On 05/09/2025 19:02, Mark Brown wrote: > On Fri, Sep 05, 2025 at 01:55:53PM -0400, Zi Yan wrote: >> On 5 Sep 2025, at 13:43, Mark Brown wrote: > >>> but the header there is getting ignored AFAICT. Probably the problem is >>> fairly obvious and I'm just being slow - I'm not quite 100% at the >>> minute. > >> prctl_thp_disable.c uses “#include <sys/mman.h>” but asm-generic/mman-common.h >> is included in asm/mman.h. And sys/mman.h gets MADV_COLLAPSE from >> bits/mman-linux.h. Maybe that is why? > > Ah, of course - if glibc is reproducing the kernel definitions rather > than including the kernel headers to get them then that'd do it. > Probably the test needs to locally define the new MADV_COLLAPSE for > glibc compatibility, IME trying to directly include the kernel headers > when glibc doesn't normally use them tends to blow up on you sooner or > later. > > I knew it'd be something simple, thanks. Hi Mark, Thanks for raising this. I think doing diff --git a/tools/testing/selftests/mm/prctl_thp_disable.c b/tools/testing/selftests/mm/prctl_thp_disable.c index 89ed0d7db1c16..0afcdbad94f3d 100644 --- a/tools/testing/selftests/mm/prctl_thp_disable.c +++ b/tools/testing/selftests/mm/prctl_thp_disable.c @@ -9,6 +9,7 @@ #include <string.h> #include <unistd.h> #include <sys/mman.h> +#include <linux/mman.h> #include <sys/prctl.h> #include <sys/wait.h> should fix this issue? Thanks Usama
