From: Pablo Neira Ayuso <pablo@netfilter.org>
To: Florian Westphal <fw@strlen.de>
Cc: netfilter-devel@vger.kernel.org
Subject: Re: [PATCH nft] netlink: fix stack buffer overrun when emitting ranged expressions
Date: Tue, 18 Mar 2025 13:24:38 +0100 [thread overview]
Message-ID: <Z9lmBhYELKyJHHOk@calendula> (raw)
In-Reply-To: <20250314124159.2131-1-fw@strlen.de>
[-- Attachment #1: Type: text/plain, Size: 4050 bytes --]
Hi Florian,
On Fri, Mar 14, 2025 at 01:41:49PM +0100, Florian Westphal wrote:
> Included bogon input generates following Sanitizer splat:
>
> AddressSanitizer: dynamic-stack-buffer-overflow on address 0x7...
> WRITE of size 2 at 0x7fffffffcbe4 thread T0
> #0 0x0000003a68b8 in __asan_memset (src/nft+0x3a68b8) (BuildId: 3678ff51a5405c77e3e0492b9a985910efee73b8)
> #1 0x0000004eb603 in __mpz_export_data src/gmputil.c:108:2
> #2 0x0000004eb603 in netlink_export_pad src/netlink.c:256:2
> #3 0x0000004eb603 in netlink_gen_range src/netlink.c:471:2
> #4 0x0000004ea250 in __netlink_gen_data src/netlink.c:523:10
> #5 0x0000004e8ee3 in alloc_nftnl_setelem src/netlink.c:205:3
> #6 0x0000004d4541 in mnl_nft_setelem_batch src/mnl.c:1816:11
>
> Problem is that the range end is emitted to the buffer at the *padded*
> location (rounded up to next register size), but buffer sizing is
> based of the expression length, not the padded length.
>
> Also extend the test script: Capture stderr and if we see
> AddressSanitizer warning, make it fail.
>
> Same bug as the one fixed in 600b84631410 ("netlink: fix stack buffer overflow with sub-reg sized prefixes"),
> just in a different function.
>
> Apply same fix: no dynamic array + add a length check.
While at it, extend it for similar code too until there is a way to
consolidate this? See attachment.
> Signed-off-by: Florian Westphal <fw@strlen.de>
> ---
> src/netlink.c | 11 +++++++----
> tests/shell/testcases/bogons/assert_failures | 17 ++++++++++++++++-
> ...an_stack_buffer_overrun_in_netlink_gen_range | 6 ++++++
> 3 files changed, 29 insertions(+), 5 deletions(-)
> create mode 100644 tests/shell/testcases/bogons/nft-f/asan_stack_buffer_overrun_in_netlink_gen_range
>
> diff --git a/src/netlink.c b/src/netlink.c
> index 8e6e2066fe2a..52c2d8009b82 100644
> --- a/src/netlink.c
> +++ b/src/netlink.c
> @@ -462,11 +462,14 @@ static void netlink_gen_verdict(const struct expr *expr,
> static void netlink_gen_range(const struct expr *expr,
> struct nft_data_linearize *nld)
> {
> - unsigned int len = div_round_up(expr->left->len, BITS_PER_BYTE) * 2;
> - unsigned char data[len];
> - unsigned int offset = 0;
> + unsigned int len = (netlink_padded_len(expr->left->len) / BITS_PER_BYTE) * 2;
> + unsigned char data[NFT_MAX_EXPR_LEN_BYTES];
> + unsigned int offset;
>
> - memset(data, 0, len);
> + if (len > sizeof(data))
> + BUG("Value export of %u bytes would overflow", len);
> +
> + memset(data, 0, sizeof(data));
> offset = netlink_export_pad(data, expr->left->value, expr->left);
> netlink_export_pad(data + offset, expr->right->value, expr->right);
> nft_data_memcpy(nld, data, len);
> diff --git a/tests/shell/testcases/bogons/assert_failures b/tests/shell/testcases/bogons/assert_failures
> index 79099427c98a..3dee63b3f97b 100755
> --- a/tests/shell/testcases/bogons/assert_failures
> +++ b/tests/shell/testcases/bogons/assert_failures
> @@ -1,12 +1,27 @@
> #!/bin/bash
>
> dir=$(dirname $0)/nft-f/
> +tmpfile=$(mktemp)
> +
> +cleanup()
> +{
> + rm -f "$tmpfile"
> +}
> +
> +trap cleanup EXIT
>
> for f in $dir/*; do
> - $NFT --check -f "$f"
> + echo "Check $f"
> + $NFT --check -f "$f" 2> "$tmpfile"
>
> if [ $? -ne 1 ]; then
> echo "Bogus input file $f did not cause expected error code" 1>&2
> exit 111
> fi
> +
> + if grep AddressSanitizer "$tmpfile"; then
> + echo "Address sanitizer splat for $f" 1>&2
> + cat "$tmpfile"
> + exit 111
> + fi
> done
> diff --git a/tests/shell/testcases/bogons/nft-f/asan_stack_buffer_overrun_in_netlink_gen_range b/tests/shell/testcases/bogons/nft-f/asan_stack_buffer_overrun_in_netlink_gen_range
> new file mode 100644
> index 000000000000..2f7872e4accd
> --- /dev/null
> +++ b/tests/shell/testcases/bogons/nft-f/asan_stack_buffer_overrun_in_netlink_gen_range
> @@ -0,0 +1,6 @@
> +table ip test {
> + chain y {
> + redirect to :tcp dport map { 83 : 80/3, 84 :4 }
> + }
> +}
> +
> --
> 2.45.3
>
>
[-- Attachment #2: x.patch --]
[-- Type: text/x-diff, Size: 2967 bytes --]
diff --git a/src/netlink.c b/src/netlink.c
index 687b0b87da3a..dfb7f4d17147 100644
--- a/src/netlink.c
+++ b/src/netlink.c
@@ -330,11 +330,15 @@ static void nft_data_memcpy(struct nft_data_linearize *nld,
static void netlink_gen_concat_key(const struct expr *expr,
struct nft_data_linearize *nld)
{
- unsigned int len = expr->len / BITS_PER_BYTE, offset = 0;
- unsigned char data[len];
+ unsigned int len = netlink_padded_len(expr->len) / BITS_PER_BYTE;
+ unsigned char data[NFT_MAX_EXPR_LEN_BYTES];
+ unsigned int offset = 0;
const struct expr *i;
- memset(data, 0, len);
+ if (len > sizeof(data))
+ BUG("Value export of %u bytes would overflow", len);
+
+ memset(data, 0, sizeof(data));
list_for_each_entry(i, &expr->expressions, list)
offset += __netlink_gen_concat_key(expr->flags, i, data + offset);
@@ -373,11 +377,15 @@ static int __netlink_gen_concat_data(int end, const struct expr *i,
static void __netlink_gen_concat_expand(const struct expr *expr,
struct nft_data_linearize *nld)
{
- unsigned int len = div_round_up(expr->len, BITS_PER_BYTE) * 2, offset = 0;
- unsigned char data[len];
+ unsigned int len = (netlink_padded_len(expr->len) / BITS_PER_BYTE) * 2;
+ unsigned char data[NFT_MAX_EXPR_LEN_BYTES];
+ unsigned int offset = 0;
const struct expr *i;
- memset(data, 0, len);
+ if (len > sizeof(data))
+ BUG("Value export of %u bytes would overflow", len);
+
+ memset(data, 0, sizeof(data));
list_for_each_entry(i, &expr->expressions, list)
offset += __netlink_gen_concat_data(false, i, data + offset);
@@ -391,11 +399,15 @@ static void __netlink_gen_concat_expand(const struct expr *expr,
static void __netlink_gen_concat(const struct expr *expr,
struct nft_data_linearize *nld)
{
- unsigned int len = expr->len / BITS_PER_BYTE, offset = 0;
- unsigned char data[len];
+ unsigned int len = netlink_padded_len(expr->len) / BITS_PER_BYTE;
+ unsigned char data[NFT_MAX_EXPR_LEN_BYTES];
+ unsigned int offset = 0;
const struct expr *i;
- memset(data, 0, len);
+ if (len > sizeof(data))
+ BUG("Value export of %u bytes would overflow", len);
+
+ memset(data, 0, sizeof(data));
list_for_each_entry(i, &expr->expressions, list)
offset += __netlink_gen_concat_data(expr->flags, i, data + offset);
@@ -1218,10 +1230,15 @@ static struct expr *range_expr_reduce(struct expr *range)
static struct expr *netlink_parse_interval_elem(const struct set *set,
struct expr *expr)
{
- unsigned int len = div_round_up(expr->len, BITS_PER_BYTE);
+ unsigned int len = netlink_padded_len(expr->len) / BITS_PER_BYTE;
const struct datatype *dtype = set->data->dtype;
struct expr *range, *left, *right;
- char data[len];
+ char data[NFT_MAX_EXPR_LEN_BYTES];
+
+ if (len > sizeof(data))
+ BUG("Value export of %u bytes would overflow", len);
+
+ memset(data, 0, sizeof(data));
mpz_export_data(data, expr->value, dtype->byteorder, len);
left = constant_expr_alloc(&internal_location, dtype,
next prev parent reply other threads:[~2025-03-18 12:24 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-14 12:41 [PATCH nft] netlink: fix stack buffer overrun when emitting ranged expressions Florian Westphal
2025-03-18 12:24 ` Pablo Neira Ayuso [this message]
2025-03-18 13:24 ` Florian Westphal
2025-03-18 15:34 ` Pablo Neira Ayuso
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=Z9lmBhYELKyJHHOk@calendula \
--to=pablo@netfilter.org \
--cc=fw@strlen.de \
--cc=netfilter-devel@vger.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;
as well as URLs for NNTP newsgroup(s).