I've had this problem since build 52 with my WRT54GL v1.1: The wireless site survey doesn't seem to work at all while the vanilla Tomato firmware and DD-WRT both work fine. Right now I'm running Tomato Firmware v1.28.8754 ND USB Ext and everything seems to be running great EXCEPT the wireless site survey. I've tried the 2.6 kernel versions as well with the same results.
Date: 02 Dec 2010 22:13
Number of posts: 18
RSS: New posts
Wireless survey works for me, running build 54, k2.6, VPN version. Router is an Asus RT-N16. You say you're running the USB version of TomatoUSB on that router? There's no USB ports on a 54GL. Might want to check your firmware version.
I also have this problem, on my WRT54GS v1.0. It worked fine in the base Tomato 1.28 build. I then switched to TomatoUSB build 51 (k2.4 NoUSB Std); I don't think I ever actually tried it with this build. Then I switched to build 52 (k2.4 NoUBS VPN) to fix the WanUP script problem and try VPN capability…and later noticed that the "Tools->Wireless Site Survey" and "Basic->Network->Wireless Channel Scan" were no longer working. This was after I had done a full NVRAM erase, and set everything back up, including heavily modified Toastman QoS setup and a dnsmasq blocking script. I didn't want to clear everything again to try it, so figured I'd wait until the next release.
So here I am, using build 54; now trying k2.6 MIPSR1 NoUSB VPN (*.9054) as well. With a full NVRAM erase, neither of the aforementioned features are working, still. Now, it -seems- that the Wireless Channel Scan finds other AP's…when hitting the 'Scan' button, the drop-down box extends, and the "Scan [#]" number increments…which seems to indicate how many other APs it finds; but then nothing actually shows up in the box. The Site Survey, when hitting 'Refresh', appears to run for a few seconds, with the 'Refresh' button graying out, then goes back to how it was.
@Boilermaker: …that you dave? If so, have you tried NoUSB VPN on your 54GL yet?
@Rotareneg: Per Boilermaker, I looked at the various build definitions to see the difference b/t Ext, VPN, etc; if you're just wanting the features in Ext, it appears that the VPN edition is based on Ext, so if you going for JFFS space, etc; you might be better off using NoUSB VPN versus USB Ext. But obviously, I'm not sure what you're after, but if it -is- JFFS, the 'build types' pages indicates, eg, for 4MB flash routers…
Std: ~300KB JFFS space (5 blocks)
Extras: ~120KB JFFS space (2 blocks) [uses 180KB/3blks vs Std]
VPN: "VPN: Extras edition merged with SgtPepperKSU' VPN build with Web GUI (no JFFS space…)" [uses 300KB/5blks vs Std]
NoUSB Std: 900KB of JFFS space (15 blocks)
This insinuates that NoUSB VPN should have ~600KB (10 blocks) JFFS space. This is naturally relative to whatever build provided the above metrics.
Wondering if something relating to free space and/or the VPN addition might be causing this, I just tried the k2.6 MIPSR1 NoUSB Std (*.9054) build; same issue as above. And then I saved off the nvram settings, tried the base Tomato 1.28 firmware, and it works fine. I saved those nvram settings, and compared TomatoUSB NoUSB Std against Tomato base. There are some added and changed settings; the default country is US vs Japan, respectively. I tried setting the base Tomato to US via console, as there's no GUI access, but rebooting sets them back to Japan. So, I put TomatoUSB back in; the nvram parameters added in Tomato remained, and setting the country parameters to Japan (JP) didn't affect anything.
So this seems to be specific to something in TomatoUSB, and apparently only on certain routers? Like I said before, the Wireless Channel Scan seems like it finds other APs, but the drop-down box shows blank. Perhaps what the driver returns is getting its pointer or memory location botched…
So this seems to be specific to something in TomatoUSB, and apparently only on certain routers?
Tomato USB uses completely different wireless driver in both - K24 and K26 branches. This is actually a much newer version of "ND" driver used in original Tomato-ND. Since this functionality comes directly from the driver this is most probably the reason.
And yes - wireless survey works on all routers I have or tested it with, including 3 different Asus models, and Linksys E3000/WRT610Nv2.
What if you run the following commands in the terminal session on your WRT-54G(L/S/etc) running Tomato USB:
wl scan --scan-type=passive; sleep 4; wl scanresults
Does it show anything? If not, try running it without the "—scan-type=passive":
wl scan; sleep 4; wl scanresults
Does it make any difference?
No results to either:
root@?:/tmp/home/root# wl scan —scan-type=passive; sleep 4; wl scanresults
root@?:/tmp/home/root# wl scan; sleep 4; wl scanresults
I spent a while messing with various args of wl, including iscan, monitor, radar, down, restart…no luck with anything. Thanks for the help though teddy - let me know if there's anything else to try.
I can confirm that Wireless Survey has stopped working on my WRT54GL v1.1 starting with Build 52. This is true for both K24 and K26 branches of NoUSB Std, neither have worked since Build 51 on this router. I tried the wl scan command with and without passive scan, and got no results that way either.
Likely related, the channel scan on the network setup page is not working either.
Are you saying that the wireless survey (and channel scan, which is the same thing essentially) on WRT54GL used to work in K24 build 51 and prior builds, but no longer works in K24 starting from build 52?? That doesn't make any sense :( - there were no changes between 51 and 52 that could possibly affect this.
Besides, judging by jwh7' test results with "wl", I'd think it never worked on this router model…
Please verify this once again - if it really ever worked with K24 builds of Tomato USB on this router, we need to know the last build which definitely worked.
I suppose you guys know for sure that the survey must show anything - i.e. you compared the results with another Tomato or dd-wrt router or inSSIDer running the survey/scan at the same location and at the same time ;) ?
To be clear:
Rotareneg is using a WRT54GL v1.1
dmb41crash is using a WRT54GL v1.1
jwh7 is using a WRT54GS v1.0
… correct? And none of you are getting survey results?
Weird. I'm running a WRT54GS v4.0 which AFAIK is almost exactly the same hardware as a GL v1.x, and survey/scan works fine for me (with K26).
Out of curiosity, for people with GL v1.1, what is your output from "wl revinfo" and "nvram get boardflags" ?
# wl revinfo vendorid 0x14e4 deviceid 0x4320 radiorev 0x22050000 chipnum 0x5352 chiprev 0x0 corerev 0x9 boardid 0x467 boardvendor 0x14e4 boardrev 1.0 driverrev 0x50a9300 ucoderev 0x1fc009d bus 0x0 phytype 0x2 phyrev 0x7 anarev 0x2 # nvram get boardflags 0x2758
Correct, GL v1.1
I'm in a very "target rich" environment, surrounded on all sides by routers broadcasting their SSID loudly and proudly with extremely strong signal relative to my location. The last version I remember this working correctly on was either K24 Build 51 or possibly 52, I'm drawing a blank now. I do know that I recently flashed to K24 Build 54 and it was still not showing any results, and I am now on K26 Build 54. Just to note, every other function of the router works perfectly normal, and yes I always clear NVRAM after each flash (since I know that's the very next thing that will be questioned)
When I get a chance, possibly later tonight, I will revert to K24 Build 50 and see if the site survey/channel scan works correctly. I'll go through the K24 and K26 builds all the way to 54 and see in which versions it does/doesn't work.
# wl revinfo vendorid 0x14e4 deviceid 0x4320 radiorev 0x22050000 chipnum 0x5352 chiprev 0x0 corerev 0x9 boardid 0x467 boardvendor 0x14e4 boardrev 1.0 driverrev 0x50a9300 ucoderev 0x1fc009d bus 0x0 phytype 0x2 phyrev 0x7 anarev 0x2 # nvram get boardflags 0x2558