Re: [PATCH] perf jvmti: Fix gcc string overflow warning

From: Arnaldo Carvalho de Melo
Date: Fri May 31 2019 - 08:09:20 EST


Em Fri, May 31, 2019 at 10:03:07AM +0200, Jiri Olsa escreveu:
> We are getting fake gcc warning when we compile with gcc9 (9.1.1):
>
> CC jvmti/libjvmti.o
> In file included from /usr/include/string.h:494,
> from jvmti/libjvmti.c:5:
> In function âstrncpyâ,
> inlined from âcopy_class_filename.constpropâ at jvmti/libjvmti.c:166:3:
> /usr/include/bits/string_fortified.h:106:10: error: â__builtin_strncpyâ specified bound depends on the length of the source argument [-Werror=stringop-overflow=]
> 106 | return __builtin___strncpy_chk (__dest, __src, __len, __bos (__dest));
> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> jvmti/libjvmti.c: In function âcopy_class_filename.constpropâ:
> jvmti/libjvmti.c:165:26: note: length computed here
> 165 | size_t file_name_len = strlen(file_name);
> | ^~~~~~~~~~~~~~~~~
> cc1: all warnings being treated as errors
>
> First I wanted to disable the check, but now I think the code
> could be more straight forward. There's no need to check the
> source size, strncpy will do that. We just need to make sure
> the string is correctly terminated.
>
> Cc: Ben Gainey <ben.gainey@xxxxxxx>
> Cc: Stephane Eranian <eranian@xxxxxxxxxx>
> Link: http://lkml.kernel.org/n/tip-sve3b63c550wr907e6ui6gx5@xxxxxxxxxxxxxx
> Signed-off-by: Jiri Olsa <jolsa@xxxxxxxxxx>
> ---
> tools/perf/jvmti/libjvmti.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/tools/perf/jvmti/libjvmti.c b/tools/perf/jvmti/libjvmti.c
> index aea7b1fe85aa..00fa0b7f1ad9 100644
> --- a/tools/perf/jvmti/libjvmti.c
> +++ b/tools/perf/jvmti/libjvmti.c
> @@ -162,8 +162,8 @@ copy_class_filename(const char * class_sign, const char * file_name, char * resu
> result[i] = '\0';
> } else {
> /* fallback case */
> - size_t file_name_len = strlen(file_name);
> - strncpy(result, file_name, file_name_len < max_length ? file_name_len : max_length);
> + strncpy(result, file_name, max_length - 1);
> + result[max_length - 1] = 0;

The usual idiom here, if we stay with strncpy is:

strncpy(result, file_name, max_length - 1)[max_length - 1] = 0;

one line instead of two, but there are other possibilities, what I've
done int these cases in tools/perf/ is to switch to strlcpy, so just
flip that 'n' to a 'l' and it should be enough.

For that we just have a copy of the kernel's strlcpy() implementation in
lib/string.c, and it has this doc:

/**
* strlcpy - Copy a C-string into a sized buffer
* @dest: Where to copy the string to
* @src: Where to copy the string from
* @size: size of destination buffer
*
* Compatible with ``*BSD``: the result is always a valid
* NUL-terminated string that fits in the buffer (unless,
* of course, the buffer size is zero). It does not pad
* out the result like strncpy() does.
*/

The kernel folks moved beyond that and in lib/string.c we have:

/**
* strscpy - Copy a C-string into a sized buffer
* @dest: Where to copy the string to
* @src: Where to copy the string from
* @count: Size of destination buffer
*
* Copy the string, or as much of it as fits, into the dest buffer. The
* behavior is undefined if the string buffers overlap. The destination
* buffer is always NUL terminated, unless it's zero-sized.
*
* Preferred to strlcpy() since the API doesn't require reading memory
* from the src string beyond the specified "count" bytes, and since
* the return value is easier to error-check than strlcpy()'s.
* In addition, the implementation is robust to the string changing out
* from underneath it, unlike the current strlcpy() implementation.
*
* Preferred to strncpy() since it always returns a valid string, and
* doesn't unnecessarily force the tail of the destination buffer to be
* zeroed. If zeroing is desired please use strscpy_pad().
*
* Return: The number of characters copied (not including the trailing
* %NUL) or -E2BIG if the destination buffer wasn't big enough.
*/
ssize_t strscpy(char *dest, const char *src, size_t count)



I think for these needs flipping that 'n' into a 'l' is good enough.

- Arnaldo

> }
> }
>
> --
> 2.21.0

--

- Arnaldo