[dpdk-dev] [PATCH v2 04/13] ethdev: support of multiple sizes of redirection table

[Thomas Monjalon]

2014-10-29 08:24, Zhang, Helin:
> 
> > -----Original Message-----
> > From: Zhang, Helin
> > Sent: Tuesday, October 28, 2014 8:01 PM
> > To: Thomas Monjalon
> > Cc: dev at dpdk.org
> > Subject: RE: [dpdk-dev] [PATCH v2 04/13] ethdev: support of multiple sizes of
> > redirection table
> > 
> > Hi Thomas
> > 
> > > -----Original Message-----
> > > From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> > > Sent: Tuesday, October 28, 2014 6:10 PM
> > > To: Zhang, Helin
> > > Cc: dev at dpdk.org
> > > Subject: Re: [dpdk-dev] [PATCH v2 04/13] ethdev: support of multiple
> > > sizes of redirection table
> > >
> > > 2014-10-28 00:33, Zhang, Helin:
> > > > From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> > > > > 2014-09-25 16:40, Helin Zhang:
> > > > > >  /* Definitions used for redirection table entry size */
> > > > > > -#define ETH_RSS_RETA_NUM_ENTRIES 128
> > > > > > -#define ETH_RSS_RETA_MAX_QUEUE   16
> > > > > > +#define ETH_RSS_RETA_SIZE_64  64 #define ETH_RSS_RETA_SIZE_128
> > > > > > +128 #define ETH_RSS_RETA_SIZE_512
> > > > > > +512
> > > > > > +
> > > > > > +#define RTE_BIT_WIDTH_64 (CHAR_BIT * sizeof(uint64_t))
> > > > >
> > > > > Are these constants really needed?
> > > >
> > > > These constants were defined for the third input parameter of
> > > > rte_eth_dev_rss_reta_update() and rte_eth_dev_rss_reta_query(). End
> > > > users need to give the correct reta size listed as above, as other
> > > > values is not valid. So it would be better to list the valid reta
> > > > sizes in macros
> > > here.
> > >
> > > OK, so you should explain that only these values are allowed.
> > > In general, it's something we explain in the comment of the function
> > It would be better to add comments for the functions.
> Now do not think explain what value is allowed for the functions in ethdev layer,
> as it might be hardware specific.
> I think the best way is to comment out end users can get the reta size by
> 'dev_infos_get' on each port, rather than telling the values directly.

Yes, it's better.

Thanks
-- 
Thomas