Re: + lib-ts_kmp-fix-integer-overflow-in-pattern-length-calculation.patch added to mm-nonmm-unstable branch

[Josh Law <hlcj1234567@gmail.com>]

9 Mar 2026 20:24:36 Andrew Morton <akpm@linux-foundation.org>:

>
> The patch titled
>      Subject: lib/ts_kmp: fix integer overflow in pattern length calculation
> has been added to the -mm mm-nonmm-unstable branch.  Its filename is
>      lib-ts_kmp-fix-integer-overflow-in-pattern-length-calculation.patch
>
> This patch will shortly appear at
>      https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/lib-ts_kmp-fix-integer-overflow-in-pattern-length-calculation.patch
>
> This patch will later appear in the mm-nonmm-unstable branch at
>     git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
>
> Before you just go and hit "reply", please:
>    a) Consider who else should be cc'ed
>    b) Prefer to cc a suitable mailing list as well
>    c) Ideally: find the original patch on the mailing list and do a
>       reply-to-all to that, adding suitable additional cc's
>
> *** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
>
> The -mm tree is included into linux-next via various
> branches at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
> and is updated there most days
>
> ------------------------------------------------------
> From: Josh Law <objecting@objecting.org>
> Subject: lib/ts_kmp: fix integer overflow in pattern length calculation
> Date: Sun, 8 Mar 2026 20:20:28 +0000
>
> The ts_kmp algorithm stores its prefix_tbl[] table and pattern in a single
> allocation sized from the pattern length.  If the prefix_tbl[] size
> calculation wraps, the resulting allocation can be too small and
> subsequent pattern copies can overflow it.
>
> Fix this by rejecting zero-length patterns and by using overflow helpers
> before calculating the combined allocation size.
>
> Link: https://lkml.kernel.org/r/20260308202028.2889285-2-objecting@objecting.org
> Signed-off-by: Josh Law <objecting@objecting.org>
> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
> ---
>
> lib/ts_kmp.c |   18 ++++++++++++++++--
> 1 file changed, 16 insertions(+), 2 deletions(-)
>
> --- a/lib/ts_kmp.c~lib-ts_kmp-fix-integer-overflow-in-pattern-length-calculation
> +++ a/lib/ts_kmp.c
> @@ -94,8 +94,22 @@ static struct ts_config *kmp_init(const
>     struct ts_config *conf;
>     struct ts_kmp *kmp;
>     int i;
> -   unsigned int prefix_tbl_len = len * sizeof(unsigned int);
> -   size_t priv_size = sizeof(*kmp) + len + prefix_tbl_len;
> +   unsigned int prefix_tbl_len;
> +   size_t priv_size;
> +
> +   /* Zero-length patterns would make kmp_find() read beyond kmp->pattern. */
> +   if (unlikely(!len))
> +       return ERR_PTR(-EINVAL);
> +
> +   /*
> +    * kmp->pattern is stored immediately after the prefix_tbl[] table.
> +    * Reject lengths that would wrap while sizing either region.
> +    */
> +   if (unlikely(check_mul_overflow(len, sizeof(*kmp->prefix_tbl),
> +                   &prefix_tbl_len) ||
> +            check_add_overflow(sizeof(*kmp), (size_t)len, &priv_size) ||
> +            check_add_overflow(priv_size, prefix_tbl_len, &priv_size)))
> +       return ERR_PTR(-EINVAL);
>
>     conf = alloc_ts_config(priv_size, gfp_mask);
>     if (IS_ERR(conf))
> _
>
> Patches currently in -mm which might be from objecting@objecting.org are
>
> lib-maple_tree-fix-swapped-arguments-in-mas_safe_pivot-call.patch
> lib-glob-fix-grammar-and-replace-non-inclusive-terminology.patch
> lib-glob-add-explicit-include-for-exporth.patch
> lib-glob-replace-bitwise-or-with-logical-operation-on-boolean.patch
> lib-glob-clean-up-bool-abuse-in-pointer-arithmetic.patch
> lib-uuid-fix-typo-reversion-to-revision-in-comment.patch
> lib-inflate-fix-memory-leak-in-inflate_fixed-on-inflate_codes-failure.patch
> lib-inflate-fix-memory-leak-in-inflate_dynamic-on-inflate_codes-failure.patch
> lib-inflate-fix-grammar-in-comment-variable-to-variables.patch
> lib-inflate-fix-typo-this-results-to-the-results-in-comment.patch
> lib-bug-fix-inconsistent-capitalization-in-bug-message.patch
> lib-bug-remove-unnecessary-variable-initializations.patch
> lib-idr-fix-ida_find_first_range-missing-ids-across-chunk-boundaries.patch
> lib-decompress_bunzip2-fix-32-bit-shift-undefined-behavior.patch
> maintainers-add-josh-law-as-reviewer-for-library-code.patch
> lib-bootconfig-fix-typo-budy-in-_xbc_exit-comment.patch
> lib-ts_bm-fix-integer-overflow-in-pattern-length-calculation.patch
> lib-ts_kmp-fix-integer-overflow-in-pattern-length-calculation.patch

Do you think these patches require Cc: stable?
Idk the severity of the bug to you, the swapped arguments got Cc stable,  overflow could cause some screwy bugs


V/R



Josh Law