* [PATCH 01/13] Bluetooth: Remove max_tx and tx_window module paramenters from L2CAP
From: Marcel Holtmann @ 2010-07-08 19:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <cover.1278619047.git.marcel@holtmann.org>
From: Gustavo F. Padovan <padovan@profusion.mobi>
We don't need these parameters anymore since we have socket options for
them.
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
---
net/bluetooth/l2cap.c | 12 ++----------
1 files changed, 2 insertions(+), 10 deletions(-)
diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c
index 1b682a5..5c636b3 100644
--- a/net/bluetooth/l2cap.c
+++ b/net/bluetooth/l2cap.c
@@ -60,8 +60,6 @@ static int enable_ertm = 1;
#else
static int enable_ertm = 0;
#endif
-static int max_transmit = L2CAP_DEFAULT_MAX_TX;
-static int tx_window = L2CAP_DEFAULT_TX_WINDOW;
static u32 l2cap_feat_mask = L2CAP_FEAT_FIXED_CHAN;
static u8 l2cap_fixed_chan[8] = { 0x02, };
@@ -808,9 +806,9 @@ static void l2cap_sock_init(struct sock *sk, struct sock *parent)
pi->mode = L2CAP_MODE_ERTM;
else
pi->mode = L2CAP_MODE_BASIC;
- pi->max_tx = max_transmit;
+ pi->max_tx = L2CAP_DEFAULT_MAX_TX;
pi->fcs = L2CAP_FCS_CRC16;
- pi->tx_win = tx_window;
+ pi->tx_win = L2CAP_DEFAULT_TX_WINDOW;
pi->sec_level = BT_SECURITY_LOW;
pi->role_switch = 0;
pi->force_reliable = 0;
@@ -4674,12 +4672,6 @@ module_exit(l2cap_exit);
module_param(enable_ertm, bool, 0644);
MODULE_PARM_DESC(enable_ertm, "Enable enhanced retransmission mode");
-module_param(max_transmit, uint, 0644);
-MODULE_PARM_DESC(max_transmit, "Max transmit value (default = 3)");
-
-module_param(tx_window, uint, 0644);
-MODULE_PARM_DESC(tx_window, "Transmission window size value (default = 63)");
-
MODULE_AUTHOR("Marcel Holtmann <marcel@holtmann.org>");
MODULE_DESCRIPTION("Bluetooth L2CAP ver " VERSION);
MODULE_VERSION(VERSION);
--
1.7.1.1
^ permalink raw reply related
* Pull request: bluetooth-2.6 2010-07-08
From: Marcel Holtmann @ 2010-07-08 19:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev
Hi Dave,
these patches fix a few bugs and crashes and also two security related
issues with the authentication procedure.
This will also be my last pull request for you. From now on all future
patches will be submitted towards John's wireless trees. That way we
can coordindate the development of Bluetooth 3.0 (Bluetooth over WiFi)
and avoid any kind of conflicts.
Regards
Marcel
Please pull from
git://git.kernel.org/pub/scm/linux/kernel/git/holtmann/bluetooth-2.6.git master
This will update the following files:
net/bluetooth/Kconfig | 13 ----
net/bluetooth/hci_conn.c | 5 ++
net/bluetooth/hci_event.c | 2 +
net/bluetooth/l2cap.c | 138 +++++++++++++++++++++++++++------------------
4 files changed, 90 insertions(+), 68 deletions(-)
through these ChangeSets:
Andrei Emeltchenko (1):
Bluetooth: Check L2CAP pending status before sending connect request
Gustavo F. Padovan (8):
Bluetooth: Remove max_tx and tx_window module paramenters from L2CAP
Bluetooth: Remove L2CAP Extended Features from Kconfig
Bluetooth: Fix drop of packets with invalid req_seq/tx_seq
Bluetooth: Fix bug with ERTM vars increment
Bluetooth: Only check SAR bits if frame is an I-frame
Bluetooth: Fix bug in l2cap_ertm_send() behavior
Bluetooth: Proper shutdown ERTM when closing the channel
Bluetooth: Fix L2CAP control bit field corruption
Johan Hedberg (1):
Bluetooth: Reset the security level after an authentication failure
João Paulo Rechi Vita (1):
Bluetooth: Fix SREJ_QUEUE corruption in L2CAP
Nathan Holstein (1):
Bluetooth: Fix bug with ERTM minimum packet length
Ville Tervo (1):
Bluetooth: Update sec_level/auth_type for already existing connections
^ permalink raw reply
* [PATCH 6/6 net-next-2.6] vxge: Version update
From: Jon Mason @ 2010-07-08 19:24 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Sreenivasa Honnur, Ramkrishna Vepa
Version update
Signed-off-by: Jon Mason <jon.mason@exar.com>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@exar.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@exar.com>
---
drivers/net/vxge/vxge-version.h | 5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/vxge/vxge-version.h b/drivers/net/vxge/vxge-version.h
index 4008e6b..7b9047c 100644
--- a/drivers/net/vxge/vxge-version.h
+++ b/drivers/net/vxge/vxge-version.h
@@ -12,12 +12,11 @@
* Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#ifndef VXGE_VERSION_H
-
#define VXGE_VERSION_H
#define VXGE_VERSION_MAJOR "2"
#define VXGE_VERSION_MINOR "0"
-#define VXGE_VERSION_FIX "8"
-#define VXGE_VERSION_BUILD "20182"
+#define VXGE_VERSION_FIX "9"
+#define VXGE_VERSION_BUILD "20670"
#define VXGE_VERSION_FOR "k"
#endif
^ permalink raw reply related
* [PATCH 5/6 net-next-2.6] vxge: Update maintainers information
From: Jon Mason @ 2010-07-08 19:23 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Sreenivasa Honnur, Ramkrishna Vepa
Update and correct maintainers information
Signed-off-by: Jon Mason <jon.mason@exar.com>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@exar.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@exar.com>
---
MAINTAINERS | 12 +++++++-----
1 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index 2ebb567..28cd736 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -3895,17 +3895,19 @@ L: netem@lists.linux-foundation.org
S: Maintained
F: net/sched/sch_netem.c
-NETERION (S2IO) 10GbE DRIVER (xframe/vxge)
-M: Ramkrishna Vepa <ram.vepa@neterion.com>
-M: Rastapur Santosh <santosh.rastapur@neterion.com>
-M: Sivakumar Subramani <sivakumar.subramani@neterion.com>
-M: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
+NETERION 10GbE DRIVERS (s2io/vxge)
+M: Ramkrishna Vepa <ramkrishna.vepa@exar.com>
+M: Sivakumar Subramani <sivakumar.subramani@exar.com>
+M: Sreenivasa Honnur <sreenivasa.honnur@exar.com>
+M: Jon Mason <jon.mason@exar.com>
L: netdev@vger.kernel.org
W: http://trac.neterion.com/cgi-bin/trac.cgi/wiki/Linux?Anonymous
W: http://trac.neterion.com/cgi-bin/trac.cgi/wiki/X3100Linux?Anonymous
S: Supported
F: Documentation/networking/s2io.txt
F: drivers/net/s2io*
+F: Documentation/networking/vxge.txt
+F: drivers/net/vxge/
NETFILTER/IPTABLES/IPCHAINS
P: Rusty Russell
^ permalink raw reply related
* [PATCH 3/6 net-next-2.6] vxge: Fix multicast issues
From: Jon Mason @ 2010-07-08 19:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Sreenivasa Honnur, Ramkrishna Vepa
Fix error in multicast flag check, add calls to restore the status of
multicast and promiscuous mode settings after change_mtu, and style
cleanups to shorten the function calls by using a temporary variable.
Signed-off-by: Jon Mason <jon.mason@exar.com>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@exar.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@exar.com>
---
drivers/net/vxge/vxge-main.c | 270 +++++++++++++++++++++++-------------------
drivers/net/vxge/vxge-main.h | 2 +
2 files changed, 148 insertions(+), 124 deletions(-)
diff --git a/drivers/net/vxge/vxge-main.c b/drivers/net/vxge/vxge-main.c
index b8eed71..f921196 100644
--- a/drivers/net/vxge/vxge-main.c
+++ b/drivers/net/vxge/vxge-main.c
@@ -1091,7 +1091,8 @@ static void vxge_set_multicast(struct net_device *dev)
struct netdev_hw_addr *ha;
struct vxgedev *vdev;
int i, mcast_cnt = 0;
- struct __vxge_hw_device *hldev;
+ struct __vxge_hw_device *hldev;
+ struct vxge_vpath *vpath;
enum vxge_hw_status status = VXGE_HW_OK;
struct macInfo mac_info;
int vpath_idx = 0;
@@ -1111,46 +1112,48 @@ static void vxge_set_multicast(struct net_device *dev)
if ((dev->flags & IFF_ALLMULTI) && (!vdev->all_multi_flg)) {
for (i = 0; i < vdev->no_of_vpath; i++) {
- vxge_assert(vdev->vpaths[i].is_open);
- status = vxge_hw_vpath_mcast_enable(
- vdev->vpaths[i].handle);
+ vpath = &vdev->vpaths[i];
+ vxge_assert(vpath->is_open);
+ status = vxge_hw_vpath_mcast_enable(vpath->handle);
+ if (status != VXGE_HW_OK)
+ vxge_debug_init(VXGE_ERR, "failed to enable "
+ "multicast, status %d", status);
vdev->all_multi_flg = 1;
}
- } else if ((dev->flags & IFF_ALLMULTI) && (vdev->all_multi_flg)) {
+ } else if (!(dev->flags & IFF_ALLMULTI) && (vdev->all_multi_flg)) {
for (i = 0; i < vdev->no_of_vpath; i++) {
- vxge_assert(vdev->vpaths[i].is_open);
- status = vxge_hw_vpath_mcast_disable(
- vdev->vpaths[i].handle);
- vdev->all_multi_flg = 1;
+ vpath = &vdev->vpaths[i];
+ vxge_assert(vpath->is_open);
+ status = vxge_hw_vpath_mcast_disable(vpath->handle);
+ if (status != VXGE_HW_OK)
+ vxge_debug_init(VXGE_ERR, "failed to disable "
+ "multicast, status %d", status);
+ vdev->all_multi_flg = 0;
}
}
- if (status != VXGE_HW_OK)
- vxge_debug_init(VXGE_ERR,
- "failed to %s multicast, status %d",
- dev->flags & IFF_ALLMULTI ?
- "enable" : "disable", status);
if (!vdev->config.addr_learn_en) {
- if (dev->flags & IFF_PROMISC) {
- for (i = 0; i < vdev->no_of_vpath; i++) {
- vxge_assert(vdev->vpaths[i].is_open);
+ for (i = 0; i < vdev->no_of_vpath; i++) {
+ vpath = &vdev->vpaths[i];
+ vxge_assert(vpath->is_open);
+
+ if (dev->flags & IFF_PROMISC)
status = vxge_hw_vpath_promisc_enable(
- vdev->vpaths[i].handle);
- }
- } else {
- for (i = 0; i < vdev->no_of_vpath; i++) {
- vxge_assert(vdev->vpaths[i].is_open);
+ vpath->handle);
+ else
status = vxge_hw_vpath_promisc_disable(
- vdev->vpaths[i].handle);
- }
+ vpath->handle);
+ if (status != VXGE_HW_OK)
+ vxge_debug_init(VXGE_ERR, "failed to %s promisc"
+ ", status %d", dev->flags&IFF_PROMISC ?
