* Re: [PATCH 5/6] drivers: net: ethernet: cpts: implement cpts hardware clock
From: Richard Cochran @ 2012-10-20 13:43 UTC (permalink / raw)
To: Mugunthan V N; +Cc: netdev, davem
In-Reply-To: <1350427518-7230-6-git-send-email-mugunthanvnm@ti.com>
I did a careful read of this driver, and I found a number of issues,
but I am not giving a line by line review, for reasons explained
below. For the most part, your driver is very similar to mine, and
your driver would probably work fine too, when given a bit of review
and revision. However, in the end, it would end up looking nearly
identical to mine.
The issues about which you have voiced concern (counter roll over,
changing input clock frequency) are by no means unique to the CPTS.
The existing PHC drivers all handle these issues in a way consistent
with each other. New drivers should not invent their own solutions to
such problems.
If I had never written a CPTS driver and had seen your present
submission first, then I would be willing to help you get your driver
merged. But in my view, this work is already done, and so I am favoring
my own driver at this point.
Here is my suggestion for moving forward with support the CPTS:
- It is too early to support the CPTS v1 found on the DM81xx, since
the CPSW support (including DT) is totally lacking. Judging from
your present driver and the TRM, the differences are minimal and
will be easy to add later on. For now, we can simply check the
version register for v2.
- If you can get the patches needed to get the CPSW working merged
during the present v3.7-rc interval, that would form a basis for
getting the CPTS into v3.8. I tried to post the patches that I had
collected, but I failed to get them all accepted. I really don't
know the issues surrounding those patches at all, and so I would
appreciate if you would take the lead and get this support merged.
- I will add a version check for CPTS v2 and integrate your CPSW
multicast patch (and anything else you think is missing) into my own
patch series. Once your background patches have appeared, I will
post the CPTS series for net-next, in time for v3.8.
- Looking ahead, once you get a patch series to fully support the
DM81xx CPSW, we can patch the CPTS too.
Thanks,
Richard
^ permalink raw reply
* Re: PROBLEM: Memory leak (at least with SLUB) from "secpath_dup" (xfrm) in 3.5+ kernels
From: Mike Kazantsev @ 2012-10-20 14:49 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, linux-mm
In-Reply-To: <CAHC9VhQ+gkAaRmwDWqzQd1U-hwH__5yxrxWa5_=koz_XTSXpjQ@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 1911 bytes --]
On Sat, 20 Oct 2012 08:42:33 -0400
Paul Moore <paul@paul-moore.com> wrote:
> Thanks for the problem report. I'm not going to be in a position to start
> looking into this until late Sunday, but hopefully it will be a quick fix.
>
> Two quick questions (my apologies, I'm not able to dig through your logs
> right now): do you see this leak on kernels < 3.5.0, and are you using any
> labeled IPsec connections?
>
As I understand, labelled connections are only used in SELinux
and SMACK LSM, which are not enabled (in Kconfig, i.e. not built) in any
of the kernels I use.
The only LSM I have enabled (and actually use on 2/4 of these machines)
is AppArmor, and though I think it doesn't attach any labels to network
connections yet (there's a "Wishlist" bug at
https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/796588, but I
can't seem to find an existing implementation).
I believe it has started with 3.5.0, according to all available logs I
have. I'm afraid laziness and other tasks have prevented me from
looking into and reporting the issue back then, but memory graph trends
start at the exact time of reboot into 3.5.0 kernels, and before that,
there're no such trends for slab memory usage.
I've been able to ignore and work around the problem for months now, so
I don't think there's any rush at all ;)
But that said, currently I've started git bisect process between v3.5
and v3.4 tags, so hopefully I'll get good-enough results of it before
you'll get to it (probably in a few hours to a few days).
Also, I've found that switching to "slab" allocator from "slub" doesn't
help the problem at all, so I guess something doesn't get freed in the
code indeed, though I hasn't been able to find anything relevant in the
logs for the sources where secpath_put and secpath_dup are used, and
decided to try bisect.
--
Mike Kazantsev // fraggod.net
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply
* [PATCH net-next 00/14] be2net updates
From: Padmanabh Ratnakar @ 2012-10-20 16:01 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
Please apply.
Thanks,
Padmanabh
Padmanabh Ratnakar (14):
be2net: Fix driver load failure for different FW configs in Lancer
be2net: Fix setting QoS for VF for Lancer
be2net: Fix change MAC operation for VF for Lancer
be2net: Wait till resources are available for VF in error recovery
be2net: Fix configuring VLAN for VF for Lancer
be2net: Fix error messages while driver load for VFs
be2net: Fix ethtool get_settings output for VF
be2net: Fix issues in error recovery due to wrong queue state
be2net: Fix unnecessary delay in PCI EEH
be2net: Fix VF driver load on newer Lancer FW
be2net: Enabling Wake-on-LAN is not supported in S5 state
be2net: Fix FW flashing on Skyhawk-R
be2net: Fix skyhawk VF PCI Device ID
be2net: Update driver version
drivers/net/ethernet/emulex/benet/be.h | 20 +-
drivers/net/ethernet/emulex/benet/be_cmds.c | 385 +++++++++++++++++-
drivers/net/ethernet/emulex/benet/be_cmds.h | 138 ++++++
drivers/net/ethernet/emulex/benet/be_ethtool.c | 80 +++--
drivers/net/ethernet/emulex/benet/be_hw.h | 3 +
drivers/net/ethernet/emulex/benet/be_main.c | 533 +++++++++++++++++-------
6 files changed, 971 insertions(+), 188 deletions(-)
^ permalink raw reply
* [PATCH net-next 01/14] be2net: Fix driver load failure for different FW configs in Lancer
From: Padmanabh Ratnakar @ 2012-10-20 16:01 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
Driver assumes FW resource counts and capabilities while creating queues and
using functionality like RSS. This causes driver load to fail in FW configs
where resources and capabilities are reduced. Fix this by querying FW
configuration during probe and using resources and capabilities accordingly.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be.h | 10 ++-
drivers/net/ethernet/emulex/benet/be_cmds.c | 150 ++++++++++++++++++++++-
drivers/net/ethernet/emulex/benet/be_cmds.h | 66 ++++++++++-
drivers/net/ethernet/emulex/benet/be_main.c | 177 +++++++++++++++++++-------
4 files changed, 349 insertions(+), 54 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index cf4c05b..dc6ec6b 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -378,7 +378,6 @@ struct be_adapter {
struct be_drv_stats drv_stats;
u16 vlans_added;
- u16 max_vlans; /* Number of vlans supported */
u8 vlan_tag[VLAN_N_VID];
u8 vlan_prio_bmap; /* Available Priority BitMap */
u16 recommended_prio; /* Recommended Priority */
@@ -434,10 +433,17 @@ struct be_adapter {
struct phy_info phy;
u8 wol_cap;
bool wol;
- u32 max_pmac_cnt; /* Max secondary UC MACs programmable */
u32 uc_macs; /* Count of secondary UC MAC programmed */
u32 msg_enable;
int be_get_temp_freq;
+ u16 max_mcast_mac;
+ u16 max_tx_queues;
+ u16 max_rss_queues;
+ u16 max_rx_queues;
+ u16 max_pmac_cnt;
+ u16 max_vlans;
+ u16 max_event_queues;
+ u32 if_cap_flags;
};
#define be_physfn(adapter) (!adapter->virtfn)
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index af60bb2..094b5f3 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -1658,9 +1658,9 @@ int be_cmd_rx_filter(struct be_adapter *adapter, u32 flags, u32 value)
/* Reset mcast promisc mode if already set by setting mask
* and not setting flags field
*/
- if (!lancer_chip(adapter) || be_physfn(adapter))
- req->if_flags_mask |=
- cpu_to_le32(BE_IF_FLAGS_MCAST_PROMISCUOUS);
+ req->if_flags_mask |=
+ cpu_to_le32(BE_IF_FLAGS_MCAST_PROMISCUOUS &
+ adapter->if_cap_flags);
req->mcast_num = cpu_to_le32(netdev_mc_count(adapter->netdev));
netdev_for_each_mc_addr(ha, adapter->netdev)
@@ -2792,6 +2792,150 @@ err:
return status;
}
+static struct be_nic_resource_desc *be_get_nic_desc(u8 *buf, u32 desc_count,
+ u32 max_buf_size)
+{
+ struct be_nic_resource_desc *desc = (struct be_nic_resource_desc *)buf;
+ int i;
+
+ for (i = 0; i < desc_count; i++) {
+ desc->desc_len = RESOURCE_DESC_SIZE;
+ if (((void *)desc + desc->desc_len) >
+ (void *)(buf + max_buf_size)) {
+ desc = NULL;
+ break;
+ }
+
+ if (desc->desc_type == NIC_RESOURCE_DESC_TYPE_ID)
+ break;
+
+ desc = (void *)desc + desc->desc_len;
+ }
+
+ if (!desc || i == MAX_RESOURCE_DESC)
+ return NULL;
+
+ return desc;
+}
+
+/* Uses Mbox */
+int be_cmd_get_func_config(struct be_adapter *adapter)
+{
+ struct be_mcc_wrb *wrb;
+ struct be_cmd_req_get_func_config *req;
+ int status;
+ struct be_dma_mem cmd;
+
+ memset(&cmd, 0, sizeof(struct be_dma_mem));
+ cmd.size = sizeof(struct be_cmd_resp_get_func_config);
+ cmd.va = pci_alloc_consistent(adapter->pdev, cmd.size,
+ &cmd.dma);
+ if (!cmd.va) {
+ dev_err(&adapter->pdev->dev, "Memory alloc failure\n");
+ return -ENOMEM;
+ }
+ if (mutex_lock_interruptible(&adapter->mbox_lock))
+ return -1;
+
+ wrb = wrb_from_mbox(adapter);
+ if (!wrb) {
+ status = -EBUSY;
+ goto err;
+ }
+
+ req = cmd.va;
+
+ be_wrb_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON,
+ OPCODE_COMMON_GET_FUNC_CONFIG,
+ cmd.size, wrb, &cmd);
+
+ status = be_mbox_notify_wait(adapter);
+ if (!status) {
+ struct be_cmd_resp_get_func_config *resp = cmd.va;
+ u32 desc_count = le32_to_cpu(resp->desc_count);
+ struct be_nic_resource_desc *desc;
+
+ desc = be_get_nic_desc(resp->func_param, desc_count,
+ sizeof(resp->func_param));
+ if (!desc) {
+ status = -EINVAL;
+ goto err;
+ }
+
+ adapter->max_pmac_cnt = le16_to_cpu(desc->unicast_mac_count);
+ adapter->max_vlans = le16_to_cpu(desc->vlan_count);
+ adapter->max_mcast_mac = le16_to_cpu(desc->mcast_mac_count);
+ adapter->max_tx_queues = le16_to_cpu(desc->txq_count);
+ adapter->max_rss_queues = le16_to_cpu(desc->rssq_count);
+ adapter->max_rx_queues = le16_to_cpu(desc->rq_count);
+
+ adapter->max_event_queues = le16_to_cpu(desc->eq_count);
+ adapter->if_cap_flags = le32_to_cpu(desc->cap_flags);
+ }
+err:
+ mutex_unlock(&adapter->mbox_lock);
+ pci_free_consistent(adapter->pdev, cmd.size,
+ cmd.va, cmd.dma);
+ return status;
+}
+
+ /* Uses sync mcc */
+int be_cmd_get_profile_config(struct be_adapter *adapter, u32 *cap_flags,
+ u8 domain)
+{
+ struct be_mcc_wrb *wrb;
+ struct be_cmd_req_get_profile_config *req;
+ int status;
+ struct be_dma_mem cmd;
+
+ memset(&cmd, 0, sizeof(struct be_dma_mem));
+ cmd.size = sizeof(struct be_cmd_resp_get_profile_config);
+ cmd.va = pci_alloc_consistent(adapter->pdev, cmd.size,
+ &cmd.dma);
+ if (!cmd.va) {
+ dev_err(&adapter->pdev->dev, "Memory alloc failure\n");
+ return -ENOMEM;
+ }
+
+ spin_lock_bh(&adapter->mcc_lock);
+
+ wrb = wrb_from_mccq(adapter);
+ if (!wrb) {
+ status = -EBUSY;
+ goto err;
+ }
+
+ req = cmd.va;
+
+ be_wrb_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON,
+ OPCODE_COMMON_GET_PROFILE_CONFIG,
+ cmd.size, wrb, &cmd);
+
+ req->type = ACTIVE_PROFILE_TYPE;
+ req->hdr.domain = domain;
+
+ status = be_mcc_notify_wait(adapter);
+ if (!status) {
+ struct be_cmd_resp_get_profile_config *resp = cmd.va;
+ u32 desc_count = le32_to_cpu(resp->desc_count);
