On Thu, May 20, 2021 at 11:22 AM Rong Chen <rong.a.chen@xxxxxxxxx> wrote: > On 5/20/21 4:28 AM, Arnd Bergmann wrote: > > On Wed, May 19, 2021 at 9:48 PM kernel test robot <lkp@xxxxxxxxx> wrote: > >> tree: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.10.y > >> head: 689e89aee55c565fe90fcdf8a7e53f2f976c5946 > >> commit: 5f2f616343b1d62d26f24316fbfc107ecf0983a9 [3189/5262] linux/compiler-clang.h: define HAVE_BUILTIN_BSWAP* > >> config: mips-randconfig-r015-20210519 (attached as .config) > >> compiler: clang version 13.0.0 (https://github.com/llvm/llvm-project 8e93d10633d751a3e9169bf9fa68326925ffa097) > >> reproduce (this is a W=1 build): > >> wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross > >> chmod +x ~/bin/make.cross > >> # install mips cross compiling tool for clang build > >> # apt-get install binutils-mips-linux-gnu > >> # https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git/commit/?id=5f2f616343b1d62d26f24316fbfc107ecf0983a9 > >> git remote add linux-stable-rc https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git > >> git fetch --no-tags linux-stable-rc linux-5.10.y > >> git checkout 5f2f616343b1d62d26f24316fbfc107ecf0983a9 > >> # save the attached .config to linux build tree > >> COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=mips > >> > >> If you fix the issue, kindly add following tag as appropriate > >> Reported-by: kernel test robot <lkp@xxxxxxxxx> > >> > >> All warnings (new ones prefixed by >>): > >> > > I analyzed another bug like this today, and I think this is again just > > CONFIG_UBSAN_ALIGNMENT=y > > on mips/clang causing tons of these warnings. > > > > I don't think this should be considered a regression, but we may want > > to change either the > > kernel or the 0day bot from test building this combination of options. > > > > Arnd > > > > Hi Arnd, > > We have disabled CONFIG_UBSAN_ALIGNMENT for mips/clang in our kconfigs > to avoid reporting such warnings again. Thanks a lot, that helps. Have you considered always enforcing CONFIG_COMPILE_TEST=y for any randconfig build? That would address this issue along with a number of similar ones. I have sent a number of patches to add COMPILE_TEST dependencies for things that otherwise break in randconfig builds but could still be useful in more specific cases when users know better. Arnd
