[v4,3/3] net/enetc: enable dpaax library
diff mbox series

Message ID 20191021105027.14792-4-g.singh@nxp.com
State Superseded, archived
Delegated to: Ferruh Yigit
Headers show
Series
  • enetc PMD specific changes
Related show

Checks

Context Check Description
ci/checkpatch success coding style OK
ci/Intel-compilation success Compilation OK
ci/travis-robot success Travis build: passed

Commit Message

Gagandeep Singh Oct. 21, 2019, 10:50 a.m. UTC
With this patch, enetc driver would start using the
physical-virtual library interfaces.

Signed-off-by: Gagandeep Singh <g.singh@nxp.com>
---
 drivers/net/enetc/Makefile       | 2 ++
 drivers/net/enetc/enetc_ethdev.c | 8 ++++++++
 drivers/net/enetc/meson.build    | 1 +
 3 files changed, 11 insertions(+)

Comments

Ferruh Yigit Oct. 21, 2019, 3:55 p.m. UTC | #1
On 10/21/2019 11:50 AM, Gagandeep Singh wrote:
> With this patch, enetc driver would start using the
> physical-virtual library interfaces.

Out of curiosity what does it mean "start using the physical-virtual library
interfaces" ?  Can you please provide a little more details how it is done?

> 
> Signed-off-by: Gagandeep Singh <g.singh@nxp.com>
> ---
>  drivers/net/enetc/Makefile       | 2 ++
>  drivers/net/enetc/enetc_ethdev.c | 8 ++++++++
>  drivers/net/enetc/meson.build    | 1 +
>  3 files changed, 11 insertions(+)
> 
> diff --git a/drivers/net/enetc/Makefile b/drivers/net/enetc/Makefile
> index 312b37245..4498bc51f 100644
> --- a/drivers/net/enetc/Makefile
> +++ b/drivers/net/enetc/Makefile
> @@ -10,6 +10,7 @@ LIB = librte_pmd_enetc.a
>  
>  CFLAGS += -O3
>  CFLAGS += $(WERROR_FLAGS)
> +CFLAGS += -I$(RTE_SDK)/drivers/common/dpaax
>  EXPORT_MAP := rte_pmd_enetc_version.map
>  LIBABIVER := 1
>  
> @@ -19,5 +20,6 @@ SRCS-$(CONFIG_RTE_LIBRTE_ENETC_PMD) += enetc_rxtx.c
>  LDLIBS += -lrte_eal -lrte_mbuf -lrte_mempool
>  LDLIBS += -lrte_ethdev -lrte_net
>  LDLIBS += -lrte_bus_pci
> +LDLIBS += -lrte_common_dpaax
>  
>  include $(RTE_SDK)/mk/rte.lib.mk
> diff --git a/drivers/net/enetc/enetc_ethdev.c b/drivers/net/enetc/enetc_ethdev.c
> index 475ec77c3..03590c6c3 100644
> --- a/drivers/net/enetc/enetc_ethdev.c
> +++ b/drivers/net/enetc/enetc_ethdev.c
> @@ -5,6 +5,7 @@
>  #include <stdbool.h>
>  #include <rte_ethdev_pci.h>
>  #include <rte_random.h>
> +#include <dpaax_iova_table.h>
>  
>  #include "enetc_logs.h"
>  #include "enetc.h"
> @@ -896,6 +897,9 @@ enetc_dev_init(struct rte_eth_dev *eth_dev)
>  	eth_dev->data->mtu = RTE_ETHER_MAX_LEN - RTE_ETHER_HDR_LEN -
>  		RTE_ETHER_CRC_LEN;
>  
> +	if (rte_eal_iova_mode() == RTE_IOVA_PA)
> +		dpaax_iova_table_populate();
> +
>  	ENETC_PMD_DEBUG("port_id %d vendorID=0x%x deviceID=0x%x",
>  			eth_dev->data->port_id, pci_dev->id.vendor_id,
>  			pci_dev->id.device_id);
> @@ -906,6 +910,10 @@ static int
>  enetc_dev_uninit(struct rte_eth_dev *eth_dev __rte_unused)
>  {
>  	PMD_INIT_FUNC_TRACE();
> +
> +	if (rte_eal_iova_mode() == RTE_IOVA_PA)
> +		dpaax_iova_table_depopulate();
> +
>  	return 0;
>  }
>  
> diff --git a/drivers/net/enetc/meson.build b/drivers/net/enetc/meson.build
> index 3bc069844..bea54bea8 100644
> --- a/drivers/net/enetc/meson.build
> +++ b/drivers/net/enetc/meson.build
> @@ -6,6 +6,7 @@ if not is_linux
