Re: [Qemu-devel] [PATCH 1/3] checkpatch: tweak the files in which TABs are checked

[Cornelia Huck]

On Wed, 10 Aug 2016 08:46:07 +0200
Markus Armbruster <address@hidden> wrote:

> I'm afraid "if not then tabs are allowed" is confusing.  We're obviously
> skipping more than just the tabs check: RCS/CVS revision markers, and a
> whole bunch of C style checks.  Makes sense, we don't want to do these
> checks for imported files.  "if not then tabs are allowed" starts to
> make some sense only once you've stared at the next "next if ..." line
> for a while.  Let's avoid that.  Minimal change:
> 
>    # check we are in a valid source file; if not then ignore this hunk
>    # make an exception from some scripts imported from other projects.
> 
> Radim's "[PATCH] checkpatch: ignore automatically imported Linux
> headers" adds a similar exception for other imported files in a
> different place:
> 
>   @@ -1312,6 +1312,9 @@ sub process {
>    # ignore non-hunk lines and lines being removed
>                   next if (!$hunk_line || $line =~ /^-/);
> 
>   +# ignore files that are being periodically imported from Linux
>   +           next if ($realfile =~ 
> /^(linux-headers|include\/standard-headers)\//);
>   +
>    #trailing whitespace
>                   if ($line =~ /^\+.*\015/) {
>                           my $herevet = "$here\n" . cat_vet($rawline) . "\n";
> 
> Should both exceptions be in the same place?

There are two cases of 'imported file':
(a) Things like the headers update where we want to copy whatever we
got from the kernel: If coding style is already messed up there, we
still want to copy.
(b) Things like checkpatch.pl which we tweak ourselves: We want to keep
close to the existing coding style, but don't want to mess up things
further.

I think in the long run it would make sense to skip any checks for case
(a) but still keep a subset of checks (like trailing whitespace) for
(b). For the sake of silencing the checkpatch bot, just skipping (a)
and (b) makes sense for now.