public inbox for passt-dev@passt.top
 help / color / mirror / code / Atom feed
From: David Gibson <david@gibson.dropbear.id.au>
To: passt-dev@passt.top
Subject: Re: [PATCH 4/7] conf, tcp, udp: Arrays for ports need 2^16 values, not 2^16-8
Date: Thu, 22 Sep 2022 16:43:24 +1000	[thread overview]
Message-ID: <YywEDOOCiT/6VoEj@yekko> (raw)
In-Reply-To: <20220921205507.2742203-5-sbrivio@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 3482 bytes --]

On Wed, Sep 21, 2022 at 10:55:04PM +0200, Stefano Brivio wrote:
> Reported by David but also by Coverity (CWE-119):
> 
> 	In conf_ports: Out-of-bounds access to a buffer
> 
> ...not in practice, because the allocation size is rounded up
> anyway, but not nice either.

So... this helps, but it's not enough.  For starters, this is still
conceptually wrong - there should be 65536 bits in the bitmap, not
65535 (USHRT_MAX).  This patch just masks it by rounding up to 65536
rather than down to 65528.

Alas, this is not the only buffer overrun caused by an off-by-one in
port numbers: the delta_to_tap etc. global arrays in tcp.c and udp.c
should also have length 65536, rather than USHRT_MAX.  So should
tcp_sock_init_lo and friends.  There are also similar problems in
icmp.c with echo request id rather than port number.

Not a buffer overrun, but the USHRT_MAX in udp_remap_to_tap() and
udp_remap_to_init() is also out by one (consider the case of
delta==0).

I have a series which makes a number of cleanups to the port mapping
handling.  It fixes this bug and adds typing stuff that should make it
harder to make similar mistakes in future.

I held off on sending, since it's currently based on a lot of my
outstanding patches.  I think it wouldn't be too hard to rebase
directly on master, should I do that so you can fix this before going
on to debug the rest of my outstanding stuff?

> 
> Reported-by: David Gibson <david(a)gibson.dropbear.id.au>
> Signed-off-by: Stefano Brivio <sbrivio(a)redhat.com>
> ---
>  conf.c | 2 +-
>  tcp.h  | 4 ++--
>  udp.h  | 4 ++--
>  3 files changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/conf.c b/conf.c
> index d80233c..7ecfa1e 100644
> --- a/conf.c
> +++ b/conf.c
> @@ -127,8 +127,8 @@ static int conf_ports(struct ctx *c, char optname, const char *optarg,
>  {
>  	int start_src, end_src, start_dst, end_dst, exclude_only = 1, i, port;
>  	char addr_buf[sizeof(struct in6_addr)] = { 0 }, *addr = addr_buf;
> +	uint8_t *map, exclude[DIV_ROUND_UP(USHRT_MAX, 8)] = { 0 };
>  	void (*remap)(in_port_t port, in_port_t delta);
> -	uint8_t *map, exclude[USHRT_MAX / 8] = { 0 };
>  	char buf[BUFSIZ], *sep, *spec, *p;
>  	sa_family_t af = AF_UNSPEC;
>  
> diff --git a/tcp.h b/tcp.h
> index 7b720c1..6431b75 100644
> --- a/tcp.h
> +++ b/tcp.h
> @@ -69,9 +69,9 @@ struct tcp_ctx {
>  	uint64_t hash_secret[2];
>  	int conn_count;
>  	int splice_conn_count;
> -	uint8_t port_to_tap	[USHRT_MAX / 8];
> +	uint8_t port_to_tap	[DIV_ROUND_UP(USHRT_MAX, 8)];
>  	int init_detect_ports;
> -	uint8_t port_to_init	[USHRT_MAX / 8];
> +	uint8_t port_to_init	[DIV_ROUND_UP(USHRT_MAX, 8)];
>  	int ns_detect_ports;
>  	struct timespec timer_run;
>  #ifdef HAS_SND_WND
> diff --git a/udp.h b/udp.h
> index f16fe5e..8f82842 100644
> --- a/udp.h
> +++ b/udp.h
> @@ -53,9 +53,9 @@ union udp_epoll_ref {
>   * @timer_run:		Timestamp of most recent timer run
>   */
>  struct udp_ctx {
> -	uint8_t port_to_tap		[USHRT_MAX / 8];
> +	uint8_t port_to_tap		[DIV_ROUND_UP(USHRT_MAX, 8)];
>  	int init_detect_ports;
> -	uint8_t port_to_init		[USHRT_MAX / 8];
> +	uint8_t port_to_init		[DIV_ROUND_UP(USHRT_MAX, 8)];
>  	int ns_detect_ports;
>  	struct timespec timer_run;
>  };

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

  reply	other threads:[~2022-09-22  6:43 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-09-21 20:55 [PATCH 0/7] Fixes and workarounds for tests, Coverity warnings Stefano Brivio
2022-09-21 20:55 ` [PATCH 1/7] test/perf: Always use /sbin/sysctl in tcp test Stefano Brivio
2022-09-22  6:28   ` David Gibson
2022-09-21 20:55 ` [PATCH 2/7] test/perf: Switch performance test duration to 10 seconds instead of 30 Stefano Brivio
2022-09-22  6:29   ` David Gibson
2022-09-21 20:55 ` [PATCH 3/7] tap: Check return value of accept4() before calling getsockopt() Stefano Brivio
2022-09-22  6:30   ` David Gibson
2022-09-21 20:55 ` [PATCH 4/7] conf, tcp, udp: Arrays for ports need 2^16 values, not 2^16-8 Stefano Brivio
2022-09-22  6:43   ` David Gibson [this message]
2022-09-22  8:21     ` Stefano Brivio
2022-09-22 23:39       ` Stefano Brivio
2022-09-23  2:09         ` David Gibson
2022-09-21 20:55 ` [PATCH 5/7] test/lib: Restore IFS while executing directives in def blocks Stefano Brivio
2022-09-22  6:44   ` David Gibson
2022-09-22  8:25     ` Stefano Brivio
2022-09-23  6:55       ` David Gibson
2022-09-21 20:55 ` [PATCH 6/7] test/lib: Wait on iperf3 clients to be done, then send SIGINT to servers Stefano Brivio
2022-09-21 20:55 ` [PATCH 7/7] test/perf: Disable periodic throughput reports to avoid vhost hang Stefano Brivio
2022-09-22  6:46   ` David Gibson

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=YywEDOOCiT/6VoEj@yekko \
    --to=david@gibson.dropbear.id.au \
    --cc=passt-dev@passt.top \
    /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.
Code repositories for project(s) associated with this public inbox

	https://passt.top/passt

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 IMAP folder(s).