* Re: [PATCH bpf-next 0/5] Add support for SKIP_BPF flag for AF_XDP sockets
From: Samudrala, Sridhar @ 2019-08-16 6:25 UTC (permalink / raw)
To: Jakub Kicinski
Cc: magnus.karlsson, bjorn.topel, netdev, bpf, intel-wired-lan,
maciej.fijalkowski, tom.herbert
In-Reply-To: <20190815122844.52eeda08@cakuba.netronome.com>
On 8/15/2019 12:28 PM, Jakub Kicinski wrote:
> On Wed, 14 Aug 2019 20:46:18 -0700, Sridhar Samudrala wrote:
>> This patch series introduces XDP_SKIP_BPF flag that can be specified
>> during the bind() call of an AF_XDP socket to skip calling the BPF
>> program in the receive path and pass the buffer directly to the socket.
>>
>> When a single AF_XDP socket is associated with a queue and a HW
>> filter is used to redirect the packets and the app is interested in
>> receiving all the packets on that queue, we don't need an additional
>> BPF program to do further filtering or lookup/redirect to a socket.
>>
>> Here are some performance numbers collected on
>> - 2 socket 28 core Intel(R) Xeon(R) Platinum 8180 CPU @ 2.50GHz
>> - Intel 40Gb Ethernet NIC (i40e)
>>
>> All tests use 2 cores and the results are in Mpps.
>>
>> turbo on (default)
>> ---------------------------------------------
>> no-skip-bpf skip-bpf
>> ---------------------------------------------
>> rxdrop zerocopy 21.9 38.5
>> l2fwd zerocopy 17.0 20.5
>> rxdrop copy 11.1 13.3
>> l2fwd copy 1.9 2.0
>>
>> no turbo : echo 1 > /sys/devices/system/cpu/intel_pstate/no_turbo
>> ---------------------------------------------
>> no-skip-bpf skip-bpf
>> ---------------------------------------------
>> rxdrop zerocopy 15.4 29.0
>> l2fwd zerocopy 11.8 18.2
>> rxdrop copy 8.2 10.5
>> l2fwd copy 1.7 1.7
>> ---------------------------------------------
>
> Could you include a third column here - namely the in-XDP performance?
> AFAIU the way to achieve better performance with AF_XDP is to move the
> fast path into the kernel's XDP program..
The in-xdp drop that can be measured with xdp1 is lower than rxdrop
zerocopy with skip-bpf although in-xdp drop uses only 1 core. af-xdp
1-core performance would improve with need-wakeup or busypoll patches
and based on early experiments so far af-xdp with need-wakeup/busypoll +
skip-bpf perf is higher than in-xdp drop.
Will include in-xdp drop data too in the next revision.
>
> Maciej's work on batching XDP program's execution should lower the
> retpoline overhead, without leaning close to the bypass model.
>
^ permalink raw reply
* [PATCH] rtw88: pci: Move a mass of jobs in hw IRQ to soft IRQ
From: Jian-Hong Pan @ 2019-08-16 6:31 UTC (permalink / raw)
To: Yan-Hsuan Chuang, Kalle Valo, David S . Miller
Cc: linux-wireless, netdev, linux-kernel, linux, Jian-Hong Pan
There is a mass of jobs between spin lock and unlock in the hardware
IRQ which will occupy much time originally. To make system work more
efficiently, this patch moves the jobs to the soft IRQ (bottom half) to
reduce the time in hardware IRQ.
Signed-off-by: Jian-Hong Pan <jian-hong@endlessm.com>
---
drivers/net/wireless/realtek/rtw88/pci.c | 36 +++++++++++++++++++-----
1 file changed, 29 insertions(+), 7 deletions(-)
diff --git a/drivers/net/wireless/realtek/rtw88/pci.c b/drivers/net/wireless/realtek/rtw88/pci.c
index 00ef229552d5..355606b167c6 100644
--- a/drivers/net/wireless/realtek/rtw88/pci.c
+++ b/drivers/net/wireless/realtek/rtw88/pci.c
@@ -866,12 +866,29 @@ static irqreturn_t rtw_pci_interrupt_handler(int irq, void *dev)
{
struct rtw_dev *rtwdev = dev;
struct rtw_pci *rtwpci = (struct rtw_pci *)rtwdev->priv;
- u32 irq_status[4];
+ unsigned long flags;
- spin_lock(&rtwpci->irq_lock);
+ spin_lock_irqsave(&rtwpci->irq_lock, flags);
if (!rtwpci->irq_enabled)
goto out;
+ /* disable RTW PCI interrupt to avoid more interrupts before the end of
+ * thread function
+ */
+ rtw_pci_disable_interrupt(rtwdev, rtwpci);
+out:
+ spin_unlock_irqrestore(&rtwpci->irq_lock, flags);
+
+ return IRQ_WAKE_THREAD;
+}
+
+static irqreturn_t rtw_pci_interrupt_threadfn(int irq, void *dev)
+{
+ struct rtw_dev *rtwdev = dev;
+ struct rtw_pci *rtwpci = (struct rtw_pci *)rtwdev->priv;
+ unsigned long flags;
+ u32 irq_status[4];
+
rtw_pci_irq_recognized(rtwdev, rtwpci, irq_status);
if (irq_status[0] & IMR_MGNTDOK)
@@ -891,8 +908,11 @@ static irqreturn_t rtw_pci_interrupt_handler(int irq, void *dev)
if (irq_status[0] & IMR_ROK)
rtw_pci_rx_isr(rtwdev, rtwpci, RTW_RX_QUEUE_MPDU);
-out:
- spin_unlock(&rtwpci->irq_lock);
+ /* all of the jobs for this interrupt have been done */
+ spin_lock_irqsave(&rtwpci->irq_lock, flags);
+ if (rtw_flag_check(rtwdev, RTW_FLAG_RUNNING))
+ rtw_pci_enable_interrupt(rtwdev, rtwpci);
+ spin_unlock_irqrestore(&rtwpci->irq_lock, flags);
return IRQ_HANDLED;
}
@@ -1152,8 +1172,10 @@ static int rtw_pci_probe(struct pci_dev *pdev,
goto err_destroy_pci;
}
- ret = request_irq(pdev->irq, &rtw_pci_interrupt_handler,
- IRQF_SHARED, KBUILD_MODNAME, rtwdev);
+ ret = devm_request_threaded_irq(rtwdev->dev, pdev->irq,
+ rtw_pci_interrupt_handler,
+ rtw_pci_interrupt_threadfn,
+ IRQF_SHARED, KBUILD_MODNAME, rtwdev);
if (ret) {
ieee80211_unregister_hw(hw);
goto err_destroy_pci;
@@ -1192,7 +1214,7 @@ static void rtw_pci_remove(struct pci_dev *pdev)
rtw_pci_disable_interrupt(rtwdev, rtwpci);
rtw_pci_destroy(rtwdev, pdev);
rtw_pci_declaim(rtwdev, pdev);
- free_irq(rtwpci->pdev->irq, rtwdev);
+ devm_free_irq(rtwdev->dev, rtwpci->pdev->irq, rtwdev);
rtw_core_deinit(rtwdev);
ieee80211_free_hw(hw);
}
--
2.20.1
^ permalink raw reply related
* Re: [PATCH] can: flexcan: disable completely the ECC mechanism
From: Marc Kleine-Budde @ 2019-08-16 6:37 UTC (permalink / raw)
To: Joakim Zhang, linux-can@vger.kernel.org
Cc: wg@grandegger.com, netdev@vger.kernel.org, stefan@agner.ch,
dl-linux-imx
In-Reply-To: <20190815075806.23212-1-qiangqing.zhang@nxp.com>
[-- Attachment #1.1: Type: text/plain, Size: 799 bytes --]
On 8/15/19 10:00 AM, Joakim Zhang wrote:
> The ECC(memory error detection and correction) mechanism can be
> activated or not, controlled by the ECCDIS bit in CAN_MECR. When
> disabled, updates on indications and reporting registers are stopped.
> So if want to disable ECC completely, had better assert ECCDIS bit,
> not just mask the related interrupts.
>
> Fixes:cdce844865be("can: flexcan: add vf610 support for FlexCAN")
> Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Added to can.
tnx
Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply
* Re: [PATCH net-next] r8152: divide the tx and rx bottom functions
From: Eric Dumazet @ 2019-08-16 6:39 UTC (permalink / raw)
To: Hayes Wang, netdev; +Cc: nic_swsd, linux-kernel
In-Reply-To: <1394712342-15778-301-Taiwan-albertk@realtek.com>
On 8/14/19 10:30 AM, Hayes Wang wrote:
> Move the tx bottom function from NAPI to a new tasklet. Then, for
> multi-cores, the bottom functions of tx and rx may be run at same
> time with different cores. This is used to improve performance.
>
>
tasklet and NAPI are scheduled on the same core (the current
cpu calling napi_schedule() or tasklet_schedule())
I would rather not add this dubious tasklet, and instead try to understand
what is wrong in this driver ;)
The various napi_schedule() calls are suspect IMO.
Also rtl8152_start_xmit() uses skb_queue_tail(&tp->tx_queue, skb);
But I see nothing really kicking the transmit if tx_free is empty ?
tx_bottom() is only called from bottom_half() (called from r8152_poll())
^ permalink raw reply
* Re: [PATCH net-next v2 4/5] r8152: support skb_add_rx_frag
From: Eric Dumazet @ 2019-08-16 6:42 UTC (permalink / raw)
To: Hayes Wang, netdev; +Cc: nic_swsd, linux-kernel, linux-usb
In-Reply-To: <1394712342-15778-299-albertk@realtek.com>
On 8/13/19 5:42 AM, Hayes Wang wrote:
> Use skb_add_rx_frag() to reduce the memory copy for rx data.
>
> Use a new list of rx_used to store the rx buffer which couldn't be
> reused yet.
>
> Besides, the total number of rx buffer may be increased or decreased
> dynamically. And it is limited by RTL8152_MAX_RX_AGG.
>
> Signed-off-by: Hayes Wang <hayeswang@realtek.com>
>
...
> skb->protocol = eth_type_trans(skb, netdev);
> rtl_rx_vlan_tag(rx_desc, skb);
> if (work_done < budget) {
> napi_gro_receive(napi, skb);
> work_done++;
> stats->rx_packets++;
> - stats->rx_bytes += pkt_len;
> + stats->rx_bytes += skb->len;
use-after-free. skb is no longer in your hands after napi_gro_receive()
^ permalink raw reply
* Re: [PATCH] can: flexcan: update hardware feature information for i.MX8QM
From: Marc Kleine-Budde @ 2019-08-16 6:45 UTC (permalink / raw)
To: Joakim Zhang, linux-can@vger.kernel.org
Cc: wg@grandegger.com, netdev@vger.kernel.org, dl-linux-imx
In-Reply-To: <20190815075638.23148-1-qiangqing.zhang@nxp.com>
[-- Attachment #1.1: Type: text/plain, Size: 497 bytes --]
On 8/15/19 9:58 AM, Joakim Zhang wrote:
> Update hardware feature information for i.MX8QM.
>
> Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Squashed into "can: flexcan: add imx8qm support"
Tnx,
Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply
* RE: [PATCH V2] can: flexcan: fix deadlock when using self wakeup
From: Joakim Zhang @ 2019-08-16 6:47 UTC (permalink / raw)
To: mkl@pengutronix.de, linux-can@vger.kernel.org, sean@geanix.com
Cc: wg@grandegger.com, netdev@vger.kernel.org, dl-linux-imx
In-Reply-To: <20190730024834.31182-1-qiangqing.zhang@nxp.com>
Kindly Ping...
Best Regards,
Joakim Zhang
> -----Original Message-----
> From: Joakim Zhang
> Sent: 2019年7月30日 10:52
> To: mkl@pengutronix.de; linux-can@vger.kernel.org; sean@geanix.com
> Cc: wg@grandegger.com; netdev@vger.kernel.org; dl-linux-imx
> <linux-imx@nxp.com>; Joakim Zhang <qiangqing.zhang@nxp.com>
> Subject: [PATCH V2] can: flexcan: fix deadlock when using self wakeup
>
> As reproted by Sean Nyekjaer below:
> When suspending, when there is still can traffic on the interfaces the flexcan
> immediately wakes the platform again. As it should :-). But it throws this error
> msg:
> [ 3169.378661] PM: noirq suspend of devices failed
>
> On the way down to suspend the interface that throws the error message does
> call flexcan_suspend but fails to call flexcan_noirq_suspend. That means the
> flexcan_enter_stop_mode is called, but on the way out of suspend the driver
> only calls flexcan_resume and skips flexcan_noirq_resume, thus it doesn't call
> flexcan_exit_stop_mode. This leaves the flexcan in stop mode, and with the
> current driver it can't recover from this even with a soft reboot, it requires a
> hard reboot.
>
> The best way to exit stop mode is in Wake Up interrupt context, and then
> suspend() and resume() functions can be symmetric. However, stop mode
> request and ack will be controlled by SCU(System Control Unit)
> firmware(manage clock,power,stop mode, etc. by Cortex-M4 core) in coming
> i.MX8(QM/QXP). And SCU firmware interface can't be available in interrupt
> context.
>
> For compatibillity, the wake up mechanism can't be symmetric, so we need
> in_stop_mode hack.
>
> Fixes: de3578c198c6 ("can: flexcan: add self wakeup support")
> Reported-by: Sean Nyekjaer <sean@geanix.com>
> Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
>
> Changelog:
> V1->V2:
> * add Reported-by tag.
> * rebase on patch: can:flexcan:fix stop mode acknowledgment.
> ---
> drivers/net/can/flexcan.c | 23 +++++++++++++++++++++++
> 1 file changed, 23 insertions(+)
>
> diff --git a/drivers/net/can/flexcan.c b/drivers/net/can/flexcan.c index
> fcec8bcb53d6..1dbec868d3ea 100644
> --- a/drivers/net/can/flexcan.c
> +++ b/drivers/net/can/flexcan.c
> @@ -282,6 +282,7 @@ struct flexcan_priv {
> const struct flexcan_devtype_data *devtype_data;
> struct regulator *reg_xceiver;
> struct flexcan_stop_mode stm;
> + bool in_stop_mode;
>
> /* Read and Write APIs */
> u32 (*read)(void __iomem *addr);
> @@ -1635,6 +1636,8 @@ static int __maybe_unused flexcan_suspend(struct
> device *device)
> err = flexcan_enter_stop_mode(priv);
> if (err)
> return err;
> +
> + priv->in_stop_mode = true;
> } else {
> err = flexcan_chip_disable(priv);
> if (err)
> @@ -1659,6 +1662,15 @@ static int __maybe_unused flexcan_resume(struct
> device *device)
> netif_device_attach(dev);
> netif_start_queue(dev);
> if (device_may_wakeup(device)) {
> + if (priv->in_stop_mode) {
> + flexcan_enable_wakeup_irq(priv, false);
> + err = flexcan_exit_stop_mode(priv);
> + if (err)
> + return err;
> +
> + priv->in_stop_mode = false;
> + }
> +
> disable_irq_wake(dev->irq);
> } else {
> err = flexcan_chip_enable(priv);
> @@ -1674,6 +1686,11 @@ static int __maybe_unused
> flexcan_noirq_suspend(struct device *device)
> struct net_device *dev = dev_get_drvdata(device);
> struct flexcan_priv *priv = netdev_priv(dev);
>
> + /* Need to enable wakeup interrupt in noirq suspend stage. Otherwise,
> + * it will trigger continuously wakeup interrupt if the wakeup event
> + * comes before noirq suspend stage, and simultaneously it has enter
> + * the stop mode.
> + */
> if (netif_running(dev) && device_may_wakeup(device))
> flexcan_enable_wakeup_irq(priv, true);
>
> @@ -1686,11 +1703,17 @@ static int __maybe_unused
> flexcan_noirq_resume(struct device *device)
> struct flexcan_priv *priv = netdev_priv(dev);
> int err;
>
> + /* Need to exit stop mode in noirq resume stage. Otherwise, it will
> + * trigger continuously wakeup interrupt if the wakeup event comes,
> + * and simultaneously it has still in stop mode.
> + */
> if (netif_running(dev) && device_may_wakeup(device)) {
> flexcan_enable_wakeup_irq(priv, false);
> err = flexcan_exit_stop_mode(priv);
> if (err)
> return err;
> +
> + priv->in_stop_mode = false;
> }
>
> return 0;
> --
> 2.17.1
^ permalink raw reply
* RE: [PATCH] can: flexcan: add LPSR mode support for i.MX7D
From: Joakim Zhang @ 2019-08-16 6:47 UTC (permalink / raw)
To: mkl@pengutronix.de, linux-can@vger.kernel.org
Cc: wg@grandegger.com, netdev@vger.kernel.org, dl-linux-imx
In-Reply-To: <20190731055401.15454-1-qiangqing.zhang@nxp.com>
Kindly Ping...
Best Regards,
Joakim Zhang
> -----Original Message-----
> From: Joakim Zhang
> Sent: 2019年7月31日 13:57
> To: mkl@pengutronix.de; linux-can@vger.kernel.org
> Cc: wg@grandegger.com; netdev@vger.kernel.org; dl-linux-imx
> <linux-imx@nxp.com>; Joakim Zhang <qiangqing.zhang@nxp.com>
> Subject: [PATCH] can: flexcan: add LPSR mode support for i.MX7D
>
> For i.MX7D LPSR mode, the controller will lost power and got the configuration
> state lost after system resume back. (coming i.MX8QM/QXP will also
> completely power off the domain, the controller state will be lost and needs
> restore).
> So we need to set pinctrl state again and re-start chip to do re-configuration
> after resume.
>
> For wakeup case, it should not set pinctrl to sleep state by
> pinctrl_pm_select_sleep_state.
> For interface is not up before suspend case, we don't need re-configure as it
> will be configured by user later by interface up.
>
> Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
> ---
> drivers/net/can/flexcan.c | 21 ++++++++++++++-------
> 1 file changed, 14 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/net/can/flexcan.c b/drivers/net/can/flexcan.c index
> c21b3507123e..228d07e84ddc 100644
> --- a/drivers/net/can/flexcan.c
> +++ b/drivers/net/can/flexcan.c
> @@ -26,6 +26,7 @@
> #include <linux/platform_device.h>
> #include <linux/pm_runtime.h>
> #include <linux/regulator/consumer.h>
> +#include <linux/pinctrl/consumer.h>
> #include <linux/regmap.h>
>
> #define DRV_NAME "flexcan"
> @@ -1889,7 +1890,7 @@ static int __maybe_unused flexcan_suspend(struct
> device *device) {
> struct net_device *dev = dev_get_drvdata(device);
> struct flexcan_priv *priv = netdev_priv(dev);
> - int err = 0;
> + int err;
>
> if (netif_running(dev)) {
> /* if wakeup is enabled, enter stop mode @@ -1899,25 +1900,27
> @@ static int __maybe_unused flexcan_suspend(struct device *device)
> enable_irq_wake(dev->irq);
> flexcan_enter_stop_mode(priv);
> } else {
> - err = flexcan_chip_disable(priv);
> + flexcan_chip_stop(dev);
> +
> + err = pm_runtime_force_suspend(device);
> if (err)
> return err;
>
> - err = pm_runtime_force_suspend(device);
> + pinctrl_pm_select_sleep_state(device);
> }
> netif_stop_queue(dev);
> netif_device_detach(dev);
> }
> priv->can.state = CAN_STATE_SLEEPING;
>
> - return err;
> + return 0;
> }
>
> static int __maybe_unused flexcan_resume(struct device *device) {
> struct net_device *dev = dev_get_drvdata(device);
> struct flexcan_priv *priv = netdev_priv(dev);
> - int err = 0;
> + int err;
>
> priv->can.state = CAN_STATE_ERROR_ACTIVE;
> if (netif_running(dev)) {
> @@ -1926,15 +1929,19 @@ static int __maybe_unused
> flexcan_resume(struct device *device)
> if (device_may_wakeup(device)) {
> disable_irq_wake(dev->irq);
> } else {
> + pinctrl_pm_select_default_state(device);
> +
> err = pm_runtime_force_resume(device);
> if (err)
> return err;
>
> - err = flexcan_chip_enable(priv);
> + err = flexcan_chip_start(dev);
> + if (err)
> + return err;
> }
> }
>
> - return err;
> + return 0;
> }
>
> static int __maybe_unused flexcan_runtime_suspend(struct device *device)
> --
> 2.17.1
^ permalink raw reply
* Re: [PATCH] can: flexcan: add LPSR mode support for i.MX7D
From: Marc Kleine-Budde @ 2019-08-16 7:05 UTC (permalink / raw)
To: Joakim Zhang, linux-can@vger.kernel.org
Cc: wg@grandegger.com, netdev@vger.kernel.org, dl-linux-imx
In-Reply-To: <DB7PR04MB46188FBD9DC87E620BBA8838E6AF0@DB7PR04MB4618.eurprd04.prod.outlook.com>
[-- Attachment #1.1: Type: text/plain, Size: 442 bytes --]
On 8/16/19 8:47 AM, Joakim Zhang wrote:
> Kindly Ping...
Can you please rebase both PM related patches to linux-can/testing and
post them in a series.
Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply
* Re: [PATCH] can: rcar_can: Remove unused platform data support
From: Marc Kleine-Budde @ 2019-08-16 7:06 UTC (permalink / raw)
To: Geert Uytterhoeven, Wolfgang Grandegger, Sergei Shtylyov
Cc: David S . Miller, Wolfram Sang, linux-can, linux-renesas-soc,
netdev
In-Reply-To: <20190814092221.12959-1-geert+renesas@glider.be>
[-- Attachment #1.1: Type: text/plain, Size: 714 bytes --]
On 8/14/19 11:22 AM, Geert Uytterhoeven wrote:
> All R-Car platforms use DT for describing CAN controllers.
> R-Car CAN platform data support was never used in any upstream kernel.
>
> Move the Clock Select Register settings enum into the driver, and remove
> platform data support and the corresponding header file.
>
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Applied to can-next-testing.
tnx,
Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply
* [PATCH] be2net: eliminate enable field from be_aic_obj
From: Ivan Vecera @ 2019-08-16 7:15 UTC (permalink / raw)
To: netdev; +Cc: sathya.perla, poros, sriharsha.basavapatna
Adaptive coalescing is managed per adapter not per event queue so it
does not needed to store 'enable' flag for each event queue.
Signed-off-by: Ivan Vecera <ivecera@redhat.com>
---
drivers/net/ethernet/emulex/benet/be.h | 2 +-
drivers/net/ethernet/emulex/benet/be_ethtool.c | 7 ++++---
drivers/net/ethernet/emulex/benet/be_main.c | 7 ++++---
3 files changed, 9 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index f287b5da5546..cf3e6f2892ff 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -192,7 +192,6 @@ struct be_eq_obj {
} ____cacheline_aligned_in_smp;
struct be_aic_obj { /* Adaptive interrupt coalescing (AIC) info */
- bool enable;
u32 min_eqd; /* in usecs */
u32 max_eqd; /* in usecs */
u32 prev_eqd; /* in usecs */
@@ -589,6 +588,7 @@ struct be_adapter {
struct be_drv_stats drv_stats;
struct be_aic_obj aic_obj[MAX_EVT_QS];
+ bool aic_enabled;
u8 vlan_prio_bmap; /* Available Priority BitMap */
u16 recommended_prio_bits;/* Recommended Priority bits in vlan tag */
struct be_dma_mem rx_filter; /* Cmd DMA mem for rx-filter */
diff --git a/drivers/net/ethernet/emulex/benet/be_ethtool.c b/drivers/net/ethernet/emulex/benet/be_ethtool.c
index 492f8769ac12..5bb5abf99588 100644
--- a/drivers/net/ethernet/emulex/benet/be_ethtool.c
+++ b/drivers/net/ethernet/emulex/benet/be_ethtool.c
@@ -329,8 +329,8 @@ static int be_get_coalesce(struct net_device *netdev,
et->tx_coalesce_usecs_high = aic->max_eqd;
et->tx_coalesce_usecs_low = aic->min_eqd;
- et->use_adaptive_rx_coalesce = aic->enable;
- et->use_adaptive_tx_coalesce = aic->enable;
+ et->use_adaptive_rx_coalesce = adapter->aic_enabled;
+ et->use_adaptive_tx_coalesce = adapter->aic_enabled;
return 0;
}
@@ -346,8 +346,9 @@ static int be_set_coalesce(struct net_device *netdev,
struct be_eq_obj *eqo;
int i;
+ adapter->aic_enabled = et->use_adaptive_rx_coalesce;
+
for_all_evt_queues(adapter, eqo, i) {
- aic->enable = et->use_adaptive_rx_coalesce;
aic->max_eqd = min(et->rx_coalesce_usecs_high, BE_MAX_EQD);
aic->min_eqd = min(et->rx_coalesce_usecs_low, aic->max_eqd);
aic->et_eqd = min(et->rx_coalesce_usecs, aic->max_eqd);
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 314e9868b861..39eb7d525043 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -2147,7 +2147,7 @@ static int be_get_new_eqd(struct be_eq_obj *eqo)
int i;
aic = &adapter->aic_obj[eqo->idx];
- if (!aic->enable) {
+ if (!adapter->aic_enabled) {
if (aic->jiffies)
aic->jiffies = 0;
eqd = aic->et_eqd;
@@ -2204,7 +2204,7 @@ static u32 be_get_eq_delay_mult_enc(struct be_eq_obj *eqo)
int eqd;
u32 mult_enc;
- if (!aic->enable)
+ if (!adapter->aic_enabled)
return 0;
if (jiffies_to_msecs(now - aic->jiffies) < 1)
@@ -2959,6 +2959,8 @@ static int be_evt_queues_create(struct be_adapter *adapter)
max(adapter->cfg_num_rx_irqs,
adapter->cfg_num_tx_irqs));
+ adapter->aic_enabled = true;
+
for_all_evt_queues(adapter, eqo, i) {
int numa_node = dev_to_node(&adapter->pdev->dev);
@@ -2966,7 +2968,6 @@ static int be_evt_queues_create(struct be_adapter *adapter)
eqo->adapter = adapter;
eqo->idx = i;
aic->max_eqd = BE_MAX_EQD;
- aic->enable = true;
eq = &eqo->q;
rc = be_queue_alloc(adapter, eq, EVNT_Q_LEN,
--
2.21.0
^ permalink raw reply related
* RE: [PATCH] rtw88: pci: Move a mass of jobs in hw IRQ to soft IRQ
From: Tony Chuang @ 2019-08-16 7:54 UTC (permalink / raw)
To: Jian-Hong Pan, Kalle Valo, David S . Miller
Cc: linux-wireless@vger.kernel.org, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, linux@endlessm.com
In-Reply-To: <20190816063109.4699-1-jian-hong@endlessm.com>
> From: Jian-Hong Pan [mailto:jian-hong@endlessm.com]
>
> There is a mass of jobs between spin lock and unlock in the hardware
> IRQ which will occupy much time originally. To make system work more
> efficiently, this patch moves the jobs to the soft IRQ (bottom half) to
> reduce the time in hardware IRQ.
>
> Signed-off-by: Jian-Hong Pan <jian-hong@endlessm.com>
> ---
> drivers/net/wireless/realtek/rtw88/pci.c | 36 +++++++++++++++++++-----
> 1 file changed, 29 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/net/wireless/realtek/rtw88/pci.c
> b/drivers/net/wireless/realtek/rtw88/pci.c
> index 00ef229552d5..355606b167c6 100644
> --- a/drivers/net/wireless/realtek/rtw88/pci.c
> +++ b/drivers/net/wireless/realtek/rtw88/pci.c
> @@ -866,12 +866,29 @@ static irqreturn_t rtw_pci_interrupt_handler(int irq,
> void *dev)
> {
> struct rtw_dev *rtwdev = dev;
> struct rtw_pci *rtwpci = (struct rtw_pci *)rtwdev->priv;
> - u32 irq_status[4];
> + unsigned long flags;
>
> - spin_lock(&rtwpci->irq_lock);
> + spin_lock_irqsave(&rtwpci->irq_lock, flags);
> if (!rtwpci->irq_enabled)
> goto out;
>
> + /* disable RTW PCI interrupt to avoid more interrupts before the end of
> + * thread function
> + */
> + rtw_pci_disable_interrupt(rtwdev, rtwpci);
> +out:
> + spin_unlock_irqrestore(&rtwpci->irq_lock, flags);
> +
> + return IRQ_WAKE_THREAD;
> +}
> +
> +static irqreturn_t rtw_pci_interrupt_threadfn(int irq, void *dev)
> +{
> + struct rtw_dev *rtwdev = dev;
> + struct rtw_pci *rtwpci = (struct rtw_pci *)rtwdev->priv;
> + unsigned long flags;
> + u32 irq_status[4];
> +
> rtw_pci_irq_recognized(rtwdev, rtwpci, irq_status);
>
> if (irq_status[0] & IMR_MGNTDOK)
> @@ -891,8 +908,11 @@ static irqreturn_t rtw_pci_interrupt_handler(int irq,
> void *dev)
> if (irq_status[0] & IMR_ROK)
> rtw_pci_rx_isr(rtwdev, rtwpci, RTW_RX_QUEUE_MPDU);
>
> -out:
> - spin_unlock(&rtwpci->irq_lock);
> + /* all of the jobs for this interrupt have been done */
> + spin_lock_irqsave(&rtwpci->irq_lock, flags);
Shouldn't we protect the ISRs above?
This patch could actually reduce the time of IRQ.
But I think I need to further test it with PCI MSI interrupt.
https://patchwork.kernel.org/patch/11081539/
Maybe we could drop the "rtw_pci_[enable/disable]_interrupt" when MSI
Is enabled with this patch.
