On 19 Dec 2016, at 19:58, Warner Losh <[email protected]> wrote: > > On Mon, Dec 19, 2016 at 1:39 AM, Ravi Pokala <[email protected]> wrote: >> -----Original Message----- >>> From: <[email protected]> on behalf of Sepherosa Ziehau >>> <[email protected]> >>> Date: 2016-12-18, Sunday at 23:02 >>> To: Dimitry Andric <[email protected]> >>> Cc: <[email protected]>, <[email protected]>, >>> <[email protected]> >>> Subject: Re: svn commit: r310171 - head/sys/sys >>> >>> The following patch unbreaks the LINT builds on amd64 for me after this >>> commit: >>> https://people.freebsd.org/~sephe/geom_sscanf.diff >> >> Wouldn't it be better to use the SCN macros? > > Are there other precedence for avoiding the SCN macros in the tree as > well, or is this new art?
I personally don't have anything against using the PRI or SCN macros, but traditionally there has been some backlash against it, if I recall correctly. It also requires including either <machine/_inttypes.h> or <inttypes.h>, depending on circumstance or preference. -Dimitry
signature.asc
Description: Message signed with OpenPGP using GPGMail
