From: "Raoul Gough" <RaoulGough@yahoo.co.uk>
To: "Elena Zannoni" <ezannoni@redhat.com>
Cc: <gdb-patches@sources.redhat.com>
Subject: Re: RFC: coffread.c relocation fixes
Date: Wed, 28 May 2003 15:04:00 -0000 [thread overview]
Message-ID: <002301c3252a$9690ec00$0200a8c0@albert> (raw)
In-Reply-To: <16078.19295.731449.579816@localhost.redhat.com>
[-- Attachment #1: Type: text/plain, Size: 1578 bytes --]
"Elena Zannoni" <ezannoni@redhat.com> wrote in message
news:16078.19295.731449.579816@localhost.redhat.com...
> Raoul Gough writes:
[snip]
> > I've added the script as a note to PR 1132 in GNATS (see
> > http://sources.redhat.com/ml/gdb-prs/2003-q2/msg00129.html).
> >
>
> I don't have much comments on the test (except that the convention
is
> to use '-' instead of '_' in file names, and that you need to update
> the gdb/config/djgpp/fnchange.lst file). I wonder if instead of
'while 1'
> you could use exp_continue. MichaelC?
Oh, DOS 8.3 filenames - I see... probably best to rename the files
everywhere - otherwise the shared-reloc.exp script would have to be
aware of the renaming issue on DJGPP (since it invokes the compiler on
the long-named .c files). I've attached the new versions to this post.
BTW, I think Andrew Cagney is sorting out CVS access for me, so I may
be able to add the test to CVS myself, if that's appropriate. I wonder
if this test works on DJGPP?
Regarding the "while 1" issue - my TCL/expect skills are so minimal, I
couldn't think of a better way of doing the loop - an improvement
would be good, as long as it doesn't involve me learning any more TCL
:-)
> > I've attached an updated diff to this posting with a fix for the
typo
> > that Elena identified. Any advice on getting these patches
checked in
> > and the PR closed? I can't do it myself, since I don't have any
write
> > access.
> >
>
> I can commit it for you.
Thanks for this - the latest CVS version now passes the shared object
relocation tests.
Regards,
Raoul Gough.
[-- Attachment #2: shreloc.c --]
[-- Type: text/plain, Size: 344 bytes --]
#if defined(_WIN32) || defined(__CYGWIN__)
# define ATTRIBUTES __attribute((__dllimport__))
#else
# define ATTRIBUTES
#endif
extern ATTRIBUTES void fn_1 (int);
extern ATTRIBUTES void fn_2 (int);
extern ATTRIBUTES int extern_var_1;
extern ATTRIBUTES int extern_var_2;
int main ()
{
fn_1 (extern_var_1);
fn_2 (extern_var_2);
return 0;
}
[-- Attachment #3: shreloc2.c --]
[-- Type: text/plain, Size: 223 bytes --]
#if defined(_WIN32) || defined(__CYGWIN__)
# define ATTRIBUTES __attribute((__dllexport__))
#else
# define ATTRIBUTES
#endif
static int static_var_2;
ATTRIBUTES void fn_2 (int unused) { }
ATTRIBUTES int extern_var_2 = 0;
[-- Attachment #4: shreloc1.c --]
[-- Type: text/plain, Size: 223 bytes --]
#if defined(_WIN32) || defined(__CYGWIN__)
# define ATTRIBUTES __attribute((__dllexport__))
#else
# define ATTRIBUTES
#endif
static int static_var_1;
ATTRIBUTES void fn_1 (int unused) { }
ATTRIBUTES int extern_var_1 = 0;
[-- Attachment #5: shreloc.exp --]
[-- Type: application/octet-stream, Size: 7161 bytes --]
# Copyright (C) 2003 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.
#
# Please email any bugs, comments, and/or additions to this file to:
# bug-gdb@prep.ai.mit.edu
# Tests for shared object file relocation. If two shared objects have
# the same load address (actually, overlapping load spaces), one of
# them gets relocated at load-time. Check that gdb gets the right
# values for the debugging and minimal symbols.
if $tracelevel then {
strace $tracelevel
}
#
# This file uses shreloc.c, shreloc1.c and shreloc2.c
#
set prms_id 0
set bug_id 0
set workdir ${objdir}/${subdir}
foreach module [list "shreloc" "shreloc1" "shreloc2"] {
if {[gdb_compile "${srcdir}/${subdir}/${module}.c" "${workdir}/${module}.o" object {debug}] != ""} {
gdb_suppress_entire_file "${module}.c compile failed, so all tests in this file will automatically fail."
return -1
}
}
set additional_flags "additional_flags=-shared"
if {([istarget "*pc-cygwin"] || [istarget "*pc-mingw32"]) } {
set additional_flags "${additional_flags} -Wl,--image-base,0x04000000"
}
foreach module [list "shreloc1" "shreloc2"] {
if {[gdb_compile "${workdir}/${module}.o" "${workdir}/${module}.dll" executable [list debug $additional_flags]] != ""} {
gdb_suppress_entire_file "${module}.dll link failed, so all tests in this file will automatically fail."
return -1
}
}
if {[gdb_compile [list "${workdir}/shreloc.o" "${workdir}/shreloc1.dll" "${workdir}/shreloc2.dll"] "${workdir}/shreloc" executable debug] != ""} {
gdb_suppress_entire_file "shreloc link failed, so all tests in this file will automatically fail."
return -1
}
gdb_exit
gdb_start
gdb_reinitialize_dir $srcdir/$subdir
gdb_load ${workdir}/shreloc
# Load up the shared objects
if ![runto_main] then {
fail "Can't run to main"
return 0
}
proc get_var_address { var } {
global gdb_prompt hex
send_gdb "print &${var}\n"
# Match output like:
# $1 = (int *) 0x0
# $5 = (int (*)()) 0
# $6 = (int (*)()) 0x24 <function_bar>
gdb_expect {
-re "\\\$\[0-9\]+ = \\(.*\\) (0|$hex)( <${var}>)?\[\r\n\]+${gdb_prompt} $"
{
pass "get address of ${var}"
if { $expect_out(1,string) == "0" } {
return "0x0"
} else {
return $expect_out(1,string)
}
}
-re "${gdb_prompt} $"
{ fail "get address of ${var} (unknown output)" }
timeout
{ fail "get address of ${var} (timeout)" }
}
return ""
}
#
# Check debugging symbol relocations
#
# Check extern function for relocation
set fn_1_addr [get_var_address fn_1]
set fn_2_addr [get_var_address fn_2]
if { "${fn_1_addr}" == "${fn_2_addr}" } {
fail "relocated extern functions have different addresses"
} else {
pass "relocated extern functions have different addresses"
}
# Check extern var for relocation
set extern_var_1_addr [get_var_address extern_var_1]
set extern_var_2_addr [get_var_address extern_var_2]
if { "${extern_var_1_addr}" == "${extern_var_2_addr}" } {
fail "relocated extern variables have different addresses"
} else {
pass "relocated extern variables have different addresses"
}
# Check static var for relocation
set static_var_1_addr [get_var_address static_var_1]
set static_var_2_addr [get_var_address static_var_2]
if { "${static_var_1_addr}" == "${static_var_2_addr}" } {
fail "relocated static variables have different addresses"
} else {
pass "relocated static variables have different addresses"
}
#
# Check minimal symbol relocations
#
proc send_gdb_discard { command } {
# Send a command to gdb and discard output up to the next prompt
global gdb_prompt
send_gdb "${command}\n"
# Discard output
gdb_expect {
-re ".*\[\r\n]+${gdb_prompt} $" {
return 1
}
timeout {
fail "{$command} (timeout)"
return 0
}
}
}
proc get_msym_addrs { var msymfile } {
# Extract the list of values for symbols matching var in the
# minimal symbol output file
global gdb_prompt hex
set result ""
send_gdb "shell grep -E \" ${var}(\[ \t\]+.*)?\$\" ${msymfile}\n"
while 1 {
gdb_expect {
-re "\[\[\]\[ 0-9\]+\] . (${hex}) ${var}(\[ \t\]+\[^\r\n\]*)?\[\r\n\]+" {
set result [concat $result $expect_out(1,string)]
}
-re "$gdb_prompt $" {
pass "get_msym_addrs ${var} (${result})"
return "${result}"
}
-re "\[^\r\n\]*\[\r\n\]+" {
# Skip
}
timeout {
fail "get_msym_addrs ${var} (timeout)"
return -1
}
}
}
}
proc check_same {var msymfile} {
# Check that the minimal symbol values matching var are the same
set len [llength [lsort -unique [get_msym_addrs "${var}" "${msymfile}"]]]
if { $len == 1 } {
return 1
} else {
return 0
}
}
proc check_different {var msymfile} {
# Check that the minimal symbol values matching var are different
set addr_list [lsort [get_msym_addrs "${var}" "${msymfile}"]]
set prev ""
if { [llength ${addr_list}] < 2 } {
return 0
}
foreach addr ${addr_list} {
if { ${prev} == ${addr} } {
return 0
}
set prev ${addr}
}
return 1
}
set msymfile "${workdir}/shreloc.txt"
if [send_gdb_discard "maint print msymbols ${msymfile}"] {
if {[check_different "static_var_\[12\]" "${msymfile}"]} {
pass "(msymbol) relocated static vars have different addresses"
} else {
fail "(msymbol) relocated static vars have different addresses"
}
if {[check_different "extern_var_\[12\]" "${msymfile}"]} {
pass "(msymbol) relocated extern vars have different addresses"
} else {
fail "(msymbol) relocated extern vars have different addresses"
}
if {[check_different "fn_\[12\]" "${msymfile}"]} {
pass "(msymbol) relocated functions have different addresses"
} else {
fail "(msymbol) relocated functions have different addresses"
}
}
if {([istarget "*pc-cygwin"] || [istarget "*pc-mingw32"]) } {
#
# We know the names of some absolute symbols included in the
# portable-executable (DLL) format. Check that they didn't get
# relocated.
#
# A better approach would be include absolute symbols via the assembler.
#
if {[check_same "_minor_os_version__" "${msymfile}"]} {
pass "Absolute symbols not relocated"
} else {
fail "Absolute symbols not relocated"
}
}
next prev parent reply other threads:[~2003-05-28 15:04 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2003-02-26 18:40 Raoul Gough
2003-04-03 21:48 ` Elena Zannoni
2003-04-07 11:54 ` Raoul Gough
2003-05-12 16:38 ` Raoul Gough
2003-05-23 16:19 ` Elena Zannoni
2003-05-28 15:04 ` Raoul Gough [this message]
2003-05-28 15:42 ` Eli Zaretskii
2003-05-23 18:42 Michael Elizabeth Chastain
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='002301c3252a$9690ec00$0200a8c0@albert' \
--to=raoulgough@yahoo.co.uk \
--cc=ezannoni@redhat.com \
--cc=gdb-patches@sources.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