Hi @deadpoolcode,
Are you building the image entirely from clean? (i.e. running make clean first, or having never built in that space before?). In release 9, an incremental build failure was identified with a similar result to what you’re describing. The dm-verity hash to validate the rootfs wasn’t being regenerated for each kernel re-build, so the mount of the rootfs is failing. If this description fits, you can try “make clean” before rebuilding (terrible, I know), or you can apply the following patch to meta-swi layer, which will be in the next release:
diff --git a/meta-swi-mdm9x28/recipes-core/initrdsripts/initramfs-mdminit.bbappend b/meta-swi-mdm9x28/recipes-core/initrdsripts/initramfs-mdminit.bbappend
index ff3a64f..0a7daae 100644
--- a/meta-swi-mdm9x28/recipes-core/initrdsripts/initramfs-mdminit.bbappend
+++ b/meta-swi-mdm9x28/recipes-core/initrdsripts/initramfs-mdminit.bbappend
@@ -9,4 +9,6 @@
install -m 0755 ${WORKDIR}/init.sh ${D}/init
}
+# force do_install task to run for synchronizing roothash
+do_install[nostamp] = "1"
do_install[depends] += "mdm9x28-image-minimal:do_image_complete"
And then independently, there seems to be an issue with swiflash not consistently detecting the device for some customers (and for me too, for what it’s worth). We’re looking into that internally as well. Rebooting the host PC seems to resolve that issue temporarily.
Ryan