This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
[RFC-v2] First scripts for ARI support inisde gdb tree
- From: "Pierre Muller" <pierre dot muller at ics-cnrs dot unistra dot fr>
- To: "'Joel Brobecker'" <brobecker at adacore dot com>
- Cc: <gdb-patches at sourceware dot org>
- Date: Fri, 18 May 2012 10:24:06 +0200
- Subject: [RFC-v2] First scripts for ARI support inisde gdb tree
- References: <003c01cd33b2$b6225130$2266f390$@muller@ics-cnrs.unistra.fr> <20120517131616.GD10253@adacore.com>
> -----Message d'origine-----
> De?: gdb-patches-owner@sourceware.org [mailto:gdb-patches-
> owner@sourceware.org] De la part de Joel Brobecker
> Envoyé?: jeudi 17 mai 2012 15:16
> À?: Pierre Muller
> Cc?: gdb-patches@sourceware.org
> Objet?: Re: [RFC] First scripts for ARI support inisde gdb tree
>
> > I do have several potential issues:
> > - permissions: Is it possible to set unix-like permissions using cvs?
> > I now the svn has such options, but I don't know for cvs...
>
> I think that CVS does record the unix permissions, or enough of it
> for our purposes, but I'm not entirely sure anymore. Make sure
> that the scripts do have the execute permission before you commit
> anything and I think we'll be good to go.
OK, let's hope for the best here...
To be aon the safe side, I will commit from a linux account.
> > ./ari/create-web-ari-in-src.sh
> > is the ultimate simple script:
> > It uses ../.. as a main source directory
> > (and thus .. as gdb source directory).
> > it create a tmp directory called /tmp/testari
> > and writes results into ~/htdocs/www/local/ari directory.
> > I have no idea if those defaults are OK for most systems...
>
> I'd like that script to work more like configure: Use the path to
> itself to determine the src dir, and create the HTML pages in
> the current directory. That way, it's easy to use out-of-tree,
> and does not force the user to use any specific directory name.
That was indeed the purpose of this script:
have a parameterless script to call that
would give update-web-ari{.sh} script default parameters.
I added some comments inside the script.
> Here is how to do it in a mostly-portable way:
>
> root=`dirname $0`
>
> # If "root" is a relative path, then convert it to absolute.
> if [ "`echo ${root} | cut -b1`" != '/' ]; then
> root="`pwd`/${root}"
> fi
>
> (works everywhere except perhaps MinGW).
I used this, thanks.
> > +# GDB script to list of problems using awk.
> > +#
> > +# Copyright (C) 2002-2005,2007,2009-2012 Free Software Foundation, Inc.
>
> You can change that to 2002-2012. Same for all the other files.
Done,
> > +# Permenant checks take the form
>
> Permanent
Corrected, as well as the trailing spaces.
> > +# Multi-line string
> > +string_p &&
>
> Trailing space...
>
> > + # Closing brace found?
>
> Ditto.
>
> > + # Closing brace found?
>
> Ditto.
>
> > +# Only function implemenation should be on first column
Also corrected.
> > +# A find that prunes files that GDB users shouldn't be interested in.
> > +# Use sort to order files alphabetically.
> > +
> > +find "$@" \
> > + -name testsuite -prune -o \
> > + -name gdbserver -prune -o \
> > + -name gnulib -prune -o \
> > + -name osf-share -prune -o \
> > + -name '*-stub.c' -prune -o \
> > + -name '*-exp.c' -prune -o \
> > + -name ada-lex.c -prune -o \
> > + -name cp-name-parser.c -prune -o \
> > + -type f -name '*.[lyhc]' -print | sort
>
> I am thinking that this script can probably be inlined in the top-level
> script (the one you called create-web-ari-in-src.sh).
It's called from within update-wevb-ari.sh,
I would like to leave that script
as close as possible to the one in ss directory
on sourceware, so that we can just call that script if it exists in the cvs
checkout made to create the daily snapshot.
> > +PATH=/bin:/usr/bin:/usr/local/bin:$HOME/bin
> > +export PATH
>
> This is really sourceware-specific, let's remove it from here.
Removed.
> > +# Really mindless usage
> > +if test $# -ne 4
> > +then
> > + echo "Usage: $0 <snapshot/sourcedir> <tmpdir> <destdir> <project>"
> 1>&2
> > + exit 1
> > +fi
> > +snapshot=$1 ; shift
> > +tmpdir=$1 ; shift
> > +wwwdir=$1 ; shift
> > +project=$1 ; shift
>
> Eventually, I'd like to simply the user interface:
> - Remove the <tmpdir> parameter and replace it by $TMP or /tmp,
> or whatever;
> - Remove the <destdir> parameter and replace it by the current
> working directory
> - Remove the <project> parameter, since the project is always gdb.
This is all done in the create-web-ari-in-src.sh script.
I just was wondering about the removal of the destination directory:
I would ideally want to be able to run
the script either from source directory
or from build directory.
In case it is run from source dir,
I would like that it doesn't generate anything in the source dir itself
to avoid polluting the cvs checkout. Anyhow, it should
be possible to run it even without write permission to the source directory.
> Also, I don't think we'll need 3 scripts in total to do the job.
> I think we can merge them all into 1, and hopefully that'll make
> things a little simpler. But maybe not.
>
> What I propose is that we avoid making the entry barrier too high,
> and refrain from making dramatic design changes for now. Once we
> have something working, we can work on those design ideas to
> simplify things as you see fit. As long as the interface to
> the entry-point script does not change, anything underneath can
> change at will...
So I will not try to put everything together now.
> > + # Check if ${diff} is not empty
> > + if [ -s ${diff} ]; then
> > + # Send an email to muller@sourceware.org
> > + mutt -s "New ARI warning ${email_suffix}" \
> > + ${warning_email} < ${wwwdir}/mail-message
> > + else
> > + if [ -s ${wwwdir}/${mail-message} ]; then
> > + # Send an email to muller@sourceware.org
> > + mutt -s "ARI warning list change ${email_suffix}" \
> > + muller@sourceware.org < ${wwwdir}/mail-message
> > + fi
> > + fi
>
> This part, on the other hand, needs to be excised from this script.
> Otherwise, users wanting to compare the ARI before and after a change
> might accidently send email to the list...
I don't want to remove it completely,
as I would like that the nightly snapshot script still generates
those messages when run.
> We'll probably want to have the bit of code that compares the ARI
> results also in the GDB repository, so perhaps a separate script?
> In the meantime, it's find to not include this part.
I added a send_email variable that defaults to false,
and should be put to true before calling the update-web-ari script
in the nightly snapshot,
does this seem reasonable?
Pierre Muller
as ARI maintainer
2012-05-17 Pierre Muller <muller@ics.u-strasbg.fr>
* ari/create-web-ari-in-src.sh: New file.
* ari/gdb_ari.sh: New file.
* ari/gdb_find.sh: New file.
* ari/update-web-ari.sh: New file.
? log
? patch
? update-web-cvs-ari
Index: create-web-ari-in-src.sh
===================================================================
RCS file: create-web-ari-in-src.sh
diff -N create-web-ari-in-src.sh
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ create-web-ari-in-src.sh 18 May 2012 07:15:58 -0000
@@ -0,0 +1,61 @@
+#! /bin/sh
+
+# GDB script to create web ARI page directly from within gdb/ari directory.
+#
+# Copyright (C) 2012 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 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/>.
+
+set -x
+
+# Determine directory of current script.
+scriptpath=`dirname $0`
+# If "scriptpath" is a relative path, then convert it to absolute.
+if [ "`echo ${scriptpath} | cut -b1`" != '/' ] ; then
+ scriptpath="`pwd`/${scriptpath}"
+fi
+
+# update-web-ari.sh script wants four parameters
+# 1: directory of checkout src or gdb-RELEASE for release sources.
+# 2: a temp directory.
+# 3: a directory for generated web page.
+# 4: The name of the current package, must be gdb here.
+# Here we provide default values for these 4 parameters
+
+# srcdir parameter
+if [ -z "${srcdir}" ] ; then
+ srcdir=${scriptpath}/../..
+fi
+
+# Determine location of a temporary directory to be used by
+# update-web-ari.sh script.
+if [ -z "${tempdir}" ] ; then
+ if [ ! -z "$TMP" ] ; then
+ tempdir=$TMP/create-ari
+ elif [ ! -z "$TEMP" ] ; then
+ tempdir=$TEMP/create-ari
+ else
+ tempdir=/tmp/create-ari
+ fi
+fi
+
+# Default location of generate index.hmtl web page.
+if [ -z "${webdir}" ] ; then
+ webdir=~/htdocs/www/local/ari
+fi
+
+# Launch update-web-ari.sh in same directory as current script.
+${scriptpath}/update-web-ari.sh ${srcdir} ${tempdir} ${webdir} gdb
Index: gdb_ari.sh
===================================================================
RCS file: gdb_ari.sh
diff -N gdb_ari.sh
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ gdb_ari.sh 18 May 2012 07:15:58 -0000
@@ -0,0 +1,1347 @@
+#!/bin/sh
+
+# GDB script to list of problems using awk.
+#
+# Copyright (C) 2002-2012 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 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/>.
+
+# Make certain that the script is not running in an internationalized
+# environment.
+
+LANG=c ; export LANG
+LC_ALL=c ; export LC_ALL
+
+# Permanent checks take the form:
+
+# Do not use XXXX, ISO C 90 implies YYYY
+# Do not use XXXX, instead use YYYY''.
+
+# and should never be removed.
+
+# Temporary checks take the form:
+
+# Replace XXXX with YYYY
+
+# and once they reach zero, can be eliminated.
+
+# FIXME: It should be able to override this on the command line
+error="regression"
+warning="regression"
+ari="regression eol code comment deprecated legacy obsolete gettext"
+all="regression eol code comment deprecated legacy obsolete gettext
deprecate internal gdbarch macro"
+print_doc=0
+print_idx=0
+
+usage ()
+{
+ cat <<EOF 1>&2
+Error: $1
+
+Usage:
+ $0 --print-doc --print-idx -Wall -Werror -W<category> <file> ...
+Options:
+ --print-doc Print a list of all potential problems, then exit.
+ --print-idx Include the problems IDX (index or key) in every message.
+ --src=file Write source lines to file.
+ -Werror Treat all problems as errors.
+ -Wall Report all problems.
+ -Wari Report problems that should be fixed in new code.
+ -W<category> Report problems in the specifed category. Vaid categories
+ are: ${all}
+EOF
+ exit 1
+}
+
+
+# Parse the various options
+Woptions=
+srclines=""
+while test $# -gt 0
+do
+ case "$1" in
+ -Wall ) Woptions="${all}" ;;
+ -Wari ) Woptions="${ari}" ;;
+ -Werror ) Werror=1 ;;
+ -W* ) Woptions="${Woptions} `echo x$1 | sed -e 's/x-W//'`" ;;
+ --print-doc ) print_doc=1 ;;
+ --print-idx ) print_idx=1 ;;
+ --src=* ) srclines="`echo $1 | sed -e 's/--src=/srclines=\"/'`\"" ;;
+ -- ) shift ; break ;;
+ - ) break ;;
+ -* ) usage "$1: unknown option" ;;
+ * ) break ;;
+ esac
+ shift
+done
+if test -n "$Woptions" ; then
+ warning="$Woptions"
+ error=
+fi
+
+
+# -Werror implies treating all warnings as errors.
+if test -n "${Werror}" ; then
+ error="${error} ${warning}"
+fi
+
+
+# Validate all errors and warnings.
+for w in ${warning} ${error}
+do
+ case " ${all} " in
+ *" ${w} "* ) ;;
+ * ) usage "Unknown option -W${w}" ;;
+ esac
+done
+
+
+# make certain that there is at least one file.
+if test $# -eq 0 -a ${print_doc} = 0
+then
+ usage "Missing file."
+fi
+
+
+# Convert the errors/warnings into corresponding array entries.
+for a in ${all}
+do
+ aris="${aris} ari_${a} = \"${a}\";"
+done
+for w in ${warning}
+do
+ warnings="${warnings} warning[ari_${w}] = 1;"
+done
+for e in ${error}
+do
+ errors="${errors} error[ari_${e}] = 1;"
+done
+
+awk -- '
+BEGIN {
+ # NOTE, for a per-file begin use "FNR == 1".
+ '"${aris}"'
+ '"${errors}"'
+ '"${warnings}"'
+ '"${srclines}"'
+ print_doc = '$print_doc'
+ print_idx = '$print_idx'
+ PWD = "'`pwd`'"
+}
+
+# Print the error message for BUG. Append SUPLEMENT if non-empty.
+function print_bug(file,line,prefix,category,bug,doc,supplement,
suffix,idx) {
+ if (print_idx) {
+ idx = bug ": "
+ } else {
+ idx = ""
+ }
+ if (supplement) {
+ suffix = " (" supplement ")"
+ } else {
+ suffix = ""
+ }
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ print file ":" line ": " prefix category ": " idx doc suffix
+ if (srclines != "") {
+ print file ":" line ":" $0 >> srclines
+ }
+}
+
+function fix(bug,file,count) {
+ skip[bug, file] = count
+ skipped[bug, file] = 0
+}
+
+function fail(bug,supplement) {
+ if (doc[bug] == "") {
+ print_bug("", 0, "internal: ", "internal", "internal", "Missing doc
for bug " bug)
+ exit
+ }
+ if (category[bug] == "") {
+ print_bug("", 0, "internal: ", "internal", "internal", "Missing
category for bug " bug)
+ exit
+ }
+
+ if (ARI_OK == bug) {
+ return
+ }
+ # Trim the filename down to just DIRECTORY/FILE so that it can be
+ # robustly used by the FIX code.
+
+ if (FILENAME ~ /^\//) {
+ canonicalname = FILENAME
+ } else {
+ canonicalname = PWD "/" FILENAME
+ }
+ shortname = gensub (/^.*\/([^\\]*\/[^\\]*)$/, "\\1", 1, canonicalname)
+
+ skipped[bug, shortname]++
+ if (skip[bug, shortname] >= skipped[bug, shortname]) {
+ # print FILENAME, FNR, skip[bug, FILENAME], skipped[bug, FILENAME],
bug
+ # Do nothing
+ } else if (error[category[bug]]) {
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ print_bug(FILENAME, FNR, "", category[bug], bug, doc[bug],
supplement)
+ } else if (warning[category[bug]]) {
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ print_bug(FILENAME, FNR, "warning: ", category[bug], bug, doc[bug],
supplement)
+ }
+}
+
+FNR == 1 {
+ seen[FILENAME] = 1
+ if (match(FILENAME, "\\.[ly]$")) {
+ # FILENAME is a lex or yacc source
+ is_yacc_or_lex = 1
+ }
+ else {
+ is_yacc_or_lex = 0
+ }
+}
+END {
+ if (print_idx) {
+ idx = bug ": "
+ } else {
+ idx = ""
+ }
+ # Did we do only a partial skip?
+ for (bug_n_file in skip) {
+ split (bug_n_file, a, SUBSEP)
+ bug = a[1]
+ file = a[2]
+ if (seen[file] && (skipped[bug_n_file] < skip[bug_n_file])) {
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ b = file " missing " bug
+ print_bug(file, 0, "", "internal", file " missing " bug,
"Expecting " skip[bug_n_file] " occurances of bug " bug " in file " file ",
only found " skipped[bug_n_file])
+ }
+ }
+}
+
+
+# Skip OBSOLETE lines
+/(^|[^_[:alnum:]])OBSOLETE([^_[:alnum:]]|$)/ { next; }
+
+# Skip ARI lines
+
+BEGIN {
+ ARI_OK = ""
+}
+
+/\/\* ARI:[[:space:]]*(.*)[[:space:]]*\*\// {
+ ARI_OK = gensub(/^.*\/\*
ARI:[[:space:]]*(.*[^[:space:]])[[:space:]]*\*\/.*$/, "\\1", 1, $0)
+ # print "ARI line found \"" $0 "\""
+ # print "ARI_OK \"" ARI_OK "\""
+}
+! /\/\* ARI:[[:space:]]*(.*)[[:space:]]*\*\// {
+ ARI_OK = ""
+}
+
+
+# Things in comments
+
+BEGIN { doc["GNU/Linux"] = "\
+Do not use `Linux'\'', instead use `Linux kernel'\'' or `GNU/Linux
system'\'';\
+ comments should clearly differentiate between the two (this test assumes
that\
+ word `Linux'\'' appears on the same line as the word `GNU'\'' or
`kernel'\''\
+ or a kernel version"
+ category["GNU/Linux"] = ari_comment
+}
+/(^|[^_[:alnum:]])Linux([^_[:alnum:]]|$)/ \
+&& !/(^|[^_[:alnum:]])Linux\[sic\]([^_[:alnum:]]|$)/ \
+&& !/(^|[^_[:alnum:]])GNU\/Linux([^_[:alnum:]]|$)/ \
+&& !/(^|[^_[:alnum:]])Linux kernel([^_[:alnum:]]|$)/ \
+&& !/(^|[^_[:alnum:]])Linux [:digit:]\.[:digit:]+)/ {
+ fail("GNU/Linux")
+}
+
+BEGIN { doc["ARGSUSED"] = "\
+Do not use ARGSUSED, unnecessary"
+ category["ARGSUSED"] = ari_regression
+}
+/(^|[^_[:alnum:]])ARGSUSED([^_[:alnum:]]|$)/ {
+ fail("ARGSUSED")
+}
+
+
+# SNIP - Strip out comments - SNIP
+
+FNR == 1 {
+ comment_p = 0
+}
+comment_p && /\*\// { gsub (/^([^\*]|\*+[^\/\*])*\*+\//, " "); comment_p =
0; }
+comment_p { next; }
+!comment_p { gsub (/\/\*([^\*]|\*+[^\/\*])*\*+\//, " "); }
+!comment_p && /(^|[^"])\/\*/ { gsub (/\/\*.*$/, " "); comment_p = 1; }
+
+
+BEGIN { doc["_ markup"] = "\
+All messages should be marked up with _."
+ category["_ markup"] = ari_gettext
+}
+/^[^"]*[[:space:]](warning|error|error_no_arg|query|perror_with_name)[[:spa
ce:]]*\([^_\(a-z]/ {
+ if (! /\("%s"/) {
+ fail("_ markup")
+ }
+}
+
+BEGIN { doc["trailing new line"] = "\
+A message should not have a trailing new line"
+ category["trailing new line"] = ari_gettext
+}
+/(^|[^_[:alnum:]])(warning|error)[[:space:]]*\(_\(".*\\n"\)[\),]/ {
+ fail("trailing new line")
+}
+
+# Include files for which GDB has a custom version.
+
+BEGIN { doc["assert.h"] = "\
+Do not include assert.h, instead include \"gdb_assert.h\"";
+ category["assert.h"] = ari_regression
+ fix("assert.h", "gdb/gdb_assert.h", 0) # it does not use it
+}
+/^#[[:space:]]*include[[:space:]]+.assert\.h./ {
+ fail("assert.h")
+}
+
+BEGIN { doc["dirent.h"] = "\
+Do not include dirent.h, instead include gdb_dirent.h"
+ category["dirent.h"] = ari_regression
+}
+/^#[[:space:]]*include[[:space:]]*.dirent\.h./ {
+ fail("dirent.h")
+}
+
+BEGIN { doc["regex.h"] = "\
+Do not include regex.h, instead include gdb_regex.h"
+ category["regex.h"] = ari_regression
+ fix("regex.h", "gdb/gdb_regex.h", 1)
+}
+/^#[[:space:]]*include[[:space:]]*.regex\.h./ {
+ fail("regex.h")
+}
+
+BEGIN { doc["xregex.h"] = "\
+Do not include xregex.h, instead include gdb_regex.h"
+ category["xregex.h"] = ari_regression
+ fix("xregex.h", "gdb/gdb_regex.h", 1)
+}
+/^#[[:space:]]*include[[:space:]]*.xregex\.h./ {
+ fail("xregex.h")
+}
+
+BEGIN { doc["gnu-regex.h"] = "\
+Do not include gnu-regex.h, instead include gdb_regex.h"
+ category["gnu-regex.h"] = ari_regression
+}
+/^#[[:space:]]*include[[:space:]]*.gnu-regex\.h./ {
+ fail("gnu regex.h")
+}
+
+BEGIN { doc["stat.h"] = "\
+Do not include stat.h or sys/stat.h, instead include gdb_stat.h"
+ category["stat.h"] = ari_regression
+ fix("stat.h", "gdb/gdb_stat.h", 1)
+}
+/^#[[:space:]]*include[[:space:]]*.stat\.h./ \
+|| /^#[[:space:]]*include[[:space:]]*.sys\/stat\.h./ {
+ fail("stat.h")
+}
+
+BEGIN { doc["wait.h"] = "\
+Do not include wait.h or sys/wait.h, instead include gdb_wait.h"
+ fix("wait.h", "gdb/gdb_wait.h", 2);
+ category["wait.h"] = ari_regression
+}
+/^#[[:space:]]*include[[:space:]]*.wait\.h./ \
+|| /^#[[:space:]]*include[[:space:]]*.sys\/wait\.h./ {
+ fail("wait.h")
+}
+
+BEGIN { doc["vfork.h"] = "\
+Do not include vfork.h, instead include gdb_vfork.h"
+ fix("vfork.h", "gdb/gdb_vfork.h", 1);
+ category["vfork.h"] = ari_regression
+}
+/^#[[:space:]]*include[[:space:]]*.vfork\.h./ {
+ fail("vfork.h")
+}
+
+BEGIN { doc["error not internal-warning"] = "\
+Do not use error(\"internal-warning\"), instead use internal_warning"
+ category["error not internal-warning"] = ari_regression
+}
+/error.*\"[Ii]nternal.warning/ {
+ fail("error not internal-warning")
+}
+
+BEGIN { doc["%p"] = "\
+Do not use printf(\"%p\"), instead use printf(\"%s\",paddr()) to dump a \
+target address, or host_address_to_string() for a host address"
+ category["%p"] = ari_code
+}
+/%p/ && !/%prec/ {
+ fail("%p")
+}
+
+BEGIN { doc["%ll"] = "\
+Do not use printf(\"%ll\"), instead use printf(\"%s\",phex()) to dump a \
+`long long'\'' value"
+ category["%ll"] = ari_code
+}
+# Allow %ll in scanf
+/%[0-9]*ll/ && !/scanf \(.*%[0-9]*ll/ {
+ fail("%ll")
+}
+
+
+# SNIP - Strip out strings - SNIP
+
+# Test on top.c, scm-valprint.c, remote-rdi.c, ada-lang.c
+FNR == 1 {
+ string_p = 0
+ trace_string = 0
+}
+# Strip escaped characters.
+{ gsub(/\\./, "."); }
+# Strip quoted quotes.
+{ gsub(/'\''.'\''/, "'\''.'\''"); }
+# End of multi-line string
+string_p && /\"/ {
+ if (trace_string) print "EOS:" FNR, $0;
+ gsub (/^[^\"]*\"/, "'\''");
+ string_p = 0;
+}
+# Middle of multi-line string, discard line.
