01:23 <farchord> <kuruczgy[m]> "For x1e, pd-mapper + adsp fw..." <- Mmmm is pd-mapper something provided by the kernel or something on the side?

01:23 <farchord> Argh sorry about the reply

01:25 <farchord> Oh it's in kernel now, mmm

01:42 <farchord> Yeah i am guessing this has something to do with the problem:

07:11 <SpieringsAE> farchord: pd-mapper should eventually go through the kernel, but to my understanding there are currently issues with that implementation so everyone(?) is using the userspace pd-mapper service for now

07:47 <macc24> there are people using in-kernel pd mapper

08:34 <jhovold> farchord, SpieringsAE, macc24: there are several bugs that have been exposed by the in-kernel pd-mapper so I've kept it disabled in my branches up until now

08:35 <jhovold> two of the issues have now been fixed in mainline, and one has a workaround in the wip branch

08:35 <jhovold> the audio regression remains, but most of the laptops do not have audio enabled yet, and hopefuly we'll get that sorted soon too

08:36 <jhovold> in any case, I decided to reenable the in-kernel pd-mapper in the latest 6.12 wip branch so anyone using that should now be using the in-kernel implementation (and can disable the user space daemon)

08:37 <jhovold> maz: sounds good, enjoy your weekend away from keyboard

08:38 <jhovold> rsalveti: I'm probably the one who has done the most boot cycles of the crd and I've never run into this

08:38 <jhovold> I'm not using the latest firmware however

08:39 <jhovold> haven't heard anything from people I know are using later fw either

08:50 <SpieringsAE> jhovold: nice looks like it is time to compile a new kernel then

09:35 <kuruczgy[m]> Steve Cossette [Farchord]: if you are still having issues, one useful debugging step is to check `/sys/class/remoteproc/*/state` to see if the aDSP is booted. It should report online regardless of whether the pd-mapper is working or not.

09:36 <kuruczgy[m]> userspace pd-mapper requires some *.jsn (JSON) firmware files that have to be manually copied, not sure about kernel pd-mapper. (Right now on my machine I actually have both enabled since I use Johan's branch but haven't disabled the userspace daemon yet, so I am still not sure how well the in-kernel one works.)

10:20 <jhovold> kuruczgy[m]: the json files are only needed with the user-space pd-mapper

10:38 <janrinze> The NVMe of my ARM laptop (HP Chromebook x360 13 , google-dojo) is somewhat slow. I get around 600-700 MB/s whereas under chromeos it is at least 3x that fast. I'm using kernel 6.12.1

10:38 <janrinze> Is there something that can be done to improve that?

10:45 <\[m]> seems no discounts on the t14s 😭

11:33 <farchord> So then the tl;dr: i should prolly still be using jhovold's kernel 😆

11:59 <macc24> janrinze: do you have any other nvme device to check if it's drive or pcie controller? check with lspci -vv if it's running at max link speed and width

13:57 <janrinze> macc24: LnkSta:Speed 8GT/s (downgraded), Width x2 (downgraded) .. this might be the cause.

13:59 <janrinze> unfortunately i cannot check what this would report on Chromeos since I am not running that anymore.

18:42 <JensGlathe[m]> <\[m]> "seems no discounts on the t14s 😭" <- nope. sad

18:53 <steev> but it says 40% off on their website /s

19:49 <JensGlathe[m]> all BS at least here in Germany

19:54 <JensGlathe[m]> my config still costs 2150

19:56 <JensGlathe[m]> list price (increased before, actually) is 2530

20:00 <HdkR> JensGlathe[m]: As per usual, wait for the "sale" so you get the real price. It isn't worth paying effectively double MSRP

20:01 <HdkR> Unless Germany just gets even worse prices :D

20:03 <\[m]> I bought it here in the end https://zakelijk.alternate.nl/Lenovo/ThinkPad-T14s-Gen-6-Snapdragon-X-Elite-(21N1000RMH)-14-inch-Copilotplus-laptop/html/product/1933086 for the qwerty

20:25 <Jasper[m]> <\[m]> "I bought it here in the end..." <- >Zakelijk

20:25 <Jasper[m]> Awww

20:25 <Jasper[m]> Better price than lenovo.com though