Howdy folks,
Asuswrt-Merlin 374.38 is now available. Note that this release is only available for SDK6-based devices, which means there is no RT-N16 builds, and no RT-N66 SDK5 build either. The latest firmware code doesn't work properly on the SDK5 platform, and I didn't want to delay this release any further so I decided to release it without these two builds. I might have to wait until the next time Asus releases firmware code for SDK5-based devices to get these fixed - we'll have to wait and see. For those devices, I recommend using 374.36 Beta 1 for the time being.
One of the big changes in this release is the merge with GPL code 374.2078 (which is almost the same as the ARM release 374.542, with a few additional changes specific to MIPS devices).
The 2078 merge brings a newer SDK for the N66 and AC66U devices. Furthermore, N66 has an even newer driver dated from late december, which is supposed to fix some of the issues encountered with this device. I'm unable to confirm so far if it did fix the signal strength issues on the RT-N66U as I was getting mixed results, so please give this build a try. See below for my recommendation as to how to proceed with switching from an SDK5 to an SDK6 build.
This newer SDK should also take care of the broken HW acceleration with PPPoE. Please give this a try as well if you are using a 200+ Mbps PPPoE connection.
Lastly, regarding this new SDK: I uploaded a new experimental build for the RT-N66U labeled "374.0-em". This build has a special switch enabled while building the firmware. At this point I am unsure if it will improve, or have no effect at all, on the wireless performance of the RT-N66U. If you can, please test both builds, and let me know if there's any difference in wifi performance between the two. I will provide more details on this experimental build once I start getting more feedback (as I want to ensure I don't issue a placebo)
Next on the list of changes is a newer wifi driver was integrated for the AC56 and AC68. Asus seems to now be compiling two different drivers for these two devices, so I'm unsure if this is to further optimize them for their respective platforms. The driver I use should be very close to the 374.542 driver (I got it a day or two before Asus released that FW, so it might be slightly different, either older or newer).
And as usual, a couple of additional fixes, all documented in the changelog (mostly dealing with regressions introduced in 374.37).
For those switching between SDK builds (N66U)
I'd like to gather some feedback concerning the wifi signal strength/performance of this new driver. To ensure we do get proper results, a factory default reset + manual reconfiguration is needed - any result done without doing this step will simply be useless. I recommend the following procedure:
1) Make a backup of your current settings, and name it "Settings-RTN66-SDK5.cfg" or something like that, so you know to which SDK those settings belonged
1a) Optional: if you want to save time, note down your DHCP static leases and port forwards. Through SSH or Run Cmd:
2) Check out the RSSI that you get from your router from a distant place in your house (somewhere where the signal is below -40dbM is probably a good place to test)
3) Flash with the new build (either regular or the -em experimental build)
4) Do a factory default reset
5) Manually reconfigure everything
5a) Optional: if you previously noted down vts_rulelist and dhcp_staticlist, reimport them:
6) Check again the RSSI from the same location (ideally with the same channel as you were previously using)
If you want to also test the EM build, just flash it, and measure again. No need to reset to factory default when switching between regular and EM builds.
As you can see, the goal of this new build was to merge new GPL code to resolve various issues regarding wifi, USB and HW acceleration.
Barring any major issue, next release will probably target merging with the next GPL release to implement the finalized USB HUB support (which isn't complete yet in 2078/542). There's two or three things I have in the pipeline as well that should debut in this next release if all goes as planned.
Enjoy! And remember to always specify which router model you are referring to when posting any feedback. Wireless feedback without knowing which router you are talking about is simply totally useless.
Asuswrt-Merlin 374.38 is now available. Note that this release is only available for SDK6-based devices, which means there is no RT-N16 builds, and no RT-N66 SDK5 build either. The latest firmware code doesn't work properly on the SDK5 platform, and I didn't want to delay this release any further so I decided to release it without these two builds. I might have to wait until the next time Asus releases firmware code for SDK5-based devices to get these fixed - we'll have to wait and see. For those devices, I recommend using 374.36 Beta 1 for the time being.
One of the big changes in this release is the merge with GPL code 374.2078 (which is almost the same as the ARM release 374.542, with a few additional changes specific to MIPS devices).
The 2078 merge brings a newer SDK for the N66 and AC66U devices. Furthermore, N66 has an even newer driver dated from late december, which is supposed to fix some of the issues encountered with this device. I'm unable to confirm so far if it did fix the signal strength issues on the RT-N66U as I was getting mixed results, so please give this build a try. See below for my recommendation as to how to proceed with switching from an SDK5 to an SDK6 build.
This newer SDK should also take care of the broken HW acceleration with PPPoE. Please give this a try as well if you are using a 200+ Mbps PPPoE connection.
Lastly, regarding this new SDK: I uploaded a new experimental build for the RT-N66U labeled "374.0-em". This build has a special switch enabled while building the firmware. At this point I am unsure if it will improve, or have no effect at all, on the wireless performance of the RT-N66U. If you can, please test both builds, and let me know if there's any difference in wifi performance between the two. I will provide more details on this experimental build once I start getting more feedback (as I want to ensure I don't issue a placebo)
Next on the list of changes is a newer wifi driver was integrated for the AC56 and AC68. Asus seems to now be compiling two different drivers for these two devices, so I'm unsure if this is to further optimize them for their respective platforms. The driver I use should be very close to the 374.542 driver (I got it a day or two before Asus released that FW, so it might be slightly different, either older or newer).
And as usual, a couple of additional fixes, all documented in the changelog (mostly dealing with regressions introduced in 374.37).
For those switching between SDK builds (N66U)
I'd like to gather some feedback concerning the wifi signal strength/performance of this new driver. To ensure we do get proper results, a factory default reset + manual reconfiguration is needed - any result done without doing this step will simply be useless. I recommend the following procedure:
1) Make a backup of your current settings, and name it "Settings-RTN66-SDK5.cfg" or something like that, so you know to which SDK those settings belonged
1a) Optional: if you want to save time, note down your DHCP static leases and port forwards. Through SSH or Run Cmd:
Code:
nvram get vts_rulelist
nvram get dhcp_staticlist
3) Flash with the new build (either regular or the -em experimental build)
4) Do a factory default reset
5) Manually reconfigure everything
5a) Optional: if you previously noted down vts_rulelist and dhcp_staticlist, reimport them:
Code:
nvram set vts_rulelist="what you noted"
nvram set dhcp_staticlist="what you noted"
nvram commit
6) Check again the RSSI from the same location (ideally with the same channel as you were previously using)
If you want to also test the EM build, just flash it, and measure again. No need to reset to factory default when switching between regular and EM builds.
As you can see, the goal of this new build was to merge new GPL code to resolve various issues regarding wifi, USB and HW acceleration.
Barring any major issue, next release will probably target merging with the next GPL release to implement the finalized USB HUB support (which isn't complete yet in 2078/542). There's two or three things I have in the pipeline as well that should debut in this next release if all goes as planned.
Enjoy! And remember to always specify which router model you are referring to when posting any feedback. Wireless feedback without knowing which router you are talking about is simply totally useless.