* [PATCH net-next 0/4] sfc: NUMA support
@ 2015-10-28 15:00 Shradha Shah
2015-10-28 15:01 ` [PATCH net-next 1/4] sfc: use __GFP_NOWARN when allocating RX pages from atomic context Shradha Shah
` (3 more replies)
0 siblings, 4 replies; 8+ messages in thread
From: Shradha Shah @ 2015-10-28 15:00 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
This patch series adds support for
- allocating rx pages local to the interrupt
- setting affinity hint to influence IRQs to be allocated on the
same NUMA node as the one where the card resides.
Alexandra Kossovsky (1):
sfc: use __GFP_NOWARN when allocating RX pages from atomic context.
Bert Kenward (2):
sfc: Use cpu_to_mem() to support memoryless nodes
sfc: set and clear interrupt affinity hints
Daniel Pieczko (1):
sfc: allocate rx pages on the same node as the interrupt
drivers/net/ethernet/sfc/efx.c | 36 +++++++++++++++++++++++++++++++++++
drivers/net/ethernet/sfc/net_driver.h | 3 +++
drivers/net/ethernet/sfc/rx.c | 18 +++++++++++++++---
3 files changed, 54 insertions(+), 3 deletions(-)
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH net-next 1/4] sfc: use __GFP_NOWARN when allocating RX pages from atomic context.
2015-10-28 15:00 [PATCH net-next 0/4] sfc: NUMA support Shradha Shah
@ 2015-10-28 15:01 ` Shradha Shah
2015-10-28 15:01 ` [PATCH net-next 2/4] sfc: allocate rx pages on the same node as the interrupt Shradha Shah
` (2 subsequent siblings)
3 siblings, 0 replies; 8+ messages in thread
From: Shradha Shah @ 2015-10-28 15:01 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
From: Alexandra Kossovsky <Alexandra.Kossovsky@oktetlabs.ru>
If we fail to allocate a page when in atomic context this is
handled by scheduling a fill in non-atomic context.
As such, a warning is not needed.
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/rx.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
index 809ea461..3f0e129 100644
--- a/drivers/net/ethernet/sfc/rx.c
+++ b/drivers/net/ethernet/sfc/rx.c
@@ -163,8 +163,15 @@ static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic)
do {
page = efx_reuse_page(rx_queue);
if (page == NULL) {
+ /* GFP_ATOMIC may fail because of various reasons,
+ * and we re-schedule rx_fill from non-atomic
+ * context in such a case. So, use __GFP_NO_WARN
+ * in case of atomic.
+ */
page = alloc_pages(__GFP_COLD | __GFP_COMP |
- (atomic ? GFP_ATOMIC : GFP_KERNEL),
+ (atomic ?
+ (GFP_ATOMIC | __GFP_NOWARN)
+ : GFP_KERNEL),
efx->rx_buffer_order);
if (unlikely(page == NULL))
return -ENOMEM;
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net-next 2/4] sfc: allocate rx pages on the same node as the interrupt
2015-10-28 15:00 [PATCH net-next 0/4] sfc: NUMA support Shradha Shah
2015-10-28 15:01 ` [PATCH net-next 1/4] sfc: use __GFP_NOWARN when allocating RX pages from atomic context Shradha Shah
@ 2015-10-28 15:01 ` Shradha Shah
2015-10-28 15:59 ` Eric Dumazet
2015-10-28 15:01 ` [PATCH net-next 3/4] sfc: Use cpu_to_mem() to support memoryless nodes Shradha Shah
2015-10-28 15:02 ` [PATCH net-next 4/4] sfc: set and clear interrupt affinity hints Shradha Shah
3 siblings, 1 reply; 8+ messages in thread
From: Shradha Shah @ 2015-10-28 15:01 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
From: Daniel Pieczko <dpieczko@solarflare.com>
When the interrupt servicing a channel is on a NUMA node that is
not local to the device, performance is improved by allocating
rx pages on the node local to the interrupt (remote to the device)
The performance-optimal case, where interrupts and applications
are pinned to CPUs on the same node as the device, is not altered
by this change.
This change gave a 1% improvement in transaction rate using Nginx
with all interrupts and Nginx threads on the node remote to the
device. It also gave a small reduction in round-trip latency,
again with the interrupt and application on a different node to
the device.
Allocating rx pages based on the channel->irq_node value is only
valid for the initial driver-load interrupt affinities; if an
interrupt is moved later, the wrong node may be used for the
allocation.
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/efx.c | 1 +
drivers/net/ethernet/sfc/net_driver.h | 3 +++
drivers/net/ethernet/sfc/rx.c | 14 +++++++++-----
3 files changed, 13 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
index 974637d..89fbd03 100644
--- a/drivers/net/ethernet/sfc/efx.c
+++ b/drivers/net/ethernet/sfc/efx.c
@@ -445,6 +445,7 @@ efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel)
channel->efx = efx;
channel->channel = i;
channel->type = &efx_default_channel_type;
+ channel->irq_node = NUMA_NO_NODE;
for (j = 0; j < EFX_TXQ_TYPES; j++) {
tx_queue = &channel->tx_queue[j];
diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h
index ad56231..0ab9080a 100644
--- a/drivers/net/ethernet/sfc/net_driver.h
+++ b/drivers/net/ethernet/sfc/net_driver.h
@@ -419,6 +419,7 @@ enum efx_sync_events_state {
* @sync_events_state: Current state of sync events on this channel
* @sync_timestamp_major: Major part of the last ptp sync event
* @sync_timestamp_minor: Minor part of the last ptp sync event
+ * @irq_node: NUMA node of interrupt
*/
struct efx_channel {
struct efx_nic *efx;
@@ -477,6 +478,8 @@ struct efx_channel {
enum efx_sync_events_state sync_events_state;
u32 sync_timestamp_major;
u32 sync_timestamp_minor;
+
+ int irq_node;
};
#ifdef CONFIG_NET_RX_BUSY_POLL
diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
index 3f0e129..c5ef1e8 100644
--- a/drivers/net/ethernet/sfc/rx.c
+++ b/drivers/net/ethernet/sfc/rx.c
@@ -168,11 +168,15 @@ static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic)
* context in such a case. So, use __GFP_NO_WARN
* in case of atomic.
*/
- page = alloc_pages(__GFP_COLD | __GFP_COMP |
- (atomic ?
- (GFP_ATOMIC | __GFP_NOWARN)
- : GFP_KERNEL),
- efx->rx_buffer_order);
+ struct efx_channel *channel;
+
+ channel = efx_rx_queue_channel(rx_queue);
+ page = alloc_pages_node(channel->irq_node, __GFP_COMP |
+ (atomic ?
+ (GFP_ATOMIC | __GFP_NOWARN)
+ : GFP_KERNEL),
+ efx->rx_buffer_order);
+
if (unlikely(page == NULL))
return -ENOMEM;
dma_addr =
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net-next 3/4] sfc: Use cpu_to_mem() to support memoryless nodes
2015-10-28 15:00 [PATCH net-next 0/4] sfc: NUMA support Shradha Shah
2015-10-28 15:01 ` [PATCH net-next 1/4] sfc: use __GFP_NOWARN when allocating RX pages from atomic context Shradha Shah
2015-10-28 15:01 ` [PATCH net-next 2/4] sfc: allocate rx pages on the same node as the interrupt Shradha Shah
@ 2015-10-28 15:01 ` Shradha Shah
2015-10-28 15:02 ` [PATCH net-next 4/4] sfc: set and clear interrupt affinity hints Shradha Shah
3 siblings, 0 replies; 8+ messages in thread
From: Shradha Shah @ 2015-10-28 15:01 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
From: Bert Kenward <bkenward@solarflare.com>
With CONFIG_HAVE_MEMORYLESS_NODES cpu_to_node() may return
nodes without memory, which is not a good choice when later
using that to allocate memory. cpu_to_mem() instead provides
the most appropriate NUMA node to allocate from.
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/efx.c | 2 +-
drivers/net/ethernet/sfc/net_driver.h | 4 ++--
drivers/net/ethernet/sfc/rx.c | 3 ++-
3 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
index 89fbd03..84f9e90 100644
--- a/drivers/net/ethernet/sfc/efx.c
+++ b/drivers/net/ethernet/sfc/efx.c
@@ -445,7 +445,7 @@ efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel)
channel->efx = efx;
channel->channel = i;
channel->type = &efx_default_channel_type;
- channel->irq_node = NUMA_NO_NODE;
+ channel->irq_mem_node = NUMA_NO_NODE;
for (j = 0; j < EFX_TXQ_TYPES; j++) {
tx_queue = &channel->tx_queue[j];
diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h
index 0ab9080a..bab6cc0 100644
--- a/drivers/net/ethernet/sfc/net_driver.h
+++ b/drivers/net/ethernet/sfc/net_driver.h
@@ -419,7 +419,7 @@ enum efx_sync_events_state {
* @sync_events_state: Current state of sync events on this channel
* @sync_timestamp_major: Major part of the last ptp sync event
* @sync_timestamp_minor: Minor part of the last ptp sync event
- * @irq_node: NUMA node of interrupt
+ * @irq_mem_node: Memory NUMA node of interrupt
*/
struct efx_channel {
struct efx_nic *efx;
@@ -479,7 +479,7 @@ struct efx_channel {
u32 sync_timestamp_major;
u32 sync_timestamp_minor;
- int irq_node;
+ int irq_mem_node;
};
#ifdef CONFIG_NET_RX_BUSY_POLL
diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
index c5ef1e8..095d1af 100644
--- a/drivers/net/ethernet/sfc/rx.c
+++ b/drivers/net/ethernet/sfc/rx.c
@@ -171,7 +171,8 @@ static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic)
struct efx_channel *channel;
channel = efx_rx_queue_channel(rx_queue);
- page = alloc_pages_node(channel->irq_node, __GFP_COMP |
+ page = alloc_pages_node(channel->irq_mem_node,
+ __GFP_COMP |
(atomic ?
(GFP_ATOMIC | __GFP_NOWARN)
: GFP_KERNEL),
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net-next 4/4] sfc: set and clear interrupt affinity hints
2015-10-28 15:00 [PATCH net-next 0/4] sfc: NUMA support Shradha Shah
` (2 preceding siblings ...)
2015-10-28 15:01 ` [PATCH net-next 3/4] sfc: Use cpu_to_mem() to support memoryless nodes Shradha Shah
@ 2015-10-28 15:02 ` Shradha Shah
2015-10-28 18:15 ` Sergei Shtylyov
3 siblings, 1 reply; 8+ messages in thread
From: Shradha Shah @ 2015-10-28 15:02 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
From: Bert Kenward <bkenward@solarflare.com>
Use cpumask_local_spread to provide interrupt affinity hints
for each queue. This will spread interrupts across NUMA local
CPUs first, extending to remote nodes if needed.
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/efx.c | 35 +++++++++++++++++++++++++++++++++++
1 file changed, 35 insertions(+)
diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
index 84f9e90..93c4c0e 100644
--- a/drivers/net/ethernet/sfc/efx.c
+++ b/drivers/net/ethernet/sfc/efx.c
@@ -1489,6 +1489,30 @@ static int efx_probe_interrupts(struct efx_nic *efx)
return 0;
}
+#if defined(CONFIG_SMP)
+static void efx_set_interrupt_affinity(struct efx_nic *efx)
+{
+ struct efx_channel *channel;
+ unsigned int cpu;
+
+ efx_for_each_channel(channel, efx) {
+ cpu = cpumask_local_spread(channel->channel,
+ pcibus_to_node(efx->pci_dev->bus));
+
+ irq_set_affinity_hint(channel->irq, cpumask_of(cpu));
+ channel->irq_mem_node = cpu_to_mem(cpu);
+ }
+}
+
+static void efx_clear_interrupt_affinity(struct efx_nic *efx)
+{
+ struct efx_channel *channel;
+
+ efx_for_each_channel(channel, efx)
+ irq_set_affinity_hint(channel->irq, NULL);
+}
+#endif /* CONFIG_SMP */
+
static int efx_soft_enable_interrupts(struct efx_nic *efx)
{
struct efx_channel *channel, *end_channel;
@@ -2932,6 +2956,9 @@ static void efx_pci_remove_main(struct efx_nic *efx)
cancel_work_sync(&efx->reset_work);
efx_disable_interrupts(efx);
+#if defined(CONFIG_SMP)
+ efx_clear_interrupt_affinity(efx);
+#endif
efx_nic_fini_interrupt(efx);
efx_fini_port(efx);
efx->type->fini(efx);
@@ -3081,6 +3108,11 @@ static int efx_pci_probe_main(struct efx_nic *efx)
rc = efx_nic_init_interrupt(efx);
if (rc)
goto fail5;
+
+#if defined(CONFIG_SMP)
+ efx_set_interrupt_affinity(efx);
+#endif
+
rc = efx_enable_interrupts(efx);
if (rc)
goto fail6;
@@ -3088,6 +3120,9 @@ static int efx_pci_probe_main(struct efx_nic *efx)
return 0;
fail6:
+#if defined(CONFIG_SMP)
+ efx_clear_interrupt_affinity(efx);
+#endif
efx_nic_fini_interrupt(efx);
fail5:
efx_fini_port(efx);
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH net-next 2/4] sfc: allocate rx pages on the same node as the interrupt
2015-10-28 15:01 ` [PATCH net-next 2/4] sfc: allocate rx pages on the same node as the interrupt Shradha Shah
@ 2015-10-28 15:59 ` Eric Dumazet
2015-11-02 13:17 ` Daniel Pieczko (dpieczko)
0 siblings, 1 reply; 8+ messages in thread
From: Eric Dumazet @ 2015-10-28 15:59 UTC (permalink / raw)
To: Shradha Shah; +Cc: David Miller, netdev, linux-net-drivers
On Wed, 2015-10-28 at 15:01 +0000, Shradha Shah wrote:
> From: Daniel Pieczko <dpieczko@solarflare.com>
>
> When the interrupt servicing a channel is on a NUMA node that is
> not local to the device, performance is improved by allocating
> rx pages on the node local to the interrupt (remote to the device)
>
> The performance-optimal case, where interrupts and applications
> are pinned to CPUs on the same node as the device, is not altered
> by this change.
>
> This change gave a 1% improvement in transaction rate using Nginx
> with all interrupts and Nginx threads on the node remote to the
> device. It also gave a small reduction in round-trip latency,
> again with the interrupt and application on a different node to
> the device.
>
> Allocating rx pages based on the channel->irq_node value is only
> valid for the initial driver-load interrupt affinities; if an
> interrupt is moved later, the wrong node may be used for the
> allocation.
>
> Signed-off-by: Shradha Shah <sshah@solarflare.com>
> ---
> drivers/net/ethernet/sfc/efx.c | 1 +
> drivers/net/ethernet/sfc/net_driver.h | 3 +++
> drivers/net/ethernet/sfc/rx.c | 14 +++++++++-----
> 3 files changed, 13 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
> index 974637d..89fbd03 100644
> --- a/drivers/net/ethernet/sfc/efx.c
> +++ b/drivers/net/ethernet/sfc/efx.c
> @@ -445,6 +445,7 @@ efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel)
> channel->efx = efx;
> channel->channel = i;
> channel->type = &efx_default_channel_type;
> + channel->irq_node = NUMA_NO_NODE;
>
> for (j = 0; j < EFX_TXQ_TYPES; j++) {
> tx_queue = &channel->tx_queue[j];
> diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h
> index ad56231..0ab9080a 100644
> --- a/drivers/net/ethernet/sfc/net_driver.h
> +++ b/drivers/net/ethernet/sfc/net_driver.h
> @@ -419,6 +419,7 @@ enum efx_sync_events_state {
> * @sync_events_state: Current state of sync events on this channel
> * @sync_timestamp_major: Major part of the last ptp sync event
> * @sync_timestamp_minor: Minor part of the last ptp sync event
> + * @irq_node: NUMA node of interrupt
> */
> struct efx_channel {
> struct efx_nic *efx;
> @@ -477,6 +478,8 @@ struct efx_channel {
> enum efx_sync_events_state sync_events_state;
> u32 sync_timestamp_major;
> u32 sync_timestamp_minor;
> +
> + int irq_node;
> };
>
> #ifdef CONFIG_NET_RX_BUSY_POLL
> diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
> index 3f0e129..c5ef1e8 100644
> --- a/drivers/net/ethernet/sfc/rx.c
> +++ b/drivers/net/ethernet/sfc/rx.c
> @@ -168,11 +168,15 @@ static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic)
> * context in such a case. So, use __GFP_NO_WARN
> * in case of atomic.
> */
> - page = alloc_pages(__GFP_COLD | __GFP_COMP |
> - (atomic ?
> - (GFP_ATOMIC | __GFP_NOWARN)
> - : GFP_KERNEL),
> - efx->rx_buffer_order);
> + struct efx_channel *channel;
> +
> + channel = efx_rx_queue_channel(rx_queue);
> + page = alloc_pages_node(channel->irq_node, __GFP_COMP |
> + (atomic ?
