On Thu, Oct 9, 2014 at 7:34 AM, Arnd Bergmann arnd@arndb.de wrote:
On Thursday 09 October 2014 07:24:18 Olof's autobooter wrote:
Full logs are available at http://arm-soc.lixom.net/bootlogs/arm-soc/v3.17-5587-g00a5e5c/
Failed boards:
capri arm-bcm_defconfig : FAILED 0:56.61 capri arm-multi_v7_defconfig : FAILED 1:39.64 colombus arm-sunxi_defconfig : FAILED 1:46.75 odroidu3 arm-exynos_defconfig : FAILED 0:58.88FWIW, this was a test I did using THUMB2_KERNEL. Apparently it broke at least capri and odroidu3, the colombus problem was preexisting.
Talking to Will at LC, it sounded like THUMB2_KERNEL isn't actually much of a desirable feature (on A-class hardware) -- it slows it down and I don't think the size advantages are enough to warrant it.
I also noticed that odroidu3 only gets tested with exynos_defconfig, not multi_v7_defconfig, is that intentional?
When I added the hardware it didn't work on multi_v7 -- there were some bug reports but I didn't go back and check. I'll do a run with multi_v7 when I deal with the farm next time (I have a bunch of hardware sitting here waiting to get added).
-Olof