> + if (rtw_flag_check(rtwdev, RTW_FLAG_RUNNING))
> + rtw_pci_enable_interrupt(rtwdev, rtwpci);
> + spin_unlock_irqrestore(&rtwpci->irq_lock, flags);
>
> return IRQ_HANDLED;
> }
> @@ -1152,8 +1172,10 @@ static int rtw_pci_probe(struct pci_dev *pdev,
> goto err_destroy_pci;
> }
>
> - ret = request_irq(pdev->irq, &rtw_pci_interrupt_handler,
> - IRQF_SHARED, KBUILD_MODNAME, rtwdev);
> + ret = devm_request_threaded_irq(rtwdev->dev, pdev->irq,
> + rtw_pci_interrupt_handler,
> + rtw_pci_interrupt_threadfn,
> + IRQF_SHARED, KBUILD_MODNAME, rtwdev);
> if (ret) {
> ieee80211_unregister_hw(hw);
> goto err_destroy_pci;
> @@ -1192,7 +1214,7 @@ static void rtw_pci_remove(struct pci_dev *pdev)
> rtw_pci_disable_interrupt(rtwdev, rtwpci);
> rtw_pci_destroy(rtwdev, pdev);
> rtw_pci_declaim(rtwdev, pdev);
> - free_irq(rtwpci->pdev->irq, rtwdev);
> + devm_free_irq(rtwdev->dev, rtwpci->pdev->irq, rtwdev);
> rtw_core_deinit(rtwdev);
> ieee80211_free_hw(hw);
> }
> --
> 2.20.1
Yan-Hsuan
^ permalink raw reply
* RE: [PATCH] rtw88: pci: Move a mass of jobs in hw IRQ to soft IRQ
From: Tony Chuang @ 2019-08-16 8:06 UTC (permalink / raw)
To: Tony Chuang, Jian-Hong Pan, Kalle Valo, David S . Miller
Cc: linux-wireless@vger.kernel.org, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, linux@endlessm.com
In-Reply-To: <F7CD281DE3E379468C6D07993EA72F84D18929BF@RTITMBSVM04.realtek.com.tw>
Hi,
A few more questions below
> > From: Jian-Hong Pan [mailto:jian-hong@endlessm.com]
> >
> > There is a mass of jobs between spin lock and unlock in the hardware
> > IRQ which will occupy much time originally. To make system work more
> > efficiently, this patch moves the jobs to the soft IRQ (bottom half) to
> > reduce the time in hardware IRQ.
> >
> > Signed-off-by: Jian-Hong Pan <jian-hong@endlessm.com>
> > ---
> > drivers/net/wireless/realtek/rtw88/pci.c | 36 +++++++++++++++++++-----
> > 1 file changed, 29 insertions(+), 7 deletions(-)
> >
> > diff --git a/drivers/net/wireless/realtek/rtw88/pci.c
> > b/drivers/net/wireless/realtek/rtw88/pci.c
> > index 00ef229552d5..355606b167c6 100644
> > --- a/drivers/net/wireless/realtek/rtw88/pci.c
> > +++ b/drivers/net/wireless/realtek/rtw88/pci.c
> > @@ -866,12 +866,29 @@ static irqreturn_t rtw_pci_interrupt_handler(int
> irq,
> > void *dev)
> > {
> > struct rtw_dev *rtwdev = dev;
> > struct rtw_pci *rtwpci = (struct rtw_pci *)rtwdev->priv;
> > - u32 irq_status[4];
> > + unsigned long flags;
> >
> > - spin_lock(&rtwpci->irq_lock);
> > + spin_lock_irqsave(&rtwpci->irq_lock, flags);
I think you can use 'spin_lock()' here as it's in IRQ context?
> > if (!rtwpci->irq_enabled)
> > goto out;
> >
> > + /* disable RTW PCI interrupt to avoid more interrupts before the end of
> > + * thread function
> > + */
> > + rtw_pci_disable_interrupt(rtwdev, rtwpci);
Why do we need rtw_pci_disable_interrupt() here.
Have you done any experiment and decided to add this.
If you have can you share your results to me?
> > +out:
> > + spin_unlock_irqrestore(&rtwpci->irq_lock, flags);
spin_unlock()
> > +
> > + return IRQ_WAKE_THREAD;
> > +}
> > +
> > +static irqreturn_t rtw_pci_interrupt_threadfn(int irq, void *dev)
> > +{
> > + struct rtw_dev *rtwdev = dev;
> > + struct rtw_pci *rtwpci = (struct rtw_pci *)rtwdev->priv;
> > + unsigned long flags;
> > + u32 irq_status[4];
> > +
> > rtw_pci_irq_recognized(rtwdev, rtwpci, irq_status);
> >
> > if (irq_status[0] & IMR_MGNTDOK)
> > @@ -891,8 +908,11 @@ static irqreturn_t rtw_pci_interrupt_handler(int
> irq,
> > void *dev)
> > if (irq_status[0] & IMR_ROK)
> > rtw_pci_rx_isr(rtwdev, rtwpci, RTW_RX_QUEUE_MPDU);
> >
> > -out:
> > - spin_unlock(&rtwpci->irq_lock);
> > + /* all of the jobs for this interrupt have been done */
> > + spin_lock_irqsave(&rtwpci->irq_lock, flags);
>
> Shouldn't we protect the ISRs above?
>
> This patch could actually reduce the time of IRQ.
> But I think I need to further test it with PCI MSI interrupt.
> https://patchwork.kernel.org/patch/11081539/
>
> Maybe we could drop the "rtw_pci_[enable/disable]_interrupt" when MSI
> Is enabled with this patch.
>
> > + if (rtw_flag_check(rtwdev, RTW_FLAG_RUNNING))
> > + rtw_pci_enable_interrupt(rtwdev, rtwpci);
> > + spin_unlock_irqrestore(&rtwpci->irq_lock, flags);
> >
> > return IRQ_HANDLED;
> > }
> > @@ -1152,8 +1172,10 @@ static int rtw_pci_probe(struct pci_dev *pdev,
> > goto err_destroy_pci;
> > }
> >
> > - ret = request_irq(pdev->irq, &rtw_pci_interrupt_handler,
> > - IRQF_SHARED, KBUILD_MODNAME, rtwdev);
> > + ret = devm_request_threaded_irq(rtwdev->dev, pdev->irq,
> > + rtw_pci_interrupt_handler,
> > + rtw_pci_interrupt_threadfn,
> > + IRQF_SHARED, KBUILD_MODNAME, rtwdev);
> > if (ret) {
> > ieee80211_unregister_hw(hw);
> > goto err_destroy_pci;
> > @@ -1192,7 +1214,7 @@ static void rtw_pci_remove(struct pci_dev
> *pdev)
> > rtw_pci_disable_interrupt(rtwdev, rtwpci);
> > rtw_pci_destroy(rtwdev, pdev);
> > rtw_pci_declaim(rtwdev, pdev);
> > - free_irq(rtwpci->pdev->irq, rtwdev);
> > + devm_free_irq(rtwdev->dev, rtwpci->pdev->irq, rtwdev);
> > rtw_core_deinit(rtwdev);
> > ieee80211_free_hw(hw);
> > }
> > --
> > 2.20.1
>
> Yan-Hsuan
>
Thanks
Yan-Hsuan
^ permalink raw reply
* Re: [PATCH net] batman-adv: fix uninit-value in batadv_netlink_get_ifindex()
From: Simon Wunderlich @ 2019-08-16 8:07 UTC (permalink / raw)
To: David Miller; +Cc: edumazet, netdev, eric.dumazet, syzkaller, mareklindner, a
In-Reply-To: <20190814.123625.170482147366456100.davem@davemloft.net>
[-- Attachment #1: Type: text/plain, Size: 605 bytes --]
Hi David,
yes, you will get it through me. I'll schedule a pull request next week.
Cheers,
Simon
On Wednesday, August 14, 2019 6:36:25 PM CEST David Miller wrote:
> From: Eric Dumazet <edumazet@google.com>
> Date: Mon, 12 Aug 2019 04:57:27 -0700
>
> > batadv_netlink_get_ifindex() needs to make sure user passed
> > a correct u32 attribute.
>
> ...
>
> > Fixes: b60620cf567b ("batman-adv: netlink: hardif query")
> > Signed-off-by: Eric Dumazet <edumazet@google.com>
> > Reported-by: syzbot <syzkaller@googlegroups.com>
>
> Simon, I assume I will get this ultimately from you.
>
> Thanks.
[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
^ permalink raw reply
* RE: [PATCH net-next] r8152: divide the tx and rx bottom functions
From: Hayes Wang @ 2019-08-16 8:10 UTC (permalink / raw)
To: Eric Dumazet, netdev@vger.kernel.org
Cc: nic_swsd, linux-kernel@vger.kernel.org
In-Reply-To: <9749764d-7815-b673-0fc4-22475601efec@gmail.com>
Eric Dumazet [mailto:eric.dumazet@gmail.com]
> Sent: Friday, August 16, 2019 2:40 PM
[...]
> tasklet and NAPI are scheduled on the same core (the current
> cpu calling napi_schedule() or tasklet_schedule())
>
> I would rather not add this dubious tasklet, and instead try to understand
> what is wrong in this driver ;)
>
> The various napi_schedule() calls are suspect IMO.
The original method as following.
static int r8152_poll(struct napi_struct *napi, int budget)
{
struct r8152 *tp = container_of(napi, struct r8152, napi);
int work_done;
work_done = rx_bottom(tp, budget); <-- RX
bottom_half(tp); <-- Tx (tx_bottom)
[...]
The rx_bottom and tx_bottom would only be called in r8152_poll.
That is, tx_bottom wouldn't be run unless rx_bottom is finished.
And, rx_bottom would be called if tx_bottom is running.
If the traffic is busy. rx_bottom or tx_bottom may take a lot
of time to deal with the packets. And the one would increase
the latency time for the other one.
Therefore, when I separate the tx_bottom and rx_bottom to
different tasklet and napi, the callback functions of tx and
rx may schedule the tasklet and napi to different cpu. Then,
the rx_bottom and tx_bottom may be run at the same time.
Take our arm platform for example. There are five cpus to
handle the interrupt of USB host controller. When the rx is
completed, cpu #1 may handle the interrupt and napi would
be scheduled. When the tx is finished, cpu #2 may handle
the interrupt and the tasklet is scheduled. Then, napi is
run on cpu #1, and tasklet is run on cpu #2.
> Also rtl8152_start_xmit() uses skb_queue_tail(&tp->tx_queue, skb);
>
> But I see nothing really kicking the transmit if tx_free is empty ?
Tx callback function "write_bulk_callback" would deal with it.
The callback function would check if there are packets waiting
to be sent.
Best Regards,
Hayes
^ permalink raw reply
* [PATCH net-next 2/6] net: hns3: modify redundant initialization of variable
From: Huazhong Tan @ 2019-08-16 8:09 UTC (permalink / raw)
To: davem
Cc: netdev, linux-kernel, salil.mehta, yisen.zhuang, linuxarm,
jakub.kicinski, Guojia Liao, Guangbin Huang, Huzhong Tan
In-Reply-To: <1565942982-12105-1-git-send-email-tanhuazhong@huawei.com>
From: Guojia Liao <liaoguojia@huawei.com>
Some temporary variables do not need to be initialized that
they will be set before used, so this patch deletes the
initialization value of these temporary variables.
Signed-off-by: Guojia Liao <liaoguojia@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huzhong Tan <tanhuazhong@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.c | 7 +++----
drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c | 2 +-
6 files changed, 8 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.c b/drivers/net/ethernet/hisilicon/hns3/hnae3.c
index 908d4f4..6bbba15 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.c
@@ -104,7 +104,6 @@ int hnae3_register_client(struct hnae3_client *client)
{
struct hnae3_client *client_tmp;
struct hnae3_ae_dev *ae_dev;
- int ret = 0;
if (!client)
return -ENODEV;
@@ -123,7 +122,7 @@ int hnae3_register_client(struct hnae3_client *client)
/* if the client could not be initialized on current port, for
* any error reasons, move on to next available port
*/
- ret = hnae3_init_client_instance(client, ae_dev);
+ int ret = hnae3_init_client_instance(client, ae_dev);
if (ret)
dev_err(&ae_dev->pdev->dev,
"match and instantiation failed for port, ret = %d\n",
@@ -164,7 +163,7 @@ void hnae3_register_ae_algo(struct hnae3_ae_algo *ae_algo)
const struct pci_device_id *id;
struct hnae3_ae_dev *ae_dev;
struct hnae3_client *client;
- int ret = 0;
+ int ret;
if (!ae_algo)
return;
@@ -258,7 +257,7 @@ int hnae3_register_ae_dev(struct hnae3_ae_dev *ae_dev)
const struct pci_device_id *id;
struct hnae3_ae_algo *ae_algo;
struct hnae3_client *client;
- int ret = 0;
+ int ret;
if (!ae_dev)
return -ENODEV;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
index 677bfe06..4260653 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
@@ -749,7 +749,7 @@ static int hns3_set_link_ksettings(struct net_device *netdev,
{
struct hnae3_handle *handle = hns3_get_handle(netdev);
const struct hnae3_ae_ops *ops = handle->ae_algo->ops;
- int ret = 0;
+ int ret;
/* Chip don't support this mode. */
if (cmd->base.speed == SPEED_1000 && cmd->base.duplex == DUPLEX_HALF)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
index 05a4cdb..5ce9a8a 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
@@ -1557,8 +1557,8 @@ int hclge_config_rocee_ras_interrupt(struct hclge_dev *hdev, bool en)
static void hclge_handle_rocee_ras_error(struct hnae3_ae_dev *ae_dev)
{
- enum hnae3_reset_type reset_type = HNAE3_NONE_RESET;
struct hclge_dev *hdev = ae_dev->priv;
+ enum hnae3_reset_type reset_type;
if (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state) ||
hdev->pdev->revision < 0x21)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
index f30d112..e829101 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
@@ -404,8 +404,8 @@ static int hclge_tm_port_shaper_cfg(struct hclge_dev *hdev)
{
struct hclge_port_shapping_cmd *shap_cfg_cmd;
struct hclge_desc desc;
- u32 shapping_para = 0;
u8 ir_u, ir_b, ir_s;
+ u32 shapping_para;
int ret;
ret = hclge_shaper_para_calc(hdev->hw.mac.speed,
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
index 55d3c78..4c2c945 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
@@ -43,7 +43,7 @@ static int hclgevf_cmd_csq_clean(struct hclgevf_hw *hw)
{
struct hclgevf_dev *hdev = container_of(hw, struct hclgevf_dev, hw);
struct hclgevf_cmq_ring *csq = &hw->cmq.csq;
- int clean = 0;
+ int clean;
u32 head;
head = hclgevf_read_dev(hw, HCLGEVF_NIC_CSQ_HEAD_REG);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
index defc905..594cae8 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
@@ -2302,7 +2302,7 @@ static void hclgevf_uninit_msi(struct hclgevf_dev *hdev)
static int hclgevf_misc_irq_init(struct hclgevf_dev *hdev)
{
- int ret = 0;
+ int ret;
hclgevf_get_misc_vector(hdev);
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 1/6] net: hns3: add or modify comments
From: Huazhong Tan @ 2019-08-16 8:09 UTC (permalink / raw)
To: davem
Cc: netdev, linux-kernel, salil.mehta, yisen.zhuang, linuxarm,
jakub.kicinski, Guojia Liao, Zhongzhu Liu, Weihang Li,
Guangbin Huang, Huazhong Tan
In-Reply-To: <1565942982-12105-1-git-send-email-tanhuazhong@huawei.com>
From: Guojia Liao <liaoguojia@huawei.com>
To explain some code, this patch adds some comments, and modifies or
merges some comments to make them more neat.
Signed-off-by: Guojia Liao <liaoguojia@huawei.com>
Signed-off-by: Zhongzhu Liu <liuzhongzhu@huawei.com>
Signed-off-by: Weihang Li <liweihang@hisilicon.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 12 ++++++------
drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 8 ++++----
drivers/net/ethernet/hisilicon/hns3/hns3_enet.h | 8 ++++----
drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c | 3 +++
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 8 ++++----
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c | 9 ++++-----
8 files changed, 27 insertions(+), 25 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index 6c9fd58..3e21533 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -85,10 +85,10 @@ struct hnae3_queue {
void __iomem *io_base;
struct hnae3_ae_algo *ae_algo;
struct hnae3_handle *handle;
- int tqp_index; /* index in a handle */
- u32 buf_size; /* size for hnae_desc->addr, preset by AE */
- u16 tx_desc_num;/* total number of tx desc */
- u16 rx_desc_num;/* total number of rx desc */
+ int tqp_index; /* index in a handle */
+ u32 buf_size; /* size for hnae_desc->addr, preset by AE */
+ u16 tx_desc_num; /* total number of tx desc */
+ u16 rx_desc_num; /* total number of rx desc */
};
struct hns3_mac_stats {
@@ -96,7 +96,7 @@ struct hns3_mac_stats {
u64 rx_pause_cnt;
};
-/*hnae3 loop mode*/
+/* hnae3 loop mode */
enum hnae3_loop {
HNAE3_LOOP_APP,
HNAE3_LOOP_SERIAL_SERDES,
@@ -621,7 +621,7 @@ struct hnae3_handle {
struct pci_dev *pdev;
void *priv;
struct hnae3_ae_algo *ae_algo; /* the class who provides this handle */
- u64 flags; /* Indicate the capabilities for this handle*/
+ u64 flags; /* Indicate the capabilities for this handle */
union {
struct net_device *netdev; /* first member */
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 1750f80..a11d514 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -229,9 +229,9 @@ static void hns3_vector_gl_rl_init(struct hns3_enet_tqp_vector *tqp_vector,
/* initialize the configuration for interrupt coalescing.
* 1. GL (Interrupt Gap Limiter)
* 2. RL (Interrupt Rate Limiter)
+ *
+ * Default: enable interrupt coalescing self-adaptive and GL
*/
-
- /* Default: enable interrupt coalescing self-adaptive and GL */
tqp_vector->tx_group.coal.gl_adapt_enable = 1;
tqp_vector->rx_group.coal.gl_adapt_enable = 1;
@@ -4207,8 +4207,8 @@ int hns3_nic_reset_all_ring(struct hnae3_handle *h)
static void hns3_store_coal(struct hns3_nic_priv *priv)
{
/* ethtool only support setting and querying one coal
- * configuation for now, so save the vector 0' coal
- * configuation here in order to restore it.
+ * configuration for now, so save the vector 0' coal
+ * configuration here in order to restore it.
*/
memcpy(&priv->tx_coal, &priv->tqp_vector[0].tx_group.coal,
sizeof(struct hns3_enet_coalesce));
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
index 5b0ee1f..e37e64e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
@@ -302,7 +302,7 @@ struct hns3_desc_cb {
dma_addr_t dma; /* dma address of this desc */
void *buf; /* cpu addr for a desc */
- /* priv data for the desc, e.g. skb when use with ip stack*/
+ /* priv data for the desc, e.g. skb when use with ip stack */
void *priv;
u32 page_offset;
u32 length; /* length of the buffer */
@@ -325,11 +325,11 @@ enum hns3_pkt_l3type {
HNS3_L3_TYPE_MAC_PAUSE,
HNS3_L3_TYPE_PFC_PAUSE,/* 0x9*/
- /* reserved for 0xA~0xB*/
+ /* reserved for 0xA~0xB */
HNS3_L3_TYPE_CNM = 0xc,
- /* reserved for 0xD~0xE*/
+ /* reserved for 0xD~0xE */
HNS3_L3_TYPE_PARSE_FAIL = 0xf /* must be last */
};
@@ -354,7 +354,7 @@ enum hns3_pkt_ol3type {
HNS3_OL3_TYPE_IPV4_OPT = 4,
HNS3_OL3_TYPE_IPV6_EXT,
- /* reserved for 0x6~0xE*/
+ /* reserved for 0x6~0xE */
HNS3_OL3_TYPE_PARSE_FAIL = 0xf /* must be last */
};
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
index 185ff32..677bfe06 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
@@ -635,7 +635,7 @@ static void hns3_get_ksettings(struct hnae3_handle *h,
&cmd->base.speed,
&cmd->base.duplex);
- /* 2.get link mode*/
+ /* 2.get link mode */
if (ops->get_link_mode)
ops->get_link_mode(h,
cmd->link_modes.supported,
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
index f0295d1..025184a 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
@@ -897,14 +897,17 @@ static void hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage,
dev_info(&hdev->pdev->dev, " read result tcam key %s(%u):\n",
sel_x ? "x" : "y", loc);
+ /* tcam_data0 ~ tcam_data1 */
req = (u32 *)req1->tcam_data;
for (i = 0; i < 2; i++)
dev_info(&hdev->pdev->dev, "%08x\n", *req++);
+ /* tcam_data2 ~ tcam_data7 */
req = (u32 *)req2->tcam_data;
for (i = 0; i < 6; i++)
dev_info(&hdev->pdev->dev, "%08x\n", *req++);
+ /* tcam_data8 ~ tcam_data12 */
req = (u32 *)req3->tcam_data;
for (i = 0; i < 5; i++)
dev_info(&hdev->pdev->dev, "%08x\n", *req++);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index a3ca0e6..1d8dee1 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -2810,9 +2810,9 @@ static u32 hclge_check_event_cause(struct hclge_dev *hdev, u32 *clearval)
* defer the processing of the mailbox events. Since, we would have not
* cleared RX CMDQ event this time we would receive again another
* interrupt from H/W just for the mailbox.
