https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120929
--- Comment #19 from Siddhesh Poyarekar <siddhesh at gcc dot gnu.org> --- (In reply to Siddhesh Poyarekar from comment #17) > (In reply to qinzhao from comment #16) > > I feel that propagating the size through _5 = *_1 might not be correct in > > general, we should only limit it to the case when the RHS is a pointer > > defined by .ACCESS_WITH_SIZE? > > That propagation is not incorrect, but it is incomplete, since (TIL) a > MEM_REF with zero offset could be the LHS of a valid gimple statement as > well. However AFAICT there's no way to directly reach it like in case of an > SSA, you'll need to walk the function to find the statement with a matching > LHS. Actually, I think you may be right; I've been turning this over in my head and I can't think of any other situation where the propagation is correct, i.e. other than the .ACCESS_WITH_SIZE case. In fact, maybe the .ACCESS_WITH_SIZE handling in objsz probably needs improvement to express it better, but that's an orthogonal matter. So it should in fact be just B as far as this issue is concerned. That is, for a _n = *_p case, only look at the size of _p if it is a .ACCESS_WITH_SIZE. There is a use case for A where we should be looking through vdefs to find memory writes to *_p that allocate memory, but it's an orthogonal case. I've filed it as bug 120947 now.
