On Wed, Dec 05, 2018 at 05:12:06PM +0800, kbuild test robot wrote: > tree: git://git.cmpxchg.org/linux-mmotm.git master > head: 7072a0ce81c613d27563eed5425727d1d8791f58 > commit: e3e68607541c60671eb3499a2c064d2f71626da4 [128/293] mm: stall movable allocations until kswapd progresses during serious external fragmentation event > config: c6x-evmc6678_defconfig (attached as .config) > compiler: c6x-elf-gcc (GCC) 8.1.0 > reproduce: > wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross > chmod +x ~/bin/make.cross > git checkout e3e68607541c60671eb3499a2c064d2f71626da4 > # save the attached .config to linux build tree > GCC_VERSION=8.1.0 make.cross ARCH=c6x > > All errors (new ones prefixed by >>): > This appears to be some sort of glitch in Andrew's tree. It works in mmots and is broken in mmotm. The problem is that with mmotm, the fragment_stall_order_sysctl_handler handler has moved below sysctl_min_slab_ratio_sysctl_handler instead of below watermark_boost_factor_sysctl_handler where it belongs. Now, while this could be fixed, in this specific instance I would prefer the patch be dropped entirely because there are some potential downsides that are potentially distracting and the supporting data is not strong enough too justify the potential downsides. Andrew? -- Mel Gorman SUSE Labs
