diff mbox series

[v2,3/3] net: mhi: Add dedicated alloc thread

Message ID 1607598951-2340-3-git-send-email-loic.poulain@linaro.org
State Superseded
Headers show
Series None | expand

Commit Message

Loic Poulain Dec. 10, 2020, 11:15 a.m. UTC
The buffer allocation for RX path is currently done by a work executed
in the system workqueue. The work to do is quite simple and consists
mostly in allocating and queueing as much as possible buffers to the MHI
RX channel.

It appears that using a dedicated kthread would be more appropriate to
prevent
1. RX allocation latency introduced by the system queue
2. Unbounded work execution, the work only returning when queue is
full, it can possibly monopolise the workqueue thread on slower systems.

This patch replaces the system work with a simple kthread that loops on
buffer allocation and sleeps when queue is full. Moreover it gets rid
of the local rx_queued variable (to track buffer count), and instead,
relies on the new mhi_get_free_desc_count helper.

After pratical testing on a x86_64 machine, this change improves
- Peek throughput (slightly, by few mbps)
- Throughput stability when concurrent loads are running (stress)
- CPU usage, less CPU cycles dedicated to the task

Below is the powertop output for RX allocation task before and after
this change, when performing UDP download at 6Gbps. Mostly to highlight
the improvement in term of CPU usage.

older (system workqueue):
Usage       Events/s    Category       Description
63,2 ms/s     134,0        kWork          mhi_net_rx_refill_work
62,8 ms/s     134,3        kWork          mhi_net_rx_refill_work
60,8 ms/s     141,4        kWork          mhi_net_rx_refill_work

newer (dedicated kthread):
Usage       Events/s    Category       Description
20,7 ms/s     155,6        Process        [PID 3360] [mhi-net-rx]
22,2 ms/s     169,6        Process        [PID 3360] [mhi-net-rx]
22,3 ms/s     150,2        Process        [PID 3360] [mhi-net-rx]

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>

---
 v2: add module parameter for changing RX refill level

 drivers/net/mhi_net.c | 110 ++++++++++++++++++++++++++++----------------------
 1 file changed, 62 insertions(+), 48 deletions(-)

-- 
2.7.4

Comments

Jakub Kicinski Dec. 12, 2020, 8:55 p.m. UTC | #1
On Thu, 10 Dec 2020 12:15:51 +0100 Loic Poulain wrote:
> The buffer allocation for RX path is currently done by a work executed

> in the system workqueue. The work to do is quite simple and consists

> mostly in allocating and queueing as much as possible buffers to the MHI

> RX channel.

> 

> It appears that using a dedicated kthread would be more appropriate to

> prevent

> 1. RX allocation latency introduced by the system queue


System work queue should not add much latency, you can also create your
own workqueue. Did you intend to modify the priority of the thread you
create?

> 2. Unbounded work execution, the work only returning when queue is

> full, it can possibly monopolise the workqueue thread on slower systems.


Is this something you observed in practice?

> This patch replaces the system work with a simple kthread that loops on

> buffer allocation and sleeps when queue is full. Moreover it gets rid

> of the local rx_queued variable (to track buffer count), and instead,

> relies on the new mhi_get_free_desc_count helper.


Seems unrelated, should probably be a separate patch.

> After pratical testing on a x86_64 machine, this change improves

> - Peek throughput (slightly, by few mbps)

> - Throughput stability when concurrent loads are running (stress)

> - CPU usage, less CPU cycles dedicated to the task


Do you have an explanation why the CPU cycles are lower?

> Below is the powertop output for RX allocation task before and after

> this change, when performing UDP download at 6Gbps. Mostly to highlight

> the improvement in term of CPU usage.

> 

> older (system workqueue):

> Usage       Events/s    Category       Description

> 63,2 ms/s     134,0        kWork          mhi_net_rx_refill_work

> 62,8 ms/s     134,3        kWork          mhi_net_rx_refill_work

> 60,8 ms/s     141,4        kWork          mhi_net_rx_refill_work

> 

> newer (dedicated kthread):

> Usage       Events/s    Category       Description

> 20,7 ms/s     155,6        Process        [PID 3360] [mhi-net-rx]

> 22,2 ms/s     169,6        Process        [PID 3360] [mhi-net-rx]

> 22,3 ms/s     150,2        Process        [PID 3360] [mhi-net-rx]

> 

> Signed-off-by: Loic Poulain <loic.poulain@linaro.org>

> ---

>  v2: add module parameter for changing RX refill level


> @@ -16,6 +17,11 @@

>  #define MHI_NET_MAX_MTU		0xffff

>  #define MHI_NET_DEFAULT_MTU	0x4000

>  

> +static unsigned int rx_refill_level = 70;

> +module_param(rx_refill_level, uint, 0600);

> +MODULE_PARM_DESC(rx_refill_level,

> +		 "The minimal RX queue level percentage (0 to 100) under which the RX queue must be refilled");


Sorry you got bad advice in v1 and I didn't catch it. Please avoid
adding module parameters. Many drivers do bulk refill, and don't need
and extra parametrization, I don't see why this one would be special -
if it is please explain.

