From: Adhemerval Zanella Netto <adhemerval.zanella@linaro.org>
To: Joe Simmons-Talbott <josimmon@redhat.com>, libc-alpha@sourceware.org
Subject: Re: [PATCH] ifaddrs: Get rid of alloca
Date: Fri, 26 May 2023 11:22:31 -0300 [thread overview]
Message-ID: <b8f218e1-efc9-33da-1bd0-b6bfa486bf89@linaro.org> (raw)
In-Reply-To: <20230526002247.1898289-1-josimmon@redhat.com>
On 25/05/23 21:22, Joe Simmons-Talbott via Libc-alpha wrote:
> Use scratch_buffers rather than alloca to avoid potential stack
> overflows.
> ---
> sysdeps/unix/sysv/linux/ifaddrs.c | 36 +++++++++++++++----------------
> 1 file changed, 17 insertions(+), 19 deletions(-)
>
> diff --git a/sysdeps/unix/sysv/linux/ifaddrs.c b/sysdeps/unix/sysv/linux/ifaddrs.c
> index 184ee224cb..ca37b4f6db 100644
> --- a/sysdeps/unix/sysv/linux/ifaddrs.c
> +++ b/sysdeps/unix/sysv/linux/ifaddrs.c
> @@ -16,13 +16,13 @@
> License along with the GNU C Library; if not, see
> <https://www.gnu.org/licenses/>. */
>
> -#include <alloca.h>
> #include <assert.h>
> #include <errno.h>
> #include <ifaddrs.h>
> #include <net/if.h>
> #include <netinet/in.h>
> #include <netpacket/packet.h>
> +#include <scratch_buffer.h>
> #include <stdbool.h>
> #include <stdint.h>
> #include <stdlib.h>
> @@ -138,19 +138,11 @@ __netlink_request (struct netlink_handle *h, int type)
> #else
> const size_t buf_size = __getpagesize ();
> #endif
> - bool use_malloc = false;
> - char *buf;
> -
> - if (__libc_use_alloca (buf_size))
> - buf = alloca (buf_size);
> - else
> - {
> - buf = malloc (buf_size);
> - if (buf != NULL)
> - use_malloc = true;
> - else
> - goto out_fail;
> - }
> + struct scratch_buffer sbuf;
> + scratch_buffer_init (&sbuf);
> + if (!scratch_buffer_set_array_size (&sbuf, 1, buf_size))
> + goto out_fail;
> + char *buf = sbuf.data;
Although netlink documentation [1] specifies that user buffer needs to
be either 8kb or the architecture page size (whichever is bigger), this
has been changed over time and now a 8Kb is suffice [2]. In fact
using larger buffers is not really effective, unless you also tune the
netlink socket buffer size [3].
So to use effectively use netlink, you will need to use at least a 8Kb
buffer which is larger than the default static one (1024). So there is
no much point in using a scratch_buffer here.
Also, I think it should remove the PAGE_SIZE and use a 8Kb buffer instead.
Something like:
--
static void
ifree (char **ptr)
{
free (*ptr);
}
int
__netlink_request (struct netlink_handle *h, int type)
{
[...]
/* Netlink requires that user buffer needs to be either 8kb or page size
(whichever is bigger), however this has been changed over time and now
a 8Kb is suffice (check NLMSG_DEFAULT_SIZE on Linux
linux/include/linux/netlink.h). */
const size_t buf_size = 8192;
char *buf __attribute__ ((__cleanup__ (ifree))) = malloc (buf_size);
if (buf == NULL)
return -1;
[...]
}
--
[1] Documentation/userspace-api/netlink/intro.rst
[2] https://patchwork.ozlabs.org/project/netdev/patch/e9409957290af5249750afa0f10de3a6@linux.vnet.ibm.com/
[3] https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=d35c99ff77ecb2eb239731b799386f3b3637a31e
>
> struct iovec iov = { buf, buf_size };
>
> @@ -229,13 +221,11 @@ __netlink_request (struct netlink_handle *h, int type)
> h->end_ptr = nlm_next;
> }
>
> - if (use_malloc)
> - free (buf);
> + scratch_buffer_free (&sbuf);
> return 0;
>
> out_fail:
> - if (use_malloc)
> - free (buf);
> + scratch_buffer_free (&sbuf);
> return -1;
> }
>
> @@ -324,6 +314,7 @@ getifaddrs_internal (struct ifaddrs **ifap)
> char *ifa_data_ptr; /* Pointer to the unused part of memory for
> ifa_data. */
> int result = 0;
> + struct scratch_buffer buf;
>
> *ifap = NULL;
>
> @@ -425,7 +416,13 @@ getifaddrs_internal (struct ifaddrs **ifap)
> }
>
> /* Table for mapping kernel index to entry in our list. */
> - map_newlink_data = alloca (newlink * sizeof (int));
> + scratch_buffer_init (&buf);
> + if (!scratch_buffer_set_array_size (&buf, 1, newlink * sizeof (int)))
> + {
> + result = -1;
> + goto exit_free;
> + }
> + map_newlink_data = buf.data;
> memset (map_newlink_data, '\xff', newlink * sizeof (int));
>
> ifa_data_ptr = (char *) &ifas[newlink + newaddr];
> @@ -820,6 +817,7 @@ getifaddrs_internal (struct ifaddrs **ifap)
> exit_free:
> __netlink_free_handle (&nh);
> __netlink_close (&nh);
> + scratch_buffer_free (&buf);
This might trigger an invalid free on uninitialized memory if
__netlink_request fails. Move the on the scratch_buffer initialization
just after the buf declaration.
>
> return result;
> }
prev parent reply other threads:[~2023-05-26 14:22 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-26 0:22 Joe Simmons-Talbott
2023-05-26 14:22 ` Adhemerval Zanella Netto [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=b8f218e1-efc9-33da-1bd0-b6bfa486bf89@linaro.org \
--to=adhemerval.zanella@linaro.org \
--cc=josimmon@redhat.com \
--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).