* [net-next 12/14] ice: update ethtool stats on-demand
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem; +Cc: Bruce Allan, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Bruce Allan <bruce.w.allan@intel.com>
Users expect ethtool statistics to be updated on-demand when invoking
'ethtool -S <iface>' instead of providing a snapshot of statistics taken
once a second (the frequency of the watchdog task where stats are currently
updated). Update stats every time 'ethtool -S <iface>' is run.
Also, fix an indentation style issue and an unnecessary local variable
initialization in ice_get_ethtool_stats() discovered while investigating
the subject issue.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice.h | 2 ++
drivers/net/ethernet/intel/ice/ice_ethtool.c | 7 +++++--
drivers/net/ethernet/intel/ice/ice_main.c | 6 ++----
3 files changed, 9 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 99e0febd8e50..97d0f61cf52b 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -447,6 +447,8 @@ ice_find_vsi_by_type(struct ice_pf *pf, enum ice_vsi_type type)
int ice_vsi_setup_tx_rings(struct ice_vsi *vsi);
int ice_vsi_setup_rx_rings(struct ice_vsi *vsi);
void ice_set_ethtool_ops(struct net_device *netdev);
+void ice_update_vsi_stats(struct ice_vsi *vsi);
+void ice_update_pf_stats(struct ice_pf *pf);
int ice_up(struct ice_vsi *vsi);
int ice_down(struct ice_vsi *vsi);
int ice_vsi_cfg(struct ice_vsi *vsi);
diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index 948a33716290..f7dd0bd03d39 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -1319,14 +1319,17 @@ ice_get_ethtool_stats(struct net_device *netdev,
struct ice_vsi *vsi = np->vsi;
struct ice_pf *pf = vsi->back;
struct ice_ring *ring;
- unsigned int j = 0;
+ unsigned int j;
int i = 0;
char *p;
+ ice_update_pf_stats(pf);
+ ice_update_vsi_stats(vsi);
+
for (j = 0; j < ICE_VSI_STATS_LEN; j++) {
p = (char *)vsi + ice_gstrings_vsi_stats[j].stat_offset;
data[i++] = (ice_gstrings_vsi_stats[j].sizeof_stat ==
- sizeof(u64)) ? *(u64 *)p : *(u32 *)p;
+ sizeof(u64)) ? *(u64 *)p : *(u32 *)p;
}
/* populate per queue stats */
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index a0d148f590c2..6dd806b763ea 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -34,8 +34,6 @@ static const struct net_device_ops ice_netdev_ops;
static void ice_rebuild(struct ice_pf *pf);
static void ice_vsi_release_all(struct ice_pf *pf);
-static void ice_update_vsi_stats(struct ice_vsi *vsi);
-static void ice_update_pf_stats(struct ice_pf *pf);
/**
* ice_get_tx_pending - returns number of Tx descriptors not processed
@@ -3254,7 +3252,7 @@ static void ice_update_vsi_ring_stats(struct ice_vsi *vsi)
* ice_update_vsi_stats - Update VSI stats counters
* @vsi: the VSI to be updated
*/
-static void ice_update_vsi_stats(struct ice_vsi *vsi)
+void ice_update_vsi_stats(struct ice_vsi *vsi)
{
struct rtnl_link_stats64 *cur_ns = &vsi->net_stats;
struct ice_eth_stats *cur_es = &vsi->eth_stats;
@@ -3290,7 +3288,7 @@ static void ice_update_vsi_stats(struct ice_vsi *vsi)
* ice_update_pf_stats - Update PF port stats counters
* @pf: PF whose stats needs to be updated
*/
-static void ice_update_pf_stats(struct ice_pf *pf)
+void ice_update_pf_stats(struct ice_pf *pf)
{
struct ice_hw_port_stats *prev_ps, *cur_ps;
struct ice_hw *hw = &pf->hw;
--
2.21.0
^ permalink raw reply related
* [net-next 09/14] ice: Fix flag used for module query
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem; +Cc: Chinh T Cao, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Chinh T Cao <chinh.t.cao@intel.com>
When checking the PHY for status, by specification, the driver
should be using "topology" mode when querying the module type.
Signed-off-by: Chinh T Cao <chinh.t.cao@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 5f9dc76699d2..15648d4a8bab 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -2031,7 +2031,7 @@ enum ice_status ice_update_link_info(struct ice_port_info *pi)
if (!pcaps)
return ICE_ERR_NO_MEMORY;
- status = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_SW_CFG,
+ status = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_TOPO_CAP,
pcaps, NULL);
if (!status)
memcpy(li->module_type, &pcaps->module_type,
--
2.21.0
^ permalink raw reply related
* [net-next 04/14] ice: Treat DCBx state NOT_STARTED as valid
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem; +Cc: Dave Ertman, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Dave Ertman <david.m.ertman@intel.com>
When a port is not cabled, but DCBx is enabled in the
firmware, the status of DCBx will be NOT_STARTED. This
is a valid state for FW enabled and should not be
treated as a is_fw_lldp true automatically.
Add the code to treat NOT_STARTED as another valid state.
Signed-off-by: Dave Ertman <david.m.ertman@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_dcb.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_dcb.c b/drivers/net/ethernet/intel/ice/ice_dcb.c
index c2002ded65f6..d60c942249e8 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb.c
+++ b/drivers/net/ethernet/intel/ice/ice_dcb.c
@@ -954,7 +954,8 @@ enum ice_status ice_init_dcb(struct ice_hw *hw)
pi->dcbx_status = ice_get_dcbx_status(hw);
if (pi->dcbx_status == ICE_DCBX_STATUS_DONE ||
- pi->dcbx_status == ICE_DCBX_STATUS_IN_PROGRESS) {
+ pi->dcbx_status == ICE_DCBX_STATUS_IN_PROGRESS ||
+ pi->dcbx_status == ICE_DCBX_STATUS_NOT_STARTED) {
/* Get current DCBX configuration */
ret = ice_get_dcb_cfg(pi);
pi->is_sw_lldp = (hw->adminq.sq_last_status == ICE_AQ_RC_EPERM);
--
2.21.0
^ permalink raw reply related
* [net-next 14/14] ice: Don't allow VSI to remove unassociated ucast filter
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem
Cc: Akeem G Abodunrin, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
If a VSI is not using a unicast filter or did not configure that
particular unicast filter, driver should not allow it to be removed
by the rogue VSI.
Signed-off-by: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_switch.c | 56 +++++++++++++++++++++
1 file changed, 56 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_switch.c b/drivers/net/ethernet/intel/ice/ice_switch.c
index 8271fd651725..99cf527d2b1a 100644
--- a/drivers/net/ethernet/intel/ice/ice_switch.c
+++ b/drivers/net/ethernet/intel/ice/ice_switch.c
@@ -2136,6 +2136,38 @@ ice_cfg_dflt_vsi(struct ice_hw *hw, u16 vsi_handle, bool set, u8 direction)
return status;
}
+/**
+ * ice_find_ucast_rule_entry - Search for a unicast MAC filter rule entry
+ * @hw: pointer to the hardware structure
+ * @recp_id: lookup type for which the specified rule needs to be searched
+ * @f_info: rule information
+ *
+ * Helper function to search for a unicast rule entry - this is to be used
+ * to remove unicast MAC filter that is not shared with other VSIs on the
+ * PF switch.
+ *
+ * Returns pointer to entry storing the rule if found
+ */
+static struct ice_fltr_mgmt_list_entry *
+ice_find_ucast_rule_entry(struct ice_hw *hw, u8 recp_id,
+ struct ice_fltr_info *f_info)
+{
+ struct ice_switch_info *sw = hw->switch_info;
+ struct ice_fltr_mgmt_list_entry *list_itr;
+ struct list_head *list_head;
+
+ list_head = &sw->recp_list[recp_id].filt_rules;
+ list_for_each_entry(list_itr, list_head, list_entry) {
+ if (!memcmp(&f_info->l_data, &list_itr->fltr_info.l_data,
+ sizeof(f_info->l_data)) &&
+ f_info->fwd_id.hw_vsi_id ==
+ list_itr->fltr_info.fwd_id.hw_vsi_id &&
+ f_info->flag == list_itr->fltr_info.flag)
+ return list_itr;
+ }
+ return NULL;
+}
+
/**
* ice_remove_mac - remove a MAC address based filter rule
* @hw: pointer to the hardware structure
@@ -2153,15 +2185,39 @@ enum ice_status
ice_remove_mac(struct ice_hw *hw, struct list_head *m_list)
{
struct ice_fltr_list_entry *list_itr, *tmp;
+ struct mutex *rule_lock; /* Lock to protect filter rule list */
if (!m_list)
return ICE_ERR_PARAM;
+ rule_lock = &hw->switch_info->recp_list[ICE_SW_LKUP_MAC].filt_rule_lock;
list_for_each_entry_safe(list_itr, tmp, m_list, list_entry) {
enum ice_sw_lkup_type l_type = list_itr->fltr_info.lkup_type;
+ u8 *add = &list_itr->fltr_info.l_data.mac.mac_addr[0];
+ u16 vsi_handle;
if (l_type != ICE_SW_LKUP_MAC)
return ICE_ERR_PARAM;
+
+ vsi_handle = list_itr->fltr_info.vsi_handle;
+ if (!ice_is_vsi_valid(hw, vsi_handle))
+ return ICE_ERR_PARAM;
+
+ list_itr->fltr_info.fwd_id.hw_vsi_id =
+ ice_get_hw_vsi_num(hw, vsi_handle);
+ if (is_unicast_ether_addr(add) && !hw->ucast_shared) {
+ /* Don't remove the unicast address that belongs to
+ * another VSI on the switch, since it is not being
+ * shared...
+ */
+ mutex_lock(rule_lock);
+ if (!ice_find_ucast_rule_entry(hw, ICE_SW_LKUP_MAC,
+ &list_itr->fltr_info)) {
+ mutex_unlock(rule_lock);
+ return ICE_ERR_DOES_NOT_EXIST;
+ }
+ mutex_unlock(rule_lock);
+ }
list_itr->status = ice_remove_rule_internal(hw,
ICE_SW_LKUP_MAC,
list_itr);
--
2.21.0
^ permalink raw reply related
* [net-next 05/14] ice: Copy dcbx configuration only if mode is correct
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem
Cc: Michal Swiatkowski, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Michal Swiatkowski <michal.swiatkowski@intel.com>
In rebuild DCB desired_dcbx_cfg was copy to local_dcbx_cfg, but
if DCBX mode is IEEE desired_dcbx_cfg is not initialized by DCBX
config from FW. Change logic to copy config value only if mode is
set to CEE.
If driver copy desired_dcbx_cfg to local_dcbx_cfg in IEEE mode there
is problem with globr. System is frozen after two or more globr.
Signed-off-by: Michal Swiatkowski <michal.swiatkowski@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_dcb_lib.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
index 22bdc244c7e0..4fc9faf5bc71 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
@@ -334,8 +334,10 @@ void ice_dcb_rebuild(struct ice_pf *pf)
devm_kfree(&pf->pdev->dev, prev_cfg);
/* Set the local desired config */
- memset(&pf->hw.port_info->local_dcbx_cfg, 0, sizeof(*local_dcbx_cfg));
- memcpy(local_dcbx_cfg, desired_dcbx_cfg, sizeof(*local_dcbx_cfg));
+ if (local_dcbx_cfg->dcbx_mode == ICE_DCBX_MODE_CEE)
+ memcpy(local_dcbx_cfg, desired_dcbx_cfg,
+ sizeof(*local_dcbx_cfg));
+
ice_cfg_etsrec_defaults(pf->hw.port_info);
ret = ice_set_dcb_cfg(pf->hw.port_info);
if (ret) {
--
2.21.0
^ permalink raw reply related
* [net-next 11/14] ice: Add input handlers for virtual channel handlers
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem
Cc: Amruth G.P, netdev, nhorman, sassmann, Nitesh B Venkatesh,
Tony Nguyen, Andrew Bowers, Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: "Amruth G.P" <amruth.gouda.parameshwarappa@intel.com>
Move the assignment to local variables after validation.
Remove unnecessary checks in ice_vc_process_vf_msg() as the respective
functions are now performing the checks.
Signed-off-by: "Amruth G.P" <amruth.gouda.parameshwarappa@intel.com>
Signed-off-by: Nitesh B Venkatesh <nitesh.b.venkatesh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
.../net/ethernet/intel/ice/ice_virtchnl_pf.c | 64 +++++++++----------
1 file changed, 32 insertions(+), 32 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
index 83e58e71081e..de0a1ef54e83 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
@@ -1734,18 +1734,18 @@ static int ice_vc_config_rss_key(struct ice_vf *vf, u8 *msg)
goto error_param;
}
- vsi = pf->vsi[vf->lan_vsi_idx];
- if (!vsi) {
+ if (vrk->key_len != ICE_VSIQF_HKEY_ARRAY_SIZE) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
goto error_param;
}
- if (vrk->key_len != ICE_VSIQF_HKEY_ARRAY_SIZE) {
+ if (!test_bit(ICE_FLAG_RSS_ENA, vf->pf->flags)) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
goto error_param;
}
- if (!test_bit(ICE_FLAG_RSS_ENA, vf->pf->flags)) {
+ vsi = pf->vsi[vf->lan_vsi_idx];
+ if (!vsi) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
goto error_param;
}
@@ -1781,18 +1781,18 @@ static int ice_vc_config_rss_lut(struct ice_vf *vf, u8 *msg)
goto error_param;
}
- vsi = pf->vsi[vf->lan_vsi_idx];
- if (!vsi) {
+ if (vrl->lut_entries != ICE_VSIQF_HLUT_ARRAY_SIZE) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
goto error_param;
}
- if (vrl->lut_entries != ICE_VSIQF_HLUT_ARRAY_SIZE) {
+ if (!test_bit(ICE_FLAG_RSS_ENA, vf->pf->flags)) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
goto error_param;
}
- if (!test_bit(ICE_FLAG_RSS_ENA, vf->pf->flags)) {
+ vsi = pf->vsi[vf->lan_vsi_idx];
+ if (!vsi) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
goto error_param;
}
@@ -1877,6 +1877,12 @@ static int ice_vc_ena_qs_msg(struct ice_vf *vf, u8 *msg)
goto error_param;
}
+ if (vqs->rx_queues > ICE_MAX_BASE_QS_PER_VF ||
+ vqs->tx_queues > ICE_MAX_BASE_QS_PER_VF) {
+ v_ret = VIRTCHNL_STATUS_ERR_PARAM;
+ goto error_param;
+ }
+
vsi = pf->vsi[vf->lan_vsi_idx];
if (!vsi) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
@@ -1932,6 +1938,12 @@ static int ice_vc_dis_qs_msg(struct ice_vf *vf, u8 *msg)
goto error_param;
}
+ if (vqs->rx_queues > ICE_MAX_BASE_QS_PER_VF ||
+ vqs->tx_queues > ICE_MAX_BASE_QS_PER_VF) {
+ v_ret = VIRTCHNL_STATUS_ERR_PARAM;
+ goto error_param;
+ }
+
vsi = pf->vsi[vf->lan_vsi_idx];
if (!vsi) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
@@ -1984,12 +1996,6 @@ static int ice_vc_cfg_irq_map_msg(struct ice_vf *vf, u8 *msg)
irqmap_info = (struct virtchnl_irq_map_info *)msg;
num_q_vectors_mapped = irqmap_info->num_vectors;
- vsi = pf->vsi[vf->lan_vsi_idx];
- if (!vsi) {
- v_ret = VIRTCHNL_STATUS_ERR_PARAM;
- goto error_param;
- }
-
/* Check to make sure number of VF vectors mapped is not greater than
* number of VF vectors originally allocated, and check that
* there is actually at least a single VF queue vector mapped
@@ -2001,6 +2007,12 @@ static int ice_vc_cfg_irq_map_msg(struct ice_vf *vf, u8 *msg)
goto error_param;
}
+ vsi = pf->vsi[vf->lan_vsi_idx];
+ if (!vsi) {
+ v_ret = VIRTCHNL_STATUS_ERR_PARAM;
+ goto error_param;
+ }
+
for (i = 0; i < num_q_vectors_mapped; i++) {
struct ice_q_vector *q_vector;
@@ -2092,10 +2104,6 @@ static int ice_vc_cfg_qs_msg(struct ice_vf *vf, u8 *msg)
goto error_param;
}
- vsi = pf->vsi[vf->lan_vsi_idx];
- if (!vsi)
- goto error_param;
-
if (qci->num_queue_pairs > ICE_MAX_BASE_QS_PER_VF) {
dev_err(&pf->pdev->dev,
"VF-%d requesting more than supported number of queues: %d\n",
@@ -2104,6 +2112,12 @@ static int ice_vc_cfg_qs_msg(struct ice_vf *vf, u8 *msg)
goto error_param;
}
+ vsi = pf->vsi[vf->lan_vsi_idx];
+ if (!vsi) {
+ v_ret = VIRTCHNL_STATUS_ERR_PARAM;
+ goto error_param;
+ }
+
for (i = 0; i < qci->num_queue_pairs; i++) {
qpi = &qci->qpair[i];
if (qpi->txq.vsi_id != qci->vsi_id ||
@@ -2755,20 +2769,6 @@ void ice_vc_process_vf_msg(struct ice_pf *pf, struct ice_rq_event_info *event)
err = -EPERM;
else
err = -EINVAL;
- goto error_handler;
- }
-
- /* Perform additional checks specific to RSS and Virtchnl */
- if (v_opcode == VIRTCHNL_OP_CONFIG_RSS_KEY) {
- struct virtchnl_rss_key *vrk = (struct virtchnl_rss_key *)msg;
-
- if (vrk->key_len != ICE_VSIQF_HKEY_ARRAY_SIZE)
- err = -EINVAL;
- } else if (v_opcode == VIRTCHNL_OP_CONFIG_RSS_LUT) {
- struct virtchnl_rss_lut *vrl = (struct virtchnl_rss_lut *)msg;
-
- if (vrl->lut_entries != ICE_VSIQF_HLUT_ARRAY_SIZE)
- err = -EINVAL;
}
error_handler:
--
2.21.0
^ permalink raw reply related
* [net-next 00/14][pull request] 100GbE Intel Wired LAN Driver Updates 2019-08-23
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem; +Cc: Jeff Kirsher, netdev, nhorman, sassmann
This series contains updates to ice driver only.
Dave adds logic for the necessary bits to be set in the VSI context for
the PF_VSI and the TX_descriptors for control packets egressing the
PF_VSI. Updated the logic to detect both DCBx and LLDP states in the
firmware engine to account for situations where DCBx is enabled and LLDP
is disabled. Fixed the driver to treat the DCBx state of "NOT_STARTED"
as a valid state and should not assume "is_fw_lldp" true automatically.
Since "enable-fw-lldp" flag was confusing and cumbersome, change the
flag to "fw-lldp-agent" with a value of on or off to help clarify
whether the LLDP agent is running or not.
Brett fixes an issue where synchronize_irq() was being called from the
host of VF's, which should not be done.
Michal fixed an issue when rebuilding the DCBx configuration while in
IEEE mode versus CEE mode, so add a check before copying the
configuration value to ensure we are only in CEE mode.
Jake fixes the PF to reject any VF request to setup head writeback since
the support has been deprecated.
Mitch adds an additional check to ensure the VF is active before sending
out an error message that a message was unable to be sent to a
particular VF.
Chinh updates the driver to use "topology" mode when checking the PHY
for status, since this mode provides us the current module type that is
available. Fixes the driver from clearing the auto_fec_enable bit which
was blocking a user from forcing non-spec compliant FEC configurations.
Amruth does a refactor on the code to first check, then assign in the
virtual channel space.
Bruce updates the driver to actually update the stats when a user runs
the ethtool command 'ethtool -S <iface>' instead of providing a snapshot
of the stats that maybe from a second ago.
Akeem fixes up the adding/removing of VSI MAC filters for VFs, so that
VFs cannot add/remove a filter from another VSI. We now track the
number of filters added right from when the VF resources get allocated
and won't get into MAC filter mis-match issue in the switch.
