build: added libexpat build support

This allows commands such as "info os files"
previously we had expat support on x86_64 only.

.gitmodules

@@ -6,3 +6,9 @@
 	path = src/submodule_packages/binutils-gdb
 	url = git@github.com:guyush1/binutils-gdb.git
 	branch = gdb-static
+[submodule "src/submodule_packages/pygments"]
+	path = src/submodule_packages/pygments
+	url = git@github.com:pygments/pygments.git
+[submodule "src/submodule_packages/libexpat"]
+	path = src/submodule_packages/libexpat
+	url = git@github.com:guyush1/libexpat.git

Dockerfile

@@ -19,6 +19,7 @@ RUN apt update && apt install -y \
     gcc-powerpc-linux-gnu \
     git \
     libncurses-dev \
+    libtool \
     m4  \
     make \
     patch \

src/compilation/build.sh

@@ -57,15 +57,18 @@ function set_compliation_variables() {
     export LDFLAGS="-s"
 }
 
-function set_ncurses_link_variables() {
-    # Set up ncurses library link variables
+function set_up_lib_search_paths() {
+    # Set up library-related linker search paths.
     #
     # Parameters:
     # $1: ncursesw build dir
+    # $2: libexpat build dir
     local ncursesw_build_dir="$1"
+    local libexpat_build_dir="$2"
 
-    # Allow tui mode by adding our custom built static ncursesw library to the linker search path.
-    export LDFLAGS="-L$ncursesw_build_dir/lib $LDFLAGS"
+    # I) Allow tui mode by adding our custom built static ncursesw library to the linker search path.
+    # II) Allow parsing xml files by adding libexpat library to the linker search path.
+    export LDFLAGS="-L$ncursesw_build_dir/lib -L$libexpat_build_dir/lib/.libs $LDFLAGS"
 }
 
 function build_iconv() {
@@ -214,12 +217,64 @@ function build_ncurses() {
     popd > /dev/null
 }
 
+function build_libexpat() {
+    # Build libexpat.
+    #
+    # Parameters:
+    # $1: libexpat package directory
+    # $2: target architecture
+    #
+    # Echoes:
+    # The libexpat build directory
+    #
+    # Returns:
+    # 0: success
+    # 1: failure
+    local libexpat_dir="$1"
+    local target_arch="$2"
+    local libexpat_build_dir="$(realpath "$libexpat_dir/build-$target_arch")"
+
+    echo "$libexpat_build_dir"
+    mkdir -p "$libexpat_build_dir"
+
+    if [[ -f "$libexpat_build_dir/lib/.libs/libexpat.a" ]]; then
+        >&2 echo "Skipping build: libexpat already built for $target_arch"
+        return 0
+    fi
+
+    pushd "$libexpat_build_dir" > /dev/null
+
+    >&2 fancy_title "Building libexpat for $target_arch"
+
+    # Generate configure if it doesnt exist.
+    if [[ ! -f "$libexpat_build_dir/../expat/configure" ]]; then
+        >&2 ../expat/buildconf.sh ../expat/
+    fi
+
+    ../expat/configure --enable-static "CC=$CC" "CXX=$CXX" "--host=$HOST" \
+        "CFLAGS=$CFLAGS" "CXXFLAGS=$CXXFLAGS" 1>&2
+    if [[ $? -ne 0 ]]; then
+        return 1
+    fi
+
+    make -j$(nproc) 1>&2
+    if [[ $? -ne 0 ]]; then
+        return 1
+    fi
+
+    >&2 fancy_title "Finished building libexpat for $target_arch"
+
+    popd > /dev/null
+}
+
 function build_python() {
     # Build python.
     #
     # Parameters:
     # $1: python package directory
     # $2: target architecture
+    # $3: gdb's python module directory parent
+    # $4: pygment's toplevel source dir.
     #
     # Echoes:
     # The python build directory
@@ -229,6 +284,8 @@ function build_python() {
     # 1: failure
     local python_dir="$1"
     local target_arch="$2"
+    local gdb_python_parent="$3"
+    local pygments_source_dir="$4"
     local python_lib_dir="$(realpath "$python_dir/build-$target_arch")"
 
     echo "$python_lib_dir"
@@ -257,6 +314,17 @@ function build_python() {
         --disable-ipv6 \
         --disable-shared
 