+ "enable" : "disable", status);
}
}
memset(&mac_info, 0, sizeof(struct macInfo));
/* Update individual M_CAST address list */
if ((!vdev->all_multi_flg) && netdev_mc_count(dev)) {
-
mcast_cnt = vdev->vpaths[0].mcast_addr_cnt;
list_head = &vdev->vpaths[0].mac_addr_list;
if ((netdev_mc_count(dev) +
@@ -1160,14 +1163,7 @@ static void vxge_set_multicast(struct net_device *dev)
/* Delete previous MC's */
for (i = 0; i < mcast_cnt; i++) {
- if (!list_empty(list_head))
- mac_entry = (struct vxge_mac_addrs *)
- list_first_entry(list_head,
- struct vxge_mac_addrs,
- item);
-
list_for_each_safe(entry, next, list_head) {
-
mac_entry = (struct vxge_mac_addrs *) entry;
/* Copy the mac address to delete */
mac_address = (u8 *)&mac_entry->macaddr;
@@ -1210,9 +1206,7 @@ _set_all_mcast:
mcast_cnt = vdev->vpaths[0].mcast_addr_cnt;
/* Delete previous MC's */
for (i = 0; i < mcast_cnt; i++) {
-
list_for_each_safe(entry, next, list_head) {
-
mac_entry = (struct vxge_mac_addrs *) entry;
/* Copy the mac address to delete */
mac_address = (u8 *)&mac_entry->macaddr;
@@ -1232,9 +1226,10 @@ _set_all_mcast:
/* Enable all multicast */
for (i = 0; i < vdev->no_of_vpath; i++) {
- vxge_assert(vdev->vpaths[i].is_open);
- status = vxge_hw_vpath_mcast_enable(
- vdev->vpaths[i].handle);
+ vpath = &vdev->vpaths[i];
+ vxge_assert(vpath->is_open);
+
+ status = vxge_hw_vpath_mcast_enable(vpath->handle);
if (status != VXGE_HW_OK) {
vxge_debug_init(VXGE_ERR,
"%s:%d Enabling all multicasts failed",
@@ -1395,6 +1390,7 @@ void vxge_vpath_intr_disable(struct vxgedev *vdev, int vp_id)
static int vxge_reset_vpath(struct vxgedev *vdev, int vp_id)
{
enum vxge_hw_status status = VXGE_HW_OK;
+ struct vxge_vpath *vpath = &vdev->vpaths[vp_id];
int ret = 0;
/* check if device is down already */
@@ -1405,12 +1401,10 @@ static int vxge_reset_vpath(struct vxgedev *vdev, int vp_id)
if (test_bit(__VXGE_STATE_RESET_CARD, &vdev->state))
return 0;
- if (vdev->vpaths[vp_id].handle) {
- if (vxge_hw_vpath_reset(vdev->vpaths[vp_id].handle)
- == VXGE_HW_OK) {
+ if (vpath->handle) {
+ if (vxge_hw_vpath_reset(vpath->handle) == VXGE_HW_OK) {
if (is_vxge_card_up(vdev) &&
- vxge_hw_vpath_recover_from_reset(
- vdev->vpaths[vp_id].handle)
+ vxge_hw_vpath_recover_from_reset(vpath->handle)
!= VXGE_HW_OK) {
vxge_debug_init(VXGE_ERR,
"vxge_hw_vpath_recover_from_reset"
@@ -1426,11 +1420,20 @@ static int vxge_reset_vpath(struct vxgedev *vdev, int vp_id)
} else
return VXGE_HW_FAIL;
- vxge_restore_vpath_mac_addr(&vdev->vpaths[vp_id]);
- vxge_restore_vpath_vid_table(&vdev->vpaths[vp_id]);
+ vxge_restore_vpath_mac_addr(vpath);
+ vxge_restore_vpath_vid_table(vpath);
/* Enable all broadcast */
- vxge_hw_vpath_bcast_enable(vdev->vpaths[vp_id].handle);
+ vxge_hw_vpath_bcast_enable(vpath->handle);
+
+ /* Enable all multicast */
+ if (vdev->all_multi_flg) {
+ status = vxge_hw_vpath_mcast_enable(vpath->handle);
+ if (status != VXGE_HW_OK)
+ vxge_debug_init(VXGE_ERR,
+ "%s:%d Enabling multicast failed",
+ __func__, __LINE__);
+ }
/* Enable the interrupts */
vxge_vpath_intr_enable(vdev, vp_id);
@@ -1438,17 +1441,17 @@ static int vxge_reset_vpath(struct vxgedev *vdev, int vp_id)
smp_wmb();
/* Enable the flow of traffic through the vpath */
- vxge_hw_vpath_enable(vdev->vpaths[vp_id].handle);
+ vxge_hw_vpath_enable(vpath->handle);
smp_wmb();
- vxge_hw_vpath_rx_doorbell_init(vdev->vpaths[vp_id].handle);
- vdev->vpaths[vp_id].ring.last_status = VXGE_HW_OK;
+ vxge_hw_vpath_rx_doorbell_init(vpath->handle);
+ vpath->ring.last_status = VXGE_HW_OK;
/* Vpath reset done */
clear_bit(vp_id, &vdev->vp_reset);
/* Start the vpath queue */
- vxge_wake_tx_queue(&vdev->vpaths[vp_id].fifo);
+ vxge_wake_tx_queue(&vpath->fifo);
return ret;
}
@@ -1482,9 +1485,9 @@ static int do_vxge_reset(struct vxgedev *vdev, int event)
vxge_debug_init(VXGE_ERR,
"%s: execution mode is debug, returning..",
vdev->ndev->name);
- clear_bit(__VXGE_STATE_CARD_UP, &vdev->state);
- netif_tx_stop_all_queues(vdev->ndev);
- return 0;
+ clear_bit(__VXGE_STATE_CARD_UP, &vdev->state);
+ netif_tx_stop_all_queues(vdev->ndev);
+ return 0;
}
}
@@ -1631,8 +1634,7 @@ out:
*/
int vxge_reset(struct vxgedev *vdev)
{
- do_vxge_reset(vdev, VXGE_LL_FULL_RESET);
- return 0;
+ return do_vxge_reset(vdev, VXGE_LL_FULL_RESET);
}
/**
@@ -1995,17 +1997,17 @@ enum vxge_hw_status vxge_restore_vpath_mac_addr(struct vxge_vpath *vpath)
/* reset vpaths */
enum vxge_hw_status vxge_reset_all_vpaths(struct vxgedev *vdev)
{
- int i;
enum vxge_hw_status status = VXGE_HW_OK;
+ struct vxge_vpath *vpath;
+ int i;
- for (i = 0; i < vdev->no_of_vpath; i++)
- if (vdev->vpaths[i].handle) {
- if (vxge_hw_vpath_reset(vdev->vpaths[i].handle)
- == VXGE_HW_OK) {
+ for (i = 0; i < vdev->no_of_vpath; i++) {
+ vpath = &vdev->vpaths[i];
+ if (vpath->handle) {
+ if (vxge_hw_vpath_reset(vpath->handle) == VXGE_HW_OK) {
if (is_vxge_card_up(vdev) &&
vxge_hw_vpath_recover_from_reset(
- vdev->vpaths[i].handle)
- != VXGE_HW_OK) {
+ vpath->handle) != VXGE_HW_OK) {
vxge_debug_init(VXGE_ERR,
"vxge_hw_vpath_recover_"
"from_reset failed for vpath: "
@@ -2019,83 +2021,87 @@ enum vxge_hw_status vxge_reset_all_vpaths(struct vxgedev *vdev)
return status;
}
}
+ }
+
return status;
}
/* close vpaths */
void vxge_close_vpaths(struct vxgedev *vdev, int index)
{
+ struct vxge_vpath *vpath;
int i;
+
for (i = index; i < vdev->no_of_vpath; i++) {
- if (vdev->vpaths[i].handle && vdev->vpaths[i].is_open) {
- vxge_hw_vpath_close(vdev->vpaths[i].handle);
+ vpath = &vdev->vpaths[i];
+
+ if (vpath->handle && vpath->is_open) {
+ vxge_hw_vpath_close(vpath->handle);
vdev->stats.vpaths_open--;
}
- vdev->vpaths[i].is_open = 0;
- vdev->vpaths[i].handle = NULL;
+ vpath->is_open = 0;
+ vpath->handle = NULL;
}
}
/* open vpaths */
int vxge_open_vpaths(struct vxgedev *vdev)
{
+ struct vxge_hw_vpath_attr attr;
enum vxge_hw_status status;
- int i;
+ struct vxge_vpath *vpath;
u32 vp_id = 0;
- struct vxge_hw_vpath_attr attr;
+ int i;
for (i = 0; i < vdev->no_of_vpath; i++) {
- vxge_assert(vdev->vpaths[i].is_configured);
- attr.vp_id = vdev->vpaths[i].device_id;
+ vpath = &vdev->vpaths[i];
+
+ vxge_assert(vpath->is_configured);
+ attr.vp_id = vpath->device_id;
attr.fifo_attr.callback = vxge_xmit_compl;
attr.fifo_attr.txdl_term = vxge_tx_term;
attr.fifo_attr.per_txdl_space = sizeof(struct vxge_tx_priv);
- attr.fifo_attr.userdata = (void *)&vdev->vpaths[i].fifo;
+ attr.fifo_attr.userdata = &vpath->fifo;
attr.ring_attr.callback = vxge_rx_1b_compl;
attr.ring_attr.rxd_init = vxge_rx_initial_replenish;
attr.ring_attr.rxd_term = vxge_rx_term;
attr.ring_attr.per_rxd_space = sizeof(struct vxge_rx_priv);
- attr.ring_attr.userdata = (void *)&vdev->vpaths[i].ring;
+ attr.ring_attr.userdata = &vpath->ring;
- vdev->vpaths[i].ring.ndev = vdev->ndev;
- vdev->vpaths[i].ring.pdev = vdev->pdev;
- status = vxge_hw_vpath_open(vdev->devh, &attr,
- &(vdev->vpaths[i].handle));
+ vpath->ring.ndev = vdev->ndev;
+ vpath->ring.pdev = vdev->pdev;
+ status = vxge_hw_vpath_open(vdev->devh, &attr, &vpath->handle);
if (status == VXGE_HW_OK) {
- vdev->vpaths[i].fifo.handle =
+ vpath->fifo.handle =
(struct __vxge_hw_fifo *)attr.fifo_attr.userdata;
- vdev->vpaths[i].ring.handle =
+ vpath->ring.handle =
(struct __vxge_hw_ring *)attr.ring_attr.userdata;
- vdev->vpaths[i].fifo.tx_steering_type =
+ vpath->fifo.tx_steering_type =
vdev->config.tx_steering_type;
- vdev->vpaths[i].fifo.ndev = vdev->ndev;
- vdev->vpaths[i].fifo.pdev = vdev->pdev;
- vdev->vpaths[i].fifo.indicate_max_pkts =
+ vpath->fifo.ndev = vdev->ndev;
+ vpath->fifo.pdev = vdev->pdev;
+ vpath->fifo.indicate_max_pkts =
vdev->config.fifo_indicate_max_pkts;
- vdev->vpaths[i].ring.rx_vector_no = 0;
- vdev->vpaths[i].ring.rx_csum = vdev->rx_csum;
- vdev->vpaths[i].is_open = 1;
- vdev->vp_handles[i] = vdev->vpaths[i].handle;
- vdev->vpaths[i].ring.gro_enable =
- vdev->config.gro_enable;
- vdev->vpaths[i].ring.vlan_tag_strip =
- vdev->vlan_tag_strip;
+ vpath->ring.rx_vector_no = 0;
+ vpath->ring.rx_csum = vdev->rx_csum;
+ vpath->is_open = 1;
+ vdev->vp_handles[i] = vpath->handle;
+ vpath->ring.gro_enable = vdev->config.gro_enable;
+ vpath->ring.vlan_tag_strip = vdev->vlan_tag_strip;
vdev->stats.vpaths_open++;
} else {
vdev->stats.vpath_open_fail++;
vxge_debug_init(VXGE_ERR,
"%s: vpath: %d failed to open "
"with status: %d",
- vdev->ndev->name, vdev->vpaths[i].device_id,
+ vdev->ndev->name, vpath->device_id,
status);
vxge_close_vpaths(vdev, 0);
return -EPERM;
}
- vp_id =
- ((struct __vxge_hw_vpath_handle *)vdev->vpaths[i].handle)->
- vpath->vp_id;
+ vp_id = vpath->handle->vpath->vp_id;
vdev->vpaths_deployed |= vxge_mBIT(vp_id);
}
return VXGE_HW_OK;
@@ -2269,7 +2275,6 @@ start:
vdev->vxge_entries[j].in_use = 0;
ret = pci_enable_msix(vdev->pdev, vdev->entries, vdev->intr_cnt);
-
if (ret > 0) {
vxge_debug_init(VXGE_ERR,
"%s: MSI-X enable failed for %d vectors, ret: %d",
@@ -2315,17 +2320,16 @@ static int vxge_enable_msix(struct vxgedev *vdev)
ret = vxge_alloc_msix(vdev);
if (!ret) {
for (i = 0; i < vdev->no_of_vpath; i++) {
+ struct vxge_vpath *vpath = &vdev->vpaths[i];
- /* If fifo or ring are not enabled
- the MSIX vector for that should be set to 0
- Hence initializeing this array to all 0s.
