Re: [PATCH] target/s390x: fix s390_probe_access to check PAGE_WRITE_ORG for writeability

[Cornelia Huck]

On Thu, 22 Apr 2021 16:44:27 +0100
Alex Bennée <alex.bennee@linaro.org> wrote:

> We can remove PAGE_WRITE when (internally) marking a page read-only
> because it contains translated code. This can get confused when we are
> executing signal return code on signal stacks.
> 
> Fixes: e56552cf07 ("target/s390x: Implement the MVPG condition-code-option 
> bit")
> Found-by: Richard Henderson <richard.henderson@linaro.org>
> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
> Cc: Cornelia Huck <cohuck@redhat.com>
> Cc: Thomas Huth <thuth@redhat.com>
> Cc: David Hildenbrand <david@redhat.com>
> Cc: Laurent Vivier <laurent@vivier.eu>
> ---
>  target/s390x/mem_helper.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/target/s390x/mem_helper.c b/target/s390x/mem_helper.c
> index 12e84a4285..f6a7d29273 100644
> --- a/target/s390x/mem_helper.c
> +++ b/target/s390x/mem_helper.c
> @@ -145,7 +145,7 @@ static int s390_probe_access(CPUArchState *env, 
> target_ulong addr, int size,
>  
>  #if defined(CONFIG_USER_ONLY)
>      flags = page_get_flags(addr);
> -    if (!(flags & (access_type == MMU_DATA_LOAD ?  PAGE_READ : PAGE_WRITE))) 
> {
> +    if (!(flags & (access_type == MMU_DATA_LOAD ?  PAGE_READ : 
> PAGE_WRITE_ORG))) {
>          env->__excp_addr = addr;
>          flags = (flags & PAGE_VALID) ? PGM_PROTECTION : PGM_ADDRESSING;
>          if (nonfault) {

What's the verdict on this one? I plan to queue this to s390-next; but
if we end up doing an -rc5, it might qualify as a regression fix.

If this is going to be in 6.1, I'll add cc:stable when queuing.