02:07 <KGB-1> https://tests.reproducible-builds.org/openwrt/openwrt_x86.html has been updated. (100.0% images and 99.9% packages reproducible in our current test framework.)

06:10 <SpectreDev_01> I need help, I'm so confused with numbers I have no clue on how to hardcode partitions

06:57 <djfe> mtd?

06:57 <djfe> this page helps calculating with hex

06:57 <djfe> https://www.calculator.net/hex-calculator.html

06:58 <djfe> what is the mtd output of logread (contains starting offsets) and what is the output of

06:58 <djfe> cat /proc/mtd

06:58 <djfe> ^ contains the size of the partitions

06:59 <djfe> both together are the two values you need for your dts

06:59 <djfe> erasesize and endoffset are irrelevant and can be ignored

08:20 <SpectreDev_01> djfe: yep it's mtd

08:23 <SpectreDev_01> https://pastebin.com/YKC0yx6X

08:54 <djfe> that's the size. please send me a logread

08:54 <djfe> the except of the mtd section would be enough, but a full logread also works

08:55 <djfe> enter "logread" 1min after boot and send me the output pls

09:15 <KGB-1> https://tests.reproducible-builds.org/openwrt/openwrt_ath79.html has been updated. (98.7% images and 99.9% packages reproducible in our current test framework.)

09:35 <SpectreDev_01> djfe: https://pastebin.com/QueDw840

09:39 <djfe> 0:APPSBLENV would be: <0xd00000 0x80000> then

09:40 <djfe> leading zeros between 0x and d00000 for example can be ignored

09:40 <djfe> I might take a look later and create it for you

09:40 <SpectreDev_01> Ok

09:40 <SpectreDev_01> Thanks

09:40 <djfe> but my suggestion: create the partition table yourself and I'll review it

09:40 <djfe> you could start with it now then.

09:41 <SpectreDev_01> Will do that, better way to learn

09:41 <djfe> if you have questions, ask here or via pm

09:41 <SpectreDev_01> Ok, thanks again

09:42 <djfe> how this could look

09:42 <djfe> https://github.com/openwrt/openwrt/blob/36746893ac690796bf7aad00a793ff4bf739d066/target/linux/ipq40xx/files/arch/arm/boot/dts/qcom-ipq4019-whw03v2.dts#L315

09:43 <djfe> or maybe better this one (same cpu unlike the first link)

09:43 <djfe> https://github.com/openwrt/openwrt/blob/36746893ac690796bf7aad00a793ff4bf739d066/target/linux/qualcommax/files/arch/arm64/boot/dts/qcom/ipq8070-cax1800.dts

09:44 <djfe> * https://github.com/openwrt/openwrt/blob/36746893ac690796bf7aad00a793ff4bf739d066/target/linux/qualcommax/files/arch/arm64/boot/dts/qcom/ipq8070-cax1800.dts#L122

09:56 <robimarko> The easiest way is to copy/paste the table from dmesg

09:56 <robimarko> Then you just copy/paste the from-to to Google and it gives you the size

09:56 <robimarko> First value is your start and thats it, you have everything you need

10:19 <\x> robimarko: re enabling m3 dump and still having nss disabled was no go, it crashes after some hours

10:19 <\x> nss region needs to be enabled still

10:19 <\x> so yeah, thats 8MB

10:20 <\x> though it aint too bad, with fw mode 2, theres around 440MB usable

10:20 <\x> 430MB*

10:28 <robimarko> Thats weird as nothing should be accesing that memory

10:29 <robimarko> Except if they did not pull yet another TZ/QSEE messing with everything move

11:29 <mrkiko> robimarko: sorry for re-asking you again ... I forgot / wante3d to understand things better. In daily use, at "normal" runtime, when are QSEE functions invoked? e.g.: I know about reboot, but then?

11:29 <mrkiko> robimarko: I mean - what are the services QSEE offers to the Linux OS and what are the services the Linux OS should mandatorily use to work

11:35 <robimarko> QSEE is the QCA version of ARM TrustZone so its running all of the time

11:35 <robimarko> For example all of the SCM calls are implemented in it

11:36 <robimarko> On ARMv8 QCA devices it implements PSCI as well

11:36 <robimarko> Probably a lot more that we have no clue about

11:36 <colo> what is SCM, and what is PSCI? (i.e., what do they expand to?)

11:37 <robimarko> QCOM Secure Channel Manager (SCM)

11:37 <robimarko> Arm Power State Coordination Interface (PSCI)

11:38 <robimarko> SCM being a fully QCA thing while PSCI is standardised

11:41 <colo> thanks

11:41 <colo> I am convinced all this "trusted" computing stuff will ruin Free Software in the long run :(

11:43 <robimarko> If they cared about Free Software we wouldnt be at the point of having FW blobs

12:38 <colo> yeah, sure, I agree. but there is a difference between the industry "not caring", and the industry making NOT using Free Software (or not being able to exercise your user rights even if you use some Free Software that was compiled and digitally signed by some bigco provider) mandatory. which is what all the recent remote attestation schemes will accomplish. it's TCPA/Palladium all over again, but this time, it's really coming.

13:08 <SpectreDev_01> robimarko: https://github.com/SpectreDev-007/openwrt/commit/dd2bb477c71f56913d4fc581db1d86a84815a41a

13:08 <SpectreDev_01> I think that's correct, I finally understood it

16:37 <SpectreDev_01> robimarko: 0.718941] qcom_nand.0: error parsing ofpart partition /soc/nand-controller@79b0000/nand@0/partition@0

16:41 <robimarko> SpectreDev_01: Those partitions need to be a subnode of partitions node

16:41 <robimarko> And you are missing the compatible for it as well as address and size cells

16:41 <tmn505> ^^ https://www.kernel.org/doc/Documentation/devicetree/bindings/mtd/partitions/fixed-partitions.yaml

16:41 <SpectreDev_01> Ah that's what I'm missing

16:58 <SpectreDev_01> robimarko: do I have todo caplock like stock does or is it fine without it

16:59 <SpectreDev_01> For the partition names

17:02 <robimarko> Both are fine, lowercase is more common so you can reuse the calibration extraction and like

17:02 <robimarko> Personally, I prefer lowercase

17:03 <SpectreDev_01> I'll stick to lowercase looks nicer

17:08 <SpectreDev_01> robimarko: uhmm usb doesn't seem to work

17:15 <robimarko> SpectreDev_01: Well, did you enable the PHY-s and USB nodes?

17:16 <SpectreDev_01> Ah no PHY is spamming in logs, ill add both

17:19 <SpectreDev_01> robimarko: it has those weird values <0x180> which u dunno what Todo with

17:19 <SpectreDev_01> Do I just multiply them

17:21 <SpectreDev_01> That'd just end in 0

17:30 <SpectreDev_01> Frankly I'm quite clueless on adding any missing nodes

17:41 <SpectreDev_01> I think I might have worked it out

17:49 <SpectreDev_01> Or not didn't work

17:52 <SpectreDev_01> Yeah I don't know what to do

18:32 <robimarko> Please just look at other devices

18:32 <robimarko> Its quite clear how to enable USB on them

18:32 <robimarko> 301W being one of them with both USB ports

18:50 <SpectreDev_01> Ok

18:55 <SpectreDev_01> I think I did it

21:09 <SpectreDev_01> robimarko: that worked but it doesn't seem to create kernel and alt_kernel partition even though I added it

21:12 <KGB-2> https://tests.reproducible-builds.org/openwrt/openwrt_lantiq.html has been updated. (95.9% images and 99.8% packages reproducible in our current test framework.)

21:36 <SpectreDev_01> Also for uboot-envtools everyone else uses appblenv but Linksys also has u_env which the other Openwrt Linksys devices are using apparently so stock can be flashed via Openwrt

21:36 <SpectreDev_01> I guess I should use u_enc

21:36 <SpectreDev_01> *u_env

21:37 <\x> i have the MR7350, yes you can flash from stock firmware

21:37 <\x> theres that small "CA" hiding there sometimes so you can access manual firmware upgrade

21:37 <\x> its on the webui

21:38 <SpectreDev_01> No I mean flash stock fw over the Openwrt one

21:38 <SpectreDev_01> But alt_kernel and kernel aren't being created is another issue

21:38 <\x> well you can if you have the partition dumps

21:38 <\x> mtd write

21:38 <\x> oh