[flext] LIBCPMT.LIB LIBCMT.LIB clash (was Re: linker error when linking a library)

[Stefano Papetti]

Hi Patrick,

thanks for the help.
I had already tried that, but only from command line, just by adding the 
line:
LDFLAGS =$(LDFLAGS)/NODEFAULTLIB:LIBCMT.LIB /NODEFAULTLIB:LIBCPMT.LIB
to the file package.txt
However, it seems that flext's build system overrides this flag, as the 
linker outputs the same error as before.

I plan to try your suggestion in VS, but my purpose at the moment is to 
get things working directly from command line.

Best,
S

On 25/11/2011 12.56, Patrick Gampp wrote:
> Hi Stefano,
> in Visual Studio, exists a flag you could set to ignore LIBCMT.LIB. Something
> like "ignore specific library".
>
> http://msdn.microsoft.com/en-us/library/3tz4da4a%28v=vs.80%29.aspx
> http://msdn.microsoft.com/en-us/library/aa278403%28v=vs.60%29.aspx
>
> Patrick
>
>
>
>
>
>
> Am Freitag, 25. November 2011, 12:13:41 schrieb Stefano Papetti:
>> Dear Thomas,
>>
>> after giving up for a while, I'm now trying again to figure out how to
>> solve the LIBCPMT.LIB LIBCMT.LIB clash reported in the file attached.
>> The same thing happens compiling in debug mode (with respect to the
>> debug libraries, of course). On the other hand, the shared version of my
>> external compiles successfully.
>> I went through all the code but unfortunately I can't really see where
>> this clash is taking place.
>> Of course if I comment:
>> LIBS=$(LIBS) LIBCPMT.LIB LIBCMT.LIB
>> in \flext\buildsys\win\nmake-msvc-flext.inc everything works... :-P
>>
>> Do you have any hints?
>>
>> Ciao,
>> S
>>
>> On 06/08/2011 13.04, Thomas Grill wrote:
>>> Am 06.08.2011 um 11:49 schrieb Stefano Papetti:
>>>> Hi Thomas,
>>>>
>>>> On 06/08/2011 11.35, Thomas Grill wrote:
>>>>> There are essentially two problems, maybe unrelated to each other
>>>>>
>>>>> - a clash of the delete operator - this is usually caused by using
>>>>> different build models for parts that are linked. In your case it
>>>>> seems that the libraries LIBCPMT.LIB and LIBCMT.LIB don't fit to each
>>>>> other.
>>>> ok, I'll try to figure out how to avoid this. Perhaps I just need to
>>>> avoid using delete and new ;)
>>> that would not be a clean solution - there might be other not so obvious
>>> clashes…. gr~~~
> .
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://grrrr.org/pipermail/flext/attachments/20111130/66316896/attachment.htm>