All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] drivers: spi: spi.c: Convert statistics to per-cpu u64_stats_t
@ 2022-05-23 14:20 David Jander
  2022-05-23 15:13 ` Andrew Lunn
  0 siblings, 1 reply; 5+ messages in thread
From: David Jander @ 2022-05-23 14:20 UTC (permalink / raw)
  To: Mark Brown; +Cc: linux-spi, Andrew Lunn, Marc Kleine-Budde, David Jander

This change gives a dramatic performance improvement in the hot path,
since many costly spin_lock_irqsave() calls can be avoided.

Suggested-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David Jander <david@protonic.nl>
---
 drivers/spi/spi.c       | 95 ++++++++++++++++++++++++-----------------
 include/linux/spi/spi.h | 80 ++++++++++++++++++++++++----------
 2 files changed, 114 insertions(+), 61 deletions(-)

diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index c4dd1200fe99..edc290e67b92 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -33,6 +33,7 @@
 #include <linux/idr.h>
 #include <linux/platform_data/x86/apple.h>
 #include <linux/ptp_clock_kernel.h>
+#include <linux/percpu.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/spi.h>
@@ -111,6 +112,25 @@ static ssize_t driver_override_show(struct device *dev,
 }
 static DEVICE_ATTR_RW(driver_override);
 
+#define spi_pcpu_stats_totalize(ret, in, field)				\
+do {									\
+	int i;								\
+	ret = 0;							\
+	for_each_possible_cpu(i) {					\
+		const struct spi_statistics *pcpu_stats;		\
+		u64 inc;						\
+		unsigned int start;					\
+		pcpu_stats = per_cpu_ptr(in, i);			\
+		do {							\
+			start = u64_stats_fetch_begin_irq(		\
+					&pcpu_stats->syncp);		\
+			inc = u64_stats_read(&pcpu_stats->field);	\
+		} while (u64_stats_fetch_retry_irq(			\
+					&pcpu_stats->syncp, start));	\
+		ret += inc;						\
+	}								\
+} while (0)
+
 #define SPI_STATISTICS_ATTRS(field, file)				\
 static ssize_t spi_controller_##field##_show(struct device *dev,	\
 					     struct device_attribute *attr, \
@@ -118,7 +138,7 @@ static ssize_t spi_controller_##field##_show(struct device *dev,	\
 {									\
 	struct spi_controller *ctlr = container_of(dev,			\
 					 struct spi_controller, dev);	\
-	return spi_statistics_##field##_show(&ctlr->statistics, buf);	\
+	return spi_statistics_##field##_show(ctlr->pcpu_statistics, buf); \
 }									\
 static struct device_attribute dev_attr_spi_controller_##field = {	\
 	.attr = { .name = file, .mode = 0444 },				\
@@ -129,7 +149,7 @@ static ssize_t spi_device_##field##_show(struct device *dev,		\
 					char *buf)			\
 {									\
 	struct spi_device *spi = to_spi_device(dev);			\
-	return spi_statistics_##field##_show(&spi->statistics, buf);	\
+	return spi_statistics_##field##_show(spi->pcpu_statistics, buf); \
 }									\
 static struct device_attribute dev_attr_spi_device_##field = {		\
 	.attr = { .name = file, .mode = 0444 },				\
@@ -140,11 +160,10 @@ static struct device_attribute dev_attr_spi_device_##field = {		\
 static ssize_t spi_statistics_##name##_show(struct spi_statistics *stat, \
 					    char *buf)			\
 {									\
-	unsigned long flags;						\
 	ssize_t len;							\
-	spin_lock_irqsave(&stat->lock, flags);				\
-	len = sysfs_emit(buf, format_string "\n", stat->field);		\
-	spin_unlock_irqrestore(&stat->lock, flags);			\
+	u64 val;							\
+	spi_pcpu_stats_totalize(val, stat, field);			\
+	len = sysfs_emit(buf, format_string "\n", val);			\
 	return len;							\
 }									\
 SPI_STATISTICS_ATTRS(name, file)
@@ -153,14 +172,14 @@ SPI_STATISTICS_ATTRS(name, file)
 	SPI_STATISTICS_SHOW_NAME(field, __stringify(field),		\
 				 field, format_string)
 
-SPI_STATISTICS_SHOW(messages, "%lu");
-SPI_STATISTICS_SHOW(transfers, "%lu");
-SPI_STATISTICS_SHOW(errors, "%lu");
-SPI_STATISTICS_SHOW(timedout, "%lu");
+SPI_STATISTICS_SHOW(messages, "%llu");
+SPI_STATISTICS_SHOW(transfers, "%llu");
+SPI_STATISTICS_SHOW(errors, "%llu");
+SPI_STATISTICS_SHOW(timedout, "%llu");
 
