From: "Doug Evans" <dje@google.com>
To: "GDB Patches" <gdb-patches@sourceware.org>
Cc: ARistovski@qnx.com
Subject: Re: [RFA] patch for DW_AT_comp_dir/DW_AT_name vs .debug_line inconsistencies
Date: Wed, 09 Apr 2008 13:33:00 -0000 [thread overview]
Message-ID: <e394668d0804081523o2f87604er7d5a5c62e6870fd4@mail.gmail.com> (raw)
In-Reply-To: <e394668d0804072241l4b92641ck72055b044c688b2c@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 592 bytes --]
On Mon, Apr 7, 2008 at 10:41 PM, Doug Evans <dje@google.com> wrote:
> I have an occasion to revisit this:
> http://sourceware.org/ml/gdb-patches/2007-11/msg00314.html
>
> Or in a different form:
> http://sourceware.org/ml/gdb-patches/2008-01/msg00103.html
>
> When we last left our story, watching for a single path mismatch of
> the main source file only seemed ok. That's what this patch does.
> It's just a minor revision of
> http://sourceware.org/ml/gdb-patches/2008-01/msg00103.html
While reading end_symtab I was inspired to try again for a
debug-format independent solution.
[-- Attachment #2: gdb-080408-buildsym-1.patch.txt --]
[-- Type: text/plain, Size: 10331 bytes --]
2008-04-08 Doug Evans <dje@google.com>
* buildsym.c (watch_main_source_file_lossage): New fn.
(end_symtab): Call it.
* gdb.base/hashline1.exp: New testcase.
* gdb.base/hashline2.exp: New testcase.
* gdb.base/hashline2.exp: New testcase.
Index: buildsym.c
===================================================================
RCS file: /cvs/src/src/gdb/buildsym.c,v
retrieving revision 1.59
diff -u -p -u -p -r1.59 buildsym.c
--- buildsym.c 1 Jan 2008 22:53:09 -0000 1.59
+++ buildsym.c 8 Apr 2008 22:16:06 -0000
@@ -889,6 +889,81 @@ start_symtab (char *name, char *dirname,
start_subfile (name, dirname);
}
+/* Subroutine of end_symtab to simplify it.
+ Look for a subfile that matches the main source file's basename.
+ If there is only one, and if the main source file doesn't have any
+ symbol or line number information, then copy this file's symtab and
+ line_vector to the main source file's subfile and discard the other subfile.
+ This can happen because of a compiler bug or from the user playing games
+ with #line or from things like a distributed build system that manipulates
+ the debug info. */
+
+static void
+watch_main_source_file_lossage ()
+{
+ struct subfile *mainsub, *subfile;
+
+ /* Find the main source file.
+ This loop could be eliminated if start_symtab saved it for us. */
+ mainsub = NULL;
+ for (subfile = subfiles; subfile; subfile = subfile->next)
+ {
+ /* The main subfile is guaranteed to be the last one. */
+ if (subfile->next == NULL)
+ mainsub = subfile;
+ }
+
+ /* If the main source file doesn't have any line number or symbol info,
+ look for an alias in another subfile.
+ We have to watch for mainsub == NULL here. It's a quirk of end_symtab,
+ it can return NULL so there may not be a main subfile. */
+
+ if (mainsub
+ && mainsub->line_vector == NULL
+ && mainsub->symtab == NULL)
+ {
+ const char *mainbase = lbasename (mainsub->name);
+ int nr_matches = 0;
+ struct subfile *prevsub;
+ struct subfile *mainsub_alias = NULL;
+ struct subfile *prev_mainsub_alias = NULL;
+
+ prevsub = NULL;
+ for (subfile = subfiles;
+ /* Stop before we get to the last one. */
+ subfile->next;
+ subfile = subfile->next)
+ {
+ if (strcmp (lbasename (subfile->name), mainbase) == 0)
+ {
+ ++nr_matches;
+ mainsub_alias = subfile;
+ prev_mainsub_alias = prevsub;
+ }
+ prevsub = subfile;
+ }
+
+ if (nr_matches == 1)
+ {
+ gdb_assert (mainsub_alias != NULL && mainsub_alias != mainsub);
+
+ /* Found a match for the main source file.
+ Copy its line_vector and symtab to the main subfile
+ and then discard it. */
+
+ mainsub->line_vector = mainsub_alias->line_vector;
+ mainsub->line_vector_length = mainsub_alias->line_vector_length;
+ mainsub->symtab = mainsub_alias->symtab;
+
+ if (prev_mainsub_alias == NULL)
+ subfiles = mainsub_alias->next;
+ else
+ prev_mainsub_alias->next = mainsub_alias->next;
+ xfree (mainsub_alias);
+ }
+ }
+}
+
/* Finish the symbol definitions for one main source file, close off
all the lexical contexts for that file (creating struct block's for
them), then make the struct symtab for that file and put it in the
@@ -1010,6 +1085,11 @@ end_symtab (CORE_ADDR end_addr, struct o
if (objfile->sf->sym_read_linetable != NULL)
objfile->sf->sym_read_linetable ();
+ /* Handle the case where the debug info specifies a different path
+ for the main source file. It can cause us to lose track of its
+ line number information. */
+ watch_main_source_file_lossage ();
+
/* Now create the symtab objects proper, one for each subfile. */
/* (The main file is the last one on the chain.) */
Index: testsuite/gdb.base/hashline1.exp
===================================================================
RCS file: testsuite/gdb.base/hashline1.exp
diff -N testsuite/gdb.base/hashline1.exp
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.base/hashline1.exp 8 Apr 2008 22:16:06 -0000
@@ -0,0 +1,57 @@
+# Copyright 2008 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+# This file is part of the gdb testsuite.
+
+# Test loading of line number information with absolute path in #line, bug 2360.
+
+if $tracelevel then {
+ strace $tracelevel
+}
+
+set prms_id 2360
+set bug_id 0
+
+# srcfile is in objdir because we need to machine generate it in order
+# to get the correct path in the #line directive.
+
+set testfile "hashline1"
+set srcfile "${testfile}.c"
+set binfile "${objdir}/${subdir}/${testfile}"
+
+set fd [open ${objdir}/${subdir}/${srcfile} w]
+puts $fd "#line 2 \"[pwd]/${subdir}/${srcfile}\""
+puts $fd "int main () { return 0; } /* set breakpoint here */"
+close $fd
+
+# The choice of path name for the source file is important in order to trigger
+# the bug. Using ${objdir}/${subdir}/${srcfile} here won't trigger the bug.
+if { [gdb_compile "./${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
+ untested hashline1.exp
+ return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+set bp_location [gdb_get_line_number "set breakpoint here" ${objdir}/${subdir}/${srcfile}]
+
+# Try to set a breakpoint on the specified file location.
+
+gdb_test "break $srcfile:$bp_location" \
+ "Breakpoint.*at.* file .*$srcfile, line.*" \
+ "set breakpoint"
Index: testsuite/gdb.base/hashline2.exp
===================================================================
RCS file: testsuite/gdb.base/hashline2.exp
diff -N testsuite/gdb.base/hashline2.exp
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.base/hashline2.exp 8 Apr 2008 22:16:06 -0000
@@ -0,0 +1,56 @@
+# Copyright 2008 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+# This file is part of the gdb testsuite.
+
+# Test loading of line number information with an absolute path with extra
+# /'s in #line, bug 2360.
+
+if $tracelevel then {
+ strace $tracelevel
+}
+
+set prms_id 2360
+set bug_id 0
+
+# srcfile is in objdir because we need to machine generate it in order
+# to get the correct path in the #line directive.
+
+set testfile "hashline2"
+set srcfile "${testfile}.c"
+set binfile "${objdir}/${subdir}/${testfile}"
+
+set fd [open ${objdir}/${subdir}/${srcfile} w]
+puts $fd "#line 2 \"///[pwd]/${subdir}/${srcfile}\""
+puts $fd "int main () { return 0; } /* set breakpoint here */"
+close $fd
+
+if { [gdb_compile "${objdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
+ untested hashline1.exp
+ return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+set bp_location [gdb_get_line_number "set breakpoint here" ${objdir}/${subdir}/${srcfile}]
+
+# Try to set a breakpoint on the specified file location.
+
+gdb_test "break $srcfile:$bp_location" \
+ "Breakpoint.*at.* file .*$srcfile, line.*" \
+ "set breakpoint"
Index: testsuite/gdb.base/hashline3.exp
===================================================================
RCS file: testsuite/gdb.base/hashline3.exp
diff -N testsuite/gdb.base/hashline3.exp
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.base/hashline3.exp 8 Apr 2008 22:16:06 -0000
@@ -0,0 +1,55 @@
+# Copyright 2008 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+# This file is part of the gdb testsuite.
+
+# Test loading of line number information with relative path in #line, bug 2360.
+
+if $tracelevel then {
+ strace $tracelevel
+}
+
+set prms_id 2360
+set bug_id 0
+
+# srcfile is in objdir because we need to machine generate it in order
+# to get the correct path in the #line directive.
+
+set testfile "hashline3"
+set srcfile "${testfile}.c"
+set binfile "${objdir}/${subdir}/${testfile}"
+
+set fd [open ${objdir}/${subdir}/${srcfile} w]
+puts $fd "#line 2 \"./${subdir}/${srcfile}\""
+puts $fd "int main () { return 0; } /* set breakpoint here */"
+close $fd
+
+if { [gdb_compile "${objdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
+ untested hashline1.exp
+ return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+set bp_location [gdb_get_line_number "set breakpoint here" ${objdir}/${subdir}/${srcfile}]
+
+# Try to set a breakpoint on the specified file location.
+
+gdb_test "break $srcfile:$bp_location" \
+ "Breakpoint.*at.* file .*$srcfile, line.*" \
+ "set breakpoint"
next prev parent reply other threads:[~2008-04-08 22:24 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-08 9:29 Doug Evans
2008-04-09 13:33 ` Doug Evans [this message]
2008-04-17 16:21 ` Daniel Jacobowitz
2008-04-17 17:28 ` Doug Evans
2008-04-17 17:37 ` Daniel Jacobowitz
-- strict thread matches above, loose matches on Subject: below --
2008-01-07 2:25 Doug Evans
2008-01-05 22:18 Doug Evans
2008-01-05 22:49 ` Doug Evans
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=e394668d0804081523o2f87604er7d5a5c62e6870fd4@mail.gmail.com \
--to=dje@google.com \
--cc=ARistovski@qnx.com \
--cc=gdb-patches@sourceware.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox