* Re: [PATCH v3] Fix SJA1000 command register writes on SMP systems
From: David Miller @ 2010-05-18 21:03 UTC (permalink / raw)
To: socketcan; +Cc: wg, netdev, socketcan-core
In-Reply-To: <4BF2E512.1090109@hartkopp.net>
From: Oliver Hartkopp <socketcan@hartkopp.net>
Date: Tue, 18 May 2010 21:05:54 +0200
> The SJA1000 command register is concurrently written in the rx-path to free
> the receive buffer _and_ in the tx-path to start the transmission.
>
> The SJA1000 data sheet, 6.4.4 COMMAND REGISTER (CMR) states:
> "Between two commands at least one internal clock cycle is needed in
> order to proceed. The internal clock is half of the external oscillator
> frequency."
>
> On SMP systems the current implementation leads to a write stall in the
> tx-path, which can be solved by adding some general locking and some time
> to settle the write_reg() operation for the command register.
>
> Thanks to Klaus Hitschler for the original fix and detailed problem
> description.
>
> This patch applies on net-2.6 and (with some offsets) on net-next-2.6 .
>
> Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
> Acked-by: Wolfgang Grandegger <wg@grandegger.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH] bfin_mac: fix memleak in mii_bus{probe|remove}
From: David Miller @ 2010-05-18 21:13 UTC (permalink / raw)
To: kirjanov; +Cc: sonic.adi, michael.hennerich, cooloney, uclinux-dist-devel,
netdev
In-Reply-To: <20100518113445.GA28386@coldcone>
From: Denis Kirjanov <kirjanov@gmail.com>
Date: Tue, 18 May 2010 15:34:46 +0400
> [PATCH] bfin_mac: fix memleak in mii_bus_{probe|remove}
> Fix memory leak with miibus->irq
>
> Signed-off-by: Denis Kirjanov <dkirjanov@kernel.org>
I've already applied the original patch.
Therefore, it makes no sense to send a fresh complete version
of the original patch to fix this.
You'll need to send a relative patch against the other one to
fix the problem.
^ permalink raw reply
* Re: [PATCH] bfin_mac: fix memleak in mii_bus{probe|remove}
From: David Miller @ 2010-05-18 21:15 UTC (permalink / raw)
To: kirjanov; +Cc: sonic.adi, michael.hennerich, cooloney, uclinux-dist-devel,
netdev
In-Reply-To: <20100518.141318.45114681.davem@davemloft.net>
From: David Miller <davem@davemloft.net>
Date: Tue, 18 May 2010 14:13:18 -0700 (PDT)
> From: Denis Kirjanov <kirjanov@gmail.com>
> Date: Tue, 18 May 2010 15:34:46 +0400
>
>> [PATCH] bfin_mac: fix memleak in mii_bus_{probe|remove}
>> Fix memory leak with miibus->irq
>>
>> Signed-off-by: Denis Kirjanov <dkirjanov@kernel.org>
>
> I've already applied the original patch.
Sorry, ignore this, I didn't see that this is a new patch.
^ permalink raw reply
* Re: [PATCH v3] Fix SJA1000 command register writes on SMP systems
From: Sam Ravnborg @ 2010-05-18 21:31 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: David Miller, SocketCAN Core Mailing List, Linux Netdev List,
Wolfgang Grandegger
In-Reply-To: <4BF2E512.1090109@hartkopp.net>
Hi Oliver.
> diff --git a/drivers/net/can/sja1000/sja1000.h b/drivers/net/can/sja1000/sja1000.h
> index 97a622b..de8e778 100644
> --- a/drivers/net/can/sja1000/sja1000.h
> +++ b/drivers/net/can/sja1000/sja1000.h
> @@ -167,6 +167,7 @@ struct sja1000_priv {
>
> void __iomem *reg_base; /* ioremap'ed address to registers */
> unsigned long irq_flags; /* for request_irq() */
> + spinlock_t cmdreg_lock; /* lock for concurrent cmd register writes */
>
> u16 flags; /* custom mode flags */
> u8 ocr; /* output control register */
You define your spinlock inside a struct so you cannot use
DEFINE_SPINLOCK().
But then you need to use spin_lock_init() - which I fail to see
you are doing in your patch.
Sam
^ permalink raw reply
* [PATCH 1/2] cnic: Convert cnic_local_flags to atomic ops.
From: Michael Chan @ 2010-05-18 21:32 UTC (permalink / raw)
To: davem; +Cc: netdev, Michael Chan
It is easier to expand the flags for other purposes because it does
not require a spin_lock. The next bug fix patch needs a flag in
cnic_local_flags.
Signed-off-by: Michael Chan <mchan@broadcom.com>
---
drivers/net/cnic.c | 6 +++---
drivers/net/cnic.h | 5 ++---
2 files changed, 5 insertions(+), 6 deletions(-)
diff --git a/drivers/net/cnic.c b/drivers/net/cnic.c
index 4b08b18..7c6d325 100644
--- a/drivers/net/cnic.c
+++ b/drivers/net/cnic.c
@@ -1143,12 +1143,12 @@ static int cnic_submit_bnx2_kwqes(struct cnic_dev *dev, struct kwqe *wqes[],
spin_lock_bh(&cp->cnic_ulp_lock);
if (num_wqes > cnic_kwq_avail(cp) &&
- !(cp->cnic_local_flags & CNIC_LCL_FL_KWQ_INIT)) {
+ !test_bit(CNIC_LCL_FL_KWQ_INIT, &cp->cnic_local_flags)) {
spin_unlock_bh(&cp->cnic_ulp_lock);
return -EAGAIN;
}
- cp->cnic_local_flags &= ~CNIC_LCL_FL_KWQ_INIT;
+ clear_bit(CNIC_LCL_FL_KWQ_INIT, &cp->cnic_local_flags);
prod = cp->kwq_prod_idx;
sw_prod = prod & MAX_KWQ_IDX;
@@ -3690,7 +3690,7 @@ static int cnic_start_bnx2_hw(struct cnic_dev *dev)
cp->max_kwq_idx = MAX_KWQ_IDX;
cp->kwq_prod_idx = 0;
cp->kwq_con_idx = 0;
- cp->cnic_local_flags |= CNIC_LCL_FL_KWQ_INIT;
+ set_bit(CNIC_LCL_FL_KWQ_INIT, &cp->cnic_local_flags);
if (CHIP_NUM(cp) == CHIP_NUM_5706 || CHIP_NUM(cp) == CHIP_NUM_5708)
cp->kwq_con_idx_ptr = &sblk->status_rx_quick_consumer_index15;
diff --git a/drivers/net/cnic.h b/drivers/net/cnic.h
index a0d853d..4422497 100644
--- a/drivers/net/cnic.h
+++ b/drivers/net/cnic.h
@@ -179,9 +179,8 @@ struct cnic_local {
#define ULP_F_CALL_PENDING 2
struct cnic_ulp_ops *ulp_ops[MAX_CNIC_ULP_TYPE];
- /* protected by ulp_lock */
- u32 cnic_local_flags;
-#define CNIC_LCL_FL_KWQ_INIT 0x00000001
+ unsigned long cnic_local_flags;
+#define CNIC_LCL_FL_KWQ_INIT 0x0
struct cnic_dev *dev;
--
1.6.4.GIT
^ permalink raw reply related
* [PATCH 2/2] cnic: Return SPQ credit to bnx2x after ring setup and shutdown.
From: Michael Chan @ 2010-05-18 21:32 UTC (permalink / raw)
To: davem; +Cc: netdev, Michael Chan
In-Reply-To: <1274218373-16324-1-git-send-email-mchan@broadcom.com>
Everytime the iSCSI ring finishes setup or shutdown, we need to return
the SPQ (slow path queue) credit to the bnx2x driver. Without this step,
the SPQ will eventually be full causing iSCSI to fail. This can happen
after 3 or 4 MTU changes for example.
Add code to wait for these slow path commands to complete in the RX ring
and return the SPQ credit to bnx2x.
Signed-off-by: Michael Chan <mchan@broadcom.com>
---
drivers/net/cnic.c | 71 +++++++++++++++++++++++++++++++++++++++++++++++++++-
drivers/net/cnic.h | 5 +++
2 files changed, 75 insertions(+), 1 deletions(-)
diff --git a/drivers/net/cnic.c b/drivers/net/cnic.c
index 7c6d325..be90d35 100644
--- a/drivers/net/cnic.c
+++ b/drivers/net/cnic.c
@@ -2145,17 +2145,56 @@ static int cnic_get_kcqes(struct cnic_dev *dev, u16 hw_prod, u16 *sw_prod)
return last_cnt;
}
+static int cnic_l2_completion(struct cnic_local *cp)
+{
+ u16 hw_cons, sw_cons;
+ union eth_rx_cqe *cqe, *cqe_ring = (union eth_rx_cqe *)
+ (cp->l2_ring + (2 * BCM_PAGE_SIZE));
+ u32 cmd;
+ int comp = 0;
+
+ if (!test_bit(CNIC_F_BNX2X_CLASS, &cp->dev->flags))
+ return 0;
+
+ hw_cons = *cp->rx_cons_ptr;
+ if ((hw_cons & BNX2X_MAX_RCQ_DESC_CNT) == BNX2X_MAX_RCQ_DESC_CNT)
+ hw_cons++;
+
+ sw_cons = cp->rx_cons;
+ while (sw_cons != hw_cons) {
+ u8 cqe_fp_flags;
+
+ cqe = &cqe_ring[sw_cons & BNX2X_MAX_RCQ_DESC_CNT];
+ cqe_fp_flags = cqe->fast_path_cqe.type_error_flags;
+ if (cqe_fp_flags & ETH_FAST_PATH_RX_CQE_TYPE) {
+ cmd = le32_to_cpu(cqe->ramrod_cqe.conn_and_cmd_data);
+ cmd >>= COMMON_RAMROD_ETH_RX_CQE_CMD_ID_SHIFT;
+ if (cmd == RAMROD_CMD_ID_ETH_CLIENT_SETUP ||
+ cmd == RAMROD_CMD_ID_ETH_HALT)
+ comp++;
+ }
+ sw_cons = BNX2X_NEXT_RCQE(sw_cons);
+ }
+ return comp;
+}
+
static void cnic_chk_pkt_rings(struct cnic_local *cp)
{
u16 rx_cons = *cp->rx_cons_ptr;
u16 tx_cons = *cp->tx_cons_ptr;
+ int comp = 0;
if (cp->tx_cons != tx_cons || cp->rx_cons != rx_cons) {
+ if (test_bit(CNIC_LCL_FL_L2_WAIT, &cp->cnic_local_flags))
+ comp = cnic_l2_completion(cp);
+
cp->tx_cons = tx_cons;
cp->rx_cons = rx_cons;
uio_event_notify(cp->cnic_uinfo);
}
+ if (comp)
+ clear_bit(CNIC_LCL_FL_L2_WAIT, &cp->cnic_local_flags);
}
static int cnic_service_bnx2(void *data, void *status_blk)
@@ -4168,6 +4207,8 @@ static void cnic_init_rings(struct cnic_dev *dev)
for (i = 0; i < sizeof(struct ustorm_eth_rx_producers) / 4; i++)
CNIC_WR(dev, off + i * 4, ((u32 *) &rx_prods)[i]);
+ set_bit(CNIC_LCL_FL_L2_WAIT, &cp->cnic_local_flags);
+
cnic_init_bnx2x_tx_ring(dev);
cnic_init_bnx2x_rx_ring(dev);
@@ -4175,6 +4216,15 @@ static void cnic_init_rings(struct cnic_dev *dev)
l5_data.phy_address.hi = 0;
cnic_submit_kwqe_16(dev, RAMROD_CMD_ID_ETH_CLIENT_SETUP,
BNX2X_ISCSI_L2_CID, ETH_CONNECTION_TYPE, &l5_data);
+ i = 0;
+ while (test_bit(CNIC_LCL_FL_L2_WAIT, &cp->cnic_local_flags) &&
+ ++i < 10)
+ msleep(1);
+
+ if (test_bit(CNIC_LCL_FL_L2_WAIT, &cp->cnic_local_flags))
+ netdev_err(dev->netdev,
+ "iSCSI CLIENT_SETUP did not complete\n");
+ cnic_kwq_completion(dev, 1);
cnic_ring_ctl(dev, BNX2X_ISCSI_L2_CID, cli, 1);
}
}
@@ -4187,14 +4237,25 @@ static void cnic_shutdown_rings(struct cnic_dev *dev)
struct cnic_local *cp = dev->cnic_priv;
u32 cli = BNX2X_ISCSI_CL_ID(CNIC_E1HVN(cp));
union l5cm_specific_data l5_data;
+ int i;
cnic_ring_ctl(dev, BNX2X_ISCSI_L2_CID, cli, 0);
+ set_bit(CNIC_LCL_FL_L2_WAIT, &cp->cnic_local_flags);
+
l5_data.phy_address.lo = cli;