+ struct be_nic_resource_desc *desc;
+
+ desc = be_get_nic_desc(resp->func_param, desc_count,
+ sizeof(resp->func_param));
+
+ if (!desc) {
+ status = -EINVAL;
+ goto err;
+ }
+ *cap_flags = le32_to_cpu(desc->cap_flags);
+ }
+err:
+ spin_unlock_bh(&adapter->mcc_lock);
+ pci_free_consistent(adapter->pdev, cmd.size,
+ cmd.va, cmd.dma);
+ return status;
+}
+
int be_roce_mcc_cmd(void *netdev_handle, void *wrb_payload,
int wrb_payload_size, u16 *cmd_status, u16 *ext_status)
{
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.h b/drivers/net/ethernet/emulex/benet/be_cmds.h
index 0936e21..c2bf4df 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.h
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.h
@@ -196,6 +196,8 @@ struct be_mcc_mailbox {
#define OPCODE_COMMON_GET_MAC_LIST 147
#define OPCODE_COMMON_SET_MAC_LIST 148
#define OPCODE_COMMON_GET_HSW_CONFIG 152
+#define OPCODE_COMMON_GET_FUNC_CONFIG 160
+#define OPCODE_COMMON_GET_PROFILE_CONFIG 164
#define OPCODE_COMMON_SET_HSW_CONFIG 153
#define OPCODE_COMMON_READ_OBJECT 171
#define OPCODE_COMMON_WRITE_OBJECT 172
@@ -1684,6 +1686,66 @@ struct be_cmd_req_set_ext_fat_caps {
struct be_fat_conf_params set_params;
};
+#define RESOURCE_DESC_SIZE 72
+#define NIC_RESOURCE_DESC_TYPE_ID 0x41
+#define MAX_RESOURCE_DESC 4
+struct be_nic_resource_desc {
+ u8 desc_type;
+ u8 desc_len;
+ u8 rsvd1;
+ u8 flags;
+ u8 vf_num;
+ u8 rsvd2;
+ u8 pf_num;
+ u8 rsvd3;
+ u16 unicast_mac_count;
+ u8 rsvd4[6];
+ u16 mcc_count;
+ u16 vlan_count;
+ u16 mcast_mac_count;
+ u16 txq_count;
+ u16 rq_count;
+ u16 rssq_count;
+ u16 lro_count;
+ u16 cq_count;
+ u16 toe_conn_count;
+ u16 eq_count;
+ u32 rsvd5;
+ u32 cap_flags;
+ u8 link_param;
+ u8 rsvd6[3];
+ u32 bw_min;
+ u32 bw_max;
+ u8 acpi_params;
+ u8 wol_param;
+ u16 rsvd7;
+ u32 rsvd8[3];
+};
+
+struct be_cmd_req_get_func_config {
+ struct be_cmd_req_hdr hdr;
+};
+
+struct be_cmd_resp_get_func_config {
+ struct be_cmd_req_hdr hdr;
+ u32 desc_count;
+ u8 func_param[MAX_RESOURCE_DESC * RESOURCE_DESC_SIZE];
+};
+
+#define ACTIVE_PROFILE_TYPE 0x2
+struct be_cmd_req_get_profile_config {
+ struct be_cmd_req_hdr hdr;
+ u8 rsvd;
+ u8 type;
+ u16 rsvd1;
+};
+
+struct be_cmd_resp_get_profile_config {
+ struct be_cmd_req_hdr hdr;
+ u32 desc_count;
+ u8 func_param[MAX_RESOURCE_DESC * RESOURCE_DESC_SIZE];
+};
+
extern int be_pci_fnum_get(struct be_adapter *adapter);
extern int be_fw_wait_ready(struct be_adapter *adapter);
extern int be_cmd_mac_addr_query(struct be_adapter *adapter, u8 *mac_addr,
@@ -1798,4 +1860,6 @@ extern int be_cmd_set_ext_fat_capabilites(struct be_adapter *adapter,
extern int lancer_wait_ready(struct be_adapter *adapter);
extern int lancer_test_and_set_rdy_state(struct be_adapter *adapter);
extern int be_cmd_query_port_name(struct be_adapter *adapter, u8 *port_name);
-
+extern int be_cmd_get_func_config(struct be_adapter *adapter);
+extern int be_cmd_get_profile_config(struct be_adapter *adapter, u32 *cap_flags,
+ u8 domain);
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index d1b6cc5..9b734a0 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -917,7 +917,7 @@ static void be_set_rx_mode(struct net_device *netdev)
/* Enable multicast promisc if num configured exceeds what we support */
if (netdev->flags & IFF_ALLMULTI ||
- netdev_mc_count(netdev) > BE_MAX_MC) {
+ netdev_mc_count(netdev) > adapter->max_mcast_mac) {
be_cmd_rx_filter(adapter, IFF_ALLMULTI, ON);
goto done;
}
@@ -1837,12 +1837,13 @@ static void be_tx_queues_destroy(struct be_adapter *adapter)
static int be_num_txqs_want(struct be_adapter *adapter)
{
- if (sriov_want(adapter) || be_is_mc(adapter) ||
- lancer_chip(adapter) || !be_physfn(adapter) ||
+ if ((!lancer_chip(adapter) && sriov_want(adapter)) ||
+ be_is_mc(adapter) ||
+ (!lancer_chip(adapter) && !be_physfn(adapter)) ||
adapter->generation == BE_GEN2)
return 1;
else
- return MAX_TX_QS;
+ return adapter->max_tx_queues;
}
static int be_tx_cqs_create(struct be_adapter *adapter)
@@ -2177,9 +2178,11 @@ static void be_msix_disable(struct be_adapter *adapter)
static uint be_num_rss_want(struct be_adapter *adapter)
{
u32 num = 0;
+
if ((adapter->function_caps & BE_FUNCTION_CAPS_RSS) &&
- !sriov_want(adapter) && be_physfn(adapter)) {
- num = (adapter->be3_native) ? BE3_MAX_RSS_QS : BE2_MAX_RSS_QS;
+ (lancer_chip(adapter) ||
+ (!sriov_want(adapter) && be_physfn(adapter)))) {
+ num = adapter->max_rss_queues;
num = min_t(u32, num, (u32)netif_get_num_default_rss_queues());
}
return num;
@@ -2579,10 +2582,30 @@ static int be_clear(struct be_adapter *adapter)
be_tx_queues_destroy(adapter);
be_evt_queues_destroy(adapter);
+ kfree(adapter->pmac_id);
+ adapter->pmac_id = NULL;
+
be_msix_disable(adapter);
return 0;
}
+static void be_get_vf_if_cap_flags(struct be_adapter *adapter,
+ u32 *cap_flags, u8 domain)
+{
+ bool profile_present = false;
+ int status;
+
+ if (lancer_chip(adapter)) {
+ status = be_cmd_get_profile_config(adapter, cap_flags, domain);
+ if (!status)
+ profile_present = true;
+ }
+
+ if (!profile_present)
+ *cap_flags = BE_IF_FLAGS_UNTAGGED | BE_IF_FLAGS_BROADCAST |
+ BE_IF_FLAGS_MULTICAST;
+}
+
static int be_vf_setup_init(struct be_adapter *adapter)
{
struct be_vf_cfg *vf_cfg;
@@ -2634,9 +2657,13 @@ static int be_vf_setup(struct be_adapter *adapter)
if (status)
goto err;
- cap_flags = en_flags = BE_IF_FLAGS_UNTAGGED | BE_IF_FLAGS_BROADCAST |
- BE_IF_FLAGS_MULTICAST;
for_all_vfs(adapter, vf_cfg, vf) {
+ be_get_vf_if_cap_flags(adapter, &cap_flags, vf + 1);
+
+ en_flags = cap_flags & (BE_IF_FLAGS_UNTAGGED |
+ BE_IF_FLAGS_BROADCAST |
+ BE_IF_FLAGS_MULTICAST);
+
status = be_cmd_if_create(adapter, cap_flags, en_flags,
&vf_cfg->if_handle, vf + 1);
if (status)
@@ -2712,12 +2739,93 @@ static int be_get_mac_addr(struct be_adapter *adapter, u8 *mac, u32 if_handle,
return status;
}
+static void be_get_resources(struct be_adapter *adapter)
+{
+ int status;
+ bool profile_present = false;
+
+ if (lancer_chip(adapter)) {
+ status = be_cmd_get_func_config(adapter);
+
+ if (!status)
+ profile_present = true;
+ }
+
+ if (profile_present) {
+ /* Sanity fixes for Lancer */
+ adapter->max_pmac_cnt = min_t(u16, adapter->max_pmac_cnt,
+ BE_UC_PMAC_COUNT);
+ adapter->max_vlans = min_t(u16, adapter->max_vlans,
+ BE_NUM_VLANS_SUPPORTED);
+ adapter->max_mcast_mac = min_t(u16, adapter->max_mcast_mac,
+ BE_MAX_MC);
+ adapter->max_tx_queues = min_t(u16, adapter->max_tx_queues,
+ MAX_TX_QS);
+ adapter->max_rss_queues = min_t(u16, adapter->max_rss_queues,
+ BE3_MAX_RSS_QS);
+ adapter->max_event_queues = min_t(u16,
+ adapter->max_event_queues,
+ BE3_MAX_RSS_QS);
+
+ if (adapter->max_rss_queues &&
+ adapter->max_rss_queues == adapter->max_rx_queues)
+ adapter->max_rss_queues -= 1;
+
+ if (adapter->max_event_queues < adapter->max_rss_queues)
+ adapter->max_rss_queues = adapter->max_event_queues;
+
+ } else {
+ if (be_physfn(adapter))
+ adapter->max_pmac_cnt = BE_UC_PMAC_COUNT;
+ else
+ adapter->max_pmac_cnt = BE_VF_UC_PMAC_COUNT;
+
+ if (adapter->function_mode & FLEX10_MODE)
+ adapter->max_vlans = BE_NUM_VLANS_SUPPORTED/8;
+ else
+ adapter->max_vlans = BE_NUM_VLANS_SUPPORTED;
+
+ adapter->max_mcast_mac = BE_MAX_MC;
+ adapter->max_tx_queues = MAX_TX_QS;
+ adapter->max_rss_queues = (adapter->be3_native) ?
+ BE3_MAX_RSS_QS : BE2_MAX_RSS_QS;
+ adapter->max_event_queues = BE3_MAX_RSS_QS;
+
+ adapter->if_cap_flags = BE_IF_FLAGS_UNTAGGED |
+ BE_IF_FLAGS_BROADCAST |
+ BE_IF_FLAGS_MULTICAST |
+ BE_IF_FLAGS_PASS_L3L4_ERRORS |
+ BE_IF_FLAGS_MCAST_PROMISCUOUS |
+ BE_IF_FLAGS_VLAN_PROMISCUOUS |
+ BE_IF_FLAGS_PROMISCUOUS;
+
+ if (adapter->function_caps & BE_FUNCTION_CAPS_RSS)
+ adapter->if_cap_flags |= BE_IF_FLAGS_RSS;
+ }
+}
+
/* Routine to query per function resource limits */
static int be_get_config(struct be_adapter *adapter)
{
- int pos;
+ int pos, status;
u16 dev_num_vfs;
+ status = be_cmd_query_fw_cfg(adapter, &adapter->port_num,
+ &adapter->function_mode,
+ &adapter->function_caps);
+ if (status)
+ goto err;
+
+ be_get_resources(adapter);
+
+ /* primary mac needs 1 pmac entry */
+ adapter->pmac_id = kcalloc(adapter->max_pmac_cnt + 1,
+ sizeof(u32), GFP_KERNEL);
+ if (!adapter->pmac_id) {
+ status = -ENOMEM;
+ goto err;
+ }
+
pos = pci_find_ext_capability(adapter->pdev, PCI_EXT_CAP_ID_SRIOV);
if (pos) {
pci_read_config_word(adapter->pdev, pos + PCI_SRIOV_TOTAL_VF,
@@ -2726,13 +2834,14 @@ static int be_get_config(struct be_adapter *adapter)
dev_num_vfs = min_t(u16, dev_num_vfs, MAX_VFS);
adapter->dev_num_vfs = dev_num_vfs;
}
- return 0;
+err:
+ return status;
}
static int be_setup(struct be_adapter *adapter)
{
struct device *dev = &adapter->pdev->dev;
- u32 cap_flags, en_flags;
+ u32 en_flags;
u32 tx_fc, rx_fc;
int status;
u8 mac[ETH_ALEN];
@@ -2740,9 +2849,12 @@ static int be_setup(struct be_adapter *adapter)
be_setup_init(adapter);
- be_get_config(adapter);
+ if (!lancer_chip(adapter))
+ be_cmd_req_native_mode(adapter);
- be_cmd_req_native_mode(adapter);
+ status = be_get_config(adapter);
+ if (status)
+ goto err;
be_msix_enable(adapter);
@@ -2764,22 +2876,13 @@ static int be_setup(struct be_adapter *adapter)
en_flags = BE_IF_FLAGS_UNTAGGED | BE_IF_FLAGS_BROADCAST |
BE_IF_FLAGS_MULTICAST | BE_IF_FLAGS_PASS_L3L4_ERRORS;
- cap_flags = en_flags | BE_IF_FLAGS_MCAST_PROMISCUOUS |
- BE_IF_FLAGS_VLAN_PROMISCUOUS | BE_IF_FLAGS_PROMISCUOUS;
- if (adapter->function_caps & BE_FUNCTION_CAPS_RSS) {
- cap_flags |= BE_IF_FLAGS_RSS;
+ if (adapter->function_caps & BE_FUNCTION_CAPS_RSS)
en_flags |= BE_IF_FLAGS_RSS;
- }
- if (lancer_chip(adapter) && !be_physfn(adapter)) {
- en_flags = BE_IF_FLAGS_UNTAGGED |
- BE_IF_FLAGS_BROADCAST |
- BE_IF_FLAGS_MULTICAST;
- cap_flags = en_flags;
- }
+ en_flags = en_flags & adapter->if_cap_flags;
- status = be_cmd_if_create(adapter, cap_flags, en_flags,
+ status = be_cmd_if_create(adapter, adapter->if_cap_flags, en_flags,
&adapter->if_handle, 0);
if (status != 0)
goto err;
@@ -3437,7 +3540,6 @@ static void be_ctrl_cleanup(struct be_adapter *adapter)
if (mem->va)
dma_free_coherent(&adapter->pdev->dev, mem->size, mem->va,
mem->dma);
- kfree(adapter->pmac_id);
}
static int be_ctrl_init(struct be_adapter *adapter)
@@ -3473,13 +3575,6 @@ static int be_ctrl_init(struct be_adapter *adapter)
goto free_mbox;
}
memset(rx_filter->va, 0, rx_filter->size);
-
- /* primary mac needs 1 pmac entry */
- adapter->pmac_id = kcalloc(adapter->max_pmac_cnt + 1,
- sizeof(*adapter->pmac_id), GFP_KERNEL);
- if (!adapter->pmac_id)
- return -ENOMEM;
-
mutex_init(&adapter->mbox_lock);
spin_lock_init(&adapter->mcc_lock);
spin_lock_init(&adapter->mcc_cq_lock);
@@ -3598,26 +3693,12 @@ u32 be_get_fw_log_level(struct be_adapter *adapter)
err:
return level;
}
+
static int be_get_initial_config(struct be_adapter *adapter)
{
int status;
u32 level;
- status = be_cmd_query_fw_cfg(adapter, &adapter->port_num,
- &adapter->function_mode, &adapter->function_caps);
- if (status)
- return status;
-
- if (adapter->function_mode & FLEX10_MODE)
- adapter->max_vlans = BE_NUM_VLANS_SUPPORTED/8;
- else
- adapter->max_vlans = BE_NUM_VLANS_SUPPORTED;
-
- if (be_physfn(adapter))
- adapter->max_pmac_cnt = BE_UC_PMAC_COUNT;
- else
- adapter->max_pmac_cnt = BE_VF_UC_PMAC_COUNT;
-
status = be_cmd_get_cntl_attributes(adapter);
if (status)
return status;
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 02/14] be2net: Fix setting QoS for VF for Lancer
From: Padmanabh Ratnakar @ 2012-10-20 16:01 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
Use Lancer specific command to set QoS for VF.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be.h | 1 +
drivers/net/ethernet/emulex/benet/be_cmds.c | 58 +++++++++++++++++++++++++++
drivers/net/ethernet/emulex/benet/be_cmds.h | 23 +++++++++++
drivers/net/ethernet/emulex/benet/be_main.c | 5 ++-
4 files changed, 86 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index dc6ec6b..3b8ef4b 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -444,6 +444,7 @@ struct be_adapter {
u16 max_vlans;
u16 max_event_queues;
u32 if_cap_flags;
+ u8 pf_number;
};
#define be_physfn(adapter) (!adapter->virtfn)
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index 094b5f3..3980b58 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -2862,6 +2862,7 @@ int be_cmd_get_func_config(struct be_adapter *adapter)
goto err;
}
+ adapter->pf_number = desc->pf_num;
adapter->max_pmac_cnt = le16_to_cpu(desc->unicast_mac_count);
adapter->max_vlans = le16_to_cpu(desc->vlan_count);
adapter->max_mcast_mac = le16_to_cpu(desc->mcast_mac_count);
@@ -2936,6 +2937,63 @@ err:
return status;
}
+/* Uses sync mcc */
+int be_cmd_set_profile_config(struct be_adapter *adapter, u32 bps,
+ u8 domain)
+{
+ struct be_mcc_wrb *wrb;
+ struct be_cmd_req_set_profile_config *req;
+ int status;
+
+ spin_lock_bh(&adapter->mcc_lock);
+
+ wrb = wrb_from_mccq(adapter);
+ if (!wrb) {
+ status = -EBUSY;
+ goto err;
+ }
+
+ req = embedded_payload(wrb);
+
+ be_wrb_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON,
+ OPCODE_COMMON_SET_PROFILE_CONFIG, sizeof(*req),
+ wrb, NULL);
+
+ req->hdr.domain = domain;
+ req->desc_count = cpu_to_le32(1);
+
+ req->nic_desc.desc_type = NIC_RESOURCE_DESC_TYPE_ID;
+ req->nic_desc.desc_len = RESOURCE_DESC_SIZE;
+ req->nic_desc.flags = (1 << QUN) | (1 << IMM) | (1 << NOSV);
+ req->nic_desc.pf_num = adapter->pf_number;
+ req->nic_desc.vf_num = domain;
+
+ /* Mark fields invalid */
+ req->nic_desc.unicast_mac_count = 0xFFFF;
+ req->nic_desc.mcc_count = 0xFFFF;
+ req->nic_desc.vlan_count = 0xFFFF;
+ req->nic_desc.mcast_mac_count = 0xFFFF;
+ req->nic_desc.txq_count = 0xFFFF;
+ req->nic_desc.rq_count = 0xFFFF;
+ req->nic_desc.rssq_count = 0xFFFF;
+ req->nic_desc.lro_count = 0xFFFF;
+ req->nic_desc.cq_count = 0xFFFF;
+ req->nic_desc.toe_conn_count = 0xFFFF;
+ req->nic_desc.eq_count = 0xFFFF;
+ req->nic_desc.link_param = 0xFF;
+ req->nic_desc.bw_min = 0xFFFFFFFF;
+ req->nic_desc.acpi_params = 0xFF;
+ req->nic_desc.wol_param = 0x0F;
+
+ /* Change BW */
+ req->nic_desc.bw_min = cpu_to_le32(bps);
+ req->nic_desc.bw_max = cpu_to_le32(bps);
+ status = be_mcc_notify_wait(adapter);
+err:
+ spin_unlock_bh(&adapter->mcc_lock);
+ return status;
+}
+
int be_roce_mcc_cmd(void *netdev_handle, void *wrb_payload,
int wrb_payload_size, u16 *cmd_status, u16 *ext_status)
{
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.h b/drivers/net/ethernet/emulex/benet/be_cmds.h
index c2bf4df..ad16d31 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.h
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.h
@@ -198,6 +198,7 @@ struct be_mcc_mailbox {
#define OPCODE_COMMON_GET_HSW_CONFIG 152
#define OPCODE_COMMON_GET_FUNC_CONFIG 160
#define OPCODE_COMMON_GET_PROFILE_CONFIG 164
+#define OPCODE_COMMON_SET_PROFILE_CONFIG 165
#define OPCODE_COMMON_SET_HSW_CONFIG 153
#define OPCODE_COMMON_READ_OBJECT 171
#define OPCODE_COMMON_WRITE_OBJECT 172
@@ -1689,6 +1690,14 @@ struct be_cmd_req_set_ext_fat_caps {
#define RESOURCE_DESC_SIZE 72
#define NIC_RESOURCE_DESC_TYPE_ID 0x41
#define MAX_RESOURCE_DESC 4
+
+/* QOS unit number */
+#define QUN 4
+/* Immediate */
+#define IMM 6
+/* No save */
+#define NOSV 7
+
struct be_nic_resource_desc {
u8 desc_type;
u8 desc_len;
@@ -1746,6 +1755,17 @@ struct be_cmd_resp_get_profile_config {
u8 func_param[MAX_RESOURCE_DESC * RESOURCE_DESC_SIZE];
};
+struct be_cmd_req_set_profile_config {
+ struct be_cmd_req_hdr hdr;
+ u32 rsvd;
+ u32 desc_count;
+ struct be_nic_resource_desc nic_desc;
+};
+
+struct be_cmd_resp_set_profile_config {
+ struct be_cmd_req_hdr hdr;
+};
+
extern int be_pci_fnum_get(struct be_adapter *adapter);
extern int be_fw_wait_ready(struct be_adapter *adapter);
extern int be_cmd_mac_addr_query(struct be_adapter *adapter, u8 *mac_addr,
@@ -1863,3 +1883,6 @@ extern int be_cmd_query_port_name(struct be_adapter *adapter, u8 *port_name);
extern int be_cmd_get_func_config(struct be_adapter *adapter);
extern int be_cmd_get_profile_config(struct be_adapter *adapter, u32 *cap_flags,
u8 domain);
+
+extern int be_cmd_set_profile_config(struct be_adapter *adapter, u32 bps,
+ u8 domain);
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 9b734a0..0b49201 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -1062,7 +1062,10 @@ static int be_set_vf_tx_rate(struct net_device *netdev,
return -EINVAL;
}
- status = be_cmd_set_qos(adapter, rate / 10, vf + 1);
+ if (lancer_chip(adapter))
+ status = be_cmd_set_profile_config(adapter, rate / 10, vf + 1);
+ else
+ status = be_cmd_set_qos(adapter, rate / 10, vf + 1);
if (status)
dev_err(&adapter->pdev->dev,
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 03/14] be2net: Fix change MAC operation for VF for Lancer
From: Padmanabh Ratnakar @ 2012-10-20 16:02 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
For changing MAC of VF from PF, delete MAC operation needs to be done before
assigning new MAC. Also in ndo_set_mac_address operation avoid delete MAC if
it has been already deleted by PF.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_main.c | 52 +++++++++++++++++++++-----
1 files changed, 42 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 0b49201..41f935a 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -237,23 +237,46 @@ static int be_mac_addr_set(struct net_device *netdev, void *p)
int status = 0;
u8 current_mac[ETH_ALEN];
u32 pmac_id = adapter->pmac_id[0];
+ bool active_mac = true;
if (!is_valid_ether_addr(addr->sa_data))
return -EADDRNOTAVAIL;
- status = be_cmd_mac_addr_query(adapter, current_mac, false,
- adapter->if_handle, 0);
+ /* For BE VF, MAC address is already activated by PF.
+ * Hence only operation left is updating netdev->devaddr.
+ * Update it if user is passing the same MAC which was used
+ * during configuring VF MAC from PF(Hypervisor).
+ */
+ if (!lancer_chip(adapter) && !be_physfn(adapter)) {
+ status = be_cmd_mac_addr_query(adapter, current_mac,
+ false, adapter->if_handle, 0);
+ if (!status && !memcmp(current_mac, addr->sa_data, ETH_ALEN))
+ goto done;
+ else
+ goto err;
+ }
+
+ if (!memcmp(addr->sa_data, netdev->dev_addr, ETH_ALEN))
+ goto done;
+
+ /* For Lancer check if any MAC is active.
+ * If active, get its mac id.
+ */
+ if (lancer_chip(adapter) && !be_physfn(adapter))
+ be_cmd_get_mac_from_list(adapter, current_mac, &active_mac,
+ &pmac_id, 0);
+
+ status = be_cmd_pmac_add(adapter, (u8 *)addr->sa_data,
+ adapter->if_handle,
+ &adapter->pmac_id[0], 0);
+
if (status)
goto err;
- if (memcmp(addr->sa_data, current_mac, ETH_ALEN)) {
- status = be_cmd_pmac_add(adapter, (u8 *)addr->sa_data,
- adapter->if_handle, &adapter->pmac_id[0], 0);
- if (status)
- goto err;
-
- be_cmd_pmac_del(adapter, adapter->if_handle, pmac_id, 0);
- }
+ if (active_mac)
+ be_cmd_pmac_del(adapter, adapter->if_handle,
+ pmac_id, 0);
+done:
memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
return 0;
err:
@@ -962,6 +985,9 @@ static int be_set_vf_mac(struct net_device *netdev, int vf, u8 *mac)
struct be_adapter *adapter = netdev_priv(netdev);
struct be_vf_cfg *vf_cfg = &adapter->vf_cfg[vf];
int status;
+ bool active_mac = false;
+ u32 pmac_id;
+ u8 old_mac[ETH_ALEN];
if (!sriov_enabled(adapter))
return -EPERM;
@@ -970,6 +996,12 @@ static int be_set_vf_mac(struct net_device *netdev, int vf, u8 *mac)
return -EINVAL;
if (lancer_chip(adapter)) {
+ status = be_cmd_get_mac_from_list(adapter, old_mac, &active_mac,
+ &pmac_id, vf + 1);
+ if (!status && active_mac)
+ be_cmd_pmac_del(adapter, vf_cfg->if_handle,
+ pmac_id, vf + 1);
+
status = be_cmd_set_mac_list(adapter, mac, 1, vf + 1);
} else {
status = be_cmd_pmac_del(adapter, vf_cfg->if_handle,
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 04/14] be2net: Wait till resources are available for VF in error recovery
From: Padmanabh Ratnakar @ 2012-10-20 16:02 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
After FW error, driver should wait for NO_RESOURCE error to disappear before
proceeding with recovery.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_cmds.c | 31 +++++++++++++++++++++++++++
drivers/net/ethernet/emulex/benet/be_hw.h | 3 ++
drivers/net/ethernet/emulex/benet/be_main.c | 5 ++-
3 files changed, 37 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index 3980b58..9a12c4b 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -452,10 +452,33 @@ int lancer_wait_ready(struct be_adapter *adapter)
return status;
}
+static bool lancer_provisioning_error(struct be_adapter *adapter)
+{
+ u32 sliport_status = 0, sliport_err1 = 0, sliport_err2 = 0;
+ sliport_status = ioread32(adapter->db + SLIPORT_STATUS_OFFSET);
+ if (sliport_status & SLIPORT_STATUS_ERR_MASK) {
+ sliport_err1 = ioread32(adapter->db +
+ SLIPORT_ERROR1_OFFSET);
+ sliport_err2 = ioread32(adapter->db +
+ SLIPORT_ERROR2_OFFSET);
+
+ if (sliport_err1 == SLIPORT_ERROR_NO_RESOURCE1 &&
+ sliport_err2 == SLIPORT_ERROR_NO_RESOURCE2)
+ return true;
+ }
+ return false;
+}
+
int lancer_test_and_set_rdy_state(struct be_adapter *adapter)
{
int status;
u32 sliport_status, err, reset_needed;
+ bool resource_error;
+
+ resource_error = lancer_provisioning_error(adapter);
+ if (resource_error)
+ return -1;
+
status = lancer_wait_ready(adapter);
if (!status) {
sliport_status = ioread32(adapter->db + SLIPORT_STATUS_OFFSET);
@@ -477,6 +500,14 @@ int lancer_test_and_set_rdy_state(struct be_adapter *adapter)
status = -1;
}
}
+ /* Stop error recovery if error is not recoverable.
+ * No resource error is temporary errors and will go away
+ * when PF provisions resources.
+ */
+ resource_error = lancer_provisioning_error(adapter);
+ if (status == -1 && !resource_error)
+ adapter->eeh_error = true;
+
return status;
}
diff --git a/drivers/net/ethernet/emulex/benet/be_hw.h b/drivers/net/ethernet/emulex/benet/be_hw.h
index b755f70..4ccbc36 100644
--- a/drivers/net/ethernet/emulex/benet/be_hw.h
+++ b/drivers/net/ethernet/emulex/benet/be_hw.h
@@ -59,6 +59,9 @@
#define PHYSDEV_CONTROL_FW_RESET_MASK 0x00000002
#define PHYSDEV_CONTROL_INP_MASK 0x40000000
+#define SLIPORT_ERROR_NO_RESOURCE1 0x2
+#define SLIPORT_ERROR_NO_RESOURCE2 0x9
+
/********* Memory BAR register ************/
#define PCICFG_MEMBAR_CTRL_INT_CTRL_OFFSET 0xfc
/* Host Interrupt Enable, if set interrupts are enabled although "PCI Interrupt
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 41f935a..7a483fd 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -3837,8 +3837,9 @@ static int lancer_recover_func(struct be_adapter *adapter)
"Adapter SLIPORT recovery succeeded\n");
return 0;
err:
- dev_err(&adapter->pdev->dev,
- "Adapter SLIPORT recovery failed\n");
+ if (adapter->eeh_error)
+ dev_err(&adapter->pdev->dev,
+ "Adapter SLIPORT recovery failed\n");
return status;
}
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 05/14] be2net: Fix configuring VLAN for VF for Lancer
From: Padmanabh Ratnakar @ 2012-10-20 16:02 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
Allow adding VLANs for Lancer VF.
