On Thu, Aug 02, 2018 at 07:37:52AM -0400, Neil Horman wrote: > On Wed, Aug 01, 2018 at 01:33:04PM -0300, Marcelo Ricardo Leitner wrote: > > I'm getting: > > nagle_snd.c: In function ‘main’: > > nagle_snd.c:275:3: warning: ‘strncpy’ output truncated before terminating nul copying 10 bytes from a string of the same length [-Wstringop-truncation] > > strncpy(message+i, "1234567890", 10); > > ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > > > Turns out the code is doing that on purpose. To silence it, lets use > > memcpy() instead. > > > > Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@xxxxxxxxx> > > --- > > src/apps/nagle_snd.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/src/apps/nagle_snd.c b/src/apps/nagle_snd.c > > index 1fc788d56c6e4dd99e89f73c8eafc829efc76dc5..bc960cca04bbc3d8e22a1757daf5fce7b76a95ce 100644 > > --- a/src/apps/nagle_snd.c > > +++ b/src/apps/nagle_snd.c > > @@ -272,7 +272,7 @@ main(int argc, char *argv[]) > > message = test_malloc((size + 1) * sizeof(u_int8_t)); > > > > for(i=0; i + 10 < size; i+= 10) > > - strncpy(message+i, "1234567890", 10); > > + memcpy(message+i, "1234567890", 10); > Are you sure you want to do that? strncpy limits the copy operation to > min(sizeof(message+i), 10), and the size of message can be specified on the > command line. memcpy just copies 10 bytes blindly, possibly overruning the end > of the array. Yes. The buffer size is handled/protected by the for() construction and by the termination of the pattern that is actually done by the subsequent strncpy() call in [1] below (and not in this one). Even if message size can be specified on the command line, the buffer is safe here. > > I think it may be more useful here to do this: > #pragma GCC diagnostic push > #pragma GCC diagnostic ignored "-Wstringop-truncation" > ... > strncpy(message+i, "1234567890", 10); > #pragma GCC diagnostic pop Interesting, > > You can wrap up those pragmas in a macro using _Pragma if you need to. Its a bit > ugly, but it clearly flags the truncation as an intended side effect. but considering the code around this particular strncpy, it sounds like an overkill to me. And too gcc-specific. > > Neil > > > strncpy(message+i, "1234567890", size-i); [1] the termination of the pattern, accommodating the end of the buffer, possibly unaligned to the pattern size. > > *(message+size) = 0; And set the NUL byte, in case the copy on [1] was interrupted by the len=size-i constraint. Marcelo > > > > -- > > 2.17.1 > > > > -- > > To unsubscribe from this list: send the line "unsubscribe linux-sctp" in > > the body of a message to majordomo@xxxxxxxxxxxxxxx > > More majordomo info at http://vger.kernel.org/majordomo-info.html > > -- To unsubscribe from this list: send the line "unsubscribe linux-sctp" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