>  	reason = 'only supported on linux'
>  endif
>  
> +deps += ['common_dpaax']
>  sources = files('enetc_ethdev.c',
>  		'enetc_rxtx.c')
>  
>
Gagandeep Singh Oct. 22, 2019, 5:31 a.m. UTC | #2
> -----Original Message-----
> From: Ferruh Yigit <ferruh.yigit@intel.com>
> Sent: Monday, October 21, 2019 9:26 PM
> To: Gagandeep Singh <G.Singh@nxp.com>; dev@dpdk.org
> Subject: Re: [PATCH v4 3/3] net/enetc: enable dpaax library
> 
> On 10/21/2019 11:50 AM, Gagandeep Singh wrote:
> > With this patch, enetc driver would start using the
> > physical-virtual library interfaces.
> 
> Out of curiosity what does it mean "start using the physical-virtual library
> interfaces" ?  Can you please provide a little more details how it is done?
> 
I was just trying to explain that from now onwards, enetc driver 
will start interact with DPAAX library for PA to VA translation in
case physical mode is enabled.
I will update the commit message with better explanation that how this
Change can be useful.
Ferruh Yigit Oct. 22, 2019, 9 a.m. UTC | #3
On 10/22/2019 6:31 AM, Gagandeep Singh wrote:
> 
> 
>> -----Original Message-----
>> From: Ferruh Yigit <ferruh.yigit@intel.com>
>> Sent: Monday, October 21, 2019 9:26 PM
>> To: Gagandeep Singh <G.Singh@nxp.com>; dev@dpdk.org
>> Subject: Re: [PATCH v4 3/3] net/enetc: enable dpaax library
>>
>> On 10/21/2019 11:50 AM, Gagandeep Singh wrote:
>>> With this patch, enetc driver would start using the
>>> physical-virtual library interfaces.
>>
>> Out of curiosity what does it mean "start using the physical-virtual library
>> interfaces" ?  Can you please provide a little more details how it is done?
>>
> I was just trying to explain that from now onwards, enetc driver 
> will start interact with DPAAX library for PA to VA translation in
> case physical mode is enabled.

If it will be used later, what about adding this change in the patch it needs to
be used?

> I will update the commit message with better explanation that how this
> Change can be useful.
>
Gagandeep Singh Oct. 22, 2019, 11:30 a.m. UTC | #4
> -----Original Message-----
> From: Ferruh Yigit <ferruh.yigit@intel.com>
> Sent: Tuesday, October 22, 2019 2:31 PM
> To: Gagandeep Singh <G.Singh@nxp.com>; dev@dpdk.org
> Subject: Re: [PATCH v4 3/3] net/enetc: enable dpaax library
> 
> On 10/22/2019 6:31 AM, Gagandeep Singh wrote:
> >
> >
> >> -----Original Message-----
> >> From: Ferruh Yigit <ferruh.yigit@intel.com>
> >> Sent: Monday, October 21, 2019 9:26 PM
> >> To: Gagandeep Singh <G.Singh@nxp.com>; dev@dpdk.org
> >> Subject: Re: [PATCH v4 3/3] net/enetc: enable dpaax library
> >>
> >> On 10/21/2019 11:50 AM, Gagandeep Singh wrote:
> >>> With this patch, enetc driver would start using the
> >>> physical-virtual library interfaces.
> >>
> >> Out of curiosity what does it mean "start using the physical-virtual library
> >> interfaces" ?  Can you please provide a little more details how it is done?
> >>
> > I was just trying to explain that from now onwards, enetc driver
> > will start interact with DPAAX library for PA to VA translation in
> > case physical mode is enabled.
> 
> If it will be used later, what about adding this change in the patch it needs to
> be used?

