The enclosed example causes a segmentation fault at runtime under Cygwin. Reducing the length of pool1 by 1 allows it to run. This length looks like 2^18 less a header block so paging/memory/stack was suspected. Note this does not occur in Linux or native Windows versions of gfortran. Setting -fmax-stack-var-size appropriately makes no difference but d:/irun/bin/gfortran -Wl,--stack=2100000 fixes the problem. -fmax-stack-var-size seems to work for non-equivalence variables but is broken for equivalenced variables - see correspondence on the list between myself and A.Pinskia dated 6-7 March 2005 under the subject: Re: Equivalence segfault in TEST_FPU.F90 Test programme and output: __________________________ program test_equivalence real*8 :: pool1(260105),pool2(1) equivalence (pool1,pool2) end program test_equivalence $ d:/irun/bin/gfortran -v test_prequi.f90 Driving: d:/irun/bin/gfortran -v test_prequi.f90 -lgfortranbegin -lgfortran Using built-in specs. Configured with: ../gcc/configure --with-mpfr=/gmp-gcc4 --with-gmp=/gmp-gcc4 -en able-languages=c,f95 --prefix=/cygdrive/d/irun Thread model: single gcc version 4.0.0 20050129 (experimental) /cygdrive/d/irun/libexec/gcc/i686-pc-cygwin/4.0.0/f951.exe test_prequi.f90 -qui et -dumpbase test_prequi.f90 -mtune=pentiumpro -auxbase test_prequi -version -o /cygdrive/c/DOCUME~1/PAULTH~1/LOCALS~1/Temp/cc1r1KUq.s GNU F95 version 4.0.0 20050209 (experimental) (i686-pc-cygwin) compiled by GNU C version 4.0.0 20050131 (experimental). GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096 as -o /cygdrive/c/DOCUME~1/PAULTH~1/LOCALS~1/Temp/ccK4DcrJ.o /cygdrive/c/DOCUME ~1/PAULTH~1/LOCALS~1/Temp/cc1r1KUq.s /cygdrive/d/irun/libexec/gcc/i686-pc-cygwin/4.0.0/collect2.exe -Bdynamic --dll- search-prefix=cyg /lib/crt0.o -L/cygdrive/d/irun/lib/gcc/i686-pc-cygwin/4.0.0 -L /cygdrive/d/irun/lib/gcc/i686-pc-cygwin/4.0.0/../../.. /cygdrive/c/DOCUME~1/PAUL TH~1/LOCALS~1/Temp/ccK4DcrJ.o -lgfortranbegin -lgfortran -lgcc -lcygwin -luser32 -lkernel32 -ladvapi32 -lshell32 -lgcc PaulThomas@Paul /cygdrive/d/gfortran $ ./a Segmentation fault (core dumped)
Just I note, my last name is Pinski, not Pinskia, the a is for my first name. Confirmed.
Subject: Bug 20361 CVSROOT: /cvs/gcc Module name: gcc Changes by: tobi@gcc.gnu.org 2005-03-12 21:44:44 Modified files: gcc/fortran : ChangeLog trans-array.c trans-array.h trans-common.c trans-decl.c trans.h gcc/testsuite : ChangeLog Added files: gcc/testsuite/gfortran.dg: largeequiv_1.f90 Log message: fortran/ PR fortran/20361 * trans-array.c (gfc_stack_space_left): Remove unused variable. (gfc_can_put_var_on_stack): Move to trans-decl.c, remove #if 0'ed code. * trans-array.h (gfc_stack_space_left, gfc_can_put_var_on_stack): Remove declaration / prototype. * trans-common.c (build_equiv_decl): Give union a name. Check if it can be put on the stack. * trans-decl.c (gfc_stack_space_left): Move function here. (gfc_build_qualified_array): Fix comment typo. * trans.h (gfc_put_var_on_stack): Add prototype. testsuite/ PR fortran/20361 * gfortran.dg/largeequiv_1.f90: New test. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/ChangeLog.diff?cvsroot=gcc&r1=1.348&r2=1.349 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans-array.c.diff?cvsroot=gcc&r1=1.39&r2=1.40 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans-array.h.diff?cvsroot=gcc&r1=1.7&r2=1.8 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans-common.c.diff?cvsroot=gcc&r1=1.23&r2=1.24 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans-decl.c.diff?cvsroot=gcc&r1=1.54&r2=1.55 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans.h.diff?cvsroot=gcc&r1=1.23&r2=1.24 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.5151&r2=1.5152 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/largeequiv_1.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1
Subject: Bug 20361 CVSROOT: /cvs/gcc Module name: gcc Branch: gcc-4_0-branch Changes by: tobi@gcc.gnu.org 2005-03-12 21:50:51 Modified files: gcc/fortran : ChangeLog trans-array.c trans-array.h trans-common.c trans-decl.c trans.h gcc/testsuite : ChangeLog Added files: gcc/testsuite/gfortran.dg: largeequiv_1.f90 Log message: fortran/ PR fortran/20361 * trans-array.c (gfc_stack_space_left): Remove unused variable. (gfc_can_put_var_on_stack): Move to trans-decl.c, remove #if 0'ed code. * trans-array.h (gfc_stack_space_left, gfc_can_put_var_on_stack): Remove declaration / prototype. * trans-common.c (build_equiv_decl): Give union a name. Check if it can be put on the stack. * trans-decl.c (gfc_stack_space_left): Move function here. (gfc_build_qualified_array): Fix comment typo. * trans.h (gfc_put_var_on_stack): Add prototype. testsuite/ PR fortran/20361 * gfortran.dg/largeequiv_1.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=1.335.2.9&r2=1.335.2.10 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans-array.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.39&r2=1.39.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans-array.h.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.7&r2=1.7.18.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans-common.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.23&r2=1.23.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans-decl.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.54&r2=1.54.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fortran/trans.h.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.23&r2=1.23.2.1 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.5084.2.37&r2=1.5084.2.38 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.dg/largeequiv_1.f90.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1
Fixed.