On Thursday 13 October 2011 13:14:34 Joe Hershberger wrote:
On Wed, Oct 12, 2011 at 8:03 PM, Mike Frysinger vapier@gentoo.org wrote:
On Wednesday 12 October 2011 20:56:09 Joe Hershberger wrote:
WARNING:NEW_TYPEDEFS: do not add new typedefs This seems rather limiting... I'm not sure why even Linux would want this, at least when it applies to typedefs of structs. It makes sense if it's a new typedef for int or something.
Linux has been discouraging new struct typedefs. but it is too broad and catches typedefs that we want (like posix_types.h).
So it's a candidate to ignore, right?
it's like the volatiles check. reviewed on a case by case basis.
WARNING:VOLATILE: Use of volatile is usually wrong: see Documentation/volatile-considered-harmful.txt Sometimes using volatile is correct... not sure how this fits in with a policy of 0 errors and 0 warnings... Should it be ignored or not?
"it depends". we'll have to see the specific instance.
That's what I was getting at... should we ignore it in checkpatch and simply review any new volatile variables?
if checkpatch ignores it, people won't notice
Or should we not ignore it in checkpatch and have exceptions to "no errors, no warnings"?
i've always advocated not using checkpatch this way. it's simply another tool which has known bugs and doesn't cover realistic nuances. -mike