Status of availability of features which allow correct and seamless support of DLLs in current GNU-Win32 releases

Paul Sokolovsky paul-ml@is.lg.ua
Wed May 3 08:23:00 GMT 2000


Hello Chris,

Wednesday, May 03, 2000, 5:33:41 PM, you wrote:

[]
>>    Ha, that's the effect of stereotypes! It is at runtime, at the point of
>>specific variable access the decision on how to access it should be
>>unavoidably made: either access var directly or via extra level of
>>inderection.

CF> Um.  This sort of needs to be done at the time of the compilation or assembly.
CF> Otherwise you have to give the linker understanding of x86 (or whaterver)
CF> opcodes.

    I do it in following way: when object is compiled, produce only
assembly. When it is linked - so all library dependencies are known -
assembly is postprocessed, assembled and linked.

CF> Do you think if it was really this easy, it wouldn't have been done this
CF> way already?

    I believe that someone already might do that. However, one failed
to contribute results to GNU libtool, the standard tool to support
builiding of shared libs in portable manner, what I hope to do,
whenever it will be possible.

    But that can be only after it will really be tested to work on
real data. One reason why I post about this here is in hope that
someone may see reasoning error in my scheme.

CF>   Microsoft would have done this long ago.

      You probaly won't believe that they kept it for someone else,
for example, me, unassuming. So, I just tell that they simply do not
have reasons to do that.

CF> cgf



Best regards,
 Paul                            mailto:paul-ml@is.lg.ua



--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com



More information about the Cygwin mailing list