-SPI_STATISTICS_SHOW(spi_sync, "%lu");
-SPI_STATISTICS_SHOW(spi_sync_immediate, "%lu");
-SPI_STATISTICS_SHOW(spi_async, "%lu");
+SPI_STATISTICS_SHOW(spi_sync, "%llu");
+SPI_STATISTICS_SHOW(spi_sync_immediate, "%llu");
+SPI_STATISTICS_SHOW(spi_async, "%llu");
 
 SPI_STATISTICS_SHOW(bytes, "%llu");
 SPI_STATISTICS_SHOW(bytes_rx, "%llu");
@@ -169,7 +188,7 @@ SPI_STATISTICS_SHOW(bytes_tx, "%llu");
 #define SPI_STATISTICS_TRANSFER_BYTES_HISTO(index, number)		\
 	SPI_STATISTICS_SHOW_NAME(transfer_bytes_histo##index,		\
 				 "transfer_bytes_histo_" number,	\
-				 transfer_bytes_histo[index],  "%lu")
+				 transfer_bytes_histo[index],  "%llu")
 SPI_STATISTICS_TRANSFER_BYTES_HISTO(0,  "0-1");
 SPI_STATISTICS_TRANSFER_BYTES_HISTO(1,  "2-3");
 SPI_STATISTICS_TRANSFER_BYTES_HISTO(2,  "4-7");
@@ -188,7 +207,7 @@ SPI_STATISTICS_TRANSFER_BYTES_HISTO(14, "16384-32767");
 SPI_STATISTICS_TRANSFER_BYTES_HISTO(15, "32768-65535");
 SPI_STATISTICS_TRANSFER_BYTES_HISTO(16, "65536+");
 
-SPI_STATISTICS_SHOW(transfers_split_maxsize, "%lu");
+SPI_STATISTICS_SHOW(transfers_split_maxsize, "%llu");
 
 static struct attribute *spi_dev_attrs[] = {
 	&dev_attr_modalias.attr,
@@ -285,30 +304,30 @@ static const struct attribute_group *spi_master_groups[] = {
 	NULL,
 };
 
-static void spi_statistics_add_transfer_stats(struct spi_statistics *stats,
+static void spi_statistics_add_transfer_stats(struct spi_statistics *pcpu_stats,
 					      struct spi_transfer *xfer,
 					      struct spi_controller *ctlr)
 {
-	unsigned long flags;
 	int l2len = min(fls(xfer->len), SPI_STATISTICS_HISTO_SIZE) - 1;
+	struct spi_statistics *stats = this_cpu_ptr(pcpu_stats);
 
 	if (l2len < 0)
 		l2len = 0;
 
-	spin_lock_irqsave(&stats->lock, flags);
+	u64_stats_update_begin(&stats->syncp);
 
-	stats->transfers++;
-	stats->transfer_bytes_histo[l2len]++;
+	u64_stats_inc(&stats->transfers);
+	u64_stats_inc(&stats->transfer_bytes_histo[l2len]);
 
-	stats->bytes += xfer->len;
+	u64_stats_add(&stats->bytes, xfer->len);
 	if ((xfer->tx_buf) &&
 	    (xfer->tx_buf != ctlr->dummy_tx))
-		stats->bytes_tx += xfer->len;
+		u64_stats_add(&stats->bytes_tx, xfer->len);
 	if ((xfer->rx_buf) &&
 	    (xfer->rx_buf != ctlr->dummy_rx))
-		stats->bytes_rx += xfer->len;
+		u64_stats_add(&stats->bytes_rx, xfer->len);
 
-	spin_unlock_irqrestore(&stats->lock, flags);
+	u64_stats_update_end(&stats->syncp);
 }
 
 /*
@@ -543,7 +562,7 @@ struct spi_device *spi_alloc_device(struct spi_controller *ctlr)
 	spi->dev.release = spidev_release;
 	spi->mode = ctlr->buswidth_override_bits;
 
-	spin_lock_init(&spi->statistics.lock);
+	spi->pcpu_statistics = spi_alloc_pcpu_stats(struct spi_statistics);
 
 	device_initialize(&spi->dev);
 	return spi;
@@ -1239,8 +1258,8 @@ static int spi_transfer_wait(struct spi_controller *ctlr,
 			     struct spi_message *msg,
 			     struct spi_transfer *xfer)
 {
-	struct spi_statistics *statm = &ctlr->statistics;
-	struct spi_statistics *stats = &msg->spi->statistics;
+	struct spi_statistics *statm = ctlr->pcpu_statistics;
+	struct spi_statistics *stats = msg->spi->pcpu_statistics;
 	u32 speed_hz = xfer->speed_hz;
 	unsigned long long ms;
 
