Ruslan Ermilov wrote:
> On Fri, May 10, 2002 at 04:41:53PM -0700, David O'Brien wrote:
> > On Fri, May 10, 2002 at 06:04:27PM +0300, Ruslan Ermilov wrote:
> > > > Bmake bits for Gcc 3.1.
> > > =20
> > > This also vanished my YACC building fixes and broke world while
> > > attempting to build `cc1plus' in a cross-tools stage. The changes
> > > below fix this and CLEANFILES.
> >=20
> > These changes are wrong.
> > =20
> These changes have been verified to produce the same binaries
> (except for cc1plus which is just broken without them), and
> they are equivalent to what was already in -CURRENT before
> your WIP_GCC31 merge commit.
I dont think you've tested them properly. They will not work.
> > > RCS file: /home/ncvs/src/gnu/usr.bin/cc/cc1/Makefile,v
> > ...
> > > -c-parse.c: c-parse.in
> > > +c-parse.y: c-parse.in
> > > sed -e "/^ifobjc$$/,/^end ifobjc$$/d" \
> > > -e "/^ifc$$/d" \
> > > -e "/^end ifc$$/d" \
> > > - ${GCCDIR}/c-parse.in > c-parse.y
> > > - ${YACC} -o c-parse.c.in c-parse.y
> > > - sed -e "s/malloc/xmalloc/g" \
> > > + -e "s/malloc/xmalloc/g" \
> > > -e "s/realloc/xrealloc/g" \
> > > - c-parse.c.in >c-parse.c
> > > + ${.ALLSRC} > ${.TARGET}
> >=20
> > The malloc usage is in the Byacc output, not the input.
> >=20
> There's no difference, [b]yacc just copies C code blocks intact.
Do not forget the driver code. There are malloc/calloc calls in the byacc
generated c-parse.c.. There are no malloc/calloc calls in the c-parse.in
file (see for yourself). If you actually test this, you will get a failure
due to 'poisoned' malloc/calloc/etc calls.
Cheers,
-Peter
--
Peter Wemm - [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED]
"All of this is for nothing if we don't go to the stars" - JMS/B5
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
