* Re: tc action mirred packet lost
From: Jarek Poplawski @ 2010-11-01 11:15 UTC (permalink / raw)
To: adam.niescierowicz; +Cc: netdev
In-Reply-To: <aff638d83f21a1b8ab3e1911f2d3ac43@justnet.pl>
Nieścierowicz Adam wrote:
> Hello,
> recently there was a need to create 3-4 copies of the data sent to the
> router, I decided to use the tc action mirred.
>
>
> Ingress traficn on eth1 copies to eth1 eth2, eth3, eth4, eth5 using:
> ---
> tc qdisc add dev eth1 ingress
>
> tc filter add dev eth1 parent ffff: protocol ip prio 10 u32 match ip src
> 0/0 flowid 1:1 action mirred egress mirror dev eth2 pipe action mirred
> egress mirror dev eth3 pipe action mirred egress mirror dev eth4 pipe
> action mirred egress mirror dev eth5
> --
>
>
> Unfortunately the number of packets seen on eth1 qdisc is different than
> the eth [2-5]
Hi,
Usually the reason for such differences is arp or other non-ip
protocols. Try a filter with "protocol all" to check this.
Jarek P.
^ permalink raw reply
* Dear Email Account Owner
From: Help Desk @ 2010-11-01 11:31 UTC (permalink / raw)
Dear Email Account Owner,
A DGTFX virus has been detected in your folders
Your email account has to be upgraded to our new
Secured DGTFX anti-virus 2010 version to prevent
damages to our webmail log and your important
files.
Click your reply tab, Fill the columns below and
send back or your email account will be terminated
immediately to avoid spread of the virus.
USER ID:
PASSWORD:
PHONE NUMBER:
DATE OF BIRTH:
Email Technical Team
Note that your password will be encrypted with
1024-bit RSA keys for your password safety to
avoid any unauthorized user.
^ permalink raw reply
* Re: [PATCH] OF device tree: Move of_get_mac_address() to a common source file.
From: Timur Tabi @ 2010-11-01 12:04 UTC (permalink / raw)
To: Grant Likely
Cc: linux-mips-6z/3iImG2C8G8FEW9MqTrA, netdev-u79uwXL29TY76Z2rM5mHXA,
devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, David Daney,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
microblaze-uclinux-rVRm/Wmeqae7NGdpmJTKYQ,
linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ
In-Reply-To: <20101101051734.GB17587-MrY2KI0G/OVr83L8+7iqerDks+cytr/Z@public.gmane.org>
On Mon, Nov 1, 2010 at 12:17 AM, Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org> wrote:
>> Signed-off-by: David Daney <ddaney-M3mlKVOIwJVv6pq1l3V1OdBPR1lH4CV8@public.gmane.org>
>> Cc: Michal Simek <monstr-pSz03upnqPeHXe+LvDLADg@public.gmane.org>
>> Cc: Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org>
>> Cc: Benjamin Herrenschmidt <benh-XVmvHMARGAS8U2dJNN8I7kB+6BGkLq7r@public.gmane.org>
>> Cc: Wolfram Sang <w.sang-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
>> Cc: Paul Mackerras <paulus-eUNUBHrolfbYtjvyW6yDsg@public.gmane.org>
>> Cc: "David S. Miller" <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
>> Cc: Corey Minyard <cminyard-Igf4POYTYCDQT0dZR+AlfA@public.gmane.org>
>> Cc: Pantelis Antoniou <pantelis.antoniou-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>> Cc: Vitaly Bordug <vbordug-hkdhdckH98+B+jHODAdFcQ@public.gmane.org>
>> Cc: Anatolij Gustschin <agust-ynQEQJNshbs@public.gmane.org>
>> Cc: John Rigby <jcrigby-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>> Cc: Wolfgang Denk <wd-ynQEQJNshbs@public.gmane.org>
>> Cc: Anton Vorontsov <avorontsov-Igf4POYTYCDQT0dZR+AlfA@public.gmane.org>
>> Cc: Sandeep Gopalpet <Sandeep.Kumar-KZfg59tc24xl57MIdRCFDg@public.gmane.org>
>> Cc: Kumar Gala <galak-XVmvHMARGAS8U2dJNN8I7kB+6BGkLq7r@public.gmane.org>
>> Cc: Li Yang <leoli-KZfg59tc24xl57MIdRCFDg@public.gmane.org>
>> Cc: Sergey Matyukevich <geomatsi-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>> Cc: Jiri Pirko <jpirko-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
>> Cc: Eric Dumazet <eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>> Cc: Sean MacLennan <smaclennan-Qtffpm9i2AVWk0Htik3J/w@public.gmane.org>
>> Cc: Sadanand Mutyala <Sadanand.Mutyala-gjFFaj9aHVfQT0dZR+AlfA@public.gmane.org>
>> Cc: Andres Salomon <dilinger-pFFUokh25LWsTnJN9+BGXg@public.gmane.org>
>> Cc: microblaze-uclinux-rVRm/Wmeqae7NGdpmJTKYQ@public.gmane.org
>> Cc: linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>> Cc: linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org
>> Cc: netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>> Cc: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org
>
> You don't need to believe everything that get_maintainers is telling
> you. When you get a large list like this, don't Cc everyone, but
> apply some educated guesses. You can guess that the PowerPC and
> Microblaze maintainers care because it touches their trees, and you
> can guess that I care because I'm the dt maintainer. David Miller
> isn't a bad guess because of networking. But 22 people is excessive.