l5_data.phy_address.hi = 0;
cnic_submit_kwqe_16(dev, RAMROD_CMD_ID_ETH_HALT,
BNX2X_ISCSI_L2_CID, ETH_CONNECTION_TYPE, &l5_data);
- msleep(10);
+ i = 0;
+ while (test_bit(CNIC_LCL_FL_L2_WAIT, &cp->cnic_local_flags) &&
+ ++i < 10)
+ msleep(1);
+
+ if (test_bit(CNIC_LCL_FL_L2_WAIT, &cp->cnic_local_flags))
+ netdev_err(dev->netdev,
+ "iSCSI CLIENT_HALT did not complete\n");
+ cnic_kwq_completion(dev, 1);
memset(&l5_data, 0, sizeof(l5_data));
cnic_submit_kwqe_16(dev, RAMROD_CMD_ID_ETH_CFC_DEL,
@@ -4315,7 +4376,15 @@ static void cnic_stop_hw(struct cnic_dev *dev)
{
if (test_bit(CNIC_F_CNIC_UP, &dev->flags)) {
struct cnic_local *cp = dev->cnic_priv;
+ int i = 0;
+ /* Need to wait for the ring shutdown event to complete
+ * before clearing the CNIC_UP flag.
+ */
+ while (cp->uio_dev != -1 && i < 15) {
+ msleep(100);
+ i++;
+ }
clear_bit(CNIC_F_CNIC_UP, &dev->flags);
rcu_assign_pointer(cp->ulp_ops[CNIC_ULP_L4], NULL);
synchronize_rcu();
diff --git a/drivers/net/cnic.h b/drivers/net/cnic.h
index 4422497..08b1235 100644
--- a/drivers/net/cnic.h
+++ b/drivers/net/cnic.h
@@ -181,6 +181,7 @@ struct cnic_local {
unsigned long cnic_local_flags;
#define CNIC_LCL_FL_KWQ_INIT 0x0
+#define CNIC_LCL_FL_L2_WAIT 0x1
struct cnic_dev *dev;
@@ -348,6 +349,10 @@ struct bnx2x_bd_chain_next {
#define BNX2X_RCQ_DESC_CNT (BCM_PAGE_SIZE / sizeof(union eth_rx_cqe))
#define BNX2X_MAX_RCQ_DESC_CNT (BNX2X_RCQ_DESC_CNT - 1)
+#define BNX2X_NEXT_RCQE(x) (((x) & BNX2X_MAX_RCQ_DESC_CNT) == \
+ (BNX2X_MAX_RCQ_DESC_CNT - 1)) ? \
+ ((x) + 2) : ((x) + 1)
+
#define BNX2X_DEF_SB_ID 16
#define BNX2X_ISCSI_RX_SB_INDEX_NUM \
--
1.6.4.GIT
^ permalink raw reply related
* Re: [PATCH REPOST] PCI: Disable MSI for MCP55 on P5N32-E SLI
From: Jesse Barnes @ 2010-05-18 21:33 UTC (permalink / raw)
To: Ben Hutchings; +Cc: linux-pci, netdev
In-Reply-To: <1273973329.2564.70.camel@localhost>
On Sun, 16 May 2010 02:28:49 +0100
Ben Hutchings <ben@decadent.org.uk> wrote:
> As reported in <http://bugs.debian.org/552299>, MSI appears to be
> broken for this on-board device. We already have a quirk for the
> P5N32-SLI Premium; extend it to cover both variants of the board.
>
> Reported-by: Romain DEGEZ <romain.degez@smartjog.com>
> Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
> Cc: stable@kernel.org
> ---
Applied to linux-next, sorry for missing it earlier.
--
Jesse Barnes, Intel Open Source Technology Center
^ permalink raw reply
* Re: [PATCH 1/3] cxgb4: fix initial addition of MAC address
From: David Miller @ 2010-05-18 22:16 UTC (permalink / raw)
To: dm; +Cc: netdev
In-Reply-To: <1274213233-26309-1-git-send-email-dm@chelsio.com>
From: Dimitris Michailidis <dm@chelsio.com>
Date: Tue, 18 May 2010 13:07:11 -0700
> The call to add the MAC address during link initialization wasn't adding the
> address to one of the HW tables. Fix.
>
> Signed-off-by: Dimitris Michailidis <dm@chelsio.com>
Applied.
^ permalink raw reply
* Re: [PATCH 2/3] cxgb4: keep interrupts available when the ports are brought down
From: David Miller @ 2010-05-18 22:17 UTC (permalink / raw)
To: dm; +Cc: netdev
In-Reply-To: <1274213233-26309-2-git-send-email-dm@chelsio.com>
From: Dimitris Michailidis <dm@chelsio.com>
Date: Tue, 18 May 2010 13:07:12 -0700
> The PF driver needs to remain alert while its ports are down to service
> requests or errors from virtual functions or FW so keep interrupts and
> queues available when the ports are brought down. The change makes
> open_device_map unnecessary so remove it.
>
> Signed-off-by: Dimitris Michailidis <dm@chelsio.com>
Applied.
^ permalink raw reply
* Re: [PATCH 3/3] cxgb4: notify upper drivers if the device is already up when they load
From: David Miller @ 2010-05-18 22:17 UTC (permalink / raw)
To: dm; +Cc: netdev
In-Reply-To: <1274213233-26309-3-git-send-email-dm@chelsio.com>
From: Dimitris Michailidis <dm@chelsio.com>
Date: Tue, 18 May 2010 13:07:13 -0700
> Upper layer drivers aren't notified that a device is ready if their modules
> load after the device becomes ready. Add the missing notification.
>
> Signed-off-by: Dimitris Michailidis <dm@chelsio.com>
Applied.
^ permalink raw reply
* Re: [PATCH 1/2] cnic: Convert cnic_local_flags to atomic ops.
From: David Miller @ 2010-05-18 22:17 UTC (permalink / raw)
To: mchan; +Cc: netdev
In-Reply-To: <1274218373-16324-1-git-send-email-mchan@broadcom.com>
From: "Michael Chan" <mchan@broadcom.com>
Date: Tue, 18 May 2010 14:32:52 -0700
> It is easier to expand the flags for other purposes because it does
> not require a spin_lock. The next bug fix patch needs a flag in
> cnic_local_flags.
>
> Signed-off-by: Michael Chan <mchan@broadcom.com>
Applied.
^ permalink raw reply
* Re: [PATCH 2/2] cnic: Return SPQ credit to bnx2x after ring setup and shutdown.
From: David Miller @ 2010-05-18 22:17 UTC (permalink / raw)
To: mchan; +Cc: netdev
In-Reply-To: <1274218373-16324-2-git-send-email-mchan@broadcom.com>
From: "Michael Chan" <mchan@broadcom.com>
Date: Tue, 18 May 2010 14:32:53 -0700
> Everytime the iSCSI ring finishes setup or shutdown, we need to return
> the SPQ (slow path queue) credit to the bnx2x driver. Without this step,
> the SPQ will eventually be full causing iSCSI to fail. This can happen
> after 3 or 4 MTU changes for example.
>
> Add code to wait for these slow path commands to complete in the RX ring
> and return the SPQ credit to bnx2x.
>
> Signed-off-by: Michael Chan <mchan@broadcom.com>
Applied.
^ permalink raw reply
* Re: [PATCH 1/4] ipv6: Replace inet6_ifaddr->dead with state
From: David Miller @ 2010-05-18 22:27 UTC (permalink / raw)
To: shemminger; +Cc: herbert, jbohac, yoshfuji, netdev
In-Reply-To: <20100518102334.70b6c0b4@nehalam>
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Tue, 18 May 2010 10:23:34 -0700
> On Tue, 18 May 2010 21:04:19 +1000
> Herbert Xu <herbert@gondor.apana.org.au> wrote:
>
>>
>> +enum {
>> + INET6_IFADDR_STATE_DAD,
>> + INET6_IFADDR_STATE_POSTDAD,
>> + INET6_IFADDR_STATE_UP,
>> + INET6_IFADDR_STATE_DEAD,
>> +};
>> +
>> #ifdef __KERNEL__
>
> Does this really need to be visible to user applications?
I'll fix this up when I apply Herbert's patches.
^ permalink raw reply
* Re: [0/4] Fix addrconf race conditions
From: David Miller @ 2010-05-18 22:27 UTC (permalink / raw)
To: shemminger; +Cc: herbert, jbohac, yoshfuji, netdev
In-Reply-To: <20100518102550.65ad3fdd@nehalam>
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Tue, 18 May 2010 10:25:50 -0700
> I wonder if so many fine grained locks are really necessary at
> all. Everything but timers looks like it is under RTNL mutex
> already.
I can't see any reasonable alternative, and it took weeks to get
a fix at all.
So I'm going to apply Herbert's fixes with the __KERNEL__ header
bit fixed up.
Thanks Herbert!
^ permalink raw reply
* Re: [0/4] Fix addrconf race conditions
From: Herbert Xu @ 2010-05-18 22:35 UTC (permalink / raw)
To: David Miller; +Cc: shemminger, jbohac, yoshfuji, netdev
In-Reply-To: <20100518.152759.218054760.davem@davemloft.net>
On Tue, May 18, 2010 at 03:27:59PM -0700, David Miller wrote:
> From: Stephen Hemminger <shemminger@vyatta.com>
> Date: Tue, 18 May 2010 10:25:50 -0700
>
> > I wonder if so many fine grained locks are really necessary at
> > all. Everything but timers looks like it is under RTNL mutex
> > already.
>
> I can't see any reasonable alternative, and it took weeks to get
> a fix at all.
Right, the issue is with forcing an immediate effect after an
NDISC packet triggers an address addition/deletion.
>From what I can gather, the hard case was with NDISC address
addition with DAD disabled. That is, you have an NDISC packet
that causes an addition followed immediately by a packet destined
(or otherwise relying on) that new address. In that case, we
simply have no chance of using the RTNL.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* [PATCH 0/3] Make vhost multi-threaded and associate each thread to its guest's cgroup/cpumask
From: Sridhar Samudrala @ 2010-05-19 0:04 UTC (permalink / raw)
To: Michael S. Tsirkin, netdev, lkml, kvm@vger.kernel.org
The following set of patches create a new API to associate a
workqueue to the current thread's cgroup and cpumask.
This API is used by multi-threaded vhost to associate each
thread to the corresponding guest's cgroup and cpumask.
Thanks
Sridhar
^ permalink raw reply
* [PATCH 1/3] cgroups: Add an API to attach a task to current task's cgroup
From: Sridhar Samudrala @ 2010-05-19 0:04 UTC (permalink / raw)
To: Michael S. Tsirkin, netdev, kvm@vger.kernel.org, lkml
Add a new kernel API to attach a task to current task's cgroup
in all the active hierarchies.
