From: Huazhong Tan <tanhuazhong@huawei.com>
To: <davem@davemloft.net>
Cc: <netdev@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
<salil.mehta@huawei.com>, <yisen.zhuang@huawei.com>,
<linuxarm@huawei.com>, <jakub.kicinski@netronome.com>,
Yunsheng Lin <linyunsheng@huawei.com>,
Huazhong Tan <tanhuazhong@huawei.com>
Subject: [PATCH net-next 3/5] net: hns3: remove unnecessary work in hclgevf_main
Date: Sat, 14 Dec 2019 10:06:39 +0800 [thread overview]
Message-ID: <1576289201-57017-4-git-send-email-tanhuazhong@huawei.com> (raw)
In-Reply-To: <1576289201-57017-1-git-send-email-tanhuazhong@huawei.com>
From: Yunsheng Lin <linyunsheng@huawei.com>
There are four work (mbx_service_task, service_task,
rst_service_task and keep_alive_task)in the hclgevf module,
mbx_service_task is for handling mailbox issue, service_task is
for periodic management issue and rst_service_task is for reset
related issue, keep_alive_task is used to keepalive between PF
and VF, which can be done in a single work.
This patch removes the mbx_service_task, rst_service_task and
keep_alive_task, and moves the related handling to the
service_task work in order to remove concurrency between the four
work and to improve efficiency.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
.../ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c | 175 ++++++++++-----------
.../ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h | 9 +-
2 files changed, 88 insertions(+), 96 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
index e48f8fe..004dbaf 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
@@ -440,6 +440,9 @@ void hclgevf_update_link_status(struct hclgevf_dev *hdev, int link_state)
struct hnae3_client *rclient;
struct hnae3_client *client;
+ if (test_and_set_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state))
+ return;
+
client = handle->client;
rclient = hdev->roce_client;
@@ -452,6 +455,8 @@ void hclgevf_update_link_status(struct hclgevf_dev *hdev, int link_state)
rclient->ops->link_status_change(rhandle, !!link_state);
hdev->hw.mac.link = link_state;
}
+
+ clear_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state);
}
static void hclgevf_update_link_mode(struct hclgevf_dev *hdev)
@@ -1767,54 +1772,38 @@ static void hclgevf_get_misc_vector(struct hclgevf_dev *hdev)
void hclgevf_reset_task_schedule(struct hclgevf_dev *hdev)
{
- if (!test_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state) &&
- !test_bit(HCLGEVF_STATE_REMOVING, &hdev->state)) {
- set_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state);
- schedule_work(&hdev->rst_service_task);
- }
+ if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state) &&
+ !test_and_set_bit(HCLGEVF_STATE_RST_SERVICE_SCHED,
+ &hdev->state))
+ mod_delayed_work(system_wq, &hdev->service_task, 0);
}
void hclgevf_mbx_task_schedule(struct hclgevf_dev *hdev)
{
- if (!test_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state) &&
- !test_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state)) {
- set_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state);
- schedule_work(&hdev->mbx_service_task);
- }
+ if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state) &&
+ !test_and_set_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED,
+ &hdev->state))
+ mod_delayed_work(system_wq, &hdev->service_task, 0);
}
-static void hclgevf_task_schedule(struct hclgevf_dev *hdev)
+static void hclgevf_task_schedule(struct hclgevf_dev *hdev,
+ unsigned long delay)
{
- if (!test_bit(HCLGEVF_STATE_DOWN, &hdev->state) &&
- !test_and_set_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state)) {