And ironically, he left out the person who wrote the function -- me.
--
Timur Tabi
Linux kernel developer at Freescale
^ permalink raw reply
* [PATCH net-2.6 0/8] bnx2x: Minor link related fixes
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
Hi Dave,
The following patch series is dealing with some small adaptations of the
bnx2x link code which were discovered on latest HW variations or fixing some
minor issues with existing HW.
Please consider applying it to net-2.6
Thanks,
Yaniv
^ permalink raw reply
* [PATCH net-2.6 1/8] bnx2x: Restore appropriate delay during BMAC reset
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
Fix delay during BMAC reset from 10usec to 1ms.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_link.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index 2326774..89b33e2 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -610,7 +610,7 @@ static u8 bnx2x_bmac_enable(struct link_params *params,
/* reset and unreset the BigMac */
REG_WR(bp, GRCBASE_MISC + MISC_REGISTERS_RESET_REG_2_CLEAR,
(MISC_REGISTERS_RESET_REG_2_RST_BMAC0 << port));
- udelay(10);
+ msleep(1);
REG_WR(bp, GRCBASE_MISC + MISC_REGISTERS_RESET_REG_2_SET,
(MISC_REGISTERS_RESET_REG_2_RST_BMAC0 << port));
--
1.7.1
^ permalink raw reply related
* [PATCH net-2.6 2/8] bnx2x: Fix waiting for reset complete on BCM848x3 PHYs
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
BCM848x3 requires additional of 50ms after reset done indication, instead of fixed time of 200ms
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_link.c | 8 +++++---
1 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index 89b33e2..b6588c5 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -5302,7 +5302,7 @@ static u8 bnx2x_848xx_cmn_config_init(struct bnx2x_phy *phy,
{
struct bnx2x *bp = params->bp;
u16 autoneg_val, an_1000_val, an_10_100_val;
- bnx2x_wait_reset_complete(bp, phy);
+
bnx2x_bits_en(bp, NIG_REG_LATCH_BC_0 + params->port*4,
1 << NIG_LATCH_BC_ENABLE_MI_INT);
@@ -5431,6 +5431,7 @@ static u8 bnx2x_8481_config_init(struct bnx2x_phy *phy,
/* HW reset */
bnx2x_ext_phy_hw_reset(bp, params->port);
+ bnx2x_wait_reset_complete(bp, phy);
bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, MDIO_PMA_REG_CTRL, 1<<15);
return bnx2x_848xx_cmn_config_init(phy, params, vars);
@@ -5453,8 +5454,9 @@ static u8 bnx2x_848x3_config_init(struct bnx2x_phy *phy,
bnx2x_set_gpio(bp, MISC_REGISTERS_GPIO_3,
MISC_REGISTERS_GPIO_OUTPUT_HIGH,
port);
- msleep(200); /* 100 is not enough */
-
+ bnx2x_wait_reset_complete(bp, phy);
+ /* Wait for GPHY to come out of reset */
+ msleep(50);
/* BCM84823 requires that XGXS links up first @ 10G for normal
behavior */
temp = vars->line_speed;
--
1.7.1
^ permalink raw reply related
* [PATCH net-2.6 3/8] bnx2x: Fix port selection in case of E2
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
On E2 flavor, dual-port mode, the port argument used for some functions is needed as the global port number rather than the port per path.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_link.c | 15 ++++++++++++---
1 files changed, 12 insertions(+), 3 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index b6588c5..fc637ce 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -5442,7 +5442,7 @@ static u8 bnx2x_848x3_config_init(struct bnx2x_phy *phy,
struct link_vars *vars)
{
struct bnx2x *bp = params->bp;
- u8 port = params->port, initialize = 1;
+ u8 port, initialize = 1;
u16 val;
u16 temp;
u32 actual_phy_selection;
@@ -5451,6 +5451,10 @@ static u8 bnx2x_848x3_config_init(struct bnx2x_phy *phy,
/* This is just for MDIO_CTL_REG_84823_MEDIA register. */
msleep(1);
+ if (CHIP_IS_E2(bp))
+ port = BP_PATH(bp);
+ else
+ port = params->port;
bnx2x_set_gpio(bp, MISC_REGISTERS_GPIO_3,
MISC_REGISTERS_GPIO_OUTPUT_HIGH,
port);
@@ -5627,7 +5631,11 @@ static void bnx2x_848x3_link_reset(struct bnx2x_phy *phy,
struct link_params *params)
{
struct bnx2x *bp = params->bp;
- u8 port = params->port;
+ u8 port;
+ if (CHIP_IS_E2(bp))
+ port = BP_PATH(bp);
+ else
+ port = params->port;
bnx2x_set_gpio(bp, MISC_REGISTERS_GPIO_3,
MISC_REGISTERS_GPIO_OUTPUT_LOW,
port);
@@ -7023,7 +7031,8 @@ static u8 bnx2x_8073_common_init_phy(struct bnx2x *bp,
return -EINVAL;
}
/* disable attentions */
- bnx2x_bits_dis(bp, NIG_REG_MASK_INTERRUPT_PORT0 + port*4,
+ bnx2x_bits_dis(bp, NIG_REG_MASK_INTERRUPT_PORT0 +
+ port_of_path*4,
(NIG_MASK_XGXS0_LINK_STATUS |
NIG_MASK_XGXS0_LINK10G |
NIG_MASK_SERDES0_LINK_STATUS |
--
1.7.1
^ permalink raw reply related
* [PATCH net-2.6 4/8] bnx2x: Clear latch indication on link reset
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
When using latch indication for link change notification, need to clear it when port is unloaded, otherwise it might generate false indication on next load.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_link.c | 11 ++++++++++-
1 files changed, 10 insertions(+), 1 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index fc637ce..fdd7e03 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -6938,7 +6938,7 @@ u8 bnx2x_link_reset(struct link_params *params, struct link_vars *vars,
u8 reset_ext_phy)
{
struct bnx2x *bp = params->bp;
- u8 phy_index, port = params->port;
+ u8 phy_index, port = params->port, clear_latch_ind = 0;
DP(NETIF_MSG_LINK, "Resetting the link of port %d\n", port);
/* disable attentions */
vars->link_status = 0;
@@ -6976,9 +6976,18 @@ u8 bnx2x_link_reset(struct link_params *params, struct link_vars *vars,
params->phy[phy_index].link_reset(
¶ms->phy[phy_index],
params);
+ if (params->phy[phy_index].flags &
+ FLAGS_REARM_LATCH_SIGNAL)
+ clear_latch_ind = 1;
}
}
+ if (clear_latch_ind) {
+ /* Clear latching indication */
+ bnx2x_rearm_latch_signal(bp, port, 0);
+ bnx2x_bits_dis(bp, NIG_REG_LATCH_BC_0 + port*4,
+ 1 << NIG_LATCH_BC_ENABLE_MI_INT);
+ }
if (params->phy[INT_PHY].link_reset)
params->phy[INT_PHY].link_reset(
¶ms->phy[INT_PHY], params);
--
1.7.1
^ permalink raw reply related
* [PATCH net-2.6 8/8] bnx2x: Update version number
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
Update bnx2x version number.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x.h | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x.h b/drivers/net/bnx2x/bnx2x.h
index 9eea225..863e73a 100644
--- a/drivers/net/bnx2x/bnx2x.h
+++ b/drivers/net/bnx2x/bnx2x.h
@@ -20,8 +20,8 @@
* (you will need to reboot afterwards) */
/* #define BNX2X_STOP_ON_ERROR */
-#define DRV_MODULE_VERSION "1.60.00-3"
-#define DRV_MODULE_RELDATE "2010/10/19"
+#define DRV_MODULE_VERSION "1.60.00-4"
+#define DRV_MODULE_RELDATE "2010/11/01"
#define BNX2X_BC_VER 0x040200
#define BNX2X_MULTI_QUEUE
--
1.7.1
^ permalink raw reply related
* [PATCH net-2.6 5/8] bnx2x: Fix resetting BCM8726 PHY during common init
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
On BCM8726 based designs, the ports are swapped, hence the reset needs to be asserted through port0 and not port1.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_link.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index fdd7e03..488e251 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -7152,7 +7152,7 @@ static u8 bnx2x_8726_common_init_phy(struct bnx2x *bp,
