Re: sparse -Wptr-subtraction-blows: still needed?

[Dave Jones <davej@redhat.com>]

On Tue, May 01, 2007 at 02:43:30PM -0700, Linus Torvalds wrote:
 > 
 > 
 > On Tue, 1 May 2007, Josh Triplett wrote:
 > > 
 > > Does this still apply?  Do current versions of GCC still have this problem?
 > > If not, can the option and warning go away?
 > 
 > Even if current versions of gcc don't triple the build time (and for the 
 > kernel, I suspect it doesn't, because we've tried to clean up our header 
 > files), the generated _code_ will invariably suck.

FWIW, I do sparse runs on the fedora development kernels as part of
our daily builds now, and of the latest ones at
http://people.redhat.com/davej/kernels/Fedora/fc7/warnings.txt
(concatenated warning logs from i586/i686/x86_64/ppc/ppc64/s390 builds)
that 'expensive pointer subtraction' turns up 3705 times.
Interestingly, 1873 of those instances are from include/linux/mm.h
on the x86-64 build.

It's complaining about this line...

static __always_inline void *lowmem_page_address(struct page *page)
{
        return __va(page_to_pfn(page) << PAGE_SHIFT);
}

...

unsigned long page_to_pfn(struct page *page)
{
        return __page_to_pfn(page);
}

...

#define __page_to_pfn(page)     ((unsigned long)((page) - mem_map) + \
                                 ARCH_PFN_OFFSET)

looks like the other two variants of __page_to_pfn also use similar arithmatic.

	Dave

-- 
http://www.codemonkey.org.uk