- */
- vdev->vpaths[i].ring.rx_vector_no =
- (vdev->vpaths[i].device_id *
- VXGE_HW_VPATH_MSIX_ACTIVE) + 1;
+ /* If fifo or ring are not enabled, the MSIX vector for
+ * it should be set to 0.
+ */
+ vpath->ring.rx_vector_no = (vpath->device_id *
+ VXGE_HW_VPATH_MSIX_ACTIVE) + 1;
- vxge_hw_vpath_msix_set(vdev->vpaths[i].handle,
- tim_msix_id, VXGE_ALARM_MSIX_ID);
+ vxge_hw_vpath_msix_set(vpath->handle, tim_msix_id,
+ VXGE_ALARM_MSIX_ID);
}
}
@@ -2540,9 +2544,10 @@ static void vxge_poll_vp_reset(unsigned long data)
static void vxge_poll_vp_lockup(unsigned long data)
{
struct vxgedev *vdev = (struct vxgedev *)data;
- int i;
- struct vxge_ring *ring;
enum vxge_hw_status status = VXGE_HW_OK;
+ struct vxge_vpath *vpath;
+ struct vxge_ring *ring;
+ int i;
for (i = 0; i < vdev->no_of_vpath; i++) {
ring = &vdev->vpaths[i].ring;
@@ -2556,13 +2561,13 @@ static void vxge_poll_vp_lockup(unsigned long data)
/* schedule vpath reset */
if (!test_and_set_bit(i, &vdev->vp_reset)) {
+ vpath = &vdev->vpaths[i];
/* disable interrupts for this vpath */
vxge_vpath_intr_disable(vdev, i);
/* stop the queue for this vpath */
- vxge_stop_tx_queue(&vdev->vpaths[i].
- fifo);
+ vxge_stop_tx_queue(&vpath->fifo);
continue;
}
}
@@ -2591,6 +2596,7 @@ vxge_open(struct net_device *dev)
enum vxge_hw_status status;
struct vxgedev *vdev;
struct __vxge_hw_device *hldev;
+ struct vxge_vpath *vpath;
int ret = 0;
int i;
u64 val64, function_mode;
@@ -2629,15 +2635,17 @@ vxge_open(struct net_device *dev)
netif_napi_add(dev, &vdev->napi, vxge_poll_inta,
vdev->config.napi_weight);
napi_enable(&vdev->napi);
- for (i = 0; i < vdev->no_of_vpath; i++)
- vdev->vpaths[i].ring.napi_p = &vdev->napi;
+ for (i = 0; i < vdev->no_of_vpath; i++) {
+ vpath = &vdev->vpaths[i];
+ vpath->ring.napi_p = &vdev->napi;
+ }
} else {
for (i = 0; i < vdev->no_of_vpath; i++) {
- netif_napi_add(dev, &vdev->vpaths[i].ring.napi,
+ vpath = &vdev->vpaths[i];
+ netif_napi_add(dev, &vpath->ring.napi,
vxge_poll_msix, vdev->config.napi_weight);
- napi_enable(&vdev->vpaths[i].ring.napi);
- vdev->vpaths[i].ring.napi_p =
- &vdev->vpaths[i].ring.napi;
+ napi_enable(&vpath->ring.napi);
+ vpath->ring.napi_p = &vpath->ring.napi;
}
}
@@ -2654,9 +2662,10 @@ vxge_open(struct net_device *dev)
}
for (i = 0; i < vdev->no_of_vpath; i++) {
+ vpath = &vdev->vpaths[i];
+
/* set initial mtu before enabling the device */
- status = vxge_hw_vpath_mtu_set(vdev->vpaths[i].handle,
- vdev->mtu);
+ status = vxge_hw_vpath_mtu_set(vpath->handle, vdev->mtu);
if (status != VXGE_HW_OK) {
vxge_debug_init(VXGE_ERR,
"%s: fatal: can not set new MTU", dev->name);
@@ -2670,10 +2679,21 @@ vxge_open(struct net_device *dev)
"%s: MTU is %d", vdev->ndev->name, vdev->mtu);
VXGE_DEVICE_DEBUG_LEVEL_SET(VXGE_ERR, VXGE_COMPONENT_LL, vdev);
- /* Reprogram the DA table with populated mac addresses */
- for (i = 0; i < vdev->no_of_vpath; i++) {
- vxge_restore_vpath_mac_addr(&vdev->vpaths[i]);
- vxge_restore_vpath_vid_table(&vdev->vpaths[i]);
+ /* Restore the DA, VID table and also multicast and promiscuous mode
+ * states
+ */
+ if (vdev->all_multi_flg) {
+ for (i = 0; i < vdev->no_of_vpath; i++) {
+ vpath = &vdev->vpaths[i];
+ vxge_restore_vpath_mac_addr(vpath);
+ vxge_restore_vpath_vid_table(vpath);
+
+ status = vxge_hw_vpath_mcast_enable(vpath->handle);
+ if (status != VXGE_HW_OK)
+ vxge_debug_init(VXGE_ERR,
+ "%s:%d Enabling multicast failed",
+ __func__, __LINE__);
+ }
}
/* Enable vpath to sniff all unicast/multicast traffic that not
@@ -2702,14 +2722,14 @@ vxge_open(struct net_device *dev)
/* Enabling Bcast and mcast for all vpath */
for (i = 0; i < vdev->no_of_vpath; i++) {
- status = vxge_hw_vpath_bcast_enable(vdev->vpaths[i].handle);
+ vpath = &vdev->vpaths[i];
+ status = vxge_hw_vpath_bcast_enable(vpath->handle);
if (status != VXGE_HW_OK)
vxge_debug_init(VXGE_ERR,
"%s : Can not enable bcast for vpath "
"id %d", dev->name, i);
if (vdev->config.addr_learn_en) {
- status =
- vxge_hw_vpath_mcast_enable(vdev->vpaths[i].handle);
+ status = vxge_hw_vpath_mcast_enable(vpath->handle);
if (status != VXGE_HW_OK)
vxge_debug_init(VXGE_ERR,
"%s : Can not enable mcast for vpath "
@@ -2744,9 +2764,11 @@ vxge_open(struct net_device *dev)
smp_wmb();
for (i = 0; i < vdev->no_of_vpath; i++) {
- vxge_hw_vpath_enable(vdev->vpaths[i].handle);
+ vpath = &vdev->vpaths[i];
+
+ vxge_hw_vpath_enable(vpath->handle);
smp_wmb();
- vxge_hw_vpath_rx_doorbell_init(vdev->vpaths[i].handle);
+ vxge_hw_vpath_rx_doorbell_init(vpath->handle);
}
netif_tx_start_all_queues(vdev->ndev);
^ permalink raw reply related
* [PATCH 4/6 net-next-2.6] vxge: Update copyright information
From: Jon Mason @ 2010-07-08 19:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Sreenivasa Honnur, Ramkrishna Vepa
Update copyright information to reflect the Exar purchase of Neterion
Signed-off-by: Jon Mason <jon.mason@exar.com>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@exar.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@exar.com>
---
drivers/net/s2io-regs.h | 2 +-
drivers/net/s2io.c | 4 ++--
drivers/net/s2io.h | 2 +-
drivers/net/vxge/Makefile | 2 +-
drivers/net/vxge/vxge-config.c | 4 ++--
drivers/net/vxge/vxge-config.h | 4 ++--
drivers/net/vxge/vxge-ethtool.c | 4 ++--
drivers/net/vxge/vxge-ethtool.h | 4 ++--
drivers/net/vxge/vxge-main.c | 6 +++---
drivers/net/vxge/vxge-main.h | 4 ++--
drivers/net/vxge/vxge-reg.h | 4 ++--
drivers/net/vxge/vxge-traffic.c | 4 ++--
drivers/net/vxge/vxge-traffic.h | 4 ++--
drivers/net/vxge/vxge-version.h | 4 ++--
14 files changed, 26 insertions(+), 26 deletions(-)
diff --git a/drivers/net/s2io-regs.h b/drivers/net/s2io-regs.h
index 416669f..3688325 100644
--- a/drivers/net/s2io-regs.h
+++ b/drivers/net/s2io-regs.h
@@ -1,6 +1,6 @@
/************************************************************************
* regs.h: A Linux PCI-X Ethernet driver for Neterion 10GbE Server NIC
- * Copyright(c) 2002-2007 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
* This software may be used and distributed according to the terms of
* the GNU General Public License (GPL), incorporated herein by reference.
diff --git a/drivers/net/s2io.c b/drivers/net/s2io.c
index d0af924..f9f9ed8 100644
--- a/drivers/net/s2io.c
+++ b/drivers/net/s2io.c
@@ -1,6 +1,6 @@
/************************************************************************
* s2io.c: A Linux PCI-X Ethernet driver for Neterion 10GbE Server NIC
- * Copyright(c) 2002-2007 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
*
* This software may be used and distributed according to the terms of
* the GNU General Public License (GPL), incorporated herein by reference.
@@ -8199,7 +8199,7 @@ s2io_init_nic(struct pci_dev *pdev, const struct pci_device_id *pre)
goto register_failed;
}
s2io_vpd_read(sp);
- DBG_PRINT(ERR_DBG, "Copyright(c) 2002-2007 Neterion Inc.\n");
+ DBG_PRINT(ERR_DBG, "Copyright(c) 2002-2010 Exar Corp.\n");
DBG_PRINT(ERR_DBG, "%s: Neterion %s (rev %d)\n", dev->name,
sp->product_name, pdev->revision);
DBG_PRINT(ERR_DBG, "%s: Driver version %s\n", dev->name,
diff --git a/drivers/net/s2io.h b/drivers/net/s2io.h
index 5e52c75..3645fb3 100644
--- a/drivers/net/s2io.h
+++ b/drivers/net/s2io.h
@@ -1,6 +1,6 @@
/************************************************************************
* s2io.h: A Linux PCI-X Ethernet driver for Neterion 10GbE Server NIC
- * Copyright(c) 2002-2007 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
* This software may be used and distributed according to the terms of
* the GNU General Public License (GPL), incorporated herein by reference.
diff --git a/drivers/net/vxge/Makefile b/drivers/net/vxge/Makefile
index 8992ca2..b625e2c 100644
--- a/drivers/net/vxge/Makefile
+++ b/drivers/net/vxge/Makefile
@@ -1,5 +1,5 @@
#
-# Makefile for Neterion Inc's X3100 Series 10 GbE PCIe # I/O
+# Makefile for Exar Corp's X3100 Series 10 GbE PCIe I/O
# Virtualized Server Adapter linux driver
obj-$(CONFIG_VXGE) += vxge.o
diff --git a/drivers/net/vxge/vxge-config.c b/drivers/net/vxge/vxge-config.c
index 297f0d2..0e6db59 100644
--- a/drivers/net/vxge/vxge-config.c
+++ b/drivers/net/vxge/vxge-config.c
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-config.c: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+ * vxge-config.c: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#include <linux/vmalloc.h>
#include <linux/etherdevice.h>
diff --git a/drivers/net/vxge/vxge-config.h b/drivers/net/vxge/vxge-config.h
index 4ae2625..1a94343 100644
--- a/drivers/net/vxge/vxge-config.h
+++ b/drivers/net/vxge/vxge-config.h
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-config.h: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+ * vxge-config.h: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#ifndef VXGE_CONFIG_H
#define VXGE_CONFIG_H
diff --git a/drivers/net/vxge/vxge-ethtool.c b/drivers/net/vxge/vxge-ethtool.c
index cadef85..05679e3 100644
--- a/drivers/net/vxge/vxge-ethtool.c
+++ b/drivers/net/vxge/vxge-ethtool.c
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-ethtool.c: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+ * vxge-ethtool.c: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#include<linux/ethtool.h>
#include <linux/slab.h>
diff --git a/drivers/net/vxge/vxge-ethtool.h b/drivers/net/vxge/vxge-ethtool.h
index 1c3df0a..6cf3044 100644
--- a/drivers/net/vxge/vxge-ethtool.h
+++ b/drivers/net/vxge/vxge-ethtool.h
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-ethtool.h: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+ * vxge-ethtool.h: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#ifndef _VXGE_ETHTOOL_H
#define _VXGE_ETHTOOL_H
diff --git a/drivers/net/vxge/vxge-main.c b/drivers/net/vxge/vxge-main.c
index f921196..df31e68 100644
--- a/drivers/net/vxge/vxge-main.c
+++ b/drivers/net/vxge/vxge-main.c
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
-* vxge-main.c: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+*vxge-main.c : Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
-* Copyright(c) 2002-2009 Neterion Inc.