VLAN ID 0 should not be added to list of VLANs sent to FW.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_main.c | 12 ++++++++++--
1 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 7a483fd..fa895bc 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -879,11 +879,15 @@ static int be_vlan_add_vid(struct net_device *netdev, u16 vid)
struct be_adapter *adapter = netdev_priv(netdev);
int status = 0;
- if (!be_physfn(adapter)) {
+ if (!lancer_chip(adapter) && !be_physfn(adapter)) {
status = -EINVAL;
goto ret;
}
+ /* Packets with VID 0 are always received by Lancer by default */
+ if (lancer_chip(adapter) && vid == 0)
+ goto ret;
+
adapter->vlan_tag[vid] = 1;
if (adapter->vlans_added <= (adapter->max_vlans + 1))
status = be_vid_config(adapter);
@@ -901,11 +905,15 @@ static int be_vlan_rem_vid(struct net_device *netdev, u16 vid)
struct be_adapter *adapter = netdev_priv(netdev);
int status = 0;
- if (!be_physfn(adapter)) {
+ if (!lancer_chip(adapter) && !be_physfn(adapter)) {
status = -EINVAL;
goto ret;
}
+ /* Packets with VID 0 are always received by Lancer by default */
+ if (lancer_chip(adapter) && vid == 0)
+ goto ret;
+
adapter->vlan_tag[vid] = 0;
if (adapter->vlans_added <= adapter->max_vlans)
status = be_vid_config(adapter);
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 06/14] be2net: Fix error messages while driver load for VFs
From: Padmanabh Ratnakar @ 2012-10-20 16:02 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
VF does not have privileges to execute many commands. When VFs try
to execute those commands there are unnecessary error messages.
Fix this by executing only those commands for which VF has privilege.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be.h | 1 +
drivers/net/ethernet/emulex/benet/be_cmds.c | 105 ++++++++++++++++++++++++
drivers/net/ethernet/emulex/benet/be_cmds.h | 43 ++++++++++
drivers/net/ethernet/emulex/benet/be_ethtool.c | 7 ++
drivers/net/ethernet/emulex/benet/be_main.c | 19 ++++-
5 files changed, 173 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index 3b8ef4b..1a05b79 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -390,6 +390,7 @@ struct be_adapter {
struct delayed_work func_recovery_work;
u32 flags;
+ u32 cmd_privileges;
/* Ethtool knobs and info */
char fw_ver[FW_VER_LEN];
int if_handle; /* Used to configure filtering */
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index 9a12c4b..9d602f4 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -19,6 +19,55 @@
#include "be.h"
#include "be_cmds.h"
+static struct be_cmd_priv_map cmd_priv_map[] = {
+ {
+ OPCODE_ETH_ACPI_WOL_MAGIC_CONFIG,
+ CMD_SUBSYSTEM_ETH,
+ BE_PRIV_LNKMGMT | BE_PRIV_VHADM |
+ BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
+ },
+ {
+ OPCODE_COMMON_GET_FLOW_CONTROL,
+ CMD_SUBSYSTEM_COMMON,
+ BE_PRIV_LNKQUERY | BE_PRIV_VHADM |
+ BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
+ },
+ {
+ OPCODE_COMMON_SET_FLOW_CONTROL,
+ CMD_SUBSYSTEM_COMMON,
+ BE_PRIV_LNKMGMT | BE_PRIV_VHADM |
+ BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
+ },
+ {
+ OPCODE_ETH_GET_PPORT_STATS,
+ CMD_SUBSYSTEM_ETH,
+ BE_PRIV_LNKMGMT | BE_PRIV_VHADM |
+ BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
+ },
+ {
+ OPCODE_COMMON_GET_PHY_DETAILS,
+ CMD_SUBSYSTEM_COMMON,
+ BE_PRIV_LNKMGMT | BE_PRIV_VHADM |
+ BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
+ }
+};
+
+static bool be_cmd_allowed(struct be_adapter *adapter, u8 opcode,
+ u8 subsystem)
+{
+ int i;
+ int num_entries = sizeof(cmd_priv_map)/sizeof(struct be_cmd_priv_map);
+ u32 cmd_privileges = adapter->cmd_privileges;
+
+ for (i = 0; i < num_entries; i++)
+ if (opcode == cmd_priv_map[i].opcode &&
+ subsystem == cmd_priv_map[i].subsystem)
+ if (!(cmd_privileges & cmd_priv_map[i].priv_mask))
+ return false;
+
+ return true;
+}
+
static inline void *embedded_payload(struct be_mcc_wrb *wrb)
{
return wrb->payload.embedded_payload;
@@ -1332,6 +1381,10 @@ int lancer_cmd_get_pport_stats(struct be_adapter *adapter,
struct lancer_cmd_req_pport_stats *req;
int status = 0;
+ if (!be_cmd_allowed(adapter, OPCODE_ETH_GET_PPORT_STATS,
+ CMD_SUBSYSTEM_ETH))
+ return -EPERM;
+
spin_lock_bh(&adapter->mcc_lock);
wrb = wrb_from_mccq(adapter);
@@ -1711,6 +1764,10 @@ int be_cmd_set_flow_control(struct be_adapter *adapter, u32 tx_fc, u32 rx_fc)
struct be_cmd_req_set_flow_control *req;
int status;
+ if (!be_cmd_allowed(adapter, OPCODE_COMMON_SET_FLOW_CONTROL,
+ CMD_SUBSYSTEM_COMMON))
+ return -EPERM;
+
spin_lock_bh(&adapter->mcc_lock);
wrb = wrb_from_mccq(adapter);
@@ -1740,6 +1797,10 @@ int be_cmd_get_flow_control(struct be_adapter *adapter, u32 *tx_fc, u32 *rx_fc)
struct be_cmd_req_get_flow_control *req;
int status;
+ if (!be_cmd_allowed(adapter, OPCODE_COMMON_GET_FLOW_CONTROL,
+ CMD_SUBSYSTEM_COMMON))
+ return -EPERM;
+
spin_lock_bh(&adapter->mcc_lock);
wrb = wrb_from_mccq(adapter);
@@ -2306,6 +2367,10 @@ int be_cmd_get_phy_info(struct be_adapter *adapter)
struct be_dma_mem cmd;
int status;
+ if (!be_cmd_allowed(adapter, OPCODE_COMMON_GET_PHY_DETAILS,
+ CMD_SUBSYSTEM_COMMON))
+ return -EPERM;
+
spin_lock_bh(&adapter->mcc_lock);
wrb = wrb_from_mccq(adapter);
@@ -2465,6 +2530,42 @@ err:
return status;
}
+/* Get privilege(s) for a function */
+int be_cmd_get_fn_privileges(struct be_adapter *adapter, u32 *privilege,
+ u32 domain)
+{
+ struct be_mcc_wrb *wrb;
+ struct be_cmd_req_get_fn_privileges *req;
+ int status;
+
+ spin_lock_bh(&adapter->mcc_lock);
+
+ wrb = wrb_from_mccq(adapter);
+ if (!wrb) {
+ status = -EBUSY;
+ goto err;
+ }
+
+ req = embedded_payload(wrb);
+
+ be_wrb_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON,
+ OPCODE_COMMON_GET_FN_PRIVILEGES, sizeof(*req),
+ wrb, NULL);
+
+ req->hdr.domain = domain;
+
+ status = be_mcc_notify_wait(adapter);
+ if (!status) {
+ struct be_cmd_resp_get_fn_privileges *resp =
+ embedded_payload(wrb);
+ *privilege = le32_to_cpu(resp->privilege_mask);
+ }
+
+err:
+ spin_unlock_bh(&adapter->mcc_lock);
+ return status;
+}
+
/* Uses synchronous MCCQ */
int be_cmd_get_mac_from_list(struct be_adapter *adapter, u8 *mac,
bool *pmac_id_active, u32 *pmac_id, u8 domain)
@@ -2682,6 +2783,10 @@ int be_cmd_get_acpi_wol_cap(struct be_adapter *adapter)
int payload_len = sizeof(*req);
struct be_dma_mem cmd;
+ if (!be_cmd_allowed(adapter, OPCODE_ETH_ACPI_WOL_MAGIC_CONFIG,
+ CMD_SUBSYSTEM_ETH))
+ return -EPERM;
+
memset(&cmd, 0, sizeof(struct be_dma_mem));
cmd.size = sizeof(struct be_cmd_resp_acpi_wol_magic_config_v1);
cmd.va = pci_alloc_consistent(adapter->pdev, cmd.size,
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.h b/drivers/net/ethernet/emulex/benet/be_cmds.h
index ad16d31..a438777 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.h
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.h
@@ -200,6 +200,7 @@ struct be_mcc_mailbox {
#define OPCODE_COMMON_GET_PROFILE_CONFIG 164
#define OPCODE_COMMON_SET_PROFILE_CONFIG 165
#define OPCODE_COMMON_SET_HSW_CONFIG 153
+#define OPCODE_COMMON_GET_FN_PRIVILEGES 170
#define OPCODE_COMMON_READ_OBJECT 171
#define OPCODE_COMMON_WRITE_OBJECT 172
@@ -1432,6 +1433,41 @@ struct be_cmd_resp_set_func_cap {
u8 rsvd[212];
};
+/*********************** Function Privileges ***********************/
+enum {
+ BE_PRIV_DEFAULT = 0x1,
+ BE_PRIV_LNKQUERY = 0x2,
+ BE_PRIV_LNKSTATS = 0x4,
+ BE_PRIV_LNKMGMT = 0x8,
+ BE_PRIV_LNKDIAG = 0x10,
+ BE_PRIV_UTILQUERY = 0x20,
+ BE_PRIV_FILTMGMT = 0x40,
+ BE_PRIV_IFACEMGMT = 0x80,
+ BE_PRIV_VHADM = 0x100,
+ BE_PRIV_DEVCFG = 0x200,
+ BE_PRIV_DEVSEC = 0x400
+};
+#define MAX_PRIVILEGES (BE_PRIV_VHADM | BE_PRIV_DEVCFG | \
+ BE_PRIV_DEVSEC)
+#define MIN_PRIVILEGES BE_PRIV_DEFAULT
+
+struct be_cmd_priv_map {
+ u8 opcode;
+ u8 subsystem;
+ u32 priv_mask;
+};
+
+struct be_cmd_req_get_fn_privileges {
+ struct be_cmd_req_hdr hdr;
+ u32 rsvd;
+};
+
+struct be_cmd_resp_get_fn_privileges {
+ struct be_cmd_resp_hdr hdr;
+ u32 privilege_mask;
+};
+
+
/******************** GET/SET_MACLIST **************************/
#define BE_MAX_MAC 64
struct be_cmd_req_get_mac_list {
@@ -1766,6 +1802,11 @@ struct be_cmd_resp_set_profile_config {
struct be_cmd_req_hdr hdr;
};
+static inline bool check_privilege(struct be_adapter *adapter, u32 flags)
+{
+ return flags & adapter->cmd_privileges ? true : false;
+}
+
extern int be_pci_fnum_get(struct be_adapter *adapter);
extern int be_fw_wait_ready(struct be_adapter *adapter);
extern int be_cmd_mac_addr_query(struct be_adapter *adapter, u8 *mac_addr,
@@ -1862,6 +1903,8 @@ extern int be_cmd_get_cntl_attributes(struct be_adapter *adapter);
extern int be_cmd_req_native_mode(struct be_adapter *adapter);
extern int be_cmd_get_reg_len(struct be_adapter *adapter, u32 *log_size);
extern void be_cmd_get_regs(struct be_adapter *adapter, u32 buf_len, void *buf);
+extern int be_cmd_get_fn_privileges(struct be_adapter *adapter,
+ u32 *privilege, u32 domain);
extern int be_cmd_get_mac_from_list(struct be_adapter *adapter, u8 *mac,
bool *pmac_id_active, u32 *pmac_id,
u8 domain);
diff --git a/drivers/net/ethernet/emulex/benet/be_ethtool.c b/drivers/net/ethernet/emulex/benet/be_ethtool.c
index 8e6fb0b..2496123 100644
--- a/drivers/net/ethernet/emulex/benet/be_ethtool.c
+++ b/drivers/net/ethernet/emulex/benet/be_ethtool.c
@@ -261,6 +261,9 @@ be_get_reg_len(struct net_device *netdev)
struct be_adapter *adapter = netdev_priv(netdev);
u32 log_size = 0;
+ if (!check_privilege(adapter, MAX_PRIVILEGES))
+ return 0;
+
if (be_physfn(adapter)) {
if (lancer_chip(adapter))
log_size = lancer_cmd_get_file_len(adapter,
@@ -787,6 +790,10 @@ static int
be_get_eeprom_len(struct net_device *netdev)
{
struct be_adapter *adapter = netdev_priv(netdev);
+
+ if (!check_privilege(adapter, MAX_PRIVILEGES))
+ return 0;
+
if (lancer_chip(adapter)) {
if (be_physfn(adapter))
return lancer_cmd_get_file_len(adapter,
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index fa895bc..cca8645 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -2745,6 +2745,11 @@ static void be_setup_init(struct be_adapter *adapter)
adapter->be3_native = false;
adapter->promiscuous = false;
adapter->eq_next_idx = 0;
+
+ if (be_physfn(adapter))
+ adapter->cmd_privileges = MAX_PRIVILEGES;
+ else
+ adapter->cmd_privileges = MIN_PRIVILEGES;
}
static int be_get_mac_addr(struct be_adapter *adapter, u8 *mac, u32 if_handle,
@@ -2917,6 +2922,13 @@ static int be_setup(struct be_adapter *adapter)
if (status)
goto err;
+ be_cmd_get_fn_privileges(adapter, &adapter->cmd_privileges, 0);
+ /* In UMC mode FW does not return right privileges.
+ * Override with correct privilege equivalent to PF.