+string_p {
+ if (trace_string) print "MOS:" FNR, $0;
+ $0 = ""
+}
+# Strip complete strings from the middle of the line
+!string_p && /\"[^\"]*\"/ {
+ if (trace_string) print "COS:" FNR, $0;
+ gsub (/\"[^\"]*\"/, "'\''");
+}
+# Start of multi-line string
+BEGIN { doc["multi-line string"] = "\
+Multi-line string must have the newline escaped"
+ category["multi-line string"] = ari_regression
+}
+!string_p && /\"/ {
+ if (trace_string) print "SOS:" FNR, $0;
+ if (/[^\\]$/) {
+ fail("multi-line string")
+ }
+ gsub (/\"[^\"]*$/, "'\''");
+ string_p = 1;
+}
+# { print }
+
+# Multi-line string
+string_p &&
+
+# Accumulate continuation lines
+FNR == 1 {
+ cont_p = 0
+}
+!cont_p { full_line = ""; }
+/[^\\]\\$/ { gsub (/\\$/, ""); full_line = full_line $0; cont_p = 1; next;
}
+cont_p { $0 = full_line $0; cont_p = 0; full_line = ""; }
+
+
+# GDB uses ISO C 90. Check for any non pure ISO C 90 code
+
+BEGIN { doc["PARAMS"] = "\
+Do not use PARAMS(), ISO C 90 implies prototypes"
+ category["PARAMS"] = ari_regression
+}
+/(^|[^_[:alnum:]])PARAMS([^_[:alnum:]]|$)/ {
+ fail("PARAMS")
+}
+
+BEGIN { doc["__func__"] = "\
+Do not use __func__, ISO C 90 does not support this macro"
+ category["__func__"] = ari_regression
+ fix("__func__", "gdb/gdb_assert.h", 1)
+}
+/(^|[^_[:alnum:]])__func__([^_[:alnum:]]|$)/ {
+ fail("__func__")
+}
+
+BEGIN { doc["__FUNCTION__"] = "\
+Do not use __FUNCTION__, ISO C 90 does not support this macro"
+ category["__FUNCTION__"] = ari_regression
+}
+/(^|[^_[:alnum:]])__FUNCTION__([^_[:alnum:]]|$)/ {
+ fail("__FUNCTION__")
+}
+
+BEGIN { doc["__CYGWIN32__"] = "\
+Do not use __CYGWIN32__, instead use __CYGWIN__ or, better, an explicit \
+autoconf tests"
+ category["__CYGWIN32__"] = ari_regression
+}
+/(^|[^_[:alnum:]])__CYGWIN32__([^_[:alnum:]]|$)/ {
+ fail("__CYGWIN32__")
+}
+
+BEGIN { doc["PTR"] = "\
+Do not use PTR, ISO C 90 implies `void *'\''"
+ category["PTR"] = ari_regression
+ #fix("PTR", "gdb/utils.c", 6)
+}
+/(^|[^_[:alnum:]])PTR([^_[:alnum:]]|$)/ {
+ fail("PTR")
+}
+
+BEGIN { doc["UCASE function"] = "\
+Function name is uppercase."
+ category["UCASE function"] = ari_code
+ possible_UCASE = 0
+ UCASE_full_line = ""
+}
+(possible_UCASE) {
+ if (ARI_OK == "UCASE function") {
+ possible_UCASE = 0
+ }
+ # Closing brace found?
+ else if (UCASE_full_line ~ \
+ /^[A-Z][[:alnum:]_]*[[:space:]]*\([^()]*\).*$/) {
+ if ((UCASE_full_line ~ \
+ /^[A-Z][[:alnum:]_]*[[:space:]]*\([^()]*\)[[:space:]]*$/) \
+ && ($0 ~ /^\{/) && (is_yacc_or_lex == 0)) {
+ store_FNR = FNR
+ FNR = possible_FNR
+ store_0 = $0;
+ $0 = UCASE_full_line;
+ fail("UCASE function")
+ FNR = store_FNR
+ $0 = store_0;
+ }
+ possible_UCASE = 0
+ UCASE_full_line = ""
+ } else {
+ UCASE_full_line = UCASE_full_line $0;
+ }
+}
+/^[A-Z][[:alnum:]_]*[[:space:]]*\([^()]*(|\))[[:space:]]*$/ {
+ possible_UCASE = 1
+ if (ARI_OK == "UCASE function") {
+ possible_UCASE = 0
+ }
+ possible_FNR = FNR
+ UCASE_full_line = $0
+}
+
+
+BEGIN { doc["editCase function"] = "\
+Function name starts lower case but has uppercased letters."
+ category["editCase function"] = ari_code
+ possible_editCase = 0
+ editCase_full_line = ""
+}
+(possible_editCase) {
+ if (ARI_OK == "ediCase function") {
+ possible_editCase = 0
+ }
+ # Closing brace found?
+ else if (editCase_full_line ~ \
+/^[a-z][a-z0-9_]*[A-Z][a-z0-9A-Z_]*[[:space:]]*\([^()]*\).*$/) {
+ if ((editCase_full_line ~ \
+/^[a-z][a-z0-9_]*[A-Z][a-z0-9A-Z_]*[[:space:]]*\([^()]*\)[[:space:]]*$/) \
+ && ($0 ~ /^\{/) && (is_yacc_or_lex == 0)) {
+ store_FNR = FNR
+ FNR = possible_FNR
+ store_0 = $0;
+ $0 = editCase_full_line;
+ fail("editCase function")
+ FNR = store_FNR
+ $0 = store_0;
+ }
+ possible_editCase = 0
+ editCase_full_line = ""
+ } else {
+ editCase_full_line = editCase_full_line $0;
+ }
+}
+/^[a-z][a-z0-9_]*[A-Z][a-z0-9A-Z_]*[[:space:]]*\([^()]*(|\))[[:space:]]*$/
{
+ possible_editCase = 1
+ if (ARI_OK == "editCase function") {
+ possible_editCase = 0
+ }
+ possible_FNR = FNR
+ editCase_full_line = $0
+}
+
+# Only function implementation should be on first column
+BEGIN { doc["function call in first column"] = "\
+Function name in first column should be restricted to function
implementation"
+ category["function call in first column"] = ari_code
+}
+/^[a-z][a-z0-9_]*[[:space:]]*\((|[^*][^()]*)\)[[:space:]]*[^ \t]+/ {
+ fail("function call in first column")
+}
+
+
+# Functions without any parameter should have (void)
+# after their name not simply ().
+BEGIN { doc["no parameter function"] = "\
+Function having no parameter should be declared with funcname (void)."
+ category["no parameter function"] = ari_code
+}
+/^[a-zA-Z][a-z0-9A-Z_]*[[:space:]]*\(\)/ {
+ fail("no parameter function")
+}
+
+BEGIN { doc["hash"] = "\
+Do not use ` #...'\'', instead use `#...'\''(some compilers only correctly
\
+parse a C preprocessor directive when `#'\'' is the first character on \
+the line)"
+ category["hash"] = ari_regression
+}
+/^[[:space:]]+#/ {
+ fail("hash")
+}
+
+BEGIN { doc["OP eol"] = "\
+Do not use &&, or || at the end of a line"
+ category["OP eol"] = ari_code
+}
+/(\|\||\&\&|==|!=)[[:space:]]*$/ {
+ fail("OP eol")
+}
+
+BEGIN { doc["strerror"] = "\
+Do not use strerror(), instead use safe_strerror()"
+ category["strerror"] = ari_regression
+ fix("strerror", "gdb/gdb_string.h", 1)
+ fix("strerror", "gdb/mingw-hdep.c", 1)
+ fix("strerror", "gdb/posix-hdep.c", 1)
+}
+/(^|[^_[:alnum:]])strerror[[:space:]]*\(/ {
+ fail("strerror")
+}
+
+BEGIN { doc["long long"] = "\
+Do not use `long long'\'', instead use LONGEST"
+ category["long long"] = ari_code
+ # defs.h needs two such patterns for LONGEST and ULONGEST definitions
+ fix("long long", "gdb/defs.h", 2)
+}
+/(^|[^_[:alnum:]])long[[:space:]]+long([^_[:alnum:]]|$)/ {
+ fail("long long")
+}
+
+BEGIN { doc["ATTRIBUTE_UNUSED"] = "\
+Do not use ATTRIBUTE_UNUSED, do not bother (GDB is compiled with -Werror
and, \
+consequently, is not able to tolerate false warnings. Since -Wunused-param
\
+produces such warnings, neither that warning flag nor ATTRIBUTE_UNUSED \
+are used by GDB"
+ category["ATTRIBUTE_UNUSED"] = ari_regression
+}
+/(^|[^_[:alnum:]])ATTRIBUTE_UNUSED([^_[:alnum:]]|$)/ {
+ fail("ATTRIBUTE_UNUSED")
+}
+
+BEGIN { doc["ATTR_FORMAT"] = "\
+Do not use ATTR_FORMAT, use ATTRIBUTE_PRINTF instead"
+ category["ATTR_FORMAT"] = ari_regression
+}
+/(^|[^_[:alnum:]])ATTR_FORMAT([^_[:alnum:]]|$)/ {
+ fail("ATTR_FORMAT")
+}
+
+BEGIN { doc["ATTR_NORETURN"] = "\
+Do not use ATTR_NORETURN, use ATTRIBUTE_NORETURN instead"
+ category["ATTR_NORETURN"] = ari_regression
+}
+/(^|[^_[:alnum:]])ATTR_NORETURN([^_[:alnum:]]|$)/ {
+ fail("ATTR_NORETURN")
+}
+
+BEGIN { doc["NORETURN"] = "\
+Do not use NORETURN, use ATTRIBUTE_NORETURN instead"
+ category["NORETURN"] = ari_regression
+}
+/(^|[^_[:alnum:]])NORETURN([^_[:alnum:]]|$)/ {
+ fail("NORETURN")
+}
+
+
+# General problems
+
+BEGIN { doc["multiple messages"] = "\
+Do not use multiple calls to warning or error, instead use a single call"
+ category["multiple messages"] = ari_gettext
+}
+FNR == 1 {
+ warning_fnr = -1
+}
+/(^|[^_[:alnum:]])(warning|error)[[:space:]]*\(/ {
+ if (FNR == warning_fnr + 1) {
+ fail("multiple messages")
+ } else {
+ warning_fnr = FNR
+ }
+}
+
+# Commented out, but left inside sources, just in case.
+# BEGIN { doc["inline"] = "\
+# Do not use the inline attribute; \
+# since the compiler generally ignores this, better algorithm selection \
+# is needed to improved performance"
+# category["inline"] = ari_code
+# }
+# /(^|[^_[:alnum:]])inline([^_[:alnum:]]|$)/ {
+# fail("inline")
+# }
+
+# This test is obsolete as this type
+# has been deprecated and finally suppressed from GDB sources
+#BEGIN { doc["obj_private"] = "\
+#Replace obj_private with objfile_data"
+# category["obj_private"] = ari_obsolete
+#}
+#/(^|[^_[:alnum:]])obj_private([^_[:alnum:]]|$)/ {
+# fail("obj_private")
+#}
+
+BEGIN { doc["abort"] = "\
+Do not use abort, instead use internal_error; GDB should never abort"
+ category["abort"] = ari_regression
+ fix("abort", "gdb/utils.c", 3)
+}
+/(^|[^_[:alnum:]])abort[[:space:]]*\(/ {
+ fail("abort")
+}
+
+BEGIN { doc["basename"] = "\
+Do not use basename, instead use lbasename"
+ category["basename"] = ari_regression
+}
+/(^|[^_[:alnum:]])basename[[:space:]]*\(/ {
+ fail("basename")
+}
+
+BEGIN { doc["assert"] = "\
+Do not use assert, instead use gdb_assert or internal_error; assert \
+calls abort and GDB should never call abort"
+ category["assert"] = ari_regression
+}
+/(^|[^_[:alnum:]])assert[[:space:]]*\(/ {
+ fail("assert")
+}
+
+BEGIN { doc["TARGET_HAS_HARDWARE_WATCHPOINTS"] = "\
+Replace TARGET_HAS_HARDWARE_WATCHPOINTS with nothing, not needed"
+ category["TARGET_HAS_HARDWARE_WATCHPOINTS"] = ari_regression
+}
+/(^|[^_[:alnum:]])TARGET_HAS_HARDWARE_WATCHPOINTS([^_[:alnum:]]|$)/ {
+ fail("TARGET_HAS_HARDWARE_WATCHPOINTS")
+}
+
+BEGIN { doc["ADD_SHARED_SYMBOL_FILES"] = "\
+Replace ADD_SHARED_SYMBOL_FILES with nothing, not needed?"