The following are changes since commit 6d24e14140053febc5ac1ce46baca6a4334c5f6c:
net/ncsi: update response packet length for GCPS/GNS/GNPTS commands
and are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue 100GbE
Akeem G Abodunrin (2):
ice: Fix issues updating VSI MAC filters
ice: Don't allow VSI to remove unassociated ucast filter
Amruth G.P (1):
ice: Add input handlers for virtual channel handlers
Brett Creeley (1):
ice: Don't call synchronize_irq() for VF's from the host
Bruce Allan (1):
ice: update ethtool stats on-demand
Chinh T Cao (2):
ice: Fix flag used for module query
ice: Don't clear auto_fec bit in ice_cfg_phy_fec()
Dave Ertman (4):
ice: Allow egress control packets from PF_VSI
ice: Account for all states of FW DCBx and LLDP
ice: Treat DCBx state NOT_STARTED as valid
ice: Rename ethtool private flag for lldp
Jacob Keller (1):
ice: reject VF attempts to enable head writeback
Michal Swiatkowski (1):
ice: Copy dcbx configuration only if mode is correct
Mitch Williams (1):
ice: silence some bogus error messages
drivers/net/ethernet/intel/ice/ice.h | 4 +-
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 1 +
drivers/net/ethernet/intel/ice/ice_common.c | 11 +-
drivers/net/ethernet/intel/ice/ice_dcb.c | 3 +-
drivers/net/ethernet/intel/ice/ice_dcb_lib.c | 42 +++----
drivers/net/ethernet/intel/ice/ice_ethtool.c | 13 +-
drivers/net/ethernet/intel/ice/ice_lib.c | 45 ++++++-
drivers/net/ethernet/intel/ice/ice_lib.h | 4 +
drivers/net/ethernet/intel/ice/ice_main.c | 70 +++--------
drivers/net/ethernet/intel/ice/ice_switch.c | 56 +++++++++
drivers/net/ethernet/intel/ice/ice_txrx.c | 11 +-
.../net/ethernet/intel/ice/ice_virtchnl_pf.c | 113 +++++++++---------
12 files changed, 225 insertions(+), 148 deletions(-)
--
2.21.0
^ permalink raw reply
* [net-next 03/14] ice: Don't call synchronize_irq() for VF's from the host
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem; +Cc: Brett Creeley, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Brett Creeley <brett.creeley@intel.com>
Currently we will call synchronize_irq() from the host for VF's. This is
not correct, so don't allow it.
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_lib.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index d6279dfe029e..c067ef6be7f4 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -2817,6 +2817,10 @@ void ice_vsi_dis_irq(struct ice_vsi *vsi)
ice_flush(hw);
+ /* don't call synchronize_irq() for VF's from the host */
+ if (vsi->type == ICE_VSI_VF)
+ return;
+
ice_for_each_q_vector(vsi, i)
synchronize_irq(pf->msix_entries[i + base].vector);
}
--
2.21.0
^ permalink raw reply related
* [net-next 10/14] ice: Don't clear auto_fec bit in ice_cfg_phy_fec()
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem; +Cc: Chinh T Cao, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Chinh T Cao <chinh.t.cao@intel.com>
The driver should never clear the auto_fec_enable bit.
Signed-off-by: Chinh T Cao <chinh.t.cao@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 9 +++------
1 file changed, 3 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 15648d4a8bab..4b43e6de847b 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -2181,27 +2181,24 @@ ice_cfg_phy_fec(struct ice_aqc_set_phy_cfg_data *cfg, enum ice_fec_mode fec)
{
switch (fec) {
case ICE_FEC_BASER:
- /* Clear auto FEC and RS bits, and AND BASE-R ability
+ /* Clear RS bits, and AND BASE-R ability
* bits and OR request bits.
*/
- cfg->caps &= ~ICE_AQC_PHY_EN_AUTO_FEC;
cfg->link_fec_opt &= ICE_AQC_PHY_FEC_10G_KR_40G_KR4_EN |
ICE_AQC_PHY_FEC_25G_KR_CLAUSE74_EN;
cfg->link_fec_opt |= ICE_AQC_PHY_FEC_10G_KR_40G_KR4_REQ |
ICE_AQC_PHY_FEC_25G_KR_REQ;
break;
case ICE_FEC_RS:
- /* Clear auto FEC and BASE-R bits, and AND RS ability
+ /* Clear BASE-R bits, and AND RS ability
* bits and OR request bits.
*/
- cfg->caps &= ~ICE_AQC_PHY_EN_AUTO_FEC;
cfg->link_fec_opt &= ICE_AQC_PHY_FEC_25G_RS_CLAUSE91_EN;
cfg->link_fec_opt |= ICE_AQC_PHY_FEC_25G_RS_528_REQ |
ICE_AQC_PHY_FEC_25G_RS_544_REQ;
break;
case ICE_FEC_NONE:
- /* Clear auto FEC and all FEC option bits. */
- cfg->caps &= ~ICE_AQC_PHY_EN_AUTO_FEC;
+ /* Clear all FEC option bits. */
cfg->link_fec_opt &= ~ICE_AQC_PHY_FEC_MASK;
break;
case ICE_FEC_AUTO:
--
2.21.0
^ permalink raw reply related
* [net-next 13/14] ice: Fix issues updating VSI MAC filters
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem
Cc: Akeem G Abodunrin, netdev, nhorman, sassmann, Tony Nguyen,
Andrew Bowers, Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
VSI, especially VF could request to add or remove filter for another VSI,
driver should really guide such request and disallow it.
However, instead of returning error for such malicious request, driver
can simply return success.
In addition, we are not tracking number of MAC filters configured per
VF correctly - and this leads to issue updating VF MAC filters whenever
they were removed and re-configured via bringing VF interface down and
up. Also, since VF could send request to update multiple MAC filters at
once, driver should program those filters individually in the switch, in
order to determine which action resulted to error, and communicate
accordingly to the VF.
So, with this changes, we now track number of filters added right from
when VF resources allocation is done, and could properly add filters for
both trusted and non_trusted VFs, without MAC filters mis-match issue in
the switch...
Also refactor code, so that driver can use new function to add or remove
MAC filters.
Signed-off-by: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_lib.c | 30 +++++++++
drivers/net/ethernet/intel/ice/ice_lib.h | 4 ++
drivers/net/ethernet/intel/ice/ice_main.c | 64 +++++--------------
.../net/ethernet/intel/ice/ice_virtchnl_pf.c | 42 ++++++------
4 files changed, 73 insertions(+), 67 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index 343d0c305423..8d5d6635a123 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -3181,3 +3181,33 @@ int ice_vsi_cfg_tc(struct ice_vsi *vsi, u8 ena_tc)
return ret;
}
#endif /* CONFIG_DCB */
+
+/**
+ * ice_vsi_cfg_mac_fltr - Add or remove a MAC address filter for a VSI
+ * @vsi: the VSI being configured MAC filter
+ * @macaddr: the MAC address to be added.
+ * @set: Add or delete a MAC filter
+ *
+ * Adds or removes MAC address filter entry for VF VSI
+ */
+enum ice_status
+ice_vsi_cfg_mac_fltr(struct ice_vsi *vsi, const u8 *macaddr, bool set)
+{
+ LIST_HEAD(tmp_add_list);
+ enum ice_status status;
+
+ /* Update MAC filter list to be added or removed for a VSI */
+ if (ice_add_mac_to_list(vsi, &tmp_add_list, macaddr)) {
+ status = ICE_ERR_NO_MEMORY;
+ goto cfg_mac_fltr_exit;
+ }
+
+ if (set)
+ status = ice_add_mac(&vsi->back->hw, &tmp_add_list);
+ else
+ status = ice_remove_mac(&vsi->back->hw, &tmp_add_list);
+
+cfg_mac_fltr_exit:
+ ice_free_fltr_list(&vsi->back->pdev->dev, &tmp_add_list);
+ return status;
+}
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.h b/drivers/net/ethernet/intel/ice/ice_lib.h
index 6e43ef03bfc3..969ba27cba95 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.h
+++ b/drivers/net/ethernet/intel/ice/ice_lib.h
@@ -95,4 +95,8 @@ void ice_vsi_free_tx_rings(struct ice_vsi *vsi);
int ice_vsi_manage_rss_lut(struct ice_vsi *vsi, bool ena);
u32 ice_intrl_usec_to_reg(u8 intrl, u8 gran);
+
+enum ice_status
+ice_vsi_cfg_mac_fltr(struct ice_vsi *vsi, const u8 *macaddr, bool set);
+
#endif /* !_ICE_LIB_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 6dd806b763ea..f3923dec32b7 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -116,10 +116,9 @@ static void ice_check_for_hang_subtask(struct ice_pf *pf)
*/
static int ice_init_mac_fltr(struct ice_pf *pf)
{
- LIST_HEAD(tmp_add_list);
+ enum ice_status status;
u8 broadcast[ETH_ALEN];
struct ice_vsi *vsi;
- int status;
vsi = ice_find_vsi_by_type(pf, ICE_VSI_PF);
if (!vsi)
@@ -130,8 +129,7 @@ static int ice_init_mac_fltr(struct ice_pf *pf)
*/
/* Add a unicast MAC filter so the VSI can get its packets */
- status = ice_add_mac_to_list(vsi, &tmp_add_list,
- vsi->port_info->mac.perm_addr);
+ status = ice_vsi_cfg_mac_fltr(vsi, vsi->port_info->mac.perm_addr, true);
if (status)
goto unregister;
@@ -139,18 +137,11 @@ static int ice_init_mac_fltr(struct ice_pf *pf)
* MAC address to the list as well.
*/
eth_broadcast_addr(broadcast);
- status = ice_add_mac_to_list(vsi, &tmp_add_list, broadcast);
- if (status)
- goto free_mac_list;
-
- /* Program MAC filters for entries in tmp_add_list */
- status = ice_add_mac(&pf->hw, &tmp_add_list);
+ status = ice_vsi_cfg_mac_fltr(vsi, broadcast, true);
if (status)
- status = -ENOMEM;
-
-free_mac_list:
- ice_free_fltr_list(&pf->pdev->dev, &tmp_add_list);
+ goto unregister;
+ return 0;
unregister:
/* We aren't useful with no MAC filters, so unregister if we
* had an error
@@ -164,7 +155,7 @@ static int ice_init_mac_fltr(struct ice_pf *pf)
vsi->netdev = NULL;
}
- return status;
+ return -EIO;
}
/**
@@ -2834,10 +2825,8 @@ static int ice_set_mac_address(struct net_device *netdev, void *pi)
struct ice_hw *hw = &pf->hw;
struct sockaddr *addr = pi;
enum ice_status status;
- LIST_HEAD(a_mac_list);
- LIST_HEAD(r_mac_list);
u8 flags = 0;
- int err;
+ int err = 0;
u8 *mac;
mac = (u8 *)addr->sa_data;
@@ -2860,42 +2849,23 @@ static int ice_set_mac_address(struct net_device *netdev, void *pi)
/* When we change the MAC address we also have to change the MAC address
* based filter rules that were created previously for the old MAC
* address. So first, we remove the old filter rule using ice_remove_mac
- * and then create a new filter rule using ice_add_mac. Note that for
- * both these operations, we first need to form a "list" of MAC
- * addresses (even though in this case, we have only 1 MAC address to be
- * added/removed) and this done using ice_add_mac_to_list. Depending on
- * the ensuing operation this "list" of MAC addresses is either to be
- * added or removed from the filter.