+ */
+ if (be_is_mc(adapter))
+ adapter->cmd_privileges = MAX_PRIVILEGES;
+
en_flags = BE_IF_FLAGS_UNTAGGED | BE_IF_FLAGS_BROADCAST |
BE_IF_FLAGS_MULTICAST | BE_IF_FLAGS_PASS_L3L4_ERRORS;
@@ -2973,8 +2985,8 @@ static int be_setup(struct be_adapter *adapter)
dev_warn(dev, "device doesn't support SRIOV\n");
}
- be_cmd_get_phy_info(adapter);
- if (be_pause_supported(adapter))
+ status = be_cmd_get_phy_info(adapter);
+ if (!status && be_pause_supported(adapter))
adapter->phy.fc_autoneg = 1;
schedule_delayed_work(&adapter->work, msecs_to_jiffies(1000));
@@ -3711,6 +3723,9 @@ u32 be_get_fw_log_level(struct be_adapter *adapter)
u32 level = 0;
int j;
+ if (lancer_chip(adapter))
+ return 0;
+
memset(&extfat_cmd, 0, sizeof(struct be_dma_mem));
extfat_cmd.size = sizeof(struct be_cmd_resp_get_ext_fat_caps);
extfat_cmd.va = pci_alloc_consistent(adapter->pdev, extfat_cmd.size,
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 07/14] be2net: Fix ethtool get_settings output for VF
From: Padmanabh Ratnakar @ 2012-10-20 16:03 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
Return default values for fields for which VFs dont have privilege to get the
required information from FW.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_ethtool.c | 73 ++++++++++++++----------
1 files changed, 42 insertions(+), 31 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_ethtool.c b/drivers/net/ethernet/emulex/benet/be_ethtool.c
index 2496123..00454a1 100644
--- a/drivers/net/ethernet/emulex/benet/be_ethtool.c
+++ b/drivers/net/ethernet/emulex/benet/be_ethtool.c
@@ -528,6 +528,10 @@ static int be_get_settings(struct net_device *netdev, struct ethtool_cmd *ecmd)
u8 link_status;
u16 link_speed = 0;
int status;
+ u32 auto_speeds;
+ u32 fixed_speeds;
+ u32 dac_cable_len;
+ u16 interface_type;
if (adapter->phy.link_speed < 0) {
status = be_cmd_link_status_query(adapter, &link_speed,
@@ -537,39 +541,46 @@ static int be_get_settings(struct net_device *netdev, struct ethtool_cmd *ecmd)
ethtool_cmd_speed_set(ecmd, link_speed);
status = be_cmd_get_phy_info(adapter);
- if (status)
- return status;
-
- ecmd->supported =
- convert_to_et_setting(adapter->phy.interface_type,
- adapter->phy.auto_speeds_supported |
- adapter->phy.fixed_speeds_supported);
- ecmd->advertising =
- convert_to_et_setting(adapter->phy.interface_type,
- adapter->phy.auto_speeds_supported);
-
- ecmd->port = be_get_port_type(adapter->phy.interface_type,
- adapter->phy.dac_cable_len);
-
- if (adapter->phy.auto_speeds_supported) {
- ecmd->supported |= SUPPORTED_Autoneg;
- ecmd->autoneg = AUTONEG_ENABLE;
- ecmd->advertising |= ADVERTISED_Autoneg;
- }
+ if (!status) {
+ interface_type = adapter->phy.interface_type;
+ auto_speeds = adapter->phy.auto_speeds_supported;
+ fixed_speeds = adapter->phy.fixed_speeds_supported;
+ dac_cable_len = adapter->phy.dac_cable_len;
+
+ ecmd->supported =
+ convert_to_et_setting(interface_type,
+ auto_speeds |
+ fixed_speeds);
+ ecmd->advertising =
+ convert_to_et_setting(interface_type,
+ auto_speeds);
+
+ ecmd->port = be_get_port_type(interface_type,
+ dac_cable_len);
+
+ if (adapter->phy.auto_speeds_supported) {
+ ecmd->supported |= SUPPORTED_Autoneg;
+ ecmd->autoneg = AUTONEG_ENABLE;
+ ecmd->advertising |= ADVERTISED_Autoneg;
+ }
- if (be_pause_supported(adapter)) {
ecmd->supported |= SUPPORTED_Pause;
- ecmd->advertising |= ADVERTISED_Pause;
- }
-
- switch (adapter->phy.interface_type) {
- case PHY_TYPE_KR_10GB:
- case PHY_TYPE_KX4_10GB:
- ecmd->transceiver = XCVR_INTERNAL;
- break;
- default:
- ecmd->transceiver = XCVR_EXTERNAL;
- break;
+ if (be_pause_supported(adapter))
+ ecmd->advertising |= ADVERTISED_Pause;
+
+ switch (adapter->phy.interface_type) {
+ case PHY_TYPE_KR_10GB:
+ case PHY_TYPE_KX4_10GB:
+ ecmd->transceiver = XCVR_INTERNAL;
+ break;
+ default:
+ ecmd->transceiver = XCVR_EXTERNAL;
+ break;
+ }
+ } else {
+ ecmd->port = PORT_OTHER;
+ ecmd->autoneg = AUTONEG_DISABLE;
+ ecmd->transceiver = XCVR_DUMMY1;
}
/* Save for future use */
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 08/14] be2net: Fix issues in error recovery due to wrong queue state
From: Padmanabh Ratnakar @ 2012-10-20 16:03 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
During recovery from a FW error, destroy queue operation may fail.
Queue should be marked as destroyed so that recovery code can recreate
the queue. Also fix queue created state not getting checked at one instance.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_cmds.c | 9 +++++----
1 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index 9d602f4..15202d0 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -681,6 +681,9 @@ static struct be_mcc_wrb *wrb_from_mccq(struct be_adapter *adapter)
struct be_queue_info *mccq = &adapter->mcc_obj.q;
struct be_mcc_wrb *wrb;
+ if (!mccq->created)
+ return NULL;
+
if (atomic_read(&mccq->used) >= mccq->len) {
dev_err(&adapter->pdev->dev, "Out of MCCQ wrbs\n");
return NULL;
@@ -1235,8 +1238,7 @@ int be_cmd_q_destroy(struct be_adapter *adapter, struct be_queue_info *q,
req->id = cpu_to_le16(q->id);
status = be_mbox_notify_wait(adapter);
- if (!status)
- q->created = false;
+ q->created = false;
mutex_unlock(&adapter->mbox_lock);
return status;
@@ -1263,8 +1265,7 @@ int be_cmd_rxq_destroy(struct be_adapter *adapter, struct be_queue_info *q)
req->id = cpu_to_le16(q->id);
status = be_mcc_notify_wait(adapter);
- if (!status)
- q->created = false;
+ q->created = false;
err:
spin_unlock_bh(&adapter->mcc_lock);
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 09/14] be2net: Fix unnecessary delay in PCI EEH
From: Padmanabh Ratnakar @ 2012-10-20 16:03 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
During PCI EEH, driver waits for all functions in the card.
Wait is needed only once per card. Fix is to wait only for the
first PCI function.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_main.c | 8 ++++++--
1 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index cca8645..410834d 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -4201,9 +4201,13 @@ static pci_ers_result_t be_eeh_err_detected(struct pci_dev *pdev,
/* The error could cause the FW to trigger a flash debug dump.
* Resetting the card while flash dump is in progress
- * can cause it not to recover; wait for it to finish
+ * can cause it not to recover; wait for it to finish.
+ * Wait only for first function as it is needed only once per
+ * adapter.
*/
- ssleep(30);
+ if (pdev->devfn == 0)
+ ssleep(30);
+
return PCI_ERS_RESULT_NEED_RESET;
}
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 10/14] be2net: Fix VF driver load on newer Lancer FW
From: Padmanabh Ratnakar @ 2012-10-20 16:03 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
PF driver should enable VF so that VF goes to ready state in
new Lancer FW.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_cmds.c | 32 +++++++++++++++++++++++++++
drivers/net/ethernet/emulex/benet/be_cmds.h | 8 ++++++
drivers/net/ethernet/emulex/benet/be_main.c | 2 +
3 files changed, 42 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index 15202d0..8d35750 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -3131,6 +3131,38 @@ err:
return status;
}
+/* Uses sync mcc */
+int be_cmd_enable_vf(struct be_adapter *adapter, u8 domain)
+{
+ struct be_mcc_wrb *wrb;
+ struct be_cmd_enable_disable_vf *req;
+ int status;
+
+ if (!lancer_chip(adapter))
+ return 0;
+
+ spin_lock_bh(&adapter->mcc_lock);
+
+ wrb = wrb_from_mccq(adapter);
+ if (!wrb) {
+ status = -EBUSY;
+ goto err;
+ }
+
+ req = embedded_payload(wrb);
+
+ be_wrb_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON,
+ OPCODE_COMMON_ENABLE_DISABLE_VF, sizeof(*req),
+ wrb, NULL);
+
+ req->hdr.domain = domain;
+ req->enable = 1;
+ status = be_mcc_notify_wait(adapter);
+err:
+ spin_unlock_bh(&adapter->mcc_lock);
+ return status;
+}
+
int be_roce_mcc_cmd(void *netdev_handle, void *wrb_payload,
int wrb_payload_size, u16 *cmd_status, u16 *ext_status)
{
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.h b/drivers/net/ethernet/emulex/benet/be_cmds.h
index a438777..68de6c6 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.h
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.h
@@ -203,6 +203,7 @@ struct be_mcc_mailbox {
#define OPCODE_COMMON_GET_FN_PRIVILEGES 170
#define OPCODE_COMMON_READ_OBJECT 171
#define OPCODE_COMMON_WRITE_OBJECT 172
+#define OPCODE_COMMON_ENABLE_DISABLE_VF 196
#define OPCODE_ETH_RSS_CONFIG 1
#define OPCODE_ETH_ACPI_CONFIG 2
@@ -1802,6 +1803,12 @@ struct be_cmd_resp_set_profile_config {
struct be_cmd_req_hdr hdr;
};
+struct be_cmd_enable_disable_vf {
+ struct be_cmd_req_hdr hdr;
+ u8 enable;
+ u8 rsvd[3];
+};
+
static inline bool check_privilege(struct be_adapter *adapter, u32 flags)
{
return flags & adapter->cmd_privileges ? true : false;
@@ -1929,3 +1936,4 @@ extern int be_cmd_get_profile_config(struct be_adapter *adapter, u32 *cap_flags,
extern int be_cmd_set_profile_config(struct be_adapter *adapter, u32 bps,
u8 domain);
+extern int be_cmd_enable_vf(struct be_adapter *adapter, u8 domain);
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 410834d..136ed64 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -2731,6 +2731,8 @@ static int be_vf_setup(struct be_adapter *adapter)
if (status)
goto err;
vf_cfg->def_vid = def_vlan;
+
+ be_cmd_enable_vf(adapter, vf + 1);
}
return 0;
err:
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 11/14] be2net: Enabling Wake-on-LAN is not supported in S5 state
From: Padmanabh Ratnakar @ 2012-10-20 16:04 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar, Vasundhara Volam
be_shutdown is enabling wake-on-lan by calling be_setup_wol.
Emulex adapter do not support wake-on-lan in S5 state.
Signed-off-by: Vasundhara Volam <vasundhara.volam@emulex.com>
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_main.c | 3 ---
1 files changed, 0 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 136ed64..ab077fb 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -4165,9 +4165,6 @@ static void be_shutdown(struct pci_dev *pdev)
netif_device_detach(adapter->netdev);
- if (adapter->wol)
- be_setup_wol(adapter, true);
-
be_cmd_reset_function(adapter);
pci_disable_device(pdev);
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 12/14] be2net: Fix FW flashing on Skyhawk-R
From: Padmanabh Ratnakar @ 2012-10-20 16:04 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar, Vasundhara Volam
FW flash layout on Skyhawk-R is different from BE3-R.
Hence the code needs to be fixed to flash FW on Skyhawk-R.
Also cleaning up code in BE3-R flashing function.
