rpms/gdb/F-12 gdb-bz559291-fortran-module-toplevel.patch, NONE, 1.1 gdb.spec, 1.423, 1.424
Jan Kratochvil
jkratoch at fedoraproject.org
Wed Feb 3 01:11:13 UTC 2010
Author: jkratoch
Update of /cvs/pkgs/rpms/gdb/F-12
In directory cvs1.fedora.phx.redhat.com:/tmp/cvs-serv27847
Modified Files:
gdb.spec
Added Files:
gdb-bz559291-fortran-module-toplevel.patch
Log Message:
* Wed Feb 3 2010 Jan Kratochvil <jan.kratochvil at redhat.com> - 7.0.1-31.fc12
- Fortran: Fix regression on setting breakpoint at toplevel symbols (BZ 559291).
gdb-bz559291-fortran-module-toplevel.patch:
f-lang.c | 48 +++++++++++++++++++--------------------
testsuite/gdb.fortran/module.exp | 10 ++++++++
testsuite/gdb.fortran/module.f90 | 2 +
3 files changed, 36 insertions(+), 24 deletions(-)
--- NEW FILE gdb-bz559291-fortran-module-toplevel.patch ---
http://sourceware.org/ml/archer/2010-q1/msg00047.html
Subject: [fortran-module] [commit] Fix a regression on setting breakpoint at
[ Backport for F-12. ]
f5c7672a52316155bc3367cbc2f0e7db22523634
https://bugzilla.redhat.com/show_bug.cgi?id=559291
gdb/
2010-02-03 Jan Kratochvil <jan.kratochvil at redhat.com>
David Moore <david.moore at intel.com>
* f-lang.c (f_lookup_symbol_nonlocal): Always fallback on
basic_lookup_symbol_nonlocal.
gdb/testsuite/
2010-02-03 Jan Kratochvil <jan.kratochvil at redhat.com>
* gdb.fortran/module.exp (show language, setting breakpoint at module):
New.
* gdb.fortran/module.f90: New statement program module.
--- ./gdb/f-lang.c 2010-02-03 01:50:07.000000000 +0100
+++ ./gdb/f-lang.c 2010-02-03 01:49:06.000000000 +0100
@@ -308,33 +308,33 @@ f_lookup_symbol_nonlocal (const char *na
const struct block *block,
const domain_enum domain)
{
- struct fortran_using *use;
-
- if (!block)
- return NULL;
-
- for (use = BLOCK_FORTRAN_USE (block); use; use = use->next)
+ if (block)
{
- struct symbol *sym;
- struct type *type;
- struct symbol *retval;
-
- sym = lookup_symbol_global (use->module_name, block, MODULE_DOMAIN);
-
- /* Module name lookup should not fail with correct debug info. */
- if (sym == NULL)
- continue;
-
- type = SYMBOL_TYPE (sym);
- gdb_assert (TYPE_CODE (type) == TYPE_CODE_MODULE);
- gdb_assert (TYPE_MODULE_BLOCK (type) != NULL);
-
- retval = lookup_block_symbol (TYPE_MODULE_BLOCK (type), name, domain);
- if (retval)
- return retval;
+ struct fortran_using *use;
+
+ for (use = BLOCK_FORTRAN_USE (block); use; use = use->next)
+ {
+ struct symbol *sym;
+ struct type *type;
+ struct symbol *retval;
+
+ sym = lookup_symbol_global (use->module_name, block, MODULE_DOMAIN);
+
+ /* Module name lookup should not fail with correct debug info. */
+ if (sym == NULL)
+ continue;
+
+ type = SYMBOL_TYPE (sym);
+ gdb_assert (TYPE_CODE (type) == TYPE_CODE_MODULE);
+ gdb_assert (TYPE_MODULE_BLOCK (type) != NULL);
+
+ retval = lookup_block_symbol (TYPE_MODULE_BLOCK (type), name, domain);
+ if (retval)
+ return retval;
+ }
}
- return NULL;
+ return basic_lookup_symbol_nonlocal (name, block, domain);
}
/* This is declared in c-lang.h but it is silly to import that file for what
--- ./gdb/testsuite/gdb.fortran/module.exp 2010-02-03 01:50:13.000000000 +0100
+++ ./gdb/testsuite/gdb.fortran/module.exp 2010-02-03 01:47:42.000000000 +0100
@@ -43,3 +43,13 @@ gdb_test "print var_b" " = 11"
gdb_test "print var_c" "No symbol \"var_c\" in current context\\."
gdb_test "print var_d" " = 12"
gdb_test "print var_i" " = 14" "print var_i value 14"
+
+# Breakpoint would work in language "c".
+gdb_test "show language" {The current source language is "(auto; currently )?fortran".}
+
+# gcc-4.4.2: The main program is always MAIN__ in .symtab so "runto" above
+# works. But DWARF DW_TAG_subprogram contains the name specified by
+# the "program" Fortran statement.
+if [gdb_breakpoint "module"] {
+ pass "setting breakpoint at module"
+}
--- ./gdb/testsuite/gdb.fortran/module.f90 2010-02-03 01:50:07.000000000 +0100
+++ ./gdb/testsuite/gdb.fortran/module.f90 2010-02-03 01:47:42.000000000 +0100
@@ -37,6 +37,8 @@ end module modmany
var_i = var_i ! i-is-2
end
+ program module
+
use modmany, only: var_b, var_d => var_c, var_i
call sub1
Index: gdb.spec
===================================================================
RCS file: /cvs/pkgs/rpms/gdb/F-12/gdb.spec,v
retrieving revision 1.423
retrieving revision 1.424
diff -u -p -r1.423 -r1.424
--- gdb.spec 31 Jan 2010 02:00:42 -0000 1.423
+++ gdb.spec 3 Feb 2010 01:11:13 -0000 1.424
@@ -36,7 +36,7 @@ Version: 7.0.1
# The release always contains a leading reserved number, start it at 1.
# `upstream' is not a part of `name' to stay fully rpm dependencies compatible for the testing.
-Release: 30%{?_with_upstream:.upstream}%{dist}
+Release: 31%{?_with_upstream:.upstream}%{dist}
License: GPLv3+
Group: Development/Debuggers
@@ -447,6 +447,9 @@ Patch402: gdb-python-cplus-crash.patch
# Fix failed gdb_assert due to the PIE patchset (BZ 559414).
Patch414: gdb-bz559414-pie-assert-fix.patch
+# Fortran: Fix regression on setting breakpoint at toplevel symbols (BZ 559291).
+Patch416: gdb-bz559291-fortran-module-toplevel.patch
+
BuildRequires: ncurses-devel%{?_isa} texinfo gettext flex bison expat-devel%{?_isa}
Requires: readline%{?_isa}
BuildRequires: readline-devel%{?_isa}
@@ -699,6 +702,7 @@ rm -f gdb/jv-exp.c gdb/m2-exp.c gdb/objc
%patch401 -p1
%patch402 -p1
%patch414 -p1
+%patch416 -p1
# Always verify their applicability.
%patch393 -p1
%patch335 -p1
@@ -1024,6 +1028,9 @@ fi
%endif
%changelog
+* Wed Feb 3 2010 Jan Kratochvil <jan.kratochvil at redhat.com> - 7.0.1-31.fc12
+- Fortran: Fix regression on setting breakpoint at toplevel symbols (BZ 559291).
+
* Sun Jan 31 2010 Jan Kratochvil <jan.kratochvil at redhat.com> - 7.0.1-30.fc12
- Fix failed gdb_assert due to the PIE patchset (BZ 559414).
More information about the scm-commits
mailing list