Are we in agreement that this is the way to go?  Below is change I
gleaned from this email thread.

diff -r dc9b77ea48d4 usr/src/Makefile.master
--- a/usr/src/Makefile.master   Wed Jan 28 15:03:17 2009 -0600
+++ b/usr/src/Makefile.master   Thu Feb 05 10:14:22 2009 -0600
@@ -652,10 +652,11 @@
 ALWAYS_LINT_DEFS +=    -erroff=E_INTEGRAL_CONST_EXP_EXPECTED
 ALWAYS_LINT_DEFS +=    -erroff=E_PASS_INT_TO_SMALL_INT
 ALWAYS_LINT_DEFS +=    -erroff=E_PTR_CONV_LOSES_BITS
+LINT_ONLY_DEFS =       -I$(SPRO_VROOT)/prod/include/lint

 SECLEVEL=      core
-LINT.c=                $(LINT) $(LINTFLAGS) $(CPPFLAGS) $(ALWAYS_LINT_DEFS)
-LINT64.c=      $(LINT) $(LINTFLAGS64) $(CPPFLAGS) $(ALWAYS_LINT_DEFS)
+LINT.c=                $(LINT) $(LINT_ONLY_DEFS) $(LINTFLAGS)
$(CPPFLAGS) $(ALWAYS_LINT_DEFS)
+LINT64.c=      $(LINT) $(LINT_ONLY_DEFS) $(LINTFLAGS64) $(CPPFLAGS)
$(ALWAYS_LINT_DEFS)
 LINT.s=                $(LINT.c)

 # For some future builds, NATIVE_MACH and MACH might be different.

If this is acceptable, should this change to in against 4875943?  I've
never actually delivered into ON before, so I might need a bit of hand
holding on the official process  (unless of course one of you wanted to
deliver it :-) ).

Thanks,
Charles Binford

Darren J Moffat wrote:
> Mark J. Nelson wrote:
>>> Why wouldn't that work ?
>>
>> This was proposed in 4393089, and was one of my earlier suggestions. 
>> I think the general objections are that we're embedding private
>> knowledge of devpro internals.
>>
>> I know that we're only shadow compiling, not shadow linting, but I'm
>> still loathe to introduce product-specific info like this into our
>> command lines.
>
> My view is that this is acceptable given that lint only exists for
> devpro and we already have lint settings in Makefile.master that
> reference devpro lint bugs.
>

Reply via email to