Signed-off-by: Vasundhara Volam <vasundhara.volam@emulex.com>
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be.h | 1 +
drivers/net/ethernet/emulex/benet/be_main.c | 248 +++++++++++++++++++--------
2 files changed, 173 insertions(+), 76 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index 1a05b79..d95825f 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -459,6 +459,7 @@ struct be_adapter {
/* BladeEngine Generation numbers */
#define BE_GEN2 2
#define BE_GEN3 3
+#define SH_HW 4
#define ON 1
#define OFF 0
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index ab077fb..2143e06 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -3088,6 +3088,47 @@ struct flash_section_info *get_fsec_info(struct be_adapter *adapter,
return NULL;
}
+static int be_flash(struct be_adapter *adapter, const u8 *img,
+ struct be_dma_mem *flash_cmd, int optype, int img_size)
+{
+ u32 total_bytes = 0, flash_op, num_bytes = 0;
+ int status = 0;
+ struct be_cmd_write_flashrom *req = flash_cmd->va;
+
+ total_bytes = img_size;
+ while (total_bytes) {
+ num_bytes = min_t(u32, 32*1024, total_bytes);
+
+ total_bytes -= num_bytes;
+
+ if (!total_bytes) {
+ if (optype == OPTYPE_PHY_FW)
+ flash_op = FLASHROM_OPER_PHY_FLASH;
+ else
+ flash_op = FLASHROM_OPER_FLASH;
+ } else {
+ if (optype == OPTYPE_PHY_FW)
+ flash_op = FLASHROM_OPER_PHY_SAVE;
+ else
+ flash_op = FLASHROM_OPER_SAVE;
+ }
+
+ memcpy(req->params.data_buf, img, num_bytes);
+ img += num_bytes;
+ status = be_cmd_write_flashrom(adapter, flash_cmd, optype,
+ flash_op, num_bytes);
+ if (status) {
+ if (status == ILLEGAL_IOCTL_REQ &&
+ optype == OPTYPE_PHY_FW)
+ break;
+ dev_err(&adapter->pdev->dev,
+ "cmd to write to flash rom failed.\n");
+ return status;
+ }
+ }
+ return 0;
+}
+
static int be_flash_data(struct be_adapter *adapter,
const struct firmware *fw,
struct be_dma_mem *flash_cmd,
@@ -3096,12 +3137,9 @@ static int be_flash_data(struct be_adapter *adapter,
{
int status = 0, i, filehdr_size = 0;
int img_hdrs_size = (num_of_images * sizeof(struct image_hdr));
- u32 total_bytes = 0, flash_op;
- int num_bytes;
const u8 *p = fw->data;
- struct be_cmd_write_flashrom *req = flash_cmd->va;
const struct flash_comp *pflashcomp;
- int num_comp, hdr_size;
+ int num_comp, redboot;
struct flash_section_info *fsec = NULL;
struct flash_comp gen3_flash_types[] = {
@@ -3170,70 +3208,105 @@ static int be_flash_data(struct be_adapter *adapter,
memcmp(adapter->fw_ver, "3.102.148.0", 11) < 0)
continue;
- if (pflashcomp[i].optype == OPTYPE_PHY_FW) {
- if (!phy_flashing_required(adapter))
+ if (pflashcomp[i].optype == OPTYPE_PHY_FW &&
+ !phy_flashing_required(adapter))
continue;
- }
-
- hdr_size = filehdr_size +
- (num_of_images * sizeof(struct image_hdr));
- if ((pflashcomp[i].optype == OPTYPE_REDBOOT) &&
- (!be_flash_redboot(adapter, fw->data, pflashcomp[i].offset,
- pflashcomp[i].size, hdr_size)))
- continue;
+ if (pflashcomp[i].optype == OPTYPE_REDBOOT) {
+ redboot = be_flash_redboot(adapter, fw->data,
+ pflashcomp[i].offset, pflashcomp[i].size,
+ filehdr_size + img_hdrs_size);
+ if (!redboot)
+ continue;
+ }
- /* Flash the component */
p = fw->data;
p += filehdr_size + pflashcomp[i].offset + img_hdrs_size;
if (p + pflashcomp[i].size > fw->data + fw->size)
return -1;
- total_bytes = pflashcomp[i].size;
- while (total_bytes) {
- if (total_bytes > 32*1024)
- num_bytes = 32*1024;
- else
- num_bytes = total_bytes;
- total_bytes -= num_bytes;
- if (!total_bytes) {
- if (pflashcomp[i].optype == OPTYPE_PHY_FW)
- flash_op = FLASHROM_OPER_PHY_FLASH;
- else
- flash_op = FLASHROM_OPER_FLASH;
- } else {
- if (pflashcomp[i].optype == OPTYPE_PHY_FW)
- flash_op = FLASHROM_OPER_PHY_SAVE;
- else
- flash_op = FLASHROM_OPER_SAVE;
- }
- memcpy(req->params.data_buf, p, num_bytes);
- p += num_bytes;
- status = be_cmd_write_flashrom(adapter, flash_cmd,
- pflashcomp[i].optype, flash_op, num_bytes);
- if (status) {
- if ((status == ILLEGAL_IOCTL_REQ) &&
- (pflashcomp[i].optype ==
- OPTYPE_PHY_FW))
- break;
- dev_err(&adapter->pdev->dev,
- "cmd to write to flash rom failed.\n");
- return -1;
- }
+
+ status = be_flash(adapter, p, flash_cmd, pflashcomp[i].optype,
+ pflashcomp[i].size);
+ if (status) {
+ dev_err(&adapter->pdev->dev,
+ "Flashing section type %d failed.\n",
+ pflashcomp[i].img_type);
+ return status;
}
}
return 0;
}
-static int get_ufigen_type(struct flash_file_hdr_g2 *fhdr)
+static int be_flash_skyhawk(struct be_adapter *adapter,
+ const struct firmware *fw,
+ struct be_dma_mem *flash_cmd, int num_of_images)
{
- if (fhdr == NULL)
- return 0;
- if (fhdr->build[0] == '3')
- return BE_GEN3;
- else if (fhdr->build[0] == '2')
- return BE_GEN2;
- else
- return 0;
+ int status = 0, i, filehdr_size = 0;
+ int img_offset, img_size, img_optype, redboot;
+ int img_hdrs_size = num_of_images * sizeof(struct image_hdr);
+ const u8 *p = fw->data;
+ struct flash_section_info *fsec = NULL;
+
+ filehdr_size = sizeof(struct flash_file_hdr_g3);
+ fsec = get_fsec_info(adapter, filehdr_size + img_hdrs_size, fw);
+ if (!fsec) {
+ dev_err(&adapter->pdev->dev,
+ "Invalid Cookie. UFI corrupted ?\n");
+ return -1;
+ }
+
+ for (i = 0; i < le32_to_cpu(fsec->fsec_hdr.num_images); i++) {
+ img_offset = le32_to_cpu(fsec->fsec_entry[i].offset);
+ img_size = le32_to_cpu(fsec->fsec_entry[i].pad_size);
+
+ switch (le32_to_cpu(fsec->fsec_entry[i].type)) {
+ case IMAGE_FIRMWARE_iSCSI:
+ img_optype = OPTYPE_ISCSI_ACTIVE;
+ break;
+ case IMAGE_BOOT_CODE:
+ img_optype = OPTYPE_REDBOOT;
+ break;
+ case IMAGE_OPTION_ROM_ISCSI:
+ img_optype = OPTYPE_BIOS;
+ break;
+ case IMAGE_OPTION_ROM_PXE:
+ img_optype = OPTYPE_PXE_BIOS;
+ break;
+ case IMAGE_OPTION_ROM_FCoE:
+ img_optype = OPTYPE_FCOE_BIOS;
+ break;
+ case IMAGE_FIRMWARE_BACKUP_iSCSI:
+ img_optype = OPTYPE_ISCSI_BACKUP;
+ break;
+ case IMAGE_NCSI:
+ img_optype = OPTYPE_NCSI_FW;
+ break;
+ default:
+ continue;
+ }
+
+ if (img_optype == OPTYPE_REDBOOT) {
+ redboot = be_flash_redboot(adapter, fw->data,
+ img_offset, img_size,
+ filehdr_size + img_hdrs_size);
+ if (!redboot)
+ continue;
+ }
+
+ p = fw->data;
+ p += filehdr_size + img_offset + img_hdrs_size;
+ if (p + img_size > fw->data + fw->size)
+ return -1;
+
+ status = be_flash(adapter, p, flash_cmd, img_optype, img_size);
+ if (status) {
+ dev_err(&adapter->pdev->dev,
+ "Flashing section type %d failed.\n",
+ fsec->fsec_entry[i].type);
+ return status;
+ }
+ }
+ return 0;
}
static int lancer_wait_idle(struct be_adapter *adapter)
@@ -3367,6 +3440,27 @@ lancer_fw_exit:
return status;
}
+static int be_get_ufi_gen(struct be_adapter *adapter,
+ struct flash_file_hdr_g2 *fhdr)
+{
+ if (fhdr == NULL)
+ goto be_get_ufi_exit;
+
+ if (adapter->generation == BE_GEN3) {
+ if (skyhawk_chip(adapter) && fhdr->build[0] == '4')
+ return SH_HW;
+ else if (!skyhawk_chip(adapter) && fhdr->build[0] == '3')
+ return BE_GEN3;
+ } else if (adapter->generation == BE_GEN2 && fhdr->build[0] == '2') {
+ return BE_GEN2;
+ }
+
+be_get_ufi_exit:
+ dev_err(&adapter->pdev->dev,
+ "UFI and Interface are not compatible for flashing\n");
+ return -1;
+}
+
static int be_fw_download(struct be_adapter *adapter, const struct firmware* fw)
{
struct flash_file_hdr_g2 *fhdr;
@@ -3374,10 +3468,7 @@ static int be_fw_download(struct be_adapter *adapter, const struct firmware* fw)
struct image_hdr *img_hdr_ptr = NULL;
struct be_dma_mem flash_cmd;
const u8 *p;
- int status = 0, i = 0, num_imgs = 0;
-
- p = fw->data;
- fhdr = (struct flash_file_hdr_g2 *) p;
+ int status = 0, i = 0, num_imgs = 0, ufi_type = 0;
flash_cmd.size = sizeof(struct be_cmd_write_flashrom) + 32*1024;
flash_cmd.va = dma_alloc_coherent(&adapter->pdev->dev, flash_cmd.size,
@@ -3389,26 +3480,31 @@ static int be_fw_download(struct be_adapter *adapter, const struct firmware* fw)
goto be_fw_exit;
}
- if ((adapter->generation == BE_GEN3) &&
- (get_ufigen_type(fhdr) == BE_GEN3)) {
- fhdr3 = (struct flash_file_hdr_g3 *) fw->data;
- num_imgs = le32_to_cpu(fhdr3->num_imgs);
- for (i = 0; i < num_imgs; i++) {
- img_hdr_ptr = (struct image_hdr *) (fw->data +
- (sizeof(struct flash_file_hdr_g3) +
- i * sizeof(struct image_hdr)));
- if (le32_to_cpu(img_hdr_ptr->imageid) == 1)
- status = be_flash_data(adapter, fw, &flash_cmd,
- num_imgs);
+ p = fw->data;
+ fhdr = (struct flash_file_hdr_g2 *)p;
+
+ ufi_type = be_get_ufi_gen(adapter, fhdr);
+
+ fhdr3 = (struct flash_file_hdr_g3 *)fw->data;
+ num_imgs = le32_to_cpu(fhdr3->num_imgs);
+ for (i = 0; i < num_imgs; i++) {
+ img_hdr_ptr = (struct image_hdr *)(fw->data +
+ (sizeof(struct flash_file_hdr_g3) +
+ i * sizeof(struct image_hdr)));
+ if (le32_to_cpu(img_hdr_ptr->imageid) == 1) {
+ if (ufi_type == SH_HW)
+ status = be_flash_skyhawk(adapter, fw,
+ &flash_cmd, num_imgs);
+ else if (ufi_type == BE_GEN3)
+ status = be_flash_data(adapter, fw,
+ &flash_cmd, num_imgs);
}
- } else if ((adapter->generation == BE_GEN2) &&
- (get_ufigen_type(fhdr) == BE_GEN2)) {
+ }
+
+ if (ufi_type == BE_GEN2)
status = be_flash_data(adapter, fw, &flash_cmd, 0);
- } else {
- dev_err(&adapter->pdev->dev,
- "UFI and Interface are not compatible for flashing\n");
+ else if (ufi_type == -1)
status = -1;
- }
dma_free_coherent(&adapter->pdev->dev, flash_cmd.size, flash_cmd.va,
flash_cmd.dma);
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 13/14] be2net: Fix skyhawk VF PCI Device ID
From: Padmanabh Ratnakar @ 2012-10-20 16:04 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be.h | 5 ++++-
drivers/net/ethernet/emulex/benet/be_main.c | 2 ++
2 files changed, 6 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index d95825f..3b66c63 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -53,6 +53,7 @@
#define OC_DEVICE_ID3 0xe220 /* Device id for Lancer cards */
#define OC_DEVICE_ID4 0xe228 /* Device id for VF in Lancer */
#define OC_DEVICE_ID5 0x720 /* Device Id for Skyhawk cards */
+#define OC_DEVICE_ID6 0x728 /* Device id for VF in SkyHawk */
#define OC_SUBSYS_DEVICE_ID1 0xE602
#define OC_SUBSYS_DEVICE_ID2 0xE642
#define OC_SUBSYS_DEVICE_ID3 0xE612
@@ -71,6 +72,7 @@ static inline char *nic_name(struct pci_dev *pdev)
case BE_DEVICE_ID2:
return BE3_NAME;
case OC_DEVICE_ID5:
+ case OC_DEVICE_ID6:
return OC_NAME_SH;
default:
return BE_NAME;
@@ -466,7 +468,8 @@ struct be_adapter {
#define lancer_chip(adapter) ((adapter->pdev->device == OC_DEVICE_ID3) || \
(adapter->pdev->device == OC_DEVICE_ID4))
-#define skyhawk_chip(adapter) (adapter->pdev->device == OC_DEVICE_ID5)
+#define skyhawk_chip(adapter) (adapter->pdev->device == OC_DEVICE_ID5 || \
+ adapter->pdev->device == OC_DEVICE_ID6)
#define be_roce_supported(adapter) ((adapter->if_type == SLI_INTF_TYPE_3 || \
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 2143e06..68665da 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -44,6 +44,7 @@ static DEFINE_PCI_DEVICE_TABLE(be_dev_ids) = {
{ PCI_DEVICE(EMULEX_VENDOR_ID, OC_DEVICE_ID3)},
{ PCI_DEVICE(EMULEX_VENDOR_ID, OC_DEVICE_ID4)},
{ PCI_DEVICE(EMULEX_VENDOR_ID, OC_DEVICE_ID5)},
+ { PCI_DEVICE(EMULEX_VENDOR_ID, OC_DEVICE_ID6)},
{ 0 }
};
MODULE_DEVICE_TABLE(pci, be_dev_ids);
@@ -3910,6 +3911,7 @@ static int be_dev_type_check(struct be_adapter *adapter)
adapter->generation = BE_GEN3;
break;
case OC_DEVICE_ID5:
+ case OC_DEVICE_ID6:
pci_read_config_dword(pdev, SLI_INTF_REG_OFFSET, &sli_intf);
if ((sli_intf & SLI_INTF_VALID_MASK) != SLI_INTF_VALID) {
dev_err(&pdev->dev, "SLI_INTF reg val is not valid\n");
--
1.6.0.2
^ permalink raw reply related
* [PATCH net-next 14/14] be2net: Update driver version
From: Padmanabh Ratnakar @ 2012-10-20 16:04 UTC (permalink / raw)
To: netdev; +Cc: Padmanabh Ratnakar
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
---
drivers/net/ethernet/emulex/benet/be.h | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index 3b66c63..f920a65 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -34,7 +34,7 @@
#include "be_hw.h"
#include "be_roce.h"
-#define DRV_VER "4.4.31.0u"
+#define DRV_VER "4.4.161.0u"
#define DRV_NAME "be2net"
#define BE_NAME "ServerEngines BladeEngine2 10Gbps NIC"
#define BE3_NAME "ServerEngines BladeEngine3 10Gbps NIC"
--
1.6.0.2
^ permalink raw reply related
* Re: [PATCH net-next 04/21] wireless: Convert is_<foo>_ether_addr uses to eth_addr_<foo>
From: Arend van Spriel @ 2012-10-20 16:19 UTC (permalink / raw)
To: Joe Perches
Cc: John W. Linville, linux-wireless@vger.kernel.org, netdev,
linux-kernel@vger.kernel.org, brcm80211 development
In-Reply-To: <bb2235c257e347354d94478bd1be35b89bf90186.1350618009.git.joe@perches.com>
On 10/19/2012 05:55 AM, Joe Perches wrote:
> Convert the old ether_addr tests to eth_addr_<foo>.
> Adds api consistency.
Acked-by: Arend van Spriel <arend@broadcom.com>
> Signed-off-by: Joe Perches <joe@perches.com>
> ---
> .../net/wireless/brcm80211/brcmfmac/dhd_linux.c | 4 ++--
> .../net/wireless/brcm80211/brcmfmac/wl_cfg80211.c | 2 +-
> .../net/wireless/brcm80211/brcmsmac/mac80211_if.c | 2 +-
> drivers/net/wireless/brcm80211/brcmsmac/main.c | 16 ++++++++--------
^ permalink raw reply
* [PATCH] ixgbe: Use is_valid_ether_addr
From: Joe Perches @ 2012-10-20 16:22 UTC (permalink / raw)
To: Jeff Kirsher; +Cc: netdev, linux-kernel
Use the normal kernel test instead of a module specific one.
