- a thin layer, using a bit boost::mpi (for the communicator mostly ...)
along the lines discussed in #12.
- implemented reduce, allreduce, bcast for arrays, simple scalars,
and any custom type that support boost serialization.
- Custom types : the operations are done recursively on members.
No change is needed in the class to use this mpi routine, as long as
serialize function is defined.
- For arrays of basic types (int, double...), a direct call to MPI C API, which works also for views
(as long as they are contiguous).
- For arrays of more complex types, we revert to boost::mpi.
- Added a simple test.
- Work still in progress :
- missing a simple scatter/gather for the arrays
- need more tests & API thinking.
- dispatch array code to array lib
- reduce is "sum" only, but do we need more.
Now the tail have a fixed size. It actually makes everything simpler. I took
order_min = -1 and order_max = 8. This makes the tails compatible with the
previous implementation. However we might want to change this to something like
-10, 10 so that they are self-contained. This commit should also fix issue #11.
- to use already a few c14 convenience details :
-> polymorphic std::plus, e.g.
boost::mpi::reduce (world, A,C, std::c14::plus<>(),0);
this plus determine the type by itself ...
-> errors on the type can be very cryptic on the gf.
-> add std::c14::make_unique
(equivalent of make_shared for unique_ptr).
- The previous version of the * operator for matrix was too clever.
It was giving a lazy object and then rewriting C = A *B into gemm (a,A,B,0,C).
The pb was in case of aliasing : when e.g. C = A, or is a part of A.
gemm is not correct that case, and as a result generic code like
a = a *b
may not be correct in matrix case, which is unacceptable.
- So we revert to a simple * operator for matrix
that does immediate computation.
Same thing for matrix* vector
- we also suppress a_x_ty class.
-> for M = a * b,
when M is a matrix, there is no overhead due to move assignment
-> however, when M is a view, there is an additionnal copy.
-Correctness comes first, hence the fix.
However, if one wants more speed and one can guarantee that
there is no aliasing possible, then one has to write a direct gemm call.
-> det_manip class was adapted, since in that case, we can show there
no alias, and we want the speed gain, so the * ops where replaced
by direct blas call (using the array blas interface).
-> also gemm, gemv, ger were overloaded in the case the return
matrix/vector (i.e. last parameter of the function) is not an lvalue,
but a temporary view created on the fly.
gf: security in the case beta<0 added in the mesh construction
gf: inline added in slice
test/triqs/gf: test of on_mesh() added
gfs: scalar for two-real_times
test/triqs/gf/ renamed in gfs, test gf_retw.cpp completed
gfs: evaluator homogeneised
two_times: evaluator corrected
test/triqs/gf/ renamed in gfs, test gf_retw.cpp completed
+ Correction after rebase
Fix a test : gf_re_im_freq_time
There is an issue with the last point.
To be fixed.
- change : all objects are by default
stored now by reference, not by copy any more.
Unless the trait force_copy_in_expr is true.
- rvalue refs are moved into the tree
- simplifies a lot the writing of lazy method, objects.
- added a macro for methods
- tests ok. Further check needed to control absence of copies...
- improved documentation
I added a version.hpp and updated the version.py so that
one can find the version of the library either from the
c++ or the python (if there is python support).
modified: CMakeLists.txt
modified: cmake/TRIQSConfig.cmake.in
modified: pytriqs/version.py.in
modified: triqs/CMakeLists.txt
new file: triqs/version.hpp.in
gcc has a pb because the template mesh<Variable,Opt>
has the name same as the gf mesh method (!).
Clang is fine however on this...
Solution : rename the template mesh<...> to gf_mesh...
Not very elegant, but ok.
- has_view_type_tag mechanism was replaced by a decltype long ago.
Cleaning since it seems to work on all supported compiler now.
- removed add_views : not used any more.
- Apparently there is a bug in release 3.3 of clang
(? on kondo cluster ?) that crash the compiler in compiling
a test. Pb not present in clang from svn.
Changed a little bit the libs as a workaround.
new policy : for non standard conforming compilers,
we can skip the tests using new features using recent c++ features.
I.e. compilers below gcc 4.8.1, clang 3.3.
- All access to grid is now using the [] operator.
() always returns const, and is an evaluation over the domain.
- Now :
* () is always a call to evaluator (or a lazy expression).
* mesh_pt accepted by [] , both const and non const,
and not by ().
- Ported the libs (fourier et al.) but this break API.
- necessary otherwise the class gf and the namespace have
the same name, leading to some confusion, and need to qualify
some functions (specially on gcc).
Same naming conventions as arrays.
- A()(0) = 1
will not work any more if A is const, as it should be.
- Introduced the notion of const_view, simply done by a flag
(flag 0, for easy debug)
- changed the slice of ISP to use it (only on C++11 compliant compiler)
- Move BoundCheck Flags to 5.
- improve the mem_block and shared_block.
- the reference counting is now done in the mem_block and shared_block, removing the need of shared_ptr.
- speed tests shows that shared_ptr is very slow (due to thread safety?)
the new version is much better, though not perfect.
- Hence introducing weak views.
- also :
-- clean the guard mechanism for python (to allow returning from python without any python ref left).
-- clean code, add documentation for mem_block
-- remove nan init, which was not working, and corresponding test
-- serialisation of view still unchanged (need to forbid serialization of view ??).
- tests ok, incl. valgrind tests.
- introducing scalar_valued gf
- Change Fourier routines to run on scalar_valued,
and then use those routines to run on matrix_valued.
- Tools for slices of 2 variables functions
