Re: [PATCH v2 2/4] Consolidate stdio-lock.h

[Adhemerval Zanella <adhemerval.zanella@linaro.org>]

On 27/04/2022 15:44, Florian Weimer wrote:
> * Adhemerval Zanella:
> 
>> You are right.  The change now uses __libc_cleanup_region_start from
>> libc-lock.h instead of the cleanup version.  I am not sure about
>> hardening, but afaiu __libc_cleanup_push_defer should handle C++
>> unwinding for fopencookie.
> 
> No, C++ unwinding will right go through these frames.  We only
> interleave in the other direction (pthread_cancel unwinding running C++
> destructors).

Right, so Hurd version is indeed broken.

> 
>> It seems that the Hurd version is indeed to the best option, so I think for
>> generic it would be better to just use:
>>
>> diff --git a/sysdeps/generic/stdio-lock.h b/sysdeps/generic/stdio-lock.h
>> index 14cf458bdd..a42131f5a5 100644
>> --- a/sysdeps/generic/stdio-lock.h
>> +++ b/sysdeps/generic/stdio-lock.h
>> @@ -45,20 +45,12 @@ __libc_lock_define_recursive (typedef, _IO_lock_t)
>>  #define _IO_cleanup_region_end(_doit) \
>>    __libc_cleanup_region_end (_doit)
>>
>> -#if defined _LIBC && IS_IN (libc)
>> -
>> -# ifdef __EXCEPTIONS
>> -# define _IO_acquire_lock(_fp) \
>> +#define _IO_acquire_lock(_fp) \
>>    do {                                                                       \
>>      FILE *_IO_acquire_lock_file                                                      \
>>         __attribute__((cleanup (_IO_acquire_lock_fct)))                       \
>>         = (_fp);                                                              \
>>      _IO_flockfile (_IO_acquire_lock_file);
>> -# else
>> -#  define _IO_acquire_lock(_fp) _IO_acquire_lock_needs_exceptions_enabled
>> -# endif
>> -# define _IO_release_lock(_fp) ; } while (0)
>> -
>> -#endif
>> +#define _IO_release_lock(_fp) ; } while (0)
>>
>>  #endif /* stdio-lock.h */
> 
> I suppose so.
> 

I will update the patch.