ec53454bff
PR fortran/14976 PR fortran/16228 * data.c (assign_substring_data_value): Remove. (create_character_intializer): New function. (gfc_assign_data_value): Track the typespec for the current subobject. Use create_character_intializer. testsuite/ * gfortran.dg/data_char_1.f90: New test. From-SVN: r86256 |
||
|---|---|---|
| .. | ||
| .cvsignore | ||
| CONTRIB | ||
| ChangeLog | ||
| Make-lang.in | ||
| NEWS | ||
| README | ||
| TODO | ||
| arith.c | ||
| arith.h | ||
| array.c | ||
| bbt.c | ||
| check.c | ||
| config-lang.in | ||
| convert.c | ||
| data.c | ||
| decl.c | ||
| dependency.c | ||
| dependency.h | ||
| dump-parse-tree.c | ||
| error.c | ||
| expr.c | ||
| f95-lang.c | ||
| gfortran.h | ||
| gfortran.texi | ||
| gfortranspec.c | ||
| interface.c | ||
| intrinsic.c | ||
| intrinsic.h | ||
| invoke.texi | ||
| io.c | ||
| iresolve.c | ||
| lang-specs.h | ||
| lang.opt | ||
| match.c | ||
| match.h | ||
| matchexp.c | ||
| mathbuiltins.def | ||
| misc.c | ||
| module.c | ||
| options.c | ||
| parse.c | ||
| parse.h | ||
| primary.c | ||
| resolve.c | ||
| scanner.c | ||
| simplify.c | ||
| st.c | ||
| symbol.c | ||
| trans-array.c | ||
| trans-array.h | ||
| trans-common.c | ||
| trans-const.c | ||
| trans-const.h | ||
| trans-decl.c | ||
| trans-expr.c | ||
| trans-intrinsic.c | ||
| trans-io.c | ||
| trans-stmt.c | ||
| trans-stmt.h | ||
| trans-types.c | ||
| trans-types.h | ||
| trans.c | ||
| trans.h | ||
README
The goal of the gcc-g95 project is to create a Free (as in speech) Fortran 95 compiler. The code has been donated to the Free Software Foundation for inclusion in GCC, thE GNU Compiler Collection. WARNING: G95 is still under development. Perusing the g77 source, we estimate that about 200,000 lines of code will be necessary to fully implement g95. Currently, G95 is about 70,000 lines long, making it about version 0.3. The current g95 can generate code for most legal Fortran 77 programs, and we're getting close to being able to compile most Fortran 95 programs as well. The generated code may still be quite poor, however. Part of this is a back-end issue, since we're using the Work-In-Progress tree-ssa infrastructure.