It is there in that post - /proc/config.gz
Ah, ok. I thought that there would be an editable version somewhere, like /boot.
Its something accomplished at compile time ā¦ so it wouldnt much matter if it were a text file.
Again at the post above the link takes you to the source where it is a text file.
Just updating here that there is no update on the freedesktop thread. Still hopeful a solution is forthcoming. This seems like kind a major downside to AMD chips.
There is some measures of progress. Linux516 now sleeps correctly, but still there are bugs in the wake up phase. From what Iāve read, a new batch of patches made their way into 5.17 kernel before the functionnality freeze, so thereās hope.
What Ryzen are you on? My 5500U does not suspend properly with 5.16.2 or 5.16.3.
Mine is an older 3700U ; I didnāt meant it worked flawlessly, but it feels ābetterā and I have high hopes the remaining bugs are going to be ironed out with the 5.17 kernel.
I noticed you already mentioned iommu=soft, which was the fix for me in grub on my Lenovo tb14 laptop.
Just adding this useful link about that, although im sure you have seen it.
Thanks for the thread, I found information here that was very useful to me.
[edit : WARNING !!! see answer below]
Finally ! The combination of linux517-rc kernel package as released tonight with āiommu=softā command line parameter works and I have a functional s2idle sleep mode !
I believe my computer was released in 2019, so now about all the functions are working after 3 years of hard work by the kernel teamā¦ (Except the fingerprint reader, but this one looks hopeless).
If HP had not messed up with the BIOS, we could have saved 2 years out of those 3, easily.
How did you install the 5.17-rc? Iāve been building kernels over for the freedesktop troubleshooting and could do without building anotherā¦
Itās in the stable updates tonight. #mhwd-kernel -i linux517 should do the trick
1 Like
Sad news after some tests : yes the laptop goes to sleep, yes it wakes up, but unlike last night, I left it in sleep mode for some time (about 1/4 of an hour), and when I grabbed it again, it was almost unbearably hot, to the point of being almost painful to the touch. I woke it up, and the CPU temperature had shot up past 80Ā°C, so I conclude that while the fans are shut off, the CPU itself IS NOT powered down, not even throttled. And the battery lost 6% in those 15 minutes, which is BAD.
So currently, s2idle sleep isnāt viable. Sorry.
Yeah, Iām with you. I tried the 5.17rc with iommu=soft - I got one good suspend, but then it was hard to wake. Back to kernel building.
1 Like