linux-kernel - [PATCH 0/3] Fix CONFIG_TEST

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: <20211129230141.228085-1-nathan@kernel.org>
Date:   Mon, 29 Nov 2021 16:01:38 -0700
From:   Nathan Chancellor <nathan@...nel.org>
To:     Andrew Morton <akpm@...ux-foundation.org>,
        Chris Mason <clm@...com>, Josef Bacik <josef@...icpanda.com>,
        David Sterba <dsterba@...e.com>,
        Luis Chamberlain <mcgrof@...nel.org>
Cc:     Nick Desaulniers <ndesaulniers@...gle.com>,
        linux-btrfs@...r.kernel.org, linux-kernel@...r.kernel.org,
        Nathan Chancellor <nathan@...nel.org>
Subject: [PATCH 0/3] Fix CONFIG_TEST_KMOD with 256kB page size

The kernel test robot reported a build error [1] from a failed assertion
in fs/btrfs/inode.c with a hexagon randconfig that includes
CONFIG_PAGE_SIZE_256KB. This error is the same one that was addressed by
commit b05fbcc36be1 ("btrfs: disable build on platforms having page size
256K") but CONFIG_TEST_KMOD selects CONFIG_BTRFS without having the
"page size less than 256kB dependency", which results in the error
reappearing.

The first patch introduces CONFIG_PAGE_SIZE_LESS_THAN_256KB by splitting
it off from CONFIG_PAGE_SIZE_LESS_THAN_64KB, which was introduced in
commit 1f0e290cc5fd ("arch: Add generic Kconfig option indicating page
size smaller than 64k") for a similar reason in 5.16-rc3.

The second patch uses that configuration option for CONFIG_BTRFS to
reduce duplication.

The third patch resolves the build error by adding
CONFIG_PAGE_SIZE_LESS_THAN_256KB as a dependency to CONFIG_TEST_KMOD so
that CONFIG_BTRFS does not get enabled under that invalid configuration.

This series could go through either btrfs or -mm, as arch/Kconfig and
lib/Kconfig.debug have no formal maintainer. I have sent it to the
maintainers of both so that they can decide who will take it.

[1]: https://lore.kernel.org/r/202111270255.UYOoN5VN-lkp@intel.com/

Nathan Chancellor (3):
  arch/Kconfig: Split PAGE_SIZE_LESS_THAN_256KB from
    PAGE_SIZE_LESS_THAN_64KB
  btrfs: Use generic Kconfig option for 256kB page size limit
  lib/Kconfig.debug: Make TEST_KMOD depend on PAGE_SIZE_LESS_THAN_256KB

 arch/Kconfig      | 4 ++++
 fs/btrfs/Kconfig  | 3 +--
 lib/Kconfig.debug | 1 +
 3 files changed, 6 insertions(+), 2 deletions(-)

base-commit: d58071a8a76d779eedab38033ae4c821c30295a5
-- 
2.34.1