Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

periodic freezes with the original driver and your fork #1

Open
lagerspetz opened this issue Dec 2, 2013 · 16 comments
Open

periodic freezes with the original driver and your fork #1

lagerspetz opened this issue Dec 2, 2013 · 16 comments

Comments

@lagerspetz
Copy link

Hi,
I am getting periodic network freezes with the original driver and also with your fork. This is mostly visible in networked real-time games where the action just freezes for a while, and then continues. It lasts 1-5 seconds. Have you noticed similar behavior? Any ideas on how to fix it?

Thanks for your great work making this possible on 3.11 👍

@gnab
Copy link
Owner

gnab commented Dec 2, 2013

I've struggled a bit lately with the network adapter either disconnecting or roaming to a '(none)' network.

My initial rewrite didn't properly create the proc entries as the void *data pointer never got handled properly. I expect this to be an error, and I've just checked in a fix about an hour ago that uses create_proc_data instead of just create_proc.

I'm not sure if this last change fixes the disconnect/roaming issue that I've had, but at least the adapter has been perfectly stable for the last hour or so.

I suggest you fetch the latest version and try it out.

@lagerspetz
Copy link
Author

Hey,
I have the latest version and still experience freezes. You can see what I mean with for example the free Zero-K game, by just spectating a random match. If all units freeze into place for 1-5 seconds, while the FPS counter is still high and scrolling etc local stuff works fine, you have seen it.

If you want to help me debug, the game is pretty easy to install, you just install mono and a couple other things, wget the binary and just run it with mono.
http://zero-k.info/Wiki/Download

@lagerspetz
Copy link
Author

So I tried to pin this on the powersaving, so I disabled that in the Makefile. No effect. Any ideas?

@gnab
Copy link
Owner

gnab commented Dec 6, 2013

No, I'm sorry. Mine is pretty stable since the latest update I did.

If you happen to have a separate wifi adapter (like a built-in one if you're on a laptop) you could check to see if it's disabled or not. I bought my D-Link adapter solely because my built-in one got completely unstable after upgrading to Ubuntu 13.10.

@lagerspetz
Copy link
Author

I don't, sorry.
This is in a desktop machine, and I have gone through 2 PCI-E cards that were useless, with abysmal connectivity and connectivity dropping all the time. This one stays connected, but just suffers freezes every couple minutes. There is one room between the router and the desktop machine. Laptops in the same location are 100% fine. I may end up changing the DWA-171 to a different USB adapter (ugh, again) if I can't get this one fixed.

@gnab
Copy link
Owner

gnab commented Dec 9, 2013

I'm still experiencing some issues with my adapter, specifically it drops out after several hours, and then tries reconnecting for quite some time without any success.

I just found another version of the rtl8812au driver that I'm currently trying out. So far, it works great, but I won't know for sure until I've used it for some days.

https://github.com/abperiasamy/rtl8812AU_8821AU_linux

@lagerspetz
Copy link
Author

Thanks for the link. I get the same freezes with that one. I am looking for another adapter, but no luck yet.

@eastmanenator
Copy link

@lagerspetz I have the same problem... :(

@Braklet
Copy link

Braklet commented Sep 17, 2014

The RTL8812AU drivers sourced from ASUS and D-Link don't have the RTL8821A extensions required for my Edimax EW-7811UTC, limiting me to the v4.2.2 drivers with associated performance problems.

Realtek was kind enough to send me a v4.3.0 driver with the RTL8821A extensions which works great with my EW-7811UTC:

https://github.com/Braklet/rtl8811AU_rtl8821A-linux/

I've only used this with 3.2.0-x kernels, and have not made any changes to support 3.10.x and up. Feel free to fork & make those modifications.

@diederikdehaas
Copy link

@Braklet Why is the repo named rtl8811AU_rtl8821A and not rtl8812AU_rtl8821A? (8811 vs 8812)
And how did you figure out you needed the RTL8821A extensions?
I have a Edimax EW-7822UAC (AC-1200) and tried the driver compiled from this repo and it seems to work. The download on the Edimax site for your card and my card are exactly the same, except for the addition of your device ID.

@Braklet
Copy link

Braklet commented Sep 17, 2014

@diederikdehaas

The repo is named rtl8811AU* because that is the name of the archive Realtek provided.

The rtl8812 driver supports multiple related devices in the Realtek stable, including RTL8811AU, RTL8812AU, certainly others.

Here is more information about your device:

https://wikidevi.com/wiki/Edimax_EW-7822UAC (note WI1 chip1: Realtek RTL8812AU)

Here is more information about my device:

https://wikidevi.com/wiki/Edimax_EW-7811UTC (note WI1 chip1: Realtek RTL8811AU)

I figured out that I needed the RTL8821A extensions after experiencing configuration and compile errors attempting to build the ASUS and D-Link sourced RTL8812 drivers. You can tell whether your driver has these extensions by checking for a hal/OUTSRC/rtl8821a directory and contents.

The driver compiled from the gnab and abperiasamy repositories as well as Edimax's download exhibit significant performance problems on my systems, lots of delays and dropouts. The driver compiled from the repository I just added performs better.

Feel free to try it, or not, as you see fit.

@diederikdehaas
Copy link

Thanks 👍

@severidaniele
Copy link

@Braklet thanks for sharing v4.3.0 driver and the additional information.

I have an Edimax EW-7811UTC too and I have built the driver v4.2.2.using this repository (gnab) because I am using Linux kernel 3.12 on a Raspberry Pi.
On my device the blue LED is constantly blinking even when there is network activity and I am not sure this is the normal expected behavior.

I am wondering if the LED is doing the same on your EW-7811UTC when using v4.3.0 driver?

Thanks.

@Braklet
Copy link

Braklet commented Sep 18, 2014

@etabeta888 - My AC600 blue LED constantly blinks, but I have a constantly active WLAN so haven't considered it a problem.

@severidaniele
Copy link

@Braklet OK I understand. Thanks for the additional info.

rhadman pushed a commit to rhadman/rtl8812au that referenced this issue May 31, 2015
…e, but by processing 0, baseIndex5G is left unset and can cause a kernel OOPS that looks like:

May 31 23:01:08 rpi2 kernel: [ 2045.373494] RTL871X: No power limit table of the specified band 1, bandwidth 0, ratesection 0, group 0, rf path 0
May 31 23:01:08 rpi2 kernel: [ 2045.373505] RTL871X: use other value 63
May 31 23:01:08 rpi2 kernel: [ 2045.373541] Unable to handle kernel paging request at virtual address 65ce6217
May 31 23:01:08 rpi2 kernel: [ 2045.375393] pgd = b2f98000
May 31 23:01:08 rpi2 kernel: [ 2045.375482] [65ce6217] *pgd=00000000
May 31 23:01:08 rpi2 kernel: [ 2045.375611] Internal error: Oops: 5 [gnab#1] PREEMPT SMP ARM
May 31 23:01:08 rpi2 kernel: [ 2045.376276] CPU: 0 PID: 1311 Comm: wpa_supplicant Tainted: G           O   3.18.0-23-rpi2 gnab#24-Ubuntu
May 31 23:01:08 rpi2 kernel: [ 2045.376450] task: b2fe5280 ti: a198c000 task.ti: a198c000
May 31 23:01:08 rpi2 kernel: [ 2045.376724] PC is at PHY_ConvertPowerLimitToPowerIndex+0x540/0x9e8 [8812au]
May 31 23:01:08 rpi2 kernel: [ 2045.376984] LR is at PHY_ConvertPowerLimitToPowerIndex+0x528/0x9e8 [8812au]
May 31 23:01:08 rpi2 kernel: [ 2045.377121] pc : [<7f15493c>]    lr : [<7f154924>]    psr: 600b0013
May 31 23:01:08 rpi2 kernel: [ 2045.377121] sp : a198dd10  ip : 80913bf0  fp : 0000345c
May 31 23:01:08 rpi2 kernel: [ 2045.377330] r10: 00000000  r9 : 00000000  r8 : b2e72872
May 31 23:01:08 rpi2 kernel: [ 2045.377433] r7 : b2e70000  r6 : b2e72872  r5 : 00000000  r4 : 0000003f
May 31 23:01:08 rpi2 kernel: [ 2045.377559] r3 : 65ce6215  r2 : 00000001  r1 : b2e73450  r0 : 0000001b
May 31 23:01:08 rpi2 kernel: [ 2045.377685] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment user
May 31 23:01:08 rpi2 kernel: [ 2045.377821] Control: 10c5387d  Table: 32f9806a  DAC: 00000015
May 31 23:01:08 rpi2 kernel: [ 2045.377933] Process wpa_supplicant (pid: 1311, stack limit = 0xa198c238)
rhadman pushed a commit to rhadman/rtl8812au that referenced this issue Jun 1, 2015
…le value using an unassigned variable.

Oops occurred on an ubuntu 14.04 LTS Raspberry Pi 2 system and looks like:
Jan  1 00:00:10 rpi2 kernel: [    7.069349] Unable to handle kernel paging request at virtual address 31fab237
Jan  1 00:00:10 rpi2 kernel: [    7.078173] pgd = b2b2c000
Jan  1 00:00:10 rpi2 kernel: [    7.087284] [31fab237] *pgd=00000000
Jan  1 00:00:10 rpi2 kernel: [    7.096431] Internal error: Oops: 5 [gnab#1] PREEMPT SMP ARM
Jan  1 00:00:10 rpi2 kernel: [    7.105658] Modules linked in: cfg80211 rfkill 8812au(O) 8192cu cp210x usbserial snd_soc_bcm2708_i2s regmap_mmio snd_soc_core bcm2708_rng spi_bcm2708 snd_com
press i2c_bcm2708 snd_bcm2835 snd_pcm_dmaengine snd_pcm snd_timer snd fuse
Jan  1 00:00:10 rpi2 kernel: [    7.135771] CPU: 0 PID: 508 Comm: wpa_supplicant Tainted: G           O   3.18.0-23-rpi2 gnab#24-Ubuntu
Jan  1 00:00:10 rpi2 kernel: [    7.157466] task: b2adb840 ti: b1832000 task.ti: b1832000
Jan  1 00:00:10 rpi2 kernel: [    7.168907] PC is at PHY_ConvertPowerLimitToPowerIndex+0x540/0x9e8 [8812au]
Jan  1 00:00:10 rpi2 kernel: [    7.180810] LR is at PHY_ConvertPowerLimitToPowerIndex+0x528/0x9e8 [8812au]
Jan  1 00:00:10 rpi2 kernel: [    7.192390] sp : b1833d10  ip : 80913bf0  fp : 0000345c
Jan  1 00:00:10 rpi2 kernel: [    7.215709] r10: 00000000  r9 : 00000000  r8 : b2d7a872
Jan  1 00:00:10 rpi2 kernel: [    7.227351] r7 : b2d78000  r6 : b2d7a872  r5 : 00000000  r4 : 0000003f
...
Jan  1 00:00:10 rpi2 kernel: [    8.107493] [<7f21193c>] (PHY_ConvertPowerLimitToPowerIndex [8812au]) from [<7f212004>] (PHY_BBConfig8812+0x220/0x378 [8812au])
Jan  1 00:00:10 rpi2 kernel: [    8.144381] [<7f212004>] (PHY_BBConfig8812 [8812au]) from [<7f21b374>] (rtl8812au_hal_init+0xcf8/0x101c [8812au])
Jan  1 00:00:10 rpi2 kernel: [    8.181404] [<7f21b374>] (rtl8812au_hal_init [8812au]) from [<7f202a60>] (rtw_hal_init+0x20/0x9c [8812au])
Jan  1 00:00:10 rpi2 kernel: [    8.218490] [<7f202a60>] (rtw_hal_init [8812au]) from [<7f1f051c>] (_netdev_open+0x74/0x2fc [8812au])
Jan  1 00:00:10 rpi2 kernel: [    8.255699] [<7f1f051c>] (_netdev_open [8812au]) from [<7f1f07c4>] (netdev_open+0x20/0x38 [8812au])
Jan  1 00:00:10 rpi2 kernel: [    8.292954] [<7f1f07c4>] (netdev_open [8812au]) from [<804e93ac>] (__dev_open+0xb8/0x130)
Jan  1 00:00:10 rpi2 kernel: [    8.330207] [<804e93ac>] (__dev_open) from [<804e9648>] (__dev_change_flags+0x94/0x158)
Jan  1 00:00:10 rpi2 kernel: [    8.367540] [<804e9648>] (__dev_change_flags) from [<804e9724>] (dev_change_flags+0x18/0x48)
Jan  1 00:00:10 rpi2 kernel: [    8.405013] [<804e9724>] (dev_change_flags) from [<8054f7b4>] (devinet_ioctl+0x6ac/0x784)
Jan  1 00:00:10 rpi2 kernel: [    8.442641] [<8054f7b4>] (devinet_ioctl) from [<804cd3b0>] (sock_ioctl+0x1c8/0x294)
Jan  1 00:00:10 rpi2 kernel: [    8.480357] [<804cd3b0>] (sock_ioctl) from [<8015ad10>] (do_vfs_ioctl+0x3f0/0x5b4)
Jan  1 00:00:10 rpi2 kernel: [    8.518159] [<8015ad10>] (do_vfs_ioctl) from [<8015af40>] (SyS_ioctl+0x6c/0x7c)
Jan  1 00:00:10 rpi2 kernel: [    8.537149] [<8015af40>] (SyS_ioctl) from [<8000ef20>] (ret_fast_syscall+0x0/0x48)
Jan  1 00:00:10 rpi2 kernel: [    8.574745] Code: 059d103c 00873001 02833db7 02833005 (05d33002)
Jan  1 00:00:10 rpi2 kernel: [    8.594684] ---[ end trace 6b00f140650346ba ]---
gnab pushed a commit that referenced this issue Aug 17, 2015
Add USB ID for Buffalo WI-U2-433DM
@deicider
Copy link

I am sorry to raise old discussion. I'he just bought an adapter based on the subject IC. I have freezes exactly every 2 minutes which last for 4-5 seconds. Hasn't this been fixed? What information can I provide to help with investigating the issue?

LanderU added a commit to erlerobot/rtl8812au that referenced this issue Mar 8, 2018
This kernel module works under 4.4.50-v7+ gnab#1 SMP PREEMPT
LanderU added a commit to erlerobot/rtl8812au that referenced this issue Mar 20, 2018
This module works under 4.4.50+ gnab#1 PREEMPT
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

7 participants