Signed-off-by: Sridhar Samudrala <sri@us.ibm.com>
diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -570,6 +570,7 @@ struct task_struct *cgroup_iter_next(struct cgroup *cgrp,
void cgroup_iter_end(struct cgroup *cgrp, struct cgroup_iter *it);
int cgroup_scan_tasks(struct cgroup_scanner *scan);
int cgroup_attach_task(struct cgroup *, struct task_struct *);
+int cgroup_attach_task_current_cg(struct task_struct *);
/*
* CSS ID is ID for cgroup_subsys_state structs under subsys. This only works
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 6d870f2..6cfeb06 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -1788,6 +1788,29 @@ out:
return retval;
}
+/**
+ * cgroup_attach_task_current_cg - attach task 'tsk' to current task's cgroup
+ * @tsk: the task to be attached
+ */
+int cgroup_attach_task_current_cg(struct task_struct *tsk)
+{
+ struct cgroupfs_root *root;
+ struct cgroup *cur_cg;
+ int retval = 0;
+
+ cgroup_lock();
+ for_each_active_root(root) {
+ cur_cg = task_cgroup_from_root(current, root);
+ retval = cgroup_attach_task(cur_cg, tsk);
+ if (retval)
+ break;
+ }
+ cgroup_unlock();
+
+ return retval;
+}
+EXPORT_SYMBOL_GPL(cgroup_attach_task_current_cg);
+
/*
* Attach task with pid 'pid' to cgroup 'cgrp'. Call with cgroup_mutex
* held. May take task_lock of task
^ permalink raw reply related
* [ PATCH 3/3] vhost: make it more scalable by creating a vhost thread per device
From: Sridhar Samudrala @ 2010-05-19 0:04 UTC (permalink / raw)
To: Michael S. Tsirkin, netdev, lkml, kvm@vger.kernel.org
Make vhost more scalable by creating a separate vhost thread per
vhost device. This provides better scaling across virtio-net interfaces
in multiple guests.
Also attach each vhost thread to the cgroup and cpumask of the
associated guest(qemu or libvirt).
Signed-off-by: Sridhar Samudrala <sri@us.ibm.com>
diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 9777583..18bf5be 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -329,19 +329,22 @@ static void handle_rx_net(struct work_struct *work)
static int vhost_net_open(struct inode *inode, struct file *f)
{
struct vhost_net *n = kmalloc(sizeof *n, GFP_KERNEL);
+ struct vhost_dev *dev;
int r;
if (!n)
return -ENOMEM;
+
+ dev = &n->dev;
n->vqs[VHOST_NET_VQ_TX].handle_kick = handle_tx_kick;
n->vqs[VHOST_NET_VQ_RX].handle_kick = handle_rx_kick;
- r = vhost_dev_init(&n->dev, n->vqs, VHOST_NET_VQ_MAX);
+ r = vhost_dev_init(dev, n->vqs, VHOST_NET_VQ_MAX);
if (r < 0) {
kfree(n);
return r;
}
- vhost_poll_init(n->poll + VHOST_NET_VQ_TX, handle_tx_net, POLLOUT);
- vhost_poll_init(n->poll + VHOST_NET_VQ_RX, handle_rx_net, POLLIN);
+ vhost_poll_init(n->poll + VHOST_NET_VQ_TX, handle_tx_net, POLLOUT, dev);
+ vhost_poll_init(n->poll + VHOST_NET_VQ_RX, handle_rx_net, POLLIN, dev);
n->tx_poll_state = VHOST_NET_POLL_DISABLED;
f->private_data = n;
@@ -644,25 +647,14 @@ static struct miscdevice vhost_net_misc = {
int vhost_net_init(void)
{
- int r = vhost_init();
- if (r)
- goto err_init;
- r = misc_register(&vhost_net_misc);
- if (r)
- goto err_reg;
- return 0;
-err_reg:
- vhost_cleanup();
-err_init:
- return r;
-
+ return misc_register(&vhost_net_misc);
}
+
module_init(vhost_net_init);
void vhost_net_exit(void)
{
misc_deregister(&vhost_net_misc);
- vhost_cleanup();
}
module_exit(vhost_net_exit);
diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
index 49fa953..b076b9b 100644
--- a/drivers/vhost/vhost.c
+++ b/drivers/vhost/vhost.c
@@ -37,8 +37,6 @@ enum {
VHOST_MEMORY_F_LOG = 0x1,
};
-static struct workqueue_struct *vhost_workqueue;
-
static void vhost_poll_func(struct file *file, wait_queue_head_t *wqh,
poll_table *pt)
{
@@ -57,18 +55,19 @@ static int vhost_poll_wakeup(wait_queue_t *wait, unsigned mode, int sync,
if (!((unsigned long)key & poll->mask))
return 0;
- queue_work(vhost_workqueue, &poll->work);
+ queue_work(poll->dev->wq, &poll->work);
return 0;
}
/* Init poll structure */
void vhost_poll_init(struct vhost_poll *poll, work_func_t func,
- unsigned long mask)
+ unsigned long mask, struct vhost_dev *dev)
{
INIT_WORK(&poll->work, func);
init_waitqueue_func_entry(&poll->wait, vhost_poll_wakeup);
init_poll_funcptr(&poll->table, vhost_poll_func);
poll->mask = mask;
+ poll->dev = dev;
}
/* Start polling a file. We add ourselves to file's wait queue. The caller must
@@ -97,7 +96,7 @@ void vhost_poll_flush(struct vhost_poll *poll)
void vhost_poll_queue(struct vhost_poll *poll)
{
- queue_work(vhost_workqueue, &poll->work);
+ queue_work(poll->dev->wq, &poll->work);
}
static void vhost_vq_reset(struct vhost_dev *dev,
@@ -129,6 +128,13 @@ long vhost_dev_init(struct vhost_dev *dev,
struct vhost_virtqueue *vqs, int nvqs)
{
int i;
+ char vhost_name[20];
+
+ snprintf(vhost_name, 20, "vhost-%d", current->pid);
+ dev->wq = create_singlethread_workqueue_in_current_cg(vhost_name);
+ if (!dev->wq)
+ return -ENOMEM;
+
dev->vqs = vqs;
dev->nvqs = nvqs;
mutex_init(&dev->mutex);
@@ -144,7 +150,7 @@ long vhost_dev_init(struct vhost_dev *dev,
if (dev->vqs[i].handle_kick)
vhost_poll_init(&dev->vqs[i].poll,
dev->vqs[i].handle_kick,
- POLLIN);
+ POLLIN, dev);
}
return 0;
}
@@ -217,6 +223,8 @@ void vhost_dev_cleanup(struct vhost_dev *dev)
if (dev->mm)
mmput(dev->mm);
dev->mm = NULL;
+
+ destroy_workqueue(dev->wq);
}
static int log_access_ok(void __user *log_base, u64 addr, unsigned long sz)
@@ -1105,16 +1113,3 @@ void vhost_disable_notify(struct vhost_virtqueue *vq)
vq_err(vq, "Failed to enable notification at %p: %d\n",
&vq->used->flags, r);
}
-
-int vhost_init(void)
-{
- vhost_workqueue = create_singlethread_workqueue("vhost");
- if (!vhost_workqueue)
- return -ENOMEM;
- return 0;
-}
-
-void vhost_cleanup(void)
-{
- destroy_workqueue(vhost_workqueue);
-}
diff --git a/drivers/vhost/vhost.h b/drivers/vhost/vhost.h
index 44591ba..60fefd0 100644
--- a/drivers/vhost/vhost.h
+++ b/drivers/vhost/vhost.h
@@ -29,10 +29,11 @@ struct vhost_poll {
/* struct which will handle all actual work. */
struct work_struct work;
unsigned long mask;
+ struct vhost_dev *dev;
};
void vhost_poll_init(struct vhost_poll *poll, work_func_t func,
- unsigned long mask);
+ unsigned long mask, struct vhost_dev *dev);
void vhost_poll_start(struct vhost_poll *poll, struct file *file);
void vhost_poll_stop(struct vhost_poll *poll);
void vhost_poll_flush(struct vhost_poll *poll);
@@ -110,6 +111,7 @@ struct vhost_dev {
int nvqs;
struct file *log_file;
struct eventfd_ctx *log_ctx;
+ struct workqueue_struct *wq;
};
long vhost_dev_init(struct vhost_dev *, struct vhost_virtqueue *vqs, int nvqs);
@@ -136,9 +138,6 @@ bool vhost_enable_notify(struct vhost_virtqueue *);
int vhost_log_write(struct vhost_virtqueue *vq, struct vhost_log *log,
unsigned int log_num, u64 len);
-int vhost_init(void);
-void vhost_cleanup(void);
-
#define vq_err(vq, fmt, ...) do { \
pr_debug(pr_fmt(fmt), ##__VA_ARGS__); \
if ((vq)->error_ctx) \
^ permalink raw reply related
* [PATCH 2/3] workqueue: Add an API to create a singlethread workqueue attached to the current task's cgroup
From: Sridhar Samudrala @ 2010-05-19 0:04 UTC (permalink / raw)
To: Michael S. Tsirkin, netdev, lkml, kvm@vger.kernel.org
Add a new kernel API to create a singlethread workqueue and attach it's
task to current task's cgroup and cpumask.
Signed-off-by: Sridhar Samudrala <sri@us.ibm.com>
diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h
index 9466e86..6d6f301 100644
--- a/include/linux/workqueue.h
+++ b/include/linux/workqueue.h
@@ -211,6 +211,7 @@ __create_workqueue_key(const char *name, int singlethread,
#define create_freezeable_workqueue(name) __create_workqueue((name), 1, 1, 0)
#define create_singlethread_workqueue(name) __create_workqueue((name), 1, 0, 0)
+extern struct workqueue_struct *create_singlethread_workqueue_in_current_cg(char *name);
extern void destroy_workqueue(struct workqueue_struct *wq);
extern int queue_work(struct workqueue_struct *wq, struct work_struct *work);
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 5bfb213..6ba226e 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -35,6 +35,7 @@
#include <linux/lockdep.h>
#define CREATE_TRACE_POINTS
#include <trace/events/workqueue.h>
+#include <linux/cgroup.h>
/*
* The per-CPU workqueue (if single thread, we always use the first
@@ -193,6 +194,45 @@ static const struct cpumask *cpu_singlethread_map __read_mostly;
*/
static cpumask_var_t cpu_populated_map __read_mostly;
+static struct task_struct *get_singlethread_wq_task(struct workqueue_struct *wq)
+{
+ return (per_cpu_ptr(wq->cpu_wq, singlethread_cpu))->thread;
+}
+
+/* Create a singlethread workqueue and attach it's task to the current task's
+ * cgroup and set it's cpumask to the current task's cpumask.
+ */
+struct workqueue_struct *create_singlethread_workqueue_in_current_cg(char *name)
+{
+ struct workqueue_struct *wq;
+ struct task_struct *task;
+ cpumask_var_t mask;
+
+ wq = create_singlethread_workqueue(name);
+ if (!wq)
+ goto out;
+
+ if (!alloc_cpumask_var(&mask, GFP_KERNEL))
+ goto err;
+
+ if (sched_getaffinity(current->pid, mask))
+ goto err;
+
+ task = get_singlethread_wq_task(wq);
+ if (sched_setaffinity(task->pid, mask))
+ goto err;
+
+ if (cgroup_attach_task_current_cg(task))
+ goto err;
+out:
+ return wq;
+err:
+ destroy_workqueue(wq);
+ wq = NULL;
+ goto out;
+}
+EXPORT_SYMBOL_GPL(create_singlethread_workqueue_in_current_cg);
+
/* If it's single threaded, it isn't in the list of workqueues. */
static inline int is_wq_single_threaded(struct workqueue_struct *wq)
{
^ permalink raw reply related
* Re: [PATCH net-next] ixgbe: return error in set_rar when index out of range
From: Jeff Kirsher @ 2010-05-19 0:19 UTC (permalink / raw)
To: Shirley Ma; +Cc: davem, kvm, netdev, e1000-devel
In-Reply-To: <1274196888.8701.2.camel@localhost.localdomain>
On Tue, May 18, 2010 at 08:34, Shirley Ma <mashirle@us.ibm.com> wrote:
> Return -1 when set_rar index is out of range
>
> Signed-off-by: Shirley Ma <xma@us.ibm.com>
> ---
>
> drivers/net/ixgbe/ixgbe_common.c | 1 +
> 1 files changed, 1 insertions(+), 0 deletions(-)
>
I think this should use IXGBE_ERR_<blah> instead and there is another
spot where this could be used. Instead I propose this patch
instead...
ixgbe: return IXGBE_ERR_RAR_INDEX when out of range
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Based on patch from Shirley Ma <xma@us.ibm.com>
Return IXGBE_ERR_RAR_INDEX when RAR index is out of range, instead of
returning IXGBE_SUCCESS.
