On Tue, Mar 07, 2017 at 11:11:20PM +0800, Christopher Li wrote: > On Sun, Mar 5, 2017 at 7:20 PM, Luc Van Oostenryck > <luc.vanoostenryck@xxxxxxxxx> wrote: > > In sparse-llvm there is the assumption that a PSEUDO_VAL is always > > of integer type. But this is not always the case: constant pointers, > > like NULL, are also of the PSEUDO_VAL kind. > > After apply this patch, the test suite has three more new test failure. > A closer look there are assert failed in three of the test cases. > > error: actual error text does not match expected error text. > error: see backend/loop2.c.error.* for further investigation. > --- backend/loop2.c.error.expected 2017-03-07 23:06:42.298291232 +0800 > +++ backend/loop2.c.error.got 2017-03-07 23:06:42.460291747 +0800 > +sparse-llvm: sparse-llvm.c:312: val_to_value: Assertion `ctype' failed. > +.././sparsec: line 35: 8495 Aborted (core dumped) > $DIRNAME/sparse-llvm $SPARSEOPTS > $TMPLLVM Mmmm, strange but it's very possible that I didn't tested patch by patch. I'll look at it. > Also the previous 1/13 patch will generate warning without this path. > > sparse-llvm.c:306:21: warning: ‘val_to_value’ defined but not used > [-Wunused-function] > static LLVMValueRef val_to_value(struct function *fn, unsigned long > long val, struct symbol *ctype) > ^~~~~~~~~~~~ > We should fix the warning. Maybe combine the previous patch with this one? Yes, if you prefer so. Generally I prefer to separate the introduction of new code and its use but it doesn't really matter much. -- Luc -- To unsubscribe from this list: send the line "unsubscribe linux-sparse" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
