Consider the following test code program format_string implicit none character(len=*), parameter :: rform='(F15.5)', & cform="(' (', F15.5, ',' F15.5, ') ')" call print_a_number(cform) contains subroutine print_a_number(style) character(len=*) :: style write(*, style) cmplx(0.0, 0.0) end subroutine print_a_number end program format_string This program compiles fine. But gives a runtime error. $gfortran-20050507-1 -v Using built-in specs. Target: i486-linux Configured with: ../src/configure -v --enable-languages=c,c++,java,f95,objc,ada,treelang --prefix=/usr/lib/gcc-snapshot --enable-shared --with-system-zlib --disable-nls --program-suffix=-20050507-1 --enable-__cxa_atexit --enable-libstdcxx-allocator=mt --enable-clocale=gnu --enable-libstdcxx-debug --enable-java-gc=boehm --enable-java-awt=gtk --enable-mpfr --disable-werror --disable-werror i486-linux Thread model: posix gcc version 4.1.0 20050507 (experimental) $gfortran format_string.f90 -o format_string $./a.out At line 10 of file format_string.f90 Fortran runtime error: Unexpected end of format string (' (', ^ I am using Debian Sid distribution, gcc-snapshot 20050507-1 package. The above program compiles and runs fine with Absoft Fortran 90/95 compiler and with intel fortran compiler 8.1 Build 20040921. The interaction with Absoft compiler looks as follows $f90 format_string.f90 rajulocal@kusumanchi:~/work/fortran/gfortran_bugs 28 527 11:21 PM $./a.out ( 0.00000, 0.00000)
Reduced testcase is following. The root of the problem is that when we declare strings of different length in the same "character(len=*)" declaration, all strings are truncated to the length of the first one. Example: $ cat pr21459.f90 character(len=*) :: c1='12345', c2='123456789' print *, c1 print *, c2 end $ gfortran pr21459.f90 && ./a.out 12345 12345
The problem is that all characters declared on the same line share the same gfc_charlen structure. This should be fixed during resolution, I'll have a look at this.
This patch passes the testsuite and fixes the bug, but creates a memory leak: * decl.c (add_init_expr_to_sym): Create new gfc_charlen for each item with length = (*) Index: decl.c =================================================================== RCS file: /cvs/gcc/gcc/gcc/fortran/decl.c,v retrieving revision 1.35 diff -u -p -r1.35 decl.c --- decl.c 29 Apr 2005 15:31:37 -0000 1.35 +++ decl.c 11 May 2005 20:39:13 -0000 @@ -740,6 +740,11 @@ add_init_expr_to_sym (const char *name, /* Update symbol character length according initializer. */ if (sym->ts.cl->length == NULL) { + /* If there are multiple CHARACTER variables declared on + the same line, we don't want them to share the same + length. */ + sym->ts.cl = gfc_get_charlen (); + if (init->expr_type == EXPR_CONSTANT) sym->ts.cl->length = gfc_int_expr (init->value.character.length); In order to evade the memory leak, we should instead of setting sym->ts.cl->length = NULL, set sym->ts.cl = &gfc_unknown_charlen, where the latter is a new global placeholder variable, and then fill in sym->ts.cl as I did in the patch. I've seen something like this in g95, but I don't remember if this was what it was used for there.
Subject: Bug 21459 CVSROOT: /cvs/gcc Module name: gcc Changes by: pault@gcc.gnu.org 2005-10-17 20:52:37 Modified files: gcc/fortran : ChangeLog gfortran.h resolve.c expr.c decl.c match.c gcc/testsuite : ChangeLog Added files: gcc/testsuite/gfortran.dg: host_dummy_index_1.f90 automatic_char_len_2.f90 recursive_statement_functions.f90 assumed_size_dt_dummy.f90 external_initializer.f90 non_module_public.f90 Log message: 2005-10-17 Paul Thomas <pault@gcc.gnu.org> PR fortran/23446 * gfortran.h: Primitive for gfc_is_formal_arg. * resolve.c(gfc_is_formal_arg): New function to signal across several function calls that formal argument lists are being processed. (resolve_formal_arglist): Set/reset the flag for gfc_is_formal_arg. *expr.c(check_restricted): Add check, via gfc_is_formal_arg, if symbol is part of an formal argument declaration. PR fortran/21459 * decl.c (add_init_expr_to_sym): Make a new character length for each variable, when the expression is NULL and link to cl_list. PR fortran/20866 * match.c (recursive_stmt_fcn): New function that tests if a statement function resurses through itself or other other statement functions. (gfc_match_st_function): Call recursive_stmt_fcn to check if this is recursive and to raise error if so. PR fortran/20849 PR fortran/20853 * resolve.c (resolve_symbol): Errors for assumed size arrays with default initializer and for external objects with an initializer. PR fortran/20837 * decl.c (match_attr_spec): Prevent PUBLIC from being used outside a module. 2005-10-17 Paul Thomas <pault@gcc.gnu.org> PR fortran/23446 * gfortran.dg/host_dummy_index_1.f90: New test. PR fortran/21459 gfortran.dg/automatic_char_len_2.f90: New test. PR fortran/20866 gfortran.dg/recursive_statement_functions.f90: New test. PR fortran/20853 gfortran.dg/assumed_size_dt_dummy.f90: New test. PR fortran/20849 gfortran.dg/external_initializer.f90: New test. PR fortran/20837 non_module_public.f90: New test. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/ChangeLog.diff?cvsroot=gcc&r1=1.587&r2=1.588 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/gfortran.h.diff?cvsroot=gcc&r1=1.89&r2=1.90 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/resolve.c.diff?cvsroot=gcc&r1=1.59&r2=1.60 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/expr.c.diff?cvsroot=gcc&r1=1.30&r2=1.31 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/decl.c.diff?cvsroot=gcc&r1=1.42&r2=1.43 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/match.c.diff?cvsroot=gcc&r1=1.46&r2=1.47 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.6205&r2=1.6206 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/host_dummy_index_1.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/automatic_char_len_2.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/recursive_statement_functions.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/assumed_size_dt_dummy.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/external_initializer.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/non_module_public.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1
Subject: Bug 21459 CVSROOT: /cvs/gcc Module name: gcc Branch: gcc-4_0-branch Changes by: pault@gcc.gnu.org 2005-10-18 05:55:23 Modified files: gcc/fortran : ChangeLog gfortran.h resolve.c expr.c decl.c match.c gcc/testsuite : ChangeLog Added files: gcc/testsuite/gfortran.dg: host_dummy_index_1.f90 automatic_char_len_2.f90 recursive_statement_functions.f90 assumed_size_dt_dummy.f90 external_initializer.f90 non_module_public.f90 Log message: 2005-10-18 Paul Thomas <pault@gcc.gnu.org> PR fortran/23446 * gfortran.h: Primitive for gfc_is_formal_arg. * resolve.c(gfc_is_formal_arg): New function to signal across several function calls that formal argument lists are being processed. (resolve_formal_arglist): Set/reset the flag for gfc_is_formal_arg. *expr.c(check_restricted): Add check, via gfc_is_formal_arg, if symbol is part of an formal argument declaration. PR fortran/21459 * decl.c (add_init_expr_to_sym): Make a new character length for each variable, when the expression is NULL and link to cl_list. PR fortran/20866 * match.c (recursive_stmt_fcn): New function that tests if a statement function resurses through itself or other other statement functions. (gfc_match_st_function): Call recursive_stmt_fcn to check if this is recursive and to raise error if so. PR fortran/20849 PR fortran/20853 * resolve.c (resolve_symbol): Errors for assumed size arrays with default initializer and for external objects with an initializer. PR fortran/20837 * decl.c (match_attr_spec): Prevent PUBLIC from being used outside a module. 2005-10-18 Paul Thomas <pault@gcc.gnu.org> PR fortran/23446 * gfortran.dg/host_dummy_index_1.f90: New test. PR fortran/21459 gfortran.dg/automatic_char_len_2.f90: New test. PR fortran/20866 gfortran.dg/recursive_statement_functions.f90: New test. PR fortran/20853 gfortran.dg/assumed_size_dt_dummy.f90: New test. PR fortran/20849 gfortran.dg/external_initializer.f90: New test. PR fortran/20837 non_module_public.f90: New test. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.335.2.130&r2=1.335.2.131 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/gfortran.h.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.58.2.18&r2=1.58.2.19 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/resolve.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.34.2.18&r2=1.34.2.19 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/expr.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.23.2.4&r2=1.23.2.5 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/decl.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.31.2.5&r2=1.31.2.6 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/match.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.31.8.13&r2=1.31.8.14 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.5084.2.465&r2=1.5084.2.466 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/host_dummy_index_1.f90.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/automatic_char_len_2.f90.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/recursive_statement_functions.f90.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/assumed_size_dt_dummy.f90.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/external_initializer.f90.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/non_module_public.f90.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1
Fixed.