[Bug 1163771] New: [RPI4] WLAN could stop running
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Bug ID: 1163771 Summary: [RPI4] WLAN could stop running Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: aarch64 OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Kernel Assignee: kernel-maintainers@forge.provo.novell.com Reporter: clin@suse.com QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- When the WLAN interface was enabled by WPA-PSK("home") via YaST, it was working for a while and then it suddenly died afterward. This symptom still exists although I have switched to different WLANs/SSIDs. In this case, I downgraded the wpa_supplicant package from v2.9 to v2.6 because somehow v2.9 can't work on ARM port images [pi2/3/4] OS: OpenSUSE-Tumbleweed-20200201 Kernel: 5.4.14-1-default Here are some kernel logs: ------------ [ 391.274635] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 707.177694] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 1023.290662] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 1339.201986] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 1655.322140] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 1971.328281] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 2287.341410] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 2603.251291] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 2919.261729] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 3235.275586] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 3551.287301] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 3867.297475] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready [ 4177.623093] ieee80211 phy0: brcmf_cfg80211_scan: Connecting: status (3) [ 4177.624656] ieee80211 phy0: brcmf_cfg80211_scan: Connecting: status (3) [ 4177.644403] ieee80211 phy0: brcmf_cfg80211_scan: Connecting: status (3) [ 4177.773375] ------------[ cut here ]------------ [ 4177.773501] WARNING: CPU: 0 PID: 5750 at net/wireless/sme.c:756 __cfg80211_connect_result+0x374/0x400 [cfg80211] [ 4177.773505] Modules linked in: joydev af_packet iscsi_ibft iscsi_boot_sysfs overlay br_netfilter bridge stp llc btsdio bluetooth ecdh_generic ecc xhci_pci brcmfmac xhci_ hcd cpufreq_dt nls_iso8859_1 nls_cp437 brcmutil vfat fat cfg80211 rfkill raspberrypi_cpufreq raspberrypi_hwmon crct10dif_ce bcm2835_wdt pcie_brcmstb uio_pdrv_genirq leds_gp io uio mmc_block dwc2 broadcom bcm_phy_lib udc_core usbcore mdio_bcm_unimac sdhci_iproc sdhci_pltfm sdhci mmc_core clk_raspberrypi i2c_bcm2835 gpio_raspberrypi_exp bcm2835_ dma genet phy_generic gpio_regulator fixed sunrpc dm_mirror dm_region_hash dm_log dm_mod sg efivarfs [ 4177.773603] CPU: 0 PID: 5750 Comm: kworker/u8:0 Not tainted 5.4.14-1-default #1 openSUSE Tumbleweed (unreleased) [ 4177.773608] Hardware name: raspberrypi rpi/rpi, BIOS 2020.01 01/30/2020 [ 4177.773693] Workqueue: cfg80211 cfg80211_event_work [cfg80211] [ 4177.773703] pstate: 60000005 (nZCv daif -PAN -UAO) [ 4177.773784] pc : __cfg80211_connect_result+0x374/0x400 [cfg80211] [ 4177.773864] lr : __cfg80211_connect_result+0x200/0x400 [cfg80211] [ 4177.773868] sp : ffff8000106abcb0 [ 4177.773873] x29: ffff8000106abcb0 x28: 0000000000000000 [ 4177.773882] x27: dead000000000100 x26: dead000000000122 [ 4177.773889] x25: 0000000000000000 x24: ffff0000ee8e80e0 [ 4177.773896] x23: ffff8000106abcf8 x22: ffff0000f69a9ec0 [ 4177.773902] x21: ffff0000f0262000 x20: ffff0000ee8e8008 [ 4177.773909] x19: ffff0000eac0a218 x18: 0000000000000000 [ 4177.773915] x17: 0000000000000000 x16: ffffc86e83ef7e38 [ 4177.773921] x15: 0000000000000000 x14: 00040f1004087f00 [ 4177.773928] x13: 0000000000000000 x12: 0000000000000028 [ 4177.773934] x11: 0101010101010101 x10: 7f7f7f7f7f7f7f7f [ 4177.773941] x9 : fefefefefefefeff x8 : ffff0000f40d4044 [ 4177.773947] x7 : 0000000000000000 x6 : ffff0000f44a00f8 [ 4177.773953] x5 : ffff8000106abbc0 x4 : 0000000000000001 [ 4177.773960] x3 : ffffc86e1dc2a9a0 x2 : 0000000000000000 [ 4177.773966] x1 : 0000000000000000 x0 : 0000000000000000 [ 4177.773973] Call trace: [ 4177.774053] __cfg80211_connect_result+0x374/0x400 [cfg80211] [ 4177.774131] cfg80211_process_wdev_events+0x1dc/0x2b8 [cfg80211] [ 4177.774208] cfg80211_process_rdev_events+0x40/0x98 [cfg80211] [ 4177.774286] cfg80211_event_work+0x24/0x38 [cfg80211] [ 4177.774300] process_one_work+0x1d0/0x468 [ 4177.774307] worker_thread+0x150/0x4f0 [ 4177.774317] kthread+0xec/0x118 [ 4177.774326] ret_from_fork+0x10/0x18 [ 4177.774331] ---[ end trace 4be4a82a659b2a9e ]--- -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Chester Lin <clin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |afaerber@suse.com, | |clin@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Chester Lin <clin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |mbrugger@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Chester Lin <clin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |ptesarik@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Chester Lin <clin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |yousaf.kaukab@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Chester Lin <clin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jlee@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Chester Lin <clin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |nsaenzjulienne@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Joey Lee <jlee@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |acho@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c1 Joey Lee <jlee@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(acho@suse.com) --- Comment #1 from Joey Lee <jlee@suse.com> --- Hi Al, Could you please help to look at this issue? It's about WiFi. Thanks -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c2 Matthias Brugger <mbrugger@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |freek@opensuse.org, | |kukuk@suse.com --- Comment #2 from Matthias Brugger <mbrugger@suse.com> --- I asked Karol the wpa_supplicant maintainer, unfortunately he has not access to his bugzilla account right now. Here his response: <paste> At a first glance there is (at the very least) an issue with the kernel. wpa_supplicant shouldn't be able to trigger kernel traces. Do you happen to know (and/or could you ask the reporter) whether this is a regression after a recent upgrade of the kernel, or is the issue only there when wpa_supplicant is upgraded (in isolation of other packages). According to the changes file the update has been done back in November (by Tomáš Chvátal) and I don't remember any such bugs back then. </paste> I remember that Freek was mentioning on the mailinglist that wifi only worked with older wpa_supplicant but didn't work with 5 GHz at all: https://lists.opensuse.org/opensuse-arm/2019-12/msg00046.html -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c3 --- Comment #3 from Chester Lin <clin@suse.com> --- (In reply to Matthias Brugger from comment #2)
I asked Karol the wpa_supplicant maintainer, unfortunately he has not access to his bugzilla account right now. Here his response:
<paste> At a first glance there is (at the very least) an issue with the kernel. wpa_supplicant shouldn't be able to trigger kernel traces.
Do you happen to know (and/or could you ask the reporter) whether this is a regression after a recent upgrade of the kernel, or is the issue only there when wpa_supplicant is upgraded (in isolation of other packages). According to the changes file the update has been done back in November (by Tomáš Chvátal) and I don't remember any such bugs back then. </paste>
I remember that Freek was mentioning on the mailinglist that wifi only worked with older wpa_supplicant but didn't work with 5 GHz at all: https://lists.opensuse.org/opensuse-arm/2019-12/msg00046.html
Hi Matthias, Yes! I downgraded the wpa_supplicant to 2.6 based on his feedback. I tried the following two WLANs: 1. My home WLAN (2.4 GHz) 2. Our corp WLAN (5 GHz) Both of them were working for a while until this issue occurred. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Thorsten Kukuk <kukuk@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC|kukuk@suse.com | -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c11 Al Cho <acho@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(chester.lin@suse. | |com) --- Comment #11 from Al Cho <acho@suse.com> --- (In reply to Chester Lin from comment #0)
When the WLAN interface was enabled by WPA-PSK("home") via YaST, it was working for a while and then it suddenly died afterward. This symptom still exists although I have switched to different WLANs/SSIDs.
In this case, I downgraded the wpa_supplicant package from v2.9 to v2.6 because somehow v2.9 can't work on ARM port images [pi2/3/4]
OS: OpenSUSE-Tumbleweed-20200201 Kernel: 5.4.14-1-default
Here are some kernel logs: ------------ [ 391.274635] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 707.177694] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 1023.290662] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 1339.201986] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 1655.322140] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 1971.328281] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 2287.341410] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 2603.251291] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 2919.261729] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 3235.275586] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 3551.287301] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 3867.297475] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 4177.623093] ieee80211 phy0: brcmf_cfg80211_scan: Connecting: status (3)
[ 4177.624656] ieee80211 phy0: brcmf_cfg80211_scan: Connecting: status (3)
[ 4177.644403] ieee80211 phy0: brcmf_cfg80211_scan: Connecting: status (3)
[ 4177.773375] ------------[ cut here ]------------
[ 4177.773501] WARNING: CPU: 0 PID: 5750 at net/wireless/sme.c:756 __cfg80211_connect_result+0x374/0x400 [cfg80211] [ 4177.773505] Modules linked in: joydev af_packet iscsi_ibft iscsi_boot_sysfs overlay br_netfilter bridge stp llc btsdio bluetooth ecdh_generic ecc xhci_pci brcmfmac xhci_ hcd cpufreq_dt nls_iso8859_1 nls_cp437 brcmutil vfat fat cfg80211 rfkill raspberrypi_cpufreq raspberrypi_hwmon crct10dif_ce bcm2835_wdt pcie_brcmstb uio_pdrv_genirq leds_gp io uio mmc_block dwc2 broadcom bcm_phy_lib udc_core usbcore mdio_bcm_unimac sdhci_iproc sdhci_pltfm sdhci mmc_core clk_raspberrypi i2c_bcm2835 gpio_raspberrypi_exp bcm2835_ dma genet phy_generic gpio_regulator fixed sunrpc dm_mirror dm_region_hash dm_log dm_mod sg efivarfs [ 4177.773603] CPU: 0 PID: 5750 Comm: kworker/u8:0 Not tainted 5.4.14-1-default #1 openSUSE Tumbleweed (unreleased) [ 4177.773608] Hardware name: raspberrypi rpi/rpi, BIOS 2020.01 01/30/2020 [ 4177.773693] Workqueue: cfg80211 cfg80211_event_work [cfg80211] [ 4177.773703] pstate: 60000005 (nZCv daif -PAN -UAO) [ 4177.773784] pc : __cfg80211_connect_result+0x374/0x400 [cfg80211] [ 4177.773864] lr : __cfg80211_connect_result+0x200/0x400 [cfg80211] [ 4177.773868] sp : ffff8000106abcb0 [ 4177.773873] x29: ffff8000106abcb0 x28: 0000000000000000 [ 4177.773882] x27: dead000000000100 x26: dead000000000122 [ 4177.773889] x25: 0000000000000000 x24: ffff0000ee8e80e0 [ 4177.773896] x23: ffff8000106abcf8 x22: ffff0000f69a9ec0 [ 4177.773902] x21: ffff0000f0262000 x20: ffff0000ee8e8008 [ 4177.773909] x19: ffff0000eac0a218 x18: 0000000000000000 [ 4177.773915] x17: 0000000000000000 x16: ffffc86e83ef7e38 [ 4177.773921] x15: 0000000000000000 x14: 00040f1004087f00 [ 4177.773928] x13: 0000000000000000 x12: 0000000000000028 [ 4177.773934] x11: 0101010101010101 x10: 7f7f7f7f7f7f7f7f [ 4177.773941] x9 : fefefefefefefeff x8 : ffff0000f40d4044 [ 4177.773947] x7 : 0000000000000000 x6 : ffff0000f44a00f8 [ 4177.773953] x5 : ffff8000106abbc0 x4 : 0000000000000001 [ 4177.773960] x3 : ffffc86e1dc2a9a0 x2 : 0000000000000000 [ 4177.773966] x1 : 0000000000000000 x0 : 0000000000000000 [ 4177.773973] Call trace: [ 4177.774053] __cfg80211_connect_result+0x374/0x400 [cfg80211] [ 4177.774131] cfg80211_process_wdev_events+0x1dc/0x2b8 [cfg80211] [ 4177.774208] cfg80211_process_rdev_events+0x40/0x98 [cfg80211] [ 4177.774286] cfg80211_event_work+0x24/0x38 [cfg80211] [ 4177.774300] process_one_work+0x1d0/0x468 [ 4177.774307] worker_thread+0x150/0x4f0 [ 4177.774317] kthread+0xec/0x118 [ 4177.774326] ret_from_fork+0x10/0x18 [ 4177.774331] ---[ end trace 4be4a82a659b2a9e ]---
Would you please attach the full log for this? only this part is not easy to find the issue. Now I have one RPi4 and connect 5Gz(Our corp WLAN (5 GHz)) and wait for trying reproduce. Please also feel free to add comments with any step what you did when you got this issue. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c12 --- Comment #12 from Chester Lin <chester.lin@suse.com> --- Created attachment 831841 --> http://bugzilla.suse.com/attachment.cgi?id=831841&action=edit wlan_log.txt (In reply to Al Cho from comment #11)
Would you please attach the full log for this? only this part is not easy to find the issue.
Now I have one RPi4 and connect 5Gz(Our corp WLAN (5 GHz)) and wait for trying reproduce.
Please also feel free to add comments with any step what you did when you got this issue.
Sorry that I didn't reserve all at that time since most of them would just normal boot logs until those unusual logs were present [as I mentioned in comment#1]. I tried to reproduce this issue again in my home-lan (2.4G) with a port image [openSUSE-Tumbleweed-ARM-GNOME-raspberrypi4.aarch64-2020.02.26-Snapshot20200226.raw.xz], and the wpa_supplicant package was still downgraded to 2.6 in order to make WPA-PSK auth work. The test result is as attached. Here are my configuration and steps: ------------------- [Configuration] - Connecting an official 5V power adapter with my rpi4-b in order to make sure that the power source is good enough. - Connecting the serial port with a PL2303 RS232-to-USB converter as a debug console. (Only TXD, RXD and ground pins were connected) [Steps] 1. Setup wlan0 via yast2. 2. Reboot and check if wlan0 is available. 3. ping to 8.8.8.8 until "ping: sendmsg: Network is unreachable" shows up. 4. check dmesg. 5. check wlan0 interface by "ip a" ------------------- This time there was no kernel warning message but wlan was still dead after icmp_seq=250. wlan0 was still there but it showed "link becomes ready" twice when I was capturing this log. Is there any way to enable more debug messages on brcmfmac? Anyway, I will try Corp wifi (5G) afterward. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Chester Lin <chester.lin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(chester.lin@suse. | |com) | -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c15 --- Comment #15 from Freek de Kruijf <freek@opensuse.org> --- I can confirm the loss of the connection via wlan0 using ping to de default gateway in my network using a fresh installation of the XFCE image. This loss was intermittent; a number of successful pings (about 8), silence (about 10 missed pings), again some successful pings, silence,... At a certain moment I got the message the network was no longer available. I also tried the JeOS image, on which the first 150 pings were successful, after which I broke off the test. In both cases I downgraded wpa_supplicant from the repository for Leap 15.2. In both cases the 5.4 GHz signal was too weak to be of use, so the tests were only with 2.4 GHz. I was unable to use the KDE image, because the wireless interface wlan0 did not appear in the system although kernel module brcmfmac was loaded. Bug#1165683. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c16 --- Comment #16 from Chester Lin <chester.lin@suse.com> --- (In reply to Freek de Kruijf from comment #15)
I can confirm the loss of the connection via wlan0 using ping to de default gateway in my network using a fresh installation of the XFCE image. This loss was intermittent; a number of successful pings (about 8), silence (about 10 missed pings), again some successful pings, silence,... At a certain moment I got the message the network was no longer available.
I also tried the JeOS image, on which the first 150 pings were successful, after which I broke off the test.
In both cases I downgraded wpa_supplicant from the repository for Leap 15.2.
In both cases the 5.4 GHz signal was too weak to be of use, so the tests were only with 2.4 GHz.
I was unable to use the KDE image, because the wireless interface wlan0 did not appear in the system although kernel module brcmfmac was loaded. Bug#1165683.
Hi Freek, Did you use wicked to configure wlan0 as well? I just found that NetworkManager works fine on wlan0 without having an abnormal disconnection. Would you like to try it? Since NetworkManager can work I think this issue should not be related to kernel. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c17 --- Comment #17 from Freek de Kruijf <freek@opensuse.org> --- (In reply to Chester Lin from comment #16)
Hi Freek,
Did you use wicked to configure wlan0 as well? I just found that NetworkManager works fine on wlan0 without having an abnormal disconnection. Would you like to try it? Since NetworkManager can work I think this issue should not be related to kernel.
