Posted: Wed Dec 27, 2017 15:23 Post subject: Re: PLEASE BE SPECIFIC ON HW REVISIONS
labrok wrote:
I see here you are talking about dir-868L ,i am sure that you already know that this router has already 3 HW revisions, as i can understand ,here you are talking about Rev A1, i also know that people with other revisions will come here to read, You are not explaining clearly if your downloads are only for rev A1 or for others too, and if someone with different revisions will have serious issues if he will try this FW, please do it now. Thank you!
Ah, well i habe B1, forgot to mention that =)
strange thing is it says A1 on the motherboard but B1 on the casing.
Where do i find it, and turn it off?
is it this
"Shortcut Forwarding Engine"? if so no change
Yes, SFE is short for Shortcut Forwarding Engine. Did you reboot the router when you turn SFE off? Changing the radio button to Off and saving does not turn it off. Best to reboot router after saving.
If you really cannot reboot, the command below will disable SFE:
rmmod shortcut-fe
just got a heads up that this release is for rev 868l rev A
i have rev B but PPTP started working but now the Upnp stopped working istead
Whas it a big problem that made NVRAM clog up?
i had to use the NVRAM erase command a few times every week and reconfig everytime
The problem wasn't NVRAM. What I found was that about a year ago, the source code that initialises the 868L and 880L were changed that resulted in both routers going into a boot loop. Erasing NVRAM probably is a work-around to allow first boot to go thru but as you found out, it returned probably on subsequent boots.
The good news is that BrainSlayer recently reverted the change that causes the issue. The next official build should once again work for the 868 and 880 routers.
By next official u mean next beta on site?
i tried the r33986 from 4th of dec, and that had the nvram bug, then i tried your ver 34050 and now 34080.
if this is for REV A and next build hopefully have the nvram bug fixed ill revert to that, cause i understand that issues wil l happen on diffrent hardware =)
By next official u mean next beta on site?
i tried the r33986 from 4th of dec, and that had the nvram bug, then i tried your ver 34050 and now 34080.
if this is for REV A and next build hopefully have the nvram bug fixed ill revert to that, cause i understand that issues wil l happen on diffrent hardware =)
but if u need REV B tester just hola
Build at or after 34093 will once again work for the 868 and 880, buy only got kernel 4.4.x.
AFAIK there’re no appreciable difference between a Rev A and B for the 868, from the firmware code perspective.
Posted: Thu Dec 28, 2017 12:10 Post subject: Dir868L rev C1
quarkysg wrote:
frodrick wrote:
By next official u mean next beta on site?
i tried the r33986 from 4th of dec, and that had the nvram bug, then i tried your ver 34050 and now 34080.
if this is for REV A and next build hopefully have the nvram bug fixed ill revert to that, cause i understand that issues wil l happen on diffrent hardware =)
but if u need REV B tester just hola
Build at or after 34093 will once again work for the 868 and 880, buy only got kernel 4.4.x.
AFAIK there’re no appreciable difference between a Rev A and B for the 868, from the firmware code perspective.
Have fun.
well , i have revision C , is it safe for me to flash this firmware?
Posted: Thu Dec 28, 2017 22:09 Post subject: Re: Dir868L rev C1
labrok wrote:
well , i have revision C , is it safe for me to flash this firmware?
868 Rev A/B/C uses the same firmware dd-wrt image, so in theory it should work. I don’t remember seeing any post of Rev C users reporting success tho. If your Rev C is already running DD-WRT then it should be safe to upgrade to my builds for the 868.
Posted: Thu Dec 28, 2017 22:18 Post subject: Re: Dir868L rev C1
quarkysg wrote:
labrok wrote:
well , i have revision C , is it safe for me to flash this firmware?
868 Rev A/B/C uses the same firmware dd-wrt image, so in theory it should work. I don’t remember seeing any post of Rev C users reporting success tho. If your Rev C is already running DD-WRT then it should be safe to upgrade to my builds for the 868.