@@ -1396,8 +1415,8 @@ static int spi_transfer_one_message(struct spi_controller *ctlr,
 	struct spi_transfer *xfer;
 	bool keep_cs = false;
 	int ret = 0;
-	struct spi_statistics *statm = &ctlr->statistics;
-	struct spi_statistics *stats = &msg->spi->statistics;
+	struct spi_statistics *statm = ctlr->pcpu_statistics;
+	struct spi_statistics *stats = msg->spi->pcpu_statistics;
 
 	spi_set_cs(msg->spi, true, false);
 
@@ -3042,7 +3061,7 @@ int spi_register_controller(struct spi_controller *ctlr)
 		}
 	}
 	/* add statistics */
-	spin_lock_init(&ctlr->statistics.lock);
+	ctlr->pcpu_statistics = devm_spi_alloc_pcpu_stats(dev, struct spi_statistics);
 
 	mutex_lock(&board_lock);
 	list_add_tail(&ctlr->list, &spi_controller_list);
@@ -3380,9 +3399,9 @@ static int __spi_split_transfer_maxsize(struct spi_controller *ctlr,
 	*xferp = &xfers[count - 1];
 
 	/* increment statistics counters */
-	SPI_STATISTICS_INCREMENT_FIELD(&ctlr->statistics,
+	SPI_STATISTICS_INCREMENT_FIELD(ctlr->pcpu_statistics,
 				       transfers_split_maxsize);
-	SPI_STATISTICS_INCREMENT_FIELD(&msg->spi->statistics,
+	SPI_STATISTICS_INCREMENT_FIELD(msg->spi->pcpu_statistics,
 				       transfers_split_maxsize);
 
 	return 0;
@@ -3769,8 +3788,8 @@ static int __spi_async(struct spi_device *spi, struct spi_message *message)
 
 	message->spi = spi;
 
-	SPI_STATISTICS_INCREMENT_FIELD(&ctlr->statistics, spi_async);
-	SPI_STATISTICS_INCREMENT_FIELD(&spi->statistics, spi_async);
+	SPI_STATISTICS_INCREMENT_FIELD(ctlr->pcpu_statistics, spi_async);
+	SPI_STATISTICS_INCREMENT_FIELD(spi->pcpu_statistics, spi_async);
 
 	trace_spi_message_submit(message);
 
@@ -3917,8 +3936,8 @@ static int __spi_sync(struct spi_device *spi, struct spi_message *message)
 	message->context = &done;
 	message->spi = spi;
 
-	SPI_STATISTICS_INCREMENT_FIELD(&ctlr->statistics, spi_sync);
-	SPI_STATISTICS_INCREMENT_FIELD(&spi->statistics, spi_sync);
+	SPI_STATISTICS_INCREMENT_FIELD(ctlr->pcpu_statistics, spi_sync);
+	SPI_STATISTICS_INCREMENT_FIELD(spi->pcpu_statistics, spi_sync);
 
 	/*
 	 * If we're not using the legacy transfer method then we will
@@ -3941,9 +3960,9 @@ static int __spi_sync(struct spi_device *spi, struct spi_message *message)
 	if (status == 0) {
 		/* Push out the messages in the calling context if we can */
 		if (ctlr->transfer == spi_queued_transfer) {
-			SPI_STATISTICS_INCREMENT_FIELD(&ctlr->statistics,
+			SPI_STATISTICS_INCREMENT_FIELD(ctlr->pcpu_statistics,
 						       spi_sync_immediate);
-			SPI_STATISTICS_INCREMENT_FIELD(&spi->statistics,
+			SPI_STATISTICS_INCREMENT_FIELD(spi->pcpu_statistics,
 						       spi_sync_immediate);
 			__spi_pump_messages(ctlr, false);
 		}
diff --git a/include/linux/spi/spi.h b/include/linux/spi/spi.h
index 5f8c063ddff4..ea7f802f5a79 100644
--- a/include/linux/spi/spi.h
+++ b/include/linux/spi/spi.h
@@ -17,6 +17,7 @@
 
 #include <uapi/linux/spi/spi.h>
 #include <linux/acpi.h>
+#include <linux/u64_stats_sync.h>
 
 struct dma_chan;
 struct software_node;
@@ -59,37 +60,70 @@ extern struct bus_type spi_bus_type;
  *                 maxsize limit
  */
 struct spi_statistics {
-	spinlock_t		lock; /* lock for the whole structure */
+	struct u64_stats_sync	syncp;
 
-	unsigned long		messages;
-	unsigned long		transfers;
-	unsigned long		errors;
-	unsigned long		timedout;
+	u64_stats_t		messages;
+	u64_stats_t		transfers;
+	u64_stats_t		errors;
+	u64_stats_t		timedout;
 
-	unsigned long		spi_sync;
-	unsigned long		spi_sync_immediate;
-	unsigned long		spi_async;
+	u64_stats_t		spi_sync;
+	u64_stats_t		spi_sync_immediate;
+	u64_stats_t		spi_async;
 
-	unsigned long long	bytes;
-	unsigned long long	bytes_rx;
-	unsigned long long	bytes_tx;
+	u64_stats_t		bytes;
+	u64_stats_t		bytes_rx;
+	u64_stats_t		bytes_tx;
 
 #define SPI_STATISTICS_HISTO_SIZE 17
-	unsigned long transfer_bytes_histo[SPI_STATISTICS_HISTO_SIZE];
+	u64_stats_t	transfer_bytes_histo[SPI_STATISTICS_HISTO_SIZE];
 
-	unsigned long transfers_split_maxsize;
+	u64_stats_t	transfers_split_maxsize;
 };
 
-#define SPI_STATISTICS_ADD_TO_FIELD(stats, field, count)	\
-	do {							\
-		unsigned long flags;				\
-		spin_lock_irqsave(&(stats)->lock, flags);	\
-		(stats)->field += count;			\
-		spin_unlock_irqrestore(&(stats)->lock, flags);	\
+#define SPI_STATISTICS_ADD_TO_FIELD(pcpu_stats, field, count)		\
+	do {								\
+		struct spi_statistics *__lstats = this_cpu_ptr(pcpu_stats); \
+		u64_stats_update_begin(&__lstats->syncp);		\
+		u64_stats_add(&__lstats->field, count);			\
+		u64_stats_update_end(&__lstats->syncp);			\
 	} while (0)
 
-#define SPI_STATISTICS_INCREMENT_FIELD(stats, field)	\
-	SPI_STATISTICS_ADD_TO_FIELD(stats, field, 1)
+#define SPI_STATISTICS_INCREMENT_FIELD(pcpu_stats, field)		\
+	do {								\
+		struct spi_statistics *__lstats = this_cpu_ptr(pcpu_stats); \
+		u64_stats_update_begin(&__lstats->syncp);		\
+		u64_stats_inc(&__lstats->field);			\
+		u64_stats_update_end(&__lstats->syncp);			\
+	} while (0)
+
+#define devm_spi_alloc_pcpu_stats(dev, type)				\
+({									\
+	typeof(type) __percpu *pcpu_stats = devm_alloc_percpu(dev, type);\
+	if (pcpu_stats) {						\
+		int __cpu;						\
+		for_each_possible_cpu(__cpu) {				\
+			typeof(type) *stat;				\
+			stat = per_cpu_ptr(pcpu_stats, __cpu);		\
+			u64_stats_init(&stat->syncp);			\
+		}							\
+	}								\
+	pcpu_stats;							\
+})
+
+#define spi_alloc_pcpu_stats(type)					\
+({									\
+	typeof(type) __percpu *pcpu_stats = alloc_percpu_gfp(type, GFP_KERNEL);\
+	if (pcpu_stats) {						\
+		int __cpu;						\
+		for_each_possible_cpu(__cpu) {				\
+			typeof(type) *stat;				\
+			stat = per_cpu_ptr(pcpu_stats, __cpu);		\
+			u64_stats_init(&stat->syncp);			\
+		}							\
+	}								\
+	pcpu_stats;							\
+})
 
 /**
  * struct spi_delay - SPI delay information
@@ -192,7 +226,7 @@ struct spi_device {
 	struct spi_delay	cs_inactive;
 
 	/* the statistics */
-	struct spi_statistics	statistics;
+	struct spi_statistics __percpu	*pcpu_statistics;
 
 	/*
 	 * likely need more hooks for more protocol options affecting how
@@ -643,7 +677,7 @@ struct spi_controller {
 	s8			max_native_cs;
 
 	/* statistics */
-	struct spi_statistics	statistics;
+	struct spi_statistics __percpu	*pcpu_statistics;
 
 	/* DMA channels for use with core dmaengine helpers */
 	struct dma_chan		*dma_tx;
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH] drivers: spi: spi.c: Convert statistics to per-cpu u64_stats_t
  2022-05-23 14:20 [PATCH] drivers: spi: spi.c: Convert statistics to per-cpu u64_stats_t David Jander
@ 2022-05-23 15:13 ` Andrew Lunn
  2022-05-23 18:56   ` Mark Brown
  2022-05-24  6:25   ` David Jander
  0 siblings, 2 replies; 5+ messages in thread
From: Andrew Lunn @ 2022-05-23 15:13 UTC (permalink / raw)
  To: David Jander; +Cc: Mark Brown, linux-spi, Marc Kleine-Budde

On Mon, May 23, 2022 at 04:20:09PM +0200, David Jander wrote:
> This change gives a dramatic performance improvement in the hot path,
> since many costly spin_lock_irqsave() calls can be avoided.

