All of lore.kernel.org
 help / color / mirror / Atom feed
From: Bruce Richardson <bruce.richardson@intel.com>
To: Stephen Hemminger <stephen@networkplumber.org>
Cc: <dev@dpdk.org>, <fengchengwen@huawei.com>
Subject: Re: [PATCH] test/argparse: fix out of bound memcpy
Date: Fri, 27 Jun 2025 19:56:57 +0100	[thread overview]
Message-ID: <aF7pef1HMfjT88-e@bricha3-mobl1.ger.corp.intel.com> (raw)
In-Reply-To: <20250627162305.340042-1-stephen@networkplumber.org>

On Fri, Jun 27, 2025 at 09:22:35AM -0700, Stephen Hemminger wrote:
> The rte_argparse API use variable length arrays for the args.
> But the test was only putting space on stack for the argparse
> part, not the args. This can lead to out of bounds writes.
> 
> The bug only gets detected if DPDK is compiled with LTO.
> In function ‘test_argparse_copy’,
>     inlined from ‘test_argparse_init_obj’ at ../app/test/test_argparse.c:108:2,
>     inlined from ‘test_argparse_opt_callback_parse_int_of_no_val’ at ../app/test/test_argparse.c:490:8:
> ../app/test/test_argparse.c:96:17: warning: ‘memcpy’ writing 56 bytes into a region of size 0 overflows the destination [-Wstringop-overflow=]
>    96 |                 memcpy(&dst->args[i], &src->args[i], sizeof(src->args[i]));
> 
> Fixes: 6c5c6571601c ("argparse: verify argument config")
> Cc: fengchengwen@huawei.com
> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
> ---

It looks to me like this is a false positive. If it's not, then the whole
method of declaring argparse arguments is broken, and the library is not
really usable.

See below for what I see in gdb for a regular (non-LTO) debug build. Looks
to me like the compiler is doing the right thing.

/Bruce

>  app/test/test_argparse.c | 56 ++++++++++++++++------------------------
>  1 file changed, 22 insertions(+), 34 deletions(-)
> 
> diff --git a/app/test/test_argparse.c b/app/test/test_argparse.c
> index 0a229752fa..f4b33e2726 100644
> --- a/app/test/test_argparse.c
> +++ b/app/test/test_argparse.c
> @@ -70,43 +70,31 @@ test_argparse_callback(uint32_t index, const char *value, void *opaque)
>  	return 0;
>  }
>  
> -/* valid templater, must contain at least two args. */
> -#define argparse_templater() { \
> -	.prog_name = "test_argparse", \
> -	.usage = "-a xx -b yy", \
> -	.descriptor = NULL, \
> -	.epilog = NULL, \
> -	.exit_on_error = false, \
> -	.callback = test_argparse_callback, \
> -	.args = { \
> -		{ "--abc", "-a", "abc argument", (void *)1, (void *)1, \
> -			RTE_ARGPARSE_VALUE_NONE, RTE_ARGPARSE_VALUE_TYPE_NONE }, \
> -		{ "--xyz", "-x", "xyz argument", (void *)1, (void *)2, \
> -			RTE_ARGPARSE_VALUE_NONE, RTE_ARGPARSE_VALUE_TYPE_NONE }, \
> -		ARGPARSE_ARG_END(), \
> -	}, \
> -}
> -
> -static void
> -test_argparse_copy(struct rte_argparse *dst, struct rte_argparse *src)
> -{
> -	uint32_t i;
> -	memcpy(dst, src, sizeof(*src));
> -	for (i = 0; /* NULL */; i++) {
> -		memcpy(&dst->args[i], &src->args[i], sizeof(src->args[i]));
> -		if (src->args[i].name_long == NULL)
> -			break;
> -	}
> -}
> -
>  static struct rte_argparse *
>  test_argparse_init_obj(void)
>  {
> -	static struct rte_argparse backup = argparse_templater();
> -	static struct rte_argparse obj = argparse_templater();
> -	/* Because obj may be overwritten, do a deep copy. */

Running gdb and querying the layout of items in this function I get:

Thread 1 "dpdk-test" hit Breakpoint 1, test_argparse_init_obj () at ../app/test/test_argparse.c:108
108		test_argparse_copy(&obj, &backup);
(gdb) print &backup
$1 = (struct rte_argparse *) 0x555556d2b8a0 <backup>
(gdb) print &obj
$2 = (struct rte_argparse *) 0x555556d2b740 <obj>
(gdb) print 0xb8a0-0xb740
$8 = 352
(gdb) print sizeof(backup)
$9 = 184
(gdb) print sizeof(backup->args[0])
$10 = 56
(gdb) print sizeof(backup->args[0])*3 + sizeof(backup)
$11 = 352
(gdb) 

So we have the space available and allocated for the full structure plus
the 3 args. This means that the memcpy is not going to overflow.

Now, the separate question arises as to whether there are better methods to
initialize things in this test. That's a different issue, and I suspect
that we don't need the memcpy at all, but for me the key thing is that the
syntax used in the templater macro is good for defining argparse arguments.


> -	test_argparse_copy(&obj, &backup);
> -	return &obj;
> +	static struct {
> +		struct rte_argparse cmd;
> +		struct rte_argparse_arg args[3];
> +	} obj;
> +
> +	obj.cmd = (struct rte_argparse) {
> +		.prog_name = "test_argparse",
> +		.usage = "-a xx -b yy",
> +		.exit_on_error = false,
> +		.callback = test_argparse_callback,
> +	};
> +	obj.args[0] = (struct rte_argparse_arg)
> +		{ "--abc", "-a", "abc argument", (void *)1, (void *)1,
> +			RTE_ARGPARSE_VALUE_NONE, RTE_ARGPARSE_VALUE_TYPE_NONE
> +		};
> +	obj.args[1] = (struct rte_argparse_arg)
> +		{ "--xyz", "-x", "xyz argument", (void *)1, (void *)2,
> +			RTE_ARGPARSE_VALUE_NONE, RTE_ARGPARSE_VALUE_TYPE_NONE
> +		};
> +	obj.args[2] = (struct rte_argparse_arg) ARGPARSE_ARG_END();
> +
> +	return &obj.cmd;
>  }
>  
>  static int
> -- 
> 2.47.2
> 

  reply	other threads:[~2025-06-27 18:57 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-06-27 16:22 [PATCH] test/argparse: fix out of bound memcpy Stephen Hemminger
2025-06-27 18:56 ` Bruce Richardson [this message]
2025-06-30 14:57   ` Stephen Hemminger
2025-06-30 14:58 ` [PATCH v2] test/argparse: change initialization to workaround LTO Stephen Hemminger
2025-06-30 15:20   ` Bruce Richardson
2025-06-30 15:23     ` Stephen Hemminger
2025-06-30 15:24     ` Stephen Hemminger
2025-07-01 15:41 ` [PATCH v3] " Stephen Hemminger
2025-07-01 15:48   ` Bruce Richardson
2025-07-07  6:42   ` fengchengwen
2025-07-07  4:12 ` [PATCH] test/argparse: fix out of bound memcpy fengchengwen
2025-09-09 13:49 ` [PATCH v4] test/argparse: change initialization to workaround LTO Stephen Hemminger
2025-10-23 19:09   ` Thomas Monjalon

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=aF7pef1HMfjT88-e@bricha3-mobl1.ger.corp.intel.com \
    --to=bruce.richardson@intel.com \
    --cc=dev@dpdk.org \
    --cc=fengchengwen@huawei.com \
    --cc=stephen@networkplumber.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.