This is right that currently enetc is not using the dpaaX table for PA->VA translation,
But with this change, caam JR driver can get benefits of dpaaX, when used with
enetc over enetc supported SoCs( e.g. in IPsec-secgw application)

Here is caam ptov code:

static inline void *
caam_jr_dma_ptov(rte_iova_t paddr)
{
	void *va;
	va = dpaax_iova_table_get_va(paddr);
	if (likely(va != NULL))
		return va;

	return rte_mem_iova2virt(paddr);
}

Currently, caam JR using the expensive API rte_mem_iova2virt in its datapath, and we can avoid this by enabling the dpaaX.

Patch
diff mbox series

diff --git a/drivers/net/enetc/Makefile b/drivers/net/enetc/Makefile
index 312b37245..4498bc51f 100644
--- a/drivers/net/enetc/Makefile
+++ b/drivers/net/enetc/Makefile
@@ -10,6 +10,7 @@  LIB = librte_pmd_enetc.a
 
 CFLAGS += -O3
 CFLAGS += $(WERROR_FLAGS)
+CFLAGS += -I$(RTE_SDK)/drivers/common/dpaax
 EXPORT_MAP := rte_pmd_enetc_version.map
 LIBABIVER := 1
 
@@ -19,5 +20,6 @@  SRCS-$(CONFIG_RTE_LIBRTE_ENETC_PMD) += enetc_rxtx.c
 LDLIBS += -lrte_eal -lrte_mbuf -lrte_mempool
 LDLIBS += -lrte_ethdev -lrte_net
 LDLIBS += -lrte_bus_pci
+LDLIBS += -lrte_common_dpaax
 
 include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/drivers/net/enetc/enetc_ethdev.c b/drivers/net/enetc/enetc_ethdev.c
index 475ec77c3..03590c6c3 100644
--- a/drivers/net/enetc/enetc_ethdev.c
+++ b/drivers/net/enetc/enetc_ethdev.c
@@ -5,6 +5,7 @@ 
 #include <stdbool.h>
 #include <rte_ethdev_pci.h>
 #include <rte_random.h>
+#include <dpaax_iova_table.h>
 
 #include "enetc_logs.h"
 #include "enetc.h"
@@ -896,6 +897,9 @@  enetc_dev_init(struct rte_eth_dev *eth_dev)
 	eth_dev->data->mtu = RTE_ETHER_MAX_LEN - RTE_ETHER_HDR_LEN -
 		RTE_ETHER_CRC_LEN;
 
+	if (rte_eal_iova_mode() == RTE_IOVA_PA)
+		dpaax_iova_table_populate();
+
 	ENETC_PMD_DEBUG("port_id %d vendorID=0x%x deviceID=0x%x",
 			eth_dev->data->port_id, pci_dev->id.vendor_id,
 			pci_dev->id.device_id);
@@ -906,6 +910,10 @@  static int
 enetc_dev_uninit(struct rte_eth_dev *eth_dev __rte_unused)
 {
 	PMD_INIT_FUNC_TRACE();
+
+	if (rte_eal_iova_mode() == RTE_IOVA_PA)
+		dpaax_iova_table_depopulate();
+
 	return 0;
 }
 
diff --git a/drivers/net/enetc/meson.build b/drivers/net/enetc/meson.build
index 3bc069844..bea54bea8 100644
--- a/drivers/net/enetc/meson.build
+++ b/drivers/net/enetc/meson.build
@@ -6,6 +6,7 @@  if not is_linux
 	reason = 'only supported on linux'
 endif
 
+deps += ['common_dpaax']
 sources = files('enetc_ethdev.c',
 		'enetc_rxtx.c')