* [RFA/testsuite] stack check
@ 2008-09-05 10:53 Jerome Guitton
2008-09-06 2:03 ` Joel Brobecker
0 siblings, 1 reply; 11+ messages in thread
From: Jerome Guitton @ 2008-09-05 10:53 UTC (permalink / raw)
To: gdb-patches
[-- Attachment #1: Type: text/plain, Size: 736 bytes --]
Follow-up to http://sourceware.org/ml/gdb-patches/2008-09/msg00007.html
Test that the different possible stack checking sequences are properly
skipped when analyzing a prologue on PPC AIX.
OK to apply?
2008-09-05 Jerome Guitton <guitton@adacore.com>
* powerpc-aix-prologue.c (stack_check_probe_1, stack_check_probe_2)
(stack_check_probe_loop_1, stack_check_probe_loop_2): New functions.
(main): Add call to these new functions.
* powerpc-aix-prologue.exp: When breaking on these functions, check
that the breakpoint is inserted at the appropriate location.
(insert_breakpoint): Slightly refine this procedure so that it can
be called several times in the test.
Thanks,
Jerome
[-- Attachment #2: powerpc-aix-prologue.diff --]
[-- Type: text/x-diff, Size: 5685 bytes --]
Index: testsuite/gdb.arch/powerpc-aix-prologue.exp
===================================================================
--- testsuite/gdb.arch/powerpc-aix-prologue.exp (revision 134842)
+++ testsuite/gdb.arch/powerpc-aix-prologue.exp (working copy)
@@ -55,7 +55,7 @@ proc insert_breakpoint {function expecte
# Insert a breakpoint using the given function name, and extract
# the breakpoint address for the output.
gdb_test_multiple "break $function" "set breakpoint in $function" {
- -re "Breakpoint 1 at ($hex).*$gdb_prompt $" {
+ -re "Breakpoint .* at ($hex).*$gdb_prompt $" {
set address $expect_out(1,string)
}
default {
@@ -79,3 +79,8 @@ proc insert_breakpoint {function expecte
}
insert_breakpoint "li_stw" 12
+insert_breakpoint "stack_check_probe_1" 16
+insert_breakpoint "stack_check_probe_2" 40
+insert_breakpoint "stack_check_probe_loop_1" 68
+insert_breakpoint "stack_check_probe_loop_2" 60
+
Index: testsuite/gdb.arch/powerpc-aix-prologue.c
===================================================================
--- testsuite/gdb.arch/powerpc-aix-prologue.c (revision 134842)
+++ testsuite/gdb.arch/powerpc-aix-prologue.c (working copy)
@@ -17,10 +17,19 @@
void li_stw (void);
+void stack_check_probe_1 (void);
+void stack_check_probe_2 (void);
+void stack_check_probe_loop_1 (void);
+void stack_check_probe_loop_2 (void);
+
int
main (void)
{
li_stw ();
+ stack_check_probe_1 ();
+ stack_check_probe_2 ();
+ stack_check_probe_loop_1 ();
+ stack_check_probe_loop_2 ();
return 0;
}
@@ -51,3 +60,125 @@ asm (" .csect .text[PR]\n"
" lwz 31,-4(1)\n"
" blr\n");
+/* Asm for procedure stack_check_probe_1().
+
+ The purpose of this function is to verify that GDB can skip the stack
+ checking probing at the beginning of the prologue. */
+
+asm (" .csect .text[PR]\n"
+ " .align 2\n"
+ " .globl stack_check_probe_1\n"
+ " .globl .stack_check_probe_1\n"
+ " .csect stack_check_probe_1[DS]\n"
+ "stack_check_probe_1:\n"
+ " .long .stack_check_probe_1, TOC[tc0], 0\n"
+ " .csect .text[PR]\n"
+ ".stack_check_probe_1:\n"
+ " stw 0,-12336(1)\n"
+ " stw 31,-4(1)\n"
+ " stwu 1,-48(1)\n"
+ " mr 31,1\n"
+ " lwz 1,0(1)\n"
+ " lwz 31,-4(1)\n"
+ " blr\n");
+
+/* Asm for procedure stack_check_probe_2 ().
+
+ Similar to stack_check_probe_1, but with a different probing sequence
+ (several probes). */
+
+asm (" .csect .text[PR]\n"
+ " .align 2\n"
+ " .globl stack_check_probe_2\n"
+ " .globl .stack_check_probe_2\n"
+ " .csect stack_check_probe_2[DS]\n"
+ "stack_check_probe_2:\n"
+ " .long .stack_check_probe_2, TOC[tc0], 0\n"
+ " .csect .text[PR]\n"
+ ".stack_check_probe_2:\n"
+ " stw 0,-16384(1)\n"
+ " stw 0,-20480(1)\n"
+ " stw 0,-24576(1)\n"
+ " stw 0,-28672(1)\n"
+ " stw 0,-28752(1)\n"
+ " mflr 0\n"
+ " stw 31,-4(1)\n"
+ " stw 0,8(1)\n"
+ " stwu 1,-16464(1)\n"
+ " mr 31,1\n"
+ " lwz 1,0(1)\n"
+ " lwz 0,8(1)\n"
+ " mtlr 0\n"
+ " lwz 31,-4(1)\n"
+ " blr\n");
+
+/* Asm for procedure stack_check_probe_loop_1() and stack_check_probe_loop_2().
