIFT 3
+static int debug = DEFAULT_DEBUG_LEVEL_SHIFT;
+module_param(debug, int, 0);
+MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all)");
@@ -386,6 +394,7 @@ ixgb_probe(struct pci_dev *pdev,
adapter->netdev = netdev;
ad
Jeff Kirsher wrote:
- This patch is to improve performance by adding prefetch to the ixgb driver
- Add driver comments
Signed-off-by: Jeff Kirsher <[EMAIL PROTECTED]>
Signed-off-by: Jesse Brandeburg <[EMAIL PROTECTED]>
Signed-off-by: John Ronciak <[EMAIL PROTECTED]>
ACK patches 1, 3-9, 11
Pat
Dale Farnsworth wrote:
From: Dale Farnsworth <[EMAIL PROTECTED]>
Signed-off-by: Dale Farnsworth <[EMAIL PROTECTED]>
---
These includes were added twice:
in commit 78a5e534758349fd3effc90ce1152b55368f52ee by Olaf Hering and
in commit b6298c22c5e9f698812e2520003ee178aad50c10 by Al Viro.
This pa
On Fri, 3 Mar 2006, Jeff Garzik wrote:
Jeff Kirsher wrote:
+#define DEFAULT_DEBUG_LEVEL_SHIFT 3
+static int debug = DEFAULT_DEBUG_LEVEL_SHIFT;
+module_param(debug, int, 0);
+MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all)");
@@ -386,6 +394,7 @@ ixgb_probe(struct pci_dev *pdev,
Jesse Brandeburg wrote:
I don't think this is cheating at all. It allows us to use the DPRINTK
macro before everything is all set up to print "eth0: blah" and instead
it prints "Bus:dev:Fn: blah" which is pretty useful, IMO. If you have
any ideas about how to do this better we're open to them
The earlier round of kobject/sysfs changes to bridge caused
it not to generate a uevent on removal. Don't think any application
cares (not sure about Xen) but since it generates add uevent
it should generate remove as well.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Index: br-2.6/net/br
Initialize the STP timers for a port when it is created,
rather than when it is enabled. This will prevent future race conditions
where timer gets started before port is enabled.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Index: br-2.6/net/bridge/br_if.c
Bridge would crash because of uninitailized timer if STP is used and
device was inserted into a bridge before bridge was up. This got
introduced when the delayed port checking was added. Fix is to not
enable STP on port unless bridge is up.
Bugzilla: http://bugzilla.kernel.org/show_bug.cgi?id=614
-stable review patch. If anyone has any objections, please let us know.
--
In 295f7324ff8d9ea58b4d3ec93b1aaa1d80e048a9 I moved defer_accept from
tcp_sock to request_queue and mistakingly reset it at reqsl_queue_alloc, causing
calls to setsockopt(TCP_DEFER_ACCEPT ) to be lost after
On Mon, 2006-02-27 at 17:10 +, Christoph Hellwig wrote:
> On Sun, Feb 26, 2006 at 12:58:02AM +, Alan Cox wrote:
> > On Sad, 2006-02-25 at 08:41 +, Christoph Hellwig wrote:
> > > the regualatory problems are not true.
> >
> > They are although the binary interpretation isn't AFAIK fro
Kasper Sandberg wrote:
the binary userspace daemon protects against nothing.
In the technical realm, true. In the legal realm, potentially false.
Jeff
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo inf
Jeff Garzik <[EMAIL PROTECTED]> wrote:
>
> > git://198.78.49.142/linux-2.6.git e1000-upstream
>
> Pulled into netdev#upstream, thanks!
Did we fix wake-on-lan?
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
Andrew Morton wrote:
Jeff Garzik <[EMAIL PROTECTED]> wrote:
> git://198.78.49.142/linux-2.6.git e1000-upstream
Pulled into netdev#upstream, thanks!
Did we fix wake-on-lan?
A bit of a generic question :/ There are indeed WOL fixes in there...
Other than that, I'll wait for the Intel guys
From: Andrew Morton <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 02:52:18 -0800
> net/xfrm/xfrm_user.c: In function `verify_sec_ctx_len':
> net/xfrm/xfrm_user.c:104: warning: comparison is always false due to limited
> range of data type
>
> if (sec_ctx->sadb_x_ctx_len > PAGE_SIZE)
>
> PAG
On Tue, Feb 28, Jesse Brandeburg wrote:
> On Tue, 28 Feb 2006, Olaf Hering wrote:
>
> >On Mon, Feb 27, Jesse Brandeburg wrote:
> >
> >>>http://ozlabs.org/pipermail/linuxppc-dev/2006-February/021308.html
> >>
> >>
> >>Olaf I reproduced this problem on a 630 and this patch fixes it for me,
> >>ple
The attached patch removes a potential problem from ieee80211_wx.c, by changing the name of routine
ipw2100_translate_scan to ieee80211_translate_scan. The problem is minor as the routine is declared
static; however, if it were made global, it would pollute the namespace.