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: Don Skidmore <donald.c.skidmore@intel.com>
---
drivers/net/ixgbe/ixgbe_common.c | 2 ++
drivers/net/ixgbe/ixgbe_type.h | 1 +
2 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_common.c b/drivers/net/ixgbe/ixgbe_common.c
index 1159d91..9595b1b 100644
--- a/drivers/net/ixgbe/ixgbe_common.c
+++ b/drivers/net/ixgbe/ixgbe_common.c
@@ -1188,6 +1188,7 @@ s32 ixgbe_set_rar_generic(struct ixgbe_hw *hw,
u32 index, u8 *addr, u32 vmdq,
IXGBE_WRITE_REG(hw, IXGBE_RAH(index), rar_high);
} else {
hw_dbg(hw, "RAR index %d is out of range.\n", index);
+ return IXGBE_ERR_RAR_INDEX;
}
return 0;
@@ -1219,6 +1220,7 @@ s32 ixgbe_clear_rar_generic(struct ixgbe_hw *hw,
u32 index)
IXGBE_WRITE_REG(hw, IXGBE_RAH(index), rar_high);
} else {
hw_dbg(hw, "RAR index %d is out of range.\n", index);
+ return IXGBE_ERR_RAR_INDEX;
}
/* clear VMDq pool/queue selection for this RAR */
diff --git a/drivers/net/ixgbe/ixgbe_type.h b/drivers/net/ixgbe/ixgbe_type.h
index bd69196..37d2807 100644
--- a/drivers/net/ixgbe/ixgbe_type.h
+++ b/drivers/net/ixgbe/ixgbe_type.h
@@ -2600,6 +2600,7 @@ struct ixgbe_info {
#define IXGBE_ERR_FDIR_REINIT_FAILED -23
#define IXGBE_ERR_EEPROM_VERSION -24
#define IXGBE_ERR_NO_SPACE -25
+#define IXGBE_ERR_RAR_INDEX -26
#define IXGBE_NOT_IMPLEMENTED 0x7FFFFFFF
#endif /* _IXGBE_TYPE_H_ */
^ permalink raw reply related
* Re: QoS weirdness : HTB accuracy
From: Philip A. Prindeville @ 2010-05-19 0:07 UTC (permalink / raw)
To: Julien Vehent; +Cc: Netdev, netfilter
In-Reply-To: <067c83163988908ef546d7ff7f560a17@localhost>
On 03/25/2010 12:06 PM, Julien Vehent wrote:
> Hello folks,
>
> I observe unused bandwidth on my QoS policy that I cannot explain.
> Conditions are: I have a HTB tree with 8 classes and a total rate of
> 768kbits. I use the ATM option so I assume the real rate to be something
> close to 675kbits (about 88% of the ethernet rate).
>
> The sum of my 8 rates is exactly 768kbits. Some have ceil values up to
> 768kbits.
>
> When class 20 "tcp_acks" starts borrowing, TC reduces the total bandwidth
> down to 595kbits/S (minus 79kbits/s). And I can't explain why....
>
> The attached graph "tc_htb_weirdness.png" shows the result: there are
> 'holes' in the sending rate.
>
> I tried to play with burst sizes, r2q value and hysteresis mode, but the
> results are the same.
>
> System is debian squeeze, kernel version is 2.6.26, iproute2 version is
> 2.6.26 - 07/25/2008.
>
> I have attached two files:
> - "tcrules.txt" : the traffic control rules
> - "tc_htb_weirdness.png" : the rrdtool graph, resolution is 1 second.
>
> And here: http://jve.linuxwall.info/ressources/code/tc_hole_analysis.html
> a sheet with some of the measures values. I used it to calculate the size
> of one of the hole. The last table (with green and red cells) shows that,
> when class 20 "tcp_acks" starts sending at unixtime 1265496813, there is a
> lot of bandwidth left over (last column is all green). During the 95
> seconds while class 20 is sending, 3880776 bits could be sent but are not.
> That's about 40kbits/s on average.
>
> Does anybody observess the same behavior? Any logical explanation to this
> or is it a bug ?
>
>
> Julien
Sorry for the late response: could this be an "aliasing" issue caused
by sampling intervals (granularity)?
-Philip
^ permalink raw reply
* Re: Network QoS support in applications
From: Philip A. Prindeville @ 2010-05-19 0:04 UTC (permalink / raw)
To: David Miller; +Cc: dunc, kalle.valo, kaber, netdev, linux-wireless
In-Reply-To: <4B9944A0.5080308@redfish-solutions.com>
On 03/11/2010 12:29 PM, Philip A. Prindeville wrote:
> On 03/11/2010 12:27 PM, David Miller wrote:
>
>> From: "Philip A. Prindeville" <philipp_subx@redfish-solutions.com>
>> Date: Thu, 11 Mar 2010 12:21:11 -0700
>>
>>
>>> And yes, there will always be misbehaving users. They are a fact of
>>> life. That doesn't mean we should lobotomize the network. We don't
>>> have an authentication mechanism on ICMP Redirects or Source-Quench,
>>>
>> Which is why most networks block those packets from the outside.
>>
>>
>>> Nor is ARP authenticated.
>>>
>> Which is why people control who can plug into their physical
>> network.
>>
>> None of the things you are saying support the idea of having
>> applications decide what the DSCP marking should be.
>>
>
> Does "decide what the DSCP marking should be" include complying to the recommendations of RFC-4594?