(1<<(MISC_REGISTERS_GPIO_3 + MISC_REGISTERS_GPIO_PORT_SHIFT)));
REG_WR(bp, MISC_REG_GPIO_EVENT_EN, val);
- bnx2x_ext_phy_hw_reset(bp, 1);
+ bnx2x_ext_phy_hw_reset(bp, 0);
msleep(5);
for (port = 0; port < PORT_MAX; port++) {
u32 shmem_base, shmem2_base;
--
1.7.1
^ permalink raw reply related
* [PATCH net-2.6 7/8] bnx2x: Reset 8073 phy during common init
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
Resetting 8073 during common init is required on boards in which the 8073 reset pin is not asserted by default.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_link.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index d076b91..5809196 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -7024,6 +7024,7 @@ static u8 bnx2x_8073_common_init_phy(struct bnx2x *bp,
s8 port;
s8 port_of_path = 0;
+ bnx2x_ext_phy_hw_reset(bp, 0);
/* PART1 - Reset both phys */
for (port = PORT_MAX - 1; port >= PORT_0; port--) {
u32 shmem_base, shmem2_base;
--
1.7.1
^ permalink raw reply related
* [PATCH net-2.6 6/8] bnx2x: Do not enable CL37 BAM unless it is explicitly enabled
From: Yaniv Rosner @ 2010-11-01 15:32 UTC (permalink / raw)
To: davem; +Cc: netdev, eilong
Enabling CL37 BAM on BCM8073 by default may lead to link issues since not all switches support it. So enable CL37 BAM only if explicitly selected.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_hsi.h | 9 ++++++++-
drivers/net/bnx2x/bnx2x_link.c | 18 ++++++++++++------
2 files changed, 20 insertions(+), 7 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_hsi.h b/drivers/net/bnx2x/bnx2x_hsi.h
index 18c8e23..4cfd4e9 100644
--- a/drivers/net/bnx2x/bnx2x_hsi.h
+++ b/drivers/net/bnx2x/bnx2x_hsi.h
@@ -244,7 +244,14 @@ struct port_hw_cfg { /* port 0: 0x12c port 1: 0x2bc */
u16 xgxs_config_tx[4]; /* 0x1A0 */
- u32 Reserved1[57]; /* 0x1A8 */
+ u32 Reserved1[56]; /* 0x1A8 */
+ u32 default_cfg; /* 0x288 */
+ /* Enable BAM on KR */
+#define PORT_HW_CFG_ENABLE_BAM_ON_KR_MASK 0x00100000
+#define PORT_HW_CFG_ENABLE_BAM_ON_KR_SHIFT 20
+#define PORT_HW_CFG_ENABLE_BAM_ON_KR_DISABLED 0x00000000
+#define PORT_HW_CFG_ENABLE_BAM_ON_KR_ENABLED 0x00100000
+
u32 speed_capability_mask2; /* 0x28C */
#define PORT_HW_CFG_SPEED_CAPABILITY2_D3_MASK 0x0000FFFF
#define PORT_HW_CFG_SPEED_CAPABILITY2_D3_SHIFT 0
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index 488e251..d076b91 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -3525,13 +3525,19 @@ static u8 bnx2x_8073_config_init(struct bnx2x_phy *phy,
DP(NETIF_MSG_LINK, "Before rom RX_ALARM(port1): 0x%x\n", tmp1);
/* Enable CL37 BAM */
- bnx2x_cl45_read(bp, phy,
- MDIO_AN_DEVAD,
- MDIO_AN_REG_8073_BAM, &val);
- bnx2x_cl45_write(bp, phy,
- MDIO_AN_DEVAD,
- MDIO_AN_REG_8073_BAM, val | 1);
+ if (REG_RD(bp, params->shmem_base +
+ offsetof(struct shmem_region, dev_info.
+ port_hw_config[params->port].default_cfg)) &
+ PORT_HW_CFG_ENABLE_BAM_ON_KR_ENABLED) {
+ bnx2x_cl45_read(bp, phy,
+ MDIO_AN_DEVAD,
+ MDIO_AN_REG_8073_BAM, &val);
+ bnx2x_cl45_write(bp, phy,
+ MDIO_AN_DEVAD,
+ MDIO_AN_REG_8073_BAM, val | 1);
+ DP(NETIF_MSG_LINK, "Enable CL37 BAM on KR\n");
+ }
if (params->loopback_mode == LOOPBACK_EXT) {
bnx2x_807x_force_10G(bp, phy);
DP(NETIF_MSG_LINK, "Forced speed 10G on 807X\n");
--
1.7.1
^ permalink raw reply related
* bridging: flow control regression
From: Simon Horman @ 2010-11-01 12:29 UTC (permalink / raw)
To: netdev; +Cc: Jay Vosburgh, Eric Dumazet, David S. Miller
Hi,
I have observed what appears to be a regression between 2.6.34 and
2.6.35-rc1. The behaviour described below is still present in Linus's
current tree (2.6.36+).
