From 16756ca4edb55cdd8c88f4e123ffa6b94501d050 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Daniel=20Gr=C3=B6ber?= Date: Wed, 24 Jun 2020 15:29:58 +0200 Subject: src: Fix nfexp_snprintf return value docs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The docs currently say "[...] Otherwise, 0 is returned." which is just completely wrong. Just like nfct_snprintf the expected buffer size is returned. Signed-off-by: Daniel Gröber Signed-off-by: Pablo Neira Ayuso --- src/expect/api.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/src/expect/api.c b/src/expect/api.c index 33099d8..39cd092 100644 --- a/src/expect/api.c +++ b/src/expect/api.c @@ -795,8 +795,9 @@ int nfexp_catch(struct nfct_handle *h) * - NFEXP_O_LAYER: include layer 3 information in the output, this is * *only* required by NFEXP_O_DEFAULT. * - * On error, -1 is returned and errno is set appropiately. Otherwise, - * 0 is returned. + * On error, -1 is returned and errno is set appropiately. Otherwise the + * size of what _would_ be written is returned, even if the size of the + * buffer is insufficient. This behaviour is similar to snprintf. */ int nfexp_snprintf(char *buf, unsigned int size, -- cgit v1.2.3