Re: [PATCH 0/8] kernel-yocto: consolidated pull request

[Richard Purdie <richard.purdie@linuxfoundation.org>]

On Mon, 2017-12-04 at 10:39 -0500, Bruce Ashfield wrote:
> Here is the latest set of kernel updates and bug fixes to the kernel-
> yocto
> build.
> 
> The 4.9 and 4.12 updates are routine -stable updates. I realize we
> are
> still fighting some boot/hang issues with 4.12, but I didn't see any
> new issues with the 4.12 updates, so I included them in this pull
> request.
> 
> I've also included Saul's in-tree defconfig fixes + one additional
> patch
> of my own.
> 
> I also bump linux-yocto-dev to 4.15-rcX, which brought in some build
> issues due to x86 ORC_UNWINDER changes, and ssl cert generation. I've
> added the new dependencies directly in the linux-yocto-dev recipe,
> but
> they will very likely have to move lower into the stack over time.
> 
> I have some additional, structural changes that are coming shortly,
> but I'm traveling for the next 5 days and will send them when I
> return.

I tried this series, minus the -dev kernel change in -next and we see:

https://autobuilder.yocto.io/builders/nightly-oe-selftest/builds/676/steps/Running%20oe-selftest/logs/stdio

The duplicate logging bug makes it hard to read but the key takeaway is:

2017-12-09 10:49:53,667 - oe-selftest - INFO - Ran 317 tests in 14099.939s
2017-12-09 10:49:53,667 - oe-selftest - INFO -  (failures=16, skipped=1)

I can reproduce easily with:

oe-selftest -r runqemu.RunqemuTests.test_boot_deploy

which fails locally too. If I revert "linux-yocto/4.12: update to
v4.12.16" it works.

The test checks if it can login over the serial port so I'm guessing
some serial driver either got turned off or regressed?

Cheers,

Richard