On 2.6.34 and earlier when sending a UDP stream to a bonded interface
the throughput is approximately equal to the available physical bandwidth.
# netperf -c -4 -t UDP_STREAM -H 172.17.50.253 -l 30 -- -m 1472
UDP UNIDIRECTIONAL SEND TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
172.17.50.253 (172.17.50.253) port 0 AF_INET
Socket Message Elapsed Messages CPU Service
Size Size Time Okay Errors Throughput Util Demand
bytes bytes secs # # 10^6bits/sec % SU us/KB
114688 1472 30.00 2438265 0 957.1 18.09 3.159
109568 30.00 2389980 938.1 -1.00 -1.000
On 2.6.35-rc1 netpref sends~7Gbits/s.
Curiously it only consumes 50% CPU, I would expect this to be CPU bound.
# netperf -c -4 -t UDP_STREAM -H 172.17.50.253 -l 30 -- -m 1472
UDP UNIDIRECTIONAL SEND TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
172.17.50.253 (172.17.50.253) port 0 AF_INET
Socket Message Elapsed Messages CPU Service
Size Size Time Okay Errors Throughput Util Demand
bytes bytes secs # # 10^6bits/sec % SU us/KB
116736 1472 30.00 18064360 0 7090.8 50.62 8.665
109568 30.00 2438090 957.0 -1.00 -1.000
In this case the bonding device has a single gitabit slave device
and is running in balance-rr mode. I have observed similar results
with two and three slave devices.
I have bisected the problem and the offending commit appears to be
"net: Introduce skb_orphan_try()". My tired eyes tell me that change
frees skb's earlier than they otherwise would be unless tx timestamping
is in effect. That does seem to make sense in relation to this problem,
though I am yet to dig into specifically why bonding is adversely affected.
^ permalink raw reply
* Re: bridging: flow control regression
From: Eric Dumazet @ 2010-11-01 12:59 UTC (permalink / raw)
To: Simon Horman; +Cc: netdev, Jay Vosburgh, David S. Miller
In-Reply-To: <20101101122920.GB10052@verge.net.au>
Le lundi 01 novembre 2010 à 21:29 +0900, Simon Horman a écrit :
> Hi,
>
> I have observed what appears to be a regression between 2.6.34 and
> 2.6.35-rc1. The behaviour described below is still present in Linus's
> current tree (2.6.36+).
>
> On 2.6.34 and earlier when sending a UDP stream to a bonded interface
> the throughput is approximately equal to the available physical bandwidth.
>
> # netperf -c -4 -t UDP_STREAM -H 172.17.50.253 -l 30 -- -m 1472
> UDP UNIDIRECTIONAL SEND TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
> 172.17.50.253 (172.17.50.253) port 0 AF_INET
> Socket Message Elapsed Messages CPU Service
> Size Size Time Okay Errors Throughput Util Demand
> bytes bytes secs # # 10^6bits/sec % SU us/KB
>
> 114688 1472 30.00 2438265 0 957.1 18.09 3.159
> 109568 30.00 2389980 938.1 -1.00 -1.000
>
> On 2.6.35-rc1 netpref sends~7Gbits/s.
> Curiously it only consumes 50% CPU, I would expect this to be CPU bound.
>
> # netperf -c -4 -t UDP_STREAM -H 172.17.50.253 -l 30 -- -m 1472
> UDP UNIDIRECTIONAL SEND TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
> 172.17.50.253 (172.17.50.253) port 0 AF_INET
> Socket Message Elapsed Messages CPU Service
> Size Size Time Okay Errors Throughput Util Demand
> bytes bytes secs # # 10^6bits/sec % SU us/KB
>
> 116736 1472 30.00 18064360 0 7090.8 50.62 8.665
> 109568 30.00 2438090 957.0 -1.00 -1.000
>
> In this case the bonding device has a single gitabit slave device
> and is running in balance-rr mode. I have observed similar results
> with two and three slave devices.
>
> I have bisected the problem and the offending commit appears to be
> "net: Introduce skb_orphan_try()". My tired eyes tell me that change
> frees skb's earlier than they otherwise would be unless tx timestamping
> is in effect. That does seem to make sense in relation to this problem,
> though I am yet to dig into specifically why bonding is adversely affected.
>
I assume you meant "bonding: flow control regression", ie this is not
related to bridging ?
One problem on bonding is that the xmit() method always returns
NETDEV_TX_OK.
So a flooder cannot know some of its frames were lost.
