Nathan Chancellor <[email protected]> writes:

> After a recent improvement in clang to warn on unsequenced accesses to
> volatile variables [1], an instance appears in show_task(), which can
> break the build with CONFIG_PPC_WERROR=y / CONFIG_WERROR=y / W=e:
>
>   arch/powerpc/xmon/xmon.c:3262:3: error: unsequenced accesses to volatile 
> qualified 'tsk' [-Werror,-Wunsequenced]
>    3261 |         printf("%16px %16lx %16px %6d %6d %c %2d %s\n", tsk,
>         |                                                         ~~~
>    3262 |                 tsk->thread.ksp, tsk->thread.regs,
>         |                 ^
>
> Per C11 5.1.2.3.2, accessing a volatile object is a side effect. Per C11
> 6.5.2, unsequenced side effects on the same scalar object are undefined
> behavior. Per C11 6.5.2.2.10/12, the evaluation of arguments in a
> function call is indeterminately sequenced, meaning any side effects in
> the arguments are unsequenced.
>
> tsk was made volatile in commit a2305e3de819 ("powerpc: mark local
> variables around longjmp as volatile") to ensure its value in
> show_tasks() was properly restored after returning from a potential
> longjmp(). Aside from that, the value is not being modified through
> other means, so rename the volatile parameter and add a local variable
> named tsk to hold its value throughout show_task(), clearing up the
> warning.

Yes. Reading vol_tsk into tsk (local variable) in show_task() function,
doesn't cause any functional change and also fixes the unsequenced
volatile accesses as you described.

>
> Cc: [email protected]
> Fixes: a2305e3de819 ("powerpc: mark local variables around longjmp as 
> volatile")
> Closes: https://github.com/ClangBuiltLinux/linux/issues/2199
> Link: 
> https://github.com/llvm/llvm-project/commit/955715460eaf9188c241c74750637fc82dfcc423
>  [1]
> Signed-off-by: Nathan Chancellor <[email protected]>
> ---
>  arch/powerpc/xmon/xmon.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>

So, LGTM. Feel free to add:
Reviewed-by: Ritesh Harjani (IBM) <[email protected]>


Reply via email to