Yes its running, but with some minor issues, tranfer errors and high cpu temperature (now 81.4c with 0% usage) , also i noticed a higher than usuall noise (-92 instead of -100 on my linksys EA8500) by upgrading to your firmware will help me? and which file to use? ddwrt-webflash-dir868-880-r34080b-patched.bin is ok?
Posted: Thu Dec 28, 2017 22:36 Post subject: Re: Dir868L rev C1
labrok wrote:
Yes its running, but with some minor issues, tranfer errors and high cpu temperature (now 81.4c with 0% usage) , also i noticed a higher than usuall noise (-92 instead of -100 on my linksys EA8500) by upgrading to your firmware will help me? and which file to use? ddwrt-webflash-dir868-880-r34080b-patched.bin is ok?
The 868 tends to run a little hot. Where I’m from, the CPU temp will shoot past 90 deg C. I added a fan to my 868 to keep it running cooler, just under 80 deg C. My builds are no different from those released by BS and Kong, other than making it bootable again for the 868 and 880, so the CPU temp figures will be similar.
As for WiFi noise, it’s probably due to different signal amps or other components used between the routers. Should not be due to drivers, although my builds may not be using the same Broadcom WiFi driver version as those released by BS and Kong. Mine will probably be an older driver, but it’s working great for me so far. In any case, noise amplitude less than -90db is considered very good, so I wouldn’t worry about it.
I would suggest all 868 and 880 owners to go back to BS or Kong’s builds from the next release.
Posted: Thu Dec 28, 2017 22:47 Post subject: Dir868L rev C1
Flashed your file, succesfully, no change on temperature as you already said and on noise amplitude too , but i think the transfer errors have dissapeared, i am tranfering big files now to test it further and i will inform you later, but till now error rate is 0!
This should work when migrating from D-Link OEM FW to DD-WRT FW. Users should be able to migrate directly to this new version without having to use an older version of WRT.
shayne85 wrote:
Thanks for all the fixes and info quarkysg!
Does this mean the New BS build r34311 12-29-17 will now work on the 880L?
Posted: Sat Dec 30, 2017 17:18 Post subject: D-LINK DIR-868L rev C1
Newest r34311 works fine on my DIR-868L rev C, but throughtout is lower than quarkysg`s builds, i could get about 65MB/sec file transfer from wired GBit Desktop to wireless 866Mbps wifi card, and now just a poor 50-55 MB/sec, same position everything! But i noticed a drop in Cpu temperature (10% lower) from 85c to 78c
PS finally its bad build for DIR-868L TX errors started again and throughout is 20% less, reverted to quarkysg`s build!!!
Posted: Sat Dec 30, 2017 19:40 Post subject: Re: D-LINK DIR-868L rev C1
labrok wrote:
Newest r34311 works fine on my DIR-868L rev C, but throughtout is lower than quarkysg`s builds, i could get about 65MB/sec file transfer from wired GBit Desktop to wireless 866Mbps wifi card, and now just a poor 50-55 MB/sec, same position everything! But i noticed a drop in Cpu temperature (10% lower) from 85c to 78c
PS finally its bad build for DIR-868L TX errors started again and throughout is 20% less, reverted to quarkysg`s build!!!
WiFi performance are very much dependent on how noisy is the environment it's being tested in, unless it's conducted inside a faraday cage, so I don't think it's due to the build. In terms of WiFi performance, the official builds should be better compared to mine as the DD-WRT devs likely have access to the driver sources and their builds should have newer drivers.
CPU temperature should not vary much between the official builds and mine as well, so it's interesting that you should see a 10% drop in temperature.
The difference between the official builds and mine are mainly in the changes in the SFE code and the driver for the LAN switches, but that shouldn't account for such a massive increase in temperature for my builds. I would have thought mine should show a slight decrease in CPU temp as the SFE module would be more efficient, since its doing less work compared to what the official releases has.
In any case, I would assume your router would be idling most of the time, so the temperature you get from both the official builds and mine should be about the same.