Signed-Off-By: Larry F
hello all-
i'm helping implement support for IPv6 in the Linux RPC server. i have
a couple of naive questions about creating a listening socket on a
kernel that has IPv6 support.
normally you will specify a sockaddr_in with address INADDR_ANY when
setting up an IPv4 listener.
what address
The correct way is to use "in6addr_any" (or IN6ADDR_ANY_INIT to initialize
an existing one).
Note, also, that the TCP port space can be shared (I expect that's the
default
for most cases), so you can use the IPv6 socket calls to send and receive
either IPv4 or IPv6 packets. The IPv4-mapped addres
On 3/3/06, Olaf Hering <[EMAIL PROTECTED]> wrote:
> On Tue, Feb 28, Jesse Brandeburg wrote:
>
> > On Tue, 28 Feb 2006, Olaf Hering wrote:
> >
> > >On Mon, Feb 27, Jesse Brandeburg wrote:
> > >
> > >>>http://ozlabs.org/pipermail/linuxppc-dev/2006-February/021308.html
> > >>
> > >>
> > >>Olaf I repr
This patch turns the RTNL from a semaphore to a new 2.6.16 mutex
and gets rid of some of the leftover legacy.
Please queue for 2.6.17
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
--- net-2.6.17.orig/drivers/net/8139too.c 2006-03-03 15:43:10.0
-0800
+++ net-2.6.17/drivers/n
Use the now standard setup_timer function.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Index: br-2.6/net/bridge/br_stp_timer.c
===
--- br-2.6.orig/net/bridge/br_stp_timer.c
+++ br-2.6/net/bridge/br_stp_timer.c
@@ -144,39 +1
Eliminate stub, just call directly.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
--- br-2.6.orig/net/bridge/br_input.c
+++ br-2.6/net/bridge/br_input.c
@@ -21,12 +21,6 @@
const unsigned char bridge_ula[6] = { 0x01, 0x80, 0xc2, 0x00, 0x00, 0x00 };
-static int br_pass_frame_up_finish(s
Optimize the forwarding and transmit paths. Both places are
called with bottom half/no preempt so there is no need to use
spin_lock_bh or rcu_read_lock.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
--- br-2.6.orig/net/bridge/br_fdb.c
+++ br-2.6/net/bridge/br_fdb.c
@@ -341,7 +341,6 @@ voi
Use kzalloc versus kmalloc+memset. Also don't need to do
memset() of bridge address since it is in netdev private data
that is already zero'd in alloc_netdev.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
--- br-2.6.orig/net/bridge/br_if.c
+++ br-2.6/net/bridge/br_if.c
@@ -210,7 +210,6 @@
Use kcalloc rather than kmalloc + memset.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Index: br-2.6/net/bridge/br_if.c
===
--- br-2.6.orig/net/bridge/br_if.c
+++ br-2.6/net/bridge/br_if.c
@@ -237,12 +237,11 @@ static int fin
The STP timers run off softirq (kernel timers), so there is no need to
disable bottom half in the spin locks.
Non-critical please queue for 2.6.17
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Index: br-2.6/net/bridge/br_stp_timer.c
===
Stephen Hemminger wrote:
Eliminate stub, just call directly.
Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
--- br-2.6.orig/net/bridge/br_input.c
+++ br-2.6/net/bridge/br_input.c
@@ -21,12 +21,6 @@
const unsigned char bridge_ula[6] = { 0x01, 0x80, 0xc2, 0x00, 0x00, 0x00 };
-static in
On Fri, 3 Mar 2006, Jeff Garzik wrote:
Andrew Morton wrote:
Did we fix wake-on-lan?
A bit of a generic question :/ There are indeed WOL fixes in there...
Other than that, I'll wait for the Intel guys to pop up and answer.
