On Thu, Nov 09, 2017 at 04:14:26AM -0800, kernelci.org bot wrote:
The armada-8040-db has been failing to boot for a little while in -next, though there were some generic arm64 issues which make it hard to tell when it started:
defconfig: armada-8040-db: lab-free-electrons: failing since 5 days (last pass: next-20170929 - first fail: next-20171103)
It gets timeouts talking to MMC and a RCU lockup:
[ 23.071083] INFO: rcu_preempt detected stalls on CPUs/tasks: [ 23.076772] 0-...0: (0 ticks this GP) idle=caa/1/4611686018427387904 softirq=408/408 fqs=2281 [ 23.085505] (detected by 1, t=5252 jiffies, g=-201, c=-202, q=20) [ 23.091714] Task dump for CPU 0: [ 23.094955] kworker/0:1 R running task 0 75 2 0x00000002 [ 23.102048] Workqueue: events deferred_probe_work_func [ 23.107208] Call trace: [ 23.109668] __switch_to+0x90/0xa8 [ 23.113084] 0xffff8001f5c95c38
with substantially less chat in the log from PCI than mainline where it seems to boot OK. Links to some builds for mainline and other useful information here: