git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
* [PATCH] Use strncpy to protect from buffer overruns.
@ 2010-06-09 10:22 Steven Michalske
  2010-06-09 12:44 ` Alex Riesen
  0 siblings, 1 reply; 5+ messages in thread
From: Steven Michalske @ 2010-06-09 10:22 UTC (permalink / raw
  To: git; +Cc: Steven Michalske

is_git_directory() uses strcpy with pointer arithmitic, protect it from
overflowing.  Even though we currently protect higher up when we have the
environment variable path passed in, we should protect the calls here.

Signed-off-by: Steven Michalske <smichalske@gmail.com>
---
 setup.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/setup.c b/setup.c
index 7e04602..0080299 100644
--- a/setup.c
+++ b/setup.c
@@ -170,22 +170,24 @@ static int is_git_directory(const char *suspect)
 	char path[PATH_MAX];
 	size_t len = strlen(suspect);
 
-	strcpy(path, suspect);
+	path[sizeof(path) - 1] = '\0';
+
+	strncpy(path, suspect, sizeof(path) - 1);
 	if (getenv(DB_ENVIRONMENT)) {
 		if (access(getenv(DB_ENVIRONMENT), X_OK))
 			return 0;
 	}
 	else {
-		strcpy(path + len, "/objects");
+		strncpy(path + len, "/objects", sizeof(path) - len - 1);
 		if (access(path, X_OK))
 			return 0;
 	}
 
-	strcpy(path + len, "/refs");
+	strncpy(path + len, "/refs", sizeof(path) - len - 1);
 	if (access(path, X_OK))
 		return 0;
 
-	strcpy(path + len, "/HEAD");
+	strncpy(path + len, "/HEAD", sizeof(path) - len - 1);
 	if (validate_headref(path))
 		return 0;
 
-- 
1.7.0.3

^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH] Use strncpy to protect from buffer overruns.
  2010-06-09 10:22 [PATCH] Use strncpy to protect from buffer overruns Steven Michalske
@ 2010-06-09 12:44 ` Alex Riesen
  2010-06-09 18:25   ` Steven Michalske
  0 siblings, 1 reply; 5+ messages in thread
From: Alex Riesen @ 2010-06-09 12:44 UTC (permalink / raw
  To: Steven Michalske; +Cc: git

On Wed, Jun 9, 2010 at 12:22, Steven Michalske <smichalske@gmail.com> wrote:
> is_git_directory() uses strcpy with pointer arithmitic, protect it from
> overflowing.  Even though we currently protect higher up when we have the
> environment variable path passed in, we should protect the calls here.

Why? The function is static.

> -       strcpy(path, suspect);
> +       path[sizeof(path) - 1] = '\0';
> +
> +       strncpy(path, suspect, sizeof(path) - 1);

And we have strlcpy for such things.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] Use strncpy to protect from buffer overruns.
  2010-06-09 12:44 ` Alex Riesen
@ 2010-06-09 18:25   ` Steven Michalske
  2010-06-09 19:31     ` Alex Riesen
  0 siblings, 1 reply; 5+ messages in thread
From: Steven Michalske @ 2010-06-09 18:25 UTC (permalink / raw
  To: Alex Riesen; +Cc: git


On Jun 9, 2010, at 5:44 AM, Alex Riesen wrote:

> On Wed, Jun 9, 2010 at 12:22, Steven Michalske <smichalske@gmail.com> wrote:
>> is_git_directory() uses strcpy with pointer arithmitic, protect it from
>> overflowing.  Even though we currently protect higher up when we have the
>> environment variable path passed in, we should protect the calls here.
> 
> Why? The function is static.
> 
The code might be locally constrained.

I always assume that a bit of code can be overwritten from other portions of code.

A small vulnerability is discovered that lets an attacker remove the length check or edit the pointer in the function call, but could not squeeze in the full shell code snippet.  But the now edited function here lets you put in arbitrarily long code.

>> -       strcpy(path, suspect);
>> +       path[sizeof(path) - 1] = '\0';
>> +
>> +       strncpy(path, suspect, sizeof(path) - 1);
> 
> And we have strlcpy for such things.

It is not portable.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] Use strncpy to protect from buffer overruns.
  2010-06-09 18:25   ` Steven Michalske
@ 2010-06-09 19:31     ` Alex Riesen
  2010-06-09 20:42       ` Steven Michalske
  0 siblings, 1 reply; 5+ messages in thread
From: Alex Riesen @ 2010-06-09 19:31 UTC (permalink / raw
  To: Steven Michalske; +Cc: git

On Wed, Jun 9, 2010 at 20:25, Steven Michalske <smichalske@gmail.com> wrote:
>> On Wed, Jun 9, 2010 at 12:22, Steven Michalske <smichalske@gmail.com> wrote:
>>> is_git_directory() uses strcpy with pointer arithmitic, protect it from
>>> overflowing.  Even though we currently protect higher up when we have the
>>> environment variable path passed in, we should protect the calls here.
>>
>> Why? The function is static.
>>
> The code might be locally constrained.
>
> I always assume that a bit of code can be overwritten from other portions of code.
>
> A small vulnerability is discovered that lets an attacker remove the length check
> or edit the pointer in the function call, but could not squeeze in the full shell code
> snippet.  But the now edited function here lets you put in arbitrarily long code.

Eh?

>>> -       strcpy(path, suspect);
>>> +       path[sizeof(path) - 1] = '\0';
>>> +
>>> +       strncpy(path, suspect, sizeof(path) - 1);
>>
>> And we have strlcpy for such things.
>
> It is not portable.

Git has its own copy of the function:

  $ git ls-files *strlcpy.c

  $

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] Use strncpy to protect from buffer overruns.
  2010-06-09 19:31     ` Alex Riesen
@ 2010-06-09 20:42       ` Steven Michalske
  0 siblings, 0 replies; 5+ messages in thread
From: Steven Michalske @ 2010-06-09 20:42 UTC (permalink / raw
  To: Alex Riesen; +Cc: git


On Jun 9, 2010, at 12:31 PM, Alex Riesen wrote:

> On Wed, Jun 9, 2010 at 20:25, Steven Michalske <smichalske@gmail.com> wrote:
>>> On Wed, Jun 9, 2010 at 12:22, Steven Michalske <smichalske@gmail.com> wrote:
>>>> is_git_directory() uses strcpy with pointer arithmitic, protect it from
>>>> overflowing.  Even though we currently protect higher up when we have the
>>>> environment variable path passed in, we should protect the calls here.
>>> 
>>> Why? The function is static.
>>> 
>> The code might be locally constrained.
>> 
>> I always assume that a bit of code can be overwritten from other portions of code.
>> 
>> A small vulnerability is discovered that lets an attacker remove the length check
>> or edit the pointer in the function call, but could not squeeze in the full shell code
>> snippet.  But the now edited function here lets you put in arbitrarily long code.
> 
> Eh?
> 
Basically the protection is not robust against malicious code.  It's armored with leather, not the modern full body armor.

>>>> -       strcpy(path, suspect);
>>>> +       path[sizeof(path) - 1] = '\0';
>>>> +
>>>> +       strncpy(path, suspect, sizeof(path) - 1);
>>> 
>>> And we have strlcpy for such things.
>> 
>> It is not portable.
> 
> Git has its own copy of the function:
> 
>  $ git ls-files *strlcpy.c
> 
>  $

Good to know, I could refactor with this.

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2010-06-09 20:42 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-06-09 10:22 [PATCH] Use strncpy to protect from buffer overruns Steven Michalske
2010-06-09 12:44 ` Alex Riesen
2010-06-09 18:25   ` Steven Michalske
2010-06-09 19:31     ` Alex Riesen
2010-06-09 20:42       ` Steven Michalske

Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/git.git

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).