From: Andrew Cagney <cagney@gnu.org>
To: Randolph Chung <randolph@tausq.org>
Cc: gdb-patches@sources.redhat.com
Subject: Re: [patch/RFA] multiarch INSTRUCTION_NULLIFIED
Date: Fri, 03 Dec 2004 18:11:00 -0000 [thread overview]
Message-ID: <41B0ABC7.4020806@gnu.org> (raw)
In-Reply-To: <20041202052417.GM6359@tausq.org>
Randolph Chung wrote:
>>PPS: A gdb.arch/ addition to tickle the basic edge case would be a
>>helpful way of capturing this knowledge.
>
>
> how about this... my first gdb testcase, so hopefully i didn't do
> anything too wrong :)
>
> 74 PASSes, 3 FAILs with the old code (assuming the instruction_nullified
> logic is removed). 77 PASSes with my previous patch.
>
> randolph
>
>
> 2004-12-01 Randolph Chung <tausq@debian.org>
>
> * gdb.arch/pa-nullify.exp: New file.
> * gdb.arch/pa-nullify.c: New file.
> +send_gdb "print foo\n"
> +gdb_expect {
> + -re ".* = {int \\(int\\)} ($hex) <foo>.*$gdb_prompt $" {
> + set baddr $expect_out(1,string)
> + verbose "Breakpoint address at $baddr"
> + pass "print foo = $baddr"
> + }
> + default {
> + fail "print foo"
> + gdb_suppress_tests
> + } timeout {
> + fail "cannot get address of foo (timed out)."
> + gdb_suppress_tests
> + }
> +}
I'd use gdb_test_multiple: it's more robust and handles more cases when
things go wrong; and we're trying to eliminate the old style (yes, there
is no marker to warn you of this, sigh). It should be a drop-in
replacement and easy to fix.
Anyway, more relevant, and as daniel asked, can it be done in assember?
A starting point for that might be the gdb.asm test case which uses
assembly source code.
Andrew
> +gdb_breakpoint "*$baddr"
> +
> +gdb_test "display /i \$pc" "1: x/i +\\\$pc.*" "display /i \$pc"
> +gdb_test "continue" ".*Breakpoint \[0-9\], foo \\(x=1\\) at .*pa-nullify.c:23.*" "break at first insn of foo"
> +
> +set done 0
> +set iter 0
> +while { $done == 0 && $iter < 20 } {
> + send_gdb "backtrace\n"
> + gdb_expect {
> + -re ".*foo \\(x=.*\\).* main \\(.*\\).*$gdb_prompt $" {
> + pass "backtrace from foo"
> + }
> + -re "$gdb_prompt $" {
> + fail "backtrace from foo"
> + gdb_suppress_tests
> + }
> + timeout {
> + fail "(timeout) backtrace from foo"
> + gdb_suppress_tests
> + }
> + }
> +
> + send_gdb "stepi\n"
> + gdb_expect {
> + -re ".*: x/i +\\\$pc.*:\[ \t\]+(.*)\r.*$gdb_prompt $" {
> + set insn $expect_out(1,string)
> + pass "x/i \$pc = $insn"
> + } timeout {
> + fail "(timeout) x/i \$pc"
> + gdb_suppress_tests
> + break
> + }
> + }
> +
> + incr iter
> +
> + if { $insn == "bv,n r0(rp)" } {
> + pass "found bv,n insn"
> + set done 1
> + }
> +}
> +
> +send_gdb "stepi\n"
> +gdb_expect {
> + -re ".*bv,n r0\\(rp\\)\r.*$gdb_prompt $" {
> + pass "restep on nullified instruction"
> + }
> + -re ".*bar.*\r.*x/i.*bar.*\r.*$gdb_prompt $" {
> + fail "step on nullification instruction ended in next function"
> + }
> + timeout {
> + fail "(timeout) restep on nullified instruction"
> + }
> +}
> +incr iter
> +
> +# Lastly, we do a single step through each instruction and generate a core file.
> +# Then, load the core file and verify that we can get a proper backtrace.