- mod_delayed_work(system_wq, &hdev->service_task,
- round_jiffies_relative(HZ));
- hdev->stats_timer++;
- }
+ if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state))
+ mod_delayed_work(system_wq, &hdev->service_task, delay);
}
-static void hclgevf_deferred_task_schedule(struct hclgevf_dev *hdev)
-{
- /* if we have any pending mailbox event then schedule the mbx task */
- if (hdev->mbx_event_pending)
- hclgevf_mbx_task_schedule(hdev);
-
- if (test_bit(HCLGEVF_RESET_PENDING, &hdev->reset_state))
- hclgevf_reset_task_schedule(hdev);
-}
-
-static void hclgevf_reset_service_task(struct work_struct *work)
+static void hclgevf_reset_service_task(struct hclgevf_dev *hdev)
{
#define HCLGEVF_MAX_RESET_ATTEMPTS_CNT 3
- struct hclgevf_dev *hdev =
- container_of(work, struct hclgevf_dev, rst_service_task);
int ret;
- if (test_and_set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state))
+ if (!test_and_clear_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state))
return;
- clear_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state);
+ if (test_and_set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state))
+ return;
if (test_and_clear_bit(HCLGEVF_RESET_PENDING,
&hdev->reset_state)) {
@@ -1877,39 +1866,24 @@ static void hclgevf_reset_service_task(struct work_struct *work)
clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state);
}
-static void hclgevf_mailbox_service_task(struct work_struct *work)
+static void hclgevf_mailbox_service_task(struct hclgevf_dev *hdev)
{
- struct hclgevf_dev *hdev;
-
- hdev = container_of(work, struct hclgevf_dev, mbx_service_task);
+ if (!test_and_clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state))
+ return;
if (test_and_set_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state))
return;
- clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state);
-
hclgevf_mbx_async_handler(hdev);
clear_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state);
}
-static void hclgevf_keep_alive_timer(struct timer_list *t)
-{
- struct hclgevf_dev *hdev = from_timer(hdev, t, keep_alive_timer);
-
- schedule_work(&hdev->keep_alive_task);
- mod_timer(&hdev->keep_alive_timer, jiffies +
- HCLGEVF_KEEP_ALIVE_TASK_INTERVAL * HZ);
-}
-
-static void hclgevf_keep_alive_task(struct work_struct *work)
+static void hclgevf_keep_alive(struct hclgevf_dev *hdev)
{
- struct hclgevf_dev *hdev;
u8 respmsg;
int ret;
- hdev = container_of(work, struct hclgevf_dev, keep_alive_task);
-
if (test_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state))
return;
@@ -1920,19 +1894,32 @@ static void hclgevf_keep_alive_task(struct work_struct *work)
"VF sends keep alive cmd failed(=%d)\n", ret);
}
-static void hclgevf_service_task(struct work_struct *work)
+static void hclgevf_periodic_service_task(struct hclgevf_dev *hdev)
{
- struct hnae3_handle *handle;
- struct hclgevf_dev *hdev;
+ unsigned long delta = round_jiffies_relative(HZ);
+ struct hnae3_handle *handle = &hdev->nic;
- hdev = container_of(work, struct hclgevf_dev, service_task.work);
- handle = &hdev->nic;
+ if (time_is_after_jiffies(hdev->last_serv_processed + HZ)) {
+ delta = jiffies - hdev->last_serv_processed;
- if (hdev->stats_timer >= HCLGEVF_STATS_TIMER_INTERVAL) {
- hclgevf_tqps_update_stats(handle);
- hdev->stats_timer = 0;
+ if (delta < round_jiffies_relative(HZ)) {
+ delta = round_jiffies_relative(HZ) - delta;
+ goto out;
+ }
+ }
+
+ hdev->serv_processed_cnt++;
+ if (!(hdev->serv_processed_cnt % HCLGEVF_KEEP_ALIVE_TASK_INTERVAL))
+ hclgevf_keep_alive(hdev);
+
+ if (test_bit(HCLGEVF_STATE_DOWN, &hdev->state)) {
+ hdev->last_serv_processed = jiffies;
+ goto out;
}
+ if (!(hdev->serv_processed_cnt % HCLGEVF_STATS_TIMER_INTERVAL))
+ hclgevf_tqps_update_stats(handle);
+
/* request the link status from the PF. PF would be able to tell VF
* about such updates in future so we might remove this later
*/
@@ -1942,11 +1929,27 @@ static void hclgevf_service_task(struct work_struct *work)
hclgevf_sync_vlan_filter(hdev);
- hclgevf_deferred_task_schedule(hdev);
+ hdev->last_serv_processed = jiffies;
- clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state);
+out:
+ hclgevf_task_schedule(hdev, delta);
+}
+
+static void hclgevf_service_task(struct work_struct *work)
+{
+ struct hclgevf_dev *hdev = container_of(work, struct hclgevf_dev,
+ service_task.work);
- hclgevf_task_schedule(hdev);
+ hclgevf_reset_service_task(hdev);
+ hclgevf_mailbox_service_task(hdev);
+ hclgevf_periodic_service_task(hdev);
+
+ /* Handle reset and mbx again in case periodical task delays the
+ * handling by calling hclgevf_task_schedule() in
+ * hclgevf_periodic_service_task()
+ */
+ hclgevf_reset_service_task(hdev);
+ hclgevf_mailbox_service_task(hdev);
}
static void hclgevf_clear_event_cause(struct hclgevf_dev *hdev, u32 regclr)
@@ -2183,16 +2186,31 @@ static int hclgevf_init_vlan_config(struct hclgevf_dev *hdev)
false);
}
+static void hclgevf_flush_link_update(struct hclgevf_dev *hdev)
+{
+#define HCLGEVF_FLUSH_LINK_TIMEOUT 100000
+
+ unsigned long last = hdev->serv_processed_cnt;
+ int i = 0;
+
+ while (test_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state) &&
+ i++ < HCLGEVF_FLUSH_LINK_TIMEOUT &&
+ last == hdev->serv_processed_cnt)
+ usleep_range(1, 1);
+}
+
static void hclgevf_set_timer_task(struct hnae3_handle *handle, bool enable)
{
struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle);
if (enable) {
- hclgevf_task_schedule(hdev);
+ hclgevf_task_schedule(hdev, 0);
} else {
set_bit(HCLGEVF_STATE_DOWN, &hdev->state);
- cancel_delayed_work_sync(&hdev->service_task);
- clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state);
+
+ /* flush memory to make sure DOWN is seen by service task */
+ smp_mb__before_atomic();
+ hclgevf_flush_link_update(hdev);
}
}
@@ -2239,16 +2257,12 @@ static int hclgevf_set_alive(struct hnae3_handle *handle, bool alive)
static int hclgevf_client_start(struct hnae3_handle *handle)
{
- struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle);
int ret;
ret = hclgevf_set_alive(handle, true);
if (ret)
return ret;
- mod_timer(&hdev->keep_alive_timer, jiffies +
- HCLGEVF_KEEP_ALIVE_TASK_INTERVAL * HZ);
-
return 0;
}
@@ -2261,22 +2275,14 @@ static void hclgevf_client_stop(struct hnae3_handle *handle)
if (ret)
dev_warn(&hdev->pdev->dev,
"%s failed %d\n", __func__, ret);
-
- del_timer_sync(&hdev->keep_alive_timer);
- cancel_work_sync(&hdev->keep_alive_task);
}
static void hclgevf_state_init(struct hclgevf_dev *hdev)
{
- /* setup tasks for the MBX */
- INIT_WORK(&hdev->mbx_service_task, hclgevf_mailbox_service_task);
clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state);
clear_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state);
INIT_DELAYED_WORK(&hdev->service_task, hclgevf_service_task);
- clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state);
-
- INIT_WORK(&hdev->rst_service_task, hclgevf_reset_service_task);
mutex_init(&hdev->mbx_resp.mbx_mutex);
@@ -2289,16 +2295,8 @@ static void hclgevf_state_uninit(struct hclgevf_dev *hdev)
set_bit(HCLGEVF_STATE_DOWN, &hdev->state);
set_bit(HCLGEVF_STATE_REMOVING, &hdev->state);
- if (hdev->keep_alive_timer.function)
- del_timer_sync(&hdev->keep_alive_timer);
- if (hdev->keep_alive_task.func)
- cancel_work_sync(&hdev->keep_alive_task);
if (hdev->service_task.work.func)
cancel_delayed_work_sync(&hdev->service_task);
- if (hdev->mbx_service_task.func)
- cancel_work_sync(&hdev->mbx_service_task);
- if (hdev->rst_service_task.func)
- cancel_work_sync(&hdev->rst_service_task);
mutex_destroy(&hdev->mbx_resp.mbx_mutex);
}
@@ -2796,6 +2794,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev)
dev_info(&hdev->pdev->dev, "finished initializing %s driver\n",
HCLGEVF_DRIVER_NAME);
+ hclgevf_task_schedule(hdev, round_jiffies_relative(HZ));
+
return 0;
err_config:
@@ -2827,7 +2827,6 @@ static void hclgevf_uninit_hdev(struct hclgevf_dev *hdev)
static int hclgevf_init_ae_dev(struct hnae3_ae_dev *ae_dev)
{
struct pci_dev *pdev = ae_dev->pdev;
- struct hclgevf_dev *hdev;
int ret;
ret = hclgevf_alloc_hdev(ae_dev);
@@ -2842,10 +2841,6 @@ static int hclgevf_init_ae_dev(struct hnae3_ae_dev *ae_dev)
return ret;
}
- hdev = ae_dev->priv;
- timer_setup(&hdev->keep_alive_timer, hclgevf_keep_alive_timer, 0);
- INIT_WORK(&hdev->keep_alive_task, hclgevf_keep_alive_task);
-
return 0;
}
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
index 991a924..450e587 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
@@ -142,12 +142,12 @@ enum hclgevf_states {
HCLGEVF_STATE_REMOVING,
HCLGEVF_STATE_NIC_REGISTERED,
/* task states */
- HCLGEVF_STATE_SERVICE_SCHED,
HCLGEVF_STATE_RST_SERVICE_SCHED,
HCLGEVF_STATE_RST_HANDLING,
HCLGEVF_STATE_MBX_SERVICE_SCHED,
HCLGEVF_STATE_MBX_HANDLING,
HCLGEVF_STATE_CMD_DISABLE,
+ HCLGEVF_STATE_LINK_UPDATING,
};
struct hclgevf_mac {
@@ -283,11 +283,7 @@ struct hclgevf_dev {
struct hclgevf_mbx_resp_status mbx_resp; /* mailbox response */
struct hclgevf_mbx_arq_ring arq; /* mailbox async rx queue */
- struct timer_list keep_alive_timer;
struct delayed_work service_task;
- struct work_struct keep_alive_task;
- struct work_struct rst_service_task;
- struct work_struct mbx_service_task;
struct hclgevf_tqp *htqp;
@@ -297,7 +293,8 @@ struct hclgevf_dev {
struct hnae3_client *nic_client;
struct hnae3_client *roce_client;
u32 flag;
- u32 stats_timer;
+ unsigned long serv_processed_cnt;
+ unsigned long last_serv_processed;
};
static inline bool hclgevf_is_reset_pending(struct hclgevf_dev *hdev)
--
2.7.4
next prev parent reply other threads:[~2019-12-14 2:06 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-14 2:06 [PATCH net-next 0/5] net: hns3: some optimizaions related to work task Huazhong Tan
2019-12-14 2:06 ` [PATCH net-next 1/5] net: hns3: schedule hclgevf_service by using delayed workqueue Huazhong Tan
2019-12-14 2:06 ` [PATCH net-next 2/5] net: hns3: remove mailbox and reset work in hclge_main Huazhong Tan
2019-12-14 2:06 ` Huazhong Tan [this message]
2019-12-14 2:06 ` [PATCH net-next 4/5] net: hns3: allocate WQ with WQ_MEM_RECLAIM flag Huazhong Tan
2019-12-14 2:06 ` [PATCH net-next 5/5] net: hns3: do not schedule the periodic task when reset fail Huazhong Tan
2019-12-17 0:12 ` [PATCH net-next 0/5] net: hns3: some optimizaions related to work task David Miller
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1576289201-57017-4-git-send-email-tanhuazhong@huawei.com \
--to=tanhuazhong@huawei.com \
--cc=davem@davemloft.net \
--cc=jakub.kicinski@netronome.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxarm@huawei.com \
--cc=linyunsheng@huawei.com \
--cc=netdev@vger.kernel.org \
--cc=salil.mehta@huawei.com \
--cc=yisen.zhuang@huawei.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).