From: Joel Brobecker <brobecker@adacore.com>
To: Richard Henderson <rth@redhat.com>
Cc: gdb-patches@gcc.gnu.org
Subject: Re: [testsuite/alpha] Add test for step over fbne instruction
Date: Thu, 26 May 2005 03:14:00 -0000 [thread overview]
Message-ID: <20050526023539.GL1525@adacore.com> (raw)
In-Reply-To: <20050525064819.GA5406@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 1112 bytes --]
> > Thanks for the couple of suggestions. Is the attached better?
>
> Yes, though I'm confused as to why you didn't use the 4 or 5 insn
> hand-coded routine that I gave you, which is more like what you would
> get from the compiler with -O2 -ffast-math.
Because I'm lazy? ;-). Seriously, that was a good suggestion, and I have
now used it in this iteration of the testcase. I took this opportunity
to actually test all forms of FP branch instructions, with branch taken
and branch not taken. That way, that should cover most cases. I left
the edge cases for another time.
2005-05-20 Joel Brobecker <brobecker@adacore.com>
* gdb.arch/alpha-step.c: New file.
* gdb.arch/alpha-step.exp: New testcase.
Tested on alpha-tru64 5.1a. We now have 21 FAILs revealed by this testcase.
Before the following patch was applied, we had 30.
http://sources.redhat.com/ml/gdb-patches/2005-05/msg00479.html
I used an ancient version of GDB, before the switch to software single-step
was made, and I have an all PASS.
Will investigate the FAILs as soon as possible.
OK to apply?
Thanks,
--
Joel
[-- Attachment #2: alpha-step.c --]
[-- Type: text/plain, Size: 3061 bytes --]
/* Copyright 2005 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. */
void gt (double a);
int
main (void)
{
gt (360.0);
gt (-360.0);
ge (360.0);
ge (-360.0);
lt (-360.0);
lt (360.0);
le (-360.0);
le (360.0);
eq (0.0);
eq (360.0);
ne (360.0);
ne (0.0);
return 0;
}
asm (" .text\n"
" .ent gt\n"
"gt:\n"
" .frame $30,0,$26,0\n"
" .prologue 0\n"
" cpys $f31,$f31,$f0\n"
" fbgt $f16,$gt_1\n" /* stop at this instruction. */
" cpysn $f16,$f16,$f0\n"
"$gt_1:\n"
" ret $31,($26),1\n"
" .end gt\n");
asm (" .text\n"
" .ent ge\n"
"ge:\n"
" .frame $30,0,$26,0\n"
" .prologue 0\n"
" cpys $f31,$f31,$f0\n"
" fbge $f16,$ge_1\n" /* stop at this instruction. */
" cpysn $f16,$f16,$f0\n"
"$ge_1:\n"
" ret $31,($26),1\n"
" .end ge\n");
asm (" .text\n"
" .ent lt\n"
"lt:\n"
" .frame $30,0,$26,0\n"
" .prologue 0\n"
" cpys $f31,$f31,$f0\n"
" fblt $f16,$lt_1\n" /* stop at this instruction. */
" cpysn $f16,$f16,$f0\n"
"$lt_1:\n"
" ret $31,($26),1\n"
" .end lt\n");
asm (" .text\n"
" .ent le\n"
"le:\n"
" .frame $30,0,$26,0\n"
" .prologue 0\n"
" cpys $f31,$f31,$f0\n"
" fble $f16,$le_1\n" /* stop at this instruction. */
" cpysn $f16,$f16,$f0\n"
"$le_1:\n"
" ret $31,($26),1\n"
" .end le\n");
asm (" .text\n"
" .ent eq\n"
"eq:\n"
" .frame $30,0,$26,0\n"
" .prologue 0\n"
" cpys $f31,$f31,$f0\n"
" fbeq $f16,$eq_1\n" /* stop at this instruction. */
" cpysn $f16,$f16,$f0\n"
"$eq_1:\n"
" ret $31,($26),1\n"
" .end eq\n");
asm (" .text\n"
" .ent ne\n"
"ne:\n"
" .frame $30,0,$26,0\n"
" .prologue 0\n"
" cpys $f31,$f31,$f0\n"
" fbne $f16,$ne_1\n" /* stop at this instruction. */
" cpysn $f16,$f16,$f0\n"
"$ne_1:\n"
" ret $31,($26),1\n"
" .end ne\n");
[-- Attachment #3: alpha-step.exp --]
[-- Type: text/plain, Size: 3797 bytes --]
# Copyright 2005 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 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.
if $tracelevel {
strace $tracelevel
}
set prms_id 0
set bug_id 0
if ![istarget "alpha*-*-*"] then {
verbose "Skipping alpha step tests."
return
}
set testfile "alpha-step"
set srcfile ${testfile}.c
set binfile ${objdir}/${subdir}/${testfile}
if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {}] != "" } {
unsupported "Testcase compile failed."
return -1
}
gdb_exit
gdb_start
gdb_reinitialize_dir $srcdir/$subdir
gdb_load ${binfile}
proc test_stepi {function } {
# Restart the program from scratch. If GDB got confused during one
# of the previous tests, this makes sure that it doesn't affect
# this series of tests.
if ![runto_main] then {
fail "Can't run to main"
return 0
}
# Insert a breakpoint on the FP branch instruction inside FUNCTION.
# Since the function has been hand written, we know this instruction
# is a "fb$function" located at FUNCTION+4.
gdb_test "break *$function+4" \
"Breakpoint .* at .*" \
"breakpoint on fb$function instruction"
gdb_test "continue" \
"Breakpoint .*, 0x\[0-9a-fA-F\]+ in $function\(\).*" \
"continue to fb$function instruction (first call)"
# Extra check to make sure we stopped on the FP branch instruction.
gdb_test "x /i \$pc" \
"0x\[0-9a-fA-F\]+ <.*>:\[ \t\]+fb$function\[ \t\]+.*" \
"Check breakpoint on fb$function instruction (first call)"
# Step test, followed by the check that we landed on the expected
# instruction (the testcase should be written in such a way that
# the branch is taken on the first call to this function.
gdb_test "stepi" \
"0x\[0-9a-fA-F\]+.*" \
"stepi on fb$function (first call)"
gdb_test "x /i \$pc" \
"0x\[0-9a-fA-F\]+ <.*>:\[ \t\]+ret\[ \t\]+.*" \
"Check stepi over fb$function stopped on ret"
# Continue again. FUNCTION should be called a second time, this time
# with an argument such that the FP branch will not be taken.
gdb_test "continue" \
"Breakpoint .*, 0x\[0-9a-fA-F\]+ in $function\(\).*" \
"continue to fb$function instruction (second call)"
# Extra check to make sure we stopped on the FP branch instruction.
gdb_test "x /i \$pc" \
"0x\[0-9a-fA-F\]+ <.*>:\[ \t\]+fb$function\[ \t\]+.*" \
"Check breakpoint on fb$function instruction (second call)"
# Step test, branch should not be taken.
gdb_test "stepi" \
"0x\[0-9a-fA-F\]+.*" \
"stepi on fb$function (branch not taken)"
# Extra check to verify that we landed on the instruction we expected.
gdb_test "x /i \$pc" \
"0x\[0-9a-fA-F\]+ <.*>:\[ \t\]+fneg\[ \t\]+.*" \
"Check stepi over fb$function stopped on fneg instruction"
}
test_stepi "gt"
test_stepi "ge"
test_stepi "lt"
test_stepi "le"
test_stepi "eq"
test_stepi "ne"
next prev parent reply other threads:[~2005-05-26 2:35 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-05-25 15:09 Richard Henderson
2005-05-26 3:14 ` Joel Brobecker [this message]
2005-05-26 3:27 ` Richard Henderson
2005-05-26 3:41 ` Daniel Jacobowitz
2005-05-26 5:02 ` Joel Brobecker
2005-05-26 6:04 ` Daniel Jacobowitz
2005-05-28 22:10 ` Daniel Jacobowitz
2005-05-29 2:53 ` Joel Brobecker
-- strict thread matches above, loose matches on Subject: below --
2005-05-20 7:02 Joel Brobecker
2005-05-20 11:31 ` Richard Henderson
2005-05-25 5:15 ` Joel Brobecker
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=20050526023539.GL1525@adacore.com \
--to=brobecker@adacore.com \
--cc=gdb-patches@gcc.gnu.org \
--cc=rth@redhat.com \
/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