Re: [PATCH] mm/page_alloc.c: fix a typo

[Li Zefan <lizf@cn.fujitsu.com>]

Li Zefan wrote:
>>> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
>>> index bdd5c43..d0ba10d 100644
>>> --- a/mm/page_alloc.c
>>> +++ b/mm/page_alloc.c
>>> @@ -303,7 +303,7 @@ static void destroy_compound_page(struct page *page, unsigned long order)
>>>  	for (i = 1; i < nr_pages; i++) {
>>>  		struct page *p = page + i;
>>>  
>>> -		if (unlikely(!PageTail(p) |
>>> +		if (unlikely(!PageTail(p) ||
>>>  				(p->first_page != page)))
>>>  			bad_page(page);
>>>  		__ClearPageTail(p);
>> I have a vague memory that the "|" was deliberate.  Most of the time,
>> "!PageTail" will be false so most of the time we won't take the first
> 
> !PageTail will be true if nothing bad happened, corrected me if I'm wrong:
> 

Silly, I was wrong...

Christoph Lameter wrote:
> I think the | there was some developers attempt to avoid gcc generating 
> too many branches. I am fine either way.
> 

Yes, I found out it's 224abf92b2f439a9030f21d2926ec8047d1ffcdb :

[PATCH] mm: bad_page optimisation
Nick Piggin [Fri, 6 Jan 2006 08:11:11 +0000 (00:11 -0800)]

Cut down size slightly by not passing bad_page the function name (it should be
able to be determined by dump_stack()).  And cut down the number of printks in
bad_page.

Also, cut down some branching in the destroy_compound_page path.