apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Allan Edwards <a...@us.ibm.com>
Subject [Fwd: Re: cvs commit: apr/atomic/win32 apr_atomic.c]
Date Mon, 20 Sep 2004 22:49:27 GMT
meant to reply to dev@apr also
-------- Original Message --------
Subject: Re: cvs commit: apr/atomic/win32 apr_atomic.c
Date: Mon, 20 Sep 2004 18:42:24 -0400
From: Allan Edwards <ake1@us.ibm.com>
To: Cliff Woolley <jwoolley@virginia.edu>
References: <20040920192013.71968.qmail@minotaur.apache.org> <Pine.GSO.4.58.0409201543200.4649@cobra.cs.Virginia.EDU>

> Doesn't this just mean that the definition of apr_atomic_win32_ptr_val_fn
> is wrong on win64?  Wouldn't it be cleaner to redefine that on win64
> instead of these two #if #else cases?

The problem is with the use of a typedef to remap the function.
It makes the compiler think there is an external function
called InterlockedExchangeAdd when it really is an
intrinsic function. So redefining for win64 does not
work (in fact no remapping would be needed for win64 since
it matches the typedef).

Allan



Mime
View raw message