selinux.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: jwcart2 <jwcart2@tycho.nsa.gov>
To: Nicolas Iooss <nicolas.iooss@m4x.org>, selinux@vger.kernel.org
Subject: Re: [Non-DoD Source] [PATCH 2/9] libsepol/cil: help static analyzers by aborting when an allocation fails
Date: Tue, 10 Sep 2019 14:48:53 -0400	[thread overview]
Message-ID: <920d5f9a-147d-1bcc-8689-0656997c6d09@tycho.nsa.gov> (raw)
In-Reply-To: <20190901180636.31586-3-nicolas.iooss@m4x.org>

On 9/1/19 2:06 PM, Nicolas Iooss wrote:
> When allocating memory with cil_* helpers, if malloc/calloc/realloc/...
> failed, (*cil_mem_error_handler)() is called. Implementations of this
> function are expected not to return to the caller, and the default one
> calls exit(1) to ensure this. In order for static analyzers to find out
> that cil_malloc/cil_realloc/... never returns a NULL pointer when
> failing to allocate some memory, introduce a call to abort().
> 
> This decreases the number of false positive warnings about null pointer
> dereferences reported by Infer static analyzer.
> 
> Signed-off-by: Nicolas Iooss <nicolas.iooss@m4x.org>
> ---
>   libsepol/cil/src/cil_mem.c | 5 +++++
>   1 file changed, 5 insertions(+)
> 
> diff --git a/libsepol/cil/src/cil_mem.c b/libsepol/cil/src/cil_mem.c
> index 12c59be21914..885431d8a8fd 100644
> --- a/libsepol/cil/src/cil_mem.c
> +++ b/libsepol/cil/src/cil_mem.c
> @@ -55,6 +55,7 @@ void *cil_malloc(size_t size)
>   			return NULL;
>   		}
>   		(*cil_mem_error_handler)();
> +		abort();
>   	}
> 

I am going to have to think about this one.

I think the real answer might be to rip out the (*cil_mem_error_handler)() 
statements and replace them with what is in the default handler. I notice now 
that even though there is a cil_set_mem_error_handler() function, it is not used 
anywhere.

>   	return mem;
> @@ -65,6 +66,7 @@ void *cil_calloc(size_t num_elements, size_t element_size)
>   	void *mem = calloc(num_elements, element_size);
>   	if (mem == NULL){
>   		(*cil_mem_error_handler)();
> +		abort();
>   	}
>   
>   	return mem;
> @@ -78,6 +80,7 @@ void *cil_realloc(void *ptr, size_t size)
>   			return NULL;
>   		}
>   		(*cil_mem_error_handler)();
> +		abort();
>   	}
>   
>   	return mem;
> @@ -95,6 +98,7 @@ char *cil_strdup(const char *str)
>   	mem = strdup(str);
>   	if (mem == NULL) {
>   		(*cil_mem_error_handler)();
> +		abort();
>   	}
>   
>   	return mem;
> @@ -111,6 +115,7 @@ __attribute__ ((format (printf, 2, 3))) int cil_asprintf(char **strp, const char
>   
>   	if (rc == -1) {
>   		(*cil_mem_error_handler)();
> +		abort();
>   	}
>   
>   	return rc;
> 


-- 
James Carter <jwcart2@tycho.nsa.gov>
National Security Agency

  reply	other threads:[~2019-09-10 18:55 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-01 18:06 [PATCH 0/9] Fix issues found by static analyzers Nicolas Iooss
2019-09-01 18:06 ` [PATCH 1/9] semodule-utils: fix comparison with argc Nicolas Iooss
2019-09-01 18:06 ` [PATCH 2/9] libsepol/cil: help static analyzers by aborting when an allocation fails Nicolas Iooss
2019-09-10 18:48   ` jwcart2 [this message]
2019-09-01 18:06 ` [PATCH 3/9] libsepol: do not dereference a failed allocated pointer Nicolas Iooss
2019-09-10 18:52   ` [Non-DoD Source] " jwcart2
2019-09-10 20:11     ` jwcart2
2019-09-01 18:06 ` [PATCH 4/9] libsepol: do not dereference scope if it can be NULL Nicolas Iooss
2019-09-01 18:06 ` [PATCH 5/9] libsepol: reset *p to NULL if sepol_module_package_create fails Nicolas Iooss
2019-09-01 18:06 ` [PATCH 6/9] libsepol/cil: do not dereference perm_value_to_cil when it has not been allocated Nicolas Iooss
2019-09-01 18:06 ` [PATCH 7/9] python/chcat: remove unnecessary assignment Nicolas Iooss
2019-09-01 18:06 ` [PATCH 8/9] python/sepolicy: remove unnecessary pass statement Nicolas Iooss
2019-09-01 18:06 ` [PATCH 9/9] libsepol/tests: do not dereference a NULL pointer Nicolas Iooss
2019-09-16 16:46 ` [Non-DoD Source] [PATCH 0/9] Fix issues found by static analyzers jwcart2
2019-09-17 15:01   ` jwcart2

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=920d5f9a-147d-1bcc-8689-0656997c6d09@tycho.nsa.gov \
    --to=jwcart2@tycho.nsa.gov \
    --cc=nicolas.iooss@m4x.org \
    --cc=selinux@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).