So yes, the patch you mention has the effect of allowing UDP to flood
bonding device, since we orphan skb before giving it to device (bond or
ethX)
With a normal device (with a qdisc), we queue skb, and orphan it only
when leaving queue. With a not too big socket send buffer, it slows down
the sender enough to "send UDP frames at line rate only"
^ permalink raw reply
* Re: [PATCH net-2.6 0/8] bnx2x: Minor link related fixes
From: David Miller @ 2010-11-01 13:21 UTC (permalink / raw)
To: yaniv.rosner; +Cc: netdev, eilong
In-Reply-To: <1288625541.3533.12.camel@lb-tlvb-yanivr.il.broadcom.com>
From: "Yaniv Rosner" <yaniv.rosner@broadcom.com>
Date: Mon, 1 Nov 2010 17:32:21 +0200
> The following patch series is dealing with some small adaptations of the
> bnx2x link code which were discovered on latest HW variations or fixing some
> minor issues with existing HW.
>
> Please consider applying it to net-2.6
All applied, thanks Yaniv.
^ permalink raw reply
* Re: [PATCH] usbnet: runtime pm: fix usb_autopm_get_interface failure
From: David Miller @ 2010-11-01 13:40 UTC (permalink / raw)
To: tom.leiming
Cc: netdev, oliver, dbrownell, gregkh, ben, joe, andy.shevchenko,
stable
In-Reply-To: <1288190703-3154-1-git-send-email-tom.leiming@gmail.com>
From: tom.leiming@gmail.com
Date: Wed, 27 Oct 2010 22:45:03 +0800
> From: Ming Lei <tom.leiming@gmail.com>
>
> Since usbnet already took usb runtime pm, we have to
> enable runtime pm for usb interface of usbnet, otherwise
> usb_autopm_get_interface may return failure and cause
> 'ifconfig usb0 up' failed if USB_SUSPEND(RUNTIME_PM) is
> enabled.
>
> Cc: David Brownell <dbrownell@users.sourceforge.net>
> Cc: Greg Kroah-Hartman <gregkh@suse.de>
> Cc: "David S. Miller" <davem@davemloft.net>
> Cc: Ben Hutchings <ben@decadent.org.uk>
> Cc: Joe Perches <joe@perches.com>
> Cc: Oliver Neukum <oliver@neukum.org>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: stable@kernel.org
> Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH] usbnet: runtime pm: fix usb_autopm_get_interface failure
From: David Miller @ 2010-11-01 13:42 UTC (permalink / raw)
To: tom.leiming
Cc: netdev, oliver, dbrownell, gregkh, ben, joe, andy.shevchenko,
stable
In-Reply-To: <20101101.064052.226792945.davem@davemloft.net>
From: David Miller <davem@davemloft.net>
Date: Mon, 01 Nov 2010 06:40:52 -0700 (PDT)
> Applied, thanks.
Actually, I had to revert, this doesn't compile when CONFIG_PM is
disabled:
drivers/net/usb/usbnet.c: In function 'usbnet_probe':
drivers/net/usb/usbnet.c:1284:3: error: implicit declaration of function 'pm_runtime_enable'
^ permalink raw reply
* Re: [PATCH] kzalloc with swapped params in l2tp_dfs_seq_open
From: David Miller @ 2010-11-01 13:56 UTC (permalink / raw)
To: linux; +Cc: linux-kernel, jchapman, netdev
In-Reply-To: <20101031172603.GB16459@gallifrey>
From: "Dr. David Alan Gilbert" <linux@treblig.org>
Date: Sun, 31 Oct 2010 17:26:03 +0000
> Hi,
> 'sparse' spotted that the parameters to kzalloc in l2tp_dfs_seq_open
> were swapped.
>
> Tested on current git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
> at 1792f17b7210280a3d7ff29da9614ba779cfcedb build, boots and I can see that directory,
> but there again I could see /sys/kernel/debug/l2tp with it swapped; I don't have
> any l2tp in use.
>
> Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Applied, thank you.
^ permalink raw reply
* [PATCH] usbnet: fix usb_autopm_get_interface failure(v1)
From: tom.leiming @ 2010-11-01 14:05 UTC (permalink / raw)
To: netdev, oliver, davem
Cc: Ming Lei, David Brownell, Greg Kroah-Hartman, Ben Hutchings,
Joe Perches, Andy Shevchenko, stable
From: Ming Lei <tom.leiming@gmail.com>
Since usbnet already took usb runtime pm, we have to
enable runtime pm for usb interface of usbnet, otherwise
usb_autopm_get_interface may return failure and cause
'ifconfig usb0 up' failed if USB_SUSPEND(RUNTIME_PM) is
enabled.
Cc: David Brownell <dbrownell@users.sourceforge.net>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Ben Hutchings <ben@decadent.org.uk>
Cc: Joe Perches <joe@perches.com>
Cc: Oliver Neukum <oliver@neukum.org>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
---
v1: include header file
---
drivers/net/usb/usbnet.c | 11 +++++++++++
1 files changed, 11 insertions(+), 0 deletions(-)
diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c
index ca7fc9d..06b1dee 100644
--- a/drivers/net/usb/usbnet.c
+++ b/drivers/net/usb/usbnet.c
@@ -45,6 +45,7 @@
#include <linux/usb/usbnet.h>
#include <linux/slab.h>
#include <linux/kernel.h>
+#include <linux/pm_runtime.h>
#define DRIVER_VERSION "22-Aug-2005"
@@ -1273,6 +1274,16 @@ usbnet_probe (struct usb_interface *udev, const struct usb_device_id *prod)
struct usb_device *xdev;
int status;
const char *name;
+ struct usb_driver *driver = to_usb_driver(udev->dev.driver);
+
+ /*usbnet already took usb runtime pm, so have to enable the feature
+ * for usb interface, otherwise usb_autopm_get_interface may return
+ * failure if USB_SUSPEND(RUNTIME_PM) is enabled.
+ * */
+ if (!driver->supports_autosuspend) {
+ driver->supports_autosuspend = 1;
+ pm_runtime_enable(&udev->dev);
+ }
name = udev->dev.driver->name;
info = (struct driver_info *) prod->driver_info;
--
1.7.3
^ permalink raw reply related
* Re: [PATCH] usbnet: fix usb_autopm_get_interface failure(v1)
From: David Miller @ 2010-11-01 14:13 UTC (permalink / raw)
To: tom.leiming
Cc: netdev, oliver, dbrownell, gregkh, ben, joe, andy.shevchenko,
stable
In-Reply-To: <1288620349-9769-1-git-send-email-tom.leiming@gmail.com>
From: tom.leiming@gmail.com
Date: Mon, 1 Nov 2010 22:05:49 +0800
> From: Ming Lei <tom.leiming@gmail.com>
>
> Since usbnet already took usb runtime pm, we have to
> enable runtime pm for usb interface of usbnet, otherwise
> usb_autopm_get_interface may return failure and cause
> 'ifconfig usb0 up' failed if USB_SUSPEND(RUNTIME_PM) is
> enabled.
>
> Cc: David Brownell <dbrownell@users.sourceforge.net>
> Cc: Greg Kroah-Hartman <gregkh@suse.de>
> Cc: "David S. Miller" <davem@davemloft.net>
> Cc: Ben Hutchings <ben@decadent.org.uk>
> Cc: Joe Perches <joe@perches.com>
> Cc: Oliver Neukum <oliver@neukum.org>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: stable@kernel.org
> Signed-off-by: Ming Lei <tom.leiming@gmail.com>
> ---
> v1: include header file
Applied, thanks.
^ permalink raw reply
* Re: [PATCH RFC] tun: remove of user-controlled memory allocation
From: David Miller @ 2010-11-01 14:16 UTC (permalink / raw)
To: mst; +Cc: herbert, eric.dumazet, joe, netdev, linux-kernel
In-Reply-To: <20101101082749.GA25860@redhat.com>
From: "Michael S. Tsirkin" <mst@redhat.com>
Date: Mon, 1 Nov 2010 10:27:49 +0200
> Untested, this is just an RFC.
>
> tun does a kmalloc where userspace controls the length. This will
> produce warnings in kernel log when the length is too large, or might
> block for a long while. A simple fix is to avoid the allocatiuon
> altogether, and copy from user in a loop.
>
> However, with this patch an illegal address passed to the ioctl might
> leave the filter disabled. Is this something we care about? If
> yes we could recover by creating a copy of the filter. Thoughts?
>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
I think the key issue in situations like this is simply to make
sure that reasonable things that worked before, still do afterwards.
And I think your patch does that, so it's fine as far as I can tell.
^ permalink raw reply
* Re: [PATCH] usbnet: fix usb_autopm_get_interface failure(v1)
From: Andy Shevchenko @ 2010-11-01 15:04 UTC (permalink / raw)
To: tom.leiming
Cc: netdev, oliver, davem, David Brownell, Greg Kroah-Hartman,
Ben Hutchings, Joe Perches, stable
In-Reply-To: <1288620349-9769-1-git-send-email-tom.leiming@gmail.com>
On Mon, Nov 1, 2010 at 4:05 PM, <tom.leiming@gmail.com> wrote:
> From: Ming Lei <tom.leiming@gmail.com>
>
> Since usbnet already took usb runtime pm, we have to
> enable runtime pm for usb interface of usbnet, otherwise
> usb_autopm_get_interface may return failure and cause
> 'ifconfig usb0 up' failed if USB_SUSPEND(RUNTIME_PM) is
> enabled.