+* Copyright(c) 2002-2010 Exar Corp.
*
* The module loadable parameters that are supported by the driver and a brief
* explanation of all the variables:
@@ -4515,7 +4515,7 @@ vxge_starter(void)
char version[32];
snprintf(version, 32, "%s", DRV_VERSION);
- printk(KERN_INFO "%s: Copyright(c) 2002-2009 Neterion Inc\n",
+ printk(KERN_CRIT "%s: Copyright(c) 2002-2010 Exar Corp.\n",
VXGE_DRIVER_NAME);
printk(KERN_INFO "%s: Driver version: %s\n",
VXGE_DRIVER_NAME, version);
diff --git a/drivers/net/vxge/vxge-main.h b/drivers/net/vxge/vxge-main.h
index 104f442..6b249a4 100644
--- a/drivers/net/vxge/vxge-main.h
+++ b/drivers/net/vxge/vxge-main.h
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-main.h: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+ * vxge-main.h: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#ifndef VXGE_MAIN_H
#define VXGE_MAIN_H
diff --git a/drivers/net/vxge/vxge-reg.h b/drivers/net/vxge/vxge-reg.h
index 9a0cf8e..3dd5c96 100644
--- a/drivers/net/vxge/vxge-reg.h
+++ b/drivers/net/vxge/vxge-reg.h
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-reg.h: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O Virtualized
+ * vxge-reg.h: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O Virtualized
* Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#ifndef VXGE_REG_H
#define VXGE_REG_H
diff --git a/drivers/net/vxge/vxge-traffic.c b/drivers/net/vxge/vxge-traffic.c
index 6cc1dd7..2a68fa3 100644
--- a/drivers/net/vxge/vxge-traffic.c
+++ b/drivers/net/vxge/vxge-traffic.c
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-traffic.c: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+ * vxge-traffic.c: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#include <linux/etherdevice.h>
diff --git a/drivers/net/vxge/vxge-traffic.h b/drivers/net/vxge/vxge-traffic.h
index c252f3d..6fa07d1 100644
--- a/drivers/net/vxge/vxge-traffic.h
+++ b/drivers/net/vxge/vxge-traffic.h
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-traffic.h: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+ * vxge-traffic.h: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#ifndef VXGE_TRAFFIC_H
#define VXGE_TRAFFIC_H
diff --git a/drivers/net/vxge/vxge-version.h b/drivers/net/vxge/vxge-version.h
index 5da7ab1..4008e6b 100644
--- a/drivers/net/vxge/vxge-version.h
+++ b/drivers/net/vxge/vxge-version.h
@@ -7,9 +7,9 @@
* system is licensed under the GPL.
* See the file COPYING in this distribution for more information.
*
- * vxge-version.h: Driver for Neterion Inc's X3100 Series 10GbE PCIe I/O
+ * vxge-version.h: Driver for Exar Corp's X3100 Series 10GbE PCIe I/O
* Virtualized Server Adapter.
- * Copyright(c) 2002-2009 Neterion Inc.
+ * Copyright(c) 2002-2010 Exar Corp.
******************************************************************************/
#ifndef VXGE_VERSION_H
^ permalink raw reply related
* [PATCH 2/6 net-next-2.6] vxge: Use fifo based trans_start time
From: Jon Mason @ 2010-07-08 19:21 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Sreenivasa Honnur, Ramkrishna Vepa
Use the fifo based netdev_queue->trans_start time in vxge_xmit, instead
of the device based netdev->trans_start time.
Signed-off-by: Jon Mason <jon.mason@exar.com>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@exar.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@exar.com>
---
drivers/net/vxge/vxge-main.c | 5 ++++-
1 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/drivers/net/vxge/vxge-main.c b/drivers/net/vxge/vxge-main.c
index e78703d..b8eed71 100644
--- a/drivers/net/vxge/vxge-main.c
+++ b/drivers/net/vxge/vxge-main.c
@@ -802,6 +802,7 @@ vxge_xmit(struct sk_buff *skb, struct net_device *dev)
unsigned long flags = 0;
int vpath_no = 0;
int do_spin_tx_lock = 1;
+ struct netdev_queue *txq;
vxge_debug_entryexit(VXGE_TRACE, "%s: %s:%d",
dev->name, __func__, __LINE__);
@@ -968,8 +969,10 @@ vxge_xmit(struct sk_buff *skb, struct net_device *dev)
VXGE_HW_FIFO_TXD_TX_CKO_UDP_EN);
vxge_hw_fifo_txdl_post(fifo_hw, dtr);
+
#ifdef NETIF_F_LLTX
- dev->trans_start = jiffies; /* NETIF_F_LLTX driver :( */
+ txq = netdev_get_tx_queue(dev, vpath_no);
+ txq->trans_start = jiffies;
#endif
spin_unlock_irqrestore(&fifo->tx_lock, flags);
^ permalink raw reply related
* [PATCH 1/6 net-next-2.6] vxge: Remove queue_state references
From: Jon Mason @ 2010-07-08 19:19 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Sreenivasa Honnur, Ramkrishna Vepa
Remove queue_state references, as they are no longer necessary.
Also, The driver needs to start/stop the queue regardless of which type
of steering is enabled. Remove checks for TX_MULTIQ_STEERING only and
start/stop for all steering types.
Signed-off-by: Jon Mason <jon.mason@exar.com>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@exar.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@exar.com>
---
drivers/net/vxge/vxge-main.c | 118 +++++++++++++++---------------------------
drivers/net/vxge/vxge-main.h | 10 +---
2 files changed, 42 insertions(+), 86 deletions(-)
diff --git a/drivers/net/vxge/vxge-main.c b/drivers/net/vxge/vxge-main.c
index ed17865..e78703d 100644
--- a/drivers/net/vxge/vxge-main.c
+++ b/drivers/net/vxge/vxge-main.c
@@ -133,75 +133,48 @@ static inline void VXGE_COMPLETE_ALL_RX(struct vxgedev *vdev)
/*
* MultiQ manipulation helper functions
*/
-void vxge_stop_all_tx_queue(struct vxgedev *vdev)
+static inline int vxge_netif_queue_stopped(struct vxge_fifo *fifo)
{
- int i;
- struct net_device *dev = vdev->ndev;
+ struct net_device *dev = fifo->ndev;
+ struct netdev_queue *txq = NULL;
+ int vpath_no = fifo->driver_id;
+ int ret = 0;
- if (vdev->config.tx_steering_type != TX_MULTIQ_STEERING) {
- for (i = 0; i < vdev->no_of_vpath; i++)
- vdev->vpaths[i].fifo.queue_state = VPATH_QUEUE_STOP;
- }
- netif_tx_stop_all_queues(dev);
+ if (fifo->tx_steering_type)
+ txq = netdev_get_tx_queue(dev, vpath_no);
+ else
+ txq = netdev_get_tx_queue(dev, 0);
+
+ ret = netif_tx_queue_stopped(txq);
+ return ret;
}
void vxge_stop_tx_queue(struct vxge_fifo *fifo)
{
struct net_device *dev = fifo->ndev;
-
struct netdev_queue *txq = NULL;
- if (fifo->tx_steering_type == TX_MULTIQ_STEERING)
+
+ if (fifo->tx_steering_type)
txq = netdev_get_tx_queue(dev, fifo->driver_id);
- else {
+ else
txq = netdev_get_tx_queue(dev, 0);
- fifo->queue_state = VPATH_QUEUE_STOP;
- }
netif_tx_stop_queue(txq);
}
-void vxge_start_all_tx_queue(struct vxgedev *vdev)
-{
- int i;
- struct net_device *dev = vdev->ndev;
-
- if (vdev->config.tx_steering_type != TX_MULTIQ_STEERING) {
- for (i = 0; i < vdev->no_of_vpath; i++)
- vdev->vpaths[i].fifo.queue_state = VPATH_QUEUE_START;
- }
- netif_tx_start_all_queues(dev);
-}
-
-static void vxge_wake_all_tx_queue(struct vxgedev *vdev)
-{
- int i;
- struct net_device *dev = vdev->ndev;
-
- if (vdev->config.tx_steering_type != TX_MULTIQ_STEERING) {
- for (i = 0; i < vdev->no_of_vpath; i++)
- vdev->vpaths[i].fifo.queue_state = VPATH_QUEUE_START;
- }
- netif_tx_wake_all_queues(dev);
-}
-
-void vxge_wake_tx_queue(struct vxge_fifo *fifo, struct sk_buff *skb)
+void vxge_wake_tx_queue(struct vxge_fifo *fifo)
{
struct net_device *dev = fifo->ndev;
-
- int vpath_no = fifo->driver_id;
struct netdev_queue *txq = NULL;
- if (fifo->tx_steering_type == TX_MULTIQ_STEERING) {
+ int vpath_no = fifo->driver_id;
+
+ if (fifo->tx_steering_type)
txq = netdev_get_tx_queue(dev, vpath_no);
- if (netif_tx_queue_stopped(txq))
- netif_tx_wake_queue(txq);
- } else {
+ else
txq = netdev_get_tx_queue(dev, 0);
- if (fifo->queue_state == VPATH_QUEUE_STOP)
- if (netif_tx_queue_stopped(txq)) {
- fifo->queue_state = VPATH_QUEUE_START;
- netif_tx_wake_queue(txq);
- }
- }
+
+ if (netif_tx_queue_stopped(txq))
+ netif_tx_wake_queue(txq);
}
/*
@@ -222,7 +195,7 @@ vxge_callback_link_up(struct __vxge_hw_device *hldev)
vdev->stats.link_up++;
netif_carrier_on(vdev->ndev);
- vxge_wake_all_tx_queue(vdev);
+ netif_tx_wake_all_queues(vdev->ndev);
vxge_debug_entryexit(VXGE_TRACE,
"%s: %s:%d Exiting...", vdev->ndev->name, __func__, __LINE__);
@@ -246,7 +219,7 @@ vxge_callback_link_down(struct __vxge_hw_device *hldev)
vdev->stats.link_down++;
netif_carrier_off(vdev->ndev);
- vxge_stop_all_tx_queue(vdev);
+ netif_tx_stop_all_queues(vdev->ndev);
vxge_debug_entryexit(VXGE_TRACE,
"%s: %s:%d Exiting...", vdev->ndev->name, __func__, __LINE__);
@@ -677,7 +650,7 @@ vxge_xmit_compl(struct __vxge_hw_fifo *fifo_hw, void *dtr,
&dtr, &t_code) == VXGE_HW_OK);
*skb_ptr = done_skb;
- vxge_wake_tx_queue(fifo, skb);
+ vxge_wake_tx_queue(fifo);
vxge_debug_entryexit(VXGE_TRACE,
"%s: %s:%d Exiting...",
@@ -881,17 +854,11 @@ vxge_xmit(struct sk_buff *skb, struct net_device *dev)
return NETDEV_TX_LOCKED;
}
- if (vdev->config.tx_steering_type == TX_MULTIQ_STEERING) {
- if (netif_subqueue_stopped(dev, skb)) {
- spin_unlock_irqrestore(&fifo->tx_lock, flags);
- return NETDEV_TX_BUSY;
- }
- } else if (unlikely(fifo->queue_state == VPATH_QUEUE_STOP)) {
- if (netif_queue_stopped(dev)) {
- spin_unlock_irqrestore(&fifo->tx_lock, flags);
- return NETDEV_TX_BUSY;
- }
+ if (vxge_netif_queue_stopped(fifo)) {
+ spin_unlock_irqrestore(&fifo->tx_lock, flags);
+ return NETDEV_TX_BUSY;
}
+
avail = vxge_hw_fifo_free_txdl_count_get(fifo_hw);
if (avail == 0) {
vxge_debug_tx(VXGE_ERR,
@@ -1478,7 +1445,7 @@ static int vxge_reset_vpath(struct vxgedev *vdev, int vp_id)
clear_bit(vp_id, &vdev->vp_reset);
/* Start the vpath queue */
- vxge_wake_tx_queue(&vdev->vpaths[vp_id].fifo, NULL);
+ vxge_wake_tx_queue(&vdev->vpaths[vp_id].fifo);
return ret;
}
@@ -1513,7 +1480,7 @@ static int do_vxge_reset(struct vxgedev *vdev, int event)
"%s: execution mode is debug, returning..",
vdev->ndev->name);
clear_bit(__VXGE_STATE_CARD_UP, &vdev->state);
- vxge_stop_all_tx_queue(vdev);
+ netif_tx_stop_all_queues(vdev->ndev);
return 0;
}
}
@@ -1523,7 +1490,7 @@ static int do_vxge_reset(struct vxgedev *vdev, int event)
switch (vdev->cric_err_event) {
case VXGE_HW_EVENT_UNKNOWN:
- vxge_stop_all_tx_queue(vdev);
+ netif_tx_stop_all_queues(vdev->ndev);
vxge_debug_init(VXGE_ERR,
"fatal: %s: Disabling device due to"
"unknown error",
@@ -1544,7 +1511,7 @@ static int do_vxge_reset(struct vxgedev *vdev, int event)
case VXGE_HW_EVENT_VPATH_ERR:
break;
case VXGE_HW_EVENT_CRITICAL_ERR:
- vxge_stop_all_tx_queue(vdev);
+ netif_tx_stop_all_queues(vdev->ndev);
vxge_debug_init(VXGE_ERR,
"fatal: %s: Disabling device due to"
"serious error",
@@ -1554,7 +1521,7 @@ static int do_vxge_reset(struct vxgedev *vdev, int event)
ret = -EPERM;
goto out;
case VXGE_HW_EVENT_SERR:
- vxge_stop_all_tx_queue(vdev);
+ netif_tx_stop_all_queues(vdev->ndev);
vxge_debug_init(VXGE_ERR,
"fatal: %s: Disabling device due to"
"serious error",
@@ -1566,7 +1533,7 @@ static int do_vxge_reset(struct vxgedev *vdev, int event)
ret = -EPERM;
goto out;
case VXGE_HW_EVENT_SLOT_FREEZE:
- vxge_stop_all_tx_queue(vdev);
+ netif_tx_stop_all_queues(vdev->ndev);
vxge_debug_init(VXGE_ERR,
"fatal: %s: Disabling device due to"
"slot freeze",
@@ -1580,7 +1547,7 @@ static int do_vxge_reset(struct vxgedev *vdev, int event)
}
if ((event == VXGE_LL_FULL_RESET) || (event == VXGE_LL_START_RESET))
- vxge_stop_all_tx_queue(vdev);
+ netif_tx_stop_all_queues(vdev->ndev);
if (event == VXGE_LL_FULL_RESET) {
status = vxge_reset_all_vpaths(vdev);
@@ -1640,7 +1607,7 @@ static int do_vxge_reset(struct vxgedev *vdev, int event)
vxge_hw_vpath_rx_doorbell_init(vdev->vpaths[i].handle);
}
- vxge_wake_all_tx_queue(vdev);
+ netif_tx_wake_all_queues(vdev->ndev);
}
out:
@@ -2779,7 +2746,7 @@ vxge_open(struct net_device *dev)
vxge_hw_vpath_rx_doorbell_init(vdev->vpaths[i].handle);
}
- vxge_start_all_tx_queue(vdev);
+ netif_tx_start_all_queues(vdev->ndev);
goto out0;
out2:
@@ -2902,7 +2869,7 @@ int do_vxge_close(struct net_device *dev, int do_io)
netif_carrier_off(vdev->ndev);
printk(KERN_NOTICE "%s: Link Down\n", vdev->ndev->name);
- vxge_stop_all_tx_queue(vdev);
+ netif_tx_stop_all_queues(vdev->ndev);
/* Note that at this point xmit() is stopped by upper layer */
if (do_io)
@@ -3215,7 +3182,7 @@ int __devinit vxge_device_register(struct __vxge_hw_device *hldev,
u64 stat;
*vdev_out = NULL;
- if (config->tx_steering_type == TX_MULTIQ_STEERING)
+ if (config->tx_steering_type)
no_of_queue = no_of_vpath;
ndev = alloc_etherdev_mq(sizeof(struct vxgedev),
@@ -3284,9 +3251,6 @@ int __devinit vxge_device_register(struct __vxge_hw_device *hldev,
if (vdev->config.gro_enable)
ndev->features |= NETIF_F_GRO;
- if (vdev->config.tx_steering_type == TX_MULTIQ_STEERING)
- ndev->real_num_tx_queues = no_of_vpath;
-
#ifdef NETIF_F_LLTX
ndev->features |= NETIF_F_LLTX;
#endif
diff --git a/drivers/net/vxge/vxge-main.h b/drivers/net/vxge/vxge-main.h
index 60276b2..a384582 100644
--- a/drivers/net/vxge/vxge-main.h
+++ b/drivers/net/vxge/vxge-main.h
@@ -228,10 +228,6 @@ struct vxge_fifo {
int tx_steering_type;
int indicate_max_pkts;
spinlock_t tx_lock;
- /* flag used to maintain queue state when MULTIQ is not enabled */
-#define VPATH_QUEUE_START 0
-#define VPATH_QUEUE_STOP 1
- int queue_state;
/* Tx stats */
struct vxge_fifo_stats stats;
@@ -447,13 +443,9 @@ int vxge_open_vpaths(struct vxgedev *vdev);
enum vxge_hw_status vxge_reset_all_vpaths(struct vxgedev *vdev);
-void vxge_stop_all_tx_queue(struct vxgedev *vdev);
-
void vxge_stop_tx_queue(struct vxge_fifo *fifo);
-void vxge_start_all_tx_queue(struct vxgedev *vdev);
-
-void vxge_wake_tx_queue(struct vxge_fifo *fifo, struct sk_buff *skb);
+void vxge_wake_tx_queue(struct vxge_fifo *fifo);
enum vxge_hw_status vxge_add_mac_addr(struct vxgedev *vdev,
struct macInfo *mac);
^ permalink raw reply related
* RE: [PATCH linux-2.6.35-rc3] ks8842 driver
From: Choi, David @ 2010-07-08 19:01 UTC (permalink / raw)
To: Simon Horman; +Cc: netdev, Li, Charles
In-Reply-To: <20100708075245.GA21371@verge.net.au>
Hi Simon,
Thank you for your comments.
The original ks8842 driver is designed to work on the customized bus
interface based on an FPGA. This patch is intended to address the more
commonly used generic bus interface available on the majority of SoC in
the market.
It is unlikely that for a system to use both FPGA based and generic bus
interface for ks8842, I am quite certain that those 2 devices are used
mutual exclusively.
Regards,
David J. Choi
-----Original Message-----
From: Simon Horman [mailto:horms@verge.net.au]
Sent: Thursday, July 08, 2010 12:53 AM
To: Choi, David
Cc: netdev@vger.kernel.org; Li, Charles
Subject: Re: [PATCH linux-2.6.35-rc3] ks8842 driver
On Wed, Jul 07, 2010 at 08:24:33AM -0700, Choi, David wrote:
> To whom it may have concerned,
>
> It seems that there are differences between Micrel ks8842 device and
Timberdale FPGA based device with generic bus interface. In order to
check the differences, I have sent several times but have not received
any response. This patch is to support ks8841/ks8842 device from Micrel.
Is it desirable that the code can never be compiled to work with both
devices?
Is this code likely to be included in a generic/distro kernel?
>
> From: David J. Choi <david.choi@micrel.com>
>
> Body of the explanation:
> -support 16bit and 32bit bus width.
> -add device reset for ks8842/8841 Micrel device.
> -set 100Mbps as a default for Micrel device.
> -set MAC address in both MAC/Switch layer with different sequence
for Micrel device,
> as mentioned in data sheet.
>
> Signed-off-by: David J. Choi <david.choi@micrel.com>
>
> ---
> --- linux-2.6.35-rc3/drivers/net/ks8842.c.orig 2010-07-01
16:26:50.000000000 -0700
> +++ linux-2.6.35-rc3/drivers/net/ks8842.c 2010-07-07
07:41:03.000000000 -0700
> @@ -18,6 +18,7 @@
>
> /* Supports:
> * The Micrel KS8842 behind the timberdale FPGA
> + * The genuine Micrel KS8841/42 device with ISA 16/32bit bus
interface
> */
>
> #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
> @@ -191,6 +192,12 @@ static inline u32 ks8842_read32(struct k
>
> static void ks8842_reset(struct ks8842_adapter *adapter)
> {
> +#ifdef CONFIG_MICREL_KS884X
> + ks8842_write16(adapter, 3, 1, REG_GRR);
> + msleep(10);
> + iowrite16(0, adapter->hw_addr + REG_GRR);
> +
> +#else
> /* The KS8842 goes haywire when doing softare reset
> * a work around in the timberdale IP is implemented to
> * do a hardware reset instead
> @@ -201,6 +208,7 @@ static void ks8842_reset(struct ks8842_a
> iowrite16(32, adapter->hw_addr + REG_SELECT_BANK);
> iowrite32(0x1, adapter->hw_addr + REG_TIMB_RST);
> msleep(20);
> +#endif
> }
>
> static void ks8842_update_link_status(struct net_device *netdev,
> @@ -269,8 +277,11 @@ static void ks8842_reset_hw(struct ks884
>
> /* restart port auto-negotiation */
> ks8842_enable_bits(adapter, 49, 1 << 13, REG_P1CR4);
> +
> +#ifndef CONFIG_MICREL_KS884X
> /* only advertise 10Mbps */
> ks8842_clear_bits(adapter, 49, 3 << 2, REG_P1CR4);
> +#endif
>
> /* Enable the transmitter */
> ks8842_enable_tx(adapter);
> @@ -296,6 +307,20 @@ static void ks8842_read_mac_addr(struct
> for (i = 0; i < ETH_ALEN; i++)
> dest[ETH_ALEN - i - 1] = ks8842_read8(adapter, 2,
REG_MARL + i);
>
> +#ifdef CONFIG_MICREL_KS884X
> + /*
> + the sequence of saving mac addr between MAC and Switch
is
> + different.
> + */
> +
> + mac = ks8842_read16(adapter, 2, REG_MARL);
> + ks8842_write16(adapter, 39, mac, REG_MACAR3);
> + mac = ks8842_read16(adapter, 2, REG_MARM);
> + ks8842_write16(adapter, 39, mac, REG_MACAR2);
> + mac = ks8842_read16(adapter, 2, REG_MARH);
> + ks8842_write16(adapter, 39, mac, REG_MACAR1);
> +#else
> +
> /* make sure the switch port uses the same MAC as the QMU */
> mac = ks8842_read16(adapter, 2, REG_MARL);
> ks8842_write16(adapter, 39, mac, REG_MACAR1);
> @@ -303,6 +328,7 @@ static void ks8842_read_mac_addr(struct
> ks8842_write16(adapter, 39, mac, REG_MACAR2);
> mac = ks8842_read16(adapter, 2, REG_MARH);
> ks8842_write16(adapter, 39, mac, REG_MACAR3);
> +#endif
> }
>
> static void ks8842_write_mac_addr(struct ks8842_adapter *adapter, u8
*mac)
> @@ -313,8 +339,13 @@ static void ks8842_write_mac_addr(struct
> spin_lock_irqsave(&adapter->lock, flags);
> for (i = 0; i < ETH_ALEN; i++) {
> ks8842_write8(adapter, 2, mac[ETH_ALEN - i - 1],
REG_MARL + i);
> +#ifdef CONFIG_MICREL_KS884X
> + ks8842_write8(adapter, 39, mac[ETH_ALEN - i - 1],
> + REG_MACAR3 + 1 - i);
> +#else
> ks8842_write8(adapter, 39, mac[ETH_ALEN - i - 1],
> REG_MACAR1 + i);
> +#endif
> }
> spin_unlock_irqrestore(&adapter->lock, flags);
> }
> @@ -328,8 +359,12 @@ static int ks8842_tx_frame(struct sk_buf
> {
> struct ks8842_adapter *adapter = netdev_priv(netdev);
> int len = skb->len;
> +#ifdef CONFIG_KS884X_16BIT
> + u16 *ptr16 = (u16 *)skb->data;
> +#else
> u32 *ptr = (u32 *)skb->data;
> u32 ctrl;
> +#endif
>
> dev_dbg(&adapter->pdev->dev,
> "%s: len %u head %p data %p tail %p end %p\n",
> @@ -340,6 +375,18 @@ static int ks8842_tx_frame(struct sk_buf
> if (ks8842_tx_fifo_space(adapter) < len + 8)
> return NETDEV_TX_BUSY;
>
> +#ifdef CONFIG_KS884X_16BIT
> + ks8842_write16(adapter, 17, 0x8000 | 0x100, REG_QMU_DATA_LO);
> + ks8842_write16(adapter, 17, (u16)len, REG_QMU_DATA_HI);
> + netdev->stats.tx_bytes += len;
> +
> + /* copy buffer */
> + while (len > 0) {
> + iowrite16(*ptr16++, adapter->hw_addr + REG_QMU_DATA_LO);
> + iowrite16(*ptr16++, adapter->hw_addr + REG_QMU_DATA_HI);
> + len -= sizeof(u32);
> + }
> +#else
> /* the control word, enable IRQ, port 1 and the length */
> ctrl = 0x8000 | 0x100 | (len << 16);
> ks8842_write32(adapter, 17, ctrl, REG_QMU_DATA_LO);
> @@ -352,6 +399,7 @@ static int ks8842_tx_frame(struct sk_buf
> len -= sizeof(u32);
> ptr++;
> }
> +#endif
>
> /* enqueue packet */
> ks8842_write16(adapter, 17, 1, REG_TXQCR);
> @@ -364,10 +412,15 @@ static int ks8842_tx_frame(struct sk_buf
> static void ks8842_rx_frame(struct net_device *netdev,
> struct ks8842_adapter *adapter)
> {
> +#ifdef CONFIG_KS884X_16BIT
> + u16 status = ks8842_read16(adapter, 17, REG_QMU_DATA_LO);
> + int len = (int)ks8842_read16(adapter, 17, REG_QMU_DATA_HI) &
0xffff;
> +#else
> u32 status = ks8842_read32(adapter, 17, REG_QMU_DATA_LO);
> int len = (status >> 16) & 0x7ff;
>
> status &= 0xffff;
> +#endif
>
> dev_dbg(&adapter->pdev->dev, "%s - rx_data: status: %x\n",
> __func__, status);
> @@ -379,13 +432,28 @@ static void ks8842_rx_frame(struct net_d
> dev_dbg(&adapter->pdev->dev, "%s, got package, len:
%d\n",
> __func__, len);
> if (skb) {
> +#ifdef CONFIG_KS884X_16BIT
> + u16 *data16;
> +#else
> u32 *data;
> +#endif
>
> netdev->stats.rx_packets++;
> netdev->stats.rx_bytes += len;
> if (status & RXSR_MULTICAST)
> netdev->stats.multicast++;
>
> +#ifdef CONFIG_KS884X_16BIT
> + data16 = (u16 *)skb_put(skb, len);
> + ks8842_select_bank(adapter, 17);
> + while (len > 0) {
> + *data16++ = ioread16(adapter->hw_addr +
> + REG_QMU_DATA_LO);
> + *data16++ = ioread16(adapter->hw_addr +
> + REG_QMU_DATA_HI);
> + len -= sizeof(u32);
> + }
> +#else
> data = (u32 *)skb_put(skb, len);
>
> ks8842_select_bank(adapter, 17);
> @@ -397,6 +465,7 @@ static void ks8842_rx_frame(struct net_d
>
> skb->protocol = eth_type_trans(skb, netdev);
> netif_rx(skb);
> +#endif
> } else
> netdev->stats.rx_dropped++;
> } else {
> --- linux-2.6.35-rc3/drivers/net/Kconfig.orig 2010-07-02
15:52:41.000000000 -0700
> +++ linux-2.6.35-rc3/drivers/net/Kconfig 2010-07-07
07:45:47.000000000 -0700
> @@ -1748,11 +1748,29 @@ config TLAN
> Please email feedback to <torben.mathiasen@compaq.com>.
>
> config KS8842
> - tristate "Micrel KSZ8842"
> + tristate "Micrel KSZ8841/42 with generic bus interface"
> depends on HAS_IOMEM
> help
> - This platform driver is for Micrel KSZ8842 / KS8842
> - 2-port ethernet switch chip (managed, VLAN, QoS).
> + This platform driver is for KSZ8841(1-port) / KS8842(2-port)
> + ethernet switch chip (managed, VLAN, QoS) from Micrel or
> + Timberdale(FPGA).
> +
> +if KS8842
> +config MICREL_KS884X
> + boolean "KSZ8841/42 device from Micrel"
> + default n
> + help
> + Say Y to use Micrel device. Otherwise Timberdale(FPGA) device
is
> + selected.
> +
> +config KS884X_16BIT
> + boolean "16bit bus width"
> + default y
> + help
> + This option specifies 16bit or 32bit bus interface. Say Y to
use
> + 16bit bus. Otherwise 32bit bus is selected.
> +
> +endif # KS8842
>
> config KS8851
> tristate "Micrel KS8851 SPI"
> ---
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH v3] vlan: allow TSO setting on vlan interfaces
From: Ben Hutchings @ 2010-07-08 18:25 UTC (permalink / raw)
To: Eric Dumazet; +Cc: David Miller, netdev, Patrick McHardy
In-Reply-To: <1278607037.2651.57.camel@edumazet-laptop>
On Thu, 2010-07-08 at 18:37 +0200, Eric Dumazet wrote:
> When we need to shape traffic using low speeds, we need to
> disable tso on network interface :
>
> ethtool -K eth0.2240 tso off
>
> It seems vlan interfaces miss the set_tso() ethtool method.
>
> Before enabling TSO, we must check real device supports
> TSO for VLAN-tagged packets and enables TSO.
>
> Note that a TSO change on real device propagates TSO setting
> on all vlans, even if admin selected a different TSO setting.
>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
> ---
> net/8021q/vlan_dev.c | 20 ++++++++++++++++++++
> 1 file changed, 20 insertions(+)
>
> diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c
> index 7865a4c..a1b8171 100644
> --- a/net/8021q/vlan_dev.c
> +++ b/net/8021q/vlan_dev.c
> @@ -836,12 +836,32 @@ static struct rtnl_link_stats64 *vlan_dev_get_stats64(struct net_device *dev, st
> return stats;
> }
>
> +static int vlan_ethtool_set_tso(struct net_device *dev, u32 data)
> +{
> + if (data) {
> + struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
> +
> + /* Underlying device must support TSO for VLAN-tagged packets
> + * and must have TSO enabled now.
> + */
> + if (!(real_dev->vlan_features & NETIF_F_TSO))
> + return -EOPNOTSUPP;
> + if (!(real_dev->features & NETIF_F_TSO))
> + return -EINVAL;
> + dev->features |= NETIF_F_TSO;
> + } else {
> + dev->features &= ~NETIF_F_TSO;
> + }
> + return 0;
> +}
> +
> static const struct ethtool_ops vlan_ethtool_ops = {
> .get_settings = vlan_ethtool_get_settings,
> .get_drvinfo = vlan_ethtool_get_drvinfo,
> .get_link = ethtool_op_get_link,
> .get_rx_csum = vlan_ethtool_get_rx_csum,
> .get_flags = vlan_ethtool_get_flags,
> + .set_tso = vlan_ethtool_set_tso,
> };
>
> static const struct net_device_ops vlan_netdev_ops = {
>
>
--
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* Re: [PATCH net-next] cxgb4vf: Implement "Unhandled Interrupts" statistic
From: Casey Leedom @ 2010-07-08 18:19 UTC (permalink / raw)
To: netdev
In-Reply-To: <201007081107.45287.leedom@chelsio.com>
By the way, if there are _any_ problems with the patch, please ignore it and
ball me out. I've tried my hardest to make it correct and follow the correct
format but if I'm still not doing it correctly that's my problem to address. I
appreciate your patience as I grind my way through this learning curve!
Casey
^ permalink raw reply
* [PATCH net-next] cxgb4vf: Implement "Unhandled Interrupts" statistic
From: Casey Leedom @ 2010-07-08 18:07 UTC (permalink / raw)
To: netdev
>From 6c9cc1b5441afb32ecfcd10ac3060452e3c4df6f Mon Sep 17 00:00:00 2001
From: Casey Leedom <leedom@chelsio.com>
Date: Thu, 8 Jul 2010 10:05:48 -0700
Subject: [PATCH net-next] cxgb4vf: Implement "Unhandled Interrupts" statistic
Implement "Unhandled Interrupts" statistic so we can detect when the
hardware tells us that it things we have work to do but we don't find
anything ...
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/cxgb4vf_main.c | 8 +++++---
drivers/net/cxgb4vf/sge.c | 3 +++
2 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index bd73ff5..e988031 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -1790,7 +1790,7 @@ static int sge_qstats_show(struct seq_file *seq, void *v)
(rxq[qs].rspq.netdev
? rxq[qs].rspq.netdev->name
: "N/A"));
- R3("u", "RspQNullInts", rspq.unhandled_irqs);
+ R3("u", "RspQNullInts:", rspq.unhandled_irqs);
R("RxPackets:", stats.pkts);
R("RxCSO:", stats.rx_cso);
R("VLANxtract:", stats.vlan_ex);
@@ -1814,14 +1814,16 @@ static int sge_qstats_show(struct seq_file *seq, void *v)
const struct sge_rspq *evtq = &adapter->sge.fw_evtq;
seq_printf(seq, "%-8s %16s\n", "QType:", "FW event queue");
- /* no real response queue statistics available to display */
+ seq_printf(seq, "%-16s %8u\n", "RspQNullInts:",
+ evtq->unhandled_irqs);
seq_printf(seq, "%-16s %8u\n", "RspQ CIdx:", evtq->cidx);
seq_printf(seq, "%-16s %8u\n", "RspQ Gen:", evtq->gen);
} else if (r == 1) {
const struct sge_rspq *intrq = &adapter->sge.intrq;
seq_printf(seq, "%-8s %16s\n", "QType:", "Interrupt Queue");
- /* no real response queue statistics available to display */
+ seq_printf(seq, "%-16s %8u\n", "RspQNullInts:",
+ intrq->unhandled_irqs);
seq_printf(seq, "%-16s %8u\n", "RspQ CIdx:", intrq->cidx);
seq_printf(seq, "%-16s %8u\n", "RspQ Gen:", intrq->gen);
}
diff --git a/drivers/net/cxgb4vf/sge.c b/drivers/net/cxgb4vf/sge.c
index 3a7c02f..5c6663a 100644
--- a/drivers/net/cxgb4vf/sge.c
+++ b/drivers/net/cxgb4vf/sge.c
@@ -1776,6 +1776,9 @@ static int napi_rx_handler(struct napi_struct *napi, int budget)
} else
intr_params = QINTR_TIMER_IDX(SGE_TIMER_UPD_CIDX);
+ if (unlikely(work_done == 0))
+ rspq->unhandled_irqs++;
+
t4_write_reg(rspq->adapter,
T4VF_SGE_BASE_ADDR + SGE_VF_GTS,
CIDXINC(work_done) |
--
1.7.0.4
^ permalink raw reply related
* Re: Bug handling devices with weird names
From: Stephen Hemminger @ 2010-07-08 17:27 UTC (permalink / raw)
To: Martín Ferrari; +Cc: netdev, Mathieu Lacage
In-Reply-To: <AANLkTinBvzfdiAldJkXfH_vOkCv_cK7hBhjmH1l2vHTd@mail.gmail.com>
On Thu, 8 Jul 2010 19:08:07 +0200
Martín Ferrari <martin.ferrari@gmail.com> wrote:
> According to dev_valid_name (net/core/dev.c), a valid device name is
> one that doesn't include spaces, slashes, and is not "." or "..". But
> if I create a device called "foo:", some operations fail:
>
> # ip link add name foo: type dummy
> # ip link list foo:
> 155: foo:: <BROADCAST,NOARP> mtu 1500 qdisc noop state DOWN ioctl(SIOCGIFXQLEN)
> failed: No such device
>
> link/ether 92:7f:4d:0d:8f:57 brd ff:ff:ff:ff:ff:ff
>
> Strace reveals that the problem occurs when executing an ioctl:
>
> ioctl(4, SIOCGIFTXQLEN, {ifr_name="foo:", ???}) = -1 ENODEV (No such device)
>
> ifconfig gets completely lost, of course:
>
> $ /sbin/ifconfig foo:
> foo:: error fetching interface information: Device not found
>
> ioctl(5, SIOCGIFFLAGS, {ifr_name="foo:", ???}) = -1 ENODEV (No such device)
>
>
> So, is this fixable or iproute should stop using the ioctl interface?
> (dunno if netlink provides everything already)
Colons are used for the old IP aliasing. IP aliasing was an older way of handling
multiple addresses per interface, and is not necessary anymore. The syntax is
retained for legacy compatibility.
^ permalink raw reply
* Bug handling devices with weird names
From: Martín Ferrari @ 2010-07-08 17:08 UTC (permalink / raw)
To: netdev; +Cc: Mathieu Lacage
According to dev_valid_name (net/core/dev.c), a valid device name is
one that doesn't include spaces, slashes, and is not "." or "..". But
if I create a device called "foo:", some operations fail:
# ip link add name foo: type dummy
# ip link list foo:
155: foo:: <BROADCAST,NOARP> mtu 1500 qdisc noop state DOWN ioctl(SIOCGIFXQLEN)
failed: No such device
link/ether 92:7f:4d:0d:8f:57 brd ff:ff:ff:ff:ff:ff
Strace reveals that the problem occurs when executing an ioctl:
ioctl(4, SIOCGIFTXQLEN, {ifr_name="foo:", ???}) = -1 ENODEV (No such device)
ifconfig gets completely lost, of course:
$ /sbin/ifconfig foo:
foo:: error fetching interface information: Device not found
ioctl(5, SIOCGIFFLAGS, {ifr_name="foo:", ???}) = -1 ENODEV (No such device)
So, is this fixable or iproute should stop using the ioctl interface?
(dunno if netlink provides everything already)
--
Martín Ferrari
^ permalink raw reply
* Re: [Bugme-new] [Bug 16257] New: sysfs changes break hwsim and bnep drivers
From: Eric W. Biederman @ 2010-07-08 16:55 UTC (permalink / raw)
To: Greg KH
Cc: Andrew Morton, Rafael J. Wysocki, Maciej W. Rozycki, Kay Sievers,
Johannes Berg, Greg KH, netdev
In-Reply-To: <20100622035631.GA3755@suse.de>
Greg KH <gregkh@suse.de> writes:
> Does this only show up if you enable network namespaces? Or is it a
> problem with the "normal" kernel configuration of no network namespaces?
It is a problem with the normal kernel configuration. In that case
we have a single network namespace.
> The only thing that changed here was your network namespace work, what
> caused this problem to show up? Was it bisectable down to a single
> patch?
The problem is that when we remove symlinks we look at the target of
the symlink to see which tag to remove it in it's parent directory.
Because the target of the symlink was not in a class directory because
of that crazy class on class stacking exception in get_device_parent
we fail to find the symlink when we attempt to remove it.
My one line patch to get_device_parent really does fix this. I have
just sent you a patch to prevent the creation of these crazy symlinks
in sysfs, which at least clearly isolates this to a handful of drivers.
> Classes on top of classes should never have originally worked, I guess
> we just let them slide by accident, and we can go and fix them up as
> they are found. But for now, for .35, it would be good to find the root
> cause of the problem here. It might be as simple as putting a
> CONFIG_BROKEN on network namespaces until we get this working, right?
Nope, it isn't as simple as disabling network namespaces.
For the mac80211_hwsim we can just remove the device parent, and all will
be well for the moment.
For the bnep code we could also not set the parent but I don't know if
that would have undesirable complications for power management or not.
I don't have a bluetooth attached network device so I don't even know how
to test that code.
My preferred fix is the one liner I set you to get_device_parent. It
is no worse than the magic device_is_not_partition checks then we
already have in sysfs. Especially when it is a bug that
get_device_parent does that early return in the first place. To fix
this we will have to introduce the missing subdirectory one way or
another.
With SYSFS_DEPRECATED enabled I don't believe this problem can actually
happen, as all network devices are placed immediately under
/sys/class/net/ in sysfs.
Eric
^ permalink raw reply
* [RFC][PATCH] mac80211_hwsim: No parent is better than an illegimate one.
From: Eric W. Biederman @ 2010-07-08 16:37 UTC (permalink / raw)
To: Johannes Berg
Cc: Andrew Morton, Rafael J. Wysocki, Maciej W. Rozycki, Kay Sievers,
Johannes Berg, Greg KH, Greg KH <gregkh@suse.de> netdev
In-Reply-To: <m1630q7x5v.fsf_-_@fess.ebiederm.org>
Don't call SET_IEEE80211_DEV. This weakens the connections between
the phy files in sysfs slightly but otherwise it makes the driver work
in the face of tagged sysfs support.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
---
Johannes does this change look usable?
drivers/net/wireless/mac80211_hwsim.c | 1 -
1 files changed, 0 insertions(+), 1 deletions(-)
diff --git a/drivers/net/wireless/mac80211_hwsim.c b/drivers/net/wireless/mac80211_hwsim.c
index 6f8cb3e..b387222 100644
--- a/drivers/net/wireless/mac80211_hwsim.c
+++ b/drivers/net/wireless/mac80211_hwsim.c
@@ -1282,7 +1282,6 @@ static int __init init_mac80211_hwsim(void)
}
data->dev->driver = &mac80211_hwsim_driver;
- SET_IEEE80211_DEV(hw, data->dev);
addr[3] = i >> 8;
addr[4] = i;
memcpy(data->addresses[0].addr, addr, ETH_ALEN);
--
1.6.5.2.143.g8cc62
^ permalink raw reply related
* [PATCH v3] vlan: allow TSO setting on vlan interfaces
From: Eric Dumazet @ 2010-07-08 16:37 UTC (permalink / raw)
To: Ben Hutchings; +Cc: David Miller, netdev, Patrick McHardy
In-Reply-To: <1278604985.2651.35.camel@edumazet-laptop>
When we need to shape traffic using low speeds, we need to
disable tso on network interface :
ethtool -K eth0.2240 tso off
It seems vlan interfaces miss the set_tso() ethtool method.
Before enabling TSO, we must check real device supports
TSO for VLAN-tagged packets and enables TSO.
Note that a TSO change on real device propagates TSO setting
on all vlans, even if admin selected a different TSO setting.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
---
net/8021q/vlan_dev.c | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)
diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c
index 7865a4c..a1b8171 100644
--- a/net/8021q/vlan_dev.c
+++ b/net/8021q/vlan_dev.c
@@ -836,12 +836,32 @@ static struct rtnl_link_stats64 *vlan_dev_get_stats64(struct net_device *dev, st
return stats;
}
+static int vlan_ethtool_set_tso(struct net_device *dev, u32 data)
+{
+ if (data) {
+ struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+
+ /* Underlying device must support TSO for VLAN-tagged packets
+ * and must have TSO enabled now.
+ */
+ if (!(real_dev->vlan_features & NETIF_F_TSO))
+ return -EOPNOTSUPP;
+ if (!(real_dev->features & NETIF_F_TSO))
+ return -EINVAL;
+ dev->features |= NETIF_F_TSO;
+ } else {
+ dev->features &= ~NETIF_F_TSO;
+ }
+ return 0;
+}
+
static const struct ethtool_ops vlan_ethtool_ops = {
.get_settings = vlan_ethtool_get_settings,
.get_drvinfo = vlan_ethtool_get_drvinfo,
.get_link = ethtool_op_get_link,
.get_rx_csum = vlan_ethtool_get_rx_csum,
.get_flags = vlan_ethtool_get_flags,
+ .set_tso = vlan_ethtool_set_tso,
};
static const struct net_device_ops vlan_netdev_ops = {
^ permalink raw reply related
* [PATCH] sysfs: Don't allow the creation of symlinks we can't remove
From: Eric W. Biederman @ 2010-07-08 16:31 UTC (permalink / raw)
To: Greg KH
Cc: Andrew Morton, Rafael J. Wysocki, Maciej W. Rozycki, Kay Sievers,
Johannes Berg, Greg KH, netdev
In-Reply-To: <20100622035631.GA3755@suse.de>
Recently my tagged sysfs support revealed a flaw in the device core
that a few rare drivers are running into such that we don't always put
network devices in a class subdirectory named net/.
Since we are not creating the class directory the network devices wind
up in a non-tagged directory, but the symlinks to the network devices
from /sys/class/net are in a tagged directory. All of which works
until we go to remove or rename the symlink. When we remove or rename
a symlink we look in the namespace of the target of the symlink.
Since the target of the symlink is in a non-tagged sysfs directory we
don't have a namespace to look in, and we fail to remove the symlink.
Detect this problem up front and simply don't create symlinks we won't
be able to remove later. This prevents symlink leakage and fails in
a much clearer and more understandable way.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
---
fs/sysfs/symlink.c | 23 ++++++++++++++++++-----
1 files changed, 18 insertions(+), 5 deletions(-)
diff --git a/fs/sysfs/symlink.c b/fs/sysfs/symlink.c
index f71246b..44bca5f 100644
--- a/fs/sysfs/symlink.c
+++ b/fs/sysfs/symlink.c
@@ -28,6 +28,7 @@ static int sysfs_do_create_link(struct kobject *kobj, struct kobject *target,
struct sysfs_dirent *target_sd = NULL;
struct sysfs_dirent *sd = NULL;
struct sysfs_addrm_cxt acxt;
+ enum kobj_ns_type ns_type;
int error;
BUG_ON(!name);
@@ -58,16 +59,28 @@ static int sysfs_do_create_link(struct kobject *kobj, struct kobject *target,
if (!sd)
goto out_put;
- if (sysfs_ns_type(parent_sd))
+ ns_type = sysfs_ns_type(parent_sd);
+ if (ns_type)
sd->s_ns = target->ktype->namespace(target);
sd->s_symlink.target_sd = target_sd;
target_sd = NULL; /* reference is now owned by the symlink */
sysfs_addrm_start(&acxt, parent_sd);
- if (warn)
- error = sysfs_add_one(&acxt, sd);
- else
- error = __sysfs_add_one(&acxt, sd);
+ /* Symlinks must be between directories with the same ns_type */
+ if (ns_type == sysfs_ns_type(sd->s_symlink.target_sd->s_parent)) {
+ if (warn)
+ error = sysfs_add_one(&acxt, sd);
+ else
+ error = __sysfs_add_one(&acxt, sd);
+ } else {
+ error = -EINVAL;
+ WARN(1, KERN_WARNING
+ "sysfs: symlink across ns_types %s/%s -> %s/%s\n",
+ parent_sd->s_name,
+ sd->s_name,
+ sd->s_symlink.target_sd->s_parent->s_name,
+ sd->s_symlink.target_sd->s_name);
+ }
sysfs_addrm_finish(&acxt);
if (error)
--
1.6.5.2.143.g8cc62
^ permalink raw reply related
* [PATCH net-next-2.6] tg3: allow TSO on vlan devices
From: Eric Dumazet @ 2010-07-08 16:14 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Matt Carlson, Michael Chan
Similar to commit 72dccb01e8632aa (bnx2: Update vlan_features)
In order to enable TSO on vlan devices, tg3 needs to update
dev->vlan_features.
Tested on HP NC326m (aka BCM5715S (rev a3))
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
---
drivers/net/tg3.c | 28 +++++++++++++++++++++-------
1 file changed, 21 insertions(+), 7 deletions(-)
diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
index 289cdc5..68329a9 100644
--- a/drivers/net/tg3.c
+++ b/drivers/net/tg3.c
@@ -12800,6 +12800,13 @@ done:
static struct pci_dev * __devinit tg3_find_peer(struct tg3 *);
+static void inline vlan_features_add(struct net_device *dev, unsigned long flags)
+{
+#if TG3_VLAN_TAG_USED
+ dev->vlan_features |= flags;
+#endif
+}
+
static int __devinit tg3_get_invariants(struct tg3 *tp)
{
static struct pci_device_id write_reorder_chipsets[] = {
@@ -13032,11 +13039,13 @@ static int __devinit tg3_get_invariants(struct tg3 *tp)
if (tp->pci_chip_rev_id == CHIPREV_ID_5700_B0)
tp->tg3_flags |= TG3_FLAG_BROKEN_CHECKSUMS;
else {
+ unsigned long features = NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_GRO;
+
tp->tg3_flags |= TG3_FLAG_RX_CHECKSUMS;
- tp->dev->features |= NETIF_F_IP_CSUM | NETIF_F_SG;
if (tp->tg3_flags3 & TG3_FLG3_5755_PLUS)
- tp->dev->features |= NETIF_F_IPV6_CSUM;
- tp->dev->features |= NETIF_F_GRO;
+ features |= NETIF_F_IPV6_CSUM;
+ tp->dev->features |= features;
+ vlan_features_add(tp->dev, features);
}
/* Determine TSO capabilities */
@@ -14525,20 +14534,25 @@ static int __devinit tg3_init_one(struct pci_dev *pdev,
* is off by default, but can be enabled using ethtool.
*/
if ((tp->tg3_flags2 & TG3_FLG2_HW_TSO) &&
- (dev->features & NETIF_F_IP_CSUM))
+ (dev->features & NETIF_F_IP_CSUM)) {
dev->features |= NETIF_F_TSO;
-
+ vlan_features_add(dev, NETIF_F_TSO);
+ }
if ((tp->tg3_flags2 & TG3_FLG2_HW_TSO_2) ||
(tp->tg3_flags2 & TG3_FLG2_HW_TSO_3)) {
- if (dev->features & NETIF_F_IPV6_CSUM)
+ if (dev->features & NETIF_F_IPV6_CSUM) {
dev->features |= NETIF_F_TSO6;
+ vlan_features_add(dev, NETIF_F_TSO6);
+ }
if ((tp->tg3_flags2 & TG3_FLG2_HW_TSO_3) ||
GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5761 ||
(GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5784 &&
GET_CHIP_REV(tp->pci_chip_rev_id) != CHIPREV_5784_AX) ||
GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5785 ||
- GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_57780)
+ GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_57780) {
dev->features |= NETIF_F_TSO_ECN;
+ vlan_features_add(dev, NETIF_F_TSO_ECN);
+ }
}
if (tp->pci_chip_rev_id == CHIPREV_ID_5705_A1 &&
^ permalink raw reply related
* Re: [PATCH v2] vlan: allow TSO setting on vlan interfaces
From: Eric Dumazet @ 2010-07-08 16:12 UTC (permalink / raw)
To: Ben Hutchings; +Cc: David Miller, netdev, Patrick McHardy
In-Reply-To: <1278605109.16013.52.camel@achroite.uk.solarflarecom.com>
Le jeudi 08 juillet 2010 à 17:05 +0100, Ben Hutchings a écrit :
>
> Yes, sorry for the noise.
No problem, it was a good point to recheck anyway.
^ permalink raw reply
* Re: [PATCH v2] vlan: allow TSO setting on vlan interfaces
From: Ben Hutchings @ 2010-07-08 16:05 UTC (permalink / raw)
To: Eric Dumazet; +Cc: David Miller, netdev, Patrick McHardy
In-Reply-To: <1278604830.2651.32.camel@edumazet-laptop>
On Thu, 2010-07-08 at 18:00 +0200, Eric Dumazet wrote:
> Le jeudi 08 juillet 2010 à 16:49 +0100, Ben Hutchings a écrit :
> > On Thu, 2010-07-08 at 16:43 +0100, Ben Hutchings wrote:
> > > On Thu, 2010-07-08 at 11:39 +0200, Eric Dumazet wrote:
> > > > When we need to shape traffic with low speeds, we need to disable tso on
> > > > network interface :
> > > >
> > > > ethtool -K eth0.2240 tso off
> > > >
> > > > It seems vlan interfaces miss the set_tso() ethtool method.
> > > > Propagating tso changes from lower device is not always wanted, some
> > > > vlans want TSO on, others want TSO off.
> > [...]
> > > I think the vlan driver should also have a netdev notifier to handle
> > > feature changes on the underlying device.
> >
> > To clarify, I think offload features should be disabled on a vlan device
> > if they are later disabled on the underlying device. Propagating
> > changes to enable features, as you say, might not be wanted.
>
> OK, but isnt it already done ?
[...]
Yes, sorry for the noise.
Ben.
--
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* Re: [PATCH v2] vlan: allow TSO setting on vlan interfaces
From: Eric Dumazet @ 2010-07-08 16:04 UTC (permalink / raw)
To: Patrick McHardy; +Cc: Ben Hutchings, David Miller, netdev
In-Reply-To: <4C35F4E4.4070606@trash.net>
Le jeudi 08 juillet 2010 à 17:55 +0200, Patrick McHardy a écrit :
> I agree. The VLAN driver should also propagate feature changes
> upwards itself.
I noticed bonding is also at fault in this area.
bond0.xxx are not noticed of TSO changes on bond0
^ permalink raw reply
* Re: [PATCH v2] vlan: allow TSO setting on vlan interfaces
From: Eric Dumazet @ 2010-07-08 16:03 UTC (permalink / raw)
To: Ben Hutchings; +Cc: David Miller, netdev, Patrick McHardy
In-Reply-To: <1278603804.16013.24.camel@achroite.uk.solarflarecom.com>
Le jeudi 08 juillet 2010 à 16:43 +0100, Ben Hutchings a écrit :
> On Thu, 2010-07-08 at 11:39 +0200, Eric Dumazet wrote:
> > When we need to shape traffic with low speeds, we need to disable tso on
> > network interface :
> >
> > ethtool -K eth0.2240 tso off
> >
> > It seems vlan interfaces miss the set_tso() ethtool method.
> > Propagating tso changes from lower device is not always wanted, some
> > vlans want TSO on, others want TSO off.
> >
> > Before enabling TSO, we must check real device supports it.
> >
> > Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> > ---
> > net/8021q/vlan_dev.c | 13 +++++++++++++
> > 1 file changed, 13 insertions(+)
> >
> > diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c
> > index c6456cb..870bc53 100644
> > --- a/net/8021q/vlan_dev.c
> > +++ b/net/8021q/vlan_dev.c
> > @@ -838,12 +838,25 @@ static struct rtnl_link_stats64 *vlan_dev_get_stats64(struct net_device *dev)
> > return stats;
> > }
> >
> > +static int vlan_ethtool_set_tso(struct net_device *dev, u32 data)
> > +{
> > + if (data) {
> > + struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
> > + unsigned long rdev_vfeatures = real_dev->features & real_dev->vlan_features;
> > +
> > + dev->features |= (NETIF_F_TSO & rdev_vfeatures);
> > + } else
> > + dev->features &= ~NETIF_F_TSO;
> > + return 0;
> > +}
> [...]
>
> This should not silently ignore attempts to enable TSO. I think it
> should be something like:
>
> static int vlan_ethtool_set_tso(struct net_device *dev, u32 data)
> {
> if (data) {
> struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
>
> /* Underlying device must support TSO for VLAN-tagged packets
> * and must have TSO enabled now. */
> if (!(real_dev->vlan_features & NETIF_F_TSO))
> return -EOPNOTSUPP;
> if (!(real_dev->features & NETIF_F_TSO))
> return -EINVAL;
>
> dev->features |= NETIF_F_TSO;
> } else {
> dev->features &= ~NETIF_F_TSO;
> }
>
> return 0;
> }
I agree.
What about other TSO flags like NETIF_F_TSO6 & NETIF_F_TSO_ECN ?
Not sure if we should manipulate them in set_tso()
^ permalink raw reply
* Re: [PATCH v2] vlan: allow TSO setting on vlan interfaces
From: Eric Dumazet @ 2010-07-08 16:00 UTC (permalink / raw)
To: Ben Hutchings; +Cc: David Miller, netdev, Patrick McHardy
In-Reply-To: <1278604198.16013.35.camel@achroite.uk.solarflarecom.com>
Le jeudi 08 juillet 2010 à 16:49 +0100, Ben Hutchings a écrit :
> On Thu, 2010-07-08 at 16:43 +0100, Ben Hutchings wrote:
> > On Thu, 2010-07-08 at 11:39 +0200, Eric Dumazet wrote:
> > > When we need to shape traffic with low speeds, we need to disable tso on
> > > network interface :
> > >
> > > ethtool -K eth0.2240 tso off
> > >
> > > It seems vlan interfaces miss the set_tso() ethtool method.
> > > Propagating tso changes from lower device is not always wanted, some
> > > vlans want TSO on, others want TSO off.
> [...]
> > I think the vlan driver should also have a netdev notifier to handle
> > feature changes on the underlying device.
>
> To clarify, I think offload features should be disabled on a vlan device
> if they are later disabled on the underlying device. Propagating
> changes to enable features, as you say, might not be wanted.
OK, but isnt it already done ?
Check vlan_transfer_features() in net/8021q/vlan.c
# ethtool -k eth3|grep tcp-segmentation
tcp-segmentation-offload: on
# ethtool -k eth3.103|grep tcp-segmentation
tcp-segmentation-offload: on
# ethtool -K eth3 tso off
# ethtool -k eth3|grep tcp-segmentation
tcp-segmentation-offload: off
# ethtool -k eth3.103|grep tcp-segmentation
tcp-segmentation-offload: off
# ethtool -K eth3 tso on
# ethtool -k eth3.103|grep tcp-segmentation
tcp-segmentation-offload: on
We should not change it to avoid ORing TSO, it might break some setups.
^ permalink raw reply
* Re: [PATCH v2] vlan: allow TSO setting on vlan interfaces
From: Patrick McHardy @ 2010-07-08 15:55 UTC (permalink / raw)
To: Ben Hutchings; +Cc: Eric Dumazet, David Miller, netdev
In-Reply-To: <1278604198.16013.35.camel@achroite.uk.solarflarecom.com>
Ben Hutchings wrote:
> On Thu, 2010-07-08 at 16:43 +0100, Ben Hutchings wrote:
>
>> On Thu, 2010-07-08 at 11:39 +0200, Eric Dumazet wrote:
>>
>>> When we need to shape traffic with low speeds, we need to disable tso on
>>> network interface :
>>>
>>> ethtool -K eth0.2240 tso off
>>>
>>> It seems vlan interfaces miss the set_tso() ethtool method.
>>> Propagating tso changes from lower device is not always wanted, some
>>> vlans want TSO on, others want TSO off.
>>>
> [...]
>
>> I think the vlan driver should also have a netdev notifier to handle
>> feature changes on the underlying device.
>>
>
> To clarify, I think offload features should be disabled on a vlan device
> if they are later disabled on the underlying device. Propagating
> changes to enable features, as you say, might not be wanted.
I agree. The VLAN driver should also propagate feature changes
upwards itself.
^ 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