It is normal to back up such claims with numbers. Please include your
numbers here.

> 
> Suggested-by: Andrew Lunn <andrew@lunn.ch>
> Signed-off-by: David Jander <david@protonic.nl>
> ---
>  drivers/spi/spi.c       | 95 ++++++++++++++++++++++++-----------------
>  include/linux/spi/spi.h | 80 ++++++++++++++++++++++++----------
>  2 files changed, 114 insertions(+), 61 deletions(-)
> 
> diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
> index c4dd1200fe99..edc290e67b92 100644
> --- a/drivers/spi/spi.c
> +++ b/drivers/spi/spi.c
> @@ -33,6 +33,7 @@
>  #include <linux/idr.h>
>  #include <linux/platform_data/x86/apple.h>
>  #include <linux/ptp_clock_kernel.h>
> +#include <linux/percpu.h>
>  
>  #define CREATE_TRACE_POINTS
>  #include <trace/events/spi.h>
> @@ -111,6 +112,25 @@ static ssize_t driver_override_show(struct device *dev,
>  }
>  static DEVICE_ATTR_RW(driver_override);
>  
> +#define spi_pcpu_stats_totalize(ret, in, field)				\
> +do {									\
> +	int i;								\
> +	ret = 0;							\
> +	for_each_possible_cpu(i) {					\
> +		const struct spi_statistics *pcpu_stats;		\
> +		u64 inc;						\
> +		unsigned int start;					\
> +		pcpu_stats = per_cpu_ptr(in, i);			\
> +		do {							\
> +			start = u64_stats_fetch_begin_irq(		\
> +					&pcpu_stats->syncp);		\
> +			inc = u64_stats_read(&pcpu_stats->field);	\
> +		} while (u64_stats_fetch_retry_irq(			\
> +					&pcpu_stats->syncp, start));	\
> +		ret += inc;						\
> +	}								\
> +} while (0)

Looking at this from the perspective of a network reviewer, that is a
pretty big macro. netdev would probably not allow this. But macro
magic does seem normal in this file.

What also makes a difference here is the API to user space. With
netdev, it is not so much the absolute values of the counters, but the
relationship with other counters that is interesting. So there is a
method to get all the counters in one atomic snapshot. For that to
work, the loop doing the addition adds up all the counters, not just
one specific counter. So there is no need for 'field' which is the
source of all this macro magic. But it seems like SPI only ever
exports individual counters.

>  #define SPI_STATISTICS_ATTRS(field, file)				\
>  static ssize_t spi_controller_##field##_show(struct device *dev,	\
>  					     struct device_attribute *attr, \
> @@ -118,7 +138,7 @@ static ssize_t spi_controller_##field##_show(struct device *dev,	\
>  {									\
>  	struct spi_controller *ctlr = container_of(dev,			\
>  					 struct spi_controller, dev);	\
> -	return spi_statistics_##field##_show(&ctlr->statistics, buf);	\
> +	return spi_statistics_##field##_show(ctlr->pcpu_statistics, buf); \
>  }									\
>  static struct device_attribute dev_attr_spi_controller_##field = {	\
>  	.attr = { .name = file, .mode = 0444 },				\
> @@ -129,7 +149,7 @@ static ssize_t spi_device_##field##_show(struct device *dev,		\
>  					char *buf)			\
>  {									\
>  	struct spi_device *spi = to_spi_device(dev);			\
> -	return spi_statistics_##field##_show(&spi->statistics, buf);	\
> +	return spi_statistics_##field##_show(spi->pcpu_statistics, buf); \
>  }									\
>  static struct device_attribute dev_attr_spi_device_##field = {		\
>  	.attr = { .name = file, .mode = 0444 },				\

Not a criticism of this patch, but this is the old way to do this. It
is better to use DEVICE_ATTR_RO(), but that is a bigger rework.

> @@ -140,11 +160,10 @@ static struct device_attribute dev_attr_spi_device_##field = {		\
>  static ssize_t spi_statistics_##name##_show(struct spi_statistics *stat, \
>  					    char *buf)			\
>  {									\
> -	unsigned long flags;						\
>  	ssize_t len;							\
> -	spin_lock_irqsave(&stat->lock, flags);				\
> -	len = sysfs_emit(buf, format_string "\n", stat->field);		\
> -	spin_unlock_irqrestore(&stat->lock, flags);			\
> +	u64 val;							\
> +	spi_pcpu_stats_totalize(val, stat, field);			\
> +	len = sysfs_emit(buf, format_string "\n", val);			\

If you have made all the statistics u64, you don't need format_string.

> @@ -543,7 +562,7 @@ struct spi_device *spi_alloc_device(struct spi_controller *ctlr)
>  	spi->dev.release = spidev_release;
>  	spi->mode = ctlr->buswidth_override_bits;
>  
> -	spin_lock_init(&spi->statistics.lock);
> +	spi->pcpu_statistics = spi_alloc_pcpu_stats(struct spi_statistics);

It looks like you are missing error checking here.

> @@ -3042,7 +3061,7 @@ int spi_register_controller(struct spi_controller *ctlr)
>  		}
>  	}
>  	/* add statistics */
> -	spin_lock_init(&ctlr->statistics.lock);
> +	ctlr->pcpu_statistics = devm_spi_alloc_pcpu_stats(dev, struct spi_statistics);

This could fail, so you should check for NULL pointer.

> +#define devm_spi_alloc_pcpu_stats(dev, type)				\
> +({									\
> +	typeof(type) __percpu *pcpu_stats = devm_alloc_percpu(dev, type);\
> +	if (pcpu_stats) {						\
> +		int __cpu;						\
> +		for_each_possible_cpu(__cpu) {				\
> +			typeof(type) *stat;				\
> +			stat = per_cpu_ptr(pcpu_stats, __cpu);		\
> +			u64_stats_init(&stat->syncp);			\
> +		}							\
> +	}								\
> +	pcpu_stats;							\
> +})
> +
> +#define spi_alloc_pcpu_stats(type)					\
> +({									\
> +	typeof(type) __percpu *pcpu_stats = alloc_percpu_gfp(type, GFP_KERNEL);\
> +	if (pcpu_stats) {						\
> +		int __cpu;						\
> +		for_each_possible_cpu(__cpu) {				\
> +			typeof(type) *stat;				\
> +			stat = per_cpu_ptr(pcpu_stats, __cpu);		\
> +			u64_stats_init(&stat->syncp);			\
> +		}							\
> +	}								\
> +	pcpu_stats;							\
> +})

Do these need to be macros? How many different types are used?

   Andrew

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] drivers: spi: spi.c: Convert statistics to per-cpu u64_stats_t
  2022-05-23 15:13 ` Andrew Lunn
@ 2022-05-23 18:56   ` Mark Brown
  2022-05-24  6:25   ` David Jander
  1 sibling, 0 replies; 5+ messages in thread
From: Mark Brown @ 2022-05-23 18:56 UTC (permalink / raw)
  To: Andrew Lunn; +Cc: David Jander, linux-spi, Marc Kleine-Budde

[-- Attachment #1: Type: text/plain, Size: 713 bytes --]

On Mon, May 23, 2022 at 05:13:16PM +0200, Andrew Lunn wrote:
> On Mon, May 23, 2022 at 04:20:09PM +0200, David Jander wrote:

> > +			inc = u64_stats_read(&pcpu_stats->field);	\
> > +		} while (u64_stats_fetch_retry_irq(			\
> > +					&pcpu_stats->syncp, start));	\
> > +		ret += inc;						\
> > +	}								\
> > +} while (0)

> Looking at this from the perspective of a network reviewer, that is a
> pretty big macro. netdev would probably not allow this. But macro
> magic does seem normal in this file.

Yes, it's following the existing idiom so it's fine.  I didn't see any
issues Andrew didn't already mention, otherwise it looks pretty good to
me - it's a nice clean change for the performance improvement.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] drivers: spi: spi.c: Convert statistics to per-cpu u64_stats_t
  2022-05-23 15:13 ` Andrew Lunn
  2022-05-23 18:56   ` Mark Brown
@ 2022-05-24  6:25   ` David Jander
  2022-05-24 11:46     ` Mark Brown
  1 sibling, 1 reply; 5+ messages in thread
From: David Jander @ 2022-05-24  6:25 UTC (permalink / raw)
  To: Andrew Lunn; +Cc: Mark Brown, linux-spi, Marc Kleine-Budde


Hi Andrew,

Thanks a lot for the review.

On Mon, 23 May 2022 17:13:16 +0200
Andrew Lunn <andrew@lunn.ch> wrote:

> On Mon, May 23, 2022 at 04:20:09PM +0200, David Jander wrote:
> > This change gives a dramatic performance improvement in the hot path,
> > since many costly spin_lock_irqsave() calls can be avoided.  
> 
> It is normal to back up such claims with numbers. Please include your
> numbers here.

Ok, will add my measurement results here.

> > Suggested-by: Andrew Lunn <andrew@lunn.ch>
> > Signed-off-by: David Jander <david@protonic.nl>
> > ---
> >  drivers/spi/spi.c       | 95 ++++++++++++++++++++++++-----------------
> >  include/linux/spi/spi.h | 80 ++++++++++++++++++++++++----------
> >  2 files changed, 114 insertions(+), 61 deletions(-)
> > 
> > diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
> > index c4dd1200fe99..edc290e67b92 100644
> > --- a/drivers/spi/spi.c
> > +++ b/drivers/spi/spi.c
> > @@ -33,6 +33,7 @@
> >  #include <linux/idr.h>
> >  #include <linux/platform_data/x86/apple.h>
> >  #include <linux/ptp_clock_kernel.h>
> > +#include <linux/percpu.h>
> >  
> >  #define CREATE_TRACE_POINTS
> >  #include <trace/events/spi.h>
> > @@ -111,6 +112,25 @@ static ssize_t driver_override_show(struct device *dev,
> >  }
> >  static DEVICE_ATTR_RW(driver_override);
> >  
> > +#define spi_pcpu_stats_totalize(ret, in, field)				\
> > +do {									\
> > +	int i;								\
> > +	ret = 0;							\
> > +	for_each_possible_cpu(i) {					\
> > +		const struct spi_statistics *pcpu_stats;		\
> > +		u64 inc;						\
> > +		unsigned int start;					\
> > +		pcpu_stats = per_cpu_ptr(in, i);			\
> > +		do {							\
> > +			start = u64_stats_fetch_begin_irq(		\
> > +					&pcpu_stats->syncp);		\
> > +			inc = u64_stats_read(&pcpu_stats->field);	\
> > +		} while (u64_stats_fetch_retry_irq(			\
> > +					&pcpu_stats->syncp, start));	\
> > +		ret += inc;						\
> > +	}								\
> > +} while (0)  
> 
> Looking at this from the perspective of a network reviewer, that is a
> pretty big macro. netdev would probably not allow this. But macro
> magic does seem normal in this file.
> 
> What also makes a difference here is the API to user space. With
> netdev, it is not so much the absolute values of the counters, but the
> relationship with other counters that is interesting. So there is a
> method to get all the counters in one atomic snapshot. For that to
> work, the loop doing the addition adds up all the counters, not just
> one specific counter. So there is no need for 'field' which is the
> source of all this macro magic. But it seems like SPI only ever
> exports individual counters.

Yes, I initially started out just totalling all the members in one loop, but
after noticing they are never used together, I changed it.
Going to go with Marks comment here and leave it as is.

> >  #define SPI_STATISTICS_ATTRS(field, file)				\
> >  static ssize_t spi_controller_##field##_show(struct device *dev,	\
> >  					     struct device_attribute *attr, \
> > @@ -118,7 +138,7 @@ static ssize_t spi_controller_##field##_show(struct device *dev,	\
> >  {									\
> >  	struct spi_controller *ctlr = container_of(dev,			\
> >  					 struct spi_controller, dev);	\
> > -	return spi_statistics_##field##_show(&ctlr->statistics, buf);	\
> > +	return spi_statistics_##field##_show(ctlr->pcpu_statistics, buf); \
> >  }									\
> >  static struct device_attribute dev_attr_spi_controller_##field = {	\
> >  	.attr = { .name = file, .mode = 0444 },				\
> > @@ -129,7 +149,7 @@ static ssize_t spi_device_##field##_show(struct device *dev,		\
> >  					char *buf)			\
> >  {									\
> >  	struct spi_device *spi = to_spi_device(dev);			\
> > -	return spi_statistics_##field##_show(&spi->statistics, buf);	\
> > +	return spi_statistics_##field##_show(spi->pcpu_statistics, buf); \
> >  }									\
> >  static struct device_attribute dev_attr_spi_device_##field = {		\
> >  	.attr = { .name = file, .mode = 0444 },				\  
> 
> Not a criticism of this patch, but this is the old way to do this. It
> is better to use DEVICE_ATTR_RO(), but that is a bigger rework.

I suppose it is better to do this in a separate patch? @Mark, should I leave
it as is, or do you require a "modernization" for this first?

> > @@ -140,11 +160,10 @@ static struct device_attribute dev_attr_spi_device_##field = {		\
> >  static ssize_t spi_statistics_##name##_show(struct spi_statistics *stat, \
> >  					    char *buf)			\
> >  {									\
> > -	unsigned long flags;						\
> >  	ssize_t len;							\
> > -	spin_lock_irqsave(&stat->lock, flags);				\
> > -	len = sysfs_emit(buf, format_string "\n", stat->field);		\
> > -	spin_unlock_irqrestore(&stat->lock, flags);			\
> > +	u64 val;							\
> > +	spi_pcpu_stats_totalize(val, stat, field);			\
> > +	len = sysfs_emit(buf, format_string "\n", val);			\  
> 
> If you have made all the statistics u64, you don't need format_string.

