Re: [PATCH 2/4] wait: Make ___wait_is_interruptible() return 0 or 1

[Vineet Gupta <vineet.gupta@linux.dev>]

On 11/5/26 1:03 AM, Peter Zijlstra wrote:
> On Wed, Nov 04, 2026 at 08:07:19AM -0800, Vineet Gupta wrote:
>> GCC 17 adds -Wconstant-logical-operand to -Wall. It warns when an operand
>> of && or || folds to a constant other than 0 or 1. With a constant state
>> such as TASK_KILLABLE, ___wait_is_interruptible() breaks the build with
>> CONFIG_WERROR=y:
>>
>>    include/linux/wait.h:285:39: error: use of logical '||' with constant operand '256' [-Werror=constant-logical-operand]
>>
>> clang has the same warning, but it does not check an operator that comes
>> from a macro expansion.
>>
>> Compare the masked state against 0. All callers test the result as a
>> truth value, so the behaviour does not change.
>>
>> Signed-off-by: Vineet Gupta <vineet.gupta@linux.dev>
>> ---
>>   include/linux/wait.h | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/include/linux/wait.h b/include/linux/wait.h
>> index c2af98b0074d..ebabdfb5b30c 100644
>> --- a/include/linux/wait.h
>> +++ b/include/linux/wait.h
>> @@ -283,7 +283,7 @@ static inline void wake_up_pollfree(struct wait_queue_head *wq_head)
>>   
>>   #define ___wait_is_interruptible(state)						\
>>   	(!__builtin_constant_p(state) ||					\
>> -	 (state & (TASK_INTERRUPTIBLE | TASK_WAKEKILL)))
>> +	 (state & (TASK_INTERRUPTIBLE | TASK_WAKEKILL)) != 0)
>>   
>>   extern void init_wait_entry(struct wait_queue_entry *wq_entry, int flags);
> What is the actual purpose of this warning? This doesn't seem to serve
> any value. Any non-zero value is a true, why is this a problem?

I think the intent of the toggle was to catch && vs. & mixups in source 
(and it only catches that in 1 direction). Clearly that is not 
applicable here and it seems it has never caught this in the kernel anyway.

> GCC doing this, don't mean we can't turn if off again.

I will send a v2 which turns the GCC version off.

Thx,
-Vineet