>
If anyone cares, here's an update:
I've submitted patches for QoS configuration for:
APR/Apache (stalled);
Proftpd (committed);
Openssh (pending review);
Firefox/Thunderbird (reviewed and on-track for commit);
Cyrus (committed);
Sendmail (submittted and acknowledged, but not yet reviewed);
Curl (stalled);
All, as per the request of the maintainers, default to either no QoS
markings or previous RFC-791 QoS markings if that's what they already
supported (Proftpd and Openssh).
If anyone can think of anything else that needs to be supported to
impact a significant portion of network (or enterprise intranet)
traffic, please call it out.
And if anyone wants to see if they can help get Apache unstalled (it's
mostly an autoconf issue with Solaris that's holding things up), please
give me a holler offline.
Thanks.
^ permalink raw reply
* [net-next PATCH 1/7] ixgbe: fix header len when unsplit packet overflows to data buffer
From: Jeff Kirsher @ 2010-05-19 2:00 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Shannon Nelson, Jeff Kirsher
From: Shannon Nelson <shannon.nelson@intel.com>
When in packet split mode, packet type is not recognized, and the packet is
larger than the header size, the 82599 overflows the packet into the data
area, but doesn't set the HDR_LEN field. We can safely assume the length
is the current header size. This fixes an obscure corner case that can be
triggered by non-ip packet headers or (more likely) by disabling the L2
packet recognition.
Signed-off-by: Shannon Nelson <shannon.nelson@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_main.c | 5 +++--
1 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 926ad8c..a9e091c 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -1201,9 +1201,10 @@ static bool ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
hdr_info = le16_to_cpu(ixgbe_get_hdr_info(rx_desc));
len = (hdr_info & IXGBE_RXDADV_HDRBUFLEN_MASK) >>
IXGBE_RXDADV_HDRBUFLEN_SHIFT;
- if (len > IXGBE_RX_HDR_SIZE)
- len = IXGBE_RX_HDR_SIZE;
upper_len = le16_to_cpu(rx_desc->wb.upper.length);
+ if ((len > IXGBE_RX_HDR_SIZE) ||
+ (upper_len && !(hdr_info & IXGBE_RXDADV_SPH)))
+ len = IXGBE_RX_HDR_SIZE;
} else {
len = le16_to_cpu(rx_desc->wb.upper.length);
}
^ permalink raw reply related
* [net-next PATCH 2/7] ixgbe: fix wrong offset to fc_frame_header in ixgbe_fcoe_ddp
From: Jeff Kirsher @ 2010-05-19 2:00 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Yi Zou, Jeff Kirsher
In-Reply-To: <20100519020003.18654.94899.stgit@localhost.localdomain>
From: Yi Zou <yi.zou@intel.com>
Make sure we point to the right offset of the fc_frame_header when
VLAN header exists and HW has VLAN stripping disabled.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_fcoe.c | 11 +++++++----
1 files changed, 7 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_fcoe.c b/drivers/net/ixgbe/ixgbe_fcoe.c
index 6493049..a82d2fc 100644
--- a/drivers/net/ixgbe/ixgbe_fcoe.c
+++ b/drivers/net/ixgbe/ixgbe_fcoe.c
@@ -32,6 +32,7 @@
#endif /* CONFIG_IXGBE_DCB */
#include <linux/if_ether.h>
#include <linux/gfp.h>
+#include <linux/if_vlan.h>
#include <scsi/scsi_cmnd.h>
#include <scsi/scsi_device.h>
#include <scsi/fc/fc_fs.h>
@@ -312,10 +313,12 @@ int ixgbe_fcoe_ddp(struct ixgbe_adapter *adapter,
if (fcerr == IXGBE_FCERR_BADCRC)
skb->ip_summed = CHECKSUM_NONE;
- skb_reset_network_header(skb);
- skb_set_transport_header(skb, skb_network_offset(skb) +
- sizeof(struct fcoe_hdr));
- fh = (struct fc_frame_header *)skb_transport_header(skb);
+ if (eth_hdr(skb)->h_proto == htons(ETH_P_8021Q))
+ fh = (struct fc_frame_header *)(skb->data +
+ sizeof(struct vlan_hdr) + sizeof(struct fcoe_hdr));
+ else
+ fh = (struct fc_frame_header *)(skb->data +
+ sizeof(struct fcoe_hdr));
fctl = ntoh24(fh->fh_f_ctl);
if (fctl & FC_FC_EX_CTX)
xid = be16_to_cpu(fh->fh_ox_id);
^ permalink raw reply related
* [net-next PATCH 3/7] ixgbe: remove some redundant code in setting FCoE FIP filter
From: Jeff Kirsher @ 2010-05-19 2:00 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Yi Zou, Jeff Kirsher
In-Reply-To: <20100519020003.18654.94899.stgit@localhost.localdomain>
From: Yi Zou <yi.zou@intel.com>
The ETQS setup for FIP out side the if..else is enough for the ETQS
setup for FIP, so remove redundant code.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_fcoe.c | 6 ------
1 files changed, 0 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_fcoe.c b/drivers/net/ixgbe/ixgbe_fcoe.c
index a82d2fc..45182ab 100644
--- a/drivers/net/ixgbe/ixgbe_fcoe.c
+++ b/drivers/net/ixgbe/ixgbe_fcoe.c
@@ -539,12 +539,6 @@ void ixgbe_configure_fcoe(struct ixgbe_adapter *adapter)
}
IXGBE_WRITE_REG(hw, IXGBE_FCRECTL, IXGBE_FCRECTL_ENA);
IXGBE_WRITE_REG(hw, IXGBE_ETQS(IXGBE_ETQF_FILTER_FCOE), 0);
- fcoe_i = f->mask;
- fcoe_i &= IXGBE_FCRETA_ENTRY_MASK;
- fcoe_q = adapter->rx_ring[fcoe_i]->reg_idx;
- IXGBE_WRITE_REG(hw, IXGBE_ETQS(IXGBE_ETQF_FILTER_FIP),
- IXGBE_ETQS_QUEUE_EN |
- (fcoe_q << IXGBE_ETQS_RX_QUEUE_SHIFT));
} else {
/* Use single rx queue for FCoE */
fcoe_i = f->mask;
^ permalink raw reply related
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