Yes, I did use wicked. I started with a fresh JeOS image. As was to be expected, wlan0 did not come up, so I replaced wpa_supplicant by the one from the 15.2 repository. After that wlan0 did not come up, but iwlist did not show the 2.4 GHz AcessPoint, only 5. I did a reboot and wlan0 did come up. Did an ifdown eth0, ifdown wlan0 and ifup wlan0 and a ping to the router. I let it run for 1000 pings without missing a single one. Did the same with the KDE version, which had a different outcome. Had a ping session of 4500 pings of which about 500 were successful. Often there was no answer on the ping, but also I got Destination not reachable. On another tty I did, in that case,------------------------ a "wicked ifup wlan0" after which ping answers did reappear, but shortly after that I got no answers, which did reappear and vanished again, until Destination not reachable did appear. Changed in the KDE image to NetworkManager. Managed to enable the wifi connection. Started a ping to the router. This ping showed the same behavior as with wicked. It looks like NetworkManager detects the Destination unreachable and restarts the connection. I see about 8 successful pings after which there are no answers on the ping until about another about 8 pings or Destination unreachable. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c18 --- Comment #18 from Nicolas Patricio Saenz Julienne <nsaenzjulienne@suse.com> --- Reproduced issue with RPi3b+ -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Nicolas Patricio Saenz Julienne <nsaenzjulienne@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Summary|[RPI4] WLAN does not work |[RPI] WLAN does not work -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Andreas Färber <afaerber@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- See Also| |http://bugzilla.suse.com/sh | |ow_bug.cgi?id=1167801 -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Dominik Heidler <dheidler@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |dheidler@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 Chester Lin <chester.lin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- See Also| |https://bugzilla.suse.com/s | |how_bug.cgi?id=1166933 -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c19 Al Cho <acho@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(freek@opensuse.or | |g) --- Comment #19 from Al Cho <acho@suse.com> --- Would you please try the latest Tumbleweed? This issue should be marked as a duplicate of bug 1156920. (In reply to Freek de Kruijf from comment #17)
Yes, I did use wicked.
I started with a fresh JeOS image. As was to be expected, wlan0 did not come up, so I replaced wpa_supplicant by the one from the 15.2 repository. After that wlan0 did not come up, but iwlist did not show the 2.4 GHz AcessPoint, only 5. I did a reboot and wlan0 did come up. Did an ifdown eth0, ifdown wlan0 and ifup wlan0 and a ping to the router. I let it run for 1000 pings without missing a single one.
Did the same with the KDE version, which had a different outcome. Had a ping session of 4500 pings of which about 500 were successful. Often there was no answer on the ping, but also I got Destination not reachable. On another tty I did, in that case,------------------------ a "wicked ifup wlan0" after which ping answers did reappear, but shortly after that I got no answers, which did reappear and vanished again, until Destination not reachable did appear.
Changed in the KDE image to NetworkManager. Managed to enable the wifi connection. Started a ping to the router. This ping showed the same behavior as with wicked. It looks like NetworkManager detects the Destination unreachable and restarts the connection. I see about 8 successful pings after which there are no answers on the ping until about another about 8 pings or Destination unreachable.
-- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c20 Freek de Kruijf <freek@opensuse.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(freek@opensuse.or | |g) | --- Comment #20 from Freek de Kruijf <freek@opensuse.org> --- (In reply to Al Cho from comment #19)
Would you please try the latest Tumbleweed?
I used openSUSE-Tumbleweed-ARM-XFCE-raspberrypi4.aarch64-2020.03.25-Build1.36.raw.xz. Used tty1 <Ctrl>+<Alt>+<F1> to configure the system. Used "yast network" to configure the wireless interface. wlan0 did not become active. So still a problem. The configuration via yast showed management via wicked. "iwlist wlan0 scan" did work and showed my access point, however no connection. I created a user and started a XFCE session. There is a system tray icon for the network. Started it and to my surprise - the configuration via yast is still management via wicked - I could configure a wireless connection, which did work. This connection is gone after leaving the XFCE session. This looks as if NetworkManager is also active. I hope this answers your request for more information. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c21 Al Cho <acho@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Assignee|acho@suse.com |wicked-maintainers@suse.de --- Comment #21 from Al Cho <acho@suse.com> --- Sorry for late to reply. (In reply to Freek de Kruijf from comment #20)
(In reply to Al Cho from comment #19)
Would you please try the latest Tumbleweed?
I used openSUSE-Tumbleweed-ARM-XFCE-raspberrypi4.aarch64-2020.03.25-Build1.36.raw. xz. Used tty1 <Ctrl>+<Alt>+<F1> to configure the system. Used "yast network" to configure the wireless interface. wlan0 did not become active. So still a problem. The configuration via yast showed management via wicked. "iwlist wlan0 scan" did work and showed my access point, however no connection.
Yes, you are right. there should be a warning message about Netwrok is currently handled by Networkmananger.
I created a user and started a XFCE session. There is a system tray icon for the network. Started it and to my surprise - the configuration via yast is still management via wicked - I could configure a wireless connection, which did work. This connection is gone after leaving the XFCE session. This looks as if NetworkManager is also active.
I also test the same version on openSUSE Tumbleweed XFCE on x86_64 laptop, there are still work fine.
I hope this answers your request for more information.
Thanks for your information. And I think that seems not a wifi driver issue, so we may ask the wicked maintainer to get help. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1163771 http://bugzilla.suse.com/show_bug.cgi?id=1163771#c22 Petr Tesařík <ptesarik@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |cfamullaconrad@suse.com Flags| |needinfo?(cfamullaconrad@su | |se.com) --- Comment #22 from Petr Tesařík <ptesarik@suse.com> --- I wonder if our SUSE wpa_supplicant maintainer could help with this bug. Clemens? -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1163771 https://bugzilla.suse.com/show_bug.cgi?id=1163771#c23 Marius Tomaschewski <mt@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |mt@suse.com --- Comment #23 from Marius Tomaschewski <mt@suse.com> --- (In reply to Freek de Kruijf from comment #5)
Part of journalctl -b shows: Feb 18 15:44:32 bachtest dbus-daemon[951]: [system] Failed to activate service > 'fi.epitest.hostap.WPASupplicant': timed out (service_start_timeout=25000ms) Feb 18 15:44:32 bachtest wickedd[1023]: ni_wpa_interface_bind(wlp1s0u1u3u3): Operation not permitted
Make sure the image contains a wpa_supplicant which also supports the old dbus api that has been removed from more recent wpa_supplicant versions (>=2.8). We're working on a wicked rewrite to the new API, but due to other tasks, it is still unfinished. See also https://build.opensuse.org/request/show/796014 - Add restore-old-dbus-interface.patch to fix wicked wlan (boo#1156920) - Restore fi.epitest.hostap.WPASupplicant.service (bsc#1167331) Thus this issue seems to be a dup of bsc#1167331. Further, make sure that only one: wicked or NetworkManager is enabled/running. There is a bug in the images, that improperly enable both, wicked + NM, which are both trying to configure the same interfaces, ... -> https://bugzilla.suse.com/show_bug.cgi?id=1141433 AFAIR. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1163771 https://bugzilla.suse.com/show_bug.cgi?id=1163771#c24 Clemens Famulla-Conrad <cfamullaconrad@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(cfamullaconrad@su |needinfo?(chester.lin@suse. |se.com) |com) --- Comment #24 from Clemens Famulla-Conrad <cfamullaconrad@suse.com> --- Yes, as Marius mentioned the latest wpa_supplicant should contain the fix for running with wicked.
was working for a while and then it suddenly died afterward.
Does this apply also if you use the latest wpa_supplicant-2.9-4.2 ? -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1163771 https://bugzilla.suse.com/show_bug.cgi?id=1163771#c25 Chester Lin <chester.lin@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |FIXED Flags|needinfo?(chester.lin@suse. | |com) | --- Comment #25 from Chester Lin <chester.lin@suse.com> --- (In reply to Clemens Famulla-Conrad from comment #24)
Yes, as Marius mentioned the latest wpa_supplicant should contain the fix for running with wicked.
was working for a while and then it suddenly died afterward.
Does this apply also if you use the latest wpa_supplicant-2.9-4.2 ?
I tested a newer version of openSUSE-TW [openSUSE-Tumbleweed-ARM-GNOME-raspberrypi4.aarch64-2020.06.15-Snapshot20200625] and confirmed that the latest wpa_supplicant-2.9-4.2 has fixed this issue. As Clemens has suggested, the NetworkManager should be disabled in order to avoid conflicts while configuring wlan interfaces. Issue closed. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1163771 https://bugzilla.suse.com/show_bug.cgi?id=1163771#c26 Freek de Kruijf <freek@opensuse.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|FIXED |--- --- Comment #26 from Freek de Kruijf <freek@opensuse.org> --- I tested Snapshot20200626 of Tumbleweed KDE and still wlan0 does NOT come up. Before configuring wlan0 I disabled NetworkManager.service and checked if wicked.service was enabled. dmesg did not show anything related to wlan0 nor did journalctl -u wicked{,d} I found the following lines in /var/log/wpa_supplicant.log, which might shed some light: 1593607968.800601: wlan0: CTRL-EVENT-ASSOC-REJECT bssid=00:00:00:00:00:00 status_code=16 1593607968.801013: wlan0: CTRL-EVENT-SSID-TEMP-DISABLED id=0 ssid="FL42MBJ" auth_failures=3 duration=30 reason=CONN_FAILED I triple checked passphrase for this AP in /etc/sysconfig/network/ifcfg-wlan0. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1163771 https://bugzilla.suse.com/show_bug.cgi?id=1163771#c27 --- Comment #27 from Clemens Famulla-Conrad <cfamullaconrad@suse.com> --- I'm not sure what cause your problem. I also tried wicked+wlan on a just updated tumbleweed and didn't got any problems with WPA_PSK config like:
BOOTPROTO='dhcp' STARTMODE='auto' WIRELESS_MODE='Managed' WIRELESS_AUTH_MODE='psk' WIRELESS_ESSID='XYZ' WIRELESS_WPA_PSK='SECRET'
You may also get some wpa_supplicant debugging possibility by enabling the control interface and use `wpa_cli` to play around:
sed -i '/^ExecStart=/c\ExecStart=/usr/sbin/wpa_supplicant -c /etc/wpa_supplicant/wpa_supplicant.conf -u -t -f /var/log/wpa_supplicant.log -O /var/run/wpa_supplicant' /usr/lib/systemd/system/wpa_supplicant.service systemctl daemon-reload systemctl restart wpa_supplicant.service rcwickedd restart
wpa_cli
log_level debug list_networks password X Y status disconnect reconnect
-- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1163771 https://bugzilla.suse.com/show_bug.cgi?id=1163771#c28 Freek de Kruijf <freek@opensuse.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution|--- |FIXED --- Comment #28 from Freek de Kruijf <freek@opensuse.org> --- (In reply to Clemens Famulla-Conrad from comment #27)
I'm not sure what cause your problem. I also tried wicked+wlan on a just updated tumbleweed and didn't got any problems with WPA_PSK config like:
BOOTPROTO='dhcp' STARTMODE='auto' WIRELESS_MODE='Managed' WIRELESS_AUTH_MODE='psk' WIRELESS_ESSID='XYZ' WIRELESS_WPA_PSK='SECRET'
I have the same setup.
You may also get some wpa_supplicant debugging possibility by enabling the control interface and use `wpa_cli` to play around:
sed -i '/^ExecStart=/c\ExecStart=/usr/sbin/wpa_supplicant -c /etc/wpa_supplicant/wpa_supplicant.conf -u -t -f /var/log/wpa_supplicant.log -O /var/run/wpa_supplicant' /usr/lib/systemd/system/wpa_supplicant.service systemctl daemon-reload systemctl restart wpa_supplicant.service rcwickedd restart
wpa_cli
log_level debug list_networks password X Y status disconnect reconnect
Played with it and most of the time the interface did not see my AP when scanning. Also used "iwlist wlan0 scan" and had the same result; only occasionally my AP was listed. My RPi4 is only 8 m from the AP, so I moved it closer to the AP and that gave me a connection. Conclusion the WiFi interface of the the RPi4 is rather weak. Nothing wrong anymore with the software. -- You are receiving this mail because: You are on the CC list for the bug.
participants (2)
-
bugzilla_noreply＠novell.com -
bugzilla_noreply＠suse.com