| 4 May 2026 |
 [0x4A6F] | * ```
| 18:34:46 |
|  @itoncek:matrix.org left the room. | 18:36:14 |
| [0x4A6F] | * On a framework 13 AMD AI 300 Series (with expansiuon cards - top-left USB-C, bottom-left USB-A):
`xhci_hcd 0000:c3:00.0: Refused to change power state from D0 to D3hot
I switched that USB-A Expansion card into my old 11th-gen Framework 13 without these log messages.
| 18:38:59 |
| [0x4A6F] | * On a framework 13 AMD AI 300 Series (with expansion cards - far-left(1) USB-C, near-left(2) USB-A):
`xhci_hcd 0000:c3:00.0: Refused to change power state from D0 to D3hot
I switched that USB-A Expansion card into my old 11th-gen Framework 13 without these log messages.
- https://knowledgebase.frame.work/en_us/expansion-card-functionality-on-framework-laptop-13-amd-ryzen-ai-300-series-Hy5SfMRs1l
| 18:46:37 |
| [0x4A6F] | * On a framework 13 AMD AI 300 Series (with expansion cards - far-left(1) USB-C, near-left(2) USB-A):
xhci_hcd 0000:c3:00.0: Refused to change power state from D0 to D3hot
I switched that USB-A Expansion card into my old 11th-gen Framework 13 without these log messages.
- https://knowledgebase.frame.work/en_us/expansion-card-functionality-on-framework-laptop-13-amd-ryzen-ai-300-series-Hy5SfMRs1l
| 18:52:15 |
 sudoforge | i think this is a pretty well-documented issue. to explain the error message, what's happening is that the kernel is trying to put the USB controller into a low power state (D3hot) but the hardware is refusing because it's still in an active state (D0). i'm pretty sure this is a power mgmt or firmware issue that FW is aware of (and hopefully working on); perhaps we can get an update here.
a workaround for you would be to adjust your port layout. because the USB-A expansion card has higher power draw in ports 1 and 3 you pretty much want to keep it in ports 3 or 4. no idea if you'll encounter the same issue in port 4 (but presumably, you would). so... i guess at the end of the day, you can either ignore the error, or make the leap into the modern era and converge on USB-C everywhere :P
| 20:58:25 |
| 7 May 2026 |
 Daniel Schaefer (Framework) | Under what conditions does that happen? | 13:02:09 |
|  tristan957 joined the room. | 21:02:19 |
| 8 May 2026 |
 Just Some Dude | Would this be a good place to troubleshoot my framework 13 not being able to fully sleep and not being able to wake up completely from sleep without using systemctl cancel? | 02:03:37 |
| Just Some Dude | I've tried it on KDE, Gnome, and Cosmic. Happens on all 3. | 02:04:28 |
| Just Some Dude | I tried using kernelPackages_latest and the enableRedistributableFirmware options and they didnt do anything | 02:05:41 |
| sudoforge | what shows up in dmesg? | 02:09:43 |
| Just Some Dude | What would I be looking for? | 02:13:54 |
| sudoforge | error messages. | 02:27:35 |
| Just Some Dude | It looks like the only messages are from me authenticating with sudo | 02:28:59 |
| sudoforge | so, there's literally nothing going on before your systemctl cancel? | 02:29:36 |
| sudoforge | and, can you go into more detail about "not being able to fully sleep" and "not being able to wake up completely from sleep"? what are the symptoms you experience? | 02:30:16 |
| Just Some Dude | It doesn't look like it to me but I don't really know how to interpret it | 02:30:35 |
| Just Some Dude | When I close it or use systemctl suspend it just turns the screen off but keeps burning battery like crazy, then when I open it back up the network adapter is dead until I use systemctl cancel | 02:31:59 |
| Just Some Dude | I can't remember where I got the "fix" of interrupting sleep because it was still ongoing it was a while ago | 02:33:12 |
| Just Some Dude | * | 03:31:17 |
| sudoforge | Just Some Dude can you share the output of cat /sys/power/mem_sleep? | 05:25:55 |
| sudoforge | and what mainboard are you using? | 05:28:43 |
| sudoforge | basically... intel and amd have both dropped support for s3 AFAIK, so i don't think the newer chips support it. you should configure hibernation to achieve the low power state that you want. | 05:44:36 |
 Marien Zwart | Iirc you'd still see something in dmesg even if it's only suspending to idle (or similar) | 05:46:56 |
| Daniel Schaefer (Framework) | Last two generations Intel and AMD don't support S3
But support s0ix/modern standby instead
Hibernate is not normally needed unless you want to have the system consume the absolute minimum power
| 06:22:29 |
| Daniel Schaefer (Framework) | Intel core ultra series 1 and 3 | 06:22:50 |
 sophronesis | In reply to @michael.zeagler:matrix.org
Would this be a good place to troubleshoot my framework 13 not being able to fully sleep and not being able to wake up completely from sleep without using systemctl cancel? https://m.youtube.com/watch?v=SewVWYVeZV4
maybe relevant | 06:45:37 |
|  jopejoe1 changed their display name from jopejoe1 (4094@epvpn) to jopejoe1. | 08:42:13 |
| Marien Zwart | after systemctl suspend, on a framework 13 (7040 series), first thing in dmesg on the way towards suspend is PM: suspend entry (s2idle) and the power led slowly blinks once suspended. Been a while since I debugged something like this, but if you're not seeing those things systemd may not be reaching suspend.target (system journal may have clues) | 10:19:04 |
