[Bug 1526] 4.10 head has warnings we consider as errors.

[bugzilla-daemon at rtems.org]

https://www.rtems.org/bugzilla/show_bug.cgi?id=1526





--- Comment #8 from Ralf Corsepius <ralf.corsepius at rtems.org>  2010-06-16 23:13:05 ---
> I did not object because the builder is showing no fails due to errors. 
This is only a matter of warning levels and conventions ;)

> On the
> AVR I see a format warning but I was not trapping on these.
Which? URL? I am currently treating all warnings as serious, because they often
indicate some more serious issue underneath.

> Also this PR is for 4.10. Is the work you are doing relating to 4.10 or 4.11 ?
Both. Actually I am going after warnings I am observing when building HEAD for
multilib-enabled builds of sparc64 (64bit), the avr (16bit) and sparc (32bit).

> The 64bit issues should be only on 4.11.
In theory, yes. In practice, this doesn't apply because fixing those
occasionally involves API changes (cf. the in_addr_t changes from a couple of
days ago) 
and unhides (often harmless) hidden issues affecting both branch and HEAD.

As I consider not fixing known bugs to be dumb, esp. in a stage of development
when branch and HEAD only differ marginally, I am often applying fixes to both
branch and HEAD.

> Do we need another PR for 4.11 so this the 4.10 can be closed ?
I don't think so. To me the appropriate time to close this PR when all
gcc-options to trigger pedantic warnings will be removed from the branch.

-- 
Configure bugmail: https://www.rtems.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.