+    # Extract the regular standard library modules that are to be frozen and include the gdb and pygments custom libraries.
+    export EXTRA_FROZEN_MODULES="$(printf "%s" "$(< ${script_dir}/frozen_python_modules.txt)" | tr $'\n' ";")"
+    export EXTRA_FROZEN_MODULES="${EXTRA_FROZEN_MODULES};<gdb.**.*>: gdb = ${gdb_python_parent};<pygments.**.*>: pygments = ${pygments_source_dir}"
+    >&2 echo "Frozen Modules: ${EXTRA_FROZEN_MODULES}"
+
+    # Regenerate frozen modules with gdb env varaible. Do it after the configure because we need
+    # the `regen-frozen` makefile.
+    >&2 python3.12 ../Tools/build/freeze_modules.py
+    >&2 make regen-frozen
+
+    # Build python after configuring the project and regnerating frozen files.
     >&2 make -j $(nproc)
     if [[ $? -ne 0 ]]; then
         return 1
@@ -373,10 +441,11 @@ function build_gdb() {
 
     ../configure -C --enable-static --with-static-standard-libraries --disable-inprocess-agent \
                  --enable-tui "$python_flag" \
+                 --with-expat --with-libexpat-type="static" \
                  "--with-libiconv-prefix=$libiconv_prefix" --with-libiconv-type=static \
                  "--with-gmp=$libgmp_prefix" \
                  "--with-mpfr=$libmpfr_prefix" \
-                 "CC=$CC" "CXX=$CXX" "--host=$HOST" \
+                 "CC=$CC" "CXX=$CXX" "LDFLAGS=$LDFLAGS" "--host=$HOST" \
                  "CFLAGS=$CFLAGS" "CXXFLAGS=$CXXFLAGS" 1>&2
     if [[ $? -ne 0 ]]; then
         return 1
@@ -515,10 +584,18 @@ function build_gdb_with_dependencies() {
     if [[ $? -ne 0 ]]; then
         return 1
     fi
-    set_ncurses_link_variables "$ncursesw_build_dir"
+
+    libexpat_build_dir="$(build_libexpat "$packages_dir/libexpat" "$target_arch")"
+    if [[ $? -ne 0 ]]; then
+        return 1
+    fi
+
+    set_up_lib_search_paths "$ncursesw_build_dir" "$libexpat_build_dir"
 
     if [[ "$with_python" == "yes" ]]; then
-        build_python "$packages_dir/cpython-static" "$target_arch"
+        local gdb_python_dir="$packages_dir/binutils-gdb/gdb/python/lib/"
+        local pygments_source_dir="$packages_dir/pygments/"
+        local python_build_dir="$(build_python "$packages_dir/cpython-static" "$target_arch" "$gdb_python_dir" "$pygments_source_dir")"
         if [[ $? -ne 0 ]]; then
             return 1
         fi

src/compilation/frozen_python_modules.txt

@@ -0,0 +1,150 @@
+abc
+_aix_support
+antigravity
+argparse
+ast
+base64
+bdb
+bisect
+calendar
+cmd
+codecs
+codeop
+code
+<collections.**.*>
+_collections_abc
+colorsys
+_compat_pickle
+compileall
+_compression
+<concurrent.**.*>
+configparser
+contextlib
+contextvars
+copy
+copyreg
+cProfile
+csv
+dataclasses
+datetime
+<dbm.**.*>
+decimal
+difflib
+dis
+<encodings.**.*>
+<ensurepip.**.*>
+enum
+filecmp
+fileinput
+fnmatch
+fractions
+ftplib
+functools
+__future__
+genericpath
+getopt
+getpass
+gettext
+glob
+graphlib
+gzip
+hashlib
+heapq
+hmac
+imaplib
+<importlib.**.*>
+inspect
+io
+ipaddress
+<json.**.*>
+keyword
+linecache
+locale
+<logging.**.*>
+lzma
+_markupbase
+mimetypes
+modulefinder
+<multiprocessing.**.*>
+netrc
+ntpath
+nturl2path
+numbers
+opcode
+operator
+optparse
+os
+_osx_support
+pathlib
+pdb
+<__phello__.**.*>
+pickle
+pickletools
+pkgutil
+platform
+plistlib
+poplib
+posixpath
+pprint
+profile
+pstats
+pty
+_py_abc
+pyclbr
+py_compile
+_pydatetime
+_pydecimal
+_pyio
+_pylong
+queue
+quopri
+random
+<re.**.*>
+reprlib
+rlcompleter
+sched
+selectors
+shelve
+shlex
+shutil
+signal
+smtplib
+socket
+socketserver
+statistics
+stat
+stringprep
+string
+_strptime
+struct
+subprocess
+symtable
+sysconfig
+tabnanny
+tempfile
+textwrap
+this
+_threading_local
+threading
+timeit
+tokenize
+token
+<tomllib.**.*>
+traceback
+tracemalloc
+trace
+tty
+types
+typing
+uuid
+warnings
+wave
+weakref
+_weakrefset
+webbrowser
+<wsgiref.**.*>
+zipapp
+<zipfile.**.*>
+<zoneinfo.**.*>
+<email.**.*>
+<urllib.**.*>