Date: Thu, 9 Jun 2011 14:40:35 -0400 From: John Baldwin <jhb@freebsd.org> To: gibbs@freebsd.org Cc: svn-src-head@freebsd.org, "Andrey V. Elsukov" <ae@freebsd.org>, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r221397 - in head/sys: kern sys Message-ID: <201106091440.36182.jhb@freebsd.org> In-Reply-To: <4DF101F9.4020600@FreeBSD.org> References: <201105031854.p43IsIJ5006816@svn.freebsd.org> <4DF101F9.4020600@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thursday, June 09, 2011 1:25:13 pm Justin T. Gibbs wrote: > On 5/3/11 12:54 PM, Andrey V. Elsukov wrote: > > Author: ae > > Date: Tue May 3 18:54:18 2011 > > New Revision: 221397 > > URL: http://svn.freebsd.org/changeset/base/221397 > > > > Log: > > Add make_dev_alias_p() function. It is similar to make_dev_alias(), > > but it may return an error like make_dev_p() does. > > Sorry for the late reply on this change. I noticed it during some recent > merge activity at work where my make_dev_credv() conflicted with your > make_dev_cred_v(). > > It seems that the existing convention in this file for "function variants" > is to append a single letter to the function name without an underscore > (e.g. make_dev_credv, make_dev_credf). make_dev_alias_v() and make_dev_p(), > break this convention. It would be nice to clean up this API so that > the naming conventions are consistent. So which should it be? Underscore > or no underscore? Hmm, could we take advantage of 9 being an API "breaking point" to simplify the API back down so there are fewer variants perhaps? Maybe we could make the 'flags' argument to make_dev() mandatory and change make_dev and make_dev_alias to always use the _p semantics (return an error and use a parameter to optionally return the new cdev)? -- John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201106091440.36182.jhb>