1204218 - Generalize init.sh to be used on all devices

Status: RESOLVED FIXED

(Firefox OS Graveyard :: GonkIntegration, defect)

Description

[Adam Farden [:adfad666]]

User Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/44.0.2403.155 Safari/537.36

Comment 1

[Adam Farden [:adfad666]]

Attachment: init.sh PR

Comment 2

[Adam Farden [:adfad666]]

Attachment: Shinano-KK PR

Comment 3

[Adam Farden [:adfad666]]

This will allow our range of sony-aosp-l devices to also make use of init.sh, and any other devices that might want to use it too.

Comment 4

[Adam Farden [:adfad666]]

Actually it will probably be easier to make a sony-aosp-l branch on the init_sh repo so we don't have to touch the KK builds.

We only need this on L builds so it makes sense, we won't be adding any more devices to KK branches.

Comment 5

[Dave Hylands [:dhylands]]

The Shinano-KK PR needs to be pushed first, so you should edit the description to have it be Part 1, and make the init.sh generalization be Part 2. If they're applied in the wrong order then the tree is broken as far as doing bisections is concerned.

The only thing I noticed was that the original script used /dev/input/event2 with a major of 13 and a minor of 67, and it looks like 13/67 actually corresponds to /dev/input/event3 so I just wanted to clarify that portion of the change.

Comment 6

[Dave Hylands [:dhylands]]

Comment on attachment 8660297 [details] [review]
Shinano-KK PR

r=me if you rename this to start wih "Part 1" (or "Pt 1") and clarification on the major/minor number.

Comment 7

[Dave Hylands [:dhylands]]

Comment on attachment 8660297 [details] [review]
Shinano-KK PR

Let's try this again with the actual r+

r=me if you rename this to start wih "Part 1" (or "Pt 1") and clarification on the major/minor number.

Comment 8

[Dave Hylands [:dhylands]]

Comment on attachment 8660296 [details] [review]
init.sh PR

r=me if you rename this to start with "Part 2" (or "Pt 2")

Comment 9

[Adam Farden [:adfad666]]

It should be [ 64 + event# ] so assuming event2 is correct it should be 66.

"event1 c 13 65" and "event3 c 13 67" are working correctly on my devices that use those respective events.

I don't have the leo/aries devices to test so Alex can you confirm the "event2 c 13 66" is working correctly?

Comment 10

[Adam Farden [:adfad666]]

If nobody has time to test on KK can we just get a sony-aosp-l branch on the init_sh repo? There's no reason to delay this any longer for L builds.

Comment 11

[Adam Farden [:adfad666]]

Attachment: init.sh L PR

Comment 12

[Adam Farden [:adfad666]]

Attachment: Manifest PR

Comment 13

[Adam Farden [:adfad666]]

Dave, you've already reviewed this, but I created a new PR just for the sony-aosp-l branch, so we don't need to test it on KK (unfortunately nobody seems to have time to do it). The new branch we created for the elf header lets us neatly sidestep that issue, and we can merge this right away for our L builds.

Comment 14

[Dave Hylands [:dhylands]]

Comment on attachment 8666056 [details] [review]
init.sh L PR

I'll need to test under KK anyways in order to finish landing busybox as source part (bug 1208288)

Comment 15

[:gerard-majax]

https://github.com/mozilla-b2g/b2g-manifest/commit/a5a85f73fb97befa04653d52a9e4b36c520488f7
https://github.com/mozilla-b2g/init_sh/commit/ca67d963b2cbfaf98c1ae887fc6b523d15263864