Re: commit into mysql-6.0-wtf:mysql-6.0-wtf branch (vvaintroub:2653) Bug#24509, WL#3049 [ifdefery]

[Davi Arnaut]

Hi Joerg,

Joerg Bruehe wrote:
> Hi Vlad, all !
> Vladislav Vaintroub wrote:
>> Hi all,
>> what Davi describes is a sane approach, that  if consistently applied can
>> result into high quality library, like Apache Portable Runtime. Hence , it
>> is not applicable for us ;)
>> Seriously, we do need a top-level mapper functions. The reason is that many
>> mysys IO functions include large chunks of code shared on all OS (error
>> handling, messing up with my_file_info array) and  it does not make sense to
>> replicate those chunks for every OS. What could make sense for Windows is to
>> create win_file_io.c with ANSI/Posix wrappers. Then, use these wrappers in
>> the mysys function.
>>
>> To illustrate, this approach would make piece of code like this 

[..]

> 
> I am perfectly ok with this style.
> It means that there will be only one place where "my_open()" is 
> implemented (using conditional compiling around pieces of code which are 
> quite small, so remain easily readable),
> and it would (on Windows) call "my_win_fdopen()".
> 
> In the same file, you could then have
> 
>    #ifdef __WIN__
>    static inline int my_win_fdopen(...) {
>      ...
>    }
>    #endif
> 
> which is not interrupted by conditional compiling, and thus reading and 
> maintaining the code is easy.
> 
> If you put that into a different file, it is "extern", but that should 
> not matter much.
> (IMNSHO, any system call will be so expensive that we need not worry too 
> much about one more function call in the chain.)

I invite you (add all colleagues that haven't seen this paper) to take
some time to read a 1992 USENIX paper about ifdef abuse:

http://www.chris-lott.org/resources/cstyle/ifdefs.pdf

H. Spencer and G. Collyer, "#ifdef Considered Harmful, or Portability
Experience with C News," in Proceedings of the Summer USENIX Conference,
(San Antonio, Texas), June 1992.

Regards,

-- Davi Arnaut