+ *
+ * check for vector0 reset event sources
*/
-
- /* check for vector0 reset event sources */
if (BIT(HCLGE_VECTOR0_IMPRESET_INT_B) & rst_src_reg) {
dev_info(&hdev->pdev->dev, "IMP reset interrupt\n");
set_bit(HNAE3_IMP_RESET, &hdev->reset_pending);
@@ -8000,7 +8000,7 @@ int hclge_set_vlan_filter(struct hnae3_handle *handle, __be16 proto,
return -EBUSY;
}
- /* When port base vlan enabled, we use port base vlan as the vlan
+ /* when port base vlan enabled, we use port base vlan as the vlan
* filter entry. In this case, we don't update vlan filter table
* when user add new vlan or remove exist vlan, just update the vport
* vlan list. The vlan id in vlan list will be writen in vlan filter
@@ -8019,7 +8019,7 @@ int hclge_set_vlan_filter(struct hnae3_handle *handle, __be16 proto,
hclge_add_vport_vlan_table(vport, vlan_id,
writen_to_tbl);
} else if (is_kill) {
- /* When remove hw vlan filter failed, record the vlan id,
+ /* when remove hw vlan filter failed, record the vlan id,
* and try to remove it from hw later, to be consistence
* with stack
*/
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
index 5a7221e..f5da28a 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
@@ -479,7 +479,7 @@ static void hclge_mbx_reset_vf_queue(struct hclge_vport *vport,
hclge_reset_vf_queue(vport, queue_id);
- /* send response msg to VF after queue reset complete*/
+ /* send response msg to VF after queue reset complete */
hclge_gen_resp_to_vf(vport, mbx_req, 0, NULL, 0);
}
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
index d8b8281..defc905 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
@@ -1269,7 +1269,7 @@ static int hclgevf_set_vlan_filter(struct hnae3_handle *handle,
HCLGE_MBX_VLAN_FILTER, msg_data,
HCLGEVF_VLAN_MBX_MSG_LEN, false, NULL, 0);
- /* When remove hw vlan filter failed, record the vlan id,
+ /* when remove hw vlan filter failed, record the vlan id,
* and try to remove it from hw later, to be consistence
* with stack.
*/
@@ -1561,7 +1561,7 @@ static int hclgevf_reset(struct hclgevf_dev *hdev)
rtnl_lock();
- /* now, re-initialize the nic client and ae device*/
+ /* now, re-initialize the nic client and ae device */
ret = hclgevf_reset_stack(hdev);
if (ret) {
dev_err(&hdev->pdev->dev, "failed to reset VF stack\n");
@@ -1784,9 +1784,8 @@ static void hclgevf_reset_service_task(struct work_struct *work)
* 1b and 2. cases but we will not get any intimation about 1a
* from PF as cmdq would be in unreliable state i.e. mailbox
* communication between PF and VF would be broken.
- */
-
- /* if we are never geting into pending state it means either:
+ *
+ * if we are never geting into pending state it means either:
* 1. PF is not receiving our request which could be due to IMP
* reset
* 2. PF is screwed
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 5/6] net: hns3: prevent unnecessary MAC TNL interrupt
From: Huazhong Tan @ 2019-08-16 8:09 UTC (permalink / raw)
To: davem
Cc: netdev, linux-kernel, salil.mehta, yisen.zhuang, linuxarm,
jakub.kicinski, Huazhong Tan
In-Reply-To: <1565942982-12105-1-git-send-email-tanhuazhong@huawei.com>
MAC TNL interrupt is used to collect statistic info about
link status changing suddenly when netdev is running.
But when stopping netdev, the enabled MAC TNL interrupt is
unnecessary, and may add some noises to the statistic info.
So this patch disables it before stopping MAC.
Fixes: a63457878b12 ("net: hns3: Add handling of MAC tunnel interruption")
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Reviewed-by: Yunsheng Lin <linyunsheng@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 24b59f0..9d64c43 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -6380,6 +6380,8 @@ static void hclge_ae_stop(struct hnae3_handle *handle)
for (i = 0; i < handle->kinfo.num_tqps; i++)
hclge_reset_tqp(handle, i);
+ hclge_config_mac_tnl_int(hdev, false);
+
/* Mac disable */
hclge_cfg_mac_mode(hdev, false);
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 3/6] net: hns3: fix error and incorrect format
From: Huazhong Tan @ 2019-08-16 8:09 UTC (permalink / raw)
To: davem
Cc: netdev, linux-kernel, salil.mehta, yisen.zhuang, linuxarm,
jakub.kicinski, Guojia Liao, Weihang Li, Jian Shen,
Guangbin Huang, Huazhong Tan
In-Reply-To: <1565942982-12105-1-git-send-email-tanhuazhong@huawei.com>
From: Guojia Liao <liaoguojia@huawei.com>
The pointer type parameter should be declare as const for preventing
from its pointed value being unexpected modified.
The uninitialized variable can not be return directly. The default
return value is 0 if no abnormal result.
This patch fixes the preceding two errors, deletes redundant
declaration of a function and align one parameter.
Signed-off-by: Guojia Liao <liaoguojia@huawei.com>
Signed-off-by: Weihang Li <liweihang@hisilicon.com>
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.c | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 7 ++++---
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 1 -
4 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.c b/drivers/net/ethernet/hisilicon/hns3/hnae3.c
index 6bbba15..528f624 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.c
@@ -46,7 +46,7 @@ void hnae3_set_client_init_flag(struct hnae3_client *client,
EXPORT_SYMBOL(hnae3_set_client_init_flag);
static int hnae3_get_client_init_flag(struct hnae3_client *client,
- struct hnae3_ae_dev *ae_dev)
+ struct hnae3_ae_dev *ae_dev)
{
int inited = 0;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
index 4260653..0332d6f 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
@@ -704,7 +704,7 @@ static int hns3_get_link_ksettings(struct net_device *netdev,
return 0;
}
-static int hns3_check_ksettings_param(struct net_device *netdev,
+static int hns3_check_ksettings_param(const struct net_device *netdev,
const struct ethtool_link_ksettings *cmd)
{
struct hnae3_handle *handle = hns3_get_handle(netdev);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 1d8dee1..24b59f0 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1128,6 +1128,7 @@ static void hclge_parse_link_mode(struct hclge_dev *hdev, u8 speed_ability)
else if (media_type == HNAE3_MEDIA_TYPE_BACKPLANE)
hclge_parse_backplane_link_mode(hdev, speed_ability);
}
+
static void hclge_parse_cfg(struct hclge_cfg *cfg, struct hclge_desc *desc)
{
struct hclge_cfg_param_cmd *req;
@@ -4364,8 +4365,8 @@ int hclge_bind_ring_with_vector(struct hclge_vport *vport,
struct hclge_dev *hdev = vport->back;
struct hnae3_ring_chain_node *node;
struct hclge_desc desc;
- struct hclge_ctrl_vector_chain_cmd *req
- = (struct hclge_ctrl_vector_chain_cmd *)desc.data;
+ struct hclge_ctrl_vector_chain_cmd *req =
+ (struct hclge_ctrl_vector_chain_cmd *)desc.data;
enum hclge_cmd_status status;
enum hclge_opcode_type op;
u16 tqp_type_and_id;
@@ -8656,7 +8657,7 @@ static int hclge_init_client_instance(struct hnae3_client *client,
}
}
- return ret;
+ return 0;
clear_nic:
hdev->nic_client = NULL;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
index f6d9b57..7c28933 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
@@ -1000,7 +1000,6 @@ int hclge_buffer_alloc(struct hclge_dev *hdev);
int hclge_rss_init_hw(struct hclge_dev *hdev);
void hclge_rss_indir_init_cfg(struct hclge_dev *hdev);
-int hclge_inform_reset_assert_to_vf(struct hclge_vport *vport);
void hclge_mbx_handler(struct hclge_dev *hdev);
int hclge_reset_tqp(struct hnae3_handle *handle, u16 queue_id);
void hclge_reset_vf_queue(struct hclge_vport *vport, u16 queue_id);
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 6/6] net: hns3: add phy_attached_info() to the hns3 driver
From: Huazhong Tan @ 2019-08-16 8:09 UTC (permalink / raw)
To: davem
Cc: netdev, linux-kernel, salil.mehta, yisen.zhuang, linuxarm,
jakub.kicinski, Yonglong Liu, Huazhong Tan
In-Reply-To: <1565942982-12105-1-git-send-email-tanhuazhong@huawei.com>
From: Yonglong Liu <liuyonglong@huawei.com>
This patch adds the call to phy_attached_info() to the hns3 driver
to identify which exact PHY drivers and models is in use.
Signed-off-by: Yonglong Liu <liuyonglong@huawei.com>
Reviewed-by: Yunsheng Lin <linyunsheng@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mdio.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mdio.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mdio.c
index abb1b43..dc4dfd4 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mdio.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mdio.c
@@ -231,6 +231,8 @@ int hclge_mac_connect_phy(struct hnae3_handle *handle)
linkmode_clear_bit(ETHTOOL_LINK_MODE_Asym_Pause_BIT,
phydev->advertising);
+ phy_attached_info(phydev);
+
return 0;
}
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 4/6] net: hns3: change print level of RAS error log from warning to error
From: Huazhong Tan @ 2019-08-16 8:09 UTC (permalink / raw)
To: davem
Cc: netdev, linux-kernel, salil.mehta, yisen.zhuang, linuxarm,
jakub.kicinski, Xiaofei Tan, Weihang Li, Huazhong Tan
In-Reply-To: <1565942982-12105-1-git-send-email-tanhuazhong@huawei.com>
From: Xiaofei Tan <tanxiaofei@huawei.com>
This patch changes print level of RAS error log from warning to error.
Because RAS error and its recovery process could cause application
failure. Also uses %u instead of %d when the parameter is unsigned.
Signed-off-by: Xiaofei Tan <tanxiaofei@huawei.com>
Signed-off-by: Weihang Li <liweihang@hisilicon.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c | 88 +++++++++++-----------
1 file changed, 44 insertions(+), 44 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
index 5ce9a8a..2425b3f 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
@@ -637,8 +637,8 @@ static void hclge_log_error(struct device *dev, char *reg,
{
while (err->msg) {
if (err->int_msk & err_sts) {
- dev_warn(dev, "%s %s found [error status=0x%x]\n",
- reg, err->msg, err_sts);
+ dev_err(dev, "%s %s found [error status=0x%x]\n",
+ reg, err->msg, err_sts);
if (err->reset_level &&
err->reset_level != HNAE3_NONE_RESET)
set_bit(err->reset_level, reset_requests);
@@ -1163,8 +1163,8 @@ static int hclge_handle_mpf_ras_error(struct hclge_dev *hdev,
status = le32_to_cpu(*(desc_data + 3)) & BIT(0);
if (status) {
- dev_warn(dev, "SSU_ECC_MULTI_BIT_INT_1 ssu_mem32_ecc_mbit_err found [error status=0x%x]\n",
- status);
+ dev_err(dev, "SSU_ECC_MULTI_BIT_INT_1 ssu_mem32_ecc_mbit_err found [error status=0x%x]\n",
+ status);
set_bit(HNAE3_GLOBAL_RESET, &ae_dev->hw_err_reset_req);
}
@@ -1200,8 +1200,8 @@ static int hclge_handle_mpf_ras_error(struct hclge_dev *hdev,
desc_data = (__le32 *)&desc[5];
status = le32_to_cpu(*(desc_data + 1));
if (status) {
- dev_warn(dev, "PPU_MPF_ABNORMAL_INT_ST1 %s found\n",
- "rpu_rx_pkt_ecc_mbit_err");
+ dev_err(dev,
+ "PPU_MPF_ABNORMAL_INT_ST1 rpu_rx_pkt_ecc_mbit_err found\n");
set_bit(HNAE3_GLOBAL_RESET, &ae_dev->hw_err_reset_req);
}
@@ -1379,17 +1379,17 @@ static int hclge_log_rocee_axi_error(struct hclge_dev *hdev)
return ret;
}
- dev_info(dev, "AXI1: %08X %08X %08X %08X %08X %08X\n",
- le32_to_cpu(desc[0].data[0]), le32_to_cpu(desc[0].data[1]),
- le32_to_cpu(desc[0].data[2]), le32_to_cpu(desc[0].data[3]),
- le32_to_cpu(desc[0].data[4]), le32_to_cpu(desc[0].data[5]));
- dev_info(dev, "AXI2: %08X %08X %08X %08X %08X %08X\n",
- le32_to_cpu(desc[1].data[0]), le32_to_cpu(desc[1].data[1]),
- le32_to_cpu(desc[1].data[2]), le32_to_cpu(desc[1].data[3]),
- le32_to_cpu(desc[1].data[4]), le32_to_cpu(desc[1].data[5]));
- dev_info(dev, "AXI3: %08X %08X %08X %08X\n",
- le32_to_cpu(desc[2].data[0]), le32_to_cpu(desc[2].data[1]),
- le32_to_cpu(desc[2].data[2]), le32_to_cpu(desc[2].data[3]));
+ dev_err(dev, "AXI1: %08X %08X %08X %08X %08X %08X\n",
+ le32_to_cpu(desc[0].data[0]), le32_to_cpu(desc[0].data[1]),
+ le32_to_cpu(desc[0].data[2]), le32_to_cpu(desc[0].data[3]),
+ le32_to_cpu(desc[0].data[4]), le32_to_cpu(desc[0].data[5]));
+ dev_err(dev, "AXI2: %08X %08X %08X %08X %08X %08X\n",
+ le32_to_cpu(desc[1].data[0]), le32_to_cpu(desc[1].data[1]),
+ le32_to_cpu(desc[1].data[2]), le32_to_cpu(desc[1].data[3]),
+ le32_to_cpu(desc[1].data[4]), le32_to_cpu(desc[1].data[5]));
+ dev_err(dev, "AXI3: %08X %08X %08X %08X\n",
+ le32_to_cpu(desc[2].data[0]), le32_to_cpu(desc[2].data[1]),
+ le32_to_cpu(desc[2].data[2]), le32_to_cpu(desc[2].data[3]));
return 0;
}
@@ -1408,12 +1408,12 @@ static int hclge_log_rocee_ecc_error(struct hclge_dev *hdev)
return ret;
}
- dev_info(dev, "ECC1: %08X %08X %08X %08X %08X %08X\n",
- le32_to_cpu(desc[0].data[0]), le32_to_cpu(desc[0].data[1]),
- le32_to_cpu(desc[0].data[2]), le32_to_cpu(desc[0].data[3]),
- le32_to_cpu(desc[0].data[4]), le32_to_cpu(desc[0].data[5]));
- dev_info(dev, "ECC2: %08X %08X %08X\n", le32_to_cpu(desc[1].data[0]),
- le32_to_cpu(desc[1].data[1]), le32_to_cpu(desc[1].data[2]));
+ dev_err(dev, "ECC1: %08X %08X %08X %08X %08X %08X\n",
+ le32_to_cpu(desc[0].data[0]), le32_to_cpu(desc[0].data[1]),
+ le32_to_cpu(desc[0].data[2]), le32_to_cpu(desc[0].data[3]),
+ le32_to_cpu(desc[0].data[4]), le32_to_cpu(desc[0].data[5]));
+ dev_err(dev, "ECC2: %08X %08X %08X\n", le32_to_cpu(desc[1].data[0]),
+ le32_to_cpu(desc[1].data[1]), le32_to_cpu(desc[1].data[2]));
return 0;
}
@@ -1442,9 +1442,9 @@ static int hclge_log_rocee_ovf_error(struct hclge_dev *hdev)
le32_to_cpu(desc[0].data[0]);
while (err->msg) {
if (err->int_msk == err_sts) {
- dev_warn(dev, "%s [error status=0x%x] found\n",
- err->msg,
- le32_to_cpu(desc[0].data[0]));
+ dev_err(dev, "%s [error status=0x%x] found\n",
+ err->msg,
+ le32_to_cpu(desc[0].data[0]));
break;
}
err++;
@@ -1452,13 +1452,13 @@ static int hclge_log_rocee_ovf_error(struct hclge_dev *hdev)
}
if (le32_to_cpu(desc[0].data[1]) & HCLGE_ROCEE_OVF_ERR_INT_MASK) {
- dev_warn(dev, "ROCEE TSP OVF [error status=0x%x] found\n",
- le32_to_cpu(desc[0].data[1]));
+ dev_err(dev, "ROCEE TSP OVF [error status=0x%x] found\n",
+ le32_to_cpu(desc[0].data[1]));
}
if (le32_to_cpu(desc[0].data[2]) & HCLGE_ROCEE_OVF_ERR_INT_MASK) {
- dev_warn(dev, "ROCEE SCC OVF [error status=0x%x] found\n",
- le32_to_cpu(desc[0].data[2]));
+ dev_err(dev, "ROCEE SCC OVF [error status=0x%x] found\n",
+ le32_to_cpu(desc[0].data[2]));
}
return 0;
@@ -1486,10 +1486,10 @@ hclge_log_and_clear_rocee_ras_error(struct hclge_dev *hdev)
if (status & HCLGE_ROCEE_AXI_ERR_INT_MASK) {
if (status & HCLGE_ROCEE_RERR_INT_MASK)
- dev_warn(dev, "ROCEE RAS AXI rresp error\n");
+ dev_err(dev, "ROCEE RAS AXI rresp error\n");
if (status & HCLGE_ROCEE_BERR_INT_MASK)
- dev_warn(dev, "ROCEE RAS AXI bresp error\n");
+ dev_err(dev, "ROCEE RAS AXI bresp error\n");
reset_type = HNAE3_FUNC_RESET;
@@ -1499,7 +1499,7 @@ hclge_log_and_clear_rocee_ras_error(struct hclge_dev *hdev)
}
if (status & HCLGE_ROCEE_ECC_INT_MASK) {
- dev_warn(dev, "ROCEE RAS 2bit ECC error\n");
+ dev_err(dev, "ROCEE RAS 2bit ECC error\n");
reset_type = HNAE3_GLOBAL_RESET;
ret = hclge_log_rocee_ecc_error(hdev);
@@ -1640,16 +1640,16 @@ pci_ers_result_t hclge_handle_hw_ras_error(struct hnae3_ae_dev *ae_dev)
/* Handling Non-fatal HNS RAS errors */
if (status & HCLGE_RAS_REG_NFE_MASK) {
- dev_warn(dev,
- "HNS Non-Fatal RAS error(status=0x%x) identified\n",
- status);
+ dev_err(dev,
+ "HNS Non-Fatal RAS error(status=0x%x) identified\n",
+ status);
hclge_handle_all_ras_errors(hdev);
}
/* Handling Non-fatal Rocee RAS errors */
if (hdev->pdev->revision >= 0x21 &&
status & HCLGE_RAS_REG_ROCEE_ERR_MASK) {
- dev_warn(dev, "ROCEE Non-Fatal RAS error identified\n");
+ dev_err(dev, "ROCEE Non-Fatal RAS error identified\n");
hclge_handle_rocee_ras_error(ae_dev);
}
@@ -1728,8 +1728,8 @@ static void hclge_handle_over_8bd_err(struct hclge_dev *hdev,
return;
}
- dev_warn(dev, "PPU_PF_ABNORMAL_INT_ST over_8bd_no_fe found, vf_id(%d), queue_id(%d)\n",
- vf_id, q_id);
+ dev_err(dev, "PPU_PF_ABNORMAL_INT_ST over_8bd_no_fe found, vf_id(%u), queue_id(%u)\n",
+ vf_id, q_id);
if (vf_id) {
if (vf_id >= hdev->num_alloc_vport) {
@@ -1746,8 +1746,8 @@ static void hclge_handle_over_8bd_err(struct hclge_dev *hdev,
ret = hclge_inform_reset_assert_to_vf(&hdev->vport[vf_id]);
if (ret)
- dev_warn(dev, "inform reset to vf(%d) failed %d!\n",
- hdev->vport->vport_id, ret);
+ dev_err(dev, "inform reset to vf(%u) failed %d!\n",
+ hdev->vport->vport_id, ret);
} else {
set_bit(HNAE3_FUNC_RESET, reset_requests);
}
@@ -1793,8 +1793,8 @@ static int hclge_handle_mpf_msix_error(struct hclge_dev *hdev,
status = le32_to_cpu(*(desc_data + 2)) &
HCLGE_PPU_MPF_INT_ST2_MSIX_MASK;
if (status)
- dev_warn(dev, "PPU_MPF_ABNORMAL_INT_ST2 rx_q_search_miss found [dfx status=0x%x\n]",
- status);
+ dev_err(dev, "PPU_MPF_ABNORMAL_INT_ST2 rx_q_search_miss found [dfx status=0x%x\n]",
+ status);
/* clear all main PF MSIx errors */
ret = hclge_clear_hw_msix_error(hdev, desc, true, mpf_bd_num);
@@ -1988,7 +1988,7 @@ void hclge_handle_all_hns_hw_errors(struct hnae3_ae_dev *ae_dev)
/* Handle Non-fatal HNS RAS errors */
if (status & HCLGE_RAS_REG_NFE_MASK) {
- dev_warn(dev, "HNS hw error(RAS) identified during init\n");
+ dev_err(dev, "HNS hw error(RAS) identified during init\n");
hclge_handle_all_ras_errors(hdev);
}
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 0/6] net: hns3: add some cleanups & bugfix
From: Huazhong Tan @ 2019-08-16 8:09 UTC (permalink / raw)
To: davem
Cc: netdev, linux-kernel, salil.mehta, yisen.zhuang, linuxarm,
jakub.kicinski, Huazhong Tan
This patch-set includes cleanups and bugfix for the HNS3 ethernet
controller driver.
[patch 01/06 - 03/06] adds some cleanups.
[patch 04/06] changes the print level of RAS.
[patch 05/06] fixes a bug related to MAC TNL.
[patch 06/06] adds phy_attached_info().
Guojia Liao (3):
net: hns3: add or modify comments
net: hns3: modify redundant initialization of variable
net: hns3: fix error and incorrect format
Huazhong Tan (1):
net: hns3: prevent unnecessary MAC TNL interrupt
Xiaofei Tan (1):
net: hns3: change print level of RAS error log from warning to error
Yonglong Liu (1):
net: hns3: add phy_attached_info() to the hns3 driver
drivers/net/ethernet/hisilicon/hns3/hnae3.c | 9 +--
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 12 +--
drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 8 +-
drivers/net/ethernet/hisilicon/hns3/hns3_enet.h | 8 +-
drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 6 +-
.../ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c | 3 +
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c | 90 +++++++++++-----------
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 17 ++--
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 1 -
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c | 2 +-
.../ethernet/hisilicon/hns3/hns3pf/hclge_mdio.c | 2 +
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 2 +-
.../ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c | 2 +-
.../ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c | 11 ++-
14 files changed, 89 insertions(+), 84 deletions(-)
--
2.7.4
^ permalink raw reply
* Re: [PATCH] arm64: do_csum: implement accelerated scalar version
From: Shaokun Zhang @ 2019-08-16 8:15 UTC (permalink / raw)
To: Will Deacon
Cc: Robin Murphy, Ard Biesheuvel, linux-arm-kernel, netdev,
ilias.apalodimas, huanglingyan (A), steve.capper
In-Reply-To: <20190815164609.GI2015@fuggles.cambridge.arm.com>
Hi Will,
On 2019/8/16 0:46, Will Deacon wrote:
> On Thu, May 16, 2019 at 11:14:35AM +0800, Zhangshaokun wrote:
>> On 2019/5/15 17:47, Will Deacon wrote:
>>> On Mon, Apr 15, 2019 at 07:18:22PM +0100, Robin Murphy wrote:
>>>> On 12/04/2019 10:52, Will Deacon wrote:
>>>>> I'm waiting for Robin to come back with numbers for a C implementation.
>>>>>
>>>>> Robin -- did you get anywhere with that?
>>>>
>>>> Still not what I would call finished, but where I've got so far (besides an
>>>> increasingly elaborate test rig) is as below - it still wants some unrolling
>>>> in the middle to really fly (and actual testing on BE), but the worst-case
>>>> performance already equals or just beats this asm version on Cortex-A53 with
>>>> GCC 7 (by virtue of being alignment-insensitive and branchless except for
>>>> the loop). Unfortunately, the advantage of C code being instrumentable does
>>>> also come around to bite me...
>>>
>>> Is there any interest from anybody in spinning a proper patch out of this?
>>> Shaokun?
>>
>> HiSilicon's Kunpeng920(Hi1620) benefits from do_csum optimization, if Ard and
>> Robin are ok, Lingyan or I can try to do it.
>> Of course, if any guy posts the patch, we are happy to test it.
>> Any will be ok.
>
> I don't mind who posts it, but Robin is super busy with SMMU stuff at the
> moment so it probably makes more sense for you or Lingyan to do it.
Thanks for restarting this topic, I or Lingyan will do it soon.
Thanks,
Shaokun
>
> Will
>
> .
>
^ permalink raw reply
* Re: [PATCH net-next] r8152: divide the tx and rx bottom functions
From: Eric Dumazet @ 2019-08-16 8:19 UTC (permalink / raw)
To: Hayes Wang, netdev@vger.kernel.org; +Cc: nic_swsd, linux-kernel@vger.kernel.org
In-Reply-To: <0835B3720019904CB8F7AA43166CEEB2F18D470D@RTITMBSVM03.realtek.com.tw>
On 8/16/19 10:10 AM, Hayes Wang wrote:
> Eric Dumazet [mailto:eric.dumazet@gmail.com]
>> Sent: Friday, August 16, 2019 2:40 PM
> [...]
>> tasklet and NAPI are scheduled on the same core (the current
>> cpu calling napi_schedule() or tasklet_schedule())
>>
>> I would rather not add this dubious tasklet, and instead try to understand
>> what is wrong in this driver ;)
>>
>> The various napi_schedule() calls are suspect IMO.
>
> The original method as following.
>
> static int r8152_poll(struct napi_struct *napi, int budget)
> {
> struct r8152 *tp = container_of(napi, struct r8152, napi);
> int work_done;
>
> work_done = rx_bottom(tp, budget); <-- RX
> bottom_half(tp); <-- Tx (tx_bottom)
> [...]
>
> The rx_bottom and tx_bottom would only be called in r8152_poll.
> That is, tx_bottom wouldn't be run unless rx_bottom is finished.
> And, rx_bottom would be called if tx_bottom is running.
>
> If the traffic is busy. rx_bottom or tx_bottom may take a lot
> of time to deal with the packets. And the one would increase
> the latency time for the other one.
>
> Therefore, when I separate the tx_bottom and rx_bottom to
> different tasklet and napi, the callback functions of tx and
> rx may schedule the tasklet and napi to different cpu. Then,
> the rx_bottom and tx_bottom may be run at the same time.
Your patch makes absolutely no guarantee of doing what you
want, I am afraid.
>
> Take our arm platform for example. There are five cpus to
> handle the interrupt of USB host controller. When the rx is
> completed, cpu #1 may handle the interrupt and napi would
> be scheduled. When the tx is finished, cpu #2 may handle
> the interrupt and the tasklet is scheduled. Then, napi is
> run on cpu #1, and tasklet is run on cpu #2.
>
>> Also rtl8152_start_xmit() uses skb_queue_tail(&tp->tx_queue, skb);
>>
>> But I see nothing really kicking the transmit if tx_free is empty ?
>
> Tx callback function "write_bulk_callback" would deal with it.
> The callback function would check if there are packets waiting
> to be sent.
Which callback ?
After an idle period (no activity, no prior packets being tx-completed ...),
a packet is sent by the upper stack, enters the ndo_start_xmit() of a network driver.
This driver ndo_start_xmit() simply adds an skb to a local list, and returns.
Where/how is scheduled this callback ?
Some kind of timer ?
An (unrelated) incoming packet ?
>
>
> Best Regards,
> Hayes
>
>
^ permalink raw reply
* [PATCH REPOST 0/2] can: flexcan: fix PM and wakeup issue
From: Joakim Zhang @ 2019-08-16 8:20 UTC (permalink / raw)
To: mkl@pengutronix.de, linux-can@vger.kernel.org, sean@geanix.com
Cc: wg@grandegger.com, netdev@vger.kernel.org, dl-linux-imx,
Joakim Zhang
This patch set intends to fix Flecan PM and wakeup issue.
Joakim Zhang (2):
can: flexcan: fix deadlock when using self wakeup
can: flexcan: add LPSR mode support for i.MX7D
drivers/net/can/flexcan.c | 34 ++++++++++++++++++++++++++++++----
1 file changed, 30 insertions(+), 4 deletions(-)
--
2.17.1
^ 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