> +
> +if ![isnative] then {
> + verbose "skipping core file tests on non-native configuration"
> + return
> +}
> +
> +send_gdb "undisplay 1\n"
> +
> +proc do_core_test { count } {
> + global gcorefile
> + global gdb_prompt
> + global hex
> +
> + if { ! [ runto_main ] } then { gdb_suppress_tests; }
> +
> + set baddr "0"
> + send_gdb "print foo\n"
> + gdb_expect {
> + -re ".* = {int \\(int\\)} ($hex) <foo>.*$gdb_prompt $" {
> + set baddr $expect_out(1,string)
> + verbose "Breakpoint address at $baddr"
> + pass "print foo = $baddr #$count"
> + }
> + default {
> + fail "print foo #$count"
> + gdb_suppress_tests
> + }
> + timeout {
> + fail "cannot get address of foo (timed out). #$count"
> + gdb_suppress_tests
> + }
> + }
> + gdb_breakpoint "*$baddr"
> +
> + gdb_test "continue" ".*Breakpoint \[0-9\]*, foo \\(x=1\\) at .*pa-nullify.c:23.*" "break at first insn of foo #$count"
> +
> + gdb_test "si $count"
> +
> + set gcore_works 0
> + set escapedfilename [string_to_regexp $gcorefile]
> + gdb_test_multiple "gcore $gcorefile" "gcore" {
> + -re "Saved corefile ${escapedfilename}\[\r\n\]+$gdb_prompt $" {
> + pass "gcore #$count"
> + set gcore_works 1
> + }
> + -re "Can't create a corefile\[\r\n\]+$gdb_prompt $" {
> + unsupported "gcore"
> + }
> + }
> +
> + if { $gcore_works } {
> + gdb_test "core $gcorefile" "Core was generated by.*" \
> + "load core file #$count" "A program is being debugged already.*" "y"
> +
> + send_gdb "backtrace\n"
> + gdb_expect {
> + -re ".*foo \\(x=.*\\).* main \\(.*\\).*$gdb_prompt $" {
> + pass "backtrace from foo #$count"
> + }
> + -re "$gdb_prompt $" {
> + fail "backtrace from foo #$count"
> + gdb_suppress_tests
> + }
> + timeout {
> + fail "(timeout) backtrace from foo in gcore #$count"
> + gdb_suppress_tests
> + }
> + }
> + }
> +}
> +
> +for { set i 1 } { $i <= $iter } {incr i} {
> + do_core_test $i
> +}
> --- /dev/null 2004-08-25 14:59:25.000000000 -0700
> +++ pa-nullify.c 2004-12-01 21:19:51.324347136 -0800
> @@ -0,0 +1,30 @@
> +/* PA nullification test program.
> +
> + Copyright 2004 Free Software Foundation, Inc.
> +
> + This file is part of GDB.
> +
> + 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 2 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, write to the Free Software
> + Foundation, Inc., 59 Temple Place - Suite 330,
> + Boston, MA 02111-1307, USA. */
> +
> +int foo(int x)
> +{ return x; }
> +
> +void bar(void) { }
> +
> +int main(int argc, char **argv)
> +{
> + return foo(argc);
> +}
>
next prev parent reply other threads:[~2004-12-03 18:11 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-11-18 0:02 Randolph Chung
2004-11-18 14:26 ` Andrew Cagney
2004-11-18 16:21 ` Randolph Chung
2004-11-18 16:56 ` Mark Kettenis
2004-11-19 9:25 ` Orjan Friberg
2004-11-23 17:50 ` Randolph Chung
2004-11-23 19:33 ` Mark Kettenis
2004-11-28 17:26 ` Andrew Cagney
2004-11-28 18:41 ` Randolph Chung
2004-11-28 19:55 ` Andrew Cagney
2004-11-29 3:30 ` Randolph Chung
2004-11-29 15:12 ` Andrew Cagney
2004-11-30 6:56 ` Randolph Chung
2004-11-30 14:51 ` Andrew Cagney
2004-11-30 16:44 ` Randolph Chung
2004-11-30 16:59 ` Andrew Cagney
2004-11-30 17:38 ` Randolph Chung
2004-12-01 21:29 ` Andrew Cagney
2004-12-01 22:33 ` Randolph Chung
2004-12-01 23:32 ` Andrew Cagney
2004-12-02 5:24 ` Randolph Chung
2004-12-02 14:27 ` Daniel Jacobowitz
2004-12-03 18:11 ` Andrew Cagney [this message]
2004-12-03 18:15 ` Randolph Chung
2004-12-03 18:57 ` Daniel Jacobowitz
2004-12-03 19:57 ` Randolph Chung
2004-12-03 21:40 ` Randolph Chung
2004-12-03 21:58 ` Andrew Cagney
2004-12-03 22:52 ` Daniel Jacobowitz
2004-12-04 0:00 ` Randolph Chung
2004-12-04 0:55 ` Randolph Chung
2004-12-04 11:27 ` Mark Kettenis
2004-12-01 6:19 ` Randolph Chung
2004-12-01 17:11 ` Daniel Jacobowitz
2004-12-01 17:17 ` Randolph Chung
2004-12-01 17:19 ` Daniel Jacobowitz
2004-12-01 17:25 ` Randolph Chung
2004-12-01 17:28 ` Daniel Jacobowitz
2004-12-01 17:30 ` Randolph Chung
2004-12-01 17:35 ` Randolph Chung
2004-12-01 18:14 ` Randolph Chung
2004-12-01 21:25 ` Andrew Cagney
2004-12-01 13:32 Paul Schlie
2004-12-01 16:25 Randolph Chung
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=41B0ABC7.4020806@gnu.org \
--to=cagney@gnu.org \
--cc=gdb-patches@sources.redhat.com \
--cc=randolph@tausq.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