Re: [BUG REPORT] brcmfmac: wrong logic in finding control/primary channel relative location

[Arend van Spriel <arend@broadcom.com>]

On 12/20/14 19:25, Rafał Miłecki wrote:
> ---
> Hi guys,
>
> Let me report something I think is a bug in this unusual-patch way ;)
> ---
>   drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c | 2 ++
>   1 file changed, 2 insertions(+)
>
> diff --git a/drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c b/drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c
> index 39b45c0..2c9e88b 100644
> --- a/drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c
> +++ b/drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c
> @@ -313,11 +313,13 @@ static u16 chandef_to_chanspec(struct brcmu_d11inf *d11inf,
>   	case NL80211_CHAN_WIDTH_80:
>   		ch_inf.bw = BRCMU_CHAN_BW_80;
>   		if (primary_offset<  0) {
> +			/* FIXME: We compare frequencies offset with a value -2. Looks like a wrong logic. */
>   			if (primary_offset<  -CH_10MHZ_APART)

You are right. The primary_offset is in MHz, but the logical expression 
is for channel spacing (1 channel == 5MHz). So here it should be -10 
instead. Thanks for reporting this.

Regards,
Arend
>   				ch_inf.sb = BRCMU_CHAN_SB_UU;
>   			else
>   				ch_inf.sb = BRCMU_CHAN_SB_UL;
>   		} else {
> +			/* FIXME: We compare frequencies offset with a value 2. Looks like a wrong logic. */
>   			if (primary_offset>  CH_10MHZ_APART)
>   				ch_inf.sb = BRCMU_CHAN_SB_LL;
>   			else