version: 0.15 date: Thu Jun 11 12:45:33 CEST 2015 changes: - improve coalescing - add isl_union_access_info_compute_flow - add mark nodes in AST - add isl_union_pw_aff and isl_multi_union_pw_aff - add schedule trees - deprecate band forests - deprecate separation_class AST generation option - introduce isl_bool and isl_stat types --- version: 0.14.1 date: Thu Apr 9 12:57:23 CEST 2015 changes: - fix bug in affine expression normalization - fix handling of conditional validity constraints --- version: 0.14 date: Sat Oct 25 16:08:47 CEST 2014 changes: - support IMath as an optional replacement for GMP - minor AST generator improvements --- version: 0.13 date: Mon Apr 14 11:08:45 CEST 2014 changes: - deprecate isl_int - improved support for multi piecewise quasi-affine expressions - allow the user to impose a bound on the number of low-level operations - add isl_id_to_ast_expr and isl_id_to_pw_aff - add isl_schedule_constraints - hide internal structure of isl_vec - remove support for piplib --- version: 0.12.2 date: Sun Jan 12 12:09:46 CET 2014 changes: - MinGW-w64 build fix - fix simplification bug --- version: 0.12.1 date: Wed Jul 24 12:54:46 CEST 2013 changes: - handle malloc returning NULL on zero-size allocation - fix regression in AST generator --- version: 0.12 date: Sun Jun 23 20:23:05 CEST 2013 changes: - add isl_val abstraction --- version: 0.11.2 date: Tue Apr 9 18:45:10 CEST 2013 changes: - make code generation output the same on Solaris - fix some hard to trigger bugs --- version: 0.11.1 date: Mon Dec 10 11:55:30 CET 2012 changes: - add LICENSE file to distribution - make code generation output independent of endianness --- version: 0.11 date: Mon Dec 3 08:17:18 CET 2012 changes: - change license from LGPL 2.1 to MIT - add support for multi piecewise quasi-affine expressions - add code generation - various minor bug fixes --- version: 0.10 date: Sun Jun 3 18:00:16 CEST 2012 changes: - support for interaction with dependence analysis - add public API for vectors - improved support for (piecewise) multi quasi-affine expressions - various minor bug fixes --- version: 0.09 date: Sat Dec 17 18:19:26 CET 2011 changes: - improved argument parsing - hide internal structure of isl_options - improved support for parameter sets - configurable scheduling --- version: 0.08 date: Fri Oct 21 12:36:20 CEST 2011 changes: - improved parsing - drop isl_div abstraction - rename isl_dim to isl_space - |- explicitly differentiate between spaces of maps, sets and parameter sets - add support for identifiers - add support for (piecewise) multi quasi-affine expressions - preliminary Python bindings --- version: 0.07 date: Tue Jul 12 19:34:51 CEST 2011 changes: - hide internal structures of isl_div and isl_constraint - preliminary scheduling - add support for local spaces and (piecewise) quasi-affine expressions --- version: 0.06 date: Fri Mar 18 15:59:16 CET 2011 changes: - improved parsing - consistency changes in API - hide internal structure of isl_ctx --- version: 0.05.1 date: Wed Jan 5 10:21:42 CET 2011 changes: - fix simple symmetry detection in parametric integer programming --- version: 0.05 date: Thu Dec 23 17:03:14 CET 2010 changes: - rename header files from isl_header.h to isl/header.h - add higher level interface for dependence analysis - improved argument parsing - optionally triangulate domains during Bernstein expansion - support extended PolyLib format - hide internal structure of some data types - improved coalescing - add simple symmetry detection in parametric integer programming --- version: 0.04 date: Fri Sep 10 12:57:50 CEST 2010 changes: - rename isl_pw_qpolynomial_fold_add - add isl_map_apply_pw_qpolynomial_fold - support named and nested spaces - support union sets and maps - add public API for matrices --- version: 0.03 date: Tue Jun 29 13:16:46 CEST 2010 changes: - new printing functions - support for "may" accesses in dependence analysis - improved coalescing - improved transitive closure - fix several hard to trigger bugs - improved argument parsing - support parametric vertex enumeration for barvinok - optionally use Bernstein expansion to compute bounds