Signed-off-by: Joe Perches <joe@perches.com>
---
found when doing that larger style conversion,
might as well submit it.
drivers/net/ethernet/intel/ixgbe/ixgbe_82599.c | 2 +-
drivers/net/ethernet/intel/ixgbe/ixgbe_common.c | 27 +------------------------
drivers/net/ethernet/intel/ixgbe/ixgbe_common.h | 1 -
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 2 +-
drivers/net/ethernet/intel/ixgbe/ixgbe_type.h | 9 ---------
drivers/net/ethernet/intel/ixgbe/ixgbe_x540.c | 2 +-
6 files changed, 4 insertions(+), 39 deletions(-)
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_82599.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_82599.c
index 1077cb2..89fe00d 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_82599.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_82599.c
@@ -1022,7 +1022,7 @@ mac_reset_top:
hw->mac.ops.get_san_mac_addr(hw, hw->mac.san_addr);
/* Add the SAN MAC address to the RAR only if it's a valid address */
- if (ixgbe_validate_mac_addr(hw->mac.san_addr) == 0) {
+ if (is_valid_ether_addr(hw->mac.san_addr)) {
hw->mac.ops.set_rar(hw, hw->mac.num_rar_entries - 1,
hw->mac.san_addr, 0, IXGBE_RAH_AV);
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.c
index dbf37e4..2d8f76d 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.c
@@ -1762,30 +1762,6 @@ s32 ixgbe_update_eeprom_checksum_generic(struct ixgbe_hw *hw)
}
/**
- * ixgbe_validate_mac_addr - Validate MAC address
- * @mac_addr: pointer to MAC address.
- *
- * Tests a MAC address to ensure it is a valid Individual Address
- **/
-s32 ixgbe_validate_mac_addr(u8 *mac_addr)
-{
- s32 status = 0;
-
- /* Make sure it is not a multicast address */
- if (IXGBE_IS_MULTICAST(mac_addr))
- status = IXGBE_ERR_INVALID_MAC_ADDR;
- /* Not a broadcast address */
- else if (IXGBE_IS_BROADCAST(mac_addr))
- status = IXGBE_ERR_INVALID_MAC_ADDR;
- /* Reject the zero address */
- else if (mac_addr[0] == 0 && mac_addr[1] == 0 && mac_addr[2] == 0 &&
- mac_addr[3] == 0 && mac_addr[4] == 0 && mac_addr[5] == 0)
- status = IXGBE_ERR_INVALID_MAC_ADDR;
-
- return status;
-}
-
-/**
* ixgbe_set_rar_generic - Set Rx address register
* @hw: pointer to hardware structure
* @index: Receive address register to write
@@ -1889,8 +1865,7 @@ s32 ixgbe_init_rx_addrs_generic(struct ixgbe_hw *hw)
* to the permanent address.
* Otherwise, use the permanent address from the eeprom.
*/
- if (ixgbe_validate_mac_addr(hw->mac.addr) ==
- IXGBE_ERR_INVALID_MAC_ADDR) {
+ if (!is_valid_ether_addr(hw->mac.addr)) {
/* Get the MAC address from the RAR0 for later reference */
hw->mac.ops.get_mac_addr(hw, hw->mac.addr);
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
index d813d11..f49ca84 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
@@ -80,7 +80,6 @@ s32 ixgbe_enable_rx_dma_generic(struct ixgbe_hw *hw, u32 regval);
s32 ixgbe_fc_enable_generic(struct ixgbe_hw *hw);
void ixgbe_fc_autoneg(struct ixgbe_hw *hw);
-s32 ixgbe_validate_mac_addr(u8 *mac_addr);
s32 ixgbe_acquire_swfw_sync(struct ixgbe_hw *hw, u16 mask);
void ixgbe_release_swfw_sync(struct ixgbe_hw *hw, u16 mask);
s32 ixgbe_get_san_mac_addr_generic(struct ixgbe_hw *hw, u8 *san_mac_addr);
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
index e2a6691..3bb3485 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
@@ -7345,7 +7345,7 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
memcpy(netdev->dev_addr, hw->mac.perm_addr, netdev->addr_len);
memcpy(netdev->perm_addr, hw->mac.perm_addr, netdev->addr_len);
- if (ixgbe_validate_mac_addr(netdev->perm_addr)) {
+ if (!is_valid_ether_addr(netdev->perm_addr)) {
e_dev_err("invalid MAC address\n");
err = -EIO;
goto err_sw_init;
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_type.h b/drivers/net/ethernet/intel/ixgbe/ixgbe_type.h
index 0722f33..9ddac64 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_type.h
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_type.h
@@ -1833,15 +1833,6 @@ enum {
/* Number of 100 microseconds we wait for PCI Express master disable */
#define IXGBE_PCI_MASTER_DISABLE_TIMEOUT 800
-/* Check whether address is multicast. This is little-endian specific check.*/
-#define IXGBE_IS_MULTICAST(Address) \
- (bool)(((u8 *)(Address))[0] & ((u8)0x01))
-
-/* Check whether an address is broadcast. */
-#define IXGBE_IS_BROADCAST(Address) \
- ((((u8 *)(Address))[0] == ((u8)0xff)) && \
- (((u8 *)(Address))[1] == ((u8)0xff)))
-
/* RAH */
#define IXGBE_RAH_VIND_MASK 0x003C0000
#define IXGBE_RAH_VIND_SHIFT 18
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_x540.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_x540.c
index de4da52..c73b929 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_x540.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_x540.c
@@ -152,7 +152,7 @@ mac_reset_top:
hw->mac.ops.get_san_mac_addr(hw, hw->mac.san_addr);
/* Add the SAN MAC address to the RAR only if it's a valid address */
- if (ixgbe_validate_mac_addr(hw->mac.san_addr) == 0) {
+ if (is_valid_ether_addr(hw->mac.san_addr)) {
hw->mac.ops.set_rar(hw, hw->mac.num_rar_entries - 1,
hw->mac.san_addr, 0, IXGBE_RAH_AV);
^ permalink raw reply related
* [PATCH net-next V2] htb: improved accuracy at high rates
From: Vimalkumar @ 2012-10-20 16:38 UTC (permalink / raw)
To: davem, Jamal Hadi Salim, Eric Dumazet, netdev; +Cc: Vimalkumar
Current HTB (and TBF) uses rate table computed by the "tc"
userspace program, which has the following issue:
The rate table has 256 entries to map packet lengths
to token (time units). With TSO sized packets, the
256 entry granularity leads to loss/gain of rate,
making the token bucket inaccurate.
Thus, instead of relying on rate table, this patch
explicitly computes the time and accounts for packet
transmission times with nanosecond granularity.
This greatly improves accuracy of HTB with a wide
range of packet sizes.
Example:
tc qdisc add dev $dev root handle 1: \
htb default 1
tc class add dev $dev classid 1:1 parent 1: \
rate 5Gbit mtu 64k
Here is an example of inaccuracy:
$ iperf -c host -t 10 -i 1
With old htb:
eth4: 34.76 Mb/s In 5827.98 Mb/s Out - 65836.0 p/s In 481273.0 p/s Out
[SUM] 9.0-10.0 sec 669 MBytes 5.61 Gbits/sec
[SUM] 0.0-10.0 sec 6.50 GBytes 5.58 Gbits/sec
With new htb:
eth4: 28.36 Mb/s In 5208.06 Mb/s Out - 53704.0 p/s In 430076.0 p/s Out
[SUM] 9.0-10.0 sec 594 MBytes 4.98 Gbits/sec
[SUM] 0.0-10.0 sec 5.80 GBytes 4.98 Gbits/sec
The bits per second on the wire is still 5200Mb/s with new HTB
because qdisc accounts for packet length using skb->len, which
is smaller than total bytes on the wire if GSO is used. But
that is for another patch regardless of how time is accounted.
Many thanks to Eric Dumazet for review and feedback.
Signed-off-by: Vimalkumar <j.vimal@gmail.com>
---
net/sched/sch_htb.c | 123 +++++++++++++++++++++++++++++++++------------------
1 files changed, 80 insertions(+), 43 deletions(-)
diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c
index 9d75b77..1f26d6b 100644
--- a/net/sched/sch_htb.c
+++ b/net/sched/sch_htb.c
@@ -55,6 +55,7 @@
static int htb_hysteresis __read_mostly = 0; /* whether to use mode hysteresis for speedup */
#define HTB_VER 0x30011 /* major must be matched with number suplied by TC as version */
+#define HTB_MIN_PKT_BYTES (64)
#if HTB_VER >> 16 != TC_HTB_PROTOVER
#error "Mismatched sch_htb.c and pkt_sch.h"
@@ -71,6 +72,12 @@ enum htb_cmode {
HTB_CAN_SEND /* class can send */
};
+struct htb_rate_cfg {
+ u64 rate_bps;
+ u32 mult;
+ u32 shift;
+};
+
/* interior & leaf nodes; props specific to leaves are marked L: */
struct htb_class {
struct Qdisc_class_common common;
@@ -118,11 +125,11 @@ struct htb_class {
int filter_cnt;
/* token bucket parameters */
- struct qdisc_rate_table *rate; /* rate table of the class itself */
- struct qdisc_rate_table *ceil; /* ceiling rate (limits borrows too) */
- long buffer, cbuffer; /* token bucket depth/rate */
+ struct htb_rate_cfg rate;
+ struct htb_rate_cfg ceil;
+ s64 buffer, cbuffer; /* token bucket depth/rate */
psched_tdiff_t mbuffer; /* max wait time */
- long tokens, ctokens; /* current number of tokens */
+ s64 tokens, ctokens; /* current number of tokens */
psched_time_t t_c; /* checkpoint time */
};
@@ -162,6 +169,30 @@ struct htb_sched {
struct work_struct work;
};
+static u64 l2t_ns(struct htb_rate_cfg *r, unsigned int len)
+{
+ return ((u64)len * r->mult) >> r->shift;
+}
+
+static void htb_precompute_ratedata(struct htb_rate_cfg *r)
+{
+ u64 factor;
+ r->shift = 0;
+ r->mult = 1;
+ /*
+ * Calibrate mult, shift so that token counting is accurate
+ * for smallest packet size (64 bytes). Token (time in ns) is
+ * computed as (bytes * 8) * NSEC_PER_SEC / rate_bps. It will
+ * work as long as the smallest packet transfer time can be
+ * accurately represented in nanosec.
+ */
+ if (r->rate_bps > 0) {
+ r->shift = 15;
+ factor = 8LLU * NSEC_PER_SEC * (1 << r->shift);
+ r->mult = div64_u64(factor, r->rate_bps);
+ }
+}
+
/* find class in global hash table using given handle */
static inline struct htb_class *htb_find(u32 handle, struct Qdisc *sch)
{
@@ -273,7 +304,7 @@ static void htb_add_to_id_tree(struct rb_root *root,
* already in the queue.
*/
static void htb_add_to_wait_tree(struct htb_sched *q,
- struct htb_class *cl, long delay)
+ struct htb_class *cl, s64 delay)
{
struct rb_node **p = &q->wait_pq[cl->level].rb_node, *parent = NULL;
@@ -441,14 +472,14 @@ static void htb_deactivate_prios(struct htb_sched *q, struct htb_class *cl)
htb_remove_class_from_row(q, cl, mask);
}
-static inline long htb_lowater(const struct htb_class *cl)
+static inline s64 htb_lowater(const struct htb_class *cl)
{
if (htb_hysteresis)
return cl->cmode != HTB_CANT_SEND ? -cl->cbuffer : 0;
else
return 0;
}
-static inline long htb_hiwater(const struct htb_class *cl)
+static inline s64 htb_hiwater(const struct htb_class *cl)
{
if (htb_hysteresis)
return cl->cmode == HTB_CAN_SEND ? -cl->buffer : 0;
@@ -469,9 +500,9 @@ static inline long htb_hiwater(const struct htb_class *cl)
* mode transitions per time unit. The speed gain is about 1/6.
*/
static inline enum htb_cmode
-htb_class_mode(struct htb_class *cl, long *diff)
+htb_class_mode(struct htb_class *cl, s64 *diff)
{
- long toks;
+ s64 toks;
if ((toks = (cl->ctokens + *diff)) < htb_lowater(cl)) {
*diff = -toks;
@@ -495,7 +526,7 @@ htb_class_mode(struct htb_class *cl, long *diff)
* to mode other than HTB_CAN_SEND (see htb_add_to_wait_tree).