+
+ Similar to stack_check_probe_1, but with a different probing sequence
+ (probing loop). */
+
+asm (" .csect .text[PR]\n"
+ " .align 2\n"
+ " .globl stack_check_probe_loop_1\n"
+ " .globl .stack_check_probe_loop_1\n"
+ " .csect stack_check_probe_loop_1[DS]\n"
+ "stack_check_probe_loop_1:\n"
+ " .long .stack_check_probe_loop_1, TOC[tc0], 0\n"
+ " .csect .text[PR]\n"
+ ".stack_check_probe_loop_1:\n"
+ " addi 12,1,-12288\n"
+ " lis 0,-8\n"
+ " ori 0,0,4096\n"
+ " add 0,12,0\n"
+ "LPSRL1..0:\n"
+ " cmpw 0,12,0\n"
+ " beq 0,LPSRE1..0\n"
+ " addi 12,12,-4096\n"
+ " stw 0,0(12)\n"
+ " b LPSRL1..0\n"
+ "LPSRE1..0:\n"
+ " stw 0,-4080(12)\n"
+ " mflr 0\n"
+ " stw 31,-4(1)\n"
+ " stw 0,8(1)\n"
+ " lis 0,0xfff8\n"
+ " ori 0,0,16\n"
+ " stwux 1,1,0\n"
+ " mr 31,1\n"
+ " lwz 1,0(1)\n"
+ " lwz 0,8(1)\n"
+ " mtlr 0\n"
+ " lwz 31,-4(1)\n"
+ " blr\n");
+
+asm (" .csect .text[PR]\n"
+ " .align 2\n"
+ " .globl stack_check_probe_loop_2\n"
+ " .globl .stack_check_probe_loop_2\n"
+ " .csect stack_check_probe_loop_2[DS]\n"
+ "stack_check_probe_loop_2:\n"
+ " .long .stack_check_probe_loop_2, TOC[tc0], 0\n"
+ " .csect .text[PR]\n"
+ ".stack_check_probe_loop_2:\n"
+ " addi 12,1,-12288\n"
+ " lis 0,-8\n"
+ " add 0,12,0\n"
+ "LPSRL2..0:\n"
+ " cmpw 0,12,0\n"
+ " beq 0,LPSRE2..0\n"
+ " addi 12,12,-4096\n"
+ " stw 0,0(12)\n"
+ " b LPSRL2..0\n"
+ "LPSRE2..0:\n"
+ " mflr 0\n"
+ " stw 31,-4(1)\n"
+ " stw 0,8(1)\n"
+ " lis 0,0xfff8\n"
+ " ori 0,0,16\n"
+ " stwux 1,1,0\n"
+ " mr 31,1\n"
+ " lwz 1,0(1)\n"
+ " lwz 0,8(1)\n"
+ " mtlr 0\n"
+ " lwz 31,-4(1)\n"
+ " blr\n");
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: [RFA/testsuite] stack check
2008-09-05 10:53 [RFA/testsuite] stack check Jerome Guitton
@ 2008-09-06 2:03 ` Joel Brobecker
2008-09-08 16:05 ` Jerome Guitton
0 siblings, 1 reply; 11+ messages in thread
From: Joel Brobecker @ 2008-09-06 2:03 UTC (permalink / raw)
To: Jerome Guitton; +Cc: gdb-patches
> 2008-09-05 Jerome Guitton <guitton@adacore.com>
>
> * powerpc-aix-prologue.c (stack_check_probe_1, stack_check_probe_2)
> (stack_check_probe_loop_1, stack_check_probe_loop_2): New functions.
> (main): Add call to these new functions.
> * powerpc-aix-prologue.exp: When breaking on these functions, check
> that the breakpoint is inserted at the appropriate location.
> (insert_breakpoint): Slightly refine this procedure so that it can
> be called several times in the test.
Nice :). The only issue is that the ChangeLog is inside gdb/testsuite,
so you need to prefix the file names with gdb.arch. In other words:
gdb.arch/powerpc-aix-prologue.c for instance. The patch itself is OK.
--
Joel
^ permalink raw reply [flat|nested] 11+ messages in thread
* [RFA/testsuite] stack check
@ 2008-09-01 17:03 Jerome Guitton
2008-09-01 17:14 ` Jerome Guitton
2008-09-02 21:02 ` Joel Brobecker
0 siblings, 2 replies; 11+ messages in thread
From: Jerome Guitton @ 2008-09-01 17:03 UTC (permalink / raw)
To: gdb-patches
[-- Attachment #1: Type: text/plain, Size: 323 bytes --]
Test that GDB is able to get a correct backtrace when the application
is compiled with -fstack-check. Related to:
http://sourceware.org/ml/gdb-patches/2008-09/msg00005.html
2008-09-01 Jerome Guitton <guitton@adacore.com>
* (stack-checking.c, stack-checking.exp): New testcase.
OK to apply?
Thanks,
- Jerome
[-- Attachment #2: stack-checking.c --]
[-- Type: text/x-csrc, Size: 389 bytes --]
/*
Test file to be compiled with -fstack-check, for testing "bt" against
different stack checking prologue sequences.
*/
int i = 0;
void
small_frame ()
{
i++; /* set breakpoint here */
}
void medium_frame ()
{
char S [16384];
small_frame ();
}
void big_frame ()
{
char S [524188];
small_frame ();
}
void
main ()
{
small_frame ();
medium_frame ();
big_frame ();
}
[-- Attachment #3: stack-checking.exp --]
[-- Type: text/plain, Size: 1665 bytes --]
# 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/>.
if $tracelevel then {
strace $tracelevel
}
set prms_id 0
set bug_id 0
set testfile "stack-checking"
set srcfile ${testfile}.c
set binfile ${objdir}/${subdir}/${testfile}
if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable [concat debug optimize=-fstack-check nowarnings]] != "" } {
untested stack-checking.exp
return -1
}
gdb_exit
gdb_start
gdb_reinitialize_dir $srcdir/$subdir
gdb_load ${binfile}
if ![runto_main] then {
fail "Can't run to main"
return 0
}
# Test that the debugger is able to unwind to "main" at the next breakpoint
# hit. The frame number of "main" should be passed into MAIN_FRAME_NUMBER."
proc backtrace_test {name main_frame_number} {
gdb_continue_to_breakpoint "breakpoint for $name"
gdb_test "bt" "#${main_frame_number}.*in main ().*"
}
gdb_breakpoint [gdb_get_line_number "set breakpoint here"]
backtrace_test "small frame" "1"
backtrace_test "medium frame" "2"
backtrace_test "big frame" "2"
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: [RFA/testsuite] stack check
2008-09-01 17:03 Jerome Guitton
@ 2008-09-01 17:14 ` Jerome Guitton
2008-09-02 21:03 ` Joel Brobecker
2008-09-02 21:02 ` Joel Brobecker
1 sibling, 1 reply; 11+ messages in thread
From: Jerome Guitton @ 2008-09-01 17:14 UTC (permalink / raw)
To: gdb-patches
Jerome Guitton (guitton@adacore.com):
>
> Test that GDB is able to get a correct backtrace when the application
> is compiled with -fstack-check.
BTW, I have a similar testcase in PPC assembly. It checks that the
different stack-checking sequences that GCC may emit for this target
are understood by GDB. It could be merged into
gdb.arch/powerpc-aix-prologue.exp. If it is of any help, I'll be happy
to post it in this list.
Thanks,
- Jerome
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [RFA/testsuite] stack check
2008-09-01 17:14 ` Jerome Guitton
@ 2008-09-02 21:03 ` Joel Brobecker
0 siblings, 0 replies; 11+ messages in thread
From: Joel Brobecker @ 2008-09-02 21:03 UTC (permalink / raw)
To: Jerome Guitton; +Cc: gdb-patches
> BTW, I have a similar testcase in PPC assembly. It checks that the
> different stack-checking sequences that GCC may emit for this target
> are understood by GDB. It could be merged into
> gdb.arch/powerpc-aix-prologue.exp. If it is of any help, I'll be happy
> to post it in this list.
Yes, I think that would be useful as well. It would allow us to make
sure that we test all variants of the stack-check code...
--
Joel
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [RFA/testsuite] stack check
2008-09-01 17:03 Jerome Guitton
2008-09-01 17:14 ` Jerome Guitton
@ 2008-09-02 21:02 ` Joel Brobecker
2008-09-03 8:53 ` Jerome Guitton
2008-09-08 10:17 ` Jerome Guitton
1 sibling, 2 replies; 11+ messages in thread
From: Joel Brobecker @ 2008-09-02 21:02 UTC (permalink / raw)
To: Jerome Guitton; +Cc: gdb-patches
> 2008-09-01 Jerome Guitton <guitton@adacore.com>
>
> * (stack-checking.c, stack-checking.exp): New testcase.
OK, except for a little nit: You forgot to add a copyright header
in the C file...
Also:
> if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable [concat debug optimize=-fstack-check nowarnings]] != "" } {
I would think that the "optimize" flag was only meant to force a certain
optimization level. Perhaps you should use additional_flags instead.
--
Joel
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2008-09-08 16:05 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-09-05 10:53 [RFA/testsuite] stack check Jerome Guitton
2008-09-06 2:03 ` Joel Brobecker
2008-09-08 16:05 ` Jerome Guitton
-- strict thread matches above, loose matches on Subject: below --
2008-09-01 17:03 Jerome Guitton
2008-09-01 17:14 ` Jerome Guitton
2008-09-02 21:03 ` Joel Brobecker
2008-09-02 21:02 ` Joel Brobecker
2008-09-03 8:53 ` Jerome Guitton
2008-09-08 10:17 ` Jerome Guitton
2008-09-08 15:42 ` Joel Brobecker
2008-09-08 15:56 ` Jerome Guitton
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox