Messages in this thread | | | Date | Thu, 19 Aug 1999 12:57:20 -0700 (PDT) | From | Linus Torvalds <> | Subject | Re: [PATCH] 2.3.14: bug-fix for raw IO error recovery |
| |
On Thu, 19 Aug 1999, Andrea Arcangeli wrote: > > Wouldn't be possible to let get_page_map to give us back more information > by returning -1 if the page_map is not good?
Don't do this. It leads to all kinds of problems for 64-bit architectures, I'm not at all sure that casting (int) -1 to a pointer is at all safe. I would not be surprised if it is entirely legal and expected on some architectures for it to become 0x00000000ffffffff, which might be a perfectly valid pointer.
If you play error games with pointers, there's just two ways to do it nicely: either the NULL pointer (right now) or by having a nicely abstracted thing that can be extended to do different things on different architectures (this is what the "dcache" error pointers are doing with the ERR_PTR/PTR_ERR/IS_ERR stuff - right now all architectures have pointers that look roughly the same and there is just one global definition for ERR_PTR, but if some strange architecture comes along some day we'll be able to handle it cleanly).
In the meantime, things like this are _definitely_ a bad idea.
Linus
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |