Re: [lmi] clang-tidy [Was: Detecting whether move semantics actually take place]

[Vadim Zeitlin]

On Wed, 20 Jul 2022 23:03:34 +0000 Greg Chicares <gchicares@sbcglobal.net> 
wrote:

GC> On 7/18/22 16:19, Vadim Zeitlin wrote:
[...]
GC> >  For testing on a single file I run it like this:
GC> > 
GC> >   % clang-tidy --config-file=.clang-tidy zero_test.cpp -- clang++-14 
...all the flags taken from "make -n" output ...
GC> 
GC> I have only 'clang++13', not 'clang++14'. Is there any reason for me to
GC> specify 'clang++13', or would 'clang' be as good when only one version
GC> is installed?

 Just clang is perfectly fine if you have it -- that is, if you've
installed the clang meta-package which depends on some concrete version of
clang. If you explicitly install clang-14 package, as I did, because I
don't want my clang to silently change to clang-15 without me noticing,
then you simply don't have clang, while you always have clang-14 (or -13,
or whatever).

 But, just a warning: if you don't install clang package, some other things
break and need to be fixed manually (I don't remember the details, but I
had to specify some paths explicitly because the default locations, with
"clang" in their name, where not symlinked to clang-14-specific ones), so
if you want to have as few problems as possible, it is almost certainly
better to install it and not just the version-specific package.

 Regards,
VZ

pgpD81atoMwIm.pgp
Description: PGP signature