You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@stdcxx.apache.org by Anton Pevtsov <An...@moscow.vdiweb.com> on 2006/06/09 16:02:50 UTC

RE: release configurations (8d, 8s, 12d, 12s) fail to compile on Windows, MSVC 7.1

Martin, what about this issue?

I tried to rebuild all configurations and found that release are still
broken.
Maybe, we should open a jira issue to not forget about the problem?

Or does this issue appear on my side only?


Thanks,
Anton Pevtsov


-----Original Message-----
From: Martin Sebor [mailto:sebor@roguewave.com] 
Sent: Friday, April 21, 2006 23:03
To: stdcxx-dev@incubator.apache.org
Subject: Re: release configurations (8d, 8s, 12d, 12s) fail to compile
on Windows, MSVC 7.1


Anton Pevtsov wrote:
> Unfortunately, I didn't build release configurations during long 
> periof of time.
> 
> But I suspect this commit: 
> http://svn.apache.org/viewcvs?rev=374551&view=rev
[...]

You're most likely right.

> It seems that there is a conflict between declarations of the swprintf

> from stdio.h and this one.
> 
> The stdio.h contains
> 
> ...
> /* This non-standard definition exists for historical reasons to avoid

> breaking old code */ _CRTIMP int __cdecl swprintf(wchar_t *, const 
> wchar_t *, ...); #ifdef  __cplusplus
> /* For C++ we have an overloaded version with the standard definition
*/
> extern "C++" _CRTIMP int __cdecl swprintf(wchar_t *, size_t, const
> wchar_t *, ...);
> #endif
> ...

Ugh. What a mess!

> 
> It is possible to avoid the error (but not warning for dll's) using 
> extern "C++" instead of extern "C" to allow overloading, but I am not 
> sure that this way is correct.

I suspect that's what we'll have to do. I.e., hardcode this into our
<cstdio> header for Windows/MSVC.

> The same is true for vswprintf and wctomb functions.

Let me look into it. Strangely, I don't see it in our nightly builds.

Martin

Re: release configurations (8d, 8s, 12d, 12s) fail to compile on Windows, MSVC 7.1

Posted by Martin Sebor <se...@roguewave.com>.
Martin Sebor wrote:
> Anton Pevtsov wrote:
> 
>> Martin, what about this issue?
>>
>> I tried to rebuild all configurations and found that release are still
>> broken.
>> Maybe, we should open a jira issue to not forget about the problem?
[...]

This should fix it:

http://svn.apache.org/viewvc?rev=413162&view=rev

Martin

Re: release configurations (8d, 8s, 12d, 12s) fail to compile on Windows, MSVC 7.1

Posted by Martin Sebor <se...@roguewave.com>.
Anton Pevtsov wrote:
> Martin, what about this issue?
> 
> I tried to rebuild all configurations and found that release are still
> broken.
> Maybe, we should open a jira issue to not forget about the problem?

Sure, it can't hurt :)

> 
> Or does this issue appear on my side only?

I suspect not. I'm behind on a number of things and this is
one of them. I'll try to get caught up by Monday. Thanks for
the reminder!

Martin

> 
> 
> Thanks,
> Anton Pevtsov
> 
> 
> -----Original Message-----
> From: Martin Sebor [mailto:sebor@roguewave.com] 
> Sent: Friday, April 21, 2006 23:03
> To: stdcxx-dev@incubator.apache.org
> Subject: Re: release configurations (8d, 8s, 12d, 12s) fail to compile
> on Windows, MSVC 7.1
> 
> 
> Anton Pevtsov wrote:
> 
>>Unfortunately, I didn't build release configurations during long 
>>periof of time.
>>
>>But I suspect this commit: 
>>http://svn.apache.org/viewcvs?rev=374551&view=rev
> 
> [...]
> 
> You're most likely right.
> 
> 
>>It seems that there is a conflict between declarations of the swprintf
> 
> 
>>from stdio.h and this one.
>>
>>The stdio.h contains
>>
>>...
>>/* This non-standard definition exists for historical reasons to avoid
> 
> 
>>breaking old code */ _CRTIMP int __cdecl swprintf(wchar_t *, const 
>>wchar_t *, ...); #ifdef  __cplusplus
>>/* For C++ we have an overloaded version with the standard definition
> 
> */
> 
>>extern "C++" _CRTIMP int __cdecl swprintf(wchar_t *, size_t, const
>>wchar_t *, ...);
>>#endif
>>...
> 
> 
> Ugh. What a mess!
> 
> 
>>It is possible to avoid the error (but not warning for dll's) using 
>>extern "C++" instead of extern "C" to allow overloading, but I am not 
>>sure that this way is correct.
> 
> 
> I suspect that's what we'll have to do. I.e., hardcode this into our
> <cstdio> header for Windows/MSVC.
> 
> 
>>The same is true for vswprintf and wctomb functions.
> 
> 
> Let me look into it. Strangely, I don't see it in our nightly builds.
> 
> Martin