+ category["ADD_SHARED_SYMBOL_FILES"] = ari_regression
+}
+/(^|[^_[:alnum:]])ADD_SHARED_SYMBOL_FILES([^_[:alnum:]]|$)/ {
+ fail("ADD_SHARED_SYMBOL_FILES")
+}
+
+BEGIN { doc["SOLIB_ADD"] = "\
+Replace SOLIB_ADD with nothing, not needed?"
+ category["SOLIB_ADD"] = ari_deprecate
+}
+/(^|[^_[:alnum:]])SOLIB_ADD([^_[:alnum:]]|$)/ {
+ fail("SOLIB_ADD")
+}
+
+BEGIN { doc["SOLIB_CREATE_INFERIOR_HOOK"] = "\
+Replace SOLIB_CREATE_INFERIOR_HOOK with nothing, not needed?"
+ category["SOLIB_CREATE_INFERIOR_HOOK"] = ari_deprecate
+}
+/(^|[^_[:alnum:]])SOLIB_CREATE_INFERIOR_HOOK([^_[:alnum:]]|$)/ {
+ fail("SOLIB_CREATE_INFERIOR_HOOK")
+}
+
+BEGIN { doc["SOLIB_LOADED_LIBRARY_PATHNAME"] = "\
+Replace SOLIB_LOADED_LIBRARY_PATHNAME with nothing, not needed?"
+ category["SOLIB_LOADED_LIBRARY_PATHNAME"] = ari_regression
+}
+/(^|[^_[:alnum:]])SOLIB_LOADED_LIBRARY_PATHNAME([^_[:alnum:]]|$)/ {
+ fail("SOLIB_LOADED_LIBRARY_PATHNAME")
+}
+
+BEGIN { doc["REGISTER_U_ADDR"] = "\
+Replace REGISTER_U_ADDR with nothing, not needed?"
+ category["REGISTER_U_ADDR"] = ari_regression
+}
+/(^|[^_[:alnum:]])REGISTER_U_ADDR([^_[:alnum:]]|$)/ {
+ fail("REGISTER_U_ADDR")
+}
+
+BEGIN { doc["PROCESS_LINENUMBER_HOOK"] = "\
+Replace PROCESS_LINENUMBER_HOOK with nothing, not needed?"
+ category["PROCESS_LINENUMBER_HOOK"] = ari_regression
+}
+/(^|[^_[:alnum:]])PROCESS_LINENUMBER_HOOK([^_[:alnum:]]|$)/ {
+ fail("PROCESS_LINENUMBER_HOOK")
+}
+
+BEGIN { doc["PC_SOLIB"] = "\
+Replace PC_SOLIB with nothing, not needed?"
+ category["PC_SOLIB"] = ari_deprecate
+}
+/(^|[^_[:alnum:]])PC_SOLIB([^_[:alnum:]]|$)/ {
+ fail("PC_SOLIB")
+}
+
+BEGIN { doc["IN_SOLIB_DYNSYM_RESOLVE_CODE"] = "\
+Replace IN_SOLIB_DYNSYM_RESOLVE_CODE with nothing, not needed?"
+ category["IN_SOLIB_DYNSYM_RESOLVE_CODE"] = ari_regression
+}
+/(^|[^_[:alnum:]])IN_SOLIB_DYNSYM_RESOLVE_CODE([^_[:alnum:]]|$)/ {
+ fail("IN_SOLIB_DYNSYM_RESOLVE_CODE")
+}
+
+BEGIN { doc["GCC_COMPILED_FLAG_SYMBOL"] = "\
+Replace GCC_COMPILED_FLAG_SYMBOL with nothing, not needed?"
+ category["GCC_COMPILED_FLAG_SYMBOL"] = ari_deprecate
+}
+/(^|[^_[:alnum:]])GCC_COMPILED_FLAG_SYMBOL([^_[:alnum:]]|$)/ {
+ fail("GCC_COMPILED_FLAG_SYMBOL")
+}
+
+BEGIN { doc["GCC2_COMPILED_FLAG_SYMBOL"] = "\
+Replace GCC2_COMPILED_FLAG_SYMBOL with nothing, not needed?"
+ category["GCC2_COMPILED_FLAG_SYMBOL"] = ari_deprecate
+}
+/(^|[^_[:alnum:]])GCC2_COMPILED_FLAG_SYMBOL([^_[:alnum:]]|$)/ {
+ fail("GCC2_COMPILED_FLAG_SYMBOL")
+}
+
+BEGIN { doc["FUNCTION_EPILOGUE_SIZE"] = "\
+Replace FUNCTION_EPILOGUE_SIZE with nothing, not needed?"
+ category["FUNCTION_EPILOGUE_SIZE"] = ari_regression
+}
+/(^|[^_[:alnum:]])FUNCTION_EPILOGUE_SIZE([^_[:alnum:]]|$)/ {
+ fail("FUNCTION_EPILOGUE_SIZE")
+}
+
+BEGIN { doc["HAVE_VFORK"] = "\
+Do not use HAVE_VFORK, instead include \"gdb_vfork.h\" and call vfork() \
+unconditionally"
+ category["HAVE_VFORK"] = ari_regression
+}
+/(^|[^_[:alnum:]])HAVE_VFORK([^_[:alnum:]]|$)/ {
+ fail("HAVE_VFORK")
+}
+
+BEGIN { doc["bcmp"] = "\
+Do not use bcmp(), ISO C 90 implies memcmp()"
+ category["bcmp"] = ari_regression
+}
+/(^|[^_[:alnum:]])bcmp[[:space:]]*\(/ {
+ fail("bcmp")
+}
+
+BEGIN { doc["setlinebuf"] = "\
+Do not use setlinebuf(), ISO C 90 implies setvbuf()"
+ category["setlinebuf"] = ari_regression
+}
+/(^|[^_[:alnum:]])setlinebuf[[:space:]]*\(/ {
+ fail("setlinebuf")
+}
+
+BEGIN { doc["bcopy"] = "\
+Do not use bcopy(), ISO C 90 implies memcpy() and memmove()"
+ category["bcopy"] = ari_regression
+}
+/(^|[^_[:alnum:]])bcopy[[:space:]]*\(/ {
+ fail("bcopy")
+}
+
+BEGIN { doc["get_frame_base"] = "\
+Replace get_frame_base with get_frame_id, get_frame_base_address, \
+get_frame_locals_address, or get_frame_args_address."
+ category["get_frame_base"] = ari_obsolete
+}
+/(^|[^_[:alnum:]])get_frame_base([^_[:alnum:]]|$)/ {
+ fail("get_frame_base")
+}
+
+BEGIN { doc["floatformat_to_double"] = "\
+Do not use floatformat_to_double() from libierty, \
+instead use floatformat_to_doublest()"
+ fix("floatformat_to_double", "gdb/doublest.c", 1)
+ category["floatformat_to_double"] = ari_regression
+}
+/(^|[^_[:alnum:]])floatformat_to_double[[:space:]]*\(/ {
+ fail("floatformat_to_double")
+}
+
+BEGIN { doc["floatformat_from_double"] = "\
+Do not use floatformat_from_double() from libierty, \
+instead use floatformat_from_doublest()"
+ category["floatformat_from_double"] = ari_regression
+}
+/(^|[^_[:alnum:]])floatformat_from_double[[:space:]]*\(/ {
+ fail("floatformat_from_double")
+}
+
+BEGIN { doc["BIG_ENDIAN"] = "\
+Do not use BIG_ENDIAN, instead use BFD_ENDIAN_BIG"
+ category["BIG_ENDIAN"] = ari_regression
+}
+/(^|[^_[:alnum:]])BIG_ENDIAN([^_[:alnum:]]|$)/ {
+ fail("BIG_ENDIAN")
+}
+
+BEGIN { doc["LITTLE_ENDIAN"] = "\
+Do not use LITTLE_ENDIAN, instead use BFD_ENDIAN_LITTLE";
+ category["LITTLE_ENDIAN"] = ari_regression
+}
+/(^|[^_[:alnum:]])LITTLE_ENDIAN([^_[:alnum:]]|$)/ {
+ fail("LITTLE_ENDIAN")
+}
+
+BEGIN { doc["BIG_ENDIAN"] = "\
+Do not use BIG_ENDIAN, instead use BFD_ENDIAN_BIG"
+ category["BIG_ENDIAN"] = ari_regression
+}
+/(^|[^_[:alnum:]])BIG_ENDIAN([^_[:alnum:]]|$)/ {
+ fail("BIG_ENDIAN")
+}
+
+BEGIN { doc["sec_ptr"] = "\
+Instead of sec_ptr, use struct bfd_section";
+ category["sec_ptr"] = ari_regression
+}
+/(^|[^_[:alnum:]])sec_ptr([^_[:alnum:]]|$)/ {
+ fail("sec_ptr")
+}
+
+BEGIN { doc["frame_unwind_unsigned_register"] = "\
+Replace frame_unwind_unsigned_register with frame_unwind_register_unsigned"
+ category["frame_unwind_unsigned_register"] = ari_regression
+}
+/(^|[^_[:alnum:]])frame_unwind_unsigned_register([^_[:alnum:]]|$)/ {
+ fail("frame_unwind_unsigned_register")
+}
+
+BEGIN { doc["frame_register_read"] = "\
+Replace frame_register_read() with get_frame_register(), or \
+possibly introduce a new method safe_get_frame_register()"
+ category["frame_register_read"] = ari_obsolete
+}
+/(^|[^_[:alnum:]])frame_register_read([^_[:alnum:]]|$)/ {
+ fail("frame_register_read")
+}
+
+BEGIN { doc["read_register"] = "\
+Replace read_register() with regcache_read() et.al."
+ category["read_register"] = ari_regression
+}
+/(^|[^_[:alnum:]])read_register([^_[:alnum:]]|$)/ {
+ fail("read_register")
+}
+
+BEGIN { doc["write_register"] = "\
+Replace write_register() with regcache_read() et.al."
+ category["write_register"] = ari_regression
+}
+/(^|[^_[:alnum:]])write_register([^_[:alnum:]]|$)/ {
+ fail("write_register")
+}
+
+function report(name) {
+ # Drop any trailing _P.