>  struct mhi_net_stats {

>  	u64_stats_t rx_packets;

>  	u64_stats_t rx_bytes;

> @@ -25,7 +31,6 @@ struct mhi_net_stats {

>  	u64_stats_t tx_bytes;

>  	u64_stats_t tx_errors;

>  	u64_stats_t tx_dropped;

> -	atomic_t rx_queued;

>  	struct u64_stats_sync tx_syncp;

>  	struct u64_stats_sync rx_syncp;

>  };

> @@ -33,17 +38,66 @@ struct mhi_net_stats {

>  struct mhi_net_dev {

>  	struct mhi_device *mdev;

>  	struct net_device *ndev;

> -	struct delayed_work rx_refill;

> +	struct task_struct *refill_task;

> +	wait_queue_head_t refill_wq;

>  	struct mhi_net_stats stats;

>  	u32 rx_queue_sz;

> +	u32 rx_refill_level;

>  };

>  

> +static int mhi_net_refill_thread(void *data)

> +{

> +	struct mhi_net_dev *mhi_netdev = data;

> +	struct net_device *ndev = mhi_netdev->ndev;

> +	struct mhi_device *mdev = mhi_netdev->mdev;

> +	int size = READ_ONCE(ndev->mtu);

> +	struct sk_buff *skb;

> +	int err;

> +

> +	while (1) {

> +		err = wait_event_interruptible(mhi_netdev->refill_wq,

> +					       !mhi_queue_is_full(mdev, DMA_FROM_DEVICE)

> +					       || kthread_should_stop());

> +		if (err || kthread_should_stop())

> +			break;

> +

> +		skb = netdev_alloc_skb(ndev, size);

> +		if (unlikely(!skb)) {

> +			/* No memory, retry later */

> +			schedule_timeout_interruptible(msecs_to_jiffies(250));


You should have a counter for this, at least for your testing. If this
condition is hit it'll probably have a large impact on the performance.

> +			continue;

> +		}

> +

> +		err = mhi_queue_skb(mdev, DMA_FROM_DEVICE, skb, size, MHI_EOT);

> +		if (unlikely(err)) {

> +			net_err_ratelimited("%s: Failed to queue RX buf (%d)\n",

> +					    ndev->name, err);

> +			kfree_skb(skb);

> +			break;

> +		}

> +

> +		/* Do not hog the CPU */

> +		cond_resched();

> +	}

> +

> +	return 0;

> +}

> +

>  static int mhi_ndo_open(struct net_device *ndev)

>  {

>  	struct mhi_net_dev *mhi_netdev = netdev_priv(ndev);

> +	unsigned int qsz = mhi_netdev->rx_queue_sz;

>  

> -	/* Feed the rx buffer pool */

> -	schedule_delayed_work(&mhi_netdev->rx_refill, 0);

> +	if (rx_refill_level >= 100)

> +		mhi_netdev->rx_refill_level = 1;

> +	else

> +		mhi_netdev->rx_refill_level = qsz - qsz * rx_refill_level / 100;


So you're switching from 50% fill level to 70%. Are you sure that's not
the reason the performance gets better? Did you experiments with higher
fill levels?

> +	mhi_netdev->refill_task = kthread_run(mhi_net_refill_thread, mhi_netdev,

> +					      "mhi-net-rx");

> +	if (IS_ERR(mhi_netdev->refill_task)) {

> +		return PTR_ERR(mhi_netdev->refill_task);

> +	}

>  

>  	/* Carrier is established via out-of-band channel (e.g. qmi) */

>  	netif_carrier_on(ndev);

> @@ -57,9 +111,9 @@ static int mhi_ndo_stop(struct net_device *ndev)

>  {

>  	struct mhi_net_dev *mhi_netdev = netdev_priv(ndev);

>  

> +	kthread_stop(mhi_netdev->refill_task);

>  	netif_stop_queue(ndev);

>  	netif_carrier_off(ndev);

> -	cancel_delayed_work_sync(&mhi_netdev->rx_refill);

>  

>  	return 0;

>  }

> @@ -138,9 +192,6 @@ static void mhi_net_dl_callback(struct mhi_device *mhi_dev,

>  {

>  	struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev);

>  	struct sk_buff *skb = mhi_res->buf_addr;

> -	int remaining;

> -

> -	remaining = atomic_dec_return(&mhi_netdev->stats.rx_queued);

>  

>  	if (unlikely(mhi_res->transaction_status)) {

>  		dev_kfree_skb_any(skb);

> @@ -163,9 +214,8 @@ static void mhi_net_dl_callback(struct mhi_device *mhi_dev,

>  		netif_rx(skb);

>  	}

>  

> -	/* Refill if RX buffers queue becomes low */

> -	if (remaining <= mhi_netdev->rx_queue_sz / 2)

> -		schedule_delayed_work(&mhi_netdev->rx_refill, 0);

> +	if (mhi_get_free_desc_count(mhi_dev, DMA_FROM_DEVICE) >= mhi_netdev->rx_refill_level)

> +		wake_up_interruptible(&mhi_netdev->refill_wq);

>  }

>  

>  static void mhi_net_ul_callback(struct mhi_device *mhi_dev,
Loic Poulain Dec. 14, 2020, 9:19 a.m. UTC | #2
Hi Jakub,

On Sat, 12 Dec 2020 at 21:55, Jakub Kicinski <kuba@kernel.org> wrote:
>
> On Thu, 10 Dec 2020 12:15:51 +0100 Loic Poulain wrote:
> > The buffer allocation for RX path is currently done by a work executed
> > in the system workqueue. The work to do is quite simple and consists
> > mostly in allocating and queueing as much as possible buffers to the MHI
> > RX channel.
> >
> > It appears that using a dedicated kthread would be more appropriate to
> > prevent
> > 1. RX allocation latency introduced by the system queue
>
> System work queue should not add much latency, you can also create your
> own workqueue. Did you intend to modify the priority of the thread you
> create?

No, and I don't, since I assume there is no reason to prioritize
network over other loads. I've considered the dedicated workqueue, but
since there is only one task to run as a while loop, I thought using a
kthread was more appropriate (and slightly lighter), but I can move to
that solution if you recommend it.

>
> > 2. Unbounded work execution, the work only returning when queue is
> > full, it can possibly monopolise the workqueue thread on slower systems.
>
> Is this something you observed in practice?

No, I've just observed that work duration is inconstant , queuing from
few buffers to several hundreeds. This unbounded behavior makes me
feel that doing that in the shared sytem workqueue is probably not the
right place. I've not tested on a slower machine though.

>
> > This patch replaces the system work with a simple kthread that loops on
> > buffer allocation and sleeps when queue is full. Moreover it gets rid
> > of the local rx_queued variable (to track buffer count), and instead,
> > relies on the new mhi_get_free_desc_count helper.
>
> Seems unrelated, should probably be a separate patch.

I can do that.

>
> > After pratical testing on a x86_64 machine, this change improves
> > - Peek throughput (slightly, by few mbps)
> > - Throughput stability when concurrent loads are running (stress)
> > - CPU usage, less CPU cycles dedicated to the task
>
> Do you have an explanation why the CPU cycles are lower?

For CPU cycles, TBH, not really, this is just observational. Regarding
throughput stability, it's certainly because the work can consume all
its dedicated kthread time.

>
> > Below is the powertop output for RX allocation task before and after
> > this change, when performing UDP download at 6Gbps. Mostly to highlight
> > the improvement in term of CPU usage.
> >
> > older (system workqueue):
> > Usage       Events/s    Category       Description
> > 63,2 ms/s     134,0        kWork          mhi_net_rx_refill_work
> > 62,8 ms/s     134,3        kWork          mhi_net_rx_refill_work
> > 60,8 ms/s     141,4        kWork          mhi_net_rx_refill_work
> >
> > newer (dedicated kthread):
> > Usage       Events/s    Category       Description
> > 20,7 ms/s     155,6        Process        [PID 3360] [mhi-net-rx]
> > 22,2 ms/s     169,6        Process        [PID 3360] [mhi-net-rx]
> > 22,3 ms/s     150,2        Process        [PID 3360] [mhi-net-rx]
> >
> > Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
> > ---
> >  v2: add module parameter for changing RX refill level
>
> > @@ -16,6 +17,11 @@
> >  #define MHI_NET_MAX_MTU              0xffff
> >  #define MHI_NET_DEFAULT_MTU  0x4000
> >
> > +static unsigned int rx_refill_level = 70;
> > +module_param(rx_refill_level, uint, 0600);
> > +MODULE_PARM_DESC(rx_refill_level,
> > +              "The minimal RX queue level percentage (0 to 100) under which the RX queue must be refilled");
>
> Sorry you got bad advice in v1 and I didn't catch it. Please avoid
> adding module parameters. Many drivers do bulk refill, and don't need
> and extra parametrization, I don't see why this one would be special -
> if it is please explain.

Ok, going to revert that.

>
> >  struct mhi_net_stats {
> >       u64_stats_t rx_packets;
> >       u64_stats_t rx_bytes;
> > @@ -25,7 +31,6 @@ struct mhi_net_stats {
> >       u64_stats_t tx_bytes;
> >       u64_stats_t tx_errors;
> >       u64_stats_t tx_dropped;
> > -     atomic_t rx_queued;
> >       struct u64_stats_sync tx_syncp;
> >       struct u64_stats_sync rx_syncp;
> >  };
> > @@ -33,17 +38,66 @@ struct mhi_net_stats {
> >  struct mhi_net_dev {
> >       struct mhi_device *mdev;
> >       struct net_device *ndev;
> > -     struct delayed_work rx_refill;
> > +     struct task_struct *refill_task;
> > +     wait_queue_head_t refill_wq;
> >       struct mhi_net_stats stats;
> >       u32 rx_queue_sz;
> > +     u32 rx_refill_level;
> >  };
> >
> > +static int mhi_net_refill_thread(void *data)
> > +{
> > +     struct mhi_net_dev *mhi_netdev = data;
> > +     struct net_device *ndev = mhi_netdev->ndev;
> > +     struct mhi_device *mdev = mhi_netdev->mdev;
> > +     int size = READ_ONCE(ndev->mtu);
> > +     struct sk_buff *skb;
> > +     int err;
> > +
> > +     while (1) {
> > +             err = wait_event_interruptible(mhi_netdev->refill_wq,
> > +                                            !mhi_queue_is_full(mdev, DMA_FROM_DEVICE)
> > +                                            || kthread_should_stop());
> > +             if (err || kthread_should_stop())
> > +                     break;
> > +
> > +             skb = netdev_alloc_skb(ndev, size);
> > +             if (unlikely(!skb)) {
> > +                     /* No memory, retry later */
> > +                     schedule_timeout_interruptible(msecs_to_jiffies(250));
>
> You should have a counter for this, at least for your testing. If this
> condition is hit it'll probably have a large impact on the performance.

Indeed, going to do that, what about a ratelimited error? I assume if
it's happen, system is really in bad shape.

>
> > +                     continue;
> > +             }
> > +
> > +             err = mhi_queue_skb(mdev, DMA_FROM_DEVICE, skb, size, MHI_EOT);
> > +             if (unlikely(err)) {
> > +                     net_err_ratelimited("%s: Failed to queue RX buf (%d)\n",
> > +                                         ndev->name, err);
> > +                     kfree_skb(skb);
> > +                     break;
> > +             }
> > +
> > +             /* Do not hog the CPU */
> > +             cond_resched();
> > +     }
> > +
> > +     return 0;
> > +}
> > +
> >  static int mhi_ndo_open(struct net_device *ndev)
> >  {
> >       struct mhi_net_dev *mhi_netdev = netdev_priv(ndev);
> > +     unsigned int qsz = mhi_netdev->rx_queue_sz;
> >
> > -     /* Feed the rx buffer pool */
> > -     schedule_delayed_work(&mhi_netdev->rx_refill, 0);
> > +     if (rx_refill_level >= 100)
> > +             mhi_netdev->rx_refill_level = 1;
> > +     else
> > +             mhi_netdev->rx_refill_level = qsz - qsz * rx_refill_level / 100;
>
> So you're switching from 50% fill level to 70%. Are you sure that's not
> the reason the performance gets better? Did you experiments with higher
> fill levels?

No, I've tested both levels with the two solutions, It's just that
after experiment, high throughput is a bit more stable with 70%. So I
can revert back to 50% to avoid confusion and keep that for a
subsequent change.

Thanks,
Loic
Jakub Kicinski Dec. 14, 2020, 7:47 p.m. UTC | #3
On Mon, 14 Dec 2020 10:19:07 +0100 Loic Poulain wrote:
> On Sat, 12 Dec 2020 at 21:55, Jakub Kicinski <kuba@kernel.org> wrote:
> > On Thu, 10 Dec 2020 12:15:51 +0100 Loic Poulain wrote:  
> > > The buffer allocation for RX path is currently done by a work executed
> > > in the system workqueue. The work to do is quite simple and consists
> > > mostly in allocating and queueing as much as possible buffers to the MHI
> > > RX channel.
> > >
> > > It appears that using a dedicated kthread would be more appropriate to
> > > prevent
> > > 1. RX allocation latency introduced by the system queue  
> >
> > System work queue should not add much latency, you can also create your
> > own workqueue. Did you intend to modify the priority of the thread you
> > create?  
> 
> No, and I don't, since I assume there is no reason to prioritize
> network over other loads. I've considered the dedicated workqueue, but
> since there is only one task to run as a while loop, I thought using a
> kthread was more appropriate (and slightly lighter), but I can move to
> that solution if you recommend it.

Not sure what to recommend TBH, if thread works better for you that's
fine. I don't understand why the thread would work better, tho. I was
just checking if there is any extra tuning that happens.

> > > 2. Unbounded work execution, the work only returning when queue is
> > > full, it can possibly monopolise the workqueue thread on slower systems.  
> >
> > Is this something you observed in practice?  
> 
> No, I've just observed that work duration is inconstant , queuing from
> few buffers to several hundreeds. This unbounded behavior makes me
> feel that doing that in the shared sytem workqueue is probably not the
> right place. I've not tested on a slower machine though.

I think long running work should not be an issue for the cmwq
implementation we have in the kernel.

