Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Joel Brobecker <brobecker@adacore.com>
To: Richard Henderson <rth@redhat.com>
Cc: gdb-patches@sources.redhat.com
Subject: Re: [testsuite/alpha] Add test for step over fbne instruction
Date: Wed, 25 May 2005 05:15:00 -0000	[thread overview]
Message-ID: <20050525022050.GE1525@adacore.com> (raw)
In-Reply-To: <20050520061810.GA29720@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 669 bytes --]

> .rdata is an ecoff thing; it won't work for elf systems.  I suggest
> that you pass in this data from main and leave the symbol work to 
> the compiler.  That is,

Thanks for the couple of suggestions. Is the attached better?

> I might also suggest that you test both branch directions.

Good idea, and it apparently revealed another problem with single
stepping. I will investigate that problem as soon as possible.

2005-05-25  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.
Currently shows 2 FAILs, which I will investigate asap.

Thanks,
-- 
Joel

[-- Attachment #2: alpha-step.c --]
[-- Type: text/plain, Size: 2319 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);

  return 0;
}

/* gt(). Obtained from the following code:

        void
        gt (double a)
        {
          if (a > 0)
            a = 0.0;
          else
            a = -a;
        }

  The purpose of the test is to stop just before the check against
  zero, and do a stepi. GDB should step one instruction and stop,
  rather than letting the program continue until the it terminates.  */

asm("        .text\n"
    "        .align 2\n"
    "        .globl gt\n"
    "        .ent gt\n"
    "$gt..ng:\n"
    "gt:\n"
    "        .frame $15,32,$26,0\n"
    "        .mask 0x4008000,-32\n"
    "$SYMBOL_LM1:\n"
    "        lda $30,-32($30)\n"
    "        stq $26,0($30)\n"
    "        stq $15,8($30)\n"
    "        bis $31,$30,$15\n"
    "        .prologue 0\n"
    "        stt $f16,16($15)\n"
    "$SYMBOL_LM2:\n"
    "        ldt $f10,16($15)\n"
    "        cmptlt $f31,$f10,$f10\n"
    "        fbne $f10,$SYMBOL_L3\n"
    "        br $31,$SYMBOL_L2\n"
    "$SYMBOL_L3:\n"
    "$SYMBOL_LM3:\n"
    "        stt $f31,16($15)\n"
    "        br $31,$SYMBOL_L4\n"
    "$SYMBOL_L2:\n"
    "$SYMBOL_LM4:\n"
    "        ldt $f10,16($15)\n"
    "        cpysn $f10,$f10,$f10\n"
    "        stt $f10,16($15)\n"
    "$SYMBOL_L4:\n"
    "$SYMBOL_LM5:\n"
    "        bis $31,$15,$30\n"
    "        ldq $26,0($30)\n"
    "        ldq $15,8($30)\n"
    "        lda $30,32($30)\n"
    "        ret $31,($26),1\n"
    "        .end gt\n");


[-- Attachment #3: alpha-step.exp --]
[-- Type: text/plain, Size: 3366 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}

if ![runto_main] then {
    fail "Can't run to main"
    return 0
}

# Insert a breakpoint on a fbne instruction inside gt that we know is
# at gt+28 (the body of gt has been written in assembly, so can not
# change from compiler to compiler). This instruction will cause a jump
# of a couple of instructions forward. Verify that this jump does not
# confuse GDB when steping...

gdb_test "break *gt+28" \
         "Breakpoint 2 at .*" \
         "breakpoint on fbne instruction"

gdb_test "continue" \
         "Breakpoint 2, 0x\[0-9a-fA-F\]+ in gt().*" \
         "continue to fbne instruction"

# Extra check to make sure we stopped on the fbne instruction.

gdb_test "x /i \$pc" \
         "0x\[0-9a-fA-F\]+ <.*>:\[ \t\]+fbne\[ \t\]+.*" \
         "Check breakpoint on fbne instruction"

# Step test, should not cause the inferior to run until completion.

gdb_test "stepi" \
         "0x\[0-9a-fA-F\]+.*" \
         "stepi on fbne"

# Extra check to verify that we landed on the instruction we expected
# (during the first call of this function, the argument is positive,
# so it should be an stt instruction).

gdb_test "x /i \$pc" \
         "0x\[0-9a-fA-F\]+ <.*>:\[ \t\]+stt\[ \t\]+.*" \
         "Check stopped on stt instruction"

# Continue again. gt() should be called a second time, this time with
# a negative parameter.

gdb_test "continue" \
         "Breakpoint 2, 0x\[0-9a-fA-F\]+ in gt().*" \
         "continue to fbne instruction (branch taken)"

# Extra check to make sure we stopped on the fbne instruction.

gdb_test "x /i \$pc" \
         "0x\[0-9a-fA-F\]+ <.*>:\[ \t\]+fbne\[ \t\]+.*" \
         "Check breakpoint on fbne instruction"

# Step test, should not cause the inferior to run until completion.

gdb_test "stepi" \
         "0x\[0-9a-fA-F\]+.*" \
         "stepi on fbne (branch not taken)"

# Extra check to verify that we landed on the instruction we expected
# (during the second call of this function, the argument is negative,
# so it should be a br instruction).

gdb_test "x /i \$pc" \
         "0x\[0-9a-fA-F\]+ <.*>:\[ \t\]+br\[ \t\]+.*" \
         "Check stopped on br instruction"


  reply	other threads:[~2005-05-25  2:21 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-05-20  7:02 Joel Brobecker
2005-05-20 11:31 ` Richard Henderson
2005-05-25  5:15   ` Joel Brobecker [this message]
2005-05-25 15:09 Richard Henderson
2005-05-26  3:14 ` Joel Brobecker
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

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=20050525022050.GE1525@adacore.com \
    --to=brobecker@adacore.com \
    --cc=gdb-patches@sources.redhat.com \
    --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