Ack. Good point. I wasn't convinced to leave it all u64 till the end, so I
didn't think of removing the format strings. Will fix that.

> > @@ -543,7 +562,7 @@ struct spi_device *spi_alloc_device(struct spi_controller *ctlr)
> >  	spi->dev.release = spidev_release;
> >  	spi->mode = ctlr->buswidth_override_bits;
> >  
> > -	spin_lock_init(&spi->statistics.lock);
> > +	spi->pcpu_statistics = spi_alloc_pcpu_stats(struct spi_statistics);  
> 
> It looks like you are missing error checking here.

Oops. Will fix that.

> > @@ -3042,7 +3061,7 @@ int spi_register_controller(struct spi_controller *ctlr)
> >  		}
> >  	}
> >  	/* add statistics */
> > -	spin_lock_init(&ctlr->statistics.lock);
> > +	ctlr->pcpu_statistics = devm_spi_alloc_pcpu_stats(dev, struct spi_statistics);  
> 
> This could fail, so you should check for NULL pointer.

Thanks. Will fix.

> > +#define devm_spi_alloc_pcpu_stats(dev, type)				\
> > +({									\
> > +	typeof(type) __percpu *pcpu_stats = devm_alloc_percpu(dev, type);\
> > +	if (pcpu_stats) {						\
> > +		int __cpu;						\
> > +		for_each_possible_cpu(__cpu) {				\
> > +			typeof(type) *stat;				\
> > +			stat = per_cpu_ptr(pcpu_stats, __cpu);		\
> > +			u64_stats_init(&stat->syncp);			\
> > +		}							\
> > +	}								\
> > +	pcpu_stats;							\
> > +})
> > +
> > +#define spi_alloc_pcpu_stats(type)					\
> > +({									\
> > +	typeof(type) __percpu *pcpu_stats = alloc_percpu_gfp(type, GFP_KERNEL);\
> > +	if (pcpu_stats) {						\
> > +		int __cpu;						\
> > +		for_each_possible_cpu(__cpu) {				\
> > +			typeof(type) *stat;				\
> > +			stat = per_cpu_ptr(pcpu_stats, __cpu);		\
> > +			u64_stats_init(&stat->syncp);			\
> > +		}							\
> > +	}								\
> > +	pcpu_stats;							\
> > +})  
> 
> Do these need to be macros? How many different types are used?

Good question. Probably not. I initially had just one macro that I intended to
use twice (the devm variant), but later realized that there is a
chicken-and-egg problem in the code, where the bus (and thus the statistics)
need to be operational before the client device is fully registered, so device
resource allocations aren't allowed at that point.
But this reminds me that not being a device resource, it needs explicit
freeing and I haven't done that. Will fix this memory leak in the next version.

Best regards,

-- 
David Jander


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] drivers: spi: spi.c: Convert statistics to per-cpu u64_stats_t
  2022-05-24  6:25   ` David Jander
@ 2022-05-24 11:46     ` Mark Brown
  0 siblings, 0 replies; 5+ messages in thread
From: Mark Brown @ 2022-05-24 11:46 UTC (permalink / raw)
  To: David Jander; +Cc: Andrew Lunn, linux-spi, Marc Kleine-Budde

[-- Attachment #1: Type: text/plain, Size: 592 bytes --]

On Tue, May 24, 2022 at 08:25:06AM +0200, David Jander wrote:
> Andrew Lunn <andrew@lunn.ch> wrote:
> > On Mon, May 23, 2022 at 04:20:09PM +0200, David Jander wrote:

> > >  	.attr = { .name = file, .mode = 0444 },				\  

> > Not a criticism of this patch, but this is the old way to do this. It
> > is better to use DEVICE_ATTR_RO(), but that is a bigger rework.

> I suppose it is better to do this in a separate patch? @Mark, should I leave
> it as is, or do you require a "modernization" for this first?

I'd do it as a followup to this patch, that's just a random cleanup.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2022-05-24 11:46 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-05-23 14:20 [PATCH] drivers: spi: spi.c: Convert statistics to per-cpu u64_stats_t David Jander
2022-05-23 15:13 ` Andrew Lunn
2022-05-23 18:56   ` Mark Brown
2022-05-24  6:25   ` David Jander
2022-05-24 11:46     ` Mark Brown

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.