ALFA AWUS036AXML - unable to run a stable AP #580
Comments
|
For others that may read this issue:
This is an issue that just popped up with the recent release of 6.12.13. One reporter has reported that backing off to 6.12.12 removes the problem making it looking like a patch that went into 6.12.13 is the cause. I have tested with 6.14 rc1 and do not see this problem. My intent is to report this to linux-wireless when I have enough information. Hi @Simon566
Let's see what we can do. I am currently reading
What band do you want to use? (2.4 GHz, 5 GHz or 6 GHz) What channel do you want to use? What throughput are you looking for? What devices are connecting to this AP and what are their capabilities?
|
|
Im running it in 2,4ghz mode with 11ax Options, 20mhz Channel width. Im able to reach 100Mbit netto Transfer....
Do you think the config for this is wrong?
13.02.2025 18:44:09 morrownr ***@***.***>:
…
[Bild][morrownr][https://avatars.githubusercontent.com/u/69053122?s=20&v=4]*morrownr* left a comment (morrownr/USB-WiFi#580)[#580 (comment)]
For others that may read this issue:
odroid-h3 kernel: BUG: kernel NULL pointer dereference, address: 0000000000000400
This is an issue that just popped up with the recent release of 6.12.13. One reporter has reported that backing off to 6.12.12 removes the problem making it looking like a patch that went into 6.12.13 is the cause. I have tested with 6.14 rc1 and do not see this problem. My intent is to report this to linux-wireless when I have enough information.
Hi @Simon566[https://github.com/Simon566]
...never able to run a stable AP... with Alfa AXML
Let's see what we can do.
I am currently reading *hostapd_alfa.conf.txt*. I am not finished reading it but I have some questions:
* Is that the hostapd.conf file you are actually using?
* Is there a reason that you are taking an adapter that is capable of AXE3000 class performance and running it is what amounts to G72. I mean, it is not even running in N150 (WiFi 4) mode, let along the WiFi 6 mode that it is capable of in all 3 bands.
* I need to know what your intent is:
What band do you want to use? (2.4 GHz, 5 GHz or 6 GHz)
What channel do you want to use?
What throughput are you looking for?
What devices are connecting to this AP and what are their capabilities?
* I am seeing some problems in the hostapd.conf. I understand that hostapd.conf can be a real challenge. Been there, done that. Let me help you. Answer my questions about what you want to do and I'll see what I can do.
—
Reply to this email directly, view it on GitHub[#580 (comment)], or unsubscribe[https://github.com/notifications/unsubscribe-auth/AKOIFAK2SFYE4XF7LTXKRWD2PTKWPAVCNFSM6AAAAABXBPP2DKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNJXGMZDENZZG4].
You are receiving this because you were mentioned.
|
|
Opps... I had to correct myself, the following will work for N150: hw_mode=g
Let me continue reading. Setting up AX on band 1 (2.4 GHz) is not as simple as it could be. Even the guide I have post for the mt7921 defaults to 5 GHz. Let me work this and let you know what I figure out.
Let me work with this. |
|
Your hostapd.conf is really pretty good. My old eyes had a hard time going through the file as it is due to all of the documentation. I'm not saying using the documentation is a bad idea. What I did is run the following to get rid of the comments and the document came to life for me: $ sed '/^#/ d' < inputFile.txt > outputFile.txt I did leave the lines in the order they were in. I have used Next: It appears that the Odroid H3 has two USB2 ports. Make sure you are plugging the AXML into a USB2 port, not a USB3 port. Why? We want to avoid the interference caused by USB3 in the 2.4 GHz band. Next: Does the back of your AXML have a line like this? 80211ax 2x2+BT5.2 If so, do you use the bluetooth? If yes, it would probably be best not to do so since you are on the 2.4 GHz band. Why? bluetooth works on the 2.4 GHz band and will interfere with wifi on the 2.4 GHz band. I have advocated for many year not to buy USB WiFi adapters that have bluetooth support. Next: There is no good reason for scatter-gather to be active in your setup and I have seen problems with it on some small board systems depending on the chip used in the USB3 hub so: To shut down scatter-gather support: Open a terminal (Ctrl+Alt+t) Final word for now: Let me know if any of the above help your AXML become stable in AP mode. |
|
Hi,
scatter gather was on , will turn it off will report back thank you |
|
here is another crash , with your adjustments made [Fr Feb 14 14:41:25 2025] br-lan: port 2(enp1s0) entered disabled state |
The problem you posted about above is specific to kernel 6.12.13. You need to move back to kernel 6.12.12 or earlier while the problem is fixed. |
|
here is a crash from 6.12.12 Feb 14 18:11:51 odroid-h3 kernel: mt7921u 1-4:1.3: Message 00020003 (seq 10) timeout |
|
Wow! Have you tried all of the usb ports on that h3? Take a look at this message: Do you have another system to test with? The more we talk, the more it seems to me that the is something more serious than a driver bug going on here. USB can be a challenge in some cases. Is your adapter plugged into a powered hub or directly into a port on the system? |
|
the Odroid H3 is a intel Atom based system with a widespread chip. i wouldnt expect issues with USB ports , especially the USB2 which im using . is xHCI handoff and things like that not only USB3 ? i tried different Ports , USB3 / USB2 but the results are the same. So far i havent got a different test system , maybe it s going to be a odroid h4 some day |
Famous last words. Does your bios have any settings for USB other than on and off? What I am seeing are things that I'm not used to seeing. Now that you went back to 6.12.12 it appears somewhat different but the only times I've seen ugliness like this is when a user has several usb devices plugged in and the usb bus is not providing adequate power for what is requested. That is a common problem on RasPis because their usb subsystem does not provide anywhere near spec power. Any thoughts on whether that could be a problem? |
|
As im using a x86 Board with a 80w or so PSU, Power is No issue
Yes,i do have BIOS settings for USB, does xhci handoff also Count for usb2? Im Happy to Test suggested settings
15.02.2025 19:48:34 morrownr ***@***.***>:
…
[Bild][morrownr][https://avatars.githubusercontent.com/u/69053122?s=20&v=4]*morrownr* left a comment (morrownr/USB-WiFi#580)[#580 (comment)]
@Simon566[https://github.com/Simon566]
i wouldnt expect issues with USB ports
Famous last words.
Does your bios have any settings for USB other than on and off?
What I am seeing are things that I'm not used to seeing. Now that you went back to 6.12.12 it appears somewhat different but the only times I've seen ugliness like this is when a user has several usb devices plugged in and the usb bus is not providing adequate power for what is requested. That is a common problem on RasPis because their usb subsystem does not provide anywhere near spec power. Any thoughts on whether that could be a problem?
—
Reply to this email directly, view it on GitHub[#580 (comment)], or unsubscribe[https://github.com/notifications/unsubscribe-auth/AKOIFAMXS4KWHFRXZGDSWMD2P6DYDAVCNFSM6AAAAABXBPP2DKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNRRGA2DGMRYGY].
You are receiving this because you were mentioned.
|
Well, I have one x86 system with a 350 w PSU and a usb subsystem that has a power problem.
Unknown but xhci handoff is a known problem so I would recommend changing the setting to test. |
|
more crashes , this time with 6.12.14 and the changed USB settings: Feb 17 16:01:38 odroid-h3 dnsmasq-dhcp[1563]: Fehler beim Senden des DHCP-Pakets an 10.10.15.138: Die Ressource ist zur Zeit nicht verfügbar |
Your 6.12.14 kernel doesn't seem to have the one line patch from the other thread, does it? |
|
it has the Patch, i will Doublecheck and come Back tomorrow
17.02.2025 21:52:53 morrownr ***@***.***>:
…
[Bild][morrownr][https://avatars.githubusercontent.com/u/69053122?s=20&v=4]*morrownr* left a comment (morrownr/USB-WiFi#580)[#580 (comment)]
@Simon566[https://github.com/Simon566]
BUG: kernel NULL pointer dereference, address: 0000000000000400
Your 6.12.14 kernel doesn't seem to have the one line patch from the other thread, does it?
—
Reply to this email directly, view it on GitHub[#580 (comment)], or unsubscribe[https://github.com/notifications/unsubscribe-auth/AKOIFAO64O6VYUQLIEIPG2D2QJD2HAVCNFSM6AAAAABXBPP2DKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNRUGA2TMOJTGQ].
You are receiving this because you were mentioned.
|
|
shame on me , patch was overwritten by git , i will repeat the test sorry |
|
It happens. We might as well laugh a little. I'll bet we figure this out. |
|
nothing happened since 24h with the NULL ptr patch, it runs . BUT:
i will keep testing it for at least a week, maybe the rtw_8812au is the one to blame |
That driver is new. We have tested as best we can but that driver simply does not have time on its side. As more people start using it, starting with kernel 6.14 and the release of Ubuntu 25.04 and other distros with kernel 6.14, much more testing will take place. I assume you are using lwfinger/rtw88, right? If so, you could just do a quick uninstall: $ sudo make uninstall And then install the following driver: https://github.com/morrownr/8812au-20210820 So as to do a comparison. That repo is pretty darn stable. I have been maintaining it for a long time and it does AP mode really well. You should be able to do a comparison.
