linux-kernel - [PATCH RFC 0/8] hardened usercopy and stacktrace improvement

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]

Message-Id: <20220418132217.1573072-1-zhe.he@windriver.com>
Date:   Mon, 18 Apr 2022 21:22:09 +0800
From:   He Zhe <zhe.he@...driver.com>
To:     catalin.marinas@....com, will@...nel.org, mark.rutland@....com,
        tglx@...utronix.de, bp@...en8.de, dave.hansen@...ux.intel.com,
        keescook@...omium.org, alexander.shishkin@...ux.intel.com,
        jolsa@...nel.org, namhyung@...nel.org, benh@...nel.crashing.org,
        paulus@...ba.org, borntraeger@...ux.ibm.com, svens@...ux.ibm.com,
        hpa@...or.com
Cc:     x86@...nel.org, linux-arm-kernel@...ts.infradead.org,
        linuxppc-dev@...ts.ozlabs.org, linux-riscv@...ts.infradead.org,
        linux-s390@...r.kernel.org, linux-perf-users@...r.kernel.org,
        linux-kernel@...r.kernel.org, zhe.he@...driver.com
Subject: [PATCH RFC 0/8] hardened usercopy and stacktrace improvement

This series is to add copy-beyond-the-frame detection for hardened usercopy
for arm64 by adding more arguments to callbacks of stackstrace and by the way
more potential checks in the future can be done. This is done by:

- Change stacktrace callback prototype to pass more information which currently
  includes pc, current stack's fp and previous frame's fp, so copying beyond the
  frame can be detected.
- Add arch_within_stack_frames for arm64 based on existing x86 algorithm, but
  via stacktrace.
- Make other arch_stack_walk callbacks use the new prototype.
- Tune lkdtm usercopy stack frame case. Currently the way to generate bad frame
  use depends on stack layout of the architecture and compiler. It happens to
  work with x86 but does not work for arm64 and it ruins the stack in
  USERCOPY_STACK_FRAME_FROM and thus causes the call chain to return to NULL
  address.

He Zhe (8):
  stacktrace: Change callback prototype to pass more information
  arm64: stacktrace: Add arch_within_stack_frames
  arm64: stacktrace: Make callbacks use new prototype with frame info
  powerpc: stacktrace: Make callbacks use new prototype with frame info
  riscv: stacktrace: Make callbacks use new prototype with frame info
  s390: stacktrace: Make callbacks use new prototype with frame info
  x86: stacktrace: Make callbacks use new prototype with frame info
  lkdtm: usercopy: Make USERCOPY_STACK_FRAME_x able to work for all
    archs

 arch/arm64/Kconfig                   |  1 +
 arch/arm64/include/asm/thread_info.h | 12 +++++
 arch/arm64/kernel/perf_callchain.c   |  4 +-
 arch/arm64/kernel/process.c          |  6 +--
 arch/arm64/kernel/return_address.c   |  4 +-
 arch/arm64/kernel/stacktrace.c       | 76 ++++++++++++++++++++++++++--
 arch/arm64/kernel/time.c             |  6 +--
 arch/powerpc/kernel/stacktrace.c     | 18 ++++---
 arch/riscv/include/asm/stacktrace.h  |  2 +-
 arch/riscv/kernel/perf_callchain.c   |  4 +-
 arch/riscv/kernel/stacktrace.c       | 22 ++++----
 arch/s390/kernel/stacktrace.c        | 14 ++---
 arch/x86/kernel/stacktrace.c         | 28 ++++++----
 drivers/misc/lkdtm/usercopy.c        | 24 +--------
 include/linux/stacktrace.h           |  9 +++-
 kernel/stacktrace.c                  | 10 ++--
 16 files changed, 161 insertions(+), 79 deletions(-)

-- 
2.25.1