*/
static void
-htb_change_class_mode(struct htb_sched *q, struct htb_class *cl, long *diff)
+htb_change_class_mode(struct htb_sched *q, struct htb_class *cl, s64 *diff)
{
enum htb_cmode new_mode = htb_class_mode(cl, diff);
@@ -558,7 +589,9 @@ static int htb_enqueue(struct sk_buff *skb, struct Qdisc *sch)
__skb_queue_tail(&q->direct_queue, skb);
q->direct_pkts++;
} else {
- return qdisc_drop(skb, sch);
+ kfree_skb(skb);
+ sch->qstats.drops++;
+ return NET_XMIT_DROP;
}
#ifdef CONFIG_NET_CLS_ACT
} else if (!cl) {
@@ -574,6 +607,7 @@ static int htb_enqueue(struct sk_buff *skb, struct Qdisc *sch)
}
return ret;
} else {
+ bstats_update(&cl->bstats, skb);
htb_activate(q, cl);
}
@@ -581,26 +615,26 @@ static int htb_enqueue(struct sk_buff *skb, struct Qdisc *sch)
return NET_XMIT_SUCCESS;
}
-static inline void htb_accnt_tokens(struct htb_class *cl, int bytes, long diff)
+static inline void htb_accnt_tokens(struct htb_class *cl, int bytes, s64 diff)
{
- long toks = diff + cl->tokens;
+ s64 toks = diff + cl->tokens;
if (toks > cl->buffer)
toks = cl->buffer;
- toks -= (long) qdisc_l2t(cl->rate, bytes);
+ toks -= (s64) l2t_ns(&cl->rate, bytes);
if (toks <= -cl->mbuffer)
toks = 1 - cl->mbuffer;
cl->tokens = toks;
}
-static inline void htb_accnt_ctokens(struct htb_class *cl, int bytes, long diff)
+static inline void htb_accnt_ctokens(struct htb_class *cl, int bytes, s64 diff)
{
- long toks = diff + cl->ctokens;
+ s64 toks = diff + cl->ctokens;
if (toks > cl->cbuffer)
toks = cl->cbuffer;
- toks -= (long) qdisc_l2t(cl->ceil, bytes);
+ toks -= (s64) l2t_ns(&cl->ceil, bytes);
if (toks <= -cl->mbuffer)
toks = 1 - cl->mbuffer;
@@ -623,10 +657,10 @@ static void htb_charge_class(struct htb_sched *q, struct htb_class *cl,
{
int bytes = qdisc_pkt_len(skb);
enum htb_cmode old_mode;
- long diff;
+ s64 diff;
while (cl) {
- diff = psched_tdiff_bounded(q->now, cl->t_c, cl->mbuffer);
+ diff = min_t(s64, q->now - cl->t_c, cl->mbuffer);
if (cl->level >= level) {
if (cl->level == level)
cl->xstats.lends++;
@@ -673,7 +707,7 @@ static psched_time_t htb_do_events(struct htb_sched *q, int level,
unsigned long stop_at = start + 2;
while (time_before(jiffies, stop_at)) {
struct htb_class *cl;
- long diff;
+ s64 diff;
struct rb_node *p = rb_first(&q->wait_pq[level]);
if (!p)
@@ -684,7 +718,7 @@ static psched_time_t htb_do_events(struct htb_sched *q, int level,
return cl->pq_key;
htb_safe_rb_erase(p, q->wait_pq + level);
- diff = psched_tdiff_bounded(q->now, cl->t_c, cl->mbuffer);
+ diff = min_t(s64, q->now - cl->t_c, cl->mbuffer);
htb_change_class_mode(q, cl, &diff);
if (cl->cmode != HTB_CAN_SEND)
htb_add_to_wait_tree(q, cl, diff);
@@ -834,7 +868,6 @@ next:
} while (cl != start);
if (likely(skb != NULL)) {
- bstats_update(&cl->bstats, skb);
cl->un.leaf.deficit[level] -= qdisc_pkt_len(skb);
if (cl->un.leaf.deficit[level] < 0) {
cl->un.leaf.deficit[level] += cl->quantum;
@@ -871,10 +904,10 @@ ok:
if (!sch->q.qlen)
goto fin;
- q->now = psched_get_time();
+ q->now = ktime_to_ns(ktime_get());
start_at = jiffies;
- next_event = q->now + 5 * PSCHED_TICKS_PER_SEC;
+ next_event = q->now + 5 * NSEC_PER_SEC;
for (level = 0; level < TC_HTB_MAXDEPTH; level++) {
/* common case optimization - skip event handler quickly */
@@ -884,7 +917,7 @@ ok:
if (q->now >= q->near_ev_cache[level]) {
event = htb_do_events(q, level, start_at);
if (!event)
- event = q->now + PSCHED_TICKS_PER_SEC;
+ event = q->now + NSEC_PER_SEC;
q->near_ev_cache[level] = event;
} else
event = q->near_ev_cache[level];
@@ -903,10 +936,17 @@ ok:
}
}
sch->qstats.overlimits++;
- if (likely(next_event > q->now))
- qdisc_watchdog_schedule(&q->watchdog, next_event);
- else
+ if (likely(next_event > q->now)) {
+ if (!test_bit(__QDISC_STATE_DEACTIVATED,
+ &qdisc_root_sleeping(q->watchdog.qdisc)->state)) {
+ ktime_t time = ns_to_ktime(next_event);
+ qdisc_throttled(q->watchdog.qdisc);
+ hrtimer_start(&q->watchdog.timer, time,
+ HRTIMER_MODE_ABS);
+ }
+ } else {
schedule_work(&q->work);
+ }
fin:
return skb;
}
@@ -1049,8 +1089,7 @@ static int htb_dump(struct Qdisc *sch, struct sk_buff *skb)
nest = nla_nest_start(skb, TCA_OPTIONS);
if (nest == NULL)
goto nla_put_failure;
- if (nla_put(skb, TCA_HTB_INIT, sizeof(gopt), &gopt))
- goto nla_put_failure;
+ NLA_PUT(skb, TCA_HTB_INIT, sizeof(gopt), &gopt);
nla_nest_end(skb, nest);
spin_unlock_bh(root_lock);
@@ -1082,15 +1121,12 @@ static int htb_dump_class(struct Qdisc *sch, unsigned long arg,
memset(&opt, 0, sizeof(opt));
- opt.rate = cl->rate->rate;
opt.buffer = cl->buffer;
- opt.ceil = cl->ceil->rate;
opt.cbuffer = cl->cbuffer;
opt.quantum = cl->quantum;
opt.prio = cl->prio;
opt.level = cl->level;
- if (nla_put(skb, TCA_HTB_PARMS, sizeof(opt), &opt))
- goto nla_put_failure;
+ NLA_PUT(skb, TCA_HTB_PARMS, sizeof(opt), &opt);
nla_nest_end(skb, nest);
spin_unlock_bh(root_lock);
@@ -1203,9 +1239,6 @@ static void htb_destroy_class(struct Qdisc *sch, struct htb_class *cl)
qdisc_destroy(cl->un.leaf.q);
}
gen_kill_estimator(&cl->bstats, &cl->rate_est);
- qdisc_put_rtab(cl->rate);
- qdisc_put_rtab(cl->ceil);
-
tcf_destroy_chain(&cl->filter_list);
kfree(cl);
}
@@ -1460,12 +1493,16 @@ static int htb_change_class(struct Qdisc *sch, u32 classid,
cl->buffer = hopt->buffer;
cl->cbuffer = hopt->cbuffer;
- if (cl->rate)
- qdisc_put_rtab(cl->rate);
- cl->rate = rtab;
- if (cl->ceil)
- qdisc_put_rtab(cl->ceil);
- cl->ceil = ctab;
+
+ cl->rate.rate_bps = (u64)rtab->rate.rate << 3;
+ cl->ceil.rate_bps = (u64)ctab->rate.rate << 3;
+
+ htb_precompute_ratedata(&cl->rate);
+ htb_precompute_ratedata(&cl->ceil);
+
+ cl->buffer = hopt->buffer << PSCHED_SHIFT;
+ cl->cbuffer = hopt->buffer << PSCHED_SHIFT;
+
sch_tree_unlock(sch);
qdisc_class_hash_grow(sch, &q->clhash);
--
1.7.5.3
^ permalink raw reply related
* Re: [PATCH] htb: improved accuracy at high rates
From: Vimal @ 2012-10-20 16:41 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Rick Jones, davem, Jamal Hadi Salim, netdev
In-Reply-To: <1350717994.13333.298.camel@edumazet-glaptop>
On 20 October 2012 00:26, Eric Dumazet <eric.dumazet@gmail.com> wrote:
>
> UFO is quite a virtual feature for the moment ;)
>
> To my knowledge only one driver uses it (neterion/s2io.c)
>
> So most probably HTB sees the fragments built by the stack before
> queueing packets to the qdisc layer.
>
> tcpdump can be used to check this.
Ah you're right. When I checked tcpdump output earlier, I checked the
IP length which was 8000 bytes, so I presumed UFO was on. When I
looked carefully again, the bytes on the wire were MTU sized.
Thanks for the suggestions! I just sent the next version of the patch
against net-next, and provided a simpler example that illustrates the
inaccuracy.
--
Vimal
^ permalink raw reply
* Re: [PATCH net-next] packet mmap : allow the user to choose tx data offset.
From: pchavent @ 2012-10-20 17:13 UTC (permalink / raw)
To: Daniel Borkmann; +Cc: davem, netdev
In-Reply-To: <CAD6jFURS6FXvZJttxEn-qPP5=sWCOHSx=BeSgiJM+ZemAkuKdQ@mail.gmail.com>
Hi Daniel
>>>> + switch (po->tp_version) {
>>>> + case TPACKET_V2:
>>>> + off = ph.h2->tp_mac;
>>>> + break;
>>>> + default:
>>>> + off = ph.h1->tp_mac;
>>>
>>>
>>> TPACKET_V1 as default is wrong since there's also TPACKET_V3. What
>>> about TPACKET_V3 in general in your patch? You simply ignore it.
>>
>>
>> I have reproduced the pattern found some lines above (1868-1875)
>>
>> switch (po->tp_version) {
>> case TPACKET_V2:
>> tp_len = ph.h2->tp_len;
>> break;
>> default:
>> tp_len = ph.h1->tp_len;
>> break;
>> }
>>
>> Do you suggest me to add a case TPACKET_V3 and get tp_len from
>> ph.h3->tp_len
>> ?
>
> If I'm not mistaken, with your current patch you can do the
> setsockopt(2) call on a tpacket v3 socket and then default to a
> tpacket v1 in your switch statement. This doesn't seem right. Thus,
> this should better be fixed in whatever way.
It seems that "Opening a Tx-ring is NOT supported in TPACKET_V3" (see
comment in packet_set_ring). So no one could be able to use the TX_OFF
feature with a TPACKET_V3.
The tpacket_snd function call some subroutines that call WARN and BUG
(like in __packet_set_status),
and other ones that simply ignore the tpacket v3 case (like in
tpacket_fill_skb).
What do you suggest ?
I add a case that call WARN and BUG like in {__packet_set_status,
__packet_get_status, packet_increment_rx_head} where we get tp_len in
the tpacket_fill_skb function ?
Or i let the code as is ?
Regards.
Paul.
^ permalink raw reply
* Re: [PATCH net-next] packet mmap : allow the user to choose tx data offset.
From: Daniel Borkmann @ 2012-10-20 17:35 UTC (permalink / raw)
To: pchavent; +Cc: davem, netdev
In-Reply-To: <535a652cffb4d52cbae20e06a95d724d@sybille.onecert.fr>
On Sat, Oct 20, 2012 at 7:13 PM, pchavent <Paul.Chavent@onera.fr> wrote:
> Hi Daniel
>
>
>>>>> + switch (po->tp_version) {
>>>>> + case TPACKET_V2:
>>>>> + off = ph.h2->tp_mac;
>>>>> + break;
>>>>> + default:
>>>>> + off = ph.h1->tp_mac;
>>>>
>>>>
>>>>
>>>> TPACKET_V1 as default is wrong since there's also TPACKET_V3. What
>>>> about TPACKET_V3 in general in your patch? You simply ignore it.
>>>
>>>
>>>
>>> I have reproduced the pattern found some lines above (1868-1875)
>>>
>>> switch (po->tp_version) {
>>> case TPACKET_V2:
>>> tp_len = ph.h2->tp_len;
>>> break;
>>> default:
>>> tp_len = ph.h1->tp_len;
>>> break;
>>> }
>>>
>>> Do you suggest me to add a case TPACKET_V3 and get tp_len from
>>> ph.h3->tp_len
>>> ?
>>
>>
>> If I'm not mistaken, with your current patch you can do the
>> setsockopt(2) call on a tpacket v3 socket and then default to a
>> tpacket v1 in your switch statement. This doesn't seem right. Thus,
>> this should better be fixed in whatever way.
>
>
> It seems that "Opening a Tx-ring is NOT supported in TPACKET_V3" (see
> comment in packet_set_ring). So no one could be able to use the TX_OFF
> feature with a TPACKET_V3.
>
> The tpacket_snd function call some subroutines that call WARN and BUG (like
> in __packet_set_status),
> and other ones that simply ignore the tpacket v3 case (like in
> tpacket_fill_skb).
>
> What do you suggest ?
>
> I add a case that call WARN and BUG like in {__packet_set_status,
> __packet_get_status, packet_increment_rx_head} where we get tp_len in the
> tpacket_fill_skb function ?
>
> Or i let the code as is ?
Hmm, ok. Then maybe for the moment having it similar like the rest
seems fine. So, if one day someone adds a TX_RING to TPACKET_V3, this
needs to be addressed as well. Probably letting the code as is might
be okay for now.
^ permalink raw reply
* Re: Re: A reliable kernel panic (3.6.2) and system crash when visiting a particular website
From: Pavel Machek @ 2012-10-20 20:32 UTC (permalink / raw)
To: Artem S. Tashkinov; +Cc: bp, linux-kernel, netdev, security
In-Reply-To: <966148591.30347.1350754909449.JavaMail.mail@webmail08>
On Sat 2012-10-20 17:41:49, Artem S. Tashkinov wrote:
> On Oct 20, 2012, Borislav Petkov wrote:
>
> > Yeah, your kernel is tainted with a proprietary module (vbox*, etc). Can
> > you reproduce your corruptions (this is what it looks like) without that
> > module?
>
> Yes, I can reproduce this panic with zero proprietary/non-free modules loaded.
>
> The problem is the kernel doesn't even print a kernel panic - the
> system just freezes completely - cursor in a text console stops
> blinking.
bugtraq? :-).
If remote website can crash your Linux, that's quite significant news.
(Cc-ed netdev@ and security@ ... this may be important).
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
^ permalink raw reply
* Re: [net-next 01/13] ixgbe: Add support for IPv6 and UDP to ixgbe_get_headlen
From: Alexander Duyck @ 2012-10-20 20:39 UTC (permalink / raw)
To: Eric Dumazet
Cc: Jeff Kirsher, davem, Alexander Duyck, netdev, gospo, sassmann
In-Reply-To: <1350720070.13333.359.camel@edumazet-glaptop>
On 10/20/2012 1:01 AM, Eric Dumazet wrote:
> On Fri, 2012-10-19 at 23:25 -0700, Jeff Kirsher wrote:
>> From: Alexander Duyck <alexander.h.duyck@intel.com>
>>
>> This change adds support for IPv6 and UDP to ixgbe_get_headlen. The
>> advantage to this is that we can now handle ipv4/UDP, ipv6/TCP, and
>> ipv6/UDP with a single memcpy instead of having to do them in multiple
>> pskb_may_pull calls.
>>
>> A quick bit of testing shows that we increase throughput for a single
>> session of netperf from 8800Mpbs to about 9300Mpbs in the case of ipv6/TCP.
>> As such overall ipv6 performance should improve with this change.
>>
>> Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
>> Tested-by: Stephen Ko <stephen.s.ko@intel.com>
>> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
>> ---
>> drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 15 ++++++++++++++-
>> 1 file changed, 14 insertions(+), 1 deletion(-)
> Thats pretty interesting.
>
> Why this work is not done generically, so that all drivers can have
> benefit of it ?
>
> Even a driver not using fragments could have benefit, because we could
> prefetch() so that GRO stack is faster (no stall to get the last 2 bytes
> of a TCP header when timestamps are used)
>
> Thanks
The ixgbe_get_headlen function originally was written to handle the case
where we were doing the hardware LRO/RSC feature but we weren't setting
gso_size. We were using it to get the size of the header so we could
determine the average payload size of the frame. When we switched over
to page based receive the function came along with it and had actually
prevented a performance regression for IPv4/TCP which this patch is
fixing for the other supported protocols. The main idea of this code is
now to emulate what our hardware was doing in determining the length of
the header when we had header-data split enabled. I hadn't really
thought of any way it could be used in a generic sense. The function is
essentially is built around several assumptions that only apply to the
igb/ixgbe case such as the fact that the entire block being linear and
the limitation of the protocols so that they match up with what we used
to do for header-data split.
That being said, if there is a way for you to use it in the generic case
feel free to copy the ixgbe version, rename it, and export it for use.
The key bit for me is that I just need to be able to get the header
length before calling eth_type_trans as I only want to do one memcpy for
the entire header.
Thanks,
Alex
^ 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