* [PATCH v2 net-next 06/10] net: hns3: Add some interface for the support of DCB feature
From: Yunsheng Lin @ 2017-09-26 2:25 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1506392718-50463-1-git-send-email-linyunsheng@huawei.com>
This patch add some interface and export some interface from
hclge_tm and hclgc_main to support the upcoming DCB feature.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 3 +-
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 3 ++
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 48 ++++++++++++++++++++--
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h | 6 +++
4 files changed, 55 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index b345070..ca7e3bc 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -30,7 +30,6 @@
#define HCLGE_64BIT_STATS_FIELD_OFF(f) (offsetof(struct hclge_64_bit_stats, f))
#define HCLGE_32BIT_STATS_FIELD_OFF(f) (offsetof(struct hclge_32_bit_stats, f))
-static int hclge_rss_init_hw(struct hclge_dev *hdev);
static int hclge_set_mta_filter_mode(struct hclge_dev *hdev,
enum hclge_mta_dmac_sel_type mta_mac_sel,
bool enable);
@@ -2655,7 +2654,7 @@ static int hclge_get_tc_size(struct hnae3_handle *handle)
return hdev->rss_size_max;
}
-static int hclge_rss_init_hw(struct hclge_dev *hdev)
+int hclge_rss_init_hw(struct hclge_dev *hdev)
{
const u8 hfunc = HCLGE_RSS_HASH_ALGO_TOEPLITZ;
struct hclge_vport *vport = hdev->vport;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
index 4fc36f0..394b587 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
@@ -515,4 +515,7 @@ static inline int hclge_get_queue_id(struct hnae3_queue *queue)
int hclge_cfg_mac_speed_dup(struct hclge_dev *hdev, int speed, u8 duplex);
int hclge_set_vf_vlan_common(struct hclge_dev *vport, int vfid,
bool is_kill, u16 vlan, u8 qos, __be16 proto);
+
+int hclge_buffer_alloc(struct hclge_dev *hdev);
+int hclge_rss_init_hw(struct hclge_dev *hdev);
#endif
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
index ea94d23..8295684 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
@@ -883,10 +883,14 @@ static int hclge_tm_pri_dwrr_cfg(struct hclge_dev *hdev)
return 0;
}
-static int hclge_tm_map_cfg(struct hclge_dev *hdev)
+int hclge_tm_map_cfg(struct hclge_dev *hdev)
{
int ret;
+ ret = hclge_up_to_tc_map(hdev);
+ if (ret)
+ return ret;
+
ret = hclge_tm_pg_to_pri_map(hdev);
if (ret)
return ret;
@@ -994,7 +998,7 @@ static int hclge_tm_lvl34_schd_mode_cfg(struct hclge_dev *hdev)
return 0;
}
-static int hclge_tm_schd_mode_hw(struct hclge_dev *hdev)
+int hclge_tm_schd_mode_hw(struct hclge_dev *hdev)
{
int ret;
@@ -1092,7 +1096,45 @@ int hclge_pause_setup_hw(struct hclge_dev *hdev)
return ret;
}
- return hclge_up_to_tc_map(hdev);
+ return 0;
+}
+
+int hclge_tm_prio_tc_info_update(struct hclge_dev *hdev, u8 *prio_tc)
+{
+ struct hclge_vport *vport = hdev->vport;
+ struct hnae3_knic_private_info *kinfo;
+ u32 i, k;
+
+ for (i = 0; i < HNAE3_MAX_USER_PRIO; i++) {
+ if (prio_tc[i] >= hdev->tm_info.num_tc)
+ return -EINVAL;
+ hdev->tm_info.prio_tc[i] = prio_tc[i];
+
+ for (k = 0; k < hdev->num_alloc_vport; k++) {
+ kinfo = &vport[k].nic.kinfo;
+ kinfo->prio_tc[i] = prio_tc[i];
+ }
+ }
+ return 0;
+}
+
+void hclge_tm_schd_info_update(struct hclge_dev *hdev, u8 num_tc)
+{
+ u8 i, bit_map = 0;
+
+ hdev->tm_info.num_tc = num_tc;
+
+ for (i = 0; i < hdev->tm_info.num_tc; i++)
+ bit_map |= BIT(i);
+
+ if (!bit_map) {
+ bit_map = 1;
+ hdev->tm_info.num_tc = 1;
+ }
+
+ hdev->hw_tc_map = bit_map;
+
+ hclge_tm_schd_info_init(hdev);
}
int hclge_tm_init_hw(struct hclge_dev *hdev)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
index 19a01e4..bf59961 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
@@ -112,4 +112,10 @@ struct hclge_port_shapping_cmd {
int hclge_tm_schd_init(struct hclge_dev *hdev);
int hclge_pause_setup_hw(struct hclge_dev *hdev);
+int hclge_tm_schd_mode_hw(struct hclge_dev *hdev);
+int hclge_tm_prio_tc_info_update(struct hclge_dev *hdev, u8 *prio_tc);
+void hclge_tm_schd_info_update(struct hclge_dev *hdev, u8 num_tc);
+int hclge_tm_dwrr_cfg(struct hclge_dev *hdev);
+int hclge_tm_map_cfg(struct hclge_dev *hdev);
+int hclge_tm_init_hw(struct hclge_dev *hdev);
#endif
--
1.9.1
^ permalink raw reply related
* [PATCH v2 net-next 05/10] net: hns3: Add tc-based TM support for sriov enabled port
From: Yunsheng Lin @ 2017-09-26 2:25 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1506392718-50463-1-git-send-email-linyunsheng@huawei.com>
When sriov is enabled and TM is in tc-based mode, vf's TM
parameters is not set in TM initialization process.
This patch add the tc_based TM support for sriov enabled
using the information in vport struct.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 49 ++++++++++++++--------
1 file changed, 31 insertions(+), 18 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
index f79cebd..ea94d23 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
@@ -388,13 +388,13 @@ static int hclge_tm_pri_schd_mode_cfg(struct hclge_dev *hdev, u8 pri_id)
return hclge_cmd_send(&hdev->hw, &desc, 1);
}
-static int hclge_tm_qs_schd_mode_cfg(struct hclge_dev *hdev, u16 qs_id)
+static int hclge_tm_qs_schd_mode_cfg(struct hclge_dev *hdev, u16 qs_id, u8 mode)
{
struct hclge_desc desc;
hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_QS_SCH_MODE_CFG, false);
- if (hdev->tm_info.tc_info[qs_id].tc_sch_mode == HCLGE_SCH_MODE_DWRR)
+ if (mode == HCLGE_SCH_MODE_DWRR)
desc.data[1] = cpu_to_le32(HCLGE_TM_TX_SCHD_DWRR_MSK);
else
desc.data[1] = 0;
@@ -638,17 +638,18 @@ static int hclge_tm_pri_q_qs_cfg(struct hclge_dev *hdev)
{
struct hclge_vport *vport = hdev->vport;
int ret;
- u32 i;
+ u32 i, k;
if (hdev->tx_sch_mode == HCLGE_FLAG_TC_BASE_SCH_MODE) {
/* Cfg qs -> pri mapping, one by one mapping */
- for (i = 0; i < hdev->tm_info.num_tc; i++) {
- ret = hclge_tm_qs_to_pri_map_cfg(hdev, i, i);
- if (ret)
- return ret;
- }
+ for (k = 0; k < hdev->num_alloc_vport; k++)
+ for (i = 0; i < hdev->tm_info.num_tc; i++) {
+ ret = hclge_tm_qs_to_pri_map_cfg(
+ hdev, vport[k].qs_offset + i, i);
+ if (ret)
+ return ret;
+ }
} else if (hdev->tx_sch_mode == HCLGE_FLAG_VNET_BASE_SCH_MODE) {
- int k;
/* Cfg qs -> pri mapping, qs = tc, pri = vf, 8 qs -> 1 pri */
for (k = 0; k < hdev->num_alloc_vport; k++)
for (i = 0; i < HNAE3_MAX_TC; i++) {
@@ -797,10 +798,11 @@ static int hclge_tm_pri_shaper_cfg(struct hclge_dev *hdev)
static int hclge_tm_pri_tc_base_dwrr_cfg(struct hclge_dev *hdev)
{
+ struct hclge_vport *vport = hdev->vport;
struct hclge_pg_info *pg_info;
u8 dwrr;
int ret;
- u32 i;
+ u32 i, k;
for (i = 0; i < hdev->tm_info.num_tc; i++) {
pg_info =
@@ -811,9 +813,13 @@ static int hclge_tm_pri_tc_base_dwrr_cfg(struct hclge_dev *hdev)
if (ret)
return ret;
- ret = hclge_tm_qs_weight_cfg(hdev, i, dwrr);
- if (ret)
- return ret;
+ for (k = 0; k < hdev->num_alloc_vport; k++) {
+ ret = hclge_tm_qs_weight_cfg(
+ hdev, vport[k].qs_offset + i,
+ vport[k].dwrr);
+ if (ret)
+ return ret;
+ }
}
return 0;
@@ -944,7 +950,10 @@ static int hclge_tm_schd_mode_vnet_base_cfg(struct hclge_vport *vport)
return ret;
for (i = 0; i < kinfo->num_tc; i++) {
- ret = hclge_tm_qs_schd_mode_cfg(hdev, vport->qs_offset + i);
+ u8 sch_mode = hdev->tm_info.tc_info[i].tc_sch_mode;
+
+ ret = hclge_tm_qs_schd_mode_cfg(hdev, vport->qs_offset + i,
+ sch_mode);
if (ret)
return ret;
}
@@ -956,7 +965,7 @@ static int hclge_tm_lvl34_schd_mode_cfg(struct hclge_dev *hdev)
{
struct hclge_vport *vport = hdev->vport;
int ret;
- u8 i;
+ u8 i, k;
if (hdev->tx_sch_mode == HCLGE_FLAG_TC_BASE_SCH_MODE) {
for (i = 0; i < hdev->tm_info.num_tc; i++) {
@@ -964,9 +973,13 @@ static int hclge_tm_lvl34_schd_mode_cfg(struct hclge_dev *hdev)
if (ret)
return ret;
- ret = hclge_tm_qs_schd_mode_cfg(hdev, i);
- if (ret)
- return ret;
+ for (k = 0; k < hdev->num_alloc_vport; k++) {
+ ret = hclge_tm_qs_schd_mode_cfg(
+ hdev, vport[k].qs_offset + i,
+ HCLGE_SCH_MODE_DWRR);
+ if (ret)
+ return ret;
+ }
}
} else {
for (i = 0; i < hdev->num_alloc_vport; i++) {
--
1.9.1
^ permalink raw reply related
* [PATCH v2 net-next 03/10] net: hns3: Add support for PFC setting in TM module
From: Yunsheng Lin @ 2017-09-26 2:25 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1506392718-50463-1-git-send-email-linyunsheng@huawei.com>
This patch add a pfc_pause_en cmd, and use it to configure
PFC option according to fc_mode in hdev->tm_info.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 68 ++++++++++++++++++++--
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h | 5 ++
2 files changed, 68 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
index 73a75d7..0b4b5d9 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
@@ -124,6 +124,20 @@ static int hclge_mac_pause_en_cfg(struct hclge_dev *hdev, bool tx, bool rx)
return hclge_cmd_send(&hdev->hw, &desc, 1);
}
+static int hclge_pfc_pause_en_cfg(struct hclge_dev *hdev, u8 tx_rx_bitmap,
+ u8 pfc_bitmap)
+{
+ struct hclge_desc desc;
+ struct hclge_pfc_en_cmd *pfc = (struct hclge_pfc_en_cmd *)&desc.data;
+
+ hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_CFG_PFC_PAUSE_EN, false);
+
+ pfc->tx_rx_en_bitmap = tx_rx_bitmap;
+ pfc->pri_en_bitmap = pfc_bitmap;
+
+ return hclge_cmd_send(&hdev->hw, &desc, 1);
+}
+
static int hclge_fill_pri_array(struct hclge_dev *hdev, u8 *pri, u8 pri_id)
{
u8 tc;
@@ -969,20 +983,64 @@ static int hclge_tm_schd_setup_hw(struct hclge_dev *hdev)
return hclge_tm_schd_mode_hw(hdev);
}
+static int hclge_pfc_setup_hw(struct hclge_dev *hdev)
+{
+ u8 enable_bitmap = 0;
+
+ if (hdev->tm_info.fc_mode == HCLGE_FC_PFC)
+ enable_bitmap = HCLGE_TX_MAC_PAUSE_EN_MSK |
+ HCLGE_RX_MAC_PAUSE_EN_MSK;
+
+ return hclge_pfc_pause_en_cfg(hdev, enable_bitmap,
+ hdev->tm_info.hw_pfc_map);
+}
+
+static int hclge_mac_pause_setup_hw(struct hclge_dev *hdev)
+{
+ bool tx_en, rx_en;
+
+ switch (hdev->tm_info.fc_mode) {
+ case HCLGE_FC_NONE:
+ tx_en = false;
+ rx_en = false;
+ break;
+ case HCLGE_FC_RX_PAUSE:
+ tx_en = false;
+ rx_en = true;
+ break;
+ case HCLGE_FC_TX_PAUSE:
+ tx_en = true;
+ rx_en = false;
+ break;
+ case HCLGE_FC_FULL:
+ tx_en = true;
+ rx_en = true;
+ break;
+ default:
+ tx_en = true;
+ rx_en = true;
+ }
+
+ return hclge_mac_pause_en_cfg(hdev, tx_en, rx_en);
+}
+
int hclge_pause_setup_hw(struct hclge_dev *hdev)
{
- bool en = hdev->tm_info.fc_mode != HCLGE_FC_PFC;
int ret;
u8 i;
- ret = hclge_mac_pause_en_cfg(hdev, en, en);
- if (ret)
- return ret;
+ if (hdev->tm_info.fc_mode != HCLGE_FC_PFC)
+ return hclge_mac_pause_setup_hw(hdev);
- /* Only DCB-supported dev supports qset back pressure setting */
+ /* Only DCB-supported dev supports qset back pressure and pfc cmd */
if (!hnae3_dev_dcb_supported(hdev))
return 0;
+ /* When MAC is GE Mode, hdev does not support pfc setting */
+ ret = hclge_pfc_setup_hw(hdev);
+ if (ret)
+ dev_warn(&hdev->pdev->dev, "set pfc pause failed:%d\n", ret);
+
for (i = 0; i < hdev->tm_info.num_tc; i++) {
ret = hclge_tm_qs_bp_cfg(hdev, i);
if (ret)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
index 85158b0..8ecd83c 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
@@ -94,6 +94,11 @@ struct hclge_bp_to_qs_map_cmd {
u32 rsvd1;
};
+struct hclge_pfc_en_cmd {
+ u8 tx_rx_en_bitmap;
+ u8 pri_en_bitmap;
+};
+
#define hclge_tm_set_field(dest, string, val) \
hnae_set_field((dest), (HCLGE_TM_SHAP_##string##_MSK), \
(HCLGE_TM_SHAP_##string##_LSH), val)
--
1.9.1
^ permalink raw reply related
* [PATCH v2 net-next 02/10] net: hns3: Add support for dynamically buffer reallocation
From: Yunsheng Lin @ 2017-09-26 2:25 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1506392718-50463-1-git-send-email-linyunsheng@huawei.com>
Current buffer allocation can only happen at init, when
doing buffer reallocation after init, care must be taken
care of memory which priv_buf points to.
This patch fixes it by using a dynamic allocated temporary
memory. Because we only do buffer reallocation at init or
when setting up the DCB parameter, and priv_buf is only
used at buffer allocation process, so it is ok to use a
dynamic allocated temporary memory.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h | 5 +
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 150 +++++++++++----------
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 2 -
3 files changed, 87 insertions(+), 70 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
index a81c6cb..6b6d28e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
@@ -322,6 +322,11 @@ struct hclge_shared_buf {
u32 buf_size;
};
+struct hclge_pkt_buf_alloc {
+ struct hclge_priv_buf priv_buf[HCLGE_MAX_TC_NUM];
+ struct hclge_shared_buf s_buf;
+};
+
#define HCLGE_RX_COM_WL_EN_B 15
struct hclge_rx_com_wl_buf {
__le16 high_wl;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 02da3d5..b345070 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1324,7 +1324,8 @@ static int hclge_alloc_vport(struct hclge_dev *hdev)
return 0;
}
-static int hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev)
+static int hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
/* TX buffer size is unit by 128 byte */
#define HCLGE_BUF_SIZE_UNIT_SHIFT 7
@@ -1338,7 +1339,7 @@ static int hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev)
hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TX_BUFF_ALLOC, 0);
for (i = 0; i < HCLGE_TC_NUM; i++) {
- u32 buf_size = hdev->priv_buf[i].tx_buf_size;
+ u32 buf_size = buf_alloc->priv_buf[i].tx_buf_size;
req->tx_pkt_buff[i] =
cpu_to_le16((buf_size >> HCLGE_BUF_SIZE_UNIT_SHIFT) |
@@ -1355,9 +1356,10 @@ static int hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev)
return 0;
}
-static int hclge_tx_buffer_alloc(struct hclge_dev *hdev)
+static int hclge_tx_buffer_alloc(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
- int ret = hclge_cmd_alloc_tx_buff(hdev);
+ int ret = hclge_cmd_alloc_tx_buff(hdev, buf_alloc);
if (ret) {
dev_err(&hdev->pdev->dev,
@@ -1390,13 +1392,14 @@ static int hclge_get_pfc_enalbe_num(struct hclge_dev *hdev)
}
/* Get the number of pfc enabled TCs, which have private buffer */
-static int hclge_get_pfc_priv_num(struct hclge_dev *hdev)
+static int hclge_get_pfc_priv_num(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
struct hclge_priv_buf *priv;
int i, cnt = 0;
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
- priv = &hdev->priv_buf[i];
+ priv = &buf_alloc->priv_buf[i];
if ((hdev->tm_info.hw_pfc_map & BIT(i)) &&
priv->enable)
cnt++;
@@ -1406,13 +1409,14 @@ static int hclge_get_pfc_priv_num(struct hclge_dev *hdev)
}
/* Get the number of pfc disabled TCs, which have private buffer */
-static int hclge_get_no_pfc_priv_num(struct hclge_dev *hdev)
+static int hclge_get_no_pfc_priv_num(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
struct hclge_priv_buf *priv;
int i, cnt = 0;
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
- priv = &hdev->priv_buf[i];
+ priv = &buf_alloc->priv_buf[i];
if (hdev->hw_tc_map & BIT(i) &&
!(hdev->tm_info.hw_pfc_map & BIT(i)) &&
priv->enable)
@@ -1422,31 +1426,33 @@ static int hclge_get_no_pfc_priv_num(struct hclge_dev *hdev)
return cnt;
}
-static u32 hclge_get_rx_priv_buff_alloced(struct hclge_dev *hdev)
+static u32 hclge_get_rx_priv_buff_alloced(struct hclge_pkt_buf_alloc *buf_alloc)
{
struct hclge_priv_buf *priv;
u32 rx_priv = 0;
int i;
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
- priv = &hdev->priv_buf[i];
+ priv = &buf_alloc->priv_buf[i];
if (priv->enable)
rx_priv += priv->buf_size;
}
return rx_priv;
}
-static u32 hclge_get_tx_buff_alloced(struct hclge_dev *hdev)
+static u32 hclge_get_tx_buff_alloced(struct hclge_pkt_buf_alloc *buf_alloc)
{
u32 i, total_tx_size = 0;
for (i = 0; i < HCLGE_MAX_TC_NUM; i++)
- total_tx_size += hdev->priv_buf[i].tx_buf_size;
+ total_tx_size += buf_alloc->priv_buf[i].tx_buf_size;
return total_tx_size;
}
-static bool hclge_is_rx_buf_ok(struct hclge_dev *hdev, u32 rx_all)
+static bool hclge_is_rx_buf_ok(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc,
+ u32 rx_all)
{
u32 shared_buf_min, shared_buf_tc, shared_std;
int tc_num, pfc_enable_num;
@@ -1467,30 +1473,31 @@ static bool hclge_is_rx_buf_ok(struct hclge_dev *hdev, u32 rx_all)
hdev->mps;
shared_std = max_t(u32, shared_buf_min, shared_buf_tc);
- rx_priv = hclge_get_rx_priv_buff_alloced(hdev);
+ rx_priv = hclge_get_rx_priv_buff_alloced(buf_alloc);
if (rx_all <= rx_priv + shared_std)
return false;
shared_buf = rx_all - rx_priv;
- hdev->s_buf.buf_size = shared_buf;
- hdev->s_buf.self.high = shared_buf;
- hdev->s_buf.self.low = 2 * hdev->mps;
+ buf_alloc->s_buf.buf_size = shared_buf;
+ buf_alloc->s_buf.self.high = shared_buf;
+ buf_alloc->s_buf.self.low = 2 * hdev->mps;
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
if ((hdev->hw_tc_map & BIT(i)) &&
(hdev->tm_info.hw_pfc_map & BIT(i))) {
- hdev->s_buf.tc_thrd[i].low = hdev->mps;
- hdev->s_buf.tc_thrd[i].high = 2 * hdev->mps;
+ buf_alloc->s_buf.tc_thrd[i].low = hdev->mps;
+ buf_alloc->s_buf.tc_thrd[i].high = 2 * hdev->mps;
} else {
- hdev->s_buf.tc_thrd[i].low = 0;
- hdev->s_buf.tc_thrd[i].high = hdev->mps;
+ buf_alloc->s_buf.tc_thrd[i].low = 0;
+ buf_alloc->s_buf.tc_thrd[i].high = hdev->mps;
}
}
return true;
}
-static int hclge_tx_buffer_calc(struct hclge_dev *hdev)
+static int hclge_tx_buffer_calc(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
u32 i, total_size;
@@ -1498,7 +1505,7 @@ static int hclge_tx_buffer_calc(struct hclge_dev *hdev)
/* alloc tx buffer for all enabled tc */
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
- struct hclge_priv_buf *priv = &hdev->priv_buf[i];
+ struct hclge_priv_buf *priv = &buf_alloc->priv_buf[i];
if (total_size < HCLGE_DEFAULT_TX_BUF)
return -ENOMEM;
@@ -1516,22 +1523,24 @@ static int hclge_tx_buffer_calc(struct hclge_dev *hdev)
/* hclge_rx_buffer_calc: calculate the rx private buffer size for all TCs
* @hdev: pointer to struct hclge_dev
+ * @buf_alloc: pointer to buffer calculation data
* @return: 0: calculate sucessful, negative: fail
*/
-int hclge_rx_buffer_calc(struct hclge_dev *hdev)
+int hclge_rx_buffer_calc(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
u32 rx_all = hdev->pkt_buf_size;
int no_pfc_priv_num, pfc_priv_num;
struct hclge_priv_buf *priv;
int i;
- rx_all -= hclge_get_tx_buff_alloced(hdev);
+ rx_all -= hclge_get_tx_buff_alloced(buf_alloc);
/* When DCB is not supported, rx private
* buffer is not allocated.
*/
if (!hnae3_dev_dcb_supported(hdev)) {
- if (!hclge_is_rx_buf_ok(hdev, rx_all))
+ if (!hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
return -ENOMEM;
return 0;
@@ -1539,7 +1548,7 @@ int hclge_rx_buffer_calc(struct hclge_dev *hdev)
/* step 1, try to alloc private buffer for all enabled tc */
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
- priv = &hdev->priv_buf[i];
+ priv = &buf_alloc->priv_buf[i];
if (hdev->hw_tc_map & BIT(i)) {
priv->enable = 1;
if (hdev->tm_info.hw_pfc_map & BIT(i)) {
@@ -1560,14 +1569,14 @@ int hclge_rx_buffer_calc(struct hclge_dev *hdev)
}
}
- if (hclge_is_rx_buf_ok(hdev, rx_all))
+ if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
return 0;
/* step 2, try to decrease the buffer size of
* no pfc TC's private buffer
*/
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
- priv = &hdev->priv_buf[i];
+ priv = &buf_alloc->priv_buf[i];
priv->enable = 0;
priv->wl.low = 0;
@@ -1590,18 +1599,18 @@ int hclge_rx_buffer_calc(struct hclge_dev *hdev)
}
}
- if (hclge_is_rx_buf_ok(hdev, rx_all))
+ if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
return 0;
/* step 3, try to reduce the number of pfc disabled TCs,
* which have private buffer
*/
/* get the total no pfc enable TC number, which have private buffer */
- no_pfc_priv_num = hclge_get_no_pfc_priv_num(hdev);
+ no_pfc_priv_num = hclge_get_no_pfc_priv_num(hdev, buf_alloc);
/* let the last to be cleared first */
for (i = HCLGE_MAX_TC_NUM - 1; i >= 0; i--) {
- priv = &hdev->priv_buf[i];
+ priv = &buf_alloc->priv_buf[i];
if (hdev->hw_tc_map & BIT(i) &&
!(hdev->tm_info.hw_pfc_map & BIT(i))) {
@@ -1613,22 +1622,22 @@ int hclge_rx_buffer_calc(struct hclge_dev *hdev)
no_pfc_priv_num--;
}
- if (hclge_is_rx_buf_ok(hdev, rx_all) ||
+ if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all) ||
no_pfc_priv_num == 0)
break;
}
- if (hclge_is_rx_buf_ok(hdev, rx_all))
+ if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
return 0;
/* step 4, try to reduce the number of pfc enabled TCs
* which have private buffer.
*/
- pfc_priv_num = hclge_get_pfc_priv_num(hdev);
+ pfc_priv_num = hclge_get_pfc_priv_num(hdev, buf_alloc);
/* let the last to be cleared first */
for (i = HCLGE_MAX_TC_NUM - 1; i >= 0; i--) {
- priv = &hdev->priv_buf[i];
+ priv = &buf_alloc->priv_buf[i];
if (hdev->hw_tc_map & BIT(i) &&
hdev->tm_info.hw_pfc_map & BIT(i)) {
@@ -1640,17 +1649,18 @@ int hclge_rx_buffer_calc(struct hclge_dev *hdev)
pfc_priv_num--;
}
- if (hclge_is_rx_buf_ok(hdev, rx_all) ||
+ if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all) ||
pfc_priv_num == 0)
break;
}
- if (hclge_is_rx_buf_ok(hdev, rx_all))
+ if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
return 0;
return -ENOMEM;
}
-static int hclge_rx_priv_buf_alloc(struct hclge_dev *hdev)
+static int hclge_rx_priv_buf_alloc(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
struct hclge_rx_priv_buff *req;
struct hclge_desc desc;
@@ -1662,7 +1672,7 @@ static int hclge_rx_priv_buf_alloc(struct hclge_dev *hdev)
/* Alloc private buffer TCs */
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
- struct hclge_priv_buf *priv = &hdev->priv_buf[i];
+ struct hclge_priv_buf *priv = &buf_alloc->priv_buf[i];
req->buf_num[i] =
cpu_to_le16(priv->buf_size >> HCLGE_BUF_UNIT_S);
@@ -1671,7 +1681,7 @@ static int hclge_rx_priv_buf_alloc(struct hclge_dev *hdev)
}
req->shared_buf =
- cpu_to_le16((hdev->s_buf.buf_size >> HCLGE_BUF_UNIT_S) |
+ cpu_to_le16((buf_alloc->s_buf.buf_size >> HCLGE_BUF_UNIT_S) |
(1 << HCLGE_TC0_PRI_BUF_EN_B));
ret = hclge_cmd_send(&hdev->hw, &desc, 1);
@@ -1686,7 +1696,8 @@ static int hclge_rx_priv_buf_alloc(struct hclge_dev *hdev)
#define HCLGE_PRIV_ENABLE(a) ((a) > 0 ? 1 : 0)
-static int hclge_rx_priv_wl_config(struct hclge_dev *hdev)
+static int hclge_rx_priv_wl_config(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
struct hclge_rx_priv_wl_buf *req;
struct hclge_priv_buf *priv;
@@ -1706,7 +1717,9 @@ static int hclge_rx_priv_wl_config(struct hclge_dev *hdev)
desc[i].flag &= ~cpu_to_le16(HCLGE_CMD_FLAG_NEXT);
for (j = 0; j < HCLGE_TC_NUM_ONE_DESC; j++) {
- priv = &hdev->priv_buf[i * HCLGE_TC_NUM_ONE_DESC + j];
+ u32 idx = i * HCLGE_TC_NUM_ONE_DESC + j;
+
+ priv = &buf_alloc->priv_buf[idx];
req->tc_wl[j].high =
cpu_to_le16(priv->wl.high >> HCLGE_BUF_UNIT_S);
req->tc_wl[j].high |=
@@ -1731,9 +1744,10 @@ static int hclge_rx_priv_wl_config(struct hclge_dev *hdev)
return 0;
}
-static int hclge_common_thrd_config(struct hclge_dev *hdev)
+static int hclge_common_thrd_config(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
- struct hclge_shared_buf *s_buf = &hdev->s_buf;
+ struct hclge_shared_buf *s_buf = &buf_alloc->s_buf;
struct hclge_rx_com_thrd *req;
struct hclge_desc desc[2];
struct hclge_tc_thrd *tc;
@@ -1777,9 +1791,10 @@ static int hclge_common_thrd_config(struct hclge_dev *hdev)
return 0;
}
-static int hclge_common_wl_config(struct hclge_dev *hdev)
+static int hclge_common_wl_config(struct hclge_dev *hdev,
+ struct hclge_pkt_buf_alloc *buf_alloc)
{
- struct hclge_shared_buf *buf = &hdev->s_buf;
+ struct hclge_shared_buf *buf = &buf_alloc->s_buf;
struct hclge_rx_com_wl *req;
struct hclge_desc desc;
int ret;
@@ -1809,69 +1824,68 @@ static int hclge_common_wl_config(struct hclge_dev *hdev)
int hclge_buffer_alloc(struct hclge_dev *hdev)
{
+ struct hclge_pkt_buf_alloc *pkt_buf;
int ret;
- hdev->priv_buf = devm_kmalloc_array(&hdev->pdev->dev, HCLGE_MAX_TC_NUM,
- sizeof(struct hclge_priv_buf),
- GFP_KERNEL | __GFP_ZERO);
- if (!hdev->priv_buf)
+ pkt_buf = kzalloc(sizeof(*pkt_buf), GFP_KERNEL);
+ if (!pkt_buf)
return -ENOMEM;
- ret = hclge_tx_buffer_calc(hdev);
+ ret = hclge_tx_buffer_calc(hdev, pkt_buf);
if (ret) {
dev_err(&hdev->pdev->dev,
"could not calc tx buffer size for all TCs %d\n", ret);
- return ret;
+ goto out;
}
- ret = hclge_tx_buffer_alloc(hdev);
+ ret = hclge_tx_buffer_alloc(hdev, pkt_buf);
if (ret) {
dev_err(&hdev->pdev->dev,
"could not alloc tx buffers %d\n", ret);
- return ret;
+ goto out;
}
- ret = hclge_rx_buffer_calc(hdev);
+ ret = hclge_rx_buffer_calc(hdev, pkt_buf);
if (ret) {
dev_err(&hdev->pdev->dev,
"could not calc rx priv buffer size for all TCs %d\n",
ret);
- return ret;
+ goto out;
}
- ret = hclge_rx_priv_buf_alloc(hdev);
+ ret = hclge_rx_priv_buf_alloc(hdev, pkt_buf);
if (ret) {
dev_err(&hdev->pdev->dev, "could not alloc rx priv buffer %d\n",
ret);
- return ret;
+ goto out;
}
if (hnae3_dev_dcb_supported(hdev)) {
- ret = hclge_rx_priv_wl_config(hdev);
+ ret = hclge_rx_priv_wl_config(hdev, pkt_buf);
if (ret) {
dev_err(&hdev->pdev->dev,
"could not configure rx private waterline %d\n",
ret);
- return ret;
+ goto out;
}
- ret = hclge_common_thrd_config(hdev);
+ ret = hclge_common_thrd_config(hdev, pkt_buf);
if (ret) {
dev_err(&hdev->pdev->dev,
"could not configure common threshold %d\n",
ret);
- return ret;
+ goto out;
}
}
- ret = hclge_common_wl_config(hdev);
- if (ret) {
+ ret = hclge_common_wl_config(hdev, pkt_buf);
+ if (ret)
dev_err(&hdev->pdev->dev,
"could not configure common waterline %d\n", ret);
- return ret;
- }
- return 0;
+out:
+ kfree(pkt_buf);
+ return ret;
}
static int hclge_init_roce_base_info(struct hclge_vport *vport)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
index 9fcfd93..4fc36f0 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
@@ -463,8 +463,6 @@ struct hclge_dev {
u32 pkt_buf_size; /* Total pf buf size for tx/rx */
u32 mps; /* Max packet size */
- struct hclge_priv_buf *priv_buf;
- struct hclge_shared_buf s_buf;
enum hclge_mta_dmac_sel_type mta_mac_sel_type;
bool enable_mta; /* Mutilcast filter enable */
--
1.9.1
^ permalink raw reply related
* [PATCH v2 net-next 01/10] net: hns3: Support for dynamically assigning tx buffer to TC
From: Yunsheng Lin @ 2017-09-26 2:25 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1506392718-50463-1-git-send-email-linyunsheng@huawei.com>
This patch add support of dynamically assigning tx buffer to
TC when the TC is enabled.
