Hi Mark,
On 30/07/2019 14:34, Mark Brown wrote:
On Tue, Jul 30, 2019 at 05:17:56AM -0700, kernelci.org bot wrote:
Boot Failures Detected:
arm: oxnas_v6_defconfig: gcc-8: ox820-cloudengines-pogoplug-series-3: 1 failed lab
For some time now -next and mainline have been failing to boot on Pogoplug 3 with the oxnas_v6_defconfig, the kernel seems to start fine but fails to parse the ramdisk it's passed:
08:50:02.086589 <6>[ 7.719854] IP-Config: Complete: 08:50:02.087213 <6>[ 7.723330] device=eth0, hwaddr=0a:a2:89:27:10:1b, ipaddr=10.201.4.144, mask=255.255.0.0, gw=10.201.0.1 08:50:02.087413 <6>[ 7.733409] host=10.201.4.144, domain=, nis-domain=(none) 08:50:02.088056 <6>[ 7.739499] bootserver=10.201.1.1, rootserver=10.201.1.1, rootpath= 08:50:02.088248 <6>[ 7.739504] nameserver0=10.201.1.1 08:50:02.129966 <5>[ 7.752025] RAMDISK: Couldn't find valid RAM disk image starting at 0. 08:50:02.130381 <4>[ 7.759616] List of all partitions: 08:50:02.131333 <4>[ 7.763363] 0100 65536 ram0
Possibly an issue with the ramdisk getting overwritten or something?
Thanks for reporting, it's my suspicion since my multiple bisect runs all point to this merge commit : a318423b61e8 Merge tag 'upstream-5.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rw/ubifs
This merge doesn't introduce notable changes for the oxnas_v6_defconfig, but disabling UBI entirely makes it work again.
Continuing my investigations...
Neil
Full details for today's -next can be seen here: