From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Monjalon Subject: Re: [PATCH 1/2] bus/dpaa: fix build with gcc 9.0 Date: Thu, 01 Nov 2018 20:45:53 +0100 Message-ID: <2352543.cRkeIK1bzI@xps> References: <20181101131951.91459-1-ferruh.yigit@intel.com> <20181101174550.GA23396@bricha3-MOBL.ger.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Cc: Bruce Richardson , Stephen Hemminger , Hemant Agrawal , Shreyansh Jain , dev@dpdk.org, stable@dpdk.org To: Ferruh Yigit Return-path: In-Reply-To: List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" 01/11/2018 19:21, Ferruh Yigit: > On 11/1/2018 5:45 PM, Bruce Richardson wrote: > > On Thu, Nov 01, 2018 at 10:42:31AM -0700, Stephen Hemminger wrote: > >> On Thu, 1 Nov 2018 13:19:50 +0000 > >> Ferruh Yigit wrote: > >> > >>> build error: > >>> In function =E2=80=98fman_if_init=E2=80=99, > >>> .../drivers/bus/dpaa/base/fman/fman.c:186:2: > >>> error: =E2=80=98strncpy=E2=80=99 output may be truncated copying = 4095 bytes from a > >>> string of length 4095 [-Werror=3Dstringop-truncation] > >>> strncpy(__if->node_path, dpa_node->full_name, PATH_MAX - 1); > >>> > >>> strncpy may result a not null-terminated string, > >>> replaced it with rte_strscpy > >>> > >>> Fixes: 5b22cf744689 ("bus/dpaa: introducing FMan configurations") > >>> Cc: stable@dpdk.org > >>> > >>> Signed-off-by: Ferruh Yigit > >> > >> Maybe just use strlcpy instead of a DPDK specific function. > >> That way if Gcc gets smarter it can check that as well. > >=20 > > +1 > > While I get the point of strscpy, if strlcpy is good enough for openbsd, > > it's good enough for me! :-) >=20 > I prefer strscpy but for both cases return value is not checked at all, s= o makes > hard to argue the benefit of the more proper return value J >=20 > So, as a rule of thumb are we saying strlcpy is dpdk preferred copy funct= ion? Yes