> + (GFP_ATOMIC | __GFP_NOWARN)
> + : GFP_KERNEL),
> + efx->rx_buffer_order);
> +
> if (unlikely(page == NULL))
> return -ENOMEM;
> dma_addr =
>
Sorry, I do not understand this patch, and why the following one is not
squashed on this one.
irq_node is always NUMA_NO_NODE (in this patch)
So you claim a 1% improvement, switching from alloc_pages(...) to
alloc_pages_node(NUMA_NO_NODE, ...) ???
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH net-next 4/4] sfc: set and clear interrupt affinity hints
2015-10-28 15:02 ` [PATCH net-next 4/4] sfc: set and clear interrupt affinity hints Shradha Shah
@ 2015-10-28 18:15 ` Sergei Shtylyov
0 siblings, 0 replies; 8+ messages in thread
From: Sergei Shtylyov @ 2015-10-28 18:15 UTC (permalink / raw)
To: Shradha Shah, David Miller; +Cc: netdev, linux-net-drivers
Hello.
On 10/28/2015 06:02 PM, Shradha Shah wrote:
> From: Bert Kenward <bkenward@solarflare.com>
>
> Use cpumask_local_spread to provide interrupt affinity hints
> for each queue. This will spread interrupts across NUMA local
> CPUs first, extending to remote nodes if needed.
>
> Signed-off-by: Shradha Shah <sshah@solarflare.com>
> ---
> drivers/net/ethernet/sfc/efx.c | 35 +++++++++++++++++++++++++++++++++++
> 1 file changed, 35 insertions(+)
>
> diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
> index 84f9e90..93c4c0e 100644
> --- a/drivers/net/ethernet/sfc/efx.c
> +++ b/drivers/net/ethernet/sfc/efx.c
> @@ -1489,6 +1489,30 @@ static int efx_probe_interrupts(struct efx_nic *efx)
> return 0;
> }
>
> +#if defined(CONFIG_SMP)
> +static void efx_set_interrupt_affinity(struct efx_nic *efx)
> +{
> + struct efx_channel *channel;
> + unsigned int cpu;
> +
> + efx_for_each_channel(channel, efx) {
> + cpu = cpumask_local_spread(channel->channel,
> + pcibus_to_node(efx->pci_dev->bus));
> +
> + irq_set_affinity_hint(channel->irq, cpumask_of(cpu));
> + channel->irq_mem_node = cpu_to_mem(cpu);
> + }
> +}
> +
> +static void efx_clear_interrupt_affinity(struct efx_nic *efx)
> +{
> + struct efx_channel *channel;
> +
> + efx_for_each_channel(channel, efx)
> + irq_set_affinity_hint(channel->irq, NULL);
> +}
> +#endif /* CONFIG_SMP */
> +
> static int efx_soft_enable_interrupts(struct efx_nic *efx)
> {
> struct efx_channel *channel, *end_channel;
> @@ -2932,6 +2956,9 @@ static void efx_pci_remove_main(struct efx_nic *efx)
> cancel_work_sync(&efx->reset_work);
>
> efx_disable_interrupts(efx);
> +#if defined(CONFIG_SMP)
> + efx_clear_interrupt_affinity(efx);
> +#endif
Please just define empty function for !SMP case instead of the ugly
#ifdef'fery in the functiojn bodies.
[...]
MBR, Sergei
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH net-next 2/4] sfc: allocate rx pages on the same node as the interrupt
2015-10-28 15:59 ` Eric Dumazet
@ 2015-11-02 13:17 ` Daniel Pieczko (dpieczko)
0 siblings, 0 replies; 8+ messages in thread
From: Daniel Pieczko (dpieczko) @ 2015-11-02 13:17 UTC (permalink / raw)
To: Eric Dumazet, Shradha Shah; +Cc: David Miller, netdev, linux-net-drivers
On 28/10/15 15:59, Eric Dumazet wrote:
> On Wed, 2015-10-28 at 15:01 +0000, Shradha Shah wrote:
>> From: Daniel Pieczko <dpieczko@solarflare.com>
>>
>> When the interrupt servicing a channel is on a NUMA node that is
>> not local to the device, performance is improved by allocating
>> rx pages on the node local to the interrupt (remote to the device)
>>
>> The performance-optimal case, where interrupts and applications
>> are pinned to CPUs on the same node as the device, is not altered
>> by this change.
>>
>> This change gave a 1% improvement in transaction rate using Nginx
>> with all interrupts and Nginx threads on the node remote to the
>> device. It also gave a small reduction in round-trip latency,
>> again with the interrupt and application on a different node to
>> the device.
>>
>> Allocating rx pages based on the channel->irq_node value is only
>> valid for the initial driver-load interrupt affinities; if an
>> interrupt is moved later, the wrong node may be used for the
>> allocation.
>>
>> Signed-off-by: Shradha Shah <sshah@solarflare.com>
>> ---
>> drivers/net/ethernet/sfc/efx.c | 1 +
>> drivers/net/ethernet/sfc/net_driver.h | 3 +++
>> drivers/net/ethernet/sfc/rx.c | 14 +++++++++-----
>> 3 files changed, 13 insertions(+), 5 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
>> index 974637d..89fbd03 100644
>> --- a/drivers/net/ethernet/sfc/efx.c
>> +++ b/drivers/net/ethernet/sfc/efx.c
>> @@ -445,6 +445,7 @@ efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel)
>> channel->efx = efx;
>> channel->channel = i;
>> channel->type = &efx_default_channel_type;
>> + channel->irq_node = NUMA_NO_NODE;
>>
>> for (j = 0; j < EFX_TXQ_TYPES; j++) {
>> tx_queue = &channel->tx_queue[j];
>> diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h
>> index ad56231..0ab9080a 100644
>> --- a/drivers/net/ethernet/sfc/net_driver.h
>> +++ b/drivers/net/ethernet/sfc/net_driver.h
>> @@ -419,6 +419,7 @@ enum efx_sync_events_state {
>> * @sync_events_state: Current state of sync events on this channel
>> * @sync_timestamp_major: Major part of the last ptp sync event
>> * @sync_timestamp_minor: Minor part of the last ptp sync event
>> + * @irq_node: NUMA node of interrupt
>> */
>> struct efx_channel {
>> struct efx_nic *efx;
>> @@ -477,6 +478,8 @@ struct efx_channel {
>> enum efx_sync_events_state sync_events_state;
>> u32 sync_timestamp_major;
>> u32 sync_timestamp_minor;
>> +
>> + int irq_node;
>> };
>>
>> #ifdef CONFIG_NET_RX_BUSY_POLL
>> diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
>> index 3f0e129..c5ef1e8 100644
>> --- a/drivers/net/ethernet/sfc/rx.c
>> +++ b/drivers/net/ethernet/sfc/rx.c
>> @@ -168,11 +168,15 @@ static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic)
>> * context in such a case. So, use __GFP_NO_WARN
>> * in case of atomic.
>> */
>> - page = alloc_pages(__GFP_COLD | __GFP_COMP |
>> - (atomic ?
>> - (GFP_ATOMIC | __GFP_NOWARN)
>> - : GFP_KERNEL),
>> - efx->rx_buffer_order);
>> + struct efx_channel *channel;
>> +
>> + channel = efx_rx_queue_channel(rx_queue);
>> + page = alloc_pages_node(channel->irq_node, __GFP_COMP |
>> + (atomic ?
>> + (GFP_ATOMIC | __GFP_NOWARN)
>> + : GFP_KERNEL),
>> + efx->rx_buffer_order);
>> +
>> if (unlikely(page == NULL))
>> return -ENOMEM;
>> dma_addr =
>>
>
> Sorry, I do not understand this patch, and why the following one is not
> squashed on this one.
>
> irq_node is always NUMA_NO_NODE (in this patch)
>
> So you claim a 1% improvement, switching from alloc_pages(...) to
> alloc_pages_node(NUMA_NO_NODE, ...) ???
>
You're correct that this doesn't make sense as it is. There is something missing in this patch (channel->irq_node should be set) and also changing the order of some patches could make this clearer. The series will need to be resent.
Daniel
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2015-11-02 13:25 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-10-28 15:00 [PATCH net-next 0/4] sfc: NUMA support Shradha Shah
2015-10-28 15:01 ` [PATCH net-next 1/4] sfc: use __GFP_NOWARN when allocating RX pages from atomic context Shradha Shah
2015-10-28 15:01 ` [PATCH net-next 2/4] sfc: allocate rx pages on the same node as the interrupt Shradha Shah
2015-10-28 15:59 ` Eric Dumazet
2015-11-02 13:17 ` Daniel Pieczko (dpieczko)
2015-10-28 15:01 ` [PATCH net-next 3/4] sfc: Use cpu_to_mem() to support memoryless nodes Shradha Shah
2015-10-28 15:02 ` [PATCH net-next 4/4] sfc: set and clear interrupt affinity hints Shradha Shah
2015-10-28 18:15 ` Sergei Shtylyov
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.