On 5/11/21 1:47 PM, Andrew Morton wrote: > On Tue, 11 May 2021 12:39:00 -0700 Randy Dunlap wrote: > >> Lots of various mm/ build errors: > > Huh, It Worked For Me (tm). Can you please share the .config? > configs (plural). All of them are randconfigs. All on x86_64. mm/slab_common.c:754:8: error: array index in initializer exceeds array bounds config-r9247 ../include/linux/hugetlb.h:340:30: error: parameter 6 (‘mode’) has incomplete type ../include/linux/hugetlb.h:335:19: error: function declaration isn’t a prototype [-Werror=strict-prototypes] config-r9251 ../include/linux/kernel.h:53:43: error: dereferencing pointer to incomplete type ‘struct module’ ../include/linux/build_bug.h:78:41: error: static assertion failed: "sizeof(typeof_member(struct module, build_id)) == 20" config-r9253 ../include/linux/compiler_types.h:328:38: error: call to ‘__compiletime_assert_253’ declared with attribute error: BUILD_BUG_ON failed: sizeof(struct page) > 80 config-r9254 The randconfig files are attached. I have been wondering lately if I could just tell you/other people the randconfig SEED that was used and then you/other people could just use that SEED to reproduce the randconfig file(s). Are rand() and srand() stable across GCC versions and h/w platforms? config-r9247:KCONFIG_SEED=0x7ABE8042 config-r9251:KCONFIG_SEED=0x67B989AA config-r9253:KCONFIG_SEED=0x7B3D63BE config-r9254:KCONFIG_SEED=0x331557C5 Usage: KCONFIG_SEED=0x331557C5 make [ARCH=x86_64] [O=builddir] randconfig thanks. -- ~Randy