Re: "introduce no new bootstrap warning" criteria. was: Loop ivdebugging patch

["Joseph S. Myers" <jsm28 at cam dot ac dot uk>]

On Fri, 12 Jan 2001, Robert Lipe wrote:

> Is it time, in the name of quality/damage control in this project, to
> make it an acceptance criteria for any commit that it introduce no new
> warnings?

 ... unless the point of the patch is to improve GCC's warnings, and this
causes previously undetected problems in GCC to be diagnosed.

What happened to the idea of a pragma to mark particular warnings as
expected in code - whether globally, scoped with push/pop, or for a
particular line of code only?  (With the preferred design from previous
discussions being, I think, a regular expression describing the expected
English language warning text, with an unresolved issue over whether the
expression should also be matched against the local language text.)  This
way, the spurious warnings that can't readily be fixed (e.g. macro strcmp
used without args on glibc systems - a -Wtraditional warning, but
presumably no system for which it is relevant has the strcmp function as a
macro) can be removed from the build.

See e.g. <URL:http://gcc.gnu.org/ml/gcc/2000-06/msg00639.html>.

-- 
Joseph S. Myers
jsm28@cam.ac.uk