Several hundred buffers means it's running concurrently with RX, right?
Since the NIC queue is 128 buffers.

> > > This patch replaces the system work with a simple kthread that loops on
> > > buffer allocation and sleeps when queue is full. Moreover it gets rid
> > > of the local rx_queued variable (to track buffer count), and instead,
> > > relies on the new mhi_get_free_desc_count helper.  
> >
> > Seems unrelated, should probably be a separate patch.  
> 
> I can do that.
> 
> >  
> > > After pratical testing on a x86_64 machine, this change improves
> > > - Peek throughput (slightly, by few mbps)
> > > - Throughput stability when concurrent loads are running (stress)
> > > - CPU usage, less CPU cycles dedicated to the task  
> >
> > Do you have an explanation why the CPU cycles are lower?  
> 
> For CPU cycles, TBH, not really, this is just observational.

Is the IRQ pinned? I wonder how often work runs on the same CPU as IRQ
processing and how often does the thread do.

> Regarding throughput stability, it's certainly because the work can
> consume all its dedicated kthread time.

Meaning workqueue implementation doesn't get enough CPU? Strange.

> > > Below is the powertop output for RX allocation task before and
> > > after this change, when performing UDP download at 6Gbps. Mostly
> > > to highlight the improvement in term of CPU usage.
> > >
> > > older (system workqueue):
> > > Usage       Events/s    Category       Description
> > > 63,2 ms/s     134,0        kWork          mhi_net_rx_refill_work
> > > 62,8 ms/s     134,3        kWork          mhi_net_rx_refill_work
> > > 60,8 ms/s     141,4        kWork          mhi_net_rx_refill_work
> > >
> > > newer (dedicated kthread):
> > > Usage       Events/s    Category       Description
> > > 20,7 ms/s     155,6        Process        [PID 3360] [mhi-net-rx]
> > > 22,2 ms/s     169,6        Process        [PID 3360] [mhi-net-rx]
> > > 22,3 ms/s     150,2        Process        [PID 3360] [mhi-net-rx]
> > >
> > > Signed-off-by: Loic Poulain <loic.poulain@linaro.org>

> > > +             skb = netdev_alloc_skb(ndev, size);
> > > +             if (unlikely(!skb)) {
> > > +                     /* No memory, retry later */
> > > +
> > > schedule_timeout_interruptible(msecs_to_jiffies(250));  
> >
> > You should have a counter for this, at least for your testing. If
> > this condition is hit it'll probably have a large impact on the
> > performance.  
> 
> Indeed, going to do that, what about a ratelimited error? I assume if
> it's happen, system is really in bad shape.

It's not that uncommon to run out of memory for a 2k allocation in an
atomic context (note that netdev_alloc_skb() uses GFP_ATOMIC).
You can add a rate-limited print if you want, tho.

> > > +                     continue;
> > > +             }
> > > +
> > > +             err = mhi_queue_skb(mdev, DMA_FROM_DEVICE, skb,
> > > size, MHI_EOT);
> > > +             if (unlikely(err)) {
> > > +                     net_err_ratelimited("%s: Failed to queue RX
> > > buf (%d)\n",
> > > +                                         ndev->name, err);
> > > +                     kfree_skb(skb);
> > > +                     break;
> > > +             }
> > > +
> > > +             /* Do not hog the CPU */
> > > +             cond_resched();
> > > +     }
> > > +
> > > +     return 0;
> > > +}
> > > +
> > >  static int mhi_ndo_open(struct net_device *ndev)
> > >  {
> > >       struct mhi_net_dev *mhi_netdev = netdev_priv(ndev);
> > > +     unsigned int qsz = mhi_netdev->rx_queue_sz;
> > >
> > > -     /* Feed the rx buffer pool */
> > > -     schedule_delayed_work(&mhi_netdev->rx_refill, 0);
> > > +     if (rx_refill_level >= 100)
> > > +             mhi_netdev->rx_refill_level = 1;
> > > +     else
> > > +             mhi_netdev->rx_refill_level = qsz - qsz *
> > > rx_refill_level / 100;  
> >
> > So you're switching from 50% fill level to 70%. Are you sure that's
> > not the reason the performance gets better? Did you experiments
> > with higher fill levels?  
> 
> No, I've tested both levels with the two solutions, It's just that
> after experiment, high throughput is a bit more stable with 70%. So I
> can revert back to 50% to avoid confusion and keep that for a
> subsequent change.

