DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] [PATCH v2] rte_ethdev: add link support flag
@ 2014-06-19 22:12 Stephen Hemminger
  2014-06-20 13:06 ` Thomas Monjalon
  0 siblings, 1 reply; 4+ messages in thread
From: Stephen Hemminger @ 2014-06-19 22:12 UTC (permalink / raw)
  To: dev

Only some devices support the link state interrupt configuration option.
Link state control does not work in virtual drivers
(virtio, vmxnet3, igbvf, and ixgbevf). Instead of having the application
try and guess whether it will work or not provide a driver flag that
can be checked instead.

Note: if device driver doesn't support link state control, what
would happen previously is that the code would never detect link
transitions. This prevents that.

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>

---
 lib/librte_eal/common/include/rte_pci.h |    2 ++
 lib/librte_ether/rte_ethdev.c           |   14 ++++++++++++++
 lib/librte_pmd_e1000/em_ethdev.c        |    2 +-
 lib/librte_pmd_e1000/igb_ethdev.c       |    2 +-
 lib/librte_pmd_ixgbe/ixgbe_ethdev.c     |    2 +-
 5 files changed, 19 insertions(+), 3 deletions(-)

--- a/lib/librte_eal/common/include/rte_pci.h	2014-06-19 14:56:44.188081500 -0700
+++ b/lib/librte_eal/common/include/rte_pci.h	2014-06-19 14:56:44.184081479 -0700
@@ -197,6 +197,8 @@ struct rte_pci_driver {
 #define RTE_PCI_DRV_MULTIPLE 0x0002
 /** Device needs to be unbound even if no module is provided */
 #define RTE_PCI_DRV_FORCE_UNBIND 0x0004
+/** Device driver supports link state interrupt */
+#define RTE_PCI_DRV_LSC	0x0008
 
 /**< Internal use only - Macro used by pci addr parsing functions **/
 #define GET_PCIADDR_FIELD(in, fd, lim, dlm)                   \
--- a/lib/librte_pmd_e1000/em_ethdev.c	2014-06-19 14:56:44.188081500 -0700
+++ b/lib/librte_pmd_e1000/em_ethdev.c	2014-06-19 14:57:50.460429276 -0700
@@ -286,7 +286,7 @@ static struct eth_driver rte_em_pmd = {
 	{
 		.name = "rte_em_pmd",
 		.id_table = pci_id_em_map,
-		.drv_flags = RTE_PCI_DRV_NEED_MAPPING,
+		.drv_flags = RTE_PCI_DRV_NEED_MAPPING | RTE_PCI_DRV_LSC,
 	},
 	.eth_dev_init = eth_em_dev_init,
 	.dev_private_size = sizeof(struct e1000_adapter),
--- a/lib/librte_pmd_e1000/igb_ethdev.c	2014-06-19 14:56:44.188081500 -0700
+++ b/lib/librte_pmd_e1000/igb_ethdev.c	2014-06-19 14:58:14.268554028 -0700
@@ -662,7 +662,7 @@ static struct eth_driver rte_igb_pmd = {
 	{
 		.name = "rte_igb_pmd",
 		.id_table = pci_id_igb_map,
-		.drv_flags = RTE_PCI_DRV_NEED_MAPPING,
+		.drv_flags = RTE_PCI_DRV_NEED_MAPPING | RTE_PCI_DRV_LSC,
 	},
 	.eth_dev_init = eth_igb_dev_init,
 	.dev_private_size = sizeof(struct e1000_adapter),
--- a/lib/librte_pmd_ixgbe/ixgbe_ethdev.c	2014-06-19 14:56:44.188081500 -0700
+++ b/lib/librte_pmd_ixgbe/ixgbe_ethdev.c	2014-06-19 14:57:31.060327549 -0700
@@ -1049,7 +1049,7 @@ static struct eth_driver rte_ixgbe_pmd =
 	{
 		.name = "rte_ixgbe_pmd",
 		.id_table = pci_id_ixgbe_map,
-		.drv_flags = RTE_PCI_DRV_NEED_MAPPING,
+		.drv_flags = RTE_PCI_DRV_NEED_MAPPING | RTE_PCI_DRV_LSC,
 	},
 	.eth_dev_init = eth_ixgbe_dev_init,
 	.dev_private_size = sizeof(struct ixgbe_adapter),
--- a/lib/librte_ether/rte_ethdev.c	2014-06-19 14:56:44.188081500 -0700
+++ b/lib/librte_ether/rte_ethdev.c	2014-06-19 14:56:44.184081479 -0700
@@ -640,6 +640,20 @@ rte_eth_dev_configure(uint8_t port_id, u
 	memcpy(&dev->data->dev_conf, dev_conf, sizeof(dev->data->dev_conf));
 
 	/*
+	 * If link state interrupt is enabled, check that the
+	 * device supports it.
+	 */
+	if (dev_conf->intr_conf.lsc == 1) {
+		const struct rte_pci_driver *pci_drv = &dev->driver->pci_drv;
+
+		if (!(pci_drv->drv_flags & RTE_PCI_DRV_LSC)) {
+			PMD_DEBUG_TRACE("driver %s does not support lsc\n",
+					pci_drv->name);
+			return (-EINVAL);
+		}
+	}
+
+	/*
 	 * If jumbo frames are enabled, check that the maximum RX packet
 	 * length is supported by the configured device.
 	 */

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [dpdk-dev] [PATCH v2] rte_ethdev: add link support flag
  2014-06-19 22:12 [dpdk-dev] [PATCH v2] rte_ethdev: add link support flag Stephen Hemminger
@ 2014-06-20 13:06 ` Thomas Monjalon
  2014-06-25  9:27   ` Thomas Monjalon
  2014-06-27  0:23   ` Thomas Monjalon
  0 siblings, 2 replies; 4+ messages in thread
From: Thomas Monjalon @ 2014-06-20 13:06 UTC (permalink / raw)
  To: Stephen Hemminger; +Cc: dev

Hi Stephen,

2014-06-19 15:12, Stephen Hemminger:
> Only some devices support the link state interrupt configuration option.
> Link state control does not work in virtual drivers
> (virtio, vmxnet3, igbvf, and ixgbevf). Instead of having the application
> try and guess whether it will work or not provide a driver flag that
> can be checked instead.
> 
> Note: if device driver doesn't support link state control, what
> would happen previously is that the code would never detect link
> transitions. This prevents that.
> 
[...]
> @@ -197,6 +197,8 @@ struct rte_pci_driver {
>  #define RTE_PCI_DRV_MULTIPLE 0x0002
>  /** Device needs to be unbound even if no module is provided */
>  #define RTE_PCI_DRV_FORCE_UNBIND 0x0004
> +/** Device driver supports link state interrupt */
> +#define RTE_PCI_DRV_LSC	0x0008

I feel RTE_PCI_DRV_INTR_LSC would be easier to understand.
Do you agree?

Note that related event is RTE_ETH_EVENT_INTR_LSC
and configuration is intr_conf.lsc.

Title should be "ethdev: add link state interrupt flag".

Thanks
-- 
Thomas

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [dpdk-dev] [PATCH v2] rte_ethdev: add link support flag
  2014-06-20 13:06 ` Thomas Monjalon
@ 2014-06-25  9:27   ` Thomas Monjalon
  2014-06-27  0:23   ` Thomas Monjalon
  1 sibling, 0 replies; 4+ messages in thread
From: Thomas Monjalon @ 2014-06-25  9:27 UTC (permalink / raw)
  To: Stephen Hemminger; +Cc: dev

Hi Stephen,

I had no answer to this comment.
Please check.

2014-06-20 15:06, Thomas Monjalon:
> 2014-06-19 15:12, Stephen Hemminger:
> > Only some devices support the link state interrupt configuration option.
> > Link state control does not work in virtual drivers
> > (virtio, vmxnet3, igbvf, and ixgbevf). Instead of having the application
> > try and guess whether it will work or not provide a driver flag that
> > can be checked instead.
> > 
> > Note: if device driver doesn't support link state control, what
> > would happen previously is that the code would never detect link
> > transitions. This prevents that.
> 
> [...]
> 
> > @@ -197,6 +197,8 @@ struct rte_pci_driver {
> > 
> >  #define RTE_PCI_DRV_MULTIPLE 0x0002
> >  /** Device needs to be unbound even if no module is provided */
> >  #define RTE_PCI_DRV_FORCE_UNBIND 0x0004
> > 
> > +/** Device driver supports link state interrupt */
> > +#define RTE_PCI_DRV_LSC	0x0008
> 
> I feel RTE_PCI_DRV_INTR_LSC would be easier to understand.
> Do you agree?
> 
> Note that related event is RTE_ETH_EVENT_INTR_LSC
> and configuration is intr_conf.lsc.
> 
> Title should be "ethdev: add link state interrupt flag".

A v3 would be nice.

Thanks
-- 
Thomas

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [dpdk-dev] [PATCH v2] rte_ethdev: add link support flag
  2014-06-20 13:06 ` Thomas Monjalon
  2014-06-25  9:27   ` Thomas Monjalon
@ 2014-06-27  0:23   ` Thomas Monjalon
  1 sibling, 0 replies; 4+ messages in thread
From: Thomas Monjalon @ 2014-06-27  0:23 UTC (permalink / raw)
  To: Stephen Hemminger; +Cc: dev

2014-06-20 15:06, Thomas Monjalon:
> 2014-06-19 15:12, Stephen Hemminger:
> > Only some devices support the link state interrupt configuration option.
> > Link state control does not work in virtual drivers
> > (virtio, vmxnet3, igbvf, and ixgbevf). Instead of having the application
> > try and guess whether it will work or not provide a driver flag that
> > can be checked instead.
> > 
> > Note: if device driver doesn't support link state control, what
> > would happen previously is that the code would never detect link
> > transitions. This prevents that.
> 
> [...]
> 
> > @@ -197,6 +197,8 @@ struct rte_pci_driver {
> > 
> >  #define RTE_PCI_DRV_MULTIPLE 0x0002
> >  /** Device needs to be unbound even if no module is provided */
> >  #define RTE_PCI_DRV_FORCE_UNBIND 0x0004
> > 
> > +/** Device driver supports link state interrupt */
> > +#define RTE_PCI_DRV_LSC	0x0008
> 
> I feel RTE_PCI_DRV_INTR_LSC would be easier to understand.
> Do you agree?
> 
> Note that related event is RTE_ETH_EVENT_INTR_LSC
> and configuration is intr_conf.lsc.

Applied with flag renamed.

Thanks
-- 
Thomas

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2014-06-27  0:23 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-06-19 22:12 [dpdk-dev] [PATCH v2] rte_ethdev: add link support flag Stephen Hemminger
2014-06-20 13:06 ` Thomas Monjalon
2014-06-25  9:27   ` Thomas Monjalon
2014-06-27  0:23   ` Thomas Monjalon

DPDK patches and discussions

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://inbox.dpdk.org/dev/0 dev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 dev dev/ https://inbox.dpdk.org/dev \
		dev@dpdk.org
	public-inbox-index dev

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://inbox.dpdk.org/inbox.dpdk.dev


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git