rtems.h broken (original) (raw)

gcc/config/sh/rtems.h and gcc/config/sh/rtemself.h still use CPP_PREDEFINES.

The defines and asserts there-in cause building gcc to emit warnings, which later causes building libiberty to abort (For some reason incomprehensible to me, this uses -pedantic).

Replacing gcc/config/sh/rtems*.h with the versions from gcc-3.4/CVS-HEAD seem to fix this issue.

This is a regression against gcc-3.2, where target=sh-rtems* had build without problems.

Comment 3 Gabriel Dos Reis 2003-12-24 22:01:20 UTC

Joel -- Can you have a look at the patch mentioned in the audit trail?

Thanks,

Comment 4 joel.sherrill 2004-01-06 18:11:28 UTC

Subject: Re: [3.3 Regression] gcc/config/sh/rtems.h broken

gdr at gcc dot gnu dot org wrote:

------- Additional Comments From gdr at gcc dot gnu dot org 2003-12-24 22:01 ------- Joel -- Can you have a look at the patch mentioned in the audit trail?

This is exactly the same patch used with the RTEMS RPMs. Ralf is just trying to work through getting our patch set submitted before 3.4 forks.

So yes.. this patch is A-OK. :)

Sorry for the delay. I took a couple of weeks off for the holidays.

Thanks,

Comment 5 Gabriel Dos Reis 2004-01-06 21:14:12 UTC

Subject: Re: [3.3 Regression] gcc/config/sh/rtems.h broken

Joel Sherrill <joel.sherrill@OARcorp.com> writes:

| So yes.. this patch is A-OK. :)

Thank you!

-- Gaby

Comment 6 Nathanael C. Nerode 2004-01-09 22:19:03 UTC

Is someone committing the patch then, or should I? :-)

Comment 7 joel.sherrill 2004-01-09 22:45:07 UTC

Subject: Re: [3.3 Regression] gcc/config/sh/rtems.h broken

neroden at gcc dot gnu dot org wrote:

------- Additional Comments From neroden at gcc dot gnu dot org 2004-01-09 22:19 ------- Is someone committing the patch then, or should I? :-)

Please commit it.

Comment 8 Nathanael C. Nerode 2004-01-10 00:01:41 UTC

Committed patch, closing.