I'm not fussed about that - it would be good tho to have the numbers in
comparisons for the same fill levels. Otherwise comparing workq at 50%
vs thread at 70% is changing two variables at the same time.
Loic Poulain Dec. 15, 2020, 12:09 p.m. UTC | #4
Hi Jakub,

On Mon, 14 Dec 2020 at 20:47, Jakub Kicinski <kuba@kernel.org> wrote:
>
> On Mon, 14 Dec 2020 10:19:07 +0100 Loic Poulain wrote:
> > On Sat, 12 Dec 2020 at 21:55, Jakub Kicinski <kuba@kernel.org> wrote:
> > > On Thu, 10 Dec 2020 12:15:51 +0100 Loic Poulain wrote:
> > > > The buffer allocation for RX path is currently done by a work executed
> > > > in the system workqueue. The work to do is quite simple and consists
> > > > mostly in allocating and queueing as much as possible buffers to the MHI
> > > > RX channel.
> > > >
> > > > It appears that using a dedicated kthread would be more appropriate to
> > > > prevent
> > > > 1. RX allocation latency introduced by the system queue
> > >
> > > System work queue should not add much latency, you can also create your
> > > own workqueue. Did you intend to modify the priority of the thread you
> > > create?
> >
> > No, and I don't, since I assume there is no reason to prioritize
> > network over other loads. I've considered the dedicated workqueue, but
> > since there is only one task to run as a while loop, I thought using a
> > kthread was more appropriate (and slightly lighter), but I can move to
> > that solution if you recommend it.
>
> Not sure what to recommend TBH, if thread works better for you that's
> fine. I don't understand why the thread would work better, tho. I was
> just checking if there is any extra tuning that happens.
>
> > > > 2. Unbounded work execution, the work only returning when queue is
> > > > full, it can possibly monopolise the workqueue thread on slower systems.
> > >
> > > Is this something you observed in practice?
> >
> > No, I've just observed that work duration is inconstant , queuing from
> > few buffers to several hundreeds. This unbounded behavior makes me
> > feel that doing that in the shared sytem workqueue is probably not the
> > right place. I've not tested on a slower machine though.
>
> I think long running work should not be an issue for the cmwq
> implementation we have in the kernel.
>
> Several hundred buffers means it's running concurrently with RX, right?
> Since the NIC queue is 128 buffers.

Exactly, buffers can be completed by the hardware before we even
finished to completely fill the MHI ring buffer, that why the loop can
queue more than 128 buffers.

> > > > This patch replaces the system work with a simple kthread that loops on
> > > > buffer allocation and sleeps when queue is full. Moreover it gets rid
> > > > of the local rx_queued variable (to track buffer count), and instead,
> > > > relies on the new mhi_get_free_desc_count helper.
> > >
> > > Seems unrelated, should probably be a separate patch.
> >
> > I can do that.
> >
> > >
> > > > After pratical testing on a x86_64 machine, this change improves
> > > > - Peek throughput (slightly, by few mbps)
> > > > - Throughput stability when concurrent loads are running (stress)
> > > > - CPU usage, less CPU cycles dedicated to the task
> > >
> > > Do you have an explanation why the CPU cycles are lower?
> >
> > For CPU cycles, TBH, not really, this is just observational.
>
> Is the IRQ pinned? I wonder how often work runs on the same CPU as IRQ
> processing and how often does the thread do.
>
> > Regarding throughput stability, it's certainly because the work can
> > consume all its dedicated kthread time.
>
> Meaning workqueue implementation doesn't get enough CPU? Strange.
>
> > > > Below is the powertop output for RX allocation task before and
> > > > after this change, when performing UDP download at 6Gbps. Mostly
> > > > to highlight the improvement in term of CPU usage.
> > > >
> > > > older (system workqueue):
> > > > Usage       Events/s    Category       Description
> > > > 63,2 ms/s     134,0        kWork          mhi_net_rx_refill_work
> > > > 62,8 ms/s     134,3        kWork          mhi_net_rx_refill_work
> > > > 60,8 ms/s     141,4        kWork          mhi_net_rx_refill_work
> > > >
> > > > newer (dedicated kthread):
> > > > Usage       Events/s    Category       Description
> > > > 20,7 ms/s     155,6        Process        [PID 3360] [mhi-net-rx]
> > > > 22,2 ms/s     169,6        Process        [PID 3360] [mhi-net-rx]
> > > > 22,3 ms/s     150,2        Process        [PID 3360] [mhi-net-rx]
> > > >
> > > > Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
>
> > > > +             skb = netdev_alloc_skb(ndev, size);
> > > > +             if (unlikely(!skb)) {
> > > > +                     /* No memory, retry later */
> > > > +
> > > > schedule_timeout_interruptible(msecs_to_jiffies(250));
> > >
> > > You should have a counter for this, at least for your testing. If
> > > this condition is hit it'll probably have a large impact on the
> > > performance.
> >
> > Indeed, going to do that, what about a ratelimited error? I assume if
> > it's happen, system is really in bad shape.
>
> It's not that uncommon to run out of memory for a 2k allocation in an
> atomic context (note that netdev_alloc_skb() uses GFP_ATOMIC).
> You can add a rate-limited print if you want, tho.
>
> > > > +                     continue;
> > > > +             }
> > > > +
> > > > +             err = mhi_queue_skb(mdev, DMA_FROM_DEVICE, skb,
> > > > size, MHI_EOT);
> > > > +             if (unlikely(err)) {
> > > > +                     net_err_ratelimited("%s: Failed to queue RX
> > > > buf (%d)\n",
> > > > +                                         ndev->name, err);
> > > > +                     kfree_skb(skb);
> > > > +                     break;
> > > > +             }
> > > > +
> > > > +             /* Do not hog the CPU */
> > > > +             cond_resched();
> > > > +     }
> > > > +
> > > > +     return 0;
> > > > +}
> > > > +
> > > >  static int mhi_ndo_open(struct net_device *ndev)
> > > >  {
> > > >       struct mhi_net_dev *mhi_netdev = netdev_priv(ndev);
> > > > +     unsigned int qsz = mhi_netdev->rx_queue_sz;
> > > >
> > > > -     /* Feed the rx buffer pool */
> > > > -     schedule_delayed_work(&mhi_netdev->rx_refill, 0);
> > > > +     if (rx_refill_level >= 100)
> > > > +             mhi_netdev->rx_refill_level = 1;
> > > > +     else
> > > > +             mhi_netdev->rx_refill_level = qsz - qsz *
> > > > rx_refill_level / 100;
> > >
> > > So you're switching from 50% fill level to 70%. Are you sure that's
> > > not the reason the performance gets better? Did you experiments
> > > with higher fill levels?
> >
> > No, I've tested both levels with the two solutions, It's just that
> > after experiment, high throughput is a bit more stable with 70%. So I
> > can revert back to 50% to avoid confusion and keep that for a
> > subsequent change.
>
> I'm not fussed about that - it would be good tho to have the numbers in
> comparisons for the same fill levels. Otherwise comparing workq at 50%
> vs thread at 70% is changing two variables at the same time.

Yes, anyway, I'm going to skip the new kthread from the series, and
I'll resubmit once I get consolidated numbers with proper comparison.

Thanks,
Loic
diff mbox series

Patch

diff --git a/drivers/net/mhi_net.c b/drivers/net/mhi_net.c
index 0333e07..bd66d51 100644
--- a/drivers/net/mhi_net.c
+++ b/drivers/net/mhi_net.c
@@ -5,6 +5,7 @@ 
  */
 
 #include <linux/if_arp.h>
+#include <linux/kthread.h>
 #include <linux/mhi.h>
 #include <linux/mod_devicetable.h>
 #include <linux/module.h>
@@ -16,6 +17,11 @@ 
 #define MHI_NET_MAX_MTU		0xffff
 #define MHI_NET_DEFAULT_MTU	0x4000
 
+static unsigned int rx_refill_level = 70;
+module_param(rx_refill_level, uint, 0600);
+MODULE_PARM_DESC(rx_refill_level,
+		 "The minimal RX queue level percentage (0 to 100) under which the RX queue must be refilled");
+
 struct mhi_net_stats {
 	u64_stats_t rx_packets;
 	u64_stats_t rx_bytes;
@@ -25,7 +31,6 @@  struct mhi_net_stats {
 	u64_stats_t tx_bytes;
 	u64_stats_t tx_errors;
 	u64_stats_t tx_dropped;
-	atomic_t rx_queued;
 	struct u64_stats_sync tx_syncp;
 	struct u64_stats_sync rx_syncp;
 };
@@ -33,17 +38,66 @@  struct mhi_net_stats {
 struct mhi_net_dev {
 	struct mhi_device *mdev;
 	struct net_device *ndev;
-	struct delayed_work rx_refill;
+	struct task_struct *refill_task;
+	wait_queue_head_t refill_wq;
 	struct mhi_net_stats stats;
 	u32 rx_queue_sz;
+	u32 rx_refill_level;
 };
 
+static int mhi_net_refill_thread(void *data)
+{
+	struct mhi_net_dev *mhi_netdev = data;
+	struct net_device *ndev = mhi_netdev->ndev;
+	struct mhi_device *mdev = mhi_netdev->mdev;
+	int size = READ_ONCE(ndev->mtu);
+	struct sk_buff *skb;
+	int err;
+
+	while (1) {
+		err = wait_event_interruptible(mhi_netdev->refill_wq,
+					       !mhi_queue_is_full(mdev, DMA_FROM_DEVICE)
+					       || kthread_should_stop());
+		if (err || kthread_should_stop())
+			break;
+
+		skb = netdev_alloc_skb(ndev, size);
+		if (unlikely(!skb)) {
+			/* No memory, retry later */
+			schedule_timeout_interruptible(msecs_to_jiffies(250));
+			continue;
+		}
+
+		err = mhi_queue_skb(mdev, DMA_FROM_DEVICE, skb, size, MHI_EOT);
+		if (unlikely(err)) {
+			net_err_ratelimited("%s: Failed to queue RX buf (%d)\n",
+					    ndev->name, err);
+			kfree_skb(skb);
+			break;
+		}
+
+		/* Do not hog the CPU */
+		cond_resched();
+	}
+
+	return 0;
+}
+
 static int mhi_ndo_open(struct net_device *ndev)
 {
 	struct mhi_net_dev *mhi_netdev = netdev_priv(ndev);
+	unsigned int qsz = mhi_netdev->rx_queue_sz;
 
-	/* Feed the rx buffer pool */
-	schedule_delayed_work(&mhi_netdev->rx_refill, 0);
+	if (rx_refill_level >= 100)
+		mhi_netdev->rx_refill_level = 1;
+	else
+		mhi_netdev->rx_refill_level = qsz - qsz * rx_refill_level / 100;
+
+	mhi_netdev->refill_task = kthread_run(mhi_net_refill_thread, mhi_netdev,
+					      "mhi-net-rx");
+	if (IS_ERR(mhi_netdev->refill_task)) {
+		return PTR_ERR(mhi_netdev->refill_task);
+	}
 
 	/* Carrier is established via out-of-band channel (e.g. qmi) */
 	netif_carrier_on(ndev);
@@ -57,9 +111,9 @@  static int mhi_ndo_stop(struct net_device *ndev)
 {
 	struct mhi_net_dev *mhi_netdev = netdev_priv(ndev);
 
+	kthread_stop(mhi_netdev->refill_task);
 	netif_stop_queue(ndev);
 	netif_carrier_off(ndev);
-	cancel_delayed_work_sync(&mhi_netdev->rx_refill);
 
 	return 0;
 }
@@ -138,9 +192,6 @@  static void mhi_net_dl_callback(struct mhi_device *mhi_dev,
 {
 	struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev);
 	struct sk_buff *skb = mhi_res->buf_addr;
-	int remaining;
-
-	remaining = atomic_dec_return(&mhi_netdev->stats.rx_queued);
 
 	if (unlikely(mhi_res->transaction_status)) {
 		dev_kfree_skb_any(skb);
@@ -163,9 +214,8 @@  static void mhi_net_dl_callback(struct mhi_device *mhi_dev,
 		netif_rx(skb);
 	}
 
-	/* Refill if RX buffers queue becomes low */
-	if (remaining <= mhi_netdev->rx_queue_sz / 2)
-		schedule_delayed_work(&mhi_netdev->rx_refill, 0);
+	if (mhi_get_free_desc_count(mhi_dev, DMA_FROM_DEVICE) >= mhi_netdev->rx_refill_level)
+		wake_up_interruptible(&mhi_netdev->refill_wq);
 }
 
 static void mhi_net_ul_callback(struct mhi_device *mhi_dev,
@@ -200,42 +250,6 @@  static void mhi_net_ul_callback(struct mhi_device *mhi_dev,
 		netif_wake_queue(ndev);
 }
 
-static void mhi_net_rx_refill_work(struct work_struct *work)
-{
-	struct mhi_net_dev *mhi_netdev = container_of(work, struct mhi_net_dev,
-						      rx_refill.work);
-	struct net_device *ndev = mhi_netdev->ndev;
-	struct mhi_device *mdev = mhi_netdev->mdev;
-	int size = READ_ONCE(ndev->mtu);
-	struct sk_buff *skb;
-	int err;
-
-	while (atomic_read(&mhi_netdev->stats.rx_queued) < mhi_netdev->rx_queue_sz) {
-		skb = netdev_alloc_skb(ndev, size);
-		if (unlikely(!skb))
-			break;
-
-		err = mhi_queue_skb(mdev, DMA_FROM_DEVICE, skb, size, MHI_EOT);
-		if (unlikely(err)) {
-			net_err_ratelimited("%s: Failed to queue RX buf (%d)\n",
-					    ndev->name, err);
-			kfree_skb(skb);
-			break;
-		}
-
-		atomic_inc(&mhi_netdev->stats.rx_queued);
-
-		/* Do not hog the CPU if rx buffers are consumed faster than
-		 * queued (unlikely).
-		 */
-		cond_resched();
-	}
-
-	/* If we're still starved of rx buffers, reschedule later */
-	if (unlikely(!atomic_read(&mhi_netdev->stats.rx_queued)))
-		schedule_delayed_work(&mhi_netdev->rx_refill, HZ / 2);
-}
-
 static int mhi_net_probe(struct mhi_device *mhi_dev,
 			 const struct mhi_device_id *id)
 {
@@ -256,7 +270,7 @@  static int mhi_net_probe(struct mhi_device *mhi_dev,
 	mhi_netdev->mdev = mhi_dev;
 	SET_NETDEV_DEV(ndev, &mhi_dev->dev);
 
-	INIT_DELAYED_WORK(&mhi_netdev->rx_refill, mhi_net_rx_refill_work);
+	init_waitqueue_head(&mhi_netdev->refill_wq);
 	u64_stats_init(&mhi_netdev->stats.rx_syncp);
 	u64_stats_init(&mhi_netdev->stats.tx_syncp);