You are not logged in.
I have been experiencing this issue since Kernel 6.1.1. Arch is installed in a second M2 SSD in my laptop. Could the issue be somehow related to that reported in the link below (in the Manjaro Forum)?
https://forum.manjaro.org/t/lenovo-t14- … d/145832/9
Last edited by cribari (2023-08-12 01:50:24)
Linux, street photography and Baroque music enthusiast.
Offline
Manjaro link superficially looks like https://wiki.archlinux.org/title/Solid_ … leshooting
Offline
Unrelated to latest posts, but I have an update. Recently I switched to a custom-built desktop PC and I cloned my root drive onto here so it has almost the same software config, but I haven't been able to replicate this issue, I'm on the latest zen kernel, let's see if I get the freeze later on.
Offline
I have the same problem with AMD RX 6600 and two monitors, one HDMI and the other DP. I tried with kernel 5.15, 6.1 and 6.4, but there is not case.
Offline
I think I have the same problem. It does not happen all the time here. And when I try manually to suspend to sleep. it works.
But sometimes, when gnome sheel do the automatic suspend, it fails.
I started https://bbs.archlinux.org/viewtopic.php?id=287610
It works for me with the lts 6.1.39 kernel.
Optiplex 7060, i5-8500, iGPU UHD 630, 2 22" screens on displayport
Last edited by solstice (2023-08-12 22:26:06)
Offline
I think I have the same problem.
Doesn't sound like the same problem to me, this problem results in the system freezing before it enters sleep, it never powers down. If it's powering down for you then it's a different issue.
Offline
no. I thought it powers down, but in fact not. because the power led is still on, and does not blink like expected.
I was just wrong. The last lines in my log are:
systemd[1]: Reached target Sleep.
systemd[1]: Starting System Suspend...
systemd-sleep[16220]: Entering sleep state 'suspend'...
kernel: PM: suspend entry (deep)
Offline
Okay, it's the same issue then, welcome to the club.
Offline
I have the same issue.
It seems that it occurs only when or more often warp-cli connect(cloudflare-warp-bin) connection is active.
My cpu is i5-10400f, it does not have iGPU,
gpu is rx 580
Offline
Same issue.
When it happens mine Caps Lock indicator starts blinking.
CPU: AMD Ryzen 5 4600H with Radeon Graphics
GPU: NVIDIA GeForce GTX 1650 (Laptop) - disabled by driver vfio-pci
iGPU is used
Offline
I can confirm that using x11 this problem is not present.
Offline
I can confirm that using x11 this problem is not present.
I'm having this issue on X11.
Offline
I have also been bisecting this and I just finished.
# first bad commit: [b006c439d58db625318bf2207feabf847510a8a6] hwrng: core - start hwrng kthread also for untrusted sources
https://git.kernel.org/pub/scm/linux/ke … 847510a8a6
I am not sure what this commit has to do with sleep (s3). Maybe someone else who has bisected can confirm?
My symptoms are same as OP. Started with kernel 6.1 where system freezes intermittently on suspend. Power light and fans remain on. Power button is required to do a hard reset. On reboot there is nothing in the log that would indicate a cause. There is no kernel panic (caps lock blinking). X11 has nothing to do with it. I think those are separate issues.
Soon I will probably post a bug at bugzilla.kernel.org. The above commit reverts without conflicts on later kernel versions.
MSI GS40 6QE
CPU: i7-6700HQ (skylake)
GPU: GeForce GTX 970M
Last edited by chriscjsus (2023-08-26 11:37:01)
Offline
# first bad commit: [b006c439d58db625318bf2207feabf847510a8a6] hwrng: core - start hwrng kthread also for untrusted sources
https://git.kernel.org/pub/scm/linux/ke … 7787b55906 since 6.4.11
Online
My system is not AMD and the commit I posted does not mention TPM.
https://git.kernel.org/pub/scm/linux/ke … 847510a8a6
This is from before 6.1 release. But I will wait and see if 6.4.11 fixes it.
Last edited by chriscjsus (2023-08-26 06:06:18)
Offline
Do you have journals of both conditions to maybe show what rng is brought into the mix w/ the bad commit?
Can you trigger the freeze w/o the bad commit but "random.trust_cpu=on" in the kernel commandline?
Offline
Awesome work chriscjsus! Hopefully we are one step closer to solving the mystery.
Offline
@seth
Arch has CONFIG_RANDOM_TRUST_CPU=y so random.trust_cpu is on by default.
Last edited by chriscjsus (2023-08-26 08:15:50)
Offline
Do you have journals of both conditions to maybe show what rng is brought into the mix w/ the bad commit?
What exactly would I search for in the journal? Searching for "random" doesn't give any random sources. Searching for "rng" gives the rngd daemon log:
systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
rngd[898]: Initializing available sources
rngd[898]: [hwrng ]: Initialized
rngd[898]: [rdrand]: Enabling RDSEED rng support
rngd[898]: [rdrand]: Initialized
rngd[898]: [jitter]: JITTER timeout set to 5 sec
rngd[898]: [jitter]: Initializing AES buffer
rngd[898]: [jitter]: Enabling JITTER rng support
rngd[898]: [jitter]: Initialized
rngd[898]: [pkcs11]: No pkcs11 slots available
rngd[898]: [pkcs11]: Initialization Failed
rngd[898]: [rtlsdr]: Initialization Failed
This is the same between both conditions.
Offline
Why are you running https://wiki.archlinux.org/title/Rng-tools at all?
Offline
My system is not AMD and the commit I posted does not mention TPM.
https://git.kernel.org/pub/scm/linux/ke … 29bcd616da which was the initial fix attempt that only disabled some AMD fTPM has more details of the issue.
Does the system have a TPM? If not please provide a list of loaded modules or devices so we can try and locate which could be feeding the RNG and blocking.
Does disabling rngd have any effect?
Edit:
If the device registers itself with devm_hwrng_register and is in tree it should be one of https://elixir.bootlin.com/linux/v6.4.1 … g_register
Last edited by loqs (2023-08-26 12:50:10)
Online
I updated the kernel from 6.4.8 to 6.4.11 on August 19th, and I haven't experienced this issue for a week since then.
It might be a coincidence, but you all might want to give it a try.
Offline
Why are you running https://wiki.archlinux.org/title/Rng-tools at all?
I have been running it for a few years to increase /dev/random speed. I will disable it for now to see if that helps this issue.
Offline
chriscjsus wrote:My system is not AMD and the commit I posted does not mention TPM.
https://git.kernel.org/pub/scm/linux/ke … 29bcd616da which was the initial fix attempt that only disabled some AMD fTPM has more details of the issue.
Does the system have a TPM? If not please provide a list of loaded modules or devices so we can try and locate which could be feeding the RNG and blocking.
Does disabling rngd have any effect?
Edit:
If the device registers itself with devm_hwrng_register and is in tree it should be one of https://elixir.bootlin.com/linux/v6.4.1 … g_register
My system has an intel firmware TPM. I have disabled rngd to see if that works. I will also try 6.4.11.
Offline
Unfortunately, the issue has resurfaced just now. Kernel version is 6.4.12.
Offline