Messages in this thread | | | Date | Fri, 27 Dec 2013 14:26:05 -0800 | From | Andrew Morton <> | Subject | Re: [PATCH] mm: dump page when hitting a VM_BUG_ON using VM_BUG_ON_PAGE |
| |
On Fri, 27 Dec 2013 17:19:44 -0500 Sasha Levin <sasha.levin@oracle.com> wrote:
> On 12/27/2013 05:38 AM, Kirill A. Shutemov wrote: > > On Thu, Dec 26, 2013 at 10:20:52PM -0500, Sasha Levin wrote: > >> Most of the VM_BUG_ON assertions are performed on a page. Usually, when > >> one of these assertions fails we'll get a BUG_ON with a call stack and > >> the registers. > >> > >> I've recently noticed based on the requests to add a small piece of code > >> that dumps the page to various VM_BUG_ON sites that the page dump is quite > >> useful to people debugging issues in mm. > >> > >> This patch adds a VM_BUG_ON_PAGE(cond, page) which beyond doing what > >> VM_BUG_ON() does, also dumps the page before executing the actual BUG_ON. > >> > >> Signed-off-by: Sasha Levin <sasha.levin@oracle.com> > > > > I like the idea. One thing I've noticed you have a lot of page flag based > > asserts, like: > > > > VM_BUG_ON_PAGE(PageLRU(page), page); > > VM_BUG_ON_PAGE(!PageLocked(page), page); > > > > What about adding per-page-flag assert macros, like: > > > > PageNotLRU_assert(page); > > PageLocked_assert(page); > > > > ? This way we will always dump right page on bug. > > > > Sure, sounds good. > > I'll send another patch on top of this one.
I think I prefer the patch as-is. To do what Kirill suggests we'd have to add a heck of a lot more macros and they add little value. And the suggested names of those macros aren't very good - there's nothing in "PageNotLRU_assert" which tells the reader that this code is conditional on CONFIG_DEBUG_VM, which is a somewhat important thing.
| |