Hello Felix,
Dnia 2012-03-05, pon o godzinie 21:36 +0100, Felix Fietkau pisze:
> On 2012-03-05 7:20 PM, Łukasz Kwestarz wrote:
> > Patch probably adds also support for 802.11a mode (not tested), fixes some
> > issues in iwinfo lib and not working WIFI activity led on WRT150N v1.1.
> > Please review it carefully, it was my first time with ash and I haven't
> > been using C for a long time.
> >
> > Patch for LuCI will follow.
> >
> > Signed-off-by: Łukasz Kwestarz <kwest...@me.com>
> >
> > Index: package/iwinfo/src/iwinfo_wl.c
> > ===================================================================
> > --- package/iwinfo/src/iwinfo_wl.c (wersja 30798)
> > +++ package/iwinfo/src/iwinfo_wl.c (kopia robocza)
> > @@ -571,7 +571,52 @@
> >
> > int wl_get_hwmodelist(const char *ifname, int *buf)
> > {
> > - return wext_get_hwmodelist(ifname, buf);
> > + int phytype;
> > + if (!wl_ioctl(ifname, WLC_GET_PHYTYPE, &phytype, sizeof(phytype)))
> > + {
> > + if(phytype == PHY_TYPE_N)
> > + {
> > + *buf |= IWINFO_80211_N;
> > + int bandlist[WLC_BAND_ALL];
> > + if (!wl_ioctl(ifname, WLC_GET_BANDLIST, bandlist,
> > sizeof(bandlist)))
> > + {
> > + for (int i = 1; i <= bandlist[0]; i++)
> > + {
> > + if(bandlist[i] == WLC_BAND_5G)
> > + {
> > + *buf |= IWINFO_80211_A;
> > + }
> > + else
> > + {
> > + *buf |= IWINFO_80211_B;
> > + *buf |= IWINFO_80211_G;
> > + }
> > + }
> > + return 0;
> > + }
> > + }
> > + else
> > + {
> > + char phylist[4];
> > + if (!wl_ioctl(ifname, WLC_GET_PHYLIST, phylist,
> > sizeof(phylist)))
> > + {
> > + for(int i = 0; phylist[i]; i++)
> > + {
> > + if(phylist[i] == 'a')
> > + {
> > + *buf |= IWINFO_80211_A;
> > + }
> > + else
> > + {
> > + *buf |= IWINFO_80211_B;
> > + *buf |= IWINFO_80211_G;
> > + }
> > + }
> > + return 0;
> > + }
> > + }
> > + }
> > + return -1;
> The indentation level there is a bit excessive, could you restructure
> the code a bit to make it more readable?
> Replace things like
> if (!wl_ioctl(ifname, WLC_GET_PHYTYPE, &phytype, sizeof(phytype))) {
> <code...>
> }
>
> with:
> if (wl_ioctl(ifname, WLC_GET_PHYTYPE, &phytype, sizeof(phytype)))
> return -1;
> <code...>
Sure can change that.
>
> > Index: package/broadcom-wl/files/lib/wifi/broadcom.sh
> > ===================================================================
> > --- package/broadcom-wl/files/lib/wifi/broadcom.sh (wersja 30798)
> > +++ package/broadcom-wl/files/lib/wifi/broadcom.sh (kopia robocza)
> > @@ -130,11 +130,19 @@
> > config_get txpower "$device" txpower
> > config_get frag "$device" frag
> > config_get rts "$device" rts
> > + config_get band "$device" band
> > config_get hwmode "$device" hwmode
> > local vif_pre_up vif_post_up vif_do_up vif_txpower
> > local doth=0
> > local wmm=1
> > -
> > + local gmode=1
> > + local nmode=0
> > + local nreqd=0
> > + local leddc=$(($(wlc leddc)))
> > + if [ "$leddc" -eq 65535 ]; then
> > + leddc=6553600;
> > + fi
> Are you sure this is generic enough to be applied to all devices?
>
If you asking about led, yes I'm pretty sure, saw similar fixes in
dd-wrt and tomato. No big risk it affects only led :)
> > @@ -161,13 +169,25 @@
> > ;;
> > esac
> >
> > + case "$band" in
> > + a) band=1;;
> > + b) band=2;;
> > + esac
> > +
> > case "$hwmode" in
> > - *b) hwmode=0;;
> > - *bg) hwmode=1;;
> > - *g) hwmode=2;;
> > - *gst) hwmode=4;;
> > - *lrs) hwmode=5;;
> > - *) hwmode=1;;
> > + *a) ;;
> > + *b) gmode=0;;
> > + *bg) ;;
> > + *g) gmode=2;;
> > + *gst) gmode=4;;
> > + *lrs) gmode=5;;
> > + *n)
> > + nreqd=1
> > + nmode=1
> > + ;;
> > + *)
> > + nmode=-1
> > + ;;
> > esac
> >
> > for vif in $vifs; do
> What's lrs? And why gst with turbo - what's that for?
> And why no support for 11ng/11na similar to how it's done in mac80211?
>
Those two settings were here before, so I didn't paid much attention to
them, but you can find them in wl gmode man, also OpenWRT wiki contains
few words about. 11ng/11na is confusing for me, there is no something
like 11ng/11na outside OpenWrt world (or I'm not aware of :)), there is
11n on 5 ghz band and 11n on 2.4 ghz band so in my opinion band
selection is reasonable and means much more for end user.
> > @@ -322,7 +342,11 @@
> > wlc ifname "$device" stdin <<EOF
> > $ifdown
> >
> > -gmode ${hwmode:-1}
> > +leddc ${leddc}
> > +band ${band:-0}
> > +gmode ${gmode}
> > +nmode ${nmode}
> > +nreqd ${nreqd}
> > apsta $apsta
> > ap $ap
> > ${mssid:+mssid $mssid}
>
> > @@ -386,6 +410,7 @@
> > option mode ap
> > option ssid OpenWrt${i#0}
> > option encryption none
> > + option wmm 1
> Not necessary, the wmm default was already changed and normally you
> should not have to turn it off anyway.
>
Correct but I think it should be marked as enabled in LuCI or other UI
interfaces.
> - Felix
Thanks for quick response and comments:)
Łukasz Kwestarz
_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel
