Stefan Manegold wrote:
> I see --- thanks! --- apparently, my clean-up attempt was a bit too eager
> ...
> 
> Also my attemt of just re-adding the *export definitions in .h header files
> does not seem to work; I still get the missing lib_streams.lib (ate laste
> with icc --- vs seems to work fine in my settings, and did so when testing
> yesterday, but also failed with nightly testing ...
> 
> What do you recon?

This probably means there are no exported functions in the module, so no
need to create a .lib file.

All functions referred to in the @mal sections using the address
construct have to be exported.

> Just undo my changes "as-is"?

That's easiest...

> Or undo them with your simplified definition of *_export in .c source files?

That could be done later as well.

> Stefan
> 
> On Mon, May 25, 2009 at 09:55:55AM +0200, Sjoerd Mullender wrote:
>> Functions that are referred to from outside the DLL they are defined in
>> *must* be declared with __declspec(dllexport) in the DLL they are
>> defined in, and with __declspec(dllimport) in the DLL they are used in.
>>  If they are looked up dynamically, then they still *must* be declared
>> with __declspec(dllexport).  Only functions that are not used outside
>> the DLL they are defined in don't need a __declspec declaration.
>>
>> The *_export defines you removed in the @c sections were there for the
>> latter bit.  They could be simplified to e.g.
>>
>> #ifdef·WIN32
>>
>> #define·XXX_export·extern·__declspec(dllexport)
>>
>> #else
>>
>> #define·XXX_export·extern
>>
>> #endif
>>
>>
>> since these defines will not be used outside the DLL they are defined in.
>>
>> Stefan Manegold wrote:
>>> Dear all,
>>>
>>> removing the *_export macro definitions from .c source files seems to have
>>> broken M5 compilation on Windows --- I'm busy fixing it by re-adding these
>>> definitions in .h header files...
>>>
>>> Stefan
>>>
>>> On Mon, May 25, 2009 at 05:06:19AM +0200, Testweb Compilation wrote:
>>>>    copy "lib_streams.lib" 
>>>> "g:\monet\monet.Int.32.32.d.1.28441\MonetDB5\.Int.32.32.d.1-Windows5.1\lib\MonetDB5\lib_streams.lib"
>>>> The system cannot find the file specified.
>>>> NMAKE : fatal error U1077: 'copy' : return code '0x1'
>>>>
>>>> http://www.ins.cwi.nl/~monet/permastore/2009-05-25/Stable/install-MonetDB5-Int.32.32.d.1-Windows5.1.errors.html#l525
>>>>
>>>> CVS committers: [email protected] [email protected] [email protected]
>>>>
>>>> module maintainers: [email protected] [email protected]
>>>>
>>>> An identical error message was detected on the following systems:
>>>> Int.32.32.d.1-Windows5.1 Mic.32.32.d.1-Windows5.1 Int.64.32.d.1-Windows5.2 
>>>> Mic.64.32.d.1-Windows5.2 Mic.64.64.d.1-Windows5.2
>>>>
>>>> -- 
>>>> At this time you cannot configure Testing mail, sorry
>>>> You are receiving this mail because:
>>>> - you checked in some code on MonetDB5 (Stable)
>>>> - you are a fixed recipient
>>>>
>>
>> -- 
>> Sjoerd Mullender
>>
> 
> 
> 


-- 
Sjoerd Mullender

Attachment: signature.asc
Description: OpenPGP digital signature

------------------------------------------------------------------------------
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
is a gathering of tech-side developers & brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing, & 
iPhoneDevCamp asthey present alongside digital heavyweights like Barbarian
Group, R/GA, & Big Spaceship. http://www.creativitycat.com 
_______________________________________________
Monetdb-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/monetdb-developers

Reply via email to