AFAIK the answer is unfortunately no. I think the driver may need
From: Stephen Hemminger <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 10:21:52 -0800
> Bridge would crash because of uninitailized timer if STP is used and
> device was inserted into a bridge before bridge was up. This got
> introduced when the delayed port checking was added. Fix is to not
> enable
From: Stephen Hemminger <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 10:23:29 -0800
> Initialize the STP timers for a port when it is created,
> rather than when it is enabled. This will prevent future race conditions
> where timer gets started before port is enabled.
>
> Signed-off-by: Stephen Hemm
From: Stephen Hemminger <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 10:24:14 -0800
> The earlier round of kobject/sysfs changes to bridge caused
> it not to generate a uevent on removal. Don't think any application
> cares (not sure about Xen) but since it generates add uevent
> it should generate r
From: Evgeniy Polyakov <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 10:28:01 +0300
> This one works.
Applied, thanks.
-
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: Baruch Even <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 17:34:37 +0200 (IST)
> Use functions to calculate jiffies from milliseconds and not the old, crude
> method of dividing HZ by a value. Ensures more accurate values even in the
> face
> of strange HZ values.
>
> Signed-Off-By: Baruch Ev
From: Baruch Even <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 17:35:39 +0200 (IST)
> Account for delayed-ACKs in H-TCP.
>
> Delayed-ACKs cause H-TCP to be less aggressive than its design calls for. It
> is especially true when the receiver is a Linux machine where the average
> delayed ack is over
From: Baruch Even <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 17:36:40 +0200 (IST)
> Instead of estimating the time since the last congestion event, count it
> directly.
>
> Signed-Off-By: Baruch Even <[EMAIL PROTECTED]>
> Cc: Doug Leith <[EMAIL PROTECTED]>
Also applied to net-2.6.17, thanks a lo
On Fri, 03 Mar 2006 18:59:12 -0500
Jeff Garzik <[EMAIL PROTECTED]> wrote:
> Stephen Hemminger wrote:
> > Eliminate stub, just call directly.
> >
> > Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
> >
> > --- br-2.6.orig/net/bridge/br_input.c
> > +++ br-2.6/net/bridge/br_input.c
> > @@ -21,
From: Stephen Hemminger <[EMAIL PROTECTED]>
Date: Fri, 3 Mar 2006 15:48:17 -0800
> This patch turns the RTNL from a semaphore to a new 2.6.16 mutex
> and gets rid of some of the leftover legacy.
>
> Please queue for 2.6.17
>
> Signed-off-by: Stephen Hemminger <[EMAIL PROTECTED]>
Applied, thanks
From: "chas williams - CONTRACTOR" <[EMAIL PROTECTED]>
Date: Thu, 02 Mar 2006 15:07:48 -0500
> please apply to 2.6.16 tree -- thanks!
>
> [ATM]: keep atmsvc failure messages quiet
>
> Signed-off-by: Chas Williams <[EMAIL PROTECTED]>
Applied, thanks Chas.
-
To unsubscribe from this list: send th
From: "chas williams - CONTRACTOR" <[EMAIL PROTECTED]>
Date: Thu, 02 Mar 2006 15:08:54 -0500
> please apply to 2.6.16 tree -- thanks!
>
> [ATM]: [fore200e] fix section mismatch warnings
>
> Signed-off-by: Sam Ravnborg <[EMAIL PROTECTED]>
> Signed-off-by: Chas Williams <[EMAIL PROTECTED]>
Also a
From: Luiz Fernando Capitulino <[EMAIL PROTECTED]>
Date: Wed, 1 Mar 2006 23:05:43 -0300
>
> pktgen's thread semaphores are strict mutexes, convert them to the mutex
> implementation.
>
> Signed-off-by: Luiz Capitulino <[EMAIL PROTECTED]>
Applied.
-
To unsubscribe from this list: send the line
From: Luiz Fernando Capitulino <[EMAIL PROTECTED]>
Date: Wed, 1 Mar 2006 23:05:48 -0300
>
> As suggested by Arnaldo, this patch replaces the
> thread_lock()/thread_unlock()
> by directly calls to mutex_lock()/mutex_unlock().
>
> This change makes the code a bit more readable, and the direct c
From: Luiz Fernando Capitulino <[EMAIL PROTECTED]>
Date: Wed, 1 Mar 2006 23:05:54 -0300
>
> Due to the thread's lock changes, we're at a new version now.
>
> Signed-off-by: Luiz Capitulino <[EMAIL PROTECTED]>
Applied, thanks a lot.
-
To unsubscribe from this list: send the line "unsubscribe ne
Jay Vosburgh wrote:
Originally submitted by Kenzo Iwami; his original description is:
The current bonding driver receives duplicate packets when broadcast/
multicast packets are sent by other devices or packets are flooded by the
switch. In this patch, new flags are added in priv_flags o
Here's couple of patches to the Devicescape 802.11 implementation.
Please consider applying to the dscape branch of wireless-2.6 tree.
--
Jouni MalinenPGP id EFC895FA
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a m
Replace local AES implementation (net/d80211/aes.c) with calls to
Crypto API.
Signed-off-by: Jouni Malinen <[EMAIL PROTECTED]>
Index: wireless-2.6/net/d80211/aes.c
===
--- wireless-2.6.orig/net/d80211/aes.c
+++ /dev/null
@@ -1,564 +
This file defines the current ioctl parameters and data structures. It
has been used in user space programs and including linux/types.h is
undesired when building for non-kernel code. In addition,
ieee80211_shared.h was renamed, so let's also get rid of the #include
for it in the non-kernel case.
On 3/3/06, Kumar Gala <[EMAIL PROTECTED]> wrote:
>
> How does this relate to Dan William's ADMA work?
I only became aware of Dan's ADMA work when he posted it last month,
and so far have not made any attempts to merge the I/OAT code with it.
Moving forward, combining these interfaces certainly se
From: Chris Leech <[EMAIL PROTECTED]>
Date: Fri, 03 Mar 2006 13:42:20 -0800
> +static spinlock_t dma_list_lock;
Please use DEFINE_SPINLOCK().
> +static void dma_chan_free_rcu(struct rcu_head *rcu) {
Newline before the brace please.
> +static void dma_async_device_cleanup(struct kref *kref) {
From: "Ian McDonald" <[EMAIL PROTECTED]>
Date: Thu, 2 Mar 2006 13:48:18 +1300
> If you get a chance can you push the ccid3 divide by zero fix upstream
> to Linus for 2.6.16 as it has no functionality changed and eliminates
> a nasty little bug...
> The commit for this is b6da19617f4ab610d3d90bcbdf
From: Alpt <[EMAIL PROTECTED]>
Date: Sat, 4 Mar 2006 02:58:52 +0100
> This patch defines the Netsukuku protocol (RTPROT_NTK) by the number
> 15 in rtnetlink.h.
>
> The Netsukuku daemon is using the same number to mark its routes,
> you can see it here:
> http://hinezumilabs.org/cgi-bin/viewcvs.cg
Adrian Bunk wrote:
Signed-off-by: Adrian Bunk <[EMAIL PROTECTED]>
---
This patch was already sent on:
- 4 Feb 2006
Documentation/feature-removal-schedule.txt |6 ++
drivers/net/eepro100.c |1 +
2 files changed, 7 insertions(+)
--- linux-2.6.15-mm4-full/Documen
This patch series is the first full release of the Intel(R) I/O
Acceleration Technology (I/OAT) for Linux. It includes an in kernel API
for offloading memory copies to hardware, a driver for the I/OAT DMA memcpy
engine, and changes to the TCP stack to offload copies of received
networking data to
Needed to be able to call tcp_cleanup_rbuf in tcp_input.c for I/OAT
Signed-off-by: Chris Leech <[EMAIL PROTECTED]>
---
include/net/tcp.h |2 ++
net/ipv4/tcp.c| 10 +-
2 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/include/net/tcp.h b/include/net/tcp.h
index fd7c
Provides for pinning user space pages in memory, copying to iovecs,
and copying from sk_buffs including fragmented and chained sk_buffs.
Signed-off-by: Chris Leech <[EMAIL PROTECTED]>
---
drivers/dma/Makefile |1
drivers/dma/iovlock.c | 320
Attempts to allocate per-CPU DMA channels
Signed-off-by: Chris Leech <[EMAIL PROTECTED]>
---
drivers/dma/Kconfig | 12 +
include/linux/netdevice.h |6 +++
include/net/netdma.h | 36
net/core/dev.c| 102 +
Adds an async_wait_queue and some additional fields to tcp_sock, and a
dma_cookie_t to sk_buff.
Signed-off-by: Chris Leech <[EMAIL PROTECTED]>
---
include/linux/skbuff.h |6 ++
include/linux/tcp.h| 10 ++
include/net/sock.h |2 ++
include/net/tcp.h |9
Any socket recv of less than this ammount will not be offloaded
Signed-off-by: Chris Leech <[EMAIL PROTECTED]>
---
include/linux/sysctl.h |1 +
include/net/tcp.h |1 +
net/core/user_dma.c|4
net/ipv4/sysctl_net_ipv4.c | 10 ++
4 files changed, 16 i
Provides an API for offloading memory copies to DMA devices
Signed-off-by: Chris Leech <[EMAIL PROTECTED]>
---
drivers/Kconfig |2
drivers/Makefile |1
drivers/dma/Kconfig | 13 ++
drivers/dma/Makefile |1
drivers/dma/dmaengine.c | 361 ++
Locks down user pages and sets up for DMA in tcp_recvmsg, then calls
dma_async_try_early_copy in tcp_v4_do_rcv
Signed-off-by: Chris Leech <[EMAIL PROTECTED]>
---
include/net/netdma.h |1
net/ipv4/tcp.c | 111 +-
net/ipv4/tcp_input.c |
Chris Leech wrote:
This patch series is the first full release of the Intel(R) I/O
Acceleration Technology (I/OAT) for Linux. It includes an in kernel API
for offloading memory copies to hardware, a driver for the I/OAT DMA memcpy
engine, and changes to the TCP stack to offload copies of receive
> Patch #2 didn't make it. Too big for the list?
Could be, it's the largest of the series. I've attached the gziped
patch. I can try and split this up for the future.
- Chris
02-ioatdma_driver.diff.gz
Description: GNU Zip compressed data
Chris Leech wrote:
Patch #2 didn't make it. Too big for the list?
Could be, it's the largest of the series. I've attached the gziped
patch. I can try and split this up for the future.
Well, for huge hunks of new code, it sometimes gets silly to split it up.
Once its not in a "reply to em
On Mar 3, 2006, at 3:40 PM, Chris Leech wrote:
This patch series is the first full release of the Intel(R) I/O
Acceleration Technology (I/OAT) for Linux. It includes an in
kernel API
for offloading memory copies to hardware, a driver for the I/OAT
DMA memcpy
engine, and changes to the TCP
On Fri, Mar 03, 2006 at 04:56:51AM -0800, Andrew Morton wrote:
>...
> Changes since 2.6.16-rc5-mm1:
>...
> git-netdev-all.patch
>...
> git trees
>...
Two remarks regarding the new IPW_QOS option:
- it should be named IPW2200_QOS (similar to the other IPW2200_*
options)
- please add a help tex
> On Mon, 20 Feb 2006 01:01:13 +, Martin Michlmayr <[EMAIL PROTECTED]>
> said:
tbm> Some Ethernet hardware implementations have no built-in storage
tbm> for allocated MAC values - an example is the Intel IXP420 chip
tbm> which has support for Ethernet but no defined way of storing
tbm
incorrect git path in MAINTAINERS.
--- linux-2.6.15/MAINTAINERS.orig 2006-02-28 10:26:25.0 -0500
+++ linux-2.6.15/MAINTAINERS2006-02-28 10:27:44.0 -0500
@@ -1839,7 +1839,7 @@
P: Patrick McHardy
M: [EMAIL PROTECTED]
L: netdev@vger.kernel.org
-T: git kern
The patch to ipip.c seems to be needed if xrfm_tunnel (INET_TUNNEL) is
defined as a module, or you can't have both the xfrm_tunnel module and
ipip modules loaded at the same time.
--- linux-2.6.15/net/ipv4/ipip.c.orig2006-02-28 10:46:37.0 -0500
+++ linux-2.6.15/net/ipv4/ipip.c
From: Roland Dreier <[EMAIL PROTECTED]>
Date: Wed, 01 Feb 2006 17:28:10 -0800
> Dave, if you want to merge this directly, that's fine. Or I'm fine
> with merging this through the IB tree if you'd prefer (if you want me
> to do that, let me know if you think it's 2.6.16 material).
Applied to net-
>Here's an updated version of these patches based on feedback. (The license
>did not change and continues to match that of the other Infiniband code.)
>Please consider for inclusion in 2.6.17.
This is just a ping for anymore feedback to this patch series, so that I can
respond to any requests be
I wanted to make doubly sure that this didn't get lost in the patch series, but
ip_dev_find() is re-exported. The use is shown below.
- Sean
>+int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr)
>+{
>+ struct net_device *dev;
>+ u32 ip = ((struct sockaddr_in *)
Sean> I can resubmit the patches if necessary.
Yes, can you please send out the patches again, with descriptive
subjects for each patch and with the ip_dev_find() re-export split
into its own patch? That would make it easier for me to pull into my
git tree.
Thanks,
Roland
-
To unsubscribe
93 matches
Mail list logo