+ name = gensub(/(_P|_p)$/, "", 1, name)
+ # Convert to lower case
+ name = tolower(name)
+ # Split into category and bug
+ cat = gensub(/^([[:alpha:]]+)_([_[:alnum:]]*)$/, "\\1", 1, name)
+ bug = gensub(/^([[:alpha:]]+)_([_[:alnum:]]*)$/, "\\2", 1, name)
+ # Report it
+ name = cat " " bug
+ doc[name] = "Do not use " cat " " bug ", see declaration for details"
+ category[name] = cat
+ fail(name)
+}
+
+/(^|[^_[:alnum:]])(DEPRECATED|deprecated|set_gdbarch_deprecated|LEGACY|lega
cy|set_gdbarch_legacy)_/ {
+ line = $0
+ # print "0 =", $0
+ while (1) {
+ name =
gensub(/^(|.*[^_[:alnum:]])((DEPRECATED|deprecated|LEGACY|legacy)_[_[:alnum:
]]*)(.*)$/, "\\2", 1, line)
+ line =
gensub(/^(|.*[^_[:alnum:]])((DEPRECATED|deprecated|LEGACY|legacy)_[_[:alnum:
]]*)(.*)$/, "\\1 \\4", 1, line)
+ # print "name =", name, "line =", line
+ if (name == line) break;
+ report(name)
+ }
+}
+
+# Count the number of times each architecture method is set
+/(^|[^_[:alnum:]])set_gdbarch_[_[:alnum:]]*([^_[:alnum:]]|$)/ {
+ name = gensub(/^.*set_gdbarch_([_[:alnum:]]*).*$/, "\\1", 1, $0)
+ doc["set " name] = "\
+Call to set_gdbarch_" name
+ category["set " name] = ari_gdbarch
+ fail("set " name)
+}
+
+# Count the number of times each tm/xm/nm macro is defined or undefined
+/^#[[:space:]]*(undef|define)[[:space:]]+[[:alnum:]_]+.*$/ \
+&&
!/^#[[:space:]]*(undef|define)[[:space:]]+[[:alnum:]_]+_H($|[[:space:]])/ \
+&& FILENAME ~ /(^|\/)config\/(|[^\/]*\/)(tm-|xm-|nm-).*\.h$/ {
+ basename = gensub(/(^|.*\/)([^\/]*)$/, "\\2", 1, FILENAME)
+ type = gensub(/^(tm|xm|nm)-.*\.h$/, "\\1", 1, basename)
+ name =
gensub(/^#[[:space:]]*(undef|define)[[:space:]]+([[:alnum:]_]+).*$/, "\\2",
1, $0)
+ if (type == basename) {
+ type = "macro"
+ }
+ doc[type " " name] = "\
+Do not define macros such as " name " in a tm, nm or xm file, \
+in fact do not provide a tm, nm or xm file"
+ category[type " " name] = ari_macro
+ fail(type " " name)
+}
+
+BEGIN { doc["deprecated_registers"] = "\
+Replace deprecated_registers with nothing, they have reached \
+end-of-life"
+ category["deprecated_registers"] = ari_eol
+}
+/(^|[^_[:alnum:]])deprecated_registers([^_[:alnum:]]|$)/ {
+ fail("deprecated_registers")
+}
+
+BEGIN { doc["read_pc"] = "\
+Replace READ_PC() with frame_pc_unwind; \
+at present the inferior function call code still uses this"
+ category["read_pc"] = ari_deprecate
+}
+/(^|[^_[:alnum:]])read_pc[[:space:]]*\(/ || \
+/(^|[^_[:alnum:]])set_gdbarch_read_pc[[:space:]]*\(/ || \
+/(^|[^_[:alnum:]])TARGET_READ_PC[[:space:]]*\(/ {
+ fail("read_pc")
+}
+
+BEGIN { doc["write_pc"] = "\
+Replace write_pc() with get_frame_base_address or get_frame_id; \
+at present the inferior function call code still uses this when doing \
+a DECR_PC_AFTER_BREAK"
+ category["write_pc"] = ari_deprecate
+}
+/(^|[^_[:alnum:]])write_pc[[:space:]]*\(/ || \
+/(^|[^_[:alnum:]])set_gdbarch_write_pc[[:space:]]*\(/ || \
+/(^|[^_[:alnum:]])TARGET_WRITE_PC[[:space:]]*\(/ {
+ fail("write_pc")
+}
+
+BEGIN { doc["generic_target_write_pc"] = "\
+Replace generic_target_write_pc with a per-architecture implementation, \
+this relies on PC_REGNUM which is being eliminated"
+ category["generic_target_write_pc"] = ari_regression
+}
+/(^|[^_[:alnum:]])generic_target_write_pc([^_[:alnum:]]|$)/ {
+ fail("generic_target_write_pc")
+}
+
+BEGIN { doc["read_sp"] = "\
+Replace read_sp() with frame_sp_unwind"
+ category["read_sp"] = ari_regression
+}
+/(^|[^_[:alnum:]])read_sp[[:space:]]*\(/ || \
+/(^|[^_[:alnum:]])set_gdbarch_read_sp[[:space:]]*\(/ || \
+/(^|[^_[:alnum:]])TARGET_READ_SP[[:space:]]*\(/ {
+ fail("read_sp")
+}
+
+BEGIN { doc["register_cached"] = "\
+Replace register_cached() with nothing, does not have a regcache parameter"
+ category["register_cached"] = ari_regression
+}
+/(^|[^_[:alnum:]])register_cached[[:space:]]*\(/ {
+ fail("register_cached")
+}
+
+BEGIN { doc["set_register_cached"] = "\
+Replace set_register_cached() with nothing, does not have a regcache
parameter"
+ category["set_register_cached"] = ari_regression
+}
+/(^|[^_[:alnum:]])set_register_cached[[:space:]]*\(/ {
+ fail("set_register_cached")
+}
+
+# Print functions: Use versions that either check for buffer overflow
+# or safely allocate a fresh buffer.
+
+BEGIN { doc["sprintf"] = "\
+Do not use sprintf, instead use xsnprintf or xstrprintf"
+ category["sprintf"] = ari_code
+}
+/(^|[^_[:alnum:]])sprintf[[:space:]]*\(/ {
+ fail("sprintf")
+}
+
+BEGIN { doc["vsprintf"] = "\
+Do not use vsprintf(), instead use xstrvprintf"
+ category["vsprintf"] = ari_regression
+}
+/(^|[^_[:alnum:]])vsprintf[[:space:]]*\(/ {
+ fail("vsprintf")
+}
+
+BEGIN { doc["asprintf"] = "\
+Do not use asprintf(), instead use xstrprintf()"
+ category["asprintf"] = ari_regression
+}
+/(^|[^_[:alnum:]])asprintf[[:space:]]*\(/ {
+ fail("asprintf")
+}
+
+BEGIN { doc["vasprintf"] = "\
+Do not use vasprintf(), instead use xstrvprintf"
+ fix("vasprintf", "gdb/utils.c", 1)
+ category["vasprintf"] = ari_regression
+}
+/(^|[^_[:alnum:]])vasprintf[[:space:]]*\(/ {
+ fail("vasprintf")
+}
+
+BEGIN { doc["xasprintf"] = "\
+Do not use xasprintf(), instead use xstrprintf"
+ fix("xasprintf", "gdb/defs.h", 1)
+ fix("xasprintf", "gdb/utils.c", 1)
+ category["xasprintf"] = ari_regression
+}
+/(^|[^_[:alnum:]])xasprintf[[:space:]]*\(/ {
+ fail("xasprintf")
+}
+
+BEGIN { doc["xvasprintf"] = "\
+Do not use xvasprintf(), instead use xstrvprintf"
+ fix("xvasprintf", "gdb/defs.h", 1)
+ fix("xvasprintf", "gdb/utils.c", 1)
+ category["xvasprintf"] = ari_regression
+}
+/(^|[^_[:alnum:]])xvasprintf[[:space:]]*\(/ {
+ fail("xvasprintf")
+}
+
+# More generic memory operations
+
+BEGIN { doc["bzero"] = "\
+Do not use bzero(), instead use memset()"
+ category["bzero"] = ari_regression
+}
+/(^|[^_[:alnum:]])bzero[[:space:]]*\(/ {
+ fail("bzero")
+}
+
+BEGIN { doc["strdup"] = "\
+Do not use strdup(), instead use xstrdup()";
+ category["strdup"] = ari_regression
+}
+/(^|[^_[:alnum:]])strdup[[:space:]]*\(/ {
+ fail("strdup")
+}
+
+BEGIN { doc["strsave"] = "\
+Do not use strsave(), instead use xstrdup() et.al."
+ category["strsave"] = ari_regression
+}
+/(^|[^_[:alnum:]])strsave[[:space:]]*\(/ {
+ fail("strsave")
+}
+
+# String compare functions
+
+BEGIN { doc["strnicmp"] = "\
+Do not use strnicmp(), instead use strncasecmp()"
+ category["strnicmp"] = ari_regression
+}
+/(^|[^_[:alnum:]])strnicmp[[:space:]]*\(/ {
+ fail("strnicmp")
+}
+
+# Boolean expressions and conditionals
+
+BEGIN { doc["boolean"] = "\
+Do not use `boolean'\'', use `int'\'' instead"
+ category["boolean"] = ari_regression
+}
+/(^|[^_[:alnum:]])boolean([^_[:alnum:]]|$)/ {
+ if (is_yacc_or_lex == 0) {
+ fail("boolean")
+ }
+}
+
+BEGIN { doc["false"] = "\
+Definitely do not use `false'\'' in boolean expressions"
+ category["false"] = ari_regression
+}
+/(^|[^_[:alnum:]])false([^_[:alnum:]]|$)/ {
+ if (is_yacc_or_lex == 0) {
+ fail("false")
+ }
+}
+
+BEGIN { doc["true"] = "\
+Do not try to use `true'\'' in boolean expressions"
+ category["true"] = ari_regression
+}
+/(^|[^_[:alnum:]])true([^_[:alnum:]]|$)/ {
+ if (is_yacc_or_lex == 0) {
+ fail("true")
+ }
+}
+
+# Typedefs that are either redundant or can be reduced to `struct
+# type *''.
+# Must be placed before if assignment otherwise ARI exceptions
+# are not handled correctly.
+
+BEGIN { doc["d_namelen"] = "\
+Do not use dirent.d_namelen, instead use NAMELEN"
+ category["d_namelen"] = ari_regression
+}
+/(^|[^_[:alnum:]])d_namelen([^_[:alnum:]]|$)/ {
+ fail("d_namelen")
+}
+
+BEGIN { doc["strlen d_name"] = "\
+Do not use strlen dirent.d_name, instead use NAMELEN"
+ category["strlen d_name"] = ari_regression
+}
+/(^|[^_[:alnum:]])strlen[[:space:]]*\(.*[^_[:alnum:]]d_name([^_[:alnum:]]|$
)/ {
+ fail("strlen d_name")
+}
+
+BEGIN { doc["var_boolean"] = "\
+Replace var_boolean with add_setshow_boolean_cmd"
+ category["var_boolean"] = ari_regression
+ fix("var_boolean", "gdb/command.h", 1)
+ # fix only uses the last directory level
+ fix("var_boolean", "cli/cli-decode.c", 2)
+}
+/(^|[^_[:alnum:]])var_boolean([^_[:alnum:]]|$)/ {
+ if ($0 !~ /(^|[^_[:alnum:]])case *var_boolean:/) {
+ fail("var_boolean")
+ }
+}
+
+BEGIN { doc["generic_use_struct_convention"] = "\
+Replace generic_use_struct_convention with nothing, \
+EXTRACT_STRUCT_VALUE_ADDRESS is a predicate"
+ category["generic_use_struct_convention"] = ari_regression
+}
+/(^|[^_[:alnum:]])generic_use_struct_convention([^_[:alnum:]]|$)/ {
+ fail("generic_use_struct_convention")
+}
+
+BEGIN { doc["if assignment"] = "\
+An IF statement'\''s expression contains an assignment (the GNU coding \
+standard discourages this)"
+ category["if assignment"] = ari_code
+}
+BEGIN { doc["if clause more than 50 lines"] = "\
+An IF statement'\''s expression expands over 50 lines"
+ category["if clause more than 50 lines"] = ari_code
+}
+#
+# Accumulate continuation lines
+FNR == 1 {
+ in_if = 0
+}
+
+/(^|[^_[:alnum:]])if / {
+ in_if = 1;
+ if_brace_level = 0;
+ if_cont_p = 0;
+ if_count = 0;
+ if_brace_end_pos = 0;
+ if_full_line = "";
+}
+(in_if) {
+ # We want everything up to closing brace of same level
+ if_count++;
+ if (if_count > 50) {
+ print "multiline if: " if_full_line $0
+ fail("if clause more than 50 lines")
+ if_brace_level = 0;
+ if_full_line = "";
+ } else {
+ if (if_count == 1) {
+ i = index($0,"if ");
+ } else {
+ i = 1;
+ }
+ for (i=i; i <= length($0); i++) {
+ char = substr($0,i,1);
+ if (char == "(") { if_brace_level++; }
+ if (char == ")") {
+ if_brace_level--;
+ if (!if_brace_level) {
+ if_brace_end_pos = i;
+ after_if = substr($0,i+1,length($0));
+ # Do not parse what is following
+ break;
+ }
+ }
+ }
+ if (if_brace_level == 0) {
+ $0 = substr($0,1,i);
+ in_if = 0;
+ } else {
+ if_full_line = if_full_line $0;
+ if_cont_p = 1;
+ next;
+ }
+ }
+}
+# if we arrive here, we need to concatenate, but we are at brace level 0
+
+(if_brace_end_pos) {
+ $0 = if_full_line substr($0,1,if_brace_end_pos);
+ if (if_count > 1) {
+ # print "IF: multi line " if_count " found at " FILENAME ":" FNR "
\"" $0 "\""
+ }
+ if_cont_p = 0;
+ if_full_line = "";
+}
+/(^|[^_[:alnum:]])if .* = / {
+ # print "fail in if " $0
+ fail("if assignment")
+}
+(if_brace_end_pos) {
+ $0 = $0 after_if;
+ if_brace_end_pos = 0;
+ in_if = 0;
+}
+
+# Printout of all found bug
+
+BEGIN {
+ if (print_doc) {
+ for (bug in doc) {
+ fail(bug)
+ }
+ exit
+ }
+}' "$@"
+
Index: gdb_find.sh
===================================================================
RCS file: gdb_find.sh
diff -N gdb_find.sh
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ gdb_find.sh 18 May 2012 07:15:58 -0000
@@ -0,0 +1,41 @@
+#!/bin/sh
+
+# GDB script to create list of files to check using gdb_ari.sh.
+#
+# Copyright (C) 2003-2012 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 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/>.
+
+# Make certain that the script is not running in an internationalized
+# environment.
+
+LANG=C ; export LANG
+LC_ALL=C ; export LC_ALL
+
+
+# A find that prunes files that GDB users shouldn't be interested in.
+# Use sort to order files alphabetically.
+
+find "$@" \
+ -name testsuite -prune -o \
+ -name gdbserver -prune -o \
+ -name gnulib -prune -o \
+ -name osf-share -prune -o \
+ -name '*-stub.c' -prune -o \
+ -name '*-exp.c' -prune -o \
+ -name ada-lex.c -prune -o \
+ -name cp-name-parser.c -prune -o \
+ -type f -name '*.[lyhc]' -print | sort
Index: update-web-ari.sh
===================================================================
RCS file: update-web-ari.sh
diff -N update-web-ari.sh
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ update-web-ari.sh 18 May 2012 07:15:58 -0000
@@ -0,0 +1,947 @@
+#!/bin/sh -x
+
+# GDB script to create GDB ARI web page.
+#
+# Copyright (C) 2001-2012 Free Software Foundation, Inc.
+#
+# This file is part of GDB.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+# TODO: setjmp.h, setjmp and longjmp.
+
+# Direct stderr into stdout but still hang onto stderr (/dev/fd/3)
+exec 3>&2 2>&1
+ECHO ()
+{
+# echo "$@" | tee /dev/fd/3 1>&2
+ echo "$@" 1>&2
+ echo "$@" 1>&3
+}
+
+# Really mindless usage
+if test $# -ne 4
+then
+ echo "Usage: $0 <snapshot/sourcedir> <tmpdir> <destdir> <project>" 1>&2
+ exit 1
+fi
+snapshot=$1 ; shift
+tmpdir=$1 ; shift
+wwwdir=$1 ; shift
+project=$1 ; shift
+
+# Try to create destination directory if it doesn't exist yet
+if [ ! -d ${wwwdir} ]
+then
+ mkdir -p ${wwwdir}
+fi
+
+# Fail if destination directory doesn't exist or is not writable
+if [ ! -w ${wwwdir} -o ! -d ${wwwdir} ]
+then
+ echo ERROR: Can not write to directory ${wwwdir} >&2
+ exit 2
+fi
+
+if [ ! -r ${snapshot} ]
+then
+ echo ERROR: Can not read snapshot file 1>&2
+ exit 1
+fi
+
+# FILE formats
+# ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+# ari.*.idx: <BUG>:<FILE>:<CATEGORY>
+# ari.*.doc: <BUG>:<COUNT>:<CATEGORY>:<DOC>
+# Where ``*'' is {source,warning,indent,doschk}
+
+unpack_source_p=true
+delete_source_p=true
+
+check_warning_p=false # broken
+check_indent_p=false # too slow, too many fail
+check_source_p=true
+check_doschk_p=true
+check_werror_p=true
+
+update_doc_p=true
+update_web_p=true
+
+if [ ! -z "$send_email" ]
+then
+ send_email=false
+fi
+
+if awk --version 2>&1 </dev/null | grep -i gnu > /dev/null
+then
+ AWK=awk
+else
+ AWK=gawk
+fi
+
+
+# Set up a few cleanups
+if ${delete_source_p}
+then
+ trap "cd /tmp; rm -rf ${tmpdir}; exit" 0 1 2 15
+fi
+
+
+# If the first parameter is a directory,
+#we just use it as the extracted source
+if [ -d ${snapshot} ]
+then
+ module=${project}
+ srcdir=${snapshot}
+ aridir=${srcdir}/${module}/ari
+ unpack_source_p=false
+ delete_source_p=false
+ version_in=${srcdir}/${module}/version.in
+else
+ # unpack the tar-ball
+ if ${unpack_source_p}
+ then
+ # Was it previously unpacked?
+ if ${delete_source_p} || test ! -d ${tmpdir}/${module}*
+ then
+ /bin/rm -rf "${tmpdir}"
+ /bin/mkdir -p ${tmpdir}
+ if [ ! -d ${tmpdir} ]
+ then
+ echo "Problem creating work directory"
+ exit 1
+ fi
+ cd ${tmpdir} || exit 1
+ echo `date`: Unpacking tar-ball ...
+ case ${snapshot} in
+ *.tar.bz2 ) bzcat ${snapshot} ;;
+ *.tar ) cat ${snapshot} ;;
+ * ) ECHO Bad file ${snapshot} ; exit 1 ;;
+ esac | tar xf -
+ fi
+ fi
+
+ module=`basename ${snapshot}`
+ module=`basename ${module} .bz2`
+ module=`basename ${module} .tar`
+ srcdir=`echo ${tmpdir}/${module}*`
+ aridir=${HOME}/ss
+ version_in=${srcdir}/gdb/version.in
+fi
+
+if [ ! -r ${version_in} ]
+then
+ echo ERROR: missing version file 1>&2
+ exit 1
+fi
+version=`cat ${version_in}`
+
+
+# THIS HAS SUFFERED BIT ROT
+if ${check_warning_p} && test -d "${srcdir}"
+then
+ echo `date`: Parsing compiler warnings 1>&2
+ cat ${root}/ari.compile | $AWK '
+BEGIN {
+ FS=":";
+}
+/^[^:]*:[0-9]*: warning:/ {
+ file = $1;
+ #sub (/^.*\//, "", file);
+ warning[file] += 1;
+}
+/^[^:]*:[0-9]*: error:/ {
+ file = $1;
+ #sub (/^.*\//, "", file);
+ error[file] += 1;
+}
+END {
+ for (file in warning) {
+ print file ":warning:" level[file]
+ }
+ for (file in error) {
+ print file ":error:" level[file]
+ }
+}
+' > ${root}/ari.warning.bug
+fi
+
+# THIS HAS SUFFERED BIT ROT
+if ${check_indent_p} && test -d "${srcdir}"
+then
+ printf "Analizing file indentation:" 1>&2
+ ( cd "${srcdir}" && /bin/sh ${aridir}/gdb_find.sh ${project} | while
read f
+ do
+ if /bin/sh ${aridir}/gdb_indent.sh < ${f} 2>/dev/null | cmp -s -
${f}
+ then
+ :
+ else
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ echo "${f}:0: info: indent: Indentation does not match GNU
indent output"
+ fi
+ done ) > ${wwwdir}/ari.indent.bug
+ echo ""
+fi
+
+if ${check_source_p} && test -d "${srcdir}"
+then
+ bugf=${wwwdir}/ari.source.bug
+ oldf=${wwwdir}/ari.source.old
+ srcf=${wwwdir}/ari.source.lines
+ oldsrcf=${wwwdir}/ari.source.lines-old
+
+ diff=${wwwdir}/ari.source.diff
+ diffin=${diff}-in
+ newf1=${bugf}1
+ oldf1=${oldf}1
+ oldpruned=${oldf1}-pruned
+ newpruned=${newf1}-pruned
+
+ cp -f ${bugf} ${oldf}
+ cp -f ${srcf} ${oldsrcf}
+ rm -f ${srcf}
+ node=`uname -n`
+ echo "`date`: Using source lines ${srcf}" 1>&2
+ echo "`date`: Checking source code" 1>&2
+ ( cd "${srcdir}" && /bin/sh ${aridir}/gdb_find.sh "${project}" | \
+ xargs /bin/sh ${aridir}/gdb_ari.sh -Werror -Wall --print-idx
--src=${srcf}
+ ) > ${bugf}
+ # Remove things we are not interested in to signal by email
+ # gdbarch changes are not important here
+ # Also convert ` into ' to avoid command substitution in script below
+ sed -e "/.*: gdbarch:.*/d" -e "s:\`:':g" ${oldf} > ${oldf1}
+ sed -e "/.*: gdbarch:.*/d" -e "s:\`:':g" ${bugf} > ${newf1}
+ # Remove line number info so that code inclusion/deletion
+ # has no impact on the result
+ sed -e "s/\([^:]*\):\([^:]*\):\(.*\)/\1:0:\3/" ${oldf1} > ${oldpruned}
+ sed -e "s/\([^:]*\):\([^:]*\):\(.*\)/\1:0:\3/" ${newf1} > ${newpruned}
+ # Use diff without option to get normal diff output that
+ # is reparsed after
+ diff ${oldpruned} ${newpruned} > ${diffin}
+ # Only keep new warnings
+ sed -n -e "/^>.*/p" ${diffin} > ${diff}
+ sedscript=${wwwdir}/sedscript
+ script=${wwwdir}/script
+ sed -n -e "s|\(^[0-9,]*\)a\(.*\)|echo \1a\2 \n \
+ sed -n \'\2s:\\\\(.*\\\\):> \\\\1:p\' ${newf1}|p" \
+ -e "s|\(^[0-9,]*\)d\(.*\)|echo \1d\2\n \
+ sed -n \'\1s:\\\\(.*\\\\):< \\\\1:p\' ${oldf1}|p" \
+ -e "s|\(^[0-9,]*\)c\(.*\)|echo \1c\2\n \
+ sed -n \'\1s:\\\\(.*\\\\):< \\\\1:p\' ${oldf1} \n \
+ sed -n \"\2s:\\\\(.*\\\\):> \\\\1:p\" ${newf1}|p" \
+ ${diffin} > ${sedscript}
+ ${SHELL} ${sedscript} > ${wwwdir}/message
+ sed -n \
+ -e "s;\(.*\);echo \\\"\1\\\";p" \
+ -e "s;.*< \([^:]*\):\([0-9]*\):.*;grep \"^\1:\2:\" ${oldsrcf};p" \
+ -e "s;.*> \([^:]*\):\([0-9]*\):.*;grep \"^\1:\2:\" ${srcf};p" \
+ ${wwwdir}/message > ${script}
+ ${SHELL} ${script} > ${wwwdir}/mail-message
+ if [ "x${branch}" != "x" ]; then
+ email_suffix="`date` in ${branch}"
+ else
+ email_suffix="`date`"
+ fi
+
+ if [ "$send_email" == "true" ]; then
+ if [ "${node}" = "sourceware.org" ]; then
+ warning_email=gdb-patches@sourceware.org
+ else
+ # Use default email
+ warning_email=${USER}@${node}
+ fi
+
+ # Check if ${diff} is not empty
+ if [ -s ${diff} ]; then
+ # Send an email $warning_email
+ mutt -s "New ARI warning ${email_suffix}" \
+ ${warning_email} < ${wwwdir}/mail-message
+ else
+ if [ -s ${wwwdir}/${mail-message} ]; then
+ # Send an email to $warning_email
+ mutt -s "ARI warning list change ${email_suffix}" \
+ ${warning_email} < ${wwwdir}/mail-message
+ fi
+ fi
+ fi
+fi
+
+
+
+
+if ${check_doschk_p} && test -d "${srcdir}"
+then
+ echo "`date`: Checking for doschk" 1>&2
+ rm -f "${wwwdir}"/ari.doschk.*
+ fnchange_lst="${srcdir}"/gdb/config/djgpp/fnchange.lst
+ fnchange_awk="${wwwdir}"/ari.doschk.awk
+ doschk_in="${wwwdir}"/ari.doschk.in
+ doschk_out="${wwwdir}"/ari.doschk.out
+ doschk_bug="${wwwdir}"/ari.doschk.bug
+ doschk_char="${wwwdir}"/ari.doschk.char
+
+ # Transform fnchange.lst into fnchange.awk. The program DJTAR
+ # does a textual substitution of each file name using the list.
