On Thursday, June 16, 2016 at 1:07:57 PM UTC-4, jean-pierre.muench wrote:
>
> First to answer your question:
>
You suffer from the fact that you actually compile cryptlib.cpp, which is 
> the main implementation for a lot of things in the library, this also 
> depends on the FIPS 140 code, which in turn wants to compile self-tests for 
> everything eligible for FIPS certification.
>

Yes, that's exactly what the problem was. cryptlib.cpp is a hub containing 
tons of cross-cutting concerns.
 

> Did you try restricting your compilation to zlib, zinflate and zdeflate 
> and only add what otherwise gives linker errors?
>

Yes, that's how I started. Things got very ugly very fast, requiring 
extensive edits to key files.

I decided to abandon my efforts to adapt cryptlib's deflate implementation 
for a couple of reasons:

* BufferedTransformation is not the ideal deflate interface, and it has too 
many members
* The implementation is not optimized
* There are features that simply aren't needed, like SecBlock
* Too many interdependent classes with cross cutting concerns

Instead I am just biting the bullet and porting zlib to a header-only C++ 
implementation, doing away with the macros and odd language constructs, and 
requiring C++11 (thus doing away with all the code needed for supporting 
ancient compilers). You can see that progress here:

https://github.com/vinniefalco/Beast/blob/59e3a4e413113e3d1c329c20cfc963b250b7d1fb/include/beast/core/impl/zistream.ipp#L15



-- 
-- 
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.
--- 
You received this message because you are subscribed to the Google Groups 
"Crypto++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to