netfilter-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Phil Sutter <phil@nwl.cc>
To: Thomas Haller <thaller@redhat.com>
Cc: NetFilter <netfilter-devel@vger.kernel.org>
Subject: Re: [PATCH nft 2/4] gmputil: add nft_gmp_free() to free strings from mpz_get_str()
Date: Wed, 20 Sep 2023 16:05:30 +0200	[thread overview]
Message-ID: <ZQr8KsFAXIT0mca9@orbyte.nwl.cc> (raw)
In-Reply-To: <20230920131554.204899-3-thaller@redhat.com>

On Wed, Sep 20, 2023 at 03:13:39PM +0200, Thomas Haller wrote:
> mpz_get_str() (with NULL as first argument) will allocate a buffer using
> the allocator functions (mp_set_memory_functions()). We should free
> those buffers with the corresponding free function.
> 
> Add nft_gmp_free() for that and use it.
> 
> The name nft_gmp_free() is chosen because "mini-gmp.c" already has an
> internal define called gmp_free(). There wouldn't be a direct conflict,
> but using the same name is confusing. And maybe our own defines should
> have a clear nft prefix.
> 
> Signed-off-by: Thomas Haller <thaller@redhat.com>
> ---
>  include/gmputil.h |  2 ++
>  src/evaluate.c    |  6 +++---
>  src/gmputil.c     | 21 ++++++++++++++++++++-
>  3 files changed, 25 insertions(+), 4 deletions(-)
> 
> diff --git a/include/gmputil.h b/include/gmputil.h
> index c524aced16ac..d1f4dcd2f1c3 100644
> --- a/include/gmputil.h
> +++ b/include/gmputil.h
> @@ -77,4 +77,6 @@ extern void __mpz_switch_byteorder(mpz_t rop, unsigned int len);
>  	__mpz_switch_byteorder(rop, len);			\
>  }
>  
> +void nft_gmp_free(void *ptr);
> +
>  #endif /* NFTABLES_GMPUTIL_H */
> diff --git a/src/evaluate.c b/src/evaluate.c
> index 03586922848a..e5c7e03a927f 100644
> --- a/src/evaluate.c
> +++ b/src/evaluate.c
> @@ -401,7 +401,7 @@ static int expr_evaluate_integer(struct eval_ctx *ctx, struct expr **exprp)
>  		expr_error(ctx->msgs, expr,
>  			   "Value %s exceeds valid range 0-%u",
>  			   valstr, ctx->ectx.maxval);
> -		free(valstr);
> +		nft_gmp_free(valstr);
>  		return -1;
>  	}
>  
> @@ -417,8 +417,8 @@ static int expr_evaluate_integer(struct eval_ctx *ctx, struct expr **exprp)
>  		expr_error(ctx->msgs, expr,
>  			   "Value %s exceeds valid range 0-%s",
>  			   valstr, rangestr);
> -		free(valstr);
> -		free(rangestr);
> +		nft_gmp_free(valstr);
> +		nft_gmp_free(rangestr);
>  		mpz_clear(mask);
>  		return -1;
>  	}
> diff --git a/src/gmputil.c b/src/gmputil.c
> index bf472c65de48..550c141294a3 100644
> --- a/src/gmputil.c
> +++ b/src/gmputil.c
> @@ -185,7 +185,7 @@ int mpz_vfprintf(FILE *fp, const char *f, va_list args)
>  
>  			str = mpz_get_str(NULL, base, *value);
>  			ok = str && fwrite(str, 1, len, fp) == len;
> -			free(str);
> +			nft_gmp_free(str);
>  
>  			if (!ok)
>  				return -1;
> @@ -197,3 +197,22 @@ int mpz_vfprintf(FILE *fp, const char *f, va_list args)
>  	return n;
>  }
>  #endif
> +
> +void nft_gmp_free(void *ptr)
> +{
> +	void (*free_fcn)(void *, size_t);
> +
> +	/* When we get allocated memory from gmp, it was allocated via the
> +	 * allocator() from mp_set_memory_functions(). We should pair the free
> +	 * with the corresponding free function, which we get via
> +	 * mp_get_memory_functions().
> +	 *
> +	 * It's not clear what the correct blk_size is. The default allocator
> +	 * function of gmp just wraps free() and ignores the extra argument.
> +	 * Assume 0 is fine.
> +	 */
> +
> +	mp_get_memory_functions(NULL, NULL, &free_fcn);

Do we have to expect the returned pointer to change at run-time? Because
if not, couldn't one make free_fcn static and call
mp_get_memory_functions() only if it's NULL?

Cheers, Phil

  reply	other threads:[~2023-09-20 14:05 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-09-20 13:13 [PATCH nft 0/4] remove xfree() and add free_const()+nft_gmp_free() Thomas Haller
2023-09-20 13:13 ` [PATCH nft 1/4] datatype: don't return a const string from cgroupv2_get_path() Thomas Haller
2023-09-20 13:13 ` [PATCH nft 2/4] gmputil: add nft_gmp_free() to free strings from mpz_get_str() Thomas Haller
2023-09-20 14:05   ` Phil Sutter [this message]
2023-09-20 14:46     ` Thomas Haller
2023-09-20 16:04       ` Phil Sutter
2023-09-20 13:13 ` [PATCH nft 3/4] all: add free_const() and use it instead of xfree() Thomas Haller
2023-09-20 14:13   ` Phil Sutter
2023-09-20 16:06     ` Pablo Neira Ayuso
2023-09-20 16:49       ` Phil Sutter
2023-09-20 16:52         ` Pablo Neira Ayuso
2023-09-20 18:03         ` Thomas Haller
2023-09-20 18:22           ` Phil Sutter
2023-09-20 19:48             ` Thomas Haller
2023-09-20 22:50               ` Phil Sutter
2023-09-21  9:08                 ` Thomas Haller
2023-09-20 13:13 ` [PATCH nft 4/4] all: remove xfree() and use plain free() Thomas Haller

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=ZQr8KsFAXIT0mca9@orbyte.nwl.cc \
    --to=phil@nwl.cc \
    --cc=netfilter-devel@vger.kernel.org \
    --cc=thaller@redhat.com \
    /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).