From mboxrd@z Thu Jan 1 00:00:00 1970 From: Julia Lawall Date: Fri, 21 Jun 2019 09:21:11 +0000 Subject: Re: Coccinelle: Add a SmPL script for the reconsideration of redundant dev_err() calls Message-Id: List-Id: References: <05d85182-7ec3-8fc1-4bcd-fd2528de3a40@web.de> <34d528db-5582-5fe2-caeb-89bcb07a1d30@web.de> <13890878-9e5f-f297-7f7c-bcc1212d83b7@web.de> In-Reply-To: <13890878-9e5f-f297-7f7c-bcc1212d83b7@web.de> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Markus Elfring Cc: Michal Marek , kernel-janitors@vger.kernel.org, Nicolas Palix , LKML , Ding Xiang , Coccinelle On Fri, 21 Jun 2019, Markus Elfring wrote: > > I think that something like > > > > if (IS_ERR(e)) > > { > > <+... > > *dev_err(...) > > ...+> > > } > > > > would be more appropriate. Whether there is a return or not doesn't > > really matter. > > Do you find the following SmPL change specification useful and acceptable? > > > @deletion depends on patch@ > expression e; > @@ > e = devm_ioremap_resource(...); > if (IS_ERR(e)) > ( > -{ > - dev_err(...); > return (...); I still don't see the point of specifying return. Why not just S, where S is a statement metavariable? julia > -} > |{ I realize that you confuse conciseness with readability, but it would really look better to have the | on a line by itself. julia > <+... > - dev_err(...); > ...+> > } > ) > > > Would this approach need a version check for the Coccinelle software? Why would that be necessary?