+ * and then create a new filter rule using ice_add_mac via
+ * ice_vsi_cfg_mac_fltr function call for both add and/or remove
+ * filters.
*/
- err = ice_add_mac_to_list(vsi, &r_mac_list, netdev->dev_addr);
- if (err) {
- err = -EADDRNOTAVAIL;
- goto free_lists;
- }
-
- status = ice_remove_mac(hw, &r_mac_list);
+ status = ice_vsi_cfg_mac_fltr(vsi, netdev->dev_addr, false);
if (status) {
err = -EADDRNOTAVAIL;
- goto free_lists;
- }
-
- err = ice_add_mac_to_list(vsi, &a_mac_list, mac);
- if (err) {
- err = -EADDRNOTAVAIL;
- goto free_lists;
+ goto err_update_filters;
}
- status = ice_add_mac(hw, &a_mac_list);
+ status = ice_vsi_cfg_mac_fltr(vsi, mac, true);
if (status) {
err = -EADDRNOTAVAIL;
- goto free_lists;
+ goto err_update_filters;
}
-free_lists:
- /* free list entries */
- ice_free_fltr_list(&pf->pdev->dev, &r_mac_list);
- ice_free_fltr_list(&pf->pdev->dev, &a_mac_list);
-
+err_update_filters:
if (err) {
netdev_err(netdev, "can't set MAC %pM. filter update failed\n",
mac);
@@ -2911,8 +2881,8 @@ static int ice_set_mac_address(struct net_device *netdev, void *pi)
flags = ICE_AQC_MAN_MAC_UPDATE_LAA_WOL;
status = ice_aq_manage_mac_write(hw, mac, flags, NULL);
if (status) {
- netdev_err(netdev, "can't set MAC %pM. write to firmware failed.\n",
- mac);
+ netdev_err(netdev, "can't set MAC %pM. write to firmware failed error %d\n",
+ mac, status);
}
return 0;
}
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
index de0a1ef54e83..86637d99ee77 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
@@ -540,7 +540,10 @@ static int ice_alloc_vsi_res(struct ice_vf *vf)
status = ice_add_mac(&pf->hw, &tmp_add_list);
if (status)
- dev_err(&pf->pdev->dev, "could not add mac filters\n");
+ dev_err(&pf->pdev->dev,
+ "could not add mac filters error %d\n", status);
+ else
+ vf->num_mac = 1;
/* Clear this bit after VF initialization since we shouldn't reclaim
* and reassign interrupts for synchronous or asynchronous VFR events.
@@ -2208,7 +2211,7 @@ ice_vc_handle_mac_addr_msg(struct ice_vf *vf, u8 *msg, bool set)
(struct virtchnl_ether_addr_list *)msg;
struct ice_pf *pf = vf->pf;
enum virtchnl_ops vc_op;
- LIST_HEAD(mac_list);
+ enum ice_status status;
struct ice_vsi *vsi;
int mac_count = 0;
int i;
@@ -2282,33 +2285,32 @@ ice_vc_handle_mac_addr_msg(struct ice_vf *vf, u8 *msg, bool set)
goto handle_mac_exit;
}
- /* get here if maddr is multicast or if VF can change MAC */
- if (ice_add_mac_to_list(vsi, &mac_list, al->list[i].addr)) {
- v_ret = VIRTCHNL_STATUS_ERR_NO_MEMORY;
+ /* program the updated filter list */
+ status = ice_vsi_cfg_mac_fltr(vsi, maddr, set);
+ if (status == ICE_ERR_DOES_NOT_EXIST ||
+ status == ICE_ERR_ALREADY_EXISTS) {
+ dev_info(&pf->pdev->dev,
+ "can't %s MAC filters %pM for VF %d, error %d\n",
+ set ? "add" : "remove", maddr, vf->vf_id,
+ status);
+ } else if (status) {
+ dev_err(&pf->pdev->dev,
+ "can't %s MAC filters for VF %d, error %d\n",
+ set ? "add" : "remove", vf->vf_id, status);
+ v_ret = ice_err_to_virt_err(status);
goto handle_mac_exit;
}
+
mac_count++;
}
- /* program the updated filter list */
+ /* Track number of MAC filters programmed for the VF VSI */
if (set)
- v_ret = ice_err_to_virt_err(ice_add_mac(&pf->hw, &mac_list));
+ vf->num_mac += mac_count;
else
- v_ret = ice_err_to_virt_err(ice_remove_mac(&pf->hw, &mac_list));
-
- if (v_ret) {
- dev_err(&pf->pdev->dev,
- "can't %s MAC filters for VF %d, error %d\n",
- set ? "add" : "remove", vf->vf_id, v_ret);
- } else {
- if (set)
- vf->num_mac += mac_count;
- else
- vf->num_mac -= mac_count;
- }
+ vf->num_mac -= mac_count;
handle_mac_exit:
- ice_free_fltr_list(&pf->pdev->dev, &mac_list);
/* send the response to the VF */
return ice_vc_send_msg_to_vf(vf, vc_op, v_ret, NULL, 0);
}
--
2.21.0
^ permalink raw reply related
* [net-next 07/14] ice: Rename ethtool private flag for lldp
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem; +Cc: Dave Ertman, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Dave Ertman <david.m.ertman@intel.com>
The current flag name of "enable-fw-lldp" is a bit cumbersome.
Change priv-flag name to "fw-lldp-agent" with a value of on or
off. This is more straight-forward in meaning.
Signed-off-by: Dave Ertman <david.m.ertman@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice.h | 2 +-
drivers/net/ethernet/intel/ice/ice_dcb_lib.c | 4 ++--
drivers/net/ethernet/intel/ice/ice_ethtool.c | 6 +++---
drivers/net/ethernet/intel/ice/ice_lib.c | 4 ++--
4 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 9f9c30d29eb5..99e0febd8e50 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -329,7 +329,7 @@ enum ice_pf_flags {
ICE_FLAG_DCB_ENA,
ICE_FLAG_LINK_DOWN_ON_CLOSE_ENA,
ICE_FLAG_NO_MEDIA,
- ICE_FLAG_ENABLE_FW_LLDP,
+ ICE_FLAG_FW_LLDP_AGENT,
ICE_FLAG_ETHTOOL_CTXT, /* set when ethtool holds RTNL lock */
ICE_PF_FLAGS_NBITS /* must be last */
};
diff --git a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
index 4fc9faf5bc71..734cef8eed9e 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
@@ -457,7 +457,7 @@ int ice_init_pf_dcb(struct ice_pf *pf, bool locked)
dev_info(&pf->pdev->dev,
"FW LLDP is disabled, DCBx/LLDP in SW mode.\n");
port_info->is_sw_lldp = true;
- clear_bit(ICE_FLAG_ENABLE_FW_LLDP, pf->flags);
+ clear_bit(ICE_FLAG_FW_LLDP_AGENT, pf->flags);
err = ice_dcb_sw_dflt_cfg(pf, locked);
if (err) {
dev_err(&pf->pdev->dev,
@@ -473,7 +473,7 @@ int ice_init_pf_dcb(struct ice_pf *pf, bool locked)
}
port_info->is_sw_lldp = false;
- set_bit(ICE_FLAG_ENABLE_FW_LLDP, pf->flags);
+ set_bit(ICE_FLAG_FW_LLDP_AGENT, pf->flags);
/* DCBX in FW and LLDP enabled in FW */
pf->dcbx_cap = DCB_CAP_DCBX_LLD_MANAGED | DCB_CAP_DCBX_VER_IEEE;
diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index 6a97ddbbda76..948a33716290 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -155,7 +155,7 @@ struct ice_priv_flag {
static const struct ice_priv_flag ice_gstrings_priv_flags[] = {
ICE_PRIV_FLAG("link-down-on-close", ICE_FLAG_LINK_DOWN_ON_CLOSE_ENA),
- ICE_PRIV_FLAG("enable-fw-lldp", ICE_FLAG_ENABLE_FW_LLDP),
+ ICE_PRIV_FLAG("fw-lldp-agent", ICE_FLAG_FW_LLDP_AGENT),
};
#define ICE_PRIV_FLAG_ARRAY_SIZE ARRAY_SIZE(ice_gstrings_priv_flags)
@@ -1201,8 +1201,8 @@ static int ice_set_priv_flags(struct net_device *netdev, u32 flags)
bitmap_xor(change_flags, pf->flags, orig_flags, ICE_PF_FLAGS_NBITS);
- if (test_bit(ICE_FLAG_ENABLE_FW_LLDP, change_flags)) {
- if (!test_bit(ICE_FLAG_ENABLE_FW_LLDP, pf->flags)) {
+ if (test_bit(ICE_FLAG_FW_LLDP_AGENT, change_flags)) {
+ if (!test_bit(ICE_FLAG_FW_LLDP_AGENT, pf->flags)) {
enum ice_status status;
/* Disable FW LLDP engine */
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index c067ef6be7f4..343d0c305423 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -2541,7 +2541,7 @@ ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
ice_cfg_sw_lldp(vsi, true, true);
/* Rx LLDP packets */
- if (!test_bit(ICE_FLAG_ENABLE_FW_LLDP, pf->flags))
+ if (!test_bit(ICE_FLAG_FW_LLDP_AGENT, pf->flags))
ice_cfg_sw_lldp(vsi, false, true);
}
@@ -2888,7 +2888,7 @@ int ice_vsi_release(struct ice_vsi *vsi)
/* The Rx rule will only exist to remove if the LLDP FW
* engine is currently stopped
*/
- if (!test_bit(ICE_FLAG_ENABLE_FW_LLDP, pf->flags))
+ if (!test_bit(ICE_FLAG_FW_LLDP_AGENT, pf->flags))
ice_cfg_sw_lldp(vsi, false, false);
}
--
2.21.0
^ permalink raw reply related
* [net-next 06/14] ice: reject VF attempts to enable head writeback
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, Andrew Bowers,
Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Jacob Keller <jacob.e.keller@intel.com>
The virtchnl interface provides a mechanism for a VF driver to request
head writeback support. This feature is deprecated as of AVF 1.0, but
older versions of a VF driver may still attempt to request the mode.
