linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Mike Rapoport <rppt@linux.ibm.com>
To: Michal Hocko <mhocko@kernel.org>
Cc: Rong Chen <rong.a.chen@intel.com>,
	Pavel Tatashin <pasha.tatashin@soleen.com>,
	linux-kernel@vger.kernel.org,
	Linux Memory Management List <linux-mm@kvack.org>,
	Andrew Morton <akpm@linux-foundation.org>, LKP <lkp@01.org>,
	Oscar Salvador <osalvador@suse.de>
Subject: Re: [LKP] efad4e475c [ 40.308255] Oops: 0000 [#1] PREEMPT SMP PTI
Date: Mon, 18 Feb 2019 19:48:29 +0200	[thread overview]
Message-ID: <20190218174828.GH25446@rapoport-lnx> (raw)
In-Reply-To: <20190218170558.GV4525@dhcp22.suse.cz>

On Mon, Feb 18, 2019 at 06:05:58PM +0100, Michal Hocko wrote:
> On Mon 18-02-19 18:48:14, Mike Rapoport wrote:
> > On Mon, Feb 18, 2019 at 04:22:13PM +0100, Michal Hocko wrote:
> [...]
> > > Thinking about it some more, is it possible that we are overflowing by 1
> > > here?
> > 
> > Looks like that, the end_pfn is actually the first pfn in the next section.
> 
> Thanks for the confirmation. I guess it also exaplains why nobody has
> noticed this off-by-one. Most people seem to use VMEMMAP SPARSE model
> and we are safe there.
> 
> > > diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> > > index 124e794867c5..6618b9d3e53a 100644
> > > --- a/mm/memory_hotplug.c
> > > +++ b/mm/memory_hotplug.c
> > > @@ -1234,10 +1234,10 @@ bool is_mem_section_removable(unsigned long start_pfn, unsigned long nr_pages)
> > >  {
> > >  	struct page *page = pfn_to_page(start_pfn);
> > >  	unsigned long end_pfn = min(start_pfn + nr_pages, zone_end_pfn(page_zone(page)));
> > > -	struct page *end_page = pfn_to_page(end_pfn);
> > > +	struct page *end_page = pfn_to_page(end_pfn - 1);
> > >  
> > >  	/* Check the starting page of each pageblock within the range */
> > > -	for (; page < end_page; page = next_active_pageblock(page)) {
> > > +	for (; page <= end_page; page = next_active_pageblock(page)) {
> > >  		if (!is_pageblock_removable_nolock(page))
> > >  			return false;
> > >  		cond_resched();
> > 
> > Works with your fix, but I think mine is more intuitive ;-)
> 
> I would rather go and rework this to pfns. What about this instead.
> Slightly larger but arguably cleared code?

Yeah, this is clearer.
 
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index 124e794867c5..a799a0bdbf34 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -1188,11 +1188,13 @@ static inline int pageblock_free(struct page *page)
>  	return PageBuddy(page) && page_order(page) >= pageblock_order;
>  }
>  
> -/* Return the start of the next active pageblock after a given page */
> -static struct page *next_active_pageblock(struct page *page)
> +/* Return the pfn of the start of the next active pageblock after a given pfn */
> +static unsigned long next_active_pageblock(unsigned long pfn)
>  {
> +	struct page *page = pfn_to_page(pfn);
> +
>  	/* Ensure the starting page is pageblock-aligned */
> -	BUG_ON(page_to_pfn(page) & (pageblock_nr_pages - 1));
> +	BUG_ON(pfn & (pageblock_nr_pages - 1));
>  
>  	/* If the entire pageblock is free, move to the end of free page */
>  	if (pageblock_free(page)) {
> @@ -1200,16 +1202,16 @@ static struct page *next_active_pageblock(struct page *page)
>  		/* be careful. we don't have locks, page_order can be changed.*/
>  		order = page_order(page);
>  		if ((order < MAX_ORDER) && (order >= pageblock_order))
> -			return page + (1 << order);
> +			return pfn + (1 << order);
>  	}
>  
> -	return page + pageblock_nr_pages;
> +	return pfn + pageblock_nr_pages;
>  }
>  
> -static bool is_pageblock_removable_nolock(struct page *page)
> +static bool is_pageblock_removable_nolock(unsigned long pfn)
>  {
> +	struct page *page = pfn_to_page(pfn);
>  	struct zone *zone;
> -	unsigned long pfn;
>  
>  	/*
>  	 * We have to be careful here because we are iterating over memory
> @@ -1232,13 +1234,14 @@ static bool is_pageblock_removable_nolock(struct page *page)
>  /* Checks if this range of memory is likely to be hot-removable. */
>  bool is_mem_section_removable(unsigned long start_pfn, unsigned long nr_pages)
>  {
> -	struct page *page = pfn_to_page(start_pfn);
> -	unsigned long end_pfn = min(start_pfn + nr_pages, zone_end_pfn(page_zone(page)));
> -	struct page *end_page = pfn_to_page(end_pfn);
> +	unsigned long end_pfn;
> +
> +	end_pfn = min(start_pfn + nr_pages,
> +			zone_end_pfn(page_zone(pfn_to_page(start_pfn))));
>  
>  	/* Check the starting page of each pageblock within the range */
> -	for (; page < end_page; page = next_active_pageblock(page)) {
> -		if (!is_pageblock_removable_nolock(page))
> +	for (; start_pfn < end_pfn; start_pfn = next_active_pageblock(start_pfn)) {
> +		if (!is_pageblock_removable_nolock(start_pfn))
>  			return false;
>  		cond_resched();
>  	}

With this on top the loop even fits into 80-chars ;-)

diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index 9cc42f3..9981ca7 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -1234,13 +1234,13 @@ static bool is_pageblock_removable_nolock(unsigned long pfn)
 /* Checks if this range of memory is likely to be hot-removable. */
 bool is_mem_section_removable(unsigned long start_pfn, unsigned long nr_pages)
 {
-	unsigned long end_pfn;
+	unsigned long end_pfn, pfn;
 
 	end_pfn = min(start_pfn + nr_pages,
 			zone_end_pfn(page_zone(pfn_to_page(start_pfn))));
 
 	/* Check the starting page of each pageblock within the range */
-	for (; start_pfn < end_pfn; start_pfn = next_active_pageblock(start_pfn)) {
+	for (pfn = start_pfn; pfn < end_pfn; pfn = next_active_pageblock(pfn)) {
 		if (!is_pageblock_removable_nolock(start_pfn))
 			return false;
 		cond_resched();

-- 
Sincerely yours,
Mike.


  reply	other threads:[~2019-02-18 17:51 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-02-18  5:28 [LKP] efad4e475c [ 40.308255] Oops: 0000 [#1] PREEMPT SMP PTI kernel test robot
2019-02-18  7:08 ` Michal Hocko
2019-02-18  8:47   ` Rong Chen
2019-02-18  9:03     ` Michal Hocko
2019-02-18  9:11       ` Rong Chen
2019-02-18  9:29         ` Michal Hocko
2019-02-18  8:55   ` Michal Hocko
     [not found]     ` <4c75d424-2c51-0d7d-5c28-78c15600e93c@intel.com>
2019-02-18 10:30       ` Michal Hocko
2019-02-18 14:05         ` Mike Rapoport
2019-02-18 15:20           ` Michal Hocko
2019-02-18 15:22             ` Michal Hocko
2019-02-18 16:48               ` Mike Rapoport
2019-02-18 17:05                 ` Michal Hocko
2019-02-18 17:48                   ` Mike Rapoport [this message]
2019-02-18 17:57                   ` Matthew Wilcox
2019-02-18 18:11                     ` Michal Hocko
2019-02-18 19:05                       ` Matthew Wilcox
2019-02-18 18:15 ` [RFC PATCH] mm, memory_hotplug: fix off-by-one in is_pageblock_removable Michal Hocko
2019-02-18 18:31   ` Mike Rapoport
2019-02-20  8:33   ` Oscar Salvador
2019-02-20 12:57   ` Michal Hocko
2019-02-21  3:18     ` [LKP] " Rong Chen
2019-02-21  7:25       ` Michal Hocko

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20190218174828.GH25446@rapoport-lnx \
    --to=rppt@linux.ibm.com \
    --cc=akpm@linux-foundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lkp@01.org \
    --cc=mhocko@kernel.org \
    --cc=osalvador@suse.de \
    --cc=pasha.tatashin@soleen.com \
    --cc=rong.a.chen@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).