test_maps_tearing_from_split times out when READ_IMPLIES_EXEC is set.
This happens by default on pre-ARMv6 CPUs, which lack no-execute support.

split_vma() re-maps the first page with mod_info->prot | PROT_EXEC to
make it differ from its neighbours.  With READ_IMPLIES_EXEC the original
mapping is already executable, so no split occurs and the test hangs
waiting for the modifier child.

Use PROT_NONE for the split mapping, which always differs from its
readable neighbours.

Fixes: beb69e817246 ("selftests/proc: add /proc/pid/maps tearing from vma split test")
Assisted-by: Codex:gpt-5.6-terra
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
---
Tested on:
- QEMU versatilepb/ARM926EJ-S: maps and smaps fail before, pass after
- QEMU realview-eb-mpcore/ARM11MPCore: maps and smaps pass before and after

Changes since v1:
- v1 went out without any recipients in To: and received no reply in 19 days.
- No functional change. reworked the commit message.

v1: https://lore.kernel.org/all/20260720061835.6290-1-kmehltretter@gmail.com/

 tools/testing/selftests/proc/proc-maps-race.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/testing/selftests/proc/proc-maps-race.c b/tools/testing/selftests/proc/proc-maps-race.c
index 1026d8c400e1..415eccb70468 100644
--- a/tools/testing/selftests/proc/proc-maps-race.c
+++ b/tools/testing/selftests/proc/proc-maps-race.c
@@ -490,7 +490,8 @@ static bool query_addr_at(int maps_fd, void *addr,
 
 static inline bool split_vma(FIXTURE_DATA(proc_maps_race) *self)
 {
-	return mmap(self->mod_info->addr, self->page_size, self->mod_info->prot | PROT_EXEC,
+	/* PROT_NONE differs from both readable neighbors. */
+	return mmap(self->mod_info->addr, self->page_size, PROT_NONE,
 		    MAP_PRIVATE | MAP_ANONYMOUS | MAP_FIXED, -1, 0) != MAP_FAILED;
 }
 
-- 
2.53.0

On Sat, Aug 8, 2026 at 1:04 PM Karl Mehltretter <kmehltretter@gmail.com> wrote:
>
> test_maps_tearing_from_split times out when READ_IMPLIES_EXEC is set.
> This happens by default on pre-ARMv6 CPUs, which lack no-execute support.
>
> split_vma() re-maps the first page with mod_info->prot | PROT_EXEC to
> make it differ from its neighbours.  With READ_IMPLIES_EXEC the original
> mapping is already executable, so no split occurs and the test hangs
> waiting for the modifier child.
>
> Use PROT_NONE for the split mapping, which always differs from its
> readable neighbours.
>
> Fixes: beb69e817246 ("selftests/proc: add /proc/pid/maps tearing from vma split test")
> Assisted-by: Codex:gpt-5.6-terra
> Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>

Sorry for missing both versions.

Acked-by: Suren Baghdasaryan <surenb@google.com>

> ---
> Tested on:
> - QEMU versatilepb/ARM926EJ-S: maps and smaps fail before, pass after
> - QEMU realview-eb-mpcore/ARM11MPCore: maps and smaps pass before and after
>
> Changes since v1:
> - v1 went out without any recipients in To: and received no reply in 19 days.
> - No functional change. reworked the commit message.
>
> v1: https://lore.kernel.org/all/20260720061835.6290-1-kmehltretter@gmail.com/
>
>  tools/testing/selftests/proc/proc-maps-race.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/tools/testing/selftests/proc/proc-maps-race.c b/tools/testing/selftests/proc/proc-maps-race.c
> index 1026d8c400e1..415eccb70468 100644
> --- a/tools/testing/selftests/proc/proc-maps-race.c
> +++ b/tools/testing/selftests/proc/proc-maps-race.c
> @@ -490,7 +490,8 @@ static bool query_addr_at(int maps_fd, void *addr,
>
>  static inline bool split_vma(FIXTURE_DATA(proc_maps_race) *self)
>  {
> -       return mmap(self->mod_info->addr, self->page_size, self->mod_info->prot | PROT_EXEC,
> +       /* PROT_NONE differs from both readable neighbors. */
> +       return mmap(self->mod_info->addr, self->page_size, PROT_NONE,
>                     MAP_PRIVATE | MAP_ANONYMOUS | MAP_FIXED, -1, 0) != MAP_FAILED;
>  }
>
> --
> 2.53.0