Catalin Marinas catalin.marinas@arm.com writes:
On Thu, Sep 05, 2019 at 04:39:38PM +0100, Mark Brown wrote:
On Wed, Sep 04, 2019 at 05:15:53PM -0700, Kevin Hilman wrote:
[ + Steven Liu who donated this board to my kernelCI lab ]
Also adding Christoph since this was bisected to his commit and Catalin and Will since this was an architecture change.
Thanks.
Bisected down to this commit[1], full bisect log here[2]. It didn't revert cleanly on top of next-20190904, so I didn't get any further.
[1] 419e2f1838819e954071dfa1d1f820ab3386ada1 is the first bad commit commit 419e2f1838819e954071dfa1d1f820ab3386ada1 Author: Christoph Hellwig hch@lst.de Date: Mon Aug 26 09:03:44 2019 +0200
I stared at this commit and the ones around it for a while but couldn't see anything obvious. AFAICT this commit only replaces some macros, no functional change.
This is a an arm64 mediatek SoC (MT7622) so hopefully the folks on linux-mediatek can take a closer look and see what's going on here.
If someone can suggest how I can enable an earlcon on this platform, I'd be happy to give that a try to see if we can get any kernel output before it hangs.
Kevin