Since the ice hardware does not support head writeback, we should not
accept Tx queue configuration which attempts to enable it.
Currently, the driver simply assumes that the headwb_enabled bit will
never be set.
If a VF driver does request head writeback, the configuration will
return successfully, even though head writeback is not enabled. This
leaves the VF driver in a non functional state since it is assuming to
be operating in head writeback mode.
Fix the PF driver to reject any attempt to setup headwb_enabled.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
index 1b1d1ea0c8f9..73ab6222d29b 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
@@ -2109,6 +2109,7 @@ static int ice_vc_cfg_qs_msg(struct ice_vf *vf, u8 *msg)
if (qpi->txq.vsi_id != qci->vsi_id ||
qpi->rxq.vsi_id != qci->vsi_id ||
qpi->rxq.queue_id != qpi->txq.queue_id ||
+ qpi->txq.headwb_enabled ||
!ice_vc_isvalid_q_id(vf, qci->vsi_id, qpi->txq.queue_id)) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
goto error_param;
--
2.21.0
^ permalink raw reply related
* [net-next 08/14] ice: silence some bogus error messages
From: Jeff Kirsher @ 2019-08-23 23:37 UTC (permalink / raw)
To: davem
Cc: Mitch Williams, netdev, nhorman, sassmann, Tony Nguyen,
Andrew Bowers, Jeff Kirsher
In-Reply-To: <20190823233750.7997-1-jeffrey.t.kirsher@intel.com>
From: Mitch Williams <mitch.a.williams@intel.com>
In some circumstances, VF devices can be deactivated while a message is
in-flight. In that case, a series of scary error message will be
printed in the log. Since these are actually harmless, check for this
case and suppress them. No harm, no foul.
Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/ice/ice_adminq_cmd.h | 1 +
drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c | 6 +++---
2 files changed, 4 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 765e3c2ed045..bf9aa533a7c6 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -1610,6 +1610,7 @@ enum ice_aq_err {
ICE_AQ_RC_EBUSY = 12, /* Device or resource busy */
ICE_AQ_RC_EEXIST = 13, /* Object already exists */
ICE_AQ_RC_ENOSPC = 16, /* No space left or allocation failure */
+ ICE_AQ_RC_ENOSYS = 17, /* Function not implemented */
};
/* Admin Queue command opcodes */
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
index 73ab6222d29b..83e58e71081e 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
@@ -1512,10 +1512,10 @@ ice_vc_send_msg_to_vf(struct ice_vf *vf, u32 v_opcode,
aq_ret = ice_aq_send_msg_to_vf(&pf->hw, vf->vf_id, v_opcode, v_retval,
msg, msglen, NULL);
- if (aq_ret) {
+ if (aq_ret && pf->hw.mailboxq.sq_last_status != ICE_AQ_RC_ENOSYS) {
dev_info(&pf->pdev->dev,
- "Unable to send the message to VF %d aq_err %d\n",
- vf->vf_id, pf->hw.mailboxq.sq_last_status);
+ "Unable to send the message to VF %d ret %d aq_err %d\n",
+ vf->vf_id, aq_ret, pf->hw.mailboxq.sq_last_status);
return -EIO;
}
--
2.21.0
^ permalink raw reply related
* [net PATCH] net: route dump netlink NLM_F_MULTI flag missing
From: John Fastabend @ 2019-08-24 0:11 UTC (permalink / raw)
To: sbrivio, davem, dsahern; +Cc: netdev, john.fastabend
An excerpt from netlink(7) man page,
In multipart messages (multiple nlmsghdr headers with associated payload
in one byte stream) the first and all following headers have the
NLM_F_MULTI flag set, except for the last header which has the type
NLMSG_DONE.
but, after (ee28906) there is a missing NLM_F_MULTI flag in the middle of a
FIB dump. The result is user space applications following above man page
excerpt may get confused and may stop parsing msg believing something went
wrong.
In the golang netlink lib [0] the library logic stops parsing believing the
message is not a multipart message. Found this running Cilium[1] against
net-next while adding a feature to auto-detect routes. I noticed with
multiple route tables we no longer could detect the default routes on net
tree kernels because the library logic was not returning them.
Fix this by handling the fib_dump_info_fnhe() case the same way the
fib_dump_info() handles it by passing the flags argument through the
call chain and adding a flags argument to rt_fill_info().
Tested with Cilium stack and auto-detection of routes works again. Also
annotated libs to dump netlink msgs and inspected NLM_F_MULTI and
NLMSG_DONE flags look correct after this.
Note: In inet_rtm_getroute() pass rt_fill_info() '0' for flags the same
as is done for fib_dump_info() so this looks correct to me.
[0] https://github.com/vishvananda/netlink/
[1] https://github.com/cilium/
Fixes: ee28906fd7a14 ("ipv4: Dump route exceptions if requested")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
include/net/route.h | 2 +-
net/ipv4/fib_trie.c | 2 +-
net/ipv4/route.c | 17 ++++++++++-------
3 files changed, 12 insertions(+), 9 deletions(-)
diff --git a/include/net/route.h b/include/net/route.h
index 630a0493f1f3..dfce19c9fa96 100644
--- a/include/net/route.h
+++ b/include/net/route.h
@@ -233,7 +233,7 @@ void rt_del_uncached_list(struct rtable *rt);
int fib_dump_info_fnhe(struct sk_buff *skb, struct netlink_callback *cb,
u32 table_id, struct fib_info *fi,
- int *fa_index, int fa_start);
+ int *fa_index, int fa_start, unsigned int flags);
static inline void ip_rt_put(struct rtable *rt)
{
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index 2b2b3d291ab0..1ab2fb6bb37d 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -2145,7 +2145,7 @@ static int fn_trie_dump_leaf(struct key_vector *l, struct fib_table *tb,
if (filter->dump_exceptions) {
err = fib_dump_info_fnhe(skb, cb, tb->tb_id, fi,
- &i_fa, s_fa);
+ &i_fa, s_fa, flags);
if (err < 0)
goto stop;
}
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 517300d587a7..b6a6f18c3dd1 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -2728,7 +2728,8 @@ EXPORT_SYMBOL_GPL(ip_route_output_flow);
/* called with rcu_read_lock held */
static int rt_fill_info(struct net *net, __be32 dst, __be32 src,
struct rtable *rt, u32 table_id, struct flowi4 *fl4,
- struct sk_buff *skb, u32 portid, u32 seq)
+ struct sk_buff *skb, u32 portid, u32 seq,
+ unsigned int flags)
{
struct rtmsg *r;
struct nlmsghdr *nlh;
@@ -2736,7 +2737,7 @@ static int rt_fill_info(struct net *net, __be32 dst, __be32 src,
u32 error;
u32 metrics[RTAX_MAX];
- nlh = nlmsg_put(skb, portid, seq, RTM_NEWROUTE, sizeof(*r), 0);
+ nlh = nlmsg_put(skb, portid, seq, RTM_NEWROUTE, sizeof(*r), flags);
if (!nlh)
return -EMSGSIZE;
@@ -2860,7 +2861,7 @@ static int rt_fill_info(struct net *net, __be32 dst, __be32 src,
static int fnhe_dump_bucket(struct net *net, struct sk_buff *skb,
struct netlink_callback *cb, u32 table_id,
struct fnhe_hash_bucket *bucket, int genid,
- int *fa_index, int fa_start)
+ int *fa_index, int fa_start, unsigned int flags)
{
int i;
@@ -2891,7 +2892,7 @@ static int fnhe_dump_bucket(struct net *net, struct sk_buff *skb,
err = rt_fill_info(net, fnhe->fnhe_daddr, 0, rt,
table_id, NULL, skb,
NETLINK_CB(cb->skb).portid,
- cb->nlh->nlmsg_seq);
+ cb->nlh->nlmsg_seq, flags);
if (err)
return err;
next:
@@ -2904,7 +2905,7 @@ static int fnhe_dump_bucket(struct net *net, struct sk_buff *skb,
int fib_dump_info_fnhe(struct sk_buff *skb, struct netlink_callback *cb,
u32 table_id, struct fib_info *fi,
- int *fa_index, int fa_start)
+ int *fa_index, int fa_start, unsigned int flags)
{
struct net *net = sock_net(cb->skb->sk);
int nhsel, genid = fnhe_genid(net);
@@ -2922,7 +2923,8 @@ int fib_dump_info_fnhe(struct sk_buff *skb, struct netlink_callback *cb,
err = 0;
if (bucket)
err = fnhe_dump_bucket(net, skb, cb, table_id, bucket,
- genid, fa_index, fa_start);
+ genid, fa_index, fa_start,
+ flags);
rcu_read_unlock();
if (err)
return err;
@@ -3183,7 +3185,8 @@ static int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh,
fl4.flowi4_tos, res.fi, 0);
} else {
err = rt_fill_info(net, dst, src, rt, table_id, &fl4, skb,
- NETLINK_CB(in_skb).portid, nlh->nlmsg_seq);
+ NETLINK_CB(in_skb).portid,
+ nlh->nlmsg_seq, 0);
}
if (err < 0)
goto errout_rcu;
^ permalink raw reply related
* pull-request: bpf 2019-08-24
From: Daniel Borkmann @ 2019-08-24 0:11 UTC (permalink / raw)
To: davem, jakub.kicinski; +Cc: daniel, ast, andrii.nakryiko, netdev, bpf
Hi David,
The following pull-request contains BPF updates for your *net* tree.
The main changes are:
1) Fix verifier precision tracking with BPF-to-BPF calls, from Alexei.
2) Fix a use-after-free in prog symbol exposure, from Daniel.
3) Several s390x JIT fixes plus BE related fixes in BPF kselftests, from Ilya.
4) Fix memory leak by unpinning XDP umem pages in error path, from Ivan.
5) Fix a potential use-after-free on flow dissector detach, from Jakub.
6) Fix bpftool to close prog fd after showing metadata, from Quentin.
7) BPF kselftest config and TEST_PROGS_EXTENDED fixes, from Anders.
Please consider pulling these changes from:
git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf.git
Thanks a lot!
----------------------------------------------------------------
The following changes since commit 125b7e0949d4e72b15c2b1a1590f8cece985a918:
net: tc35815: Explicitly check NET_IP_ALIGN is not zero in tc35815_rx (2019-08-11 21:41:48 -0700)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf.git
for you to fetch changes up to 2c238177bd7f4b14bdf7447cc1cd9bb791f147e6:
bpf: allow narrow loads of some sk_reuseport_md fields with offset > 0 (2019-08-24 01:25:41 +0200)
----------------------------------------------------------------
Alexei Starovoitov (1):
bpf: fix precision tracking in presence of bpf2bpf calls
Anders Roxell (2):
selftests/bpf: add config fragment BPF_JIT
selftests/bpf: install files test_xdp_vlan.sh
Daniel Borkmann (1):
bpf: fix use after free in prog symbol exposure
Ilya Leoshkevich (6):
s390/bpf: fix lcgr instruction encoding
s390/bpf: use 32-bit index for tail calls
selftests/bpf: fix "bind{4, 6} deny specific IP & port" on s390
selftests/bpf: fix test_cgroup_storage on s390
selftests/bpf: fix test_btf_dump with O=
bpf: allow narrow loads of some sk_reuseport_md fields with offset > 0
Ivan Khoronzhuk (1):
xdp: unpin xdp umem pages in error path
Jakub Sitnicki (1):
flow_dissector: Fix potential use-after-free on BPF_PROG_DETACH
Quentin Monnet (1):
tools: bpftool: close prog FD before exit on showing a single program
arch/s390/net/bpf_jit_comp.c | 12 +++++----
kernel/bpf/syscall.c | 30 ++++++++++++++---------
kernel/bpf/verifier.c | 9 ++++---
net/core/filter.c | 8 +++---
net/core/flow_dissector.c | 2 +-
net/xdp/xdp_umem.c | 4 ++-
tools/bpf/bpftool/prog.c | 4 ++-
tools/testing/selftests/bpf/Makefile | 6 ++++-
tools/testing/selftests/bpf/config | 1 +
tools/testing/selftests/bpf/test_btf_dump.c | 7 ++++++
tools/testing/selftests/bpf/test_cgroup_storage.c | 6 ++---
tools/testing/selftests/bpf/test_sock.c | 7 ++++--
12 files changed, 62 insertions(+), 34 deletions(-)
^ permalink raw reply
* Re: [PATCHv4 net 1/2] ipv4/icmp: fix rt dst dev null pointer dereference
From: Jonathan Lemon @ 2019-08-23 22:32 UTC (permalink / raw)
To: Hangbin Liu
Cc: netdev, Stefano Brivio, wenxu, Alexei Starovoitov,
David S . Miller, Eric Dumazet, Julian Anastasov
In-Reply-To: <20190822141949.29561-2-liuhangbin@gmail.com>
On 22 Aug 2019, at 7:19, Hangbin Liu wrote:
> In __icmp_send() there is a possibility that the rt->dst.dev is NULL,
> e,g, with tunnel collect_md mode, which will cause kernel crash.
> Here is what the code path looks like, for GRE:
>
> - ip6gre_tunnel_xmit
> - ip6gre_xmit_ipv4
> - __gre6_xmit
> - ip6_tnl_xmit
> - if skb->len - t->tun_hlen - eth_hlen > mtu; return -EMSGSIZE
> - icmp_send
> - net = dev_net(rt->dst.dev); <-- here
>
> The reason is __metadata_dst_init() init dst->dev to NULL by default.
> We could not fix it in __metadata_dst_init() as there is no dev supplied.
> On the other hand, the reason we need rt->dst.dev is to get the net.
> So we can just try get it from skb->dev when rt->dst.dev is NULL.
>
> v4: Julian Anastasov remind skb->dev also could be NULL. We'd better
> still use dst.dev and do a check to avoid crash.
>
> v3: No changes.
>
> v2: fix the issue in __icmp_send() instead of updating shared dst dev
> in {ip_md, ip6}_tunnel_xmit.
>
> Fixes: c8b34e680a09 ("ip_tunnel: Add tnl_update_pmtu in ip_md_tunnel_xmit")
> Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Jonathan Lemon <jonathan.lemon@gmail.com>
^ permalink raw reply
* Re: [PATCH net-next v2 03/10] net: sched: refactor block offloads counter usage
From: Jakub Kicinski @ 2019-08-24 0:26 UTC (permalink / raw)
To: Vlad Buslov; +Cc: netdev, jhs, xiyou.wangcong, jiri, davem, pablo
In-Reply-To: <20190823185056.12536-4-vladbu@mellanox.com>
On Fri, 23 Aug 2019 21:50:49 +0300, Vlad Buslov wrote:
> @@ -1201,14 +1199,11 @@ static int u32_reoffload_knode(struct tcf_proto *tp, struct tc_u_knode *n,
> cls_u32.knode.link_handle = ht->handle;
> }
>
> - err = cb(TC_SETUP_CLSU32, &cls_u32, cb_priv);
> - if (err) {
> - if (add && tc_skip_sw(n->flags))
> - return err;
> - return 0;
> - }
> -
> - tc_cls_offload_cnt_update(block, &n->in_hw_count, &n->flags, add);
> + err = tc_setup_cb_reoffload(block, tp, add, cb, TC_SETUP_CLSU32,
> + &cls_u32, cb_priv, &n->flags,
> + &n->in_hw_count);
> + if (err && add && tc_skip_sw(n->flags))
> + return err;
Could this be further simplified by adding something along the lines of:
if (!add || !tc_skip_sw(*flags))
err = 0;
to tc_setup_cb_reoffload() ?
>
> return 0;
> }
^ permalink raw reply
* Re: [PATCH v2] riscv: add support for SECCOMP and SECCOMP_FILTER
From: Paul Walmsley @ 2019-08-24 0:30 UTC (permalink / raw)
To: David Abdurachmanov, Tycho Andersen
Cc: Palmer Dabbelt, Albert Ou, Oleg Nesterov, Kees Cook,
Andy Lutomirski, Will Drewry, Shuah Khan, Alexei Starovoitov,
Daniel Borkmann, Martin KaFai Lau, Song Liu, Yonghong Song,
David Abdurachmanov, Thomas Gleixner, Allison Randal,
Alexios Zavras, Anup Patel, Vincent Chen, Alan Kao, linux-riscv,
linux-kernel, linux-kselftest, netdev, bpf, me
In-Reply-To: <20190822205533.4877-1-david.abdurachmanov@sifive.com>
On Thu, 22 Aug 2019, David Abdurachmanov wrote:
> There is one failing kernel selftest: global.user_notification_signal
Is this the only failing test? Or are the rest of the selftests skipped
when this test fails, and no further tests are run, as seems to be shown
here:
https://lore.kernel.org/linux-riscv/CADnnUqcmDMRe1f+3jG8SPR6jRrnBsY8VVD70VbKEm0NqYeoicA@mail.gmail.com/
For example, looking at the source, I'd naively expect to see the
user_notification_closed_listener test result -- which follows right
after the failing test in the selftest source. But there aren't any
results?
Also - could you follow up with the author of this failing test to see if
we can get some more clarity about what might be going wrong here? It
appears that the failing test was added in commit 6a21cc50f0c7f ("seccomp:
add a return code to trap to userspace") by Tycho Andersen
<tycho@tycho.ws>.
- Paul
^ permalink raw reply
* Re: pull-request: bpf 2019-08-24
From: David Miller @ 2019-08-24 0:34 UTC (permalink / raw)
To: daniel; +Cc: jakub.kicinski, ast, andrii.nakryiko, netdev, bpf
In-Reply-To: <20190824001157.16043-1-daniel@iogearbox.net>
From: Daniel Borkmann <daniel@iogearbox.net>
Date: Sat, 24 Aug 2019 02:11:57 +0200
> The following pull-request contains BPF updates for your *net* tree.
>
> The main changes are:
>
> 1) Fix verifier precision tracking with BPF-to-BPF calls, from Alexei.
>
> 2) Fix a use-after-free in prog symbol exposure, from Daniel.
>
> 3) Several s390x JIT fixes plus BE related fixes in BPF kselftests, from Ilya.
>
> 4) Fix memory leak by unpinning XDP umem pages in error path, from Ivan.
>
> 5) Fix a potential use-after-free on flow dissector detach, from Jakub.
>
> 6) Fix bpftool to close prog fd after showing metadata, from Quentin.
>
> 7) BPF kselftest config and TEST_PROGS_EXTENDED fixes, from Anders.
>
> Please consider pulling these changes from:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf.git
Pulled, thanks Daniel.
^ permalink raw reply
* [PATCHv2 1/1] net: rds: add service level support in rds-info
From: Zhu Yanjun @ 2019-08-24 1:04 UTC (permalink / raw)
To: santosh.shilimkar, davem, netdev, linux-rdma, rds-devel,
gerd.rausch
From IB specific 7.6.5 SERVICE LEVEL, Service Level (SL)
is used to identify different flows within an IBA subnet.
It is carried in the local route header of the packet.
Before this commit, run "rds-info -I". The outputs are as
below:
"
RDS IB Connections:
LocalAddr RemoteAddr Tos SL LocalDev RemoteDev
192.2.95.3 192.2.95.1 2 0 fe80::21:28:1a:39 fe80::21:28:10:b9
192.2.95.3 192.2.95.1 1 0 fe80::21:28:1a:39 fe80::21:28:10:b9
192.2.95.3 192.2.95.1 0 0 fe80::21:28:1a:39 fe80::21:28:10:b9
"
After this commit, the output is as below:
"
RDS IB Connections:
LocalAddr RemoteAddr Tos SL LocalDev RemoteDev
192.2.95.3 192.2.95.1 2 2 fe80::21:28:1a:39 fe80::21:28:10:b9
192.2.95.3 192.2.95.1 1 1 fe80::21:28:1a:39 fe80::21:28:10:b9
192.2.95.3 192.2.95.1 0 0 fe80::21:28:1a:39 fe80::21:28:10:b9
"
The commit fe3475af3bdf ("net: rds: add per rds connection cache
statistics") adds cache_allocs in struct rds_info_rdma_connection
as below:
struct rds_info_rdma_connection {
...
__u32 rdma_mr_max;
__u32 rdma_mr_size;
__u8 tos;
__u32 cache_allocs;
};
The peer struct in rds-tools of struct rds_info_rdma_connection is as
below:
struct rds_info_rdma_connection {
...
uint32_t rdma_mr_max;
uint32_t rdma_mr_size;
uint8_t tos;
uint8_t sl;
uint32_t cache_allocs;
};
The difference between userspace and kernel is the member variable sl.
In the kernel struct, the member variable sl is missing. This will
introduce risks. So it is necessary to use this commit to avoid this risk.
Fixes: fe3475af3bdf ("net: rds: add per rds connection cache statistics")
CC: Joe Jin <joe.jin@oracle.com>
CC: JUNXIAO_BI <junxiao.bi@oracle.com>
Suggested-by: Gerd Rausch <gerd.rausch@oracle.com>
Signed-off-by: Zhu Yanjun <yanjun.zhu@oracle.com>
---
V1->V2: fix typos in commit logs.
---
include/uapi/linux/rds.h | 2 ++
net/rds/ib.c | 16 ++++++++++------
net/rds/ib.h | 1 +
net/rds/ib_cm.c | 3 +++
net/rds/rdma_transport.c | 10 ++++++++--
5 files changed, 24 insertions(+), 8 deletions(-)
diff --git a/include/uapi/linux/rds.h b/include/uapi/linux/rds.h
index fd6b5f6..cba368e 100644
--- a/include/uapi/linux/rds.h
+++ b/include/uapi/linux/rds.h
@@ -250,6 +250,7 @@ struct rds_info_rdma_connection {
__u32 rdma_mr_max;
__u32 rdma_mr_size;
__u8 tos;
+ __u8 sl;
__u32 cache_allocs;
};
@@ -265,6 +266,7 @@ struct rds6_info_rdma_connection {
__u32 rdma_mr_max;
__u32 rdma_mr_size;
__u8 tos;
+ __u8 sl;
__u32 cache_allocs;
};
diff --git a/net/rds/ib.c b/net/rds/ib.c
index ec05d91..45acab2 100644
--- a/net/rds/ib.c
+++ b/net/rds/ib.c
@@ -291,7 +291,7 @@ static int rds_ib_conn_info_visitor(struct rds_connection *conn,
void *buffer)
{
struct rds_info_rdma_connection *iinfo = buffer;
- struct rds_ib_connection *ic;
+ struct rds_ib_connection *ic = conn->c_transport_data;
/* We will only ever look at IB transports */
if (conn->c_trans != &rds_ib_transport)
@@ -301,15 +301,16 @@ static int rds_ib_conn_info_visitor(struct rds_connection *conn,
iinfo->src_addr = conn->c_laddr.s6_addr32[3];
iinfo->dst_addr = conn->c_faddr.s6_addr32[3];
- iinfo->tos = conn->c_tos;
+ if (ic) {
+ iinfo->tos = conn->c_tos;
+ iinfo->sl = ic->i_sl;
+ }
memset(&iinfo->src_gid, 0, sizeof(iinfo->src_gid));
memset(&iinfo->dst_gid, 0, sizeof(iinfo->dst_gid));
if (rds_conn_state(conn) == RDS_CONN_UP) {
struct rds_ib_device *rds_ibdev;
- ic = conn->c_transport_data;
-
rdma_read_gids(ic->i_cm_id, (union ib_gid *)&iinfo->src_gid,
(union ib_gid *)&iinfo->dst_gid);
@@ -329,7 +330,7 @@ static int rds6_ib_conn_info_visitor(struct rds_connection *conn,
void *buffer)
{
struct rds6_info_rdma_connection *iinfo6 = buffer;
- struct rds_ib_connection *ic;
+ struct rds_ib_connection *ic = conn->c_transport_data;
/* We will only ever look at IB transports */
if (conn->c_trans != &rds_ib_transport)
@@ -337,6 +338,10 @@ static int rds6_ib_conn_info_visitor(struct rds_connection *conn,
iinfo6->src_addr = conn->c_laddr;
iinfo6->dst_addr = conn->c_faddr;
+ if (ic) {
+ iinfo6->tos = conn->c_tos;
+ iinfo6->sl = ic->i_sl;
+ }
memset(&iinfo6->src_gid, 0, sizeof(iinfo6->src_gid));
memset(&iinfo6->dst_gid, 0, sizeof(iinfo6->dst_gid));
@@ -344,7 +349,6 @@ static int rds6_ib_conn_info_visitor(struct rds_connection *conn,
if (rds_conn_state(conn) == RDS_CONN_UP) {
struct rds_ib_device *rds_ibdev;
- ic = conn->c_transport_data;
rdma_read_gids(ic->i_cm_id, (union ib_gid *)&iinfo6->src_gid,
(union ib_gid *)&iinfo6->dst_gid);
rds_ibdev = ic->rds_ibdev;
diff --git a/net/rds/ib.h b/net/rds/ib.h
index 303c6ee..f2b558e 100644
--- a/net/rds/ib.h
+++ b/net/rds/ib.h
@@ -220,6 +220,7 @@ struct rds_ib_connection {
/* Send/Recv vectors */
int i_scq_vector;
int i_rcq_vector;
+ u8 i_sl;
};
/* This assumes that atomic_t is at least 32 bits */
diff --git a/net/rds/ib_cm.c b/net/rds/ib_cm.c
index fddaa09..233f136 100644
--- a/net/rds/ib_cm.c
+++ b/net/rds/ib_cm.c
@@ -152,6 +152,9 @@ void rds_ib_cm_connect_complete(struct rds_connection *conn, struct rdma_cm_even
RDS_PROTOCOL_MINOR(conn->c_version),
ic->i_flowctl ? ", flow control" : "");
+ /* receive sl from the peer */
+ ic->i_sl = ic->i_cm_id->route.path_rec->sl;
+
atomic_set(&ic->i_cq_quiesce, 0);
/* Init rings and fill recv. this needs to wait until protocol
diff --git a/net/rds/rdma_transport.c b/net/rds/rdma_transport.c
index ff74c4b..28668ad 100644
--- a/net/rds/rdma_transport.c
+++ b/net/rds/rdma_transport.c
@@ -43,6 +43,9 @@
static struct rdma_cm_id *rds6_rdma_listen_id;
#endif
+/* Per IB specification 7.7.3, service level is a 4-bit field. */
+#define TOS_TO_SL(tos) ((tos) & 0xF)
+
static int rds_rdma_cm_event_handler_cmn(struct rdma_cm_id *cm_id,
struct rdma_cm_event *event,
bool isv6)
@@ -97,10 +100,13 @@ static int rds_rdma_cm_event_handler_cmn(struct rdma_cm_id *cm_id,
struct rds_ib_connection *ibic;
ibic = conn->c_transport_data;
- if (ibic && ibic->i_cm_id == cm_id)
+ if (ibic && ibic->i_cm_id == cm_id) {
+ cm_id->route.path_rec[0].sl =
+ TOS_TO_SL(conn->c_tos);
ret = trans->cm_initiate_connect(cm_id, isv6);
- else
+ } else {
rds_conn_drop(conn);
+ }
}
break;
--
1.7.1
^ permalink raw reply related
* Re: [PATCH v2] riscv: add support for SECCOMP and SECCOMP_FILTER
From: David Abdurachmanov @ 2019-08-24 1:04 UTC (permalink / raw)
To: Paul Walmsley
Cc: Tycho Andersen, Palmer Dabbelt, Albert Ou, Oleg Nesterov,
Kees Cook, Andy Lutomirski, Will Drewry, Shuah Khan,
Alexei Starovoitov, Daniel Borkmann, Martin KaFai Lau, Song Liu,
Yonghong Song, David Abdurachmanov, Thomas Gleixner,
Allison Randal, Alexios Zavras, Anup Patel, Vincent Chen,
Alan Kao, linux-riscv, linux-kernel, linux-kselftest, netdev, bpf,
me
In-Reply-To: <alpine.DEB.2.21.9999.1908231717550.25649@viisi.sifive.com>
On Fri, Aug 23, 2019 at 5:30 PM Paul Walmsley <paul.walmsley@sifive.com> wrote:
>
> On Thu, 22 Aug 2019, David Abdurachmanov wrote:
>
> > There is one failing kernel selftest: global.user_notification_signal
>
> Is this the only failing test? Or are the rest of the selftests skipped
> when this test fails, and no further tests are run, as seems to be shown
> here:
>
> https://lore.kernel.org/linux-riscv/CADnnUqcmDMRe1f+3jG8SPR6jRrnBsY8VVD70VbKEm0NqYeoicA@mail.gmail.com/
Yes, it's a single test failing. After removing global.user_notification_signal
test everything else pass and you get the results printed.
>
> For example, looking at the source, I'd naively expect to see the
> user_notification_closed_listener test result -- which follows right
> after the failing test in the selftest source. But there aren't any
> results?
Yes, it hangs at this point. You have to manually terminate it.
>
> Also - could you follow up with the author of this failing test to see if
> we can get some more clarity about what might be going wrong here? It
> appears that the failing test was added in commit 6a21cc50f0c7f ("seccomp:
> add a return code to trap to userspace") by Tycho Andersen
> <tycho@tycho.ws>.
Well the code states ".. and hope that it doesn't break when there
is actually a signal :)". Maybe we are just unlucky. I don't have results
from other architectures to compare.
I found that Linaro is running selftests, but SECCOMP is disabled
and thus it's failing. Is there another CI which tracks selftests?
https://qa-reports.linaro.org/lkft/linux-next-oe/tests/kselftest/seccomp_seccomp_bpf?top=next-20190823
>
>
> - Paul
^ permalink raw reply
* Re: [net PATCH] net: route dump netlink NLM_F_MULTI flag missing
From: Stefano Brivio @ 2019-08-24 1:06 UTC (permalink / raw)
To: John Fastabend; +Cc: davem, dsahern, netdev
In-Reply-To: <156660549861.5753.7912871726096518275.stgit@john-XPS-13-9370>
On Fri, 23 Aug 2019 17:11:38 -0700
John Fastabend <john.fastabend@gmail.com> wrote:
> An excerpt from netlink(7) man page,
>
> In multipart messages (multiple nlmsghdr headers with associated payload
> in one byte stream) the first and all following headers have the
> NLM_F_MULTI flag set, except for the last header which has the type
> NLMSG_DONE.
>
> but, after (ee28906) there is a missing NLM_F_MULTI flag in the middle of a
> FIB dump.
In your case (see below), it can be zero or more, depending on how many
exception routes you have.
> The result is user space applications following above man page
> excerpt may get confused and may stop parsing msg believing something went
> wrong.
Worse yet, also RFC 3459 says:
[...] For multipart
messages, the first and all following headers have the NLM_F_MULTI
Netlink header flag set, except for the last header which has the
Netlink header type NLMSG_DONE.
But iproute2 doesn't check for this, so the selftests I added didn't
notice. Thanks for fixing this!
> In the golang netlink lib [0] the library logic stops parsing believing the
> message is not a multipart message. Found this running Cilium[1] against
> net-next while adding a feature to auto-detect routes. I noticed with
> multiple route tables we no longer could detect the default routes on net
> tree kernels because the library logic was not returning them.
However, note that if strict netlink checking is requested (I think the
library should be updated), and RTM_F_CLONED is not set (which should
be the case if you are just looking for "regular" routes), you won't
hit this.
> Fix this by handling the fib_dump_info_fnhe() case the same way the
> fib_dump_info() handles it by passing the flags argument through the
> call chain and adding a flags argument to rt_fill_info().
>
> Tested with Cilium stack and auto-detection of routes works again. Also
> annotated libs to dump netlink msgs and inspected NLM_F_MULTI and
> NLMSG_DONE flags look correct after this.
>
> Note: In inet_rtm_getroute() pass rt_fill_info() '0' for flags the same
> as is done for fib_dump_info() so this looks correct to me.
Yes, that's correct, because if the buffer is too small for a single
route dumped by a single rt_fill_info() call, we'll just fail, so that
will never be a multipart message.
> [0] https://github.com/vishvananda/netlink/
> [1] https://github.com/cilium/
>
> Fixes: ee28906fd7a14 ("ipv4: Dump route exceptions if requested")
> Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
--
Stefano
^ permalink raw reply
* Re: [PATCH v2] riscv: add support for SECCOMP and SECCOMP_FILTER
From: David Abdurachmanov @ 2019-08-24 1:10 UTC (permalink / raw)
To: Paul Walmsley
Cc: Tycho Andersen, Palmer Dabbelt, Albert Ou, Oleg Nesterov,
Kees Cook, Andy Lutomirski, Will Drewry, Shuah Khan,
Alexei Starovoitov, Daniel Borkmann, Martin KaFai Lau, Song Liu,
Yonghong Song, David Abdurachmanov, Thomas Gleixner,
Allison Randal, Alexios Zavras, Anup Patel, Vincent Chen,
Alan Kao, linux-riscv, linux-kernel, linux-kselftest, netdev, bpf,
me
In-Reply-To: <CAEn-LTp=ss0Dfv6J00=rCAy+N78U2AmhqJNjfqjr2FDpPYjxEQ@mail.gmail.com>
On Fri, Aug 23, 2019 at 6:04 PM David Abdurachmanov
<david.abdurachmanov@gmail.com> wrote:
>
> On Fri, Aug 23, 2019 at 5:30 PM Paul Walmsley <paul.walmsley@sifive.com> wrote:
> >
> > On Thu, 22 Aug 2019, David Abdurachmanov wrote:
> >
> > > There is one failing kernel selftest: global.user_notification_signal
> >
> > Is this the only failing test? Or are the rest of the selftests skipped
> > when this test fails, and no further tests are run, as seems to be shown
> > here:
> >
> > https://lore.kernel.org/linux-riscv/CADnnUqcmDMRe1f+3jG8SPR6jRrnBsY8VVD70VbKEm0NqYeoicA@mail.gmail.com/
>
> Yes, it's a single test failing. After removing global.user_notification_signal
> test everything else pass and you get the results printed.
>
> >
> > For example, looking at the source, I'd naively expect to see the
> > user_notification_closed_listener test result -- which follows right
> > after the failing test in the selftest source. But there aren't any
> > results?
>
> Yes, it hangs at this point. You have to manually terminate it.
>
> >
> > Also - could you follow up with the author of this failing test to see if
> > we can get some more clarity about what might be going wrong here? It
> > appears that the failing test was added in commit 6a21cc50f0c7f ("seccomp:
> > add a return code to trap to userspace") by Tycho Andersen
> > <tycho@tycho.ws>.
>
> Well the code states ".. and hope that it doesn't break when there
> is actually a signal :)". Maybe we are just unlucky. I don't have results
> from other architectures to compare.
>
> I found that Linaro is running selftests, but SECCOMP is disabled
> and thus it's failing. Is there another CI which tracks selftests?
>
> https://qa-reports.linaro.org/lkft/linux-next-oe/tests/kselftest/seccomp_seccomp_bpf?top=next-20190823
Actually it seems that seccomp is enabled in kernel, but not in
systemd, and somehow seccomp_bpf is missing on all arches thus
causing automatic failure.
> >
> >
> > - Paul
^ permalink raw reply
* Re: [PATCH v2] riscv: add support for SECCOMP and SECCOMP_FILTER
From: Paul Walmsley @ 2019-08-24 1:18 UTC (permalink / raw)
To: David Abdurachmanov
Cc: Tycho Andersen, Palmer Dabbelt, Albert Ou, Oleg Nesterov,
Kees Cook, Andy Lutomirski, Will Drewry, Shuah Khan,
Alexei Starovoitov, Daniel Borkmann, Martin KaFai Lau, Song Liu,
Yonghong Song, David Abdurachmanov, Thomas Gleixner,
Allison Randal, Alexios Zavras, Anup Patel, Vincent Chen,
Alan Kao, linux-riscv, linux-kernel, linux-kselftest, netdev, bpf,
me
In-Reply-To: <CAEn-LTp=ss0Dfv6J00=rCAy+N78U2AmhqJNjfqjr2FDpPYjxEQ@mail.gmail.com>
On Fri, 23 Aug 2019, David Abdurachmanov wrote:
> On Fri, Aug 23, 2019 at 5:30 PM Paul Walmsley <paul.walmsley@sifive.com> wrote:
> >
> > On Thu, 22 Aug 2019, David Abdurachmanov wrote:
> >
> > > There is one failing kernel selftest: global.user_notification_signal
> >
> > Is this the only failing test? Or are the rest of the selftests skipped
> > when this test fails, and no further tests are run, as seems to be shown
> > here:
> >
> > https://lore.kernel.org/linux-riscv/CADnnUqcmDMRe1f+3jG8SPR6jRrnBsY8VVD70VbKEm0NqYeoicA@mail.gmail.com/
>
> Yes, it's a single test failing. After removing global.user_notification_signal
> test everything else pass and you get the results printed.
OK.
> Well the code states ".. and hope that it doesn't break when there
> is actually a signal :)". Maybe we are just unlucky. I don't have results
> from other architectures to compare.
>
> I found that Linaro is running selftests, but SECCOMP is disabled
> and thus it's failing. Is there another CI which tracks selftests?
0day runs the kselftests, and at least on some architectures/Kconfigs,
it's succeeding:
https://lore.kernel.org/lkml/20190726083740.GG22106@shao2-debian/
https://lore.kernel.org/lkml/20190712064850.GC20848@shao2-debian/
https://lore.kernel.org/lkml/20190311074115.GC10839@shao2-debian/
etc.
- Paul
^ permalink raw reply
* Re: [PATCHv2 1/1] net: rds: add service level support in rds-info
From: santosh.shilimkar @ 2019-08-24 1:25 UTC (permalink / raw)
To: Zhu Yanjun, davem, netdev, linux-rdma, rds-devel, gerd.rausch
In-Reply-To: <1566608656-30836-1-git-send-email-yanjun.zhu@oracle.com>
On 8/23/19 6:04 PM, Zhu Yanjun wrote:
> From IB specific 7.6.5 SERVICE LEVEL, Service Level (SL)
> is used to identify different flows within an IBA subnet.
> It is carried in the local route header of the packet.
>
> Before this commit, run "rds-info -I". The outputs are as
> below:
> "
> RDS IB Connections:
> LocalAddr RemoteAddr Tos SL LocalDev RemoteDev
> 192.2.95.3 192.2.95.1 2 0 fe80::21:28:1a:39 fe80::21:28:10:b9
> 192.2.95.3 192.2.95.1 1 0 fe80::21:28:1a:39 fe80::21:28:10:b9
> 192.2.95.3 192.2.95.1 0 0 fe80::21:28:1a:39 fe80::21:28:10:b9
> "
> After this commit, the output is as below:
> "
> RDS IB Connections:
> LocalAddr RemoteAddr Tos SL LocalDev RemoteDev
> 192.2.95.3 192.2.95.1 2 2 fe80::21:28:1a:39 fe80::21:28:10:b9
> 192.2.95.3 192.2.95.1 1 1 fe80::21:28:1a:39 fe80::21:28:10:b9
> 192.2.95.3 192.2.95.1 0 0 fe80::21:28:1a:39 fe80::21:28:10:b9
> "
>
> The commit fe3475af3bdf ("net: rds: add per rds connection cache
> statistics") adds cache_allocs in struct rds_info_rdma_connection
> as below:
> struct rds_info_rdma_connection {
> ...
> __u32 rdma_mr_max;
> __u32 rdma_mr_size;
> __u8 tos;
> __u32 cache_allocs;
> };
> The peer struct in rds-tools of struct rds_info_rdma_connection is as
> below:
> struct rds_info_rdma_connection {
> ...
> uint32_t rdma_mr_max;
> uint32_t rdma_mr_size;
> uint8_t tos;
> uint8_t sl;
> uint32_t cache_allocs;
> };
> The difference between userspace and kernel is the member variable sl.
> In the kernel struct, the member variable sl is missing. This will
> introduce risks. So it is necessary to use this commit to avoid this risk.
>
> Fixes: fe3475af3bdf ("net: rds: add per rds connection cache statistics")
> CC: Joe Jin <joe.jin@oracle.com>
> CC: JUNXIAO_BI <junxiao.bi@oracle.com>
> Suggested-by: Gerd Rausch <gerd.rausch@oracle.com>
> Signed-off-by: Zhu Yanjun <yanjun.zhu@oracle.com>
> ---
> V1->V2: fix typos in commit logs.
> ---
I did ask you when ypu posted the patch about whether you did
backward compatibility tests for which you said, you did all the
tests and said "So do not worry about backward compatibility. This
commit will work well with older rds-tools2.0.5 and 2.0.6."
https://www.spinics.net/lists/netdev/msg574691.html
I was worried about exactly such issue as described in commit.
Anyways thanks for the fixup patch. Should be applied to stable
as well.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Regards,
Santosh
^ 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