Replies: 30 comments 7 replies
-
|
Hi @bjlockie
I did a quick look and this is interesting. It is a USB WiFi adapter but it is neither Mediatek or Realtek. Could we have another chip company getting into the usb chip business?
It appears that the Linux driver is based on mac80211, which is what we want.
It appears the vid-pid is saying the maker is aicsemi and the product is 8800. I'd say this is the chip maker default vid/pid for the chipset in use. My guess is that we are looking for a company called AIC Semiconductor or something close to that. That lsusb came up with information is curious. It had to be added to the ids file. Let me investigate the company and check in the kernel. |
Beta Was this translation helpful? Give feedback.
-
|
AIC Semiconductor: |
Beta Was this translation helpful? Give feedback.
-
|
I was looking at that github repo that has a driver. It does not look like a fresh driver. You might want to go the web site of PIX-Link to see if you can download a recent version of the driver. Send a copy to me. I was looking around the Makefile. I can probably talk you through installing it. Some observations: lsusb showed USB2, was it in a USB2 port? Interesting project you have there. Heck, I like the idea of another usb competitor showing up but they need to get that driver in the kernel if they want to compete. |
Beta Was this translation helpful? Give feedback.
-
|
It is a Aicsemi aic8800 chipset. https://github.com/lynxlikenation/aic8800 This repo has less errors but I still can't compile it: There seems to be a Gentoo package: |
Beta Was this translation helpful? Give feedback.
-
This being out-of-kernel will suffer from all of the same crap that the Realtek out-of-kernel drivers suffer from. Given the development model of Linux, in-kernel is the only real solution. I found the following repo: https://github.com/radxa-pkg/aic8800 The many firmware files are quite the mess, If we can sort out what is the latest release, I can try to help you get it going. I don't need any adapter to compile the driver. Heck, I don't even need to install the firmware files. You might want to see about contacting aicsemi to see if you can make it clear to someone that will listen that these out-of-kernel drivers are going make this product unpopular in the world Linux. If they need help connecting to the right people to start getting the driver in the kernel, I can help. |
Beta Was this translation helpful? Give feedback.
-
|
I couldn't compile the radxa repo but they do have assets which I downloaded. :-) It is still stuck at USB2. These are relative iperfs. `$ iperf3 -c 192.168.1.1 [ ID] Interval Transfer Bitrate Retr iperf Done. $ iperf3 -c 192.168.1.1 [ ID] Interval Transfer Bitrate Retr modinfo for the aic8800: |
Beta Was this translation helpful? Give feedback.
-
|
Here is a little info that might add to what we have:
Based on the reading I have been doing, I beginning to think the 8800 based adapters are only USB2 capable. I'm seeing some adapters that are single band only and are advertised as AX300. Much research needs to be done.
You got something to compile? Do tell. What do you get with: $ iw list and $ iw dev |
Beta Was this translation helpful? Give feedback.
-
|
I didn't compile anything, they had precompiled .deb packages. I should have done iw dev first. :-) The Intel has 2 streams and the aic is only 1 stream. |
Beta Was this translation helpful? Give feedback.
-
|
The Pix-Link.com website says it is USB2 but aliexpress ad says USB3. |
Beta Was this translation helpful? Give feedback.
-
You can't really precompile a driver that could be installed on different kernels so I'll bet it is a dkms installation. Point me to that driver, please.
Well, that is way beyond anything Realtek has provided so that is a good sign. However, that we are getting an out-of-kernel driver, whether it is fully standards compliant or not, is a sign that aicsemi does not understand Linux and its development model. This driver may be fully mac80211 compliant and I hope it is but that is only part of how things need to be done. They need to working this through linux-wireless or they will just end up with pissed users. |
Beta Was this translation helpful? Give feedback.
-
|
You're right, it is dkms because I had to install the dkms package. I wonder if they have more recent USB devices with a later aic chipset. :-) |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
The version in the above link is the most recent release according to the person that packaged it. Have you switched out and tried the driver in the link above? I'm curious what you are finding. |
Beta Was this translation helpful? Give feedback.
-
|
The manufacturer webpage says it is USB2. |
Beta Was this translation helpful? Give feedback.
-
The author of the package that I provided the link for also provided the package to AIC Semi so, you are correct, it is likely the same. I'm doing my best to get to the bottom of this and I will know more of the next couple of weeks. I need to test compiling the code. The problem with a deb is that it only works for Debian based distros while leaving the rest of the world out in the cold... so maybe I can start a repo so we can get smart on this. Now, my question was actually related to how the driver and adapter are performing? |
Beta Was this translation helpful? Give feedback.
-
It is half as good as the built in Intel AX201 so I'm not using it. |
Beta Was this translation helpful? Give feedback.
-
Ouch! Compiling drivers for OpenWRT is way into pain territory. |
Beta Was this translation helpful? Give feedback.
-
|
do this driver work on linux even if its not in-kernel driver? |
Beta Was this translation helpful? Give feedback.
-
|
i also have this wireless adapter and im new here so i dont know much about this usb wireless adapter |
Beta Was this translation helpful? Give feedback.
-
|
I have successfully run the AIC8800 wireless card on LEDE, but the signal strength and connected devices do not show AP mode and cannot drive AX. |
Beta Was this translation helpful? Give feedback.
-
|
An adapter maker recently sent me an adapter based on an AIC Semi chip and I am aware of the Linux driver but I simply do not have time to investigate and report on what is going on for the time being. I think it is a good idea to have another competitor for USB WiFi chips but what I have not seen is an AIC Semi presence on linux-wireless. This causes me to be a little less than enthusiastic about their Linux support as the last thing we need is another company kicking out out-of-kernel drivers. If anyone in this thread has the time or desire to contact AIC Semi, please do and explain to them that they need to join the linux-wireless mailing list and do driver development there. Until that happens, it is hard to get excited about their offerings. If they ask why, tell them to show up here at github.com/morrownr/USB-WiFi and post a message in Issues requesting information. |
Beta Was this translation helpful? Give feedback.
-
|
Im on Ubuntu kernel 5.15, tried the dkms debs but is stucked as USB 2.0 device not switching to wifi adapter. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @MR0000001 Need more detail. What adapter? What driver? What computer? Provide the results of: $ lsusb |
Beta Was this translation helpful? Give feedback.
-
|
Identical to bjlockie. |
Beta Was this translation helpful? Give feedback.
-
|
any news on the AIC8800? :( |
Beta Was this translation helpful? Give feedback.
-
|
I have attempted to contact the company but have been ignored. The message I tried to get across is that while Linux support is appreciated, they need to get plugged into linux-wireless and do the work and maintenance at linux-wireless. The last thing we need is another supplier of Linux wireless drivers that are out-of-kernel as that does not work with the Linux dev model. Anyway, I have no intention to add any information about their products on this site until such time as I see them connected into linux-wireless. |
Beta Was this translation helpful? Give feedback.
-
Could we consider contacting them again, perhaps by requesting assistance from someone else? For instance, in this comment, it seems that @RadxaYuntian has an established communication channel with the company:
Maybe @RadxaYuntian could help relay your message about the importance of submitting the driver to the upstream Linux kernel? |
Beta Was this translation helpful? Give feedback.
-
In this forum thread, a user said that the company doesn't publish the datasheets. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @antermin
I've actually done this as I have contacts at a couple of adapter makers. I have actually been sent some adapters that use AICSemi chips but I have decided that until AICSemi starts doing their Linux drivers the right way (in-kernel), they will not get any mention on this site. If you or anyone else wants to try to make contact with them, please do. |
Beta Was this translation helpful? Give feedback.
-
|
Currently on openwrt For clarification IP is from Riviera Network INC more here https://github.com/ulli-kroll/aic8800 I don't answer question, there If someone owns a "strange" device.
bluetooth on SDIO is working, sort of. every time I look at the code I need to drink |
Beta Was this translation helpful? Give feedback.
-
|
I looked into a thread in linux-wireless. The submitter does not appear to be from AIC / CEVA / RivieraWaves. Last week (Dec 3), @Cryolitia said the following in a reply:
The unclear redistribution license for firmware gives us another reason to avoid the AIC chips, I guess. |
Beta Was this translation helpful? Give feedback.
-
I am aware of this. I monitor linux-wireless. There was also a previous series of patch where someone tried to submit the driver code as released by AICSemi. It did not go well because the code needs a lot of work for it to be included in the kernel. The kernel devs that responded were nice and offered help but I think it overwhelmed the person that submitted the code. The submission needs to be done by an AICSemi dev, period. AICSemi needs to know that what they are doing is not Linux support. Out of kernel drivers do not work because of the Linux development model. The model is like it is because that is how resources can be maximized. Don't fight it, work with it. |
Beta Was this translation helpful? Give feedback.
-
I don't think so. We have just force their to clarify the driver's source code GPLv2 last month. So everyone could submit it. The AIC's developers are basically completely ignorant of the mainline kernel; they don't even understand Incidentally, part of their driver was clearly copied from the mainline kernel[2], and after the community pointed this out, they modified the copyright information in most of the file headers in the September release[3]. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @Cryolitia
I said the submission NEEDS to be done by AICSemi, not that it must be... There will be a lot of challenges for anyone external to AICSemi that tries to submit the code. Even if someone does get the code in the kernel, it will be a challenge to maintain. Without an insider that knows the code in the firmware, it can be a long series of trail and error efforts only to come to the conclusion that there must be a bug in the firmware.
So, how long until they go under? We have two established companies providing USB chipsets. Intel and Qualcomm and everyone besides Mediatek and Realtek have made the decision to not participate in the USB Wireless market. That statement is not about Linux, it is about the entire USB market. I wish them well but give them about a zero percent chance of making it unless they allow some of us to help them do Linux support the right way. |
Beta Was this translation helpful? Give feedback.
-
As a kernel developer with no vested interest in AIC, I want to say that they are cheap enough, while many Chinese embedded hardware manufacturers only care about cost.
They don't disallow us, they just don't know about allow. If you want to do that, you could start at Radxa's repo on GitHub. At least we have a GPLv2 now. But I don't think they have the ability to even help you in the true way about mainline kernel. |
Beta Was this translation helpful? Give feedback.
-
|
Aside from being cheap, do they actually work well? Assuming you're okay with the out-of-kernel driver, how's the Wi-Fi performance and stability? Are they reliable for daily use? |
Beta Was this translation helpful? Give feedback.
-
I get one on Radxa Q6A. I haven't specifically tested its performance, but it's not bad enough to make me notice. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I found this: https://forum.openwrt.org/t/support-usb-adapter-wifi-6-ax-aic8800-dc/199187/10
I couldn't find out what chipset it is so I bought it because it is cheap.
I hoped it was a Mediatek because it comes with a small CD. :-)
Does anyone understand this commit:
kooiot/openwrt@f93e65b
My guess it builds a package but is not part of the official openwrt git?
` sudo lsusb -d a69c:8800 -vv
Bus 001 Device 007: ID a69c:8800 aicsemi AIC Wlan
Device Descriptor:
bLength 18
bDescriptorType 1
bcdUSB 2.00
bDeviceClass 0 [unknown]
bDeviceSubClass 0 [unknown]
bDeviceProtocol 0
bMaxPacketSize0 64
idVendor 0xa69c aicsemi
idProduct 0x8800 AIC Wlan
bcdDevice 1.00
iManufacturer 1 aicsemi
iProduct 2 AIC Wlan
iSerial 3 20190227
bNumConfigurations 1
Configuration Descriptor:
bLength 9
bDescriptorType 2
wTotalLength 0x0020
bNumInterfaces 1
bConfigurationValue 1
iConfiguration 0
bmAttributes 0xa0
(Bus Powered)
Remote Wakeup
MaxPower 500mA
Interface Descriptor:
bLength 9
bDescriptorType 4
bInterfaceNumber 0
bAlternateSetting 0
bNumEndpoints 2
bInterfaceClass 255 Vendor Specific Class
bInterfaceSubClass 255 Vendor Specific Subclass
bInterfaceProtocol 255 Vendor Specific Protocol
iInterface 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x01 EP 1 OUT
bmAttributes 2
Transfer Type Bulk
Synch Type None
Usage Type Data
wMaxPacketSize 0x0200 1x 512 bytes
bInterval 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x82 EP 2 IN
bmAttributes 2
Transfer Type Bulk
Synch Type None
Usage Type Data
wMaxPacketSize 0x0200 1x 512 bytes
bInterval 0
Device Qualifier (for other device speed):
bLength 10
bDescriptorType 6
bcdUSB 2.00
bDeviceClass 0 [unknown]
bDeviceSubClass 0 [unknown]
bDeviceProtocol 0
bMaxPacketSize0 64
bNumConfigurations 0
can't get debug descriptor: Resource temporarily unavailable
Device Status: 0x0000
(Bus Powered)
`
Beta Was this translation helpful? Give feedback.
All reactions