From: Josh Law <objecting@objecting.org>
To: Masami Hiramatsu <mhiramat@kernel.org>,
Andrew Morton <akpm@linux-foundation.org>
Cc: linux-kernel@vger.kernel.org, linux-trace-kernel@vger.kernel.org,
Josh Law <objecting@objecting.org>
Subject: [PATCH v8 00/13] bootconfig: cleanups, correctness, and modernization
Date: Wed, 18 Mar 2026 15:59:06 +0000 [thread overview]
Message-ID: <20260318155919.78168-1-objecting@objecting.org> (raw)
This series addresses a collection of issues found during a review of
lib/bootconfig.c, include/linux/bootconfig.h, and tools/bootconfig,
covering off-by-one errors, coding style, signedness/type cleanup, and
API modernization.
The two patches with Fixes tags (xbc_init_node() override check and
load_xbc_file() fd leak) have been split into a separate series for
bootconfig/fixes per Masami's request.
Changes since v7:
- Split fixes from cleanups/improvements into separate series per
maintainer request, so fixes can go into bootconfig/fixes and the
rest (this series) into bootconfig/for-next.
Bug fixes:
- Fix off-by-one in xbc_verify_tree() where a next-node index equal
to xbc_node_num passes the bounds check despite being out of range;
a malformed bootconfig could cause an out-of-bounds read of kernel
memory during tree traversal at boot time (patch 3).
- Move xbc_node_num increment to after xbc_init_node() validation so
a failed init does not leave a partially initialized node counted
in the array; on a maximum-size bootconfig, the uninitialized node
could be traversed leading to unpredictable boot behavior (patch 4).
- Validate child node indices in xbc_verify_tree() alongside the
existing next-node check; without this, a corrupt bootconfig could
trigger an out-of-bounds memory access via an invalid child index
during tree traversal (patch 8).
Correctness:
- Narrow the flag parameter in node creation helpers from uint32_t to
uint16_t to match the xbc_node.data field width (patch 2).
- Constify the xbc_calc_checksum() data parameter since it only reads
the buffer (patch 6).
- Fix strict-GCC signedness and narrowing warnings by aligning local
types with strlen() APIs and the node index/data storage in
xbc_node_get_data(), xbc_node_match_prefix(), xbc_init_node(),
xbc_verify_tree(), and xbc_node_index() (patches 9-13).
Cleanups:
- Fix comment typos, missing blank line before kerneldoc, and
inconsistent if/else bracing (patch 1).
- Drop redundant memset after memblock_alloc which already returns
zeroed memory; switch the userspace path from malloc to calloc to
match (patch 5).
Modernization:
- Replace the catch-all linux/kernel.h include with the specific
headers needed: linux/cache.h, linux/compiler.h, and
linux/sprintf.h (patch 7).
Build-tested with both the in-kernel build (lib/bootconfig.o,
init/main.o) and the userspace tools/bootconfig build. All 70
tools/bootconfig test cases pass.
Josh Law (13):
lib/bootconfig: clean up comment typos and bracing
lib/bootconfig: narrow flag parameter type from uint32_t to uint16_t
lib/bootconfig: fix off-by-one in xbc_verify_tree() next node check
lib/bootconfig: increment xbc_node_num after node init succeeds
lib/bootconfig: drop redundant memset of xbc_nodes
bootconfig: constify xbc_calc_checksum() data parameter
lib/bootconfig: replace linux/kernel.h with specific includes
lib/bootconfig: validate child node index in xbc_verify_tree()
lib/bootconfig: fix signed comparison in xbc_node_get_data()
lib/bootconfig: use size_t for strlen result in
xbc_node_match_prefix()
lib/bootconfig: use signed type for offset in xbc_init_node()
lib/bootconfig: use size_t for key length tracking in
xbc_verify_tree()
lib/bootconfig: change xbc_node_index() return type to uint16_t
include/linux/bootconfig.h | 6 ++--
lib/bootconfig.c | 62 ++++++++++++++++++++++----------------
2 files changed, 39 insertions(+), 29 deletions(-)
--
2.34.1
next reply other threads:[~2026-03-18 15:59 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-18 15:59 Josh Law [this message]
2026-03-18 15:59 ` [PATCH v8 01/13] lib/bootconfig: clean up comment typos and bracing Josh Law
2026-03-18 15:59 ` [PATCH v8 02/13] lib/bootconfig: narrow flag parameter type from uint32_t to uint16_t Josh Law
2026-03-18 15:59 ` [PATCH v8 03/13] lib/bootconfig: fix off-by-one in xbc_verify_tree() next node check Josh Law
2026-03-18 15:59 ` [PATCH v8 04/13] lib/bootconfig: increment xbc_node_num after node init succeeds Josh Law
2026-03-18 15:59 ` [PATCH v8 05/13] lib/bootconfig: drop redundant memset of xbc_nodes Josh Law
2026-03-18 20:22 ` Markus Elfring
2026-03-18 15:59 ` [PATCH v8 06/13] bootconfig: constify xbc_calc_checksum() data parameter Josh Law
2026-03-18 15:59 ` [PATCH v8 07/13] lib/bootconfig: replace linux/kernel.h with specific includes Josh Law
2026-03-18 15:59 ` [PATCH v8 08/13] lib/bootconfig: validate child node index in xbc_verify_tree() Josh Law
2026-03-18 15:59 ` [PATCH v8 09/13] lib/bootconfig: fix signed comparison in xbc_node_get_data() Josh Law
2026-03-18 15:59 ` [PATCH v8 10/13] lib/bootconfig: use size_t for strlen result in xbc_node_match_prefix() Josh Law
2026-03-18 15:59 ` [PATCH v8 11/13] lib/bootconfig: use signed type for offset in xbc_init_node() Josh Law
2026-03-18 15:59 ` [PATCH v8 12/13] lib/bootconfig: use size_t for key length tracking in xbc_verify_tree() Josh Law
2026-03-18 15:59 ` [PATCH v8 13/13] lib/bootconfig: change xbc_node_index() return type to uint16_t Josh Law
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=20260318155919.78168-1-objecting@objecting.org \
--to=objecting@objecting.org \
--cc=akpm@linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-trace-kernel@vger.kernel.org \
--cc=mhiramat@kernel.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