From: Simon Marchi <simon.marchi@efficios.com>
To: gdb-patches@sourceware.org
Cc: Simon Marchi <simon.marchi@efficios.com>
Subject: [PATCH] gdb: fix MS_WIN64 redefinition error with Python >= 3.12 on MinGW
Date: Tue, 13 Oct 2026 10:08:56 -0400 [thread overview]
Message-ID: <20261013140907.1346231-1-simon.marchi@efficios.com> (raw)
When trying to cross-build GDB for x86_64-w64-mingw32 on Arch Linux
against Python 3.14 [1], I see:
CXX python/py-arch.o
In file included from /usr/x86_64-w64-mingw32/include/python314/Python.h:14,
from /home/smarchi/src/binutils-gdb/gdb/python/python-internal.h:59,
from /home/smarchi/src/binutils-gdb/gdb/python/py-arch.c:23:
/usr/x86_64-w64-mingw32/include/python314/pyconfig.h:253:9: error: ‘MS_WIN64’ redefined [-Werror]
253 | #define MS_WIN64
| ^~~~~~~~
<command-line>: note: this is the location of the previous definition
Our configure.ac defines MS_WIN64 because of this:
# On x64 Windows, Python's include headers, and pyconfig.h in
# particular, rely on MS_WIN64 macro to detect that it's a 64bit
# version of Windows. Unfortunately, MS_WIN64 is only defined if
# _MSC_VER, a Microsoft-specific macro, is defined. So, when
# building on x64 Windows with GCC, we define MS_WIN64 ourselves.
# The issue was reported to the Python community, but still isn't
# solved as of 2012-10-02 (http://bugs.python.org/issue4709).
Actually, since CPython commit 158b8a07212c ("gh-79218: Define
`MS_WIN64` macro for Mingw-w64 64bit on Windows (GH-100137)",
2022-12-12), which is in Python 3.12, pyconfig.h does define this macro
when building with MinGW. However, it defines it empty, whereas our
`-DMS_WIN64` defines it to 1, causing the macro redefinition error.
Change it to `-DMS_WIN64=`, which defines an empty macro, and avoids the
redefinition warning on Python >= 3.12. Older Python versions still get
the macro they need.
Regenerate configure.
[1] https://aur.archlinux.org/packages/mingw-w64-python314-bin
Change-Id: I67d9b6cb22d66812b4230d0cd5a51cfdc96dc379
---
gdb/configure | 15 +++++++++------
gdb/configure.ac | 15 +++++++++------
2 files changed, 18 insertions(+), 12 deletions(-)
diff --git a/gdb/configure b/gdb/configure
index 810e9ed7a378..90a8ea0b7ebd 100755
--- a/gdb/configure
+++ b/gdb/configure
@@ -28677,16 +28677,19 @@ $as_echo "${PYTHON_CFLAGS}" >&6; }
# On x64 Windows, Python's include headers, and pyconfig.h in
# particular, rely on MS_WIN64 macro to detect that it's a 64bit
- # version of Windows. Unfortunately, MS_WIN64 is only defined if
- # _MSC_VER, a Microsoft-specific macro, is defined. So, when
- # building on x64 Windows with GCC, we define MS_WIN64 ourselves.
- # The issue was reported to the Python community, but still isn't
- # solved as of 2012-10-02 (http://bugs.python.org/issue4709).
+ # version of Windows. Before Python 3.12, MS_WIN64 is only defined
+ # if _MSC_VER, a Microsoft-specific macro, is defined. So, when
+ # building on x64 Windows with GCC, we define MS_WIN64 ourselves
+ # (http://bugs.python.org/issue4709). Python 3.12 and later define
+ # it as empty for MinGW too, so define it as empty as well here too.
+ #
+ # This can be removed when support for Python 3.12 and below is
+ # dropped.
case "$gdb_host" in
mingw64)
if test "${GCC}" = yes; then
- CPPFLAGS="$CPPFLAGS -DMS_WIN64"
+ CPPFLAGS="$CPPFLAGS -DMS_WIN64="
fi
;;
esac
diff --git a/gdb/configure.ac b/gdb/configure.ac
index 0d88a280b163..3c526c9c108e 100644
--- a/gdb/configure.ac
+++ b/gdb/configure.ac
@@ -1080,16 +1080,19 @@ if test "${have_libpython}" != no; then
# On x64 Windows, Python's include headers, and pyconfig.h in
# particular, rely on MS_WIN64 macro to detect that it's a 64bit
- # version of Windows. Unfortunately, MS_WIN64 is only defined if
- # _MSC_VER, a Microsoft-specific macro, is defined. So, when
- # building on x64 Windows with GCC, we define MS_WIN64 ourselves.
- # The issue was reported to the Python community, but still isn't
- # solved as of 2012-10-02 (http://bugs.python.org/issue4709).
+ # version of Windows. Before Python 3.12, MS_WIN64 is only defined
+ # if _MSC_VER, a Microsoft-specific macro, is defined. So, when
+ # building on x64 Windows with GCC, we define MS_WIN64 ourselves
+ # (http://bugs.python.org/issue4709). Python 3.12 and later define
+ # it as empty for MinGW too, so define it as empty as well here too.
+ #
+ # This can be removed when support for Python 3.12 and below is
+ # dropped.
case "$gdb_host" in
mingw64)
if test "${GCC}" = yes; then
- CPPFLAGS="$CPPFLAGS -DMS_WIN64"
+ CPPFLAGS="$CPPFLAGS -DMS_WIN64="
fi
;;
esac
base-commit: 80b91cb32b083868945f8da8fdeb2c20b5c9f099
--
2.56.0
next reply other threads:[~2026-10-13 14:09 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-13 14:08 Simon Marchi [this message]
2026-10-14 11:57 ` Tom Tromey
2026-10-14 12:52 ` Simon Marchi
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=20261013140907.1346231-1-simon.marchi@efficios.com \
--to=simon.marchi@efficios.com \
--cc=gdb-patches@sourceware.org \
/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