On Sun, Jan 22, 2012 at 8:40 PM, Roman Yeryomin <leroi.li...@gmail.com> wrote: > On 22 January 2012 14:13, Marco Antonio Mauro <marcu...@gmail.com> wrote: >> On Sat, Jan 21, 2012 at 6:36 PM, Roman Yeryomin <leroi.li...@gmail.com> >> wrote: >>> On 21 January 2012 00:53, Marco Antonio Mauro <marcu...@gmail.com> wrote: >>>> This patch adds support for the Sitecom WL-341 v3 and other Sercomm >>>> IP1006RRv2 based boards for sysupgrade support and for the initial >>>> flash through pushbutton initiated recovery mode with the special >>>> partition table and fixes for the quirks and things required by the >>>> modified bootloader. >>>> >>>> There are two known bugs: >>>> >>>> 1. The LED configuration is broken -- only one led is working, an >>>> amber led, as I found no way to know which GPIOs were used for those. >>>> The board has 6 unlabeled leds, of which 1 is blue only and 5 are >>>> amber/blue. If I understand which GPIOs go to which leds I'll submit a >>>> patch later. >>> >>> I had this problem too and wrote a script to test gpios for leds and >>> buttons (see attachment). >>> Hope it'll help you (and maybe someone else). >>> Note that it will not help if your leds are connected via shift >>> registers (as on Belkin boards). >> >> Your script helped me (but I had to modify it slightly, the ls had to >> be ls -d else it wouldn't work) > > you are right, it should be with '-d' > >> I found out 4 more leds, but I still miss 7 leds. Also, your script >> errors out on some pins: this is the log. > > sure, it returns error for those pins which are occupied by other modules > >> Here the pins that return 0s are leds. > > Leds are tested in output mode which you are skipping. > If you are testing leds only you should be skipping input mode, not > output, and see which led blinks (it should blink 3 times if you don't > interrupt it).
The thing is that it won't change gpio if I don't skip input and output manually I figured out some more leds, now I only miss 4, one of which is turned on by default. Some pins error out even though no modules apart from network are loaded. I'm trying to find those too and once done I'll update the patch -- you can disregard this one. > Testing for buttons you should press it, while in input mode, and see > if pin state changes. Buttons should work as is as I found the pins for them in the source code released by the manufacturer. >>>> 2. Wi-Fi is not working on my board probably because of the lack of >>>> RAM (the board only has 16MiB ram, but there is an empty slot for >>>> another ram chip,) but I don't know for sure. The driver loads but >>>> hostapd fails to load so I think it's not related to the specific >>>> device except for the lack of RAM. >>> >>> That's very strange, I don't have such problem with my 4/16 MB >>> Sparklan WCR-150GN. >>> I don't quite understand how it could be that it's low RAM problem if >>> all kernel modules are loaded _before_ everything else? >> >> It's hostapd that fails executing, not the .kos that fail loading. >> > > ah.. well, no problem with that anyway.. I can't seem to make it run at boot time instead -- if I free enough ram it starts after the device boots. -- Marcus905 GPG pubkey: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x1FC0ECC932FE5FAC _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
