From: Szabolcs Nagy <szabolcs.nagy@arm.com>
To: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Cc: libc-alpha@sourceware.org
Subject: Re: [PATCH] string: Fix OOB read on generic strncmp
Date: Wed, 22 Feb 2023 12:18:18 +0000 [thread overview]
Message-ID: <Y/YIChcSJhKEas/y@arm.com> (raw)
In-Reply-To: <20230221190612.2034413-1-adhemerval.zanella@linaro.org>
The 02/21/2023 16:06, Adhemerval Zanella wrote:
> For unaligned case, reading ahead can only be done if parting reads
> matches the aligned input.
>
> Also extend the stratcliff tests to check such cases.
>
> Checked on x86_64-linux-gnu, i686-linux-gnu, powerpc64-linux-gnu,
> and powerpc-linux-gnu by removing the arch-specific assembly
> implementation and disabling multi-arch (it covers both LE and BE
> for 64 and 32 bits).
> ---
> string/stratcliff.c | 17 +++++++++++++++++
> string/strncmp.c | 13 ++++++++++++-
> 2 files changed, 29 insertions(+), 1 deletion(-)
>
> diff --git a/string/stratcliff.c b/string/stratcliff.c
> index 74d64cc03d..864d856921 100644
> --- a/string/stratcliff.c
> +++ b/string/stratcliff.c
> @@ -409,6 +409,23 @@ do_test (void)
> }
> }
>
> + for (outer = 1; outer < 32; ++outer)
> + for (middle = 0; middle < 16; ++middle)
> + {
> + MEMSET (adr + middle, L('T'), 256);
> + adr[256] = L('\0');
> + MEMSET (dest + nchars - outer, L('T'), outer - 1);
> + dest[nchars - outer] = L('U');
> +
> + if (STRNCMP (adr + middle, &dest[nchars - middle - 1], outer) > 0)
> + {
> + printf ("%s 1 flunked for outer = %zu, middle = %zu, "
> + "inner = %zu\n",
> + STRINGIFY (STRNCMP), outer, middle, inner);
> + result = 1;
> + }
> + }
> +
this depends on dest[nchars-1] != 'T' when outer > middle+1
and dest[nchars-middle-1] <= 'T' when outer < middle
which is not clear from the context.
below the existing
if (STRNCMP (adr + middle, dest + nchars - outer, outer) >= 0) ...
i'd just add
if (STRNCMP (adr + middle, dest + nchars - outer, outer + 99) >= 0) ...
and then with flipped args too.
> diff --git a/string/strncmp.c b/string/strncmp.c
> index 4c8bf36bb9..751bf53d55 100644
> --- a/string/strncmp.c
> +++ b/string/strncmp.c
> @@ -73,7 +73,11 @@ strncmp_unaligned_loop (const op_t *x1, const op_t *x2, op_t w1, uintptr_t ofs,
> uintptr_t sh_2 = sizeof(op_t) * CHAR_BIT - sh_1;
>
> op_t w2 = MERGE (w2a, sh_1, (op_t)-1, sh_2);
> - if (!has_zero (w2) && n > (sizeof (op_t) - ofs))
> +
> + /* Reading ahead is wrong if w1 and w2 already differs. */
> + op_t w1a = MERGE (w1, 0, (op_t)-1, sh_2);
> +
> + if (!has_zero (w2) && w2 == w1a && n >= (sizeof (op_t) - ofs))
> {
> op_t w2b;
>
> @@ -90,6 +94,13 @@ strncmp_unaligned_loop (const op_t *x1, const op_t *x2, op_t w1, uintptr_t ofs,
> if (has_zero (w2b) || n <= (sizeof (op_t) - ofs))
> break;
> w1 = *x1++;
> +
> + /* Reading ahead is wrong if w1 and w2 already differs. */
> + w2 = MERGE (w2b, sh_1, (op_t)-1, sh_2);
> + w1a = MERGE (w1, 0, (op_t)-1, sh_2);
> + if (w2 != w1a)
> + return final_cmp (w1a, w2, n);
> +
> w2a = w2b;
> }
i have difficulty following this code, but it looks ok to me.
prev parent reply other threads:[~2023-02-22 12:18 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-21 19:06 Adhemerval Zanella
2023-02-22 12:18 ` Szabolcs Nagy [this message]
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=Y/YIChcSJhKEas/y@arm.com \
--to=szabolcs.nagy@arm.com \
--cc=adhemerval.zanella@linaro.org \
--cc=libc-alpha@sourceware.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).