[PATCH infiniband-diags] Add Connect-X5 device ID to device white lists

[PATCH infiniband-diags] Add Connect-X5 device ID to device white lists

[Hal Rosenstock]

Connect-X5 is device ID 0x1017

Signed-off-by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
 libibnetdisc/src/ibnetdisc.c | 2 +-
 src/ibdiag_common.c          | 2 +-
 src/vendstat.c               | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/libibnetdisc/src/ibnetdisc.c b/libibnetdisc/src/ibnetdisc.c
index 7ba5c91..57fc33e 100644
--- a/libibnetdisc/src/ibnetdisc.c
+++ b/libibnetdisc/src/ibnetdisc.c
@@ -199,7 +199,7 @@ static int is_mlnx_ext_port_info_supported(ibnd_port_t * port)
 
 	if ((devid >= 0xc738 && devid <= 0xc73b) || devid == 0xcb20 || devid == 0xcf08 || devid == 0x1b02)
 		return 1;
-	if (devid >= 0x1003 && devid <= 0x1016)
+	if (devid >= 0x1003 && devid <= 0x1017)
 		return 1;
 	return 0;
 }
diff --git a/src/ibdiag_common.c b/src/ibdiag_common.c
index f74eed4..981f82d 100644
--- a/src/ibdiag_common.c
+++ b/src/ibdiag_common.c
@@ -532,7 +532,7 @@ int is_mlnx_ext_port_info_supported(uint32_t devid)
 	if (ibd_ibnetdisc_flags & IBND_CONFIG_MLX_EPI) {
 		if ((devid >= 0xc738 && devid <= 0xc73b) || devid == 0xcb20 || devid == 0xcf08 || devid == 0x1b02)
 			return 1;
-		if (devid >= 0x1003 && devid <= 0x1016)
+		if (devid >= 0x1003 && devid <= 0x1017)
 			return 1;
 	}
 	return 0;
diff --git a/src/vendstat.c b/src/vendstat.c
index 284ef93..7f88bf3 100644
--- a/src/vendstat.c
+++ b/src/vendstat.c
@@ -148,7 +148,7 @@ static uint16_t ext_fw_info_device[][2] = {
 	{0xcb20, 0xcb20},	/* Switch-IB */
 	{0xcf08, 0xcf08},	/* Switch-IB2 */
 	{0x01b3, 0x01b3},	/* IS-4 */
-	{0x1003, 0x1016},	/* Connect-X */
+	{0x1003, 0x1017},	/* Connect-X */
 	{0x1b02, 0x1b02},	/* Bull */
 	{0x0000, 0x0000}};
 
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH infiniband-diags] Add Connect-X5 device ID to device white lists

[Or Gerlitz]

What is the semantics of the white-list?

What happens if you see a device which is not there?

Why do we need this list?

Or.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH infiniband-diags] Add Connect-X5 device ID to device white lists

[Hal Rosenstock]

On 12/20/2016 2:25 AM, Or Gerlitz wrote:
> What is the semantics of the white-list?

The white lists control the issuing/checking of several Mellanox vendor
specific MADs such as SM class MLNX ExtendedPortInfo (MEPI) which
handles FDR10 and vendor class 0x0A GeneralInfo interpretation as to
whether extended version fields are supported.

> What happens if you see a device which is not there?

It will affect FDR10 reporting or version reporting done via Mellanox
vendor specific MADs.

Do you see any that you think are missing ?

> Why do we need this list?

To know whether MEPI can be issued or not without timeout and to
properly interpret the version fields in the GeneralInfo response.

-- Hal

> Or.
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH infiniband-diags] Add Connect-X5 device ID to device white lists

[ira.weiny]

On Mon, Dec 19, 2016 at 04:24:34PM -0500, Hal Rosenstock wrote:
> 
> Connect-X5 is device ID 0x1017
> 
> Signed-off-by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>

After the patch from Vincent I had to apply this by hand.

So basically applied.

But after I push all this could you double check that this is ok?

Thanks,
Ira


> ---
>  libibnetdisc/src/ibnetdisc.c | 2 +-
>  src/ibdiag_common.c          | 2 +-
>  src/vendstat.c               | 2 +-
>  3 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/libibnetdisc/src/ibnetdisc.c b/libibnetdisc/src/ibnetdisc.c
> index 7ba5c91..57fc33e 100644
> --- a/libibnetdisc/src/ibnetdisc.c
> +++ b/libibnetdisc/src/ibnetdisc.c
> @@ -199,7 +199,7 @@ static int is_mlnx_ext_port_info_supported(ibnd_port_t * port)
>  
>  	if ((devid >= 0xc738 && devid <= 0xc73b) || devid == 0xcb20 || devid == 0xcf08 || devid == 0x1b02)
>  		return 1;
> -	if (devid >= 0x1003 && devid <= 0x1016)
> +	if (devid >= 0x1003 && devid <= 0x1017)
>  		return 1;
>  	return 0;
>  }
> diff --git a/src/ibdiag_common.c b/src/ibdiag_common.c
> index f74eed4..981f82d 100644
> --- a/src/ibdiag_common.c
> +++ b/src/ibdiag_common.c
> @@ -532,7 +532,7 @@ int is_mlnx_ext_port_info_supported(uint32_t devid)
>  	if (ibd_ibnetdisc_flags & IBND_CONFIG_MLX_EPI) {
>  		if ((devid >= 0xc738 && devid <= 0xc73b) || devid == 0xcb20 || devid == 0xcf08 || devid == 0x1b02)
>  			return 1;
> -		if (devid >= 0x1003 && devid <= 0x1016)
> +		if (devid >= 0x1003 && devid <= 0x1017)
>  			return 1;
>  	}
>  	return 0;
> diff --git a/src/vendstat.c b/src/vendstat.c
> index 284ef93..7f88bf3 100644
> --- a/src/vendstat.c
> +++ b/src/vendstat.c
> @@ -148,7 +148,7 @@ static uint16_t ext_fw_info_device[][2] = {
>  	{0xcb20, 0xcb20},	/* Switch-IB */
>  	{0xcf08, 0xcf08},	/* Switch-IB2 */
>  	{0x01b3, 0x01b3},	/* IS-4 */
> -	{0x1003, 0x1016},	/* Connect-X */
> +	{0x1003, 0x1017},	/* Connect-X */
>  	{0x1b02, 0x1b02},	/* Bull */
>  	{0x0000, 0x0000}};
>  
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH infiniband-diags] Add Connect-X5 device ID to device white lists

[Hal Rosenstock]

On 12/20/2016 11:57 AM, ira.weiny wrote:
> On Mon, Dec 19, 2016 at 04:24:34PM -0500, Hal Rosenstock wrote:
>>
>> Connect-X5 is device ID 0x1017
>>
>> Signed-off-by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
> 
> After the patch from Vincent I had to apply this by hand.
> 
> So basically applied.
> 
> But after I push all this could you double check that this is ok?

It looks fine to me.

Thanks!

-- Hal

> Thanks,
> Ira
> 
> 
>> ---
>>  libibnetdisc/src/ibnetdisc.c | 2 +-
>>  src/ibdiag_common.c          | 2 +-
>>  src/vendstat.c               | 2 +-
>>  3 files changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/libibnetdisc/src/ibnetdisc.c b/libibnetdisc/src/ibnetdisc.c
>> index 7ba5c91..57fc33e 100644
>> --- a/libibnetdisc/src/ibnetdisc.c
>> +++ b/libibnetdisc/src/ibnetdisc.c
>> @@ -199,7 +199,7 @@ static int is_mlnx_ext_port_info_supported(ibnd_port_t * port)
>>  
>>  	if ((devid >= 0xc738 && devid <= 0xc73b) || devid == 0xcb20 || devid == 0xcf08 || devid == 0x1b02)
>>  		return 1;
>> -	if (devid >= 0x1003 && devid <= 0x1016)
>> +	if (devid >= 0x1003 && devid <= 0x1017)
>>  		return 1;
>>  	return 0;
>>  }
>> diff --git a/src/ibdiag_common.c b/src/ibdiag_common.c
>> index f74eed4..981f82d 100644
>> --- a/src/ibdiag_common.c
>> +++ b/src/ibdiag_common.c
>> @@ -532,7 +532,7 @@ int is_mlnx_ext_port_info_supported(uint32_t devid)
>>  	if (ibd_ibnetdisc_flags & IBND_CONFIG_MLX_EPI) {
>>  		if ((devid >= 0xc738 && devid <= 0xc73b) || devid == 0xcb20 || devid == 0xcf08 || devid == 0x1b02)
>>  			return 1;
>> -		if (devid >= 0x1003 && devid <= 0x1016)
>> +		if (devid >= 0x1003 && devid <= 0x1017)
>>  			return 1;
>>  	}
>>  	return 0;
>> diff --git a/src/vendstat.c b/src/vendstat.c
>> index 284ef93..7f88bf3 100644
>> --- a/src/vendstat.c
>> +++ b/src/vendstat.c
>> @@ -148,7 +148,7 @@ static uint16_t ext_fw_info_device[][2] = {
>>  	{0xcb20, 0xcb20},	/* Switch-IB */
>>  	{0xcf08, 0xcf08},	/* Switch-IB2 */
>>  	{0x01b3, 0x01b3},	/* IS-4 */
>> -	{0x1003, 0x1016},	/* Connect-X */
>> +	{0x1003, 0x1017},	/* Connect-X */
>>  	{0x1b02, 0x1b02},	/* Bull */
>>  	{0x0000, 0x0000}};
>>  
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH infiniband-diags] Add Connect-X5 device ID to device white lists

[Or Gerlitz]

On Tue, Dec 20, 2016 at 6:40 PM, Hal Rosenstock <hal-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org> wrote:
> On 12/20/2016 2:25 AM, Or Gerlitz wrote:
>> What is the semantics of the white-list?
>
> The white lists control the issuing/checking of several Mellanox vendor
> specific MADs such as SM class MLNX ExtendedPortInfo (MEPI) which
> handles FDR10 and vendor class 0x0A GeneralInfo interpretation as to
> whether extended version fields are supported.
>
>> What happens if you see a device which is not there?
>
> It will affect FDR10 reporting or version reporting done via Mellanox
> vendor specific MADs.
>
> Do you see any that you think are missing ?

no, thanks for explaining
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html