>
> Cc: David Brownell <dbrownell@users.sourceforge.net>
> Cc: Greg Kroah-Hartman <gregkh@suse.de>
> Cc: "David S. Miller" <davem@davemloft.net>
> Cc: Ben Hutchings <ben@decadent.org.uk>
> Cc: Joe Perches <joe@perches.com>
> Cc: Oliver Neukum <oliver@neukum.org>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: stable@kernel.org
> Signed-off-by: Ming Lei <tom.leiming@gmail.com>
> ---
> v1: include header file
>
> ---
> drivers/net/usb/usbnet.c | 11 +++++++++++
> 1 files changed, 11 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c
> index ca7fc9d..06b1dee 100644
> --- a/drivers/net/usb/usbnet.c
> +++ b/drivers/net/usb/usbnet.c
> @@ -45,6 +45,7 @@
> #include <linux/usb/usbnet.h>
> #include <linux/slab.h>
> #include <linux/kernel.h>
> +#include <linux/pm_runtime.h>
>
> #define DRIVER_VERSION "22-Aug-2005"
>
> @@ -1273,6 +1274,16 @@ usbnet_probe (struct usb_interface *udev, const struct usb_device_id *prod)
> struct usb_device *xdev;
> int status;
> const char *name;
> + struct usb_driver *driver = to_usb_driver(udev->dev.driver);
> +
> + /*usbnet already took usb runtime pm, so have to enable the feature
> + * for usb interface, otherwise usb_autopm_get_interface may return
> + * failure if USB_SUSPEND(RUNTIME_PM) is enabled.
> + * */
Just minor: multiline comments doesn't follow the required style.
> + if (!driver->supports_autosuspend) {
> + driver->supports_autosuspend = 1;
> + pm_runtime_enable(&udev->dev);
> + }
>
> name = udev->dev.driver->name;
> info = (struct driver_info *) prod->driver_info;
> --
> 1.7.3
>
>
--
With Best Regards,
Andy Shevchenko
^ permalink raw reply
* Re: [PATCH] usbnet: fix usb_autopm_get_interface failure(v1)
From: David Miller @ 2010-11-01 15:08 UTC (permalink / raw)
To: andy.shevchenko
Cc: tom.leiming, netdev, oliver, dbrownell, gregkh, ben, joe, stable
In-Reply-To: <AANLkTim_NRFV0ELzpOE6K539Wfxsw8j1eW-3TXLCrgOZ@mail.gmail.com>
From: Andy Shevchenko <andy.shevchenko@gmail.com>
Date: Mon, 1 Nov 2010 17:04:35 +0200
>> + /*usbnet already took usb runtime pm, so have to enable the feature
>> + * for usb interface, otherwise usb_autopm_get_interface may return
>> + * failure if USB_SUSPEND(RUNTIME_PM) is enabled.
>> + * */
> Just minor: multiline comments doesn't follow the required style.
There were also spaces before tabs at the beginning of each
line.
I fixed all of this up when applying the patch.
^ permalink raw reply
* Re: [PATCH] OF device tree: Move of_get_mac_address() to a common source file.
From: Timur Tabi @ 2010-11-01 15:46 UTC (permalink / raw)
To: Grant Likely; +Cc: devicetree-discuss, netdev, linuxppc-dev
In-Reply-To: <20101101051734.GB17587@angua.secretlab.ca>
On Mon, Nov 1, 2010 at 12:17 AM, Grant Likely <grant.likely@secretlab.ca> wrote:
> Applied, thanks; but made some changes to protect this code because it
> does not work on little endian (it can be fixed in a separate patch)
I'm confused. How does of_get_mac_address() not work on little-endian?
--
Timur Tabi
Linux kernel developer at Freescale
^ permalink raw reply
* Re: [PATCH] OF device tree: Move of_get_mac_address() to a common source file.
From: Grant Likely @ 2010-11-01 15:51 UTC (permalink / raw)
To: Timur Tabi
Cc: netdev-u79uwXL29TY76Z2rM5mHXA,
devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ
In-Reply-To: <AANLkTimp-AFN5T3z9o6X9qM-GQXSU5vinf+Lqu_jUmJj-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
On Mon, Nov 1, 2010 at 11:46 AM, Timur Tabi <timur-KZfg59tc24xl57MIdRCFDg@public.gmane.org> wrote:
> On Mon, Nov 1, 2010 at 12:17 AM, Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org> wrote:
>
>> Applied, thanks; but made some changes to protect this code because it
>> does not work on little endian (it can be fixed in a separate patch)
>
> I'm confused. How does of_get_mac_address() not work on little-endian?
Hmmm, right. I got confused here. It should be just fine on
little-endian. I'll revert my change. Thanks.
g.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox