Most of the initscripts in rcS.d are not being executed

[robintemme]

Summary

We found out that most of the initscripts in /etc/rcS.d are not executed. In fact only S00psplash is executed as far as we could find out. We discovered this because S04udev seemed not to be running.
We could not find the cause though, if we work around this by starting the scripts manually it works like a charm (but this doesn't feel right).

Steps to reproduce

We wanted to build our Yocto-based firmware that already existed for another machine now for the raspberrypi2, so we added this layer and baked an image. We were on dizzy back then. Then we found this bug which did not exist on our Freescale iMX.6 based machine we used before.
We baked a few different images for debugging (rpi-hwup-image, rpi-test-image, core-image-x11) and experienced the bug on every one of them.
We updated the whole project to jethro afterwards, hoping that it would fix the bug. It did not.

Expected result

The initscripts should all have been executed after boot.

Actual result

We put some echo-logs in the different S0X-files and it showed that only S00psplash got executed.

Logs / files etc.

Build configuration

Build Configuration:
BB_VERSION        = "1.28.0"
BUILD_SYS         = "x86_64-linux"
NATIVELSBSTRING   = "Debian-8.0"
TARGET_SYS        = "arm-poky-linux-gnueabi"
MACHINE           = "raspberrypi2"
DISTRO            = "poky"
DISTRO_VERSION    = "2.0.1"
TUNE_FEATURES     = "arm armv7a vfp thumb neon callconvention-hard vfpv4 cortexa7"
TARGET_FPU        = "vfp-vfpv4-neon"
meta              
meta-yocto        = "jethro:5b12268f6e17574999f91628a60e21711cf62ee4"
meta-browser      = "jethro:e114d625d4bd23a52cc1108a45d96ffd8dc0ab7f"
meta-oe           
meta-gnome        
meta-networking   
meta-multimedia   
meta-python       
meta-webserver    = "jethro:dc5634968b270dde250690609f0015f881db81f2"
meta-angstrom     = "angstrom-v2015.12-yocto2.0:6a31b67fbbe9e0ca9f597ab8b248d9fa980b156a"
meta-raspberrypi  = "jethro:f2cff839f52a6e6211337fc45c7c3eabf0fac113"
meta-fsl-arm      = "jethro:35b8b9bd9863de208ab60e33b55f10ee43e2619b"
meta-yocto-bsp    = "jethro:5b12268f6e17574999f91628a60e21711cf62ee4"
meta-qt5          = "jethro:d5536e34ec985c82b621448ab4325e5cbba38560"

I would be happy to provide further information as needed.

[agherzan]

One interesting information would be what is your DISTRO_FEATURES.

[robintemme]

bitbake -e our-image:

DISTRO_FEATURES="alsa argp ext2 irda largefile pcmcia usbgadget usbhost wifi xattr nfs zeroconf pci nfc x11 ipv4 ipv6 libc-backtrace libc-big-macros libc-bsd libc-cxx-tests libc-catgets libc-charsets libc-crypt libc-crypt-ufc libc-db-aliases libc-envz libc-fcvt libc-fmtmsg libc-fstab libc-ftraverse libc-getlogin libc-idn libc-inet-anl libc-libm libc-locales libc-locale-code libc-memusage libc-nis libc-nsswitch libc-rcmd libc-rtld-debug libc-spawn libc-streams libc-sunrpc libc-utmp libc-utmpx libc-wordexp libc-posix-clang-wchar libc-posix-regexp libc-posix-regexp-glibc libc-posix-wchar-io largefile opengl ptest multiarch wayland pulseaudio sysvinit bluez5"

But these seem to be way too many. We really don't need many of them I think.

[robintemme]

We kind of fixed the issue with udev by changing it's position to a later one - S20udev worked for us.
But it still is very strange.

[robintemme]

We didn't come across any other bugs so far, so this seems to be related to udev only. Maybe, to work around the issue, you could add a udev_182.bbappend containing this:

INITSCRIPT_PARAMS_udev = "start 20 S ."

You can close this now if you want.

[agherzan]

Hm. I haven't seen this issue at all. I actually tested by creating a dummy init script in rcS which echoed something. It was started at next reboot and the message was shown.