+ # Generate an awk script that does the equivalent - matches an
+ # exact line and then outputs the replacement.
+
+ sed -e 's;@[^@]*@[/]*\([^ ]*\) @[^@]*@[/]*\([^ ]*\);\$0 == "\1" { print
"\2"\; next\; };' \
+ < "${fnchange_lst}" > "${fnchange_awk}"
+ echo '{ print }' >> "${fnchange_awk}"
+
+ # Do the raw analysis - transform the list of files into the DJGPP
+ # equivalents putting it in the .in file
+ ( cd "${srcdir}" && find * \
+ -name '*.info-[0-9]*' -prune \
+ -o -name tcl -prune \
+ -o -name itcl -prune \
+ -o -name tk -prune \
+ -o -name libgui -prune \
+ -o -name tix -prune \
+ -o -name dejagnu -prune \
+ -o -name expect -prune \
+ -o -type f -print ) \
+ | $AWK -f ${fnchange_awk} > ${doschk_in}
+
+ # Start with a clean slate
+ rm -f ${doschk_bug}
+
+ # Check for any invalid characters.
+ grep '[\+\,\;\=\[\]\|\<\>\\\"\:\?\*]' < ${doschk_in} > ${doschk_char}
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ sed < ${doschk_char} >> ${doschk_bug} \
+ -e s'/$/:0: dos: DOSCHK: Invalid DOS character/'
+
+ # Magic to map ari.doschk.out to ari.doschk.bug goes here
+ doschk < ${doschk_in} > ${doschk_out}
+ cat ${doschk_out} | $AWK >> ${doschk_bug} '
+BEGIN {
+ state = 1;
+ invalid_dos = state++; bug[invalid_dos] = "invalid DOS file name";
category[invalid_dos] = "dos";
+ same_dos = state++; bug[same_dos] = "DOS 8.3";
category[same_dos] = "dos";
+ same_sysv = state++; bug[same_sysv] = "SysV";
+ long_sysv = state++; bug[long_sysv] = "long SysV";
+ internal = state++; bug[internal] = "internal doschk";
category[internal] = "internal";
+ state = 0;
+}
+/^$/ { state = 0; next; }
+/^The .* not valid DOS/ { state = invalid_dos; next; }
+/^The .* same DOS/ { state = same_dos; next; }
+/^The .* same SysV/ { state = same_sysv; next; }
+/^The .* too long for SysV/ { state = long_sysv; next; }
+/^The .* / { state = internal; next; }
+
+NF == 0 { next }
+
+NF == 3 { name = $1 ; file = $3 }
+NF == 1 { file = $1 }
+NF > 3 && $2 == "-" { file = $1 ; name = gensub(/^.* - /, "", 1) }
+
+state == same_dos {
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ print file ":0: " category[state] ": " \
+ name " " bug[state] " " " dup: " \
+ " DOSCHK - the names " name " and " file " resolve to the same" \
+ " file on a " bug[state] \
+ " system.<br>For DOS, this can be fixed by modifying the file" \
+ " fnchange.lst."
+ next
+}
+state == invalid_dos {
+ # ari.*.bug: <FILE>:<LINE>: <SEVERITY>: <CATEGORY>: <DOC>
+ print file ":0: " category[state] ": " name ": DOSCHK - " name
+ next
+}
+state == internal {
+ # ari.*.bug: <FILE>:<LINE>: <SEVERITY>: <CATEGORY>: <DOC>
+ print file ":0: " category[state] ": " bug[state] ": DOSCHK - a " \
+ bug[state] " problem"
+}
+'
+fi
+
+
+
+if ${check_werror_p} && test -d "${srcdir}"
+then
+ echo "`date`: Checking Makefile.in for non- -Werror rules"
+ rm -f ${wwwdir}/ari.werror.*
+ cat "${srcdir}/${project}/Makefile.in" | $AWK >
${wwwdir}/ari.werror.bug '
+BEGIN {
+ count = 0
+ cont_p = 0
+ full_line = ""
+}
+/^[-_[:alnum:]]+\.o:/ {
+ file = gensub(/.o:.*/, "", 1) ".c"
+}
+
+/[^\\]\\$/ { gsub (/\\$/, ""); full_line = full_line $0; cont_p = 1; next;
}
+cont_p { $0 = full_line $0; cont_p = 0; full_line = ""; }
+
+/\$\(COMPILE\.pre\)/ {
+ print file " has line " $0
+ if (($0 !~ /\$\(.*ERROR_CFLAGS\)/) && ($0 !~ /\$\(INTERNAL_CFLAGS\)/))
{
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ print "'"${project}"'/" file ":0: info: Werror: The file is not
being compiled with -Werror"
+ }
+}
+'
+fi
+
+
+# From the warnings, generate the doc and indexed bug files
+if ${update_doc_p}
+then
+ cd ${wwwdir}
+ rm -f ari.doc ari.idx ari.doc.bug
+ # Generate an extra file containing all the bugs that the ARI can
detect.
+ /bin/sh ${aridir}/gdb_ari.sh -Werror -Wall --print-idx --print-doc >>
ari.doc.bug
+ cat ari.*.bug | $AWK > ari.idx '
+BEGIN {
+ FS=": *"
+}
+{
+ # ari.*.bug: <FILE>:<LINE>: <CATEGORY>: <BUG>: <DOC>
+ file = $1
+ line = $2
+ category = $3
+ bug = $4
+ if (! (bug in cat)) {
+ cat[bug] = category
+ # strip any trailing .... (supplement)
+ doc[bug] = gensub(/ \([^\)]*\)$/, "", 1, $5)
+ count[bug] = 0
+ }
+ if (file != "") {
+ count[bug] += 1
+ # ari.*.idx: <BUG>:<FILE>:<CATEGORY>
+ print bug ":" file ":" category
+ }
+ # Also accumulate some categories as obsolete
+ if (category == "deprecated") {
+ # ari.*.idx: <BUG>:<FILE>:<CATEGORY>
+ if (file != "") {
+ print category ":" file ":" "obsolete"
+ }
+ #count[category]++
+ #doc[category] = "Contains " category " code"
+ }
+}
+END {
+ i = 0;
+ for (bug in count) {
+ # ari.*.doc: <BUG>:<COUNT>:<CATEGORY>:<DOC>
+ print bug ":" count[bug] ":" cat[bug] ":" doc[bug] >> "ari.doc"
+ }
+}
+'
+fi
+
+
+# print_toc BIAS MIN_COUNT CATEGORIES TITLE
+
+# Print a table of contents containing the bugs CATEGORIES. If the
+# BUG count >= MIN_COUNT print it in the table-of-contents. If
+# MIN_COUNT is non -ve, also include a link to the table.Adjust the
+# printed BUG count by BIAS.
+
+all=
+
+print_toc ()
+{
+ bias="$1" ; shift
+ min_count="$1" ; shift
+
+ all=" $all $1 "
+ categories=""
+ for c in $1; do
+ categories="${categories} categories[\"${c}\"] = 1 ;"
+ done
+ shift
+
+ title="$@" ; shift
+
+ echo "<p>" >> ${newari}
+ echo "<a name=${title}>" | tr '[A-Z]' '[a-z]' >> ${newari}
+ echo "<h3>${title}</h3>" >> ${newari}
+ cat >> ${newari} # description
+
+ cat >> ${newari} <<EOF
+<p>
+<table>
+<tr><th align=left>BUG</th><th>Total</th><th
align=left>Description</th></tr>
+EOF
+ # ari.*.doc: <BUG>:<COUNT>:<CATEGORY>:<DOC>
+ cat ${wwwdir}/ari.doc \
+ | sort -t: +1rn -2 +0d \
+ | $AWK >> ${newari} '
+BEGIN {
+ FS=":"
+ '"$categories"'
+ MIN_COUNT = '${min_count}'
+ BIAS = '${bias}'
+ total = 0
+ nr = 0
+}
+{
+ # ari.*.doc: <BUG>:<COUNT>:<CATEGORY>:<DOC>
+ bug = $1
+ count = $2
+ category = $3
+ doc = $4
+ if (count < MIN_COUNT) next
+ if (!(category in categories)) next
+ nr += 1
+ total += count
+ printf "<tr>"
+ printf "<th align=left valign=top><a name=\"%s\">", bug
+ printf "%s", gensub(/_/, " ", "g", bug)
+ printf "</a></th>"
+ printf "<td align=right valign=top>"
+ if (count > 0 && MIN_COUNT >= 0) {
+ printf "<a href=\"#,%s\">%d</a></td>", bug, count + BIAS
+ } else {
+ printf "%d", count + BIAS
+ }
+ printf "</td>"
+ printf "<td align=left valign=top>%s</td>", doc
+ printf "</tr>"
+ print ""
+}
+END {
+ print "<tr><th align=right valign=top>" nr "</th><th align=right
valign=top>" total "</th><td></td></tr>"
+}
+'
+cat >> ${newari} <<EOF
+</table>
+<p>
+EOF
+}
+
+
+print_table ()
+{
+ categories=""
+ for c in $1; do
+ categories="${categories} categories[\"${c}\"] = 1 ;"
+ done
+ # Remember to prune the dir prefix from projects files
+ # ari.*.idx: <BUG>:<FILE>:<CATEGORY>
+ cat ${wwwdir}/ari.idx | $AWK >> ${newari} '
+function qsort (table,
+ middle, tmp, left, nr_left, right, nr_right, result) {
+ middle = ""
+ for (middle in table) { break; }
+ nr_left = 0;
+ nr_right = 0;
+ for (tmp in table) {
+ if (tolower(tmp) < tolower(middle)) {
+ nr_left++
+ left[tmp] = tmp
+ } else if (tolower(tmp) > tolower(middle)) {
+ nr_right++
+ right[tmp] = tmp
+ }
+ }
+ #print "qsort " nr_left " " middle " " nr_right > "/dev/stderr"
+ result = ""
+ if (nr_left > 0) {
+ result = qsort(left) SUBSEP
+ }
+ result = result middle
+ if (nr_right > 0) {
+ result = result SUBSEP qsort(right)
+ }
+ return result
+}
+function print_heading (where, bug_i) {
+ print ""
+ print "<tr border=1>"
+ print "<th align=left>File</th>"
+ print "<th align=left><em>Total</em></th>"
+ print "<th></th>"
+ for (bug_i = 1; bug_i <= nr_bug; bug_i++) {
+ bug = i2bug[bug_i];
+ printf "<th>"
+ # The title names are offset by one. Otherwize, when the browser
+ # jumps to the name it leaves out half the relevant column.
