On 14/03/12 12:33, Joe Perches wrote:

> On Tue, 2012-03-13 at 20:58 -0400, Andrew Miller wrote:
>> Fix long line coding style issue
>> Signed-off-by: Andrew Miller <[email protected]>
> 
> Hi Andrew.
> 
> Please strive for clarity instead of just fixing
> random 80 char warnings.
> 
>> diff --git a/drivers/staging/rtl8187se/r8180_core.c 
>> b/drivers/staging/rtl8187se/r8180_core.c
> []
>> @@ -1607,17 +1609,20 @@ void rtl8180_rx(struct net_device *dev)
>>              /* printk("==========================>rx : RXAGC is 
>> %d,signalstrength is %d\n",RXAGC,stats.signalstrength); */
>>              stats.rssi = priv->wstats.qual.qual = priv->SignalQuality;
>>              stats.noise = priv->wstats.qual.noise = 100 - 
>> priv->wstats.qual.qual;
>> -            bHwError = (((*(priv->rxringtail)) & (0x00000fff)) == 4080) | 
>> (((*(priv->rxringtail)) & (0x04000000)) != 0)
>> -                    | (((*(priv->rxringtail)) & (0x08000000)) != 0) | 
>> (((~(*(priv->rxringtail))) & (0x10000000)) != 0) | 
>> (((~(*(priv->rxringtail))) & (0x20000000)) != 0);
>> +            bHwError = (((*(priv->rxringtail)) & (0x00000fff)) == 4080) |
>> +                       (((*(priv->rxringtail)) & (0x04000000)) != 0) |
>> +                       (((*(priv->rxringtail)) & (0x08000000)) != 0) |
>> +                       (((~(*(priv->rxringtail))) & (0x10000000)) != 0) |
>> +                       (((~(*(priv->rxringtail))) & (0x20000000)) != 0);
> 
> Likely these | uses should be ||
> 
>>              bCRC = ((*(priv->rxringtail)) & (0x00002000)) >> 13;
>>              bICV = ((*(priv->rxringtail)) & (0x00001000)) >> 12;
>>              hdr = (struct ieee80211_hdr_4addr *)priv->rxbuffer->buf;
>>                  fc = le16_to_cpu(hdr->frame_ctl);
>>              type = WLAN_FC_GET_TYPE(fc);
>>  
>> -                    if ((IEEE80211_FTYPE_CTL != type) &&
>> -                            
>> (eqMacAddr(priv->ieee80211->current_network.bssid, (fc & 
>> IEEE80211_FCTL_TODS) ? hdr->addr1 : (fc & IEEE80211_FCTL_FROMDS) ? 
>> hdr->addr2 : hdr->addr3))
>> -                             && (!bHwError) && (!bCRC) && (!bICV)) {
>> +                    if ((IEEE80211_FTYPE_CTL != type) && 
>> (eqMacAddr(priv->ieee80211->current_network.bssid,
>> +                        (fc & IEEE80211_FCTL_TODS) ? hdr->addr1 : (fc & 
>> IEEE80211_FCTL_FROMDS) ? hdr->addr2 : hdr->addr3))
>> +                        && (!bHwError) && (!bCRC) && (!bICV)) {
> 
> This is difficult to read and could be better written
> removing unnecessary parentheses and making the egMacAddr
> clearer as a function/macro call:
> 
>                       if (IEEE80211_FTYPE_CTL != type &&
>                           eqMacAddr(priv->ieee80211->current_network.bssid,
>                                     fc & IEEE80211_FCTL_TODS ? hdr->addr1 :
>                                     fc & IEEE80211_FCTL_FROMDS ? hdr->addr2 :
>                                     hdr->addr3) &&
>                           !bHwError &&
>                           !bCRC &&
>                           !bICV)


That's still pretty nasty. Eight lines for an if expression! It took me
a couple of glances to realise that there were only two arguments to
eqMacAddr. If you add some temporary variables then you can make the
code a lot more sane, without making it overly verbose:

        u8 *addr, *bssid = priv->ieee80211->current_network.bssid;

        if (fc & IEEE80211_FCTL_TODS)
                addr = hdr->addr1;
        else if (fc & IEEE80211_FCTL_FROMDS)
                addr = hdr->addr2;
        else
                addr = hdr->addr3;

        if (!bHwError && !bCRC && !bICV &&
            type != IEEE80211_FTYPE_CTL && eqMacAddr(bssid, addr)) {
                ...

Also note that the whole if block is indented one too many tab stops.
Would be best to fix that in a separate patch though. If you fix that
first, then you will have more horizontal space to re-organise the
expression inside the if :-).

~Ryan

_______________________________________________
devel mailing list
[email protected]
http://driverdev.linuxdriverproject.org/mailman/listinfo/devel

Reply via email to