1
1
mirror of https://github.com/NixOS/mobile-nixos.git synced 2024-09-19 07:47:20 +03:00
Commit Graph

352 Commits

Author SHA1 Message Date
Zhaofeng Li
38814f75ea module/lib: Actually pass specialArgs to eval 2021-07-13 17:22:58 -07:00
Samuel Dionne-Riel
365a93eedb
Merge pull request #369 from samueldr-wip/device/motorola-rav/init
motorola-rav: init
2021-06-30 15:52:00 -04:00
Niklas Hambüchen
993f7e9c01 device-notes: Fix missing spaces in front of -A 2021-06-30 03:46:47 +02:00
Samuel Dionne-Riel
a9b90deb9a hardware-qualcomm: Add SM6125 2021-06-29 21:37:04 -04:00
Samuel Dionne-Riel
9139a1c9a1
Merge pull request #359 from telent/beatrix
motorola-potter: init
2021-06-28 19:16:29 -04:00
Daniel Barlow
632fa13be0 fix touchscreen without need of quirk
mdss_fb_release_all runs when the display blanks between the splash
screen and launching the main menu (in "hello" example). If ref_cnt is
zero, it calls fb_notifier_call_chain with the payload of
FB_BLANK_POWERDOWN. The synaptics_dsx_i2c touchscreen driver is
subscribed to this notifier, and disables touch.

This is all fine except that the corresponding opposite operation - to
notify the subscribers with FB_BLANK_UNBLANK when the display
is *opened* - is not done anywhere. Add a call to mdss_fb_open
to do this whenever it's calling mdss_fb_blank_sub

For context, this behaviour was introduced in

   commit a12be25aec3fd648f2a9c4c27e88d470bed41f67
   Author: parasur <parasur@motorola.com>
   Date:   Mon Nov 7 09:45:54 2016 +0800
   a12be25aec (diff-eab94fa9ea3e15183d23b93fbaf4e08bc137ae4ebbdc589fc106355b753d1b6a)

but probably (speculating) goes unnoticed in Androd because the
framebuffer is never closed in Android unless something crashes, at
which point most people probably reboot anyway
2021-06-27 23:24:52 +01:00
Daniel Barlow
163c436f1d activate the touchscreen at boot
This is needed for potter: I found the fix in
https://wiki.postmarketos.org/wiki/Troubleshooting#Touch_screen_device_present_but_not_outputting_any_touch_events

which suggests that other motorola devices have the same issue, so I
implemented it as a quirk. I don't yet have any insight on how to find
the correct device path portably
2021-06-27 23:24:52 +01:00
Daniel Barlow
7aecb4656b use systemdMinimal for udev in initrd
saves 649216 bytes on Potter
2021-06-27 23:24:52 +01:00
Samuel Dionne-Riel
f4441658ee
Merge pull request #311 from samueldr-wip/wip/amazon-austin
amazon-austin: Init
2021-06-19 01:46:28 -04:00
Samuel Dionne-Riel
365e174f5c hardware-mediatek: Add MT8127 2021-06-18 21:26:06 -04:00
Samuel Dionne-Riel
af8260e632
Merge pull request #356 from lheckemann/btrfs
imageBuilder: add btrfs support
2021-06-18 17:45:19 -04:00
Linus Heckemann
10000edcda imageBuilder: add btrfs support 2021-05-21 17:45:32 +02:00
Pacman99
27683f2d52 module/lib: try to pass specialArgs to eval 2021-05-06 16:30:46 -07:00
Samuel Dionne-Riel
b07dc19062 cross-workarounds: Don't apply overlay on pkgsBuildBuild
As described near the change, this would cause qemu to become an overly
costly build.
2021-03-07 17:52:04 -05:00
Samuel Dionne-Riel
280d872897 make-flashable-zip: Add to overlay
There is no reason to stow this in the system-type module. Other than
making it harder for external uses.
2021-02-28 15:53:12 -05:00
Samuel Dionne-Riel
3941294673 system-types/android: Review default output name
`android-device` was a poor name choice. Especially now that we have
flashable zips, which are equally as usable as fastboot-flashable
outputs.
2021-02-28 15:53:12 -05:00
Samuel Dionne-Riel
d17e09c075 system-types/android: Review config and flashing scripts 2021-02-28 15:53:12 -05:00
Samuel Dionne-Riel
3980354c21 system-types/android: Update documentation 2021-02-28 15:53:12 -05:00
Samuel Dionne-Riel
939dd6e9dc android-flashable-zip-binaries: Always build static 2021-02-28 15:53:12 -05:00
Samuel Dionne-Riel
0b15ab04cf system-types/android: Add flashable zips 2021-02-28 15:53:12 -05:00
Samuel Dionne-Riel
98e2ab907f
Merge pull request #305 from samueldr-wip/feature/lvgui-pan
lvgui: "pan" framebuffers
2021-02-23 20:59:51 -05:00
Samuel Dionne-Riel
90167b0b64 cross-workarounds: Work around libselinux dependency issue
It fails with:

```
/nix/store/m7080pw0ryjk0jhljp55rq1hd2qy8gki-python3-3.8.6/include/python3.8/pyport.h:726:2: error: #error "LONG_BIT definition appears wrong for platform (bad gcc/glibc config?)."
  726 | #error "LONG_BIT definition appears wrong for platform (bad gcc/glibc config?)."

[...]

builder for '/nix/store/cdysjc47jc4z49r25xzwf2sazqd4w4rg-libselinux-3.0-armv7l-unknown-linux-gnueabihf.drv' failed with exit code 2
```
2021-02-20 17:23:01 -05:00
Samuel Dionne-Riel
e5ca35f45e quirks/framebuffer: fb-refresher is enabled separately for stage-1/0 2021-02-20 16:41:23 -05:00
Samuel Dionne-Riel
62297b2e10
Merge pull request #300 from samueldr-wip/feature/stage-0-finishing-touches
Finish stage-0 boot features
2021-02-15 16:19:44 -05:00
Samuel Dionne-Riel
34ad416e77 stage-1/usb-gadget: Fix error message error
`#constantize` does not exist on `String`.

This would only have been an issue in the improbable cause that a daemon
is not configured for an FFS gadget.
2021-02-11 19:52:38 -05:00
Samuel Dionne-Riel
ff73bb6ccf stage-0: Allow adding nodes and props to forward 2021-02-11 18:15:34 -05:00
Samuel Dionne-Riel
3014bcb858 stage-0: Add fdt-forward 2021-02-11 18:15:34 -05:00
Samuel Dionne-Riel
c81e858ea5 initrd: Add kexectool in stage-0 module 2021-02-11 18:15:34 -05:00
Samuel Dionne-Riel
c85712a663 system-build: Add DTB mapping 2021-02-11 18:14:30 -05:00
Samuel Dionne-Riel
826315aa11 system-build: Init module adding details to toplevel
For now, device-name... but see following commit.
2021-02-11 18:14:30 -05:00
Samuel Dionne-Riel
ed93aa4678 modules/initrd-kernel: Fix mocked kernelPackages
.ci/instantiate-all.nix wouldn't pass otherwise.
2021-02-02 05:37:31 -05:00
Samuel Dionne-Riel
cc77670be1 modules: Remove kernel nixos-disintegration
The solution is, ugh, to add `...` to allow additional params on kernel
builder call derivations.
2021-02-02 01:58:41 -05:00
Samuel Dionne-Riel
25b2d8fbba modules/stage-0: Use composeConfig 2021-02-02 01:58:41 -05:00
Samuel Dionne-Riel
84b7e5eead modules/recovery: Use composeConfig 2021-02-02 01:58:41 -05:00
Samuel Dionne-Riel
34b033330f Replace evalConfig with config.lib.mobile-nixos.composeConfig 2021-02-02 01:58:41 -05:00
Samuel Dionne-Riel
b57495c363 release-tools: Remove _mobile-nixos.nixpkgsPath
`modulesPath` is enough for our needs.
2021-02-02 01:58:41 -05:00
Edward Amsden
c507b39099 Make the package set a parameter instead of using <nixpkgs> everywhere.
Co-authored-by: Samuel Dionne-Riel <samuel@dionne-riel.com>
Co-authored-by: Edward Amsden <edward@blackriversoft.com>
2021-02-02 01:58:41 -05:00
Samuel Dionne-Riel
a96c2b845c boot/init: Add support for FunctionFS on GadgetFS
FunctionFS on GadgetFS requires this to happen in order:

 - (1) preparing the gadget
 - (2) preparing the function
 - (3) mounting the functionfs
 - (4) starting the daemon
 - (5) enabling the device

Doing any of these steps in the wrong order *will* fail. You cannot
mount functionfs if the function name hasn't been allocated yet. You
cannot start the daemon without functionfs being mounted. You cannot
enable the device without the daemon running on the USB endpoints.

This is why the GadgetFS stuff here *has* to know about mounting the
filesystem and starting the daemon.

In fact, the generic library to be written to handle USB gadget mode
**will** have to have that knowledge!

The added TODOs are to be done when extracting as a standalone library.
2021-01-27 22:16:55 -05:00
Samuel Dionne-Riel
6f6940226a modules/adb: ADB works just fine on any usb gadget
(Though it may require fiddling with udev perms)
2021-01-27 22:16:55 -05:00
Samuel Dionne-Riel
4d7bd722a6 modules/adb: Decouple implementation details
The service and mount points are more of an implementation detail of the
GadgetFS handling library and not an implementation detail of the init
library.

In turn, this means that we are decoupling this from the init, follow-up
on the next commit...
2021-01-27 19:09:07 -05:00
Samuel Dionne-Riel
5ba41dd963 boot/init: Fix shellOnFail
- It was running too late.
 - Wasn't actually using `shellOnFail` config
2021-01-27 19:09:07 -05:00
Samuel Dionne-Riel
10fc8bc0bb initrd-shell: Fix job control 2021-01-24 23:38:05 -05:00
Samuel Dionne-Riel
3a48ed5fa1 quirks/wifi: Add quirk to disable MAC address randomization 2021-01-10 21:08:09 -05:00
Samuel Dionne-Riel
62d00f17fc system-types/android: Add odin-specific device-notes fragment 2021-01-09 19:23:21 -05:00
Samuel Dionne-Riel
7381c783a7 system-types/android: Rename fastboot device notes fragment 2021-01-09 19:23:20 -05:00
Samuel Dionne-Riel
2b674af725 system-types/android: Choose the appropriate device-notes fragment 2021-01-09 19:23:20 -05:00
Samuel Dionne-Riel
f5fbeb1b76 system-types/uefi: Add device-notes fragment 2021-01-09 19:23:20 -05:00
Samuel Dionne-Riel
7a4ddb41ff system-types/u-boot: Add device-notes fragment 2021-01-09 19:23:20 -05:00
Samuel Dionne-Riel
ec7a2d120d devices-metadata: Export documentation config
Since *anyway* this is mainly intended to be consuemd by the
documentation.
2021-01-09 19:23:20 -05:00
Samuel Dionne-Riel
8d2af65cc4 modules/documentation: Init 2021-01-09 19:23:20 -05:00