+ #printf "<a name=\",%s\"> </a>", bug
+ printf "<a name=\",%s\"> </a>", i2bug[bug_i-1]
+ printf "<a href=\"#%s\">", bug
+ printf "%s", gensub(/_/, " ", "g", bug)
+ printf "</a>\n"
+ printf "</th>\n"
+ }
+ #print "<th></th>"
+ printf "<th><a name=\"%s,\"> </a></th>\n", i2bug[bug_i-1]
+ print "<th align=left><em>Total</em></th>"
+ print "<th align=left>File</th>"
+ print "</tr>"
+}
+function print_totals (where, bug_i) {
+ print "<th align=left><em>Totals</em></th>"
+ printf "<th align=right>"
+ printf "<em>%s</em>", total
+ printf ">"
+ printf "</th>\n"
+ print "<th></th>";
+ for (bug_i = 1; bug_i <= nr_bug; bug_i++) {
+ bug = i2bug[bug_i];
+ printf "<th align=right>"
+ printf "<em>"
+ printf "<a href=\"#%s\">%d</a>", bug, bug_total[bug]
+ printf "</em>";
+ printf "<a href=\"#%s,%s\">^</a>", prev_file[bug, where], bug
+ printf "<a href=\"#%s,%s\">v</a>", next_file[bug, where], bug
+ printf "<a name=\"%s,%s\"> </a>", where, bug
+ printf "</th>";
+ print ""
+ }
+ print "<th></th>"
+ printf "<th align=right>"
+ printf "<em>%s</em>", total
+ printf "<"
+ printf "</th>\n"
+ print "<th align=left><em>Totals</em></th>"
+ print "</tr>"
+}
+BEGIN {
+ FS = ":"
+ '"${categories}"'
+ nr_file = 0;
+ nr_bug = 0;
+}
+{
+ # ari.*.idx: <BUG>:<FILE>:<CATEGORY>
+ bug = $1
+ file = $2
+ category = $3
+ # Interested in this
+ if (!(category in categories)) next
+ # Totals
+ db[bug, file] += 1
+ bug_total[bug] += 1
+ file_total[file] += 1
+ total += 1
+}
+END {
+
+ # Sort the files and bugs creating indexed lists.
+ nr_bug = split(qsort(bug_total), i2bug, SUBSEP);
+ nr_file = split(qsort(file_total), i2file, SUBSEP);
+
+ # Dummy entries for first/last
+ i2file[0] = 0
+ i2file[-1] = -1
+ i2bug[0] = 0
+ i2bug[-1] = -1
+
+ # Construct a cycle of next/prev links. The file/bug "0" and "-1"
+ # are used to identify the start/end of the cycle. Consequently,
+ # prev(0) = -1 (prev of start is the end) and next(-1) = 0 (next
+ # of end is the start).
+
+ # For all the bugs, create a cycle that goes to the prev / next file.
+ for (bug_i = 1; bug_i <= nr_bug; bug_i++) {
+ bug = i2bug[bug_i]
+ prev = 0
+ prev_file[bug, 0] = -1
+ next_file[bug, -1] = 0
+ for (file_i = 1; file_i <= nr_file; file_i++) {
+ file = i2file[file_i]
+ if ((bug, file) in db) {
+ prev_file[bug, file] = prev
+ next_file[bug, prev] = file
+ prev = file
+ }
+ }
+ prev_file[bug, -1] = prev
+ next_file[bug, prev] = -1
+ }
+
+ # For all the files, create a cycle that goes to the prev / next bug.
+ for (file_i = 1; file_i <= nr_file; file_i++) {
+ file = i2file[file_i]
+ prev = 0
+ prev_bug[file, 0] = -1
+ next_bug[file, -1] = 0
+ for (bug_i = 1; bug_i <= nr_bug; bug_i++) {
+ bug = i2bug[bug_i]
+ if ((bug, file) in db) {
+ prev_bug[file, bug] = prev
+ next_bug[file, prev] = bug
+ prev = bug
+ }
+ }
+ prev_bug[file, -1] = prev
+ next_bug[file, prev] = -1
+ }
+
+ print "<table border=1 cellspacing=0>"
+ print "<tr></tr>"
+ print_heading(0);
+ print "<tr></tr>"
+ print_totals(0);
+ print "<tr></tr>"
+
+ for (file_i = 1; file_i <= nr_file; file_i++) {
+ file = i2file[file_i];
+ pfile = gensub(/^'${project}'\//, "", 1, file)
+ print ""
+ print "<tr>"
+ print "<th align=left><a name=\"" file ",\">" pfile "</a></th>"
+ printf "<th align=right>"
+ printf "%s", file_total[file]
+ printf "<a href=\"#%s,%s\">></a>", file, next_bug[file, 0]
+ printf "</th>\n"
+ print "<th></th>"
+ for (bug_i = 1; bug_i <= nr_bug; bug_i++) {
+ bug = i2bug[bug_i];
+ if ((bug, file) in db) {
+ printf "<td align=right>"
+ printf "<a href=\"#%s\">%d</a>", bug, db[bug, file]
+ printf "<a href=\"#%s,%s\">^</a>", prev_file[bug, file], bug
+ printf "<a href=\"#%s,%s\">v</a>", next_file[bug, file], bug
+ printf "<a name=\"%s,%s\"> </a>", file, bug
+ printf "</td>"
+ print ""
+ } else {
+ print "<td> </td>"
+ #print "<td></td>"
+ }
+ }
+ print "<th></th>"
+ printf "<th align=right>"
+ printf "%s", file_total[file]
+ printf "<a href=\"#%s,%s\"><</a>", file, prev_bug[file, -1]
+ printf "</th>\n"
+ print "<th align=left>" pfile "</th>"
+ print "</tr>"
+ }
+
+ print "<tr></tr>"
+ print_totals(-1)
+ print "<tr></tr>"
+ print_heading(-1);
+ print "<tr></tr>"
+ print ""
+ print "</table>"
+ print ""
+}
+'
+}
+
+
+# Make the scripts available
+cp ${aridir}/gdb_*.sh ${wwwdir}
+
+# Compute the ARI index - ratio of zero vs non-zero problems.
+indexes=`awk '
+BEGIN {
+ FS=":"
+}
+{
+ # ari.*.doc: <BUG>:<COUNT>:<CATEGORY>:<DOC>
+ bug = $1; count = $2; category = $3; doc = $4
+
+ if (bug ~ /^legacy_/) legacy++
+ if (bug ~ /^deprecated_/) deprecated++
+
+ if (category !~ /^gdbarch$/) {
+ bugs += count
+ }
+ if (count == 0) {
+ oks++
+ }
+}
+END {
+ #print "tests/ok:", nr / ok
+ #print "bugs/tests:", bugs / nr
+ #print "bugs/ok:", bugs / ok
+ print bugs / ( oks + legacy + deprecated )
+}
+' ${wwwdir}/ari.doc`
+
+# Merge, generating the ARI tables.
+if ${update_web_p}
+then
+ echo "Create the ARI table" 1>&2
+ oldari=${wwwdir}/old.html
+ ari=${wwwdir}/index.html
+ newari=${wwwdir}/new.html
+ rm -f ${newari} ${newari}.gz
+ cat <<EOF >> ${newari}
+<html>
+<head>
+<title>A.R. Index for GDB version ${version}</title>
+</head>
+<body>
+
+<center><h2>A.R. Index for GDB version ${version}<h2></center>
+
+<!-- body, update above using ../index.sh -->
+
+<!-- Navigation. This page contains the following anchors.
+"BUG": The definition of the bug.
+"FILE,BUG": The row/column containing FILEs BUG count
+"0,BUG", "-1,BUG": The top/bottom total for BUGs column.
+"FILE,O", "FILE,-1": The left/right total for FILEs row.
+",BUG": The top title for BUGs column.
+"FILE,": The left title for FILEs row.
+-->
+
+<center><h3>${indexes}</h3></center>
+<center><h3>You can not take this seriously!</h3></center>
+
+<center>
+Also available:
+<a href="../gdb/ari/">most recent branch</a>
+|
+<a href="../gdb/current/ari/">current</a>
+|
+<a href="../gdb/download/ari/">last release</a>
+</center>
+
+<center>
+Last updated: `date -u`
+</center>
+EOF
+
+ print_toc 0 1 "internal regression" Critical <<EOF
+Things previously eliminated but returned. This should always be empty.
+EOF
+
+ print_table "regression code comment obsolete gettext"
+
+ print_toc 0 0 code Code <<EOF
+Coding standard problems, portability problems, readability problems.
+EOF
+
+ print_toc 0 0 comment Comments <<EOF
+Problems concerning comments in source files.
+EOF
+
+ print_toc 0 0 gettext GetText <<EOF
+Gettext related problems.
+EOF
+
+ print_toc 0 -1 dos DOS 8.3 File Names <<EOF
+File names with problems on 8.3 file systems.
+EOF
+
+ print_toc -2 -1 deprecated Deprecated <<EOF
+Mechanisms that have been replaced with something better, simpler,
+cleaner; or are no longer required by core-GDB. New code should not
+use deprecated mechanisms. Existing code, when touched, should be
+updated to use non-deprecated mechanisms. See obsolete and deprecate.
+(The declaration and definition are hopefully excluded from count so
+zero should indicate no remaining uses).
+EOF
+
+ print_toc 0 0 obsolete Obsolete <<EOF
+Mechanisms that have been replaced, but have not yet been marked as
+such (using the deprecated_ prefix). See deprecate and deprecated.
+EOF
+
+ print_toc 0 -1 deprecate Deprecate <<EOF
+Mechanisms that are a candidate for being made obsolete. Once core
+GDB no longer depends on these mechanisms and/or there is a
+replacement available, these mechanims can be deprecated (adding the
+deprecated prefix) obsoleted (put into category obsolete) or deleted.
+See obsolete and deprecated.
+EOF
+
+ print_toc -2 -1 legacy Legacy <<EOF
+Methods used to prop up targets using targets that still depend on
+deprecated mechanisms. (The method's declaration and definition are
+hopefully excluded from count).
+EOF
+
+ print_toc -2 -1 gdbarch Gdbarch <<EOF
+Count of calls to the gdbarch set methods. (Declaration and
+definition hopefully excluded from count).
+EOF
+
+ print_toc 0 -1 macro Macro <<EOF
+Breakdown of macro definitions (and #undef) in configuration files.
+EOF
+
+ print_toc 0 0 regression Fixed <<EOF
+Problems that have been expunged from the source code.
+EOF
+
+ # Check for invalid categories
+ for a in $all; do
+ alls="$alls all[$a] = 1 ;"
+ done
+ cat ari.*.doc | $AWK >> ${newari} '
+BEGIN {
+ FS = ":"
+ '"$alls"'
+}
+{
+ # ari.*.doc: <BUG>:<COUNT>:<CATEGORY>:<DOC>
+ bug = $1
+ count = $2
+ category = $3
+ doc = $4
+ if (!(category in all)) {
+ print "<b>" category "</b>: no documentation<br>"
+ }
+}
+'
+
+ cat >> ${newari} <<EOF
+<center>
+Input files:
+`( cd ${wwwdir} && ls ari.*.bug ari.idx ari.doc ) | while read f
+do
+ echo "<a href=\"${f}\">${f}</a>"
+done`
+</center>
+
+<center>
+Scripts:
+`( cd ${wwwdir} && ls *.sh ) | while read f
+do
+ echo "<a href=\"${f}\">${f}</a>"
+done`
+</center>
+
+<!-- /body, update below using ../index.sh -->
+</body>
+</html>
+EOF
+
+ for i in . .. ../..; do
+ x=${wwwdir}/${i}/index.sh
+ if test -x $x; then
+ $x ${newari}
+ break
+ fi
+ done
+
+ gzip -c -v -9 ${newari} > ${newari}.gz
+
+ cp ${ari} ${oldari}
+ cp ${ari}.gz ${oldari}.gz
+ cp ${newari} ${ari}
+ cp ${newari}.gz ${ari}.gz
+
+fi # update_web_p
+
+# ls -l ${wwwdir}
+
+exit 0