linux-kernel - [PATCH 0/2] x86: Avoid CONFIG_X86_X32

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 14 Mar 2022 12:48:40 -0700
From:   Nathan Chancellor <nathan@...nel.org>
To:     Peter Zijlstra <peterz@...radead.org>, x86@...nel.org
Cc:     Andy Lutomirski <luto@...nel.org>,
        "H. Peter Anvin" <hpa@...or.com>,
        Nick Desaulniers <ndesaulniers@...gle.com>,
        Masahiro Yamada <masahiroy@...nel.org>,
        linux-kernel@...r.kernel.org, alsa-devel@...a-project.org,
        linux-fsdevel@...r.kernel.org, linux-xfs@...r.kernel.org,
        llvm@...ts.linux.dev, Nathan Chancellor <nathan@...nel.org>
Subject: [PATCH 0/2] x86: Avoid CONFIG_X86_X32_ABI=y with llvm-objcopy

This series disables CONFIG_X86_X32_ABI=y with llvm-objcopy, which has
had two outstanding issues, which are further outlined in the second
patch:

https://github.com/ClangBuiltLinux/linux/issues/514
https://github.com/ClangBuiltLinux/linux/issues/1141

The first patch is from Masahiro, which moves the CONFIG_X86_X32_ABI
back into Kconfig proper. It was initially pushed as an RFC:

https://lore.kernel.org/r/20210227183910.221873-1-masahiroy@kernel.org/

The conclusion of that thread was that the check was still needed
because of the two issues above. However, with the introduction of
IBT, specifically commit 41c5ef31ad71 ("x86/ibt: Base IBT bits"), the
second issue linked above becomes visible with allmodconfig, which
heavily impacts automated testing.

The second patch builds on the first by just universally disabling
CONFIG_X86_X32_ABI when using llvm-objcopy at configuration time so that
neither issue is visible.

While it is unlikely that these issues will be fixed in LLVM, this could
eventually become a version check.

This is based on -tip x86/core and I would like for both patches to go
with the IBT changes so that there is no build breakage.

Masahiro Yamada (1):
  x86: Remove toolchain check for X32 ABI capability

Nathan Chancellor (1):
  x86/Kconfig: Do not allow CONFIG_X86_X32_ABI=y with llvm-objcopy

 arch/x86/Kconfig                       | 13 +++++++------
 arch/x86/Makefile                      | 16 ----------------
 arch/x86/entry/syscalls/Makefile       |  2 +-
 arch/x86/include/asm/syscall_wrapper.h |  6 +++---
 arch/x86/include/asm/vdso.h            |  2 +-
 arch/x86/kernel/process_64.c           |  2 +-
 fs/fuse/ioctl.c                        |  2 +-
 fs/xfs/xfs_ioctl32.c                   |  2 +-
 sound/core/control_compat.c            | 16 ++++++++--------
 sound/core/pcm_compat.c                | 20 ++++++++++----------
 10 files changed, 33 insertions(+), 48 deletions(-)

base-commit: f8afc9d88e65d189653f363eacc1f3131216ef7c
-- 
2.35.1