When you have something to report, please do at: |
|
it looks pretty stable 6.12.14 with the NULL ptr patch + the repo for the 8812au adapter here no dmesg warnings none , just runs ..... will report back in 2-3 weeks |
Alfa adapter:
Feb 12 11:43:58 odroid-h3 kernel: BUG: kernel NULL pointer dereference, address: 0000000000000400
Feb 12 11:43:58 odroid-h3 kernel: #PF: supervisor read access in kernel mode
Feb 12 11:43:58 odroid-h3 kernel: #PF: error_code(0x0000) - not-present page
Feb 12 11:43:58 odroid-h3 kernel: PGD 0 P4D 0
Feb 12 11:43:58 odroid-h3 kernel: Oops: Oops: 0000 [https://github.com//issues/1] PREEMPT SMP NOPTI
Feb 12 11:43:58 odroid-h3 kernel: CPU: 3 UID: 0 PID: 235904 Comm: mt76-usb-rx phy Tainted: G U W OE 6.12.13 #14
Feb 12 11:43:58 odroid-h3 kernel: Tainted: [U]=USER, [W]=WARN, [O]=OOT_MODULE, [E]=UNSIGNED_MODULE
Feb 12 11:43:58 odroid-h3 kernel: Hardware name: HARDKERNEL ODROID-H3/ODROID-H3, BIOS 5.19 07/19/2023
Feb 12 11:43:58 odroid-h3 kernel: RIP: 0010:mt792x_rx_get_wcid+0x38/0x70 [mt792x_lib]
Feb 12 11:43:58 odroid-h3 kernel: Code: 77 48 0f b7 f6 48 81 c6 84 06 00 00 48 8b 44 f7 08 48 85 c0 74 36 84 d2 75 32 f6 80 bc 00 00 00 01 74 27>
Feb 12 11:43:58 odroid-h3 kernel: RSP: 0018:ffffa8e68334bdb8 EFLAGS: 00010202
Feb 12 11:43:58 odroid-h3 kernel: RAX: 0000000000000000 RBX: ffff8d05e3b83d00 RCX: 0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: RDX: 0000000000000000 RSI: 000000000000068a RDI: ffff8d05ce022000
Feb 12 11:43:58 odroid-h3 kernel: RBP: ffff8d05cad9c800 R08: 0000000000000000 R09: 0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: R10: ffff8d05e3b83d00 R11: 0000000000000000 R12: ffff8d05c5591000
Feb 12 11:43:58 odroid-h3 kernel: R13: 0000000038010148 R14: 0000000080843806 R15: ffff8d05ce022000
Feb 12 11:43:58 odroid-h3 kernel: FS: 0000000000000000(0000) GS:ffff8d092ff80000(0000) knlGS:0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Feb 12 11:43:58 odroid-h3 kernel: CR2: 0000000000000400 CR3: 000000020f222000 CR4: 0000000000352ef0
Feb 12 11:43:58 odroid-h3 kernel: Call Trace:
Feb 12 11:43:58 odroid-h3 kernel:
Feb 12 11:43:58 odroid-h3 kernel: ? __die+0x23/0x80
Feb 12 11:43:58 odroid-h3 kernel: ? page_fault_oops+0x173/0x5d0
Feb 12 11:43:58 odroid-h3 kernel: ? exc_page_fault+0x76/0x1b0
Feb 12 11:43:58 odroid-h3 kernel: ? asm_exc_page_fault+0x26/0x30
Feb 12 11:43:58 odroid-h3 kernel: ? mt792x_rx_get_wcid+0x38/0x70 [mt792x_lib]
Feb 12 11:43:58 odroid-h3 kernel: mt7921_queue_rx_skb+0x1df/0x9a0 [mt7921_common]
Feb 12 11:43:58 odroid-h3 kernel: mt76u_rx_worker+0x5be/0x5f0 [mt76_usb]
Feb 12 11:43:58 odroid-h3 kernel: ? __pfx___mt76_worker_fn+0x10/0x10 [mt76]
Feb 12 11:43:58 odroid-h3 kernel: __mt76_worker_fn+0x4c/0x80 [mt76]
Feb 12 11:43:58 odroid-h3 kernel: kthread+0xde/0x110
Feb 12 11:43:58 odroid-h3 kernel: ? __pfx_kthread+0x10/0x10
Feb 12 11:43:58 odroid-h3 kernel: ret_from_fork+0x31/0x50
Feb 12 11:43:58 odroid-h3 kernel: ? __pfx_kthread+0x10/0x10
Feb 12 11:43:58 odroid-h3 kernel: ret_from_fork_asm+0x1a/0x30
Feb 12 11:43:58 odroid-h3 kernel:
Feb 12 11:43:58 odroid-h3 kernel: Modules linked in: rtw_8812au(OE) ftdi_sio mt7921u mt792x_usb mt7921_common mt792x_lib mt76_connac_lib mt76_us>
Feb 12 11:43:58 odroid-h3 kernel: snd_soc_acpi snd_soc_core x86_pkg_temp_thermal nls_ascii snd_compress nls_cp437 soundwire_bus intel_powerclam>
Feb 12 11:43:58 odroid-h3 kernel: sunrpc garp stp mrp loop dm_mod llc efi_pstore configfs efivarfs ip_tables x_tables autofs4 hid_logitech_hidp>
Feb 12 11:43:58 odroid-h3 kernel: CR2: 0000000000000400
Feb 12 11:43:58 odroid-h3 kernel: ---[ end trace 0000000000000000 ]---
Feb 12 11:43:58 odroid-h3 kernel: RIP: 0010:mt792x_rx_get_wcid+0x38/0x70 [mt792x_lib]
Feb 12 11:43:58 odroid-h3 kernel: Code: 77 48 0f b7 f6 48 81 c6 84 06 00 00 48 8b 44 f7 08 48 85 c0 74 36 84 d2 75 32 f6 80 bc 00 00 00 01 74 27>
Feb 12 11:43:58 odroid-h3 kernel: RSP: 0018:ffffa8e68334bdb8 EFLAGS: 00010202
Feb 12 11:43:58 odroid-h3 kernel: RAX: 0000000000000000 RBX: ffff8d05e3b83d00 RCX: 0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: RDX: 0000000000000000 RSI: 000000000000068a RDI: ffff8d05ce022000
Feb 12 11:43:58 odroid-h3 kernel: RBP: ffff8d05cad9c800 R08: 0000000000000000 R09: 0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: R10: ffff8d05e3b83d00 R11: 0000000000000000 R12: ffff8d05c5591000
Feb 12 11:43:58 odroid-h3 kernel: R13: 0000000038010148 R14: 0000000080843806 R15: ffff8d05ce022000
Feb 12 11:43:58 odroid-h3 kernel: FS: 0000000000000000(0000) GS:ffff8d092ff80000(0000) knlGS:0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Feb 12 11:43:58 odroid-h3 kernel: CR2: 0000000000000400 CR3: 000000018337c000 CR4: 0000000000352ef0
Feb 12 11:43:58 odroid-h3 kernel: note: mt76-usb-rx phy[235904] exited with irqs disabled
and
Feb 12 11:43:58 odroid-h3 kernel: ---[ end trace 0000000000000000 ]---
Feb 12 11:43:58 odroid-h3 kernel: RIP: 0010:mt792x_rx_get_wcid+0x38/0x70 [mt792x_lib]
Feb 12 11:43:58 odroid-h3 kernel: Code: 77 48 0f b7 f6 48 81 c6 84 06 00 00 48 8b 44 f7 08 48 85 c0 74 36 84 d2 75 32 f6 80 bc 00 00 00 01 74 27>
Feb 12 11:43:58 odroid-h3 kernel: RSP: 0018:ffffa8e68334bdb8 EFLAGS: 00010202
Feb 12 11:43:58 odroid-h3 kernel: RAX: 0000000000000000 RBX: ffff8d05e3b83d00 RCX: 0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: RDX: 0000000000000000 RSI: 000000000000068a RDI: ffff8d05ce022000
Feb 12 11:43:58 odroid-h3 kernel: RBP: ffff8d05cad9c800 R08: 0000000000000000 R09: 0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: R10: ffff8d05e3b83d00 R11: 0000000000000000 R12: ffff8d05c5591000
Feb 12 11:43:58 odroid-h3 kernel: R13: 0000000038010148 R14: 0000000080843806 R15: ffff8d05ce022000
Feb 12 11:43:58 odroid-h3 kernel: FS: 0000000000000000(0000) GS:ffff8d092ff80000(0000) knlGS:0000000000000000
Feb 12 11:43:58 odroid-h3 kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Feb 12 11:43:58 odroid-h3 kernel: CR2: 0000000000000400 CR3: 000000018337c000 CR4: 0000000000352ef0
Feb 12 11:43:58 odroid-h3 kernel: note: mt76-usb-rx phy[235904] exited with irqs disabled
Feb 12 11:43:58 odroid-h3 dnsmasq-dhcp[1572]: Fehler beim Senden des DHCP-Pakets an 10.10.15.138: Die Ressource ist zur Zeit nicht verfügbar
no patches , vanilla kernel 6.12.13 on Odroid H3 ( X86-64) with the USB Alfa adapter
running Debian Bookworm , no network manager
it was never able to run a stable AP
tx urb failed issues
null ptr accesses
device is unusable since i bought it ( one year ) passing over 5.19 , 6.6 and 6.12 kernel versions
bluetooth crash is avoided by your FAQ and blacklisting
hostapd_alfa.conf.txt
The text was updated successfully, but these errors were encountered: