Bug 2067108
Summary: | ath9k WiFi problem >5.16.14 | ||
---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | FSA <fsa> |
Component: | kernel | Assignee: | Kernel Maintainer List <kernel-maint> |
Status: | CLOSED ERRATA | QA Contact: | Fedora Extras Quality Assurance <extras-qa> |
Severity: | unspecified | Docs Contact: | |
Priority: | unspecified | ||
Version: | 35 | CC: | abecker, acaringi, adscvr, airlied, alciregi, andrew, ansari.ibrahim1, awfslot, bskeggs, doctor.mixmid, ego.cordatus, hdegoede, jarodwilson, jeremy, jforbes, jglisse, joaoluiz, jonathan, josef, jr.om, kernel-maint, lgoncalv, linville, ltrpgshard, masami256, mchehab, mroche, nkudriavtsev, olegon, ptalbert, steved, victorffrock, williambader |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | Unspecified | ||
OS: | Unspecified | ||
Whiteboard: | |||
Fixed In Version: | kernel-5.16.18-100.fc34 kernel-5.16.18-200.fc35 | Doc Type: | If docs needed, set a value |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2022-03-30 01:11:26 UTC | Type: | Bug |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: |
Description
FSA
2022-03-23 10:31:39 UTC
So, it seems I have the exact same problem in "Qualcomm Atheros QCA9565 / AR9565" since Monday, and my outputs are pretty much the same except I don't get this message: мар 22 19:49:29 fsa-k75de kernel: ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 dmesg output: ``` [qua mar 23 10:05:44 2022] wlp2s0: authenticate with xx:xx:xx:xx:xx:xx [qua mar 23 10:05:44 2022] wlp2s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [qua mar 23 10:05:44 2022] wlp2s0: authenticated [qua mar 23 10:05:44 2022] wlp2s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [qua mar 23 10:05:44 2022] wlp2s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0xc11 status=0 aid=1) [qua mar 23 10:05:44 2022] wlp2s0: associated [qua mar 23 10:05:44 2022] ath: EEPROM regdomain: 0x804c [qua mar 23 10:05:44 2022] ath: EEPROM indicates we should expect a country code [qua mar 23 10:05:44 2022] ath: doing EEPROM country->regdmn map search [qua mar 23 10:05:44 2022] ath: country maps to regdmn code: 0x3b [qua mar 23 10:05:44 2022] ath: Country alpha2 being used: BR [qua mar 23 10:05:44 2022] ath: Regpair used: 0x3b [qua mar 23 10:05:44 2022] ath: regdomain 0x804c dynamically updated by country element [qua mar 23 10:07:43 2022] i915 0000:00:02.0: [drm] HPD interrupt storm detected on connector DP-1: switching from hotplug detection to polling [qua mar 23 10:12:20 2022] i915 0000:00:02.0: [drm] HPD interrupt storm detected on connector DP-1: switching from hotplug detection to polling ``` To reproduce: enable wireless and begin working. Atheros wireless pauses and freezes after some bit of traffic, restarts. ...problem started in kernels after 5.16.14-200.fc35.x86_64 (which still works fine). lspci-n: 07:00.0 0280: 168c:0032 (rev 01) lspci: 07:00.0 Network controller: Qualcomm Atheros AR9485 Wireless Network Adapter (rev 01) DeviceName: Maryann Subsystem: Hewlett-Packard Company Device 1785 Flags: bus master, fast devsel, latency 0, IRQ 17 Memory at f0300000 (64-bit, non-prefetchable) [size=512K] Expansion ROM at f0380000 [disabled] [size=64K] Capabilities: [40] Power Management version 2 Capabilities: [50] MSI: Enable- Count=1/4 Maskable+ 64bit+ Capabilities: [70] Express Endpoint, MSI 00 Capabilities: [100] Advanced Error Reporting Capabilities: [140] Virtual Channel Capabilities: [160] Device Serial Number 00-00-00-00-00-00-00-00 Kernel driver in use: ath9k Kernel modules: ath9k dmesg: wlo1: authenticate with b0:95:75:13:32:3c [ 632.135381] wlo1: send auth to b0:95:75:13:32:3c (try 1/3) [ 632.244219] wlo1: send auth to b0:95:75:13:32:3c (try 2/3) [ 632.247997] wlo1: authenticated [ 632.249146] wlo1: associate with b0:95:75:13:32:3c (try 1/3) [ 632.256531] wlo1: RX AssocResp from b0:95:75:13:32:3c (capab=0x1031 status=0 aid=2) [ 632.256714] wlo1: associated [ 638.176296] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_S <<--- fail W=0x02000020 DMADBG_7=0x00006100 journalctl: kernel: ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x020 00020 DMADBG_7=0x00006100 There is some kind of related work being done upstream, as noted on kernel mailing list. I have a similar symptom with wifi pausing and ksoftirqd getting a lot of cpu with kernel 5.16.15-101.fc34.x86_64 but not with kernel 5.16.11-100.fc34.x86_64. My laptop is Sony Vaio VPCCB4Q1E. dmesg has "ath9k 0000:02:00.0 wlp2s0: renamed from wlan0" and "usbcore: registered new interface driver ath3k" *** Bug 2068272 has been marked as a duplicate of this bug. *** I had the same issue. Fedora 35 was working fine. On march 22nd wifi began to behave inconsistently immediately after upgrading the linux kernel from 5.16.15 to 5.16.16. I am using a Dell Inspiron 7520 and my wifi adapter is Qualcomm Atheros AR9485. Last night and this morning wifi seems to be working well, so I wonder if this has been fixed? (In reply to Joao Carvalho from comment #5) > Last night and this morning wifi seems to be > working well, so I wonder if this has been fixed? Wifi disconnected a couple of minutes after I posted this, so no, it hasn't been fixed. Problem remains. *** Bug 2068350 has been marked as a duplicate of this bug. *** The same issue with kernels >5.16.14. Network controller: Qualcomm Atheros AR9485 Wireless Network Adapter (rev 01). After about 1 minute the connection hangs. Pings do not go. The reconnection saves for yet another 1 minute. The excerpt of my dmesg: [ 1818.528053] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 [ 1819.580136] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 1819.602005] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 1819.604705] wlp3s0: authenticated [ 1819.606535] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 1819.617234] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 1819.617424] wlp3s0: associated [ 2247.546341] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 [ 2248.598664] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 2248.620287] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 2248.623891] wlp3s0: authenticated [ 2248.625147] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 2248.636073] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 2248.636294] wlp3s0: associated [ 2435.570679] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 [ 2436.620959] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 2436.641688] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 2436.645636] wlp3s0: authenticated [ 2436.647099] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 2436.657830] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 2436.658060] wlp3s0: associated [ 2780.941517] wlp3s0: deauthenticating from xx:xx:xx:xx:xx:xx by local choice (Reason: 3=DEAUTH_LEAVING) [ 2782.508032] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 2782.522917] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 2782.524811] wlp3s0: authenticated [ 2782.525098] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 2782.536162] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 2782.536324] wlp3s0: associated [ 2782.661372] IPv6: ADDRCONF(NETDEV_CHANGE): wlp3s0: link becomes ready [ 2943.697780] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 [ 2944.747908] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 2944.769749] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 2944.777727] wlp3s0: authenticated [ 2944.778259] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 2944.790043] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 2944.790321] wlp3s0: associated [ 3003.746021] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 [ 3004.796532] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 3004.817204] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 3004.824695] wlp3s0: authenticated [ 3004.825411] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 3004.839099] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 3004.839289] wlp3s0: associated [ 3014.840096] wlp3s0: deauthenticating from xx:xx:xx:xx:xx:xx by local choice (Reason: 3=DEAUTH_LEAVING) [ 3014.868889] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 [ 3015.924090] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 3015.945144] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 3015.947781] wlp3s0: authenticated [ 3015.948183] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 3015.959638] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 3015.959816] wlp3s0: associated [ 3018.729934] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 [ 3019.778887] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 3019.800849] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 3019.805622] wlp3s0: authenticated [ 3019.806178] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 3019.818719] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 3019.818909] wlp3s0: associated Thanks in advance. If anyone wants to give https://koji.fedoraproject.org/koji/taskinfo?taskID=84697781 a try, and let me know if it works for you. It is not secure boot signed because it is a scratch build, but it contains a patch that upstream is tossing around. There is a still a considerable amount of discussion on it. Would like to add that I'm running Fedora 36 Workstation Pre-release, and this issue is happening with me after the update from Linux 5.17-rc7 to 5.17 as well, this issue is not just limited to 5.16.x series. lspci: 09:00.0 Network controller: Qualcomm Atheros AR9485 Wireless Network Adapter (rev 01) dmesg: [12193.613742] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006100 [12194.669386] wlp9s0: authenticate with xx:xx:xx:xx:xx:xx [12194.669427] wlp9s0: bad VHT capabilities, disabling VHT [12194.669432] wlp9s0: 80 MHz not supported, disabling VHT [12194.687759] wlp9s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [12194.701654] wlp9s0: authenticated [12194.701806] wlp9s0: associating with AP with corrupt probe response [12194.702140] wlp9s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [12194.708789] wlp9s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [12194.708917] wlp9s0: associated [12194.709086] ath: EEPROM regdomain: 0x8164 [12194.709090] ath: EEPROM indicates we should expect a country code [12194.709091] ath: doing EEPROM country->regdmn map search [12194.709093] ath: country maps to regdmn code: 0x5b [12194.709094] ath: Country alpha2 being used: IN [12194.709096] ath: Regpair used: 0x5b [12194.709098] ath: regdomain 0x8164 dynamically updated by country element Essentially identical to the above reports. Yes, that would make since given that upstream has not accepted a fix yet. (In reply to Justin M. Forbes from comment #9) > If anyone wants to give > https://koji.fedoraproject.org/koji/taskinfo?taskID=84697781 a try, and let > me know if it works for you. It is not secure boot signed because it is a > scratch build, but it contains a patch that upstream is tossing around. > There is a still a considerable amount of discussion on it. Had to recompile it for Fedora 34 using $ rpmbuild -ba -without=debug -without=debuginfo kernel.spec Now the connection stays stable, pings go, but the string "ath: phy0: DMA failed to stop" is still here. [ 319.773168] Bluetooth: BNEP (Ethernet Emulation) ver 1.3 [ 319.773174] Bluetooth: BNEP filters: protocol multicast [ 319.773179] Bluetooth: BNEP socket layer initialized [ 331.965969] usb 1-1.5: Failed to query (GET_DEF) UVC control 13 on unit 1: -32 (exp. 8). [ 367.816339] wlp3s0: authenticate with xx:xx:xx:xx:xx:xx [ 367.837359] wlp3s0: send auth to xx:xx:xx:xx:xx:xx (try 1/3) [ 367.840020] wlp3s0: authenticated [ 367.840645] wlp3s0: associate with xx:xx:xx:xx:xx:xx (try 1/3) [ 367.851709] wlp3s0: RX AssocResp from xx:xx:xx:xx:xx:xx (capab=0x411 status=0 aid=1) [ 367.851909] wlp3s0: associated [ 367.986877] IPv6: ADDRCONF(NETDEV_CHANGE): wlp3s0: link becomes ready [ 791.983102] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x0000a400 [ 818.670005] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000028 AR_DIAG_SW=0x02000020 DMADBG_7=0x0000a400 [ 936.491263] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000028 AR_DIAG_SW=0x02000020 DMADBG_7=0x0000a400 (In reply to Justin M. Forbes from comment #9) This fixes my, Qualcomm Atheros QCA9565 / AR9565 Wireless Network Adapter - many thanks. Before this patch, when streaming videos, my Acer Aspire E1-570 laptop's log repeatedly contained the message wpa_supplicant[959]: wlp3s0: CTRL-EVENT-BEACON-LOSS when using any of the 3 following kernel versions 5.16.15-201.fc35.x86_64 5.16.16-200.fc35.x86_64 5.16.17-200.fc35.x86_64 The patch in the 5.16.17-201.fc35.x86_64 kernel, fixes the above ath9k driver issue for my laptop. Now I don't have to revert back to kernel version 5.16.14-200.fc35.x86_64, to get perfect Wi-Fi access. FEDORA-2022-eb323bcd80 has been submitted as an update to Fedora 35. https://bodhi.fedoraproject.org/updates/FEDORA-2022-eb323bcd80 FEDORA-2022-8e3ac65667 has been submitted as an update to Fedora 34. https://bodhi.fedoraproject.org/updates/FEDORA-2022-8e3ac65667 FEDORA-2022-eb323bcd80 has been pushed to the Fedora 35 testing repository. Soon you'll be able to install the update with the following command: `sudo dnf upgrade --enablerepo=updates-testing --advisory=FEDORA-2022-eb323bcd80` You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2022-eb323bcd80 See also https://fedoraproject.org/wiki/QA:Updates_Testing for more information on how to test updates. FEDORA-2022-8e3ac65667 has been pushed to the Fedora 34 testing repository. Soon you'll be able to install the update with the following command: `sudo dnf upgrade --enablerepo=updates-testing --advisory=FEDORA-2022-8e3ac65667` You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2022-8e3ac65667 See also https://fedoraproject.org/wiki/QA:Updates_Testing for more information on how to test updates. This bug seems to be fixed for me since updating to Linux 5.17.1 kernel as well. $ uname -a Linux fedora 5.16.18-100.fc34.x86_64 #1 SMP PREEMPT Mon Mar 28 14:46:06 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux The string "ath: phy0: DMA failed to stop" is no more. $ dmesg | grep ath [ 10.436355] systemd[1]: Listening on multipathd control socket. [ 19.277669] ath: phy0: ASPM enabled: 0x43 [ 19.277678] ath: EEPROM regdomain: 0x60 [ 19.277679] ath: EEPROM indicates we should expect a direct regpair map [ 19.277682] ath: Country alpha2 being used: 00 [ 19.277683] ath: Regpair used: 0x60 [ 19.533424] ath9k 0000:03:00.0 wlp3s0: renamed from wlan0 [ 21.764876] usbcore: registered new interface driver ath3k Thanks again. (In reply to Fedora Update System from comment #14) Thanks for back porting the fix into Fedora from mainline. FEDORA-2022-8e3ac65667 has been pushed to the Fedora 34 stable repository. If problem still persists, please make note of it in this bug report. FEDORA-2022-eb323bcd80 has been pushed to the Fedora 35 stable repository. If problem still persists, please make note of it in this bug report. |