This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Problem with nested functions, aliasing, ssa rewriting, etc


In message <20040310215246.GA2798@redhat.com>, Richard Henderson writes:
 >On Wed, Mar 10, 2004 at 11:44:18AM -0700, law@redhat.com wrote:
 >>   CHAIN.1_2->p = &CHAIN.1->buffer;
 >
 >The problem here is that this used to be "&buffer", which
 >had the TREE_INVARANT bit set.  Failing to clear that meant
 >that we incorrectly treated the entire rhs as min_invariant.
 >
 >Fixed thus.  I'll put this through the testsuite momentarily.
OK.

 >Also noticed a problem in note_addressable where we'd consider
 >
 >	&x->y
 >
 >to address X, when this is in fact a normal use.  Any use of
 >get_base_decl should be viewed with suspicion...
Funny.  I had noticed that as well while looking at this problem.  I had
put it on my todo list since it didn't appear to directly influence the
correctness of the testcase I was looking at.

jeff


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]