03:12 <steev> jhovold: i just had my x13 reboot, and i'm not sure what caused it, because it was sitting idle. i just noticed it because i'm watching Exploit Engineering – Attacking the Linux Kernel on a different system and saw the LENOVO logo pop up on it out of the corner of my eye, it hasn't been touched in a couple hours

03:32 <steev> absolutely nothing in the logs

03:32 <steev> 2023-06-08T22:07:37.192105-05:00 wintermute kernel: [ 0.000000] Booting Linux on physical CPU 0x0000000000 [0x410fd4b0]

03:32 <steev> 2023-06-08T21:35:10.938285-05:00 wintermute kernel: [118005.852619] ath11k_pci 0006:01:00.0: msdu_done bit in attention is not set

04:16 <steev> https://youtu.be/9dfISRD6hwk - a talk at offensivecon, they used the windows devkit to do their arm uefi research apparently

06:01 <jhovold> steev: never seen anything like that, any idea of what may have been running when it happened (e.g. is blutooth connected)?

06:02 <steev> nothing connected to bluetooth, nvme ssd connected via usb via apple-c usb/hdmi adapter; apps running were chromium, firefox, alacritty and possibly gnome-terminal

06:03 <steev> hdmi was not connected at all

06:04 <steev> i *might* have been ssh'd in but i don't think i was

06:25 <jhovold> steev: ok, thanks

06:26 <steev> It’s not any sort of new setup, been running this way since forever

06:33 <jhovold> yeah, it's a bit troubling. Nothing new and scary you added on top of my rc5 branch I assume?

08:20 <jhovold> steev: btw, why do you cherry-pick all the patches from my wip branches instead of just rebasing on top?

08:21 <jhovold> doing so would make it a little clearer what comes from my tree and what you have applied on top (e.g. when skimming the log on github)

14:28 <clover[m]> I think I dropped only 40% battery leaving it in suspend overnight. This feels like a bit better power savings than earlier kernels?

16:02 <amstan> only?? that's huge

16:03 <clover[m]> haha

16:11 <konradybcio> clover and that's on which soc?

16:11 <konradybcio> 8280?

16:20 <robclark> amstan: I'm sure still being on clk_ignore_unused pd_ignore_unused isn't really helping things

16:20 <amstan> heh

16:21 <amstan> i'm just amused at double digit percentages loss overnight being okish (looking at my framework laptop too) while we're sitting at over 14 days on chromebooks

16:21 <konradybcio> not as big as lacking interconnect sync state.. however clocks do eat power too

16:21 <konradybcio> thankfully we have sync_state on 8280

16:21 <clover[m]> Snapdragon (TM) 8cx Gen 3 @ 3.0 GHz

16:22 <konradybcio> amstan it would really help if the soc entered proper power collapse, indeed ;)

16:22 <clover[m]> is that the soc or just the cpu

16:22 <amstan> it's all those androidy features of your system doing who knows what when sleeping

16:22 <konradybcio> cpu is like the least meaningful part of snapdragons (and any other soc fwiw) clover

16:22 <amstan> no, i don't want my laptop to stepcount or check tiktok while it's in my bag

16:23 <amstan> well, not that YOUR os does that, but the hardware is very "flexible" to such needs

16:23 <clover[m]> konradybcio: ok, well its a sc8280xp idk if the 'xp' is meaningful or not

16:24 <konradybcio> sc - snapdragon compute, 8 means high end, 2 means 2nd major gen, 80 is sort of product denomination

16:24 <konradybcio> x is not clear, p is likely "no modem"

16:24 <clover[m]> cool :)

16:24 <robclark> sc7180p was the slightly faster variant of sc7180

16:25 <robclark> there was no special designation for modem vs no-modem in the part #

16:25 <robclark> so kinda figured xp == extra performance?

16:32 <clover[m]> and the next one will be sc8380xxp for extra, extra, performance ;)

16:34 <clover[m]> well they will probably switch to Nuvia chips with their next gen so we might see completely different naming?

17:29 <ajhalaney[m]> The fun part to me is (outside of looking at dts inheritance) figuring out which platforms are related.. i.e. sc8280xp, sa8295p, sa8540p

17:29 <ajhalaney[m]> or where sa8775p falls in with other chips

17:29 <ajhalaney[m]> /me shrugs and goes back to not making sense of versioning

18:21 <qzed> sa ist automotive?

18:21 <steev> tak

18:22 <konradybcio> tak means yes btw

18:22 <steev> si

18:22 <qzed> I wasn't sure, but assumed as much xD

18:26 <steev> disclaimer: i'm just assuming yes

18:28 <clover[m]> oui

18:52 <ajhalaney[m]> they're all automotive socs, so I think that's a good guess :)

21:31 <steev> fedora in my car

21:57 <clover[m]> my car runs arch btw

23:55 <bamse> jenneron[m], konradybcio: doesn't sounds like bus scaling, but either another interrupt problem or something similar to the usb iommu issue we're seeing under load on sc8280xp

23:55 <bamse> jenneron[m]: i presume you have 100% load on CPU0 when you run this benchmark?

23:56 <jenneron[m]> bamse: yes

23:59 <bamse> jenneron[m]: i've not booted 8180 in quite a while...but i used to use the lenovo machine as my desktop, and i didn't have that problem back then...