2014-10-02 22:38:23

by Stephen Boyd

Subject: Re: [RFC 5/7] soc: qcom: Add Shared Memory Driver

On 09/29/14 17:34, Bjorn Andersson wrote:
> +
> +#define GET_RX_CHANNEL_INFO(channel, param) \
> + (channel->rx_info_word ? \
> + channel->rx_info_word->param : \
> + channel->rx_info->param)
> +
> +#define GET_TX_CHANNEL_INFO(channel, param) \
> + (channel->rx_info_word ? \
> + channel->tx_info_word->param : \
> + channel->tx_info->param)
> +
> +#define SET_RX_CHANNEL_INFO(channel, param, value) \
> + (channel->rx_info_word ? \
> + (channel->rx_info_word->param = value) : \
> + (channel->rx_info->param = value))
> +
> +#define SET_TX_CHANNEL_INFO(channel, param, value) \
> + (channel->rx_info_word ? \

Drive-by review: Should this be tx_info_word? Given that it works I
wonder why not just have a flag indicating if we should use word aligned
read/write vs. byte aligned.

> + (channel->tx_info_word->param = value) : \
> + (channel->tx_info->param = value))
> +


--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation

2014-10-04 00:02:43

by Bjorn Andersson

Subject: Re: [RFC 5/7] soc: qcom: Add Shared Memory Driver

On Thu 02 Oct 15:38 PDT 2014, Stephen Boyd wrote:

> On 09/29/14 17:34, Bjorn Andersson wrote:
> > +
> > +#define GET_RX_CHANNEL_INFO(channel, param) \
> > + (channel->rx_info_word ? \
> > + channel->rx_info_word->param : \
> > + channel->rx_info->param)
> > +
> > +#define GET_TX_CHANNEL_INFO(channel, param) \
> > + (channel->rx_info_word ? \
> > + channel->tx_info_word->param : \
> > + channel->tx_info->param)
> > +
> > +#define SET_RX_CHANNEL_INFO(channel, param, value) \
> > + (channel->rx_info_word ? \
> > + (channel->rx_info_word->param = value) : \
> > + (channel->rx_info->param = value))
> > +
> > +#define SET_TX_CHANNEL_INFO(channel, param, value) \
> > + (channel->rx_info_word ? \
>
> Drive-by review: Should this be tx_info_word? Given that it works I
> wonder why not just have a flag indicating if we should use word aligned
> read/write vs. byte aligned.
>

You're right, that should be tx - but from the way things both channels will
always be of the same type, so it will simply work.

I had a separate flag, but instead of having 4 members in the struct to
indicate if I was dealing with word aligned access I had 5. So I dropped it.

> > + (channel->tx_info_word->param = value) : \
> > + (channel->tx_info->param = value))
> > +
>

Regards,
Bjorn