On Tue, Sep 01, 2026 at 06:06:40PM +0200, Stefano Garzarella wrote: > On Tue, Sep 01, 2026 at 05:29:46PM +0300, Jarkko Sakkinen wrote: > > In all the pre-existing call sites both @iomem and @iobase_ptr are > > either NULL or non-NULL. > > I don't know this code, but I'm a bit worried about iobase_ptr and > *iobase_ptr. IIUC it is true that iobase_ptr and iores are either NULL or > non-NULL, but here we are removing the case where *iobase_ptr is NULL. > > Now looking at crb_map_io(), IIUC iobase_array is initialized with NULL > pointers and the code we are removing was the only one initializing those > pointers IIUC, or am I missing something?
crb_map_io() sets both to non-NULL value, or leaves both as NULL. crb_map_pluton() explicitly calls both explicitly with NULL. If anything else will arrive too crb_map_res, that'd be unexpected input, which without this patch will go unnoticed and will lead to undefined behavior. Not sure what is the argument here really. > > Thanks, > Stefano BR, Jarkko
