[dpdk-dev,v2] examples/vhost: reduce number of hugepages needed

  Change MAX_QUEUES from 512 to 128 to reduce the number of hugepages
required by the vhost-switch program.

Changes in v2:
remove comment added before #define MAX_QUEUES in v1 patch.

Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
---
 examples/vhost/main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
  

> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Bernard Iremonger
> Sent: Thursday, December 10, 2015 1:53 PM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v2] examples/vhost: reduce number of hugepages needed
> 
> Change MAX_QUEUES from 512 to 128 to reduce the number of hugepages
> required by the vhost-switch program.
> 
> Changes in v2:
> remove comment added before #define MAX_QUEUES in v1 patch.
> 
> Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
> Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>


Wasn't it increased a while ago, because someone complained that
128 queues might not be enough on FVL?
From git log I can see that it was first increased from 128 to 256,
then from 256 to 512.
The reason mentioned - HW that has bigger number of queues. 
Isn't it not the case anymore?
If yes, why?

BTW, shouldn't it be then at least:

+#ifndef MAX_QUEUES
+#define MAX_QUEUES 128
+#endif

So people can just do -D MAX_QUEUES=X at build time if they like(need) to.
Konstantin

> ---
>  examples/vhost/main.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/examples/vhost/main.c b/examples/vhost/main.c
> index b30f1bd..5ce5926 100644
> --- a/examples/vhost/main.c
> +++ b/examples/vhost/main.c
> @@ -1,7 +1,7 @@
>  /*-
>   *   BSD LICENSE
>   *
> - *   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
> + *   Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
>   *   All rights reserved.
>   *
>   *   Redistribution and use in source and binary forms, with or without
> @@ -53,7 +53,7 @@
> 
>  #include "main.h"
> 
> -#define MAX_QUEUES 512
> +#define MAX_QUEUES 128
> 
>  /* the maximum number of external ports supported */
>  #define MAX_SUP_PORTS 1
> --
> 2.6.3
  

Hi Konstantin,

<snip>

> > Subject: [dpdk-dev] [PATCH v2] examples/vhost: reduce number of
> > hugepages needed
> >
> > Change MAX_QUEUES from 512 to 128 to reduce the number of hugepages
> > required by the vhost-switch program.
> >
> > Changes in v2:
> > remove comment added before #define MAX_QUEUES in v1 patch.
> >
> > Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
> > Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
> 
> 
> Wasn't it increased a while ago, because someone complained that
> 128 queues might not be enough on FVL?
> From git log I can see that it was first increased from 128 to 256, then from
> 256 to 512.
> The reason mentioned - HW that has bigger number of queues.
> Isn't it not the case anymore?
> If yes, why?
> 
> BTW, shouldn't it be then at least:
> 
> +#ifndef MAX_QUEUES
> +#define MAX_QUEUES 128
> +#endif
> 
> So people can just do -D MAX_QUEUES=X at build time if they like(need) to.
> Konstantin
> 
<snip>

I will add #ifndef MAX_QUEUES and #endif.

Regards,

Bernard.
  

On 12/10/2015 04:50 PM, Ananyev, Konstantin wrote:
>
>
>> -----Original Message-----
>> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Bernard Iremonger
>> Sent: Thursday, December 10, 2015 1:53 PM
>> To: dev@dpdk.org
>> Subject: [dpdk-dev] [PATCH v2] examples/vhost: reduce number of hugepages needed
>>
>> Change MAX_QUEUES from 512 to 128 to reduce the number of hugepages
>> required by the vhost-switch program.
>>
>> Changes in v2:
>> remove comment added before #define MAX_QUEUES in v1 patch.
>>
>> Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
>> Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
>
>
> Wasn't it increased a while ago, because someone complained that
> 128 queues might not be enough on FVL?
>  From git log I can see that it was first increased from 128 to 256,
> then from 256 to 512.
> The reason mentioned - HW that has bigger number of queues.
> Isn't it not the case anymore?
> If yes, why?
>
> BTW, shouldn't it be then at least:
>
> +#ifndef MAX_QUEUES
> +#define MAX_QUEUES 128
> +#endif
>
> So people can just do -D MAX_QUEUES=X at build time if they like(need) to.

Being subject to constant changes back and forth suggests this really 
should be a runtime tunable rather than build time constant.

	- Panu -