It will save buffer for rx direction to avoid packet loss.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
V2:
Fix for not defining variables in local loop.
V1:
Initial submit.
---
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h | 1 +
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 64 ++++++++++++++++++----
2 files changed, 55 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
index 758cf39..a81c6cb 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
@@ -311,6 +311,7 @@ struct hclge_tc_thrd {
struct hclge_priv_buf {
struct hclge_waterline wl; /* Waterline for low and high*/
u32 buf_size; /* TC private buffer size */
+ u32 tx_buf_size;
u32 enable; /* Enable TC private buffer or not */
};
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index d27618b..02da3d5 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1324,7 +1324,7 @@ static int hclge_alloc_vport(struct hclge_dev *hdev)
return 0;
}
-static int hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev, u16 buf_size)
+static int hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev)
{
/* TX buffer size is unit by 128 byte */
#define HCLGE_BUF_SIZE_UNIT_SHIFT 7
@@ -1337,10 +1337,13 @@ static int hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev, u16 buf_size)
req = (struct hclge_tx_buff_alloc *)desc.data;
hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TX_BUFF_ALLOC, 0);
- for (i = 0; i < HCLGE_TC_NUM; i++)
+ for (i = 0; i < HCLGE_TC_NUM; i++) {
+ u32 buf_size = hdev->priv_buf[i].tx_buf_size;
+
req->tx_pkt_buff[i] =
cpu_to_le16((buf_size >> HCLGE_BUF_SIZE_UNIT_SHIFT) |
HCLGE_BUF_SIZE_UPDATE_EN_MSK);
+ }
ret = hclge_cmd_send(&hdev->hw, &desc, 1);
if (ret) {
@@ -1352,9 +1355,9 @@ static int hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev, u16 buf_size)
return 0;
}
-static int hclge_tx_buffer_alloc(struct hclge_dev *hdev, u32 buf_size)
+static int hclge_tx_buffer_alloc(struct hclge_dev *hdev)
{
- int ret = hclge_cmd_alloc_tx_buff(hdev, buf_size);
+ int ret = hclge_cmd_alloc_tx_buff(hdev);
if (ret) {
dev_err(&hdev->pdev->dev,
@@ -1433,6 +1436,16 @@ static u32 hclge_get_rx_priv_buff_alloced(struct hclge_dev *hdev)
return rx_priv;
}
+static u32 hclge_get_tx_buff_alloced(struct hclge_dev *hdev)
+{
+ u32 i, total_tx_size = 0;
+
+ for (i = 0; i < HCLGE_MAX_TC_NUM; i++)
+ total_tx_size += hdev->priv_buf[i].tx_buf_size;
+
+ return total_tx_size;
+}
+
static bool hclge_is_rx_buf_ok(struct hclge_dev *hdev, u32 rx_all)
{
u32 shared_buf_min, shared_buf_tc, shared_std;
@@ -1477,18 +1490,43 @@ static bool hclge_is_rx_buf_ok(struct hclge_dev *hdev, u32 rx_all)
return true;
}
+static int hclge_tx_buffer_calc(struct hclge_dev *hdev)
+{
+ u32 i, total_size;
+
+ total_size = hdev->pkt_buf_size;
+
+ /* alloc tx buffer for all enabled tc */
+ for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
+ struct hclge_priv_buf *priv = &hdev->priv_buf[i];
+
+ if (total_size < HCLGE_DEFAULT_TX_BUF)
+ return -ENOMEM;
+
+ if (hdev->hw_tc_map & BIT(i))
+ priv->tx_buf_size = HCLGE_DEFAULT_TX_BUF;
+ else
+ priv->tx_buf_size = 0;
+
+ total_size -= priv->tx_buf_size;
+ }
+
+ return 0;
+}
+
/* hclge_rx_buffer_calc: calculate the rx private buffer size for all TCs
* @hdev: pointer to struct hclge_dev
- * @tx_size: the allocated tx buffer for all TCs
* @return: 0: calculate sucessful, negative: fail
*/
-int hclge_rx_buffer_calc(struct hclge_dev *hdev, u32 tx_size)
+int hclge_rx_buffer_calc(struct hclge_dev *hdev)
{
- u32 rx_all = hdev->pkt_buf_size - tx_size;
+ u32 rx_all = hdev->pkt_buf_size;
int no_pfc_priv_num, pfc_priv_num;
struct hclge_priv_buf *priv;
int i;
+ rx_all -= hclge_get_tx_buff_alloced(hdev);
+
/* When DCB is not supported, rx private
* buffer is not allocated.
*/
@@ -1771,7 +1809,6 @@ static int hclge_common_wl_config(struct hclge_dev *hdev)
int hclge_buffer_alloc(struct hclge_dev *hdev)
{
- u32 tx_buf_size = HCLGE_DEFAULT_TX_BUF;
int ret;
hdev->priv_buf = devm_kmalloc_array(&hdev->pdev->dev, HCLGE_MAX_TC_NUM,
@@ -1780,14 +1817,21 @@ int hclge_buffer_alloc(struct hclge_dev *hdev)
if (!hdev->priv_buf)
return -ENOMEM;
- ret = hclge_tx_buffer_alloc(hdev, tx_buf_size);
+ ret = hclge_tx_buffer_calc(hdev);
+ if (ret) {
+ dev_err(&hdev->pdev->dev,
+ "could not calc tx buffer size for all TCs %d\n", ret);
+ return ret;
+ }
+
+ ret = hclge_tx_buffer_alloc(hdev);
if (ret) {
dev_err(&hdev->pdev->dev,
"could not alloc tx buffers %d\n", ret);
return ret;
}
- ret = hclge_rx_buffer_calc(hdev, tx_buf_size);
+ ret = hclge_rx_buffer_calc(hdev);
if (ret) {
dev_err(&hdev->pdev->dev,
"could not calc rx priv buffer size for all TCs %d\n",
--
1.9.1
^ permalink raw reply related
* [PATCH v2 net-next 00/10] Add support for DCB feature in hns3 driver
From: Yunsheng Lin @ 2017-09-26 2:25 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
The patchset contains some enhancement related to DCB before
adding support for DCB feature.
This patchset depends on the following patchset:
https://patchwork.ozlabs.org/cover/815646/
https://patchwork.ozlabs.org/cover/816145/
High Level Architecture:
[ tc qdisc ] [ lldpad ]
| |
| |
| |
[ hns3_enet ] [ hns3_dcbnl ]
\ /
\ /
\ /
[ hclge_dcb ]
/ \
/ \
/ \
[ hclge_main ] [ hclge_tm ]
Current patch-set support following functionality:
1. Use of tc qdisc to configure the tc num and prio_tc_map.
2. Use of lldptool to configure the tc schedule mode, tc
bandwidth(if schedule mode is ETS), prio_tc_map and
PFC parameter.
---
V2: Fix for not defining variables in local loop.
V1: Initial Submit.
Yunsheng Lin (10):
net: hns3: Support for dynamically assigning tx buffer to TC
net: hns3: Add support for dynamically buffer reallocation
net: hns3: Add support for PFC setting in TM module
net: hns3: Add support for port shaper setting in TM module
net: hns3: Add tc-based TM support for sriov enabled port
net: hns3: Add some interface for the support of DCB feature
net: hns3: Add hclge_dcb module for the support of DCB feature
net: hns3: Add dcb netlink interface for the support of DCB feature
net: hns3: Setting for fc_mode and dcb enable flag in TM module
net: hns3: Add mqprio support when interacting with network stack
drivers/net/ethernet/hisilicon/Kconfig | 9 +
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 20 ++
.../net/ethernet/hisilicon/hns3/hns3pf/Makefile | 4 +
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h | 6 +
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c | 327 +++++++++++++++++++++
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.h | 21 ++
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 214 +++++++++-----
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 8 +-
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 231 +++++++++++++--
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h | 15 +
.../ethernet/hisilicon/hns3/hns3pf/hns3_dcbnl.c | 106 +++++++
.../net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c | 137 +++++++--
.../net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.h | 7 +
13 files changed, 977 insertions(+), 128 deletions(-)
create mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c
create mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.h
create mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_dcbnl.c
--
1.9.1
^ permalink raw reply
* Re: [PATCH net] ipv6: remove incorrect WARN_ON() in fib6_del()
From: Eric Dumazet @ 2017-09-26 2:23 UTC (permalink / raw)
To: Martin KaFai Lau; +Cc: Wei Wang, David Miller, Linux Kernel Network Developers
In-Reply-To: <20170926020720.gnvfz6xch4frf2dy@kafai-mbp>
On Mon, Sep 25, 2017 at 7:07 PM, Martin KaFai Lau <kafai@fb.com> wrote:
> I am probably still missing something.
>
> Considering the del operation should be under the writer lock,
> if rt->rt6i_node should be NULL (for rt that has already been
> removed from fib6), why this WARN_ON() is triggered?
>
> An example may help.
>
Look at the stack trace, you'll find the answers...
ip6_link_failure() -> ip6_del_rt()
Note that rt might have been deleted from the _tree_ already.
^ permalink raw reply
* Re: [PATCH net] ipv6: remove incorrect WARN_ON() in fib6_del()
From: Martin KaFai Lau @ 2017-09-26 2:07 UTC (permalink / raw)
To: Wei Wang; +Cc: David Miller, Linux Kernel Network Developers, Eric Dumazet
In-Reply-To: <CAEA6p_CvtmNaYcgSSyFNEbip1ckKwLjJH3PJTOqym0hCt-x2uw@mail.gmail.com>
On Tue, Sep 26, 2017 at 01:16:05AM +0000, Wei Wang wrote:
> On Mon, Sep 25, 2017 at 5:56 PM, Martin KaFai Lau <kafai@fb.com> wrote:
> > On Mon, Sep 25, 2017 at 05:35:22PM +0000, Wei Wang wrote:
> >> From: Wei Wang <weiwan@google.com>
> >>
> >> fib6_del() generates WARN_ON() when rt->dst.obsolete > 0. This does not
> >> make sense because it is possible that the route passed in is already
> >> deleted by some other thread and rt->dst.obsolete is set to
> >> DST_OBSOLETE_DEAD.
> >> So this commit deletes this WARN_ON() and also remove the
> >> "#ifdef RT6_DEBUG >= 2" condition so that if the route is already
> >> obsolete, we return right at the beginning of fib6_del().
> >>
> >>
> >> diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
> >> index e5308d7cbd75..693bcd7ef6d2 100644
> >> --- a/net/ipv6/ip6_fib.c
> >> +++ b/net/ipv6/ip6_fib.c
> >> @@ -1592,13 +1592,7 @@ int fib6_del(struct rt6_info *rt, struct nl_info *info)
> >> struct net *net = info->nl_net;
> >> struct rt6_info **rtp;
> >>
> >> -#if RT6_DEBUG >= 2
> >> - if (rt->dst.obsolete > 0) {
> >> - WARN_ON(fn);
> > fn should have already been set to NULL if it is removed
> > from the fib6 tree?
> >
>
> That is true. rt->rt6i_node (fn) should already be marked as NULL.
I am probably still missing something.
Considering the del operation should be under the writer lock,
if rt->rt6i_node should be NULL (for rt that has already been
removed from fib6), why this WARN_ON() is triggered?
An example may help.
> That means the check on rt->dst.obsolete is redundant.
> I will remove it in v2.
> Thanks Martin.
>
>
> >> - return -ENOENT;
> >> - }
> >> -#endif
> >> - if (!fn || rt == net->ipv6.ip6_null_entry)
> >> + if (!fn || rt->dst.obsolete > 0 || rt == net->ipv6.ip6_null_entry)
> >> return -ENOENT;
> >>
> >> WARN_ON(!(fn->fn_flags & RTN_RTINFO));
> >> --
> >> 2.14.1.821.g8fa685d3b7-goog
> >>
^ permalink raw reply
* Re: [PATCH net-next] bpf/verifier: improve disassembly of BPF_END instructions
From: Alexei Starovoitov @ 2017-09-26 1:33 UTC (permalink / raw)
To: Daniel Borkmann
Cc: Y Song, Edward Cree, David Miller, netdev, Jiong Wang,
Jakub Kicinski
In-Reply-To: <59C978A2.6070405@iogearbox.net>
On Mon, Sep 25, 2017 at 11:44:02PM +0200, Daniel Borkmann wrote:
> On 09/24/2017 07:50 AM, Alexei Starovoitov wrote:
> > On Fri, Sep 22, 2017 at 09:49:10PM -0700, Y Song wrote:
> > > On Fri, Sep 22, 2017 at 9:23 AM, Edward Cree <ecree@solarflare.com> wrote:
> > > > On 22/09/17 16:16, Alexei Starovoitov wrote:
> > > > > looks like we're converging on
> > > > > "be16/be32/be64/le16/le32/le64 #register" for BPF_END.
> > > > > I guess it can live with that. I would prefer more C like syntax
> > > > > to match the rest, but llvm parsing point is a strong one.
> > > > Yep, agreed. I'll post a v2 once we've settled BPF_NEG.
> > > > > For BPG_NEG I prefer to do it in C syntax like interpreter does:
> > > > > ALU_NEG:
> > > > > DST = (u32) -DST;
> > > > > ALU64_NEG:
> > > > > DST = -DST;
> > > > > Yonghong, does it mean that asmparser will equally suffer?
> > > > Correction to my earlier statements: verifier will currently disassemble
> > > > neg as:
> > > > (87) r0 neg 0
> > > > (84) (u32) r0 neg (u32) 0
> > > > because it pretends 'neg' is a compound-assignment operator like +=.
> > > > The analogy with be16 and friends would be to use
> > > > neg64 r0
> > > > neg32 r0
> > > > whereas the analogy with everything else would be
> > > > r0 = -r0
> > > > r0 = (u32) -r0
> > > > as Alexei says.
> > > > I'm happy to go with Alexei's version if it doesn't cause problems for llvm.
> > >
> > > I got some time to do some prototyping in llvm and it looks like that
> > > I am able to
> > > resolve the issue and we are able to use more C-like syntax. That is:
> > > for bswap:
> > > r1 = (be16) (u16) r1
> > > or
> > > r1 = (be16) r1
> > > or
> > > r1 = be16 r1
> > > for neg:
> > > r0 = -r0
> > > (for 32bit support, llvm may output "w0 = -w0" in the future. But
> > > since it is not
> > > enabled yet, you can continue to output "r0 = (u32) -r0".)
> > >
> > > Not sure which syntax is best for bswap. The "r1 = (be16) (u16) r1" is most
> > > explicit in its intention.
> > >
> > > Attaching my llvm patch as well and cc'ing Jiong and Jakub so they can see my
> > > implementation and the relative discussion here. (In this patch, I did
> > > not implement
> > > bswap for little endian yet.) Maybe they can provide additional comments.
> >
> > This is awesome. In such case I'd like to swing back to the C syntax for bpf_end :)
> > Any of these
> > r1 = (be16) (u16) r1
> > or
> > r1 = (be16) r1
> > or
> > r1 = be16 r1
> > are better than just
> > be16 r1
> > I like 1st the most, since it's explicit in terms of what happens with upper bits,
> > but 2nd is also ok. 3rd is not quite C-like.
>
> But above cast to be16 also doesn't seem quite C-like in terms
> of what we're actually doing... 3rd option would be my personal
> preference even if it doesn't look C-like, but otoh we also have
> 'call' etc which is neither.
well, we have not quite C, but C-like syntax already
(u32) r0 += (u32) r1;
Clearly first cast isn't quite C, but it shows that src reg is only
using 32-bit in the alu and imo that's better than using
w0 += w1; syntax. It's not quite C, but it's intent is way more clear
for folks that know C and don't know assembler than 'w0 = w1'.
And imo it's better than r0 = (u32) r0 + (u32) r1;
that why I did this format long ago and not because of laziness.
The 'endian' part, have to confess, was me being lazy,
but I'd rather fix it to match the rest of the 'not quite C, but C-like' style.
In that sense (be16) cast is pretty much self explanatory.
So I'd like to continue bikesheding in hopes to convince you
to accept either 1 or 2 above ;)
^ permalink raw reply
* Re: [PATCH v4 2/3] ipv4: Namespaceify tcp_fastopen_key knob
From: 严海双 @ 2017-09-26 1:25 UTC (permalink / raw)
To: David Miller; +Cc: kuznet, edumazet, weiwan, lucab, netdev, linux-kernel
In-Reply-To: <20170925.162445.234890912211240693.davem@davemloft.net>
> On 2017年9月26日, at 上午7:24, David Miller <davem@davemloft.net> wrote:
>
> From: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>
> Date: Fri, 22 Sep 2017 21:48:43 +0800
>
>> @@ -9,13 +9,18 @@
>> #include <net/inetpeer.h>
>> #include <net/tcp.h>
>>
>> -struct tcp_fastopen_context __rcu *tcp_fastopen_ctx;
>> -
>> -static DEFINE_SPINLOCK(tcp_fastopen_ctx_lock);
>> -
>> -void tcp_fastopen_init_key_once(bool publish)
>> +void tcp_fastopen_init_key_once(struct net *net)
>
> Why did you remove the 'publish' logic from this function?
>
I think this logic is not necessary now, in proc_tcp_fastopen_key, I have removed
tcp_fastopen_init_key_once(false) where the ‘publish’ is false:
- /* Generate a dummy secret but don't publish it. This
- * is needed so we don't regenerate a new key on the
- * first invocation of tcp_fastopen_cookie_gen
- */
- tcp_fastopen_init_key_once(false);
- tcp_fastopen_reset_cipher(user_key, TCP_FASTOPEN_KEY_LENGTH);
+ tcp_fastopen_reset_cipher(net, user_key, TCP_FASTOPEN_KEY_LENGTH);
It said we don't regenerate a new key on first invocation of tcp_fastopen_cookie_gen,
but in tcp_fastopen_cookie_gen，it didn’t call tcp_fastopen_init_key_once since
from commit dfea2aa654243 (tcp: Do not call tcp_fastopen_reset_cipher from interrupt context)：
And in other places where call tcp_fastopen_init_key_once, the ‘publish’ is always true:
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -222,7 +222,7 @@ int inet_listen(struct socket *sock, int backlog)
(tcp_fastopen & TFO_SERVER_ENABLE) &&
!inet_csk(sk)->icsk_accept_queue.fastopenq.max_qlen) {
fastopen_queue_tune(sk, backlog);
- tcp_fastopen_init_key_once(true);
+ tcp_fastopen_init_key_once(sock_net(sk));
}
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2749,7 +2749,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
case TCP_FASTOPEN:
if (val >= 0 && ((1 << sk->sk_state) & (TCPF_CLOSE |
TCPF_LISTEN))) {
- tcp_fastopen_init_key_once(true);
+ tcp_fastopen_init_key_once(net);
fastopen_queue_tune(sk, val);
} else {
So I deleted ‘publish’ logic to ensure it was always true.
^ permalink raw reply
* Re: [PATCH net] ipv6: remove incorrect WARN_ON() in fib6_del()
From: Wei Wang @ 2017-09-26 1:16 UTC (permalink / raw)
To: Martin KaFai Lau
Cc: David Miller, Linux Kernel Network Developers, Eric Dumazet
In-Reply-To: <20170926005618.z4ymlry7jl47nkjm@kafai-mbp.dhcp.thefacebook.com>
On Mon, Sep 25, 2017 at 5:56 PM, Martin KaFai Lau <kafai@fb.com> wrote:
> On Mon, Sep 25, 2017 at 05:35:22PM +0000, Wei Wang wrote:
>> From: Wei Wang <weiwan@google.com>
>>
>> fib6_del() generates WARN_ON() when rt->dst.obsolete > 0. This does not
>> make sense because it is possible that the route passed in is already
>> deleted by some other thread and rt->dst.obsolete is set to
>> DST_OBSOLETE_DEAD.
>> So this commit deletes this WARN_ON() and also remove the
>> "#ifdef RT6_DEBUG >= 2" condition so that if the route is already
>> obsolete, we return right at the beginning of fib6_del().
>>
>>
>> diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
>> index e5308d7cbd75..693bcd7ef6d2 100644
>> --- a/net/ipv6/ip6_fib.c
>> +++ b/net/ipv6/ip6_fib.c
>> @@ -1592,13 +1592,7 @@ int fib6_del(struct rt6_info *rt, struct nl_info *info)
>> struct net *net = info->nl_net;
>> struct rt6_info **rtp;
>>
>> -#if RT6_DEBUG >= 2
>> - if (rt->dst.obsolete > 0) {
>> - WARN_ON(fn);
> fn should have already been set to NULL if it is removed
> from the fib6 tree?
>
That is true. rt->rt6i_node (fn) should already be marked as NULL.
That means the check on rt->dst.obsolete is redundant.
I will remove it in v2.
Thanks Martin.
>> - return -ENOENT;
>> - }
>> -#endif
>> - if (!fn || rt == net->ipv6.ip6_null_entry)
>> + if (!fn || rt->dst.obsolete > 0 || rt == net->ipv6.ip6_null_entry)
>> return -ENOENT;
>>
>> WARN_ON(!(fn->fn_flags & RTN_RTINFO));
>> --
>> 2.14.1.821.g8fa685d3b7-goog
>>
^ permalink raw reply
* Re: [PATCH] r8152: add Linksys USB3GIGV1 id
From: Grant Grundler @ 2017-09-26 1:12 UTC (permalink / raw)
To: Grant Grundler
Cc: Oliver Neukum, Hayes Wang, David S . Miller, LKML, linux-usb,
netdev
In-Reply-To: <CANEJEGtM+gFC9Ofmp=UmGn5pKys9NRbsC6+ks_VqaLKWkEBS8A@mail.gmail.com>
On Mon, Sep 25, 2017 at 1:17 PM, Grant Grundler <grundler@chromium.org> wrote:
...
> I didn't realize cdc_ether has a blacklist to make sure
> RTL8152|RTL8153 devices are not picked up by cdc_ether. Would you
> prefer I add this device to the blacklist in the same patch?
I've sent a V2 which also updates the blacklist in cdc_ether.
cheers,
grant
^ permalink raw reply
* [PATCH V2] r8152: add Linksys USB3GIGV1 id
From: Grant Grundler @ 2017-09-26 1:09 UTC (permalink / raw)
To: Hayes Wang; +Cc: linux-usb, David S . Miller, LKML, netdev, Grant Grundler
This linksys dongle by default comes up in cdc_ether mode.
This patch allows r8152 to claim the device:
Bus 002 Device 002: ID 13b1:0041 Linksys
Signed-off-by: Grant Grundler <grundler@chromium.org>
---
drivers/net/usb/cdc_ether.c | 8 ++++++++
drivers/net/usb/r8152.c | 2 ++
2 files changed, 10 insertions(+)
V2: add LINKSYS_VENDOR_ID to cdc_ether blacklist
diff --git a/drivers/net/usb/cdc_ether.c b/drivers/net/usb/cdc_ether.c
index 8ab281b478f2..fa5c2e7aff1a 100644
--- a/drivers/net/usb/cdc_ether.c
+++ b/drivers/net/usb/cdc_ether.c
@@ -546,6 +546,7 @@ static const struct driver_info wwan_info = {
#define DELL_VENDOR_ID 0x413C
#define REALTEK_VENDOR_ID 0x0bda
#define SAMSUNG_VENDOR_ID 0x04e8
+#define LINKSYS_VENDOR_ID 0x13b1
#define LENOVO_VENDOR_ID 0x17ef
#define NVIDIA_VENDOR_ID 0x0955
#define HP_VENDOR_ID 0x03f0
@@ -737,6 +738,13 @@ static const struct usb_device_id products[] = {
.driver_info = 0,
},
+/* Linksys USB3GIGV1 Ethernet Adapter */
+{
+ USB_DEVICE_AND_INTERFACE_INFO(LINKSYS_VENDOR_ID, 0x0041, USB_CLASS_COMM,
+ USB_CDC_SUBCLASS_ETHERNET, USB_CDC_PROTO_NONE),
+ .driver_info = 0,
+},
+
/* ThinkPad USB-C Dock (based on Realtek RTL8153) */
{
USB_DEVICE_AND_INTERFACE_INFO(LENOVO_VENDOR_ID, 0x3062, USB_CLASS_COMM,
diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c
index ceb78e2ea4f0..941ece08ba78 100644
--- a/drivers/net/usb/r8152.c
+++ b/drivers/net/usb/r8152.c
@@ -613,6 +613,7 @@ enum rtl8152_flags {
#define VENDOR_ID_MICROSOFT 0x045e
#define VENDOR_ID_SAMSUNG 0x04e8
#define VENDOR_ID_LENOVO 0x17ef
+#define VENDOR_ID_LINKSYS 0x13b1
#define VENDOR_ID_NVIDIA 0x0955
#define MCU_TYPE_PLA 0x0100
@@ -5316,6 +5317,7 @@ static const struct usb_device_id rtl8152_table[] = {
{REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x7205)},
{REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x720c)},
{REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x7214)},
+ {REALTEK_USB_DEVICE(VENDOR_ID_LINKSYS, 0x0041)},
{REALTEK_USB_DEVICE(VENDOR_ID_NVIDIA, 0x09ff)},
{}
};
--
2.14.1.821.g8fa685d3b7-goog
^ permalink raw reply related
* Re: [PATCH net] ipv6: remove incorrect WARN_ON() in fib6_del()
From: Martin KaFai Lau @ 2017-09-26 0:56 UTC (permalink / raw)
To: Wei Wang; +Cc: David Miller, netdev, Eric Dumazet
In-Reply-To: <20170925173522.99892-1-tracywwnj@gmail.com>
On Mon, Sep 25, 2017 at 05:35:22PM +0000, Wei Wang wrote:
> From: Wei Wang <weiwan@google.com>
>
> fib6_del() generates WARN_ON() when rt->dst.obsolete > 0. This does not
> make sense because it is possible that the route passed in is already
> deleted by some other thread and rt->dst.obsolete is set to
> DST_OBSOLETE_DEAD.
> So this commit deletes this WARN_ON() and also remove the
> "#ifdef RT6_DEBUG >= 2" condition so that if the route is already
> obsolete, we return right at the beginning of fib6_del().
>
>
> diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
> index e5308d7cbd75..693bcd7ef6d2 100644
> --- a/net/ipv6/ip6_fib.c
> +++ b/net/ipv6/ip6_fib.c
> @@ -1592,13 +1592,7 @@ int fib6_del(struct rt6_info *rt, struct nl_info *info)
> struct net *net = info->nl_net;
> struct rt6_info **rtp;
>
> -#if RT6_DEBUG >= 2
> - if (rt->dst.obsolete > 0) {
> - WARN_ON(fn);
fn should have already been set to NULL if it is removed
from the fib6 tree?
> - return -ENOENT;
> - }
> -#endif
> - if (!fn || rt == net->ipv6.ip6_null_entry)
> + if (!fn || rt->dst.obsolete > 0 || rt == net->ipv6.ip6_null_entry)
> return -ENOENT;
>
> WARN_ON(!(fn->fn_flags & RTN_RTINFO));
> --
> 2.14.1.821.g8fa685d3b7-goog
>
^ permalink raw reply
* Re: [Patch net-next v2] net_sched: use idr to allocate bpf filter handles
From: Daniel Borkmann @ 2017-09-25 23:47 UTC (permalink / raw)
To: Cong Wang; +Cc: Linux Kernel Network Developers, Chris Mi, Jamal Hadi Salim
In-Reply-To: <CAM_iQpUKURMfZOyTMbgVFG43rgUg8-3yAAVtEmPja5T6xkZpjA@mail.gmail.com>
On 09/26/2017 01:11 AM, Cong Wang wrote:
> On Mon, Sep 25, 2017 at 2:16 PM, Daniel Borkmann <daniel@iogearbox.net> wrote:
>> On 09/25/2017 07:13 PM, Cong Wang wrote:
[...]
>> There's this condition in the code before above idr allocations,
>> I think also in other classifiers:
>>
>> if (oldprog) {
>> if (handle && oldprog->handle != handle) {
>> ret = -EINVAL;
>> goto errout;
>> }
>> }
>
> Sure. If we use handle to find oldprog, it should have the
> same handle. cls_bpf_get() guarantees it. This check is
> redundant.
Good point, we should just test for 'oldprog && oldprog->handle !=
handle' and bail out then, otherwise it's just irritating. I can
see to fix this up later. Seems fine to me then, thanks!
^ permalink raw reply
* Re: [PATCH] [for 4.14] net: qcom/emac: specify the correct size when mapping a DMA buffer
From: David Miller @ 2017-09-25 23:35 UTC (permalink / raw)
To: timur; +Cc: netdev, stable
In-Reply-To: <1506112364-26787-1-git-send-email-timur@codeaurora.org>
From: Timur Tabi <timur@codeaurora.org>
Date: Fri, 22 Sep 2017 15:32:44 -0500
> When mapping the RX DMA buffers, the driver was accidentally specifying
> zero for the buffer length. Under normal circumstances, SWIOTLB does not
> need to allocate a bounce buffer, so the address is just mapped without
> checking the size field. This is why the error was not detected earlier.
>
> Fixes: b9b17debc69d ("net: emac: emac gigabit ethernet controller driver")
> Cc: stable@vger.kernel.org
> Signed-off-by: Timur Tabi <timur@codeaurora.org>
Applied and queued up for -stable, thanks.
^ permalink raw reply
* Re: [net-next] net: remove MTU limits for dummy and ifb device
From: David Miller @ 2017-09-25 23:26 UTC (permalink / raw)
To: zhangshengju; +Cc: sd, edumazet, netdev
In-Reply-To: <1506095869-5465-1-git-send-email-zhangshengju@cmss.chinamobile.com>
From: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Date: Fri, 22 Sep 2017 23:57:49 +0800
> These two drivers (dummy and ifb) call ether_setup(), after commit
> 61e84623ace3 ("net: centralize net_device min/max MTU checking"), the
> range of mtu is [min_mtu, max_mtu], which is [68, 1500] by default.
>
> These two devices should not have limits on MTU. This patch set their
> min_mtu/max_mtu to 0. So that dev_set_mtu() will not check the mtu range,
> and can be set with any value.
>
> CC: Eric Dumazet <edumazet@google.com>
> CC: Sabrina Dubroca <sd@queasysnail.net>
> Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Applied.
^ permalink raw reply
* Re: [PATCH] hv_netvsc: make const array ver_list static, reduces object code size
From: David Miller @ 2017-09-25 23:25 UTC (permalink / raw)
To: colin.king
Cc: kys, haiyangz, sthemmin, devel, netdev, kernel-janitors,
linux-kernel
In-Reply-To: <20170922155023.15821-1-colin.king@canonical.com>
From: Colin King <colin.king@canonical.com>
Date: Fri, 22 Sep 2017 16:50:23 +0100
> From: Colin Ian King <colin.king@canonical.com>
>
> Don't populate const array ver_list on the stack, instead make it
> static. Makes the object code smaller by over 400 bytes:
>
> Before:
> text data bss dec hex filename
> 18444 3168 320 21932 55ac drivers/net/hyperv/netvsc.o
>
> After:
> text data bss dec hex filename
> 17950 3224 320 21494 53f6 drivers/net/hyperv/netvsc.o
>
> (gcc 6.3.0, x86-64)
>
> Signed-off-by: Colin Ian King <colin.king@canonical.com>
Applied to net-next, thanks.
^ permalink raw reply
* Re: [PATCH v4 2/3] ipv4: Namespaceify tcp_fastopen_key knob
From: David Miller @ 2017-09-25 23:24 UTC (permalink / raw)
To: yanhaishuang; +Cc: kuznet, edumazet, weiwan, lucab, netdev, linux-kernel
In-Reply-To: <1506088124-12650-2-git-send-email-yanhaishuang@cmss.chinamobile.com>
From: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>
Date: Fri, 22 Sep 2017 21:48:43 +0800
> @@ -9,13 +9,18 @@
> #include <net/inetpeer.h>
> #include <net/tcp.h>
>
> -struct tcp_fastopen_context __rcu *tcp_fastopen_ctx;
> -
> -static DEFINE_SPINLOCK(tcp_fastopen_ctx_lock);
> -
> -void tcp_fastopen_init_key_once(bool publish)
> +void tcp_fastopen_init_key_once(struct net *net)
Why did you remove the 'publish' logic from this function?
^ permalink raw reply
* Re: [Patch net-next v2] net_sched: use idr to allocate u32 filter handles
From: Cong Wang @ 2017-09-25 23:13 UTC (permalink / raw)
To: Jiri Pirko; +Cc: Linux Kernel Network Developers, Chris Mi, Jamal Hadi Salim
In-Reply-To: <20170925190031.GB1936@nanopsycho.orion>
On Mon, Sep 25, 2017 at 12:00 PM, Jiri Pirko <jiri@resnulli.us> wrote:
> Interesting, any idea why this is not 0x7FFFFFFF as well?
>
> I wonder if we could have 0x7FFFFFFF magic defined somewhere.
I have no idea, it just exists for a rather long time. Probably too late
to change, or at least requires a separate patch to change it.
^ permalink raw reply
* Re: [Patch net-next v2] net_sched: use idr to allocate bpf filter handles
From: Cong Wang @ 2017-09-25 23:11 UTC (permalink / raw)
To: Daniel Borkmann
Cc: Linux Kernel Network Developers, Chris Mi, Jamal Hadi Salim
In-Reply-To: <59C97244.30301@iogearbox.net>
On Mon, Sep 25, 2017 at 2:16 PM, Daniel Borkmann <daniel@iogearbox.net> wrote:
> On 09/25/2017 07:13 PM, Cong Wang wrote:
>> ret = cls_bpf_offload(tp, prog, oldprog);
>> if (ret) {
>> + if (!oldprog)
>> + idr_remove_ext(&head->handle_idr, prog->handle);
>
>
> Shouldn't we also call idr_remove_ext() when there was an
> oldprog, but we didn't care about reusing the same handle,
> so it was handle == 0 initially?
When oldprog is non-NULL, we are replacing the oldprog with
a new one, therefore we should call idr_replace_ext() which
happens after this. So no need to call idr_remove_ext() at
this point.
>
> There's this condition in the code before above idr allocations,
> I think also in other classifiers:
>
> if (oldprog) {
> if (handle && oldprog->handle != handle) {
> ret = -EINVAL;
> goto errout;
> }
> }
Sure. If we use handle to find oldprog, it should have the
same handle. cls_bpf_get() guarantees it. This check is
redundant.
>
>> __cls_bpf_delete_prog(prog);
>> return ret;
>> }
>> @@ -499,6 +494,7 @@ static int cls_bpf_change(struct net *net, struct
>> sk_buff *in_skb,
>> prog->gen_flags |= TCA_CLS_FLAGS_NOT_IN_HW;
>>
>> if (oldprog) {
>> + idr_replace_ext(&head->handle_idr, prog, handle);
>
>
> And here, we should probably use prog->handle for the above
> mentioned case as well, no?
Since are replacing oldprog with a new one, prog->handle is
same with handle.
>
> Would be great if all this (and e.g. the fact that we use idr itself)
> could optionally be hidden behind some handle generator api given
> we could reuse that api also for cls_basic and cls_u32. Could also
> be followed-up perhaps.
>
Yeah, the idr_alloc_ext(.., handle, handle+1,) is ugly. Ideally we should
specify the range during initialization rather than in each idr_alloc_ext().
Commit c15ab236d69d already did the same thing. We can refactor
this later.
^ permalink raw reply
* [PATCH net] net: dsa: Fix network device registration order
From: Florian Fainelli @ 2017-09-25 22:55 UTC (permalink / raw)
To: netdev; +Cc: davem, Florian Fainelli, Andrew Lunn, Vivien Didelot, open list
We cannot be registering the network device first, then setting its
carrier off and finally connecting it to a PHY, doing that leaves a
window during which the carrier is at best inconsistent, and at worse
the device is not usable without a down/up sequence since the network
device is visible to user space with possibly no PHY device attached.
Re-order steps so that they make logical sense. This fixes some devices
where the port was not usable after e.g: an unbind then bind of the
driver.
Fixes: 0071f56e46da ("dsa: Register netdev before phy")
Fixes: 91da11f870f0 ("net: Distributed Switch Architecture protocol support")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
net/dsa/slave.c | 31 ++++++++++++++++++-------------
1 file changed, 18 insertions(+), 13 deletions(-)
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index 2afa99506f8b..865e29e62bad 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -1301,28 +1301,33 @@ int dsa_slave_create(struct dsa_port *port, const char *name)
p->old_duplex = -1;
port->netdev = slave_dev;
- ret = register_netdev(slave_dev);
- if (ret) {
- netdev_err(master, "error %d registering interface %s\n",
- ret, slave_dev->name);
- port->netdev = NULL;
- free_percpu(p->stats64);
- free_netdev(slave_dev);
- return ret;
- }
netif_carrier_off(slave_dev);
ret = dsa_slave_phy_setup(p, slave_dev);
if (ret) {
netdev_err(master, "error %d setting up slave phy\n", ret);
- unregister_netdev(slave_dev);
- free_percpu(p->stats64);
- free_netdev(slave_dev);
- return ret;
+ goto out_free;
+ }
+
+ ret = register_netdev(slave_dev);
+ if (ret) {
+ netdev_err(master, "error %d registering interface %s\n",
+ ret, slave_dev->name);
+ goto out_phy;
}
return 0;
+
+out_phy:
+ phy_disconnect(p->phy);
+ if (of_phy_is_fixed_link(p->dp->dn))
+ of_phy_deregister_fixed_link(p->dp->dn);
+out_free:
+ free_percpu(p->stats64);
+ free_netdev(slave_dev);
+ port->netdev = NULL;
+ return ret;
}
void dsa_slave_destroy(struct net_device *slave_dev)
--
2.9.3
^ permalink raw reply related
* [jkirsher/next-queue PATCH] ixgbe: Update adaptive ITR algorithm
From: Alexander Duyck @ 2017-09-25 21:55 UTC (permalink / raw)
To: netdev, intel-wired-lan; +Cc: john.fastabend, brouer
From: Alexander Duyck <alexander.h.duyck@intel.com>
The following change is meant to update the adaptive ITR algorithm to
better support the needs of the network. Specifically with this change what
I have done is make it so that our ITR algorithm will try to prevent either
starving a socket buffer for memory in the case of Tx, or overruing an Rx
socket buffer on receive.
In addition a side effect of the calculations used is that we should
function better with new features such as XDP which can handle small
packets at high rates without needing to lock us into NAPI polling mode.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
---
So I am putting this out to a wider distribution list than normal for a
patch like this in order to get feedback on if there are any areas I may
have overlooked. With this patch is should address many of the performance
limitations seen with pktgen and XDP in terms of workloads that the old
adaptive scheme wasn't handling.
drivers/net/ethernet/intel/ixgbe/ixgbe.h | 7 +
drivers/net/ethernet/intel/ixgbe/ixgbe_lib.c | 11 +
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 215 +++++++++++++++++++------
3 files changed, 178 insertions(+), 55 deletions(-)
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe.h b/drivers/net/ethernet/intel/ixgbe/ixgbe.h
index 56039d04b38d..555eb80d8a08 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe.h
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe.h
@@ -435,8 +435,15 @@ static inline unsigned int ixgbe_rx_pg_order(struct ixgbe_ring *ring)
}
#define ixgbe_rx_pg_size(_ring) (PAGE_SIZE << ixgbe_rx_pg_order(_ring))
+#define IXGBE_ITR_ADAPTIVE_MIN_INC 2
+#define IXGBE_ITR_ADAPTIVE_MIN_USECS 10
+#define IXGBE_ITR_ADAPTIVE_MAX_USECS 126
+#define IXGBE_ITR_ADAPTIVE_LATENCY 0x80
+#define IXGBE_ITR_ADAPTIVE_BULK 0x00
+
struct ixgbe_ring_container {
struct ixgbe_ring *ring; /* pointer to linked list of rings */
+ unsigned long next_update; /* jiffies value of last update */
unsigned int total_bytes; /* total bytes processed this int */
unsigned int total_packets; /* total packets processed this int */
u16 work_limit; /* total work allowed per interrupt */
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_lib.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_lib.c
index f1bfae0c41d0..8e2a957aca18 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_lib.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_lib.c
@@ -806,6 +806,7 @@ static void ixgbe_add_ring(struct ixgbe_ring *ring,
ring->next = head->ring;
head->ring = ring;
head->count++;
+ head->next_update = jiffies + 1;
}
/**
@@ -879,8 +880,11 @@ static int ixgbe_alloc_q_vector(struct ixgbe_adapter *adapter,
/* initialize work limits */
q_vector->tx.work_limit = adapter->tx_work_limit;
- /* initialize pointer to rings */
- ring = q_vector->ring;
+ /* Initialize setting for adaptive ITR */
+ q_vector->tx.itr = IXGBE_ITR_ADAPTIVE_MAX_USECS |
+ IXGBE_ITR_ADAPTIVE_LATENCY;
+ q_vector->rx.itr = IXGBE_ITR_ADAPTIVE_MAX_USECS |
+ IXGBE_ITR_ADAPTIVE_LATENCY;
/* intialize ITR */
if (txr_count && !rxr_count) {
@@ -897,6 +901,9 @@ static int ixgbe_alloc_q_vector(struct ixgbe_adapter *adapter,
q_vector->itr = adapter->rx_itr_setting;
}
+ /* initialize pointer to rings */
+ ring = q_vector->ring;
+
while (txr_count) {
/* assign generic ring traits */
ring->dev = &adapter->pdev->dev;
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
index 3d3739f103af..44a96878075b 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
@@ -2517,50 +2517,174 @@ enum latency_range {
static void ixgbe_update_itr(struct ixgbe_q_vector *q_vector,
struct ixgbe_ring_container *ring_container)
{
- int bytes = ring_container->total_bytes;
- int packets = ring_container->total_packets;
- u32 timepassed_us;
- u64 bytes_perint;
- u8 itr_setting = ring_container->itr;
+ unsigned int itr = IXGBE_ITR_ADAPTIVE_MIN_USECS |
+ IXGBE_ITR_ADAPTIVE_LATENCY;
+ unsigned int avg_wire_size, packets, bytes;
+ unsigned long next_update = jiffies;
- if (packets == 0)
+ /* If we don't have any rings just leave ourselves set for maximum
+ * possible latency so we take ourselves out of the equation.
+ */
+ if (!ring_container->ring)
return;
- /* simple throttlerate management
- * 0-10MB/s lowest (100000 ints/s)
- * 10-20MB/s low (20000 ints/s)
- * 20-1249MB/s bulk (12000 ints/s)
+ /* If we didn't update within up to 1 - 2 jiffies we can assume
+ * that either packets are coming in so slow there hasn't been
+ * any work, or that there is so much work that NAPI is dealing
+ * with interrupt moderation and we don't need to do anything.
*/
- /* what was last interrupt timeslice? */
- timepassed_us = q_vector->itr >> 2;
- if (timepassed_us == 0)
- return;
+ if (time_after(next_update, ring_container->next_update))
+ goto clear_counts;
- bytes_perint = bytes / timepassed_us; /* bytes/usec */
+ packets = ring_container->total_packets;
- switch (itr_setting) {
- case lowest_latency:
- if (bytes_perint > 10)
- itr_setting = low_latency;
- break;
- case low_latency:
- if (bytes_perint > 20)
- itr_setting = bulk_latency;
- else if (bytes_perint <= 10)
- itr_setting = lowest_latency;
+ /* We have no packets to actually measure against. This means
+ * either one of the other queues on this vector is active or
+ * we are a Tx queue doing TSO with too high of an interrupt rate.
+ *
+ * When this occurs just tick up our delay by the minimum value
+ * and hope that this extra delay will prevent us from being called
+ * without any work on our queue.
+ */
+ if (!packets) {
+ itr = (q_vector->itr >> 2) + IXGBE_ITR_ADAPTIVE_MIN_INC;
+ if (itr > IXGBE_ITR_ADAPTIVE_MAX_USECS)
+ itr = IXGBE_ITR_ADAPTIVE_MAX_USECS;
+ itr += ring_container->itr & IXGBE_ITR_ADAPTIVE_LATENCY;
+ goto clear_counts;
+ }
+
+ bytes = ring_container->total_bytes;
+
+ /* If packets are less than 4 or bytes are less than 9000 assume
+ * insufficient data to use bulk rate limiting approach. We are
+ * likely latency driven.
+ */
+ if (packets < 4 && bytes < 9000) {
+ itr = IXGBE_ITR_ADAPTIVE_LATENCY;
+ goto adjust_by_size;
+ }
+
+ /* Between 4 and 48 we can assume that our current interrupt delay
+ * is only slightly too low. As such we should increase it by a small
+ * fixed amount.
+ */
+ if (packets < 48) {
+ itr = (q_vector->itr >> 2) + IXGBE_ITR_ADAPTIVE_MIN_INC;
+ if (itr > IXGBE_ITR_ADAPTIVE_MAX_USECS)
+ itr = IXGBE_ITR_ADAPTIVE_MAX_USECS;
+ goto clear_counts;
+ }
+
+ /* Between 48 and 96 is our "goldilocks" zone where we are working
+ * out "just right". Just report that our current ITR is good for us.
+ */
+ if (packets < 96) {
+ itr = q_vector->itr >> 2;
+ goto clear_counts;
+ }
+
+ /* If packet count is 96 or greater we are likely looking at a slight
+ * overrun of the delay we want. Try halving our delay to see if that
+ * will cut the number of packets in half per interrupt.
+ */
+ if (packets < 256) {
+ itr = q_vector->itr >> 3;
+ if (itr < IXGBE_ITR_ADAPTIVE_MIN_USECS)
+ itr = IXGBE_ITR_ADAPTIVE_MIN_USECS;
+ goto clear_counts;
+ }
+
+ /* The paths below assume we are dealing with a bulk ITR since number
+ * of packets is 256 or greater. We are just going to have to compute
+ * a value and try to bring the count under control, though for smaller
+ * packet sizes there isn't much we can do as NAPI polling will likely
+ * be kicking in sooner rather than later.
+ */
+ itr = IXGBE_ITR_ADAPTIVE_BULK;
+
+adjust_by_size:
+ /* If packet counts are 256 or greater we can assume we have a gross
+ * overestimation of what the rate should be. Instead of trying to fine
+ * tune it just use the formula below to try and dial in an exact value
+ * give the current packet size of the frame.
+ */
+ avg_wire_size = bytes / packets;
+
+ /* The following is a crude approximation of:
+ * wmem_default / (size + overhead) = desired_pkts_per_int
+ * rate / bits_per_byte / (size + ethernet overhead) = pkt_rate
+ * (desired_pkt_rate / pkt_rate) * usecs_per_sec = ITR value
+ *
+ * Assuming wmem_default is 212992 and overhead is 640 bytes per
+ * packet, (256 skb, 64 headroom, 320 shared info), we can reduce the
+ * formula down to
+ *
+ * (170 * (size + 24)) / (size + 640) = ITR
+ *
+ * We first do some math on the packet size and then finally bitshift
+ * by 8 after rounding up. We also have to account for PCIe link speed
+ * difference as ITR scales based on this.
+ */
+ if (avg_wire_size <= 60) {
+ /* Start at 50k ints/sec */
+ avg_wire_size = 5120;
+ } else if (avg_wire_size <= 316) {
+ /* 50K ints/sec to 16K ints/sec */
+ avg_wire_size *= 40;
+ avg_wire_size += 2720;
+ } else if (avg_wire_size <= 1084) {
+ /* 16K ints/sec to 9.2K ints/sec */
+ avg_wire_size *= 15;
+ avg_wire_size += 11452;
+ } else if (avg_wire_size <= 1980) {
+ /* 9.2K ints/sec to 8K ints/sec */
+ avg_wire_size *= 5;
+ avg_wire_size += 22420;
+ } else {
+ /* plateau at a limit of 8K ints/sec */
+ avg_wire_size = 32256;
+ }
+
+ /* If we are in low latency mode half our delay which doubles the rate
+ * to somewhere between 100K to 16K ints/sec
+ */
+ if (itr & IXGBE_ITR_ADAPTIVE_LATENCY)
+ avg_wire_size >>= 1;
+
+ /* Resultant value is 256 times larger than it needs to be. This
+ * gives us room to adjust the value as needed to either increase
+ * or decrease the value based on link speeds of 10G, 2.5G, 1G, etc.
+ *
+ * Use addition as we have already recorded the new latency flag
+ * for the ITR value.
+ */
+ switch (q_vector->adapter->link_speed) {
+ case IXGBE_LINK_SPEED_10GB_FULL:
+ case IXGBE_LINK_SPEED_100_FULL:
+ default:
+ itr += DIV_ROUND_UP(avg_wire_size,
+ IXGBE_ITR_ADAPTIVE_MIN_INC * 256) *
+ IXGBE_ITR_ADAPTIVE_MIN_INC;
break;
- case bulk_latency:
- if (bytes_perint <= 20)
- itr_setting = low_latency;
+ case IXGBE_LINK_SPEED_2_5GB_FULL:
+ case IXGBE_LINK_SPEED_1GB_FULL:
+ case IXGBE_LINK_SPEED_10_FULL:
+ itr += DIV_ROUND_UP(avg_wire_size,
+ IXGBE_ITR_ADAPTIVE_MIN_INC * 64) *
+ IXGBE_ITR_ADAPTIVE_MIN_INC;
break;
}
- /* clear work counters since we have the values we need */
+clear_counts:
+ /* write back value */
+ ring_container->itr = itr;
+
+ /* next update should occur within next jiffy */
+ ring_container->next_update = next_update + 1;
+
ring_container->total_bytes = 0;
ring_container->total_packets = 0;
-
- /* write updated itr to ring container */
- ring_container->itr = itr_setting;
}
/**
@@ -2602,34 +2726,19 @@ void ixgbe_write_eitr(struct ixgbe_q_vector *q_vector)
static void ixgbe_set_itr(struct ixgbe_q_vector *q_vector)
{
- u32 new_itr = q_vector->itr;
- u8 current_itr;
+ u32 new_itr;
ixgbe_update_itr(q_vector, &q_vector->tx);
ixgbe_update_itr(q_vector, &q_vector->rx);
- current_itr = max(q_vector->rx.itr, q_vector->tx.itr);
+ /* use the smallest value of new ITR delay calculations */
+ new_itr = min(q_vector->rx.itr, q_vector->tx.itr);
- switch (current_itr) {
- /* counts and packets in update_itr are dependent on these numbers */
- case lowest_latency:
- new_itr = IXGBE_100K_ITR;
- break;
- case low_latency:
- new_itr = IXGBE_20K_ITR;
- break;
- case bulk_latency:
- new_itr = IXGBE_12K_ITR;
- break;
- default:
- break;
- }
+ /* Clear latency flag if set, shift into correct position */
+ new_itr &= ~IXGBE_ITR_ADAPTIVE_LATENCY;
+ new_itr <<= 2;
if (new_itr != q_vector->itr) {
- /* do an exponential smoothing */
- new_itr = (10 * new_itr * q_vector->itr) /
- ((9 * new_itr) + q_vector->itr);
-
/* save the algorithm value here */
q_vector->itr = new_itr;
^ permalink raw reply related
* Re: [PATCH net-next] sch_netem: faster rb tree removal
From: David Ahern @ 2017-09-25 21:52 UTC (permalink / raw)
To: David Miller; +Cc: eric.dumazet, netdev, stephen
In-Reply-To: <20170925.131101.478664536826772174.davem@davemloft.net>
On 9/25/17 2:11 PM, David Miller wrote:
> From: David Ahern <dsahern@gmail.com>
> Date: Mon, 25 Sep 2017 10:14:23 -0600
>
>> I made a simple program this morning and ran it under perf.
>
> If possible please submit this for selftests.
>
It is more of a microbenchmark of options to flush an rbtree than a
self-test. Further, it relies on the tools/lib/rbtree.c versus
lib/rbtree.c. The tools/lib version was imported by Arnaldo in July 2015
and is a out of date, though it is good enough to show the intent w.r.t.
flushing options.
^ permalink raw reply
* Re: [PATCH net 0/2] l2tp: fix some races in session deletion
From: David Miller @ 2017-09-25 21:45 UTC (permalink / raw)
To: g.nault; +Cc: netdev, jchapman, tparkin, sd
In-Reply-To: <cover.1506086081.git.g.nault@alphalink.fr>
From: Guillaume Nault <g.nault@alphalink.fr>
Date: Fri, 22 Sep 2017 15:39:22 +0200
> L2TP provides several interfaces for deleting sessions. Using two of
> them concurrently can lead to use-after-free bugs.
>
> Patch #2 uses a flag to prevent double removal of L2TP sessions.
> Patch #1 fixes a bug found in the way. Fixing this bug is also
> necessary for patch #2 to handle all cases.
>
> This issue is similar to the tunnel deletion bug being worked on by
> Sabrina: https://patchwork.ozlabs.org/patch/814173/
Series applied and queued up for -stable, thanks.
^ 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