Re: [PATCH for-8.0 1/7] qemu/main-loop: Introduce QEMU_IOTHREAD_LOCK_GUARD

[Richard Henderson <richard.henderson@linaro.org>]

On 11/18/22 05:38, Alex Bennée wrote:
> 
> Richard Henderson <richard.henderson@linaro.org> writes:
> 
>> Create a wrapper for locking/unlocking the iothread lock.
>>
>> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
>> ---
>> Cc: Paolo Bonzini <pbonzini@redhat.com> (maintainer:Main loop)
>> ---
>>   include/qemu/main-loop.h | 29 +++++++++++++++++++++++++++++
>>   1 file changed, 29 insertions(+)
>>
>> diff --git a/include/qemu/main-loop.h b/include/qemu/main-loop.h
>> index 3c9a9a982d..c25f390696 100644
>> --- a/include/qemu/main-loop.h
>> +++ b/include/qemu/main-loop.h
>> @@ -343,6 +343,35 @@ void qemu_mutex_lock_iothread_impl(const char *file, int line);
>>    */
>>   void qemu_mutex_unlock_iothread(void);
>>   
>> +/**
>> + * QEMU_IOTHREAD_LOCK_GUARD
>> + *
>> + * Wrap a block of code in a conditional qemu_mutex_{lock,unlock}_iothread.
>> + */
>> +typedef struct IOThreadLockAuto IOThreadLockAuto;
>> +
>> +static inline IOThreadLockAuto *qemu_iothread_auto_lock(const char *file,
>> +                                                        int line)
>> +{
>> +    if (qemu_mutex_iothread_locked()) {
>> +        return NULL;
>> +    }
>> +    qemu_mutex_lock_iothread_impl(file, line);
>> +    /* Anything non-NULL causes the cleanup function to be called */
>> +    return (IOThreadLockAuto *)(uintptr_t)1;
> 
> Oh hang on, what black magic is this. Does the compiler do a NULL check
> before calling the cleanup?

Not the compiler, but...

>> +G_DEFINE_AUTOPTR_CLEANUP_FUNC(IOThreadLockAuto, qemu_iothread_auto_unlock)

... this does.  Follow the macros down and you get

>   static G_GNUC_UNUSED inline void _GLIB_AUTOPTR_CLEAR_FUNC_NAME(TypeName) (TypeName *_ptr)
>     { if (_ptr) (cleanup) ((ParentName *) _ptr); }                                                       

r~