Re: [patch 06/26] mm/memory_hotplug: don't access uninitialized memmaps in shrink_pgdat_span()

[Michal Hocko]

Has this been reviewed properly? I do not see any Acks nor Reviewed-bys.
Did Aneesh gave it some testing?

On Fri 18-10-19 20:19:33, Andrew Morton wrote:
> From: David Hildenbrand <david@xxxxxxxxxx>
> Subject: mm/memory_hotplug: don't access uninitialized memmaps in shrink_pgdat_span()
> 
> We might use the nid of memmaps that were never initialized.  For example,
> if the memmap was poisoned, we will crash the kernel in pfn_to_nid() right
> now.  Let's use the calculated boundaries of the separate zones instead. 
> This now also avoids having to iterate over a whole bunch of subsections
> again, after shrinking one zone.
> 
> Before commit d0dc12e86b31 ("mm/memory_hotplug: optimize memory hotplug"),
> the memmap was initialized to 0 and the node was set to the right value. 
> After that commit, the node might be garbage.
> 
> We'll have to fix shrink_zone_span() next.
> 
> Link: http://lkml.kernel.org/r/20191006085646.5768-4-david@xxxxxxxxxx
> Fixes: f1dd2cd13c4b ("mm, memory_hotplug: do not associate hotadded memory to zones until online")	[d0dc12e86b319]
> Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>
> Reported-by: Aneesh Kumar K.V <aneesh.kumar@xxxxxxxxxxxxx>
> Cc: Oscar Salvador <osalvador@xxxxxxx>
> Cc: David Hildenbrand <david@xxxxxxxxxx>
> Cc: Michal Hocko <mhocko@xxxxxxxx>
> Cc: Pavel Tatashin <pasha.tatashin@xxxxxxxxxx>
> Cc: Dan Williams <dan.j.williams@xxxxxxxxx>
> Cc: Wei Yang <richardw.yang@xxxxxxxxxxxxxxx>
> Cc: Alexander Duyck <alexander.h.duyck@xxxxxxxxxxxxxxx>
> Cc: Alexander Potapenko <glider@xxxxxxxxxx>
> Cc: Andy Lutomirski <luto@xxxxxxxxxx>
> Cc: Anshuman Khandual <anshuman.khandual@xxxxxxx>
> Cc: Benjamin Herrenschmidt <benh@xxxxxxxxxxxxxxxxxxx>
> Cc: Borislav Petkov <bp@xxxxxxxxx>
> Cc: Catalin Marinas <catalin.marinas@xxxxxxx>
> Cc: Christian Borntraeger <borntraeger@xxxxxxxxxx>
> Cc: Christophe Leroy <christophe.leroy@xxxxxx>
> Cc: Damian Tometzki <damian.tometzki@xxxxxxxxx>
> Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
> Cc: Fenghua Yu <fenghua.yu@xxxxxxxxx>
> Cc: Gerald Schaefer <gerald.schaefer@xxxxxxxxxx>
> Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
> Cc: Halil Pasic <pasic@xxxxxxxxxxxxx>
> Cc: Heiko Carstens <heiko.carstens@xxxxxxxxxx>
> Cc: "H. Peter Anvin" <hpa@xxxxxxxxx>
> Cc: Ingo Molnar <mingo@xxxxxxxxxx>
> Cc: Ira Weiny <ira.weiny@xxxxxxxxx>
> Cc: Jason Gunthorpe <jgg@xxxxxxxx>
> Cc: Jun Yao <yaojun8558363@xxxxxxxxx>
> Cc: Logan Gunthorpe <logang@xxxxxxxxxxxx>
> Cc: Mark Rutland <mark.rutland@xxxxxxx>
> Cc: Masahiro Yamada <yamada.masahiro@xxxxxxxxxxxxx>
> Cc: "Matthew Wilcox (Oracle)" <willy@xxxxxxxxxxxxx>
> Cc: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
> Cc: Michael Ellerman <mpe@xxxxxxxxxxxxxx>
> Cc: Mike Rapoport <rppt@xxxxxxxxxxxxx>
> Cc: Pankaj Gupta <pagupta@xxxxxxxxxx>
> Cc: Paul Mackerras <paulus@xxxxxxxxx>
> Cc: Pavel Tatashin <pavel.tatashin@xxxxxxxxxxxxx>
> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Cc: Qian Cai <cai@xxxxxx>
> Cc: Rich Felker <dalias@xxxxxxxx>
> Cc: Robin Murphy <robin.murphy@xxxxxxx>
> Cc: Steve Capper <steve.capper@xxxxxxx>
> Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> Cc: Tom Lendacky <thomas.lendacky@xxxxxxx>
> Cc: Tony Luck <tony.luck@xxxxxxxxx>
> Cc: Vasily Gorbik <gor@xxxxxxxxxxxxx>
> Cc: Vlastimil Babka <vbabka@xxxxxxx>
> Cc: Wei Yang <richard.weiyang@xxxxxxxxx>
> Cc: Will Deacon <will@xxxxxxxxxx>
> Cc: Yoshinori Sato <ysato@xxxxxxxxxxxxxxxxxxxx>
> Cc: Yu Zhao <yuzhao@xxxxxxxxxx>
> Cc: <stable@xxxxxxxxxxxxxxx>	[4.13+]
> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> ---
> 
>  mm/memory_hotplug.c |   74 +++++++++---------------------------------
>  1 file changed, 16 insertions(+), 58 deletions(-)
> 
> --- a/mm/memory_hotplug.c~mm-memory_hotplug-dont-access-uninitialized-memmaps-in-shrink_pgdat_span
> +++ a/mm/memory_hotplug.c
> @@ -436,67 +436,25 @@ static void shrink_zone_span(struct zone
>  	zone_span_writeunlock(zone);
>  }
>  
> -static void shrink_pgdat_span(struct pglist_data *pgdat,
> -			      unsigned long start_pfn, unsigned long end_pfn)
> +static void update_pgdat_span(struct pglist_data *pgdat)
>  {
> -	unsigned long pgdat_start_pfn = pgdat->node_start_pfn;
> -	unsigned long p = pgdat_end_pfn(pgdat); /* pgdat_end_pfn namespace clash */
> -	unsigned long pgdat_end_pfn = p;
> -	unsigned long pfn;
> -	int nid = pgdat->node_id;
> -
> -	if (pgdat_start_pfn == start_pfn) {
> -		/*
> -		 * If the section is smallest section in the pgdat, it need
> -		 * shrink pgdat->node_start_pfn and pgdat->node_spanned_pages.
> -		 * In this case, we find second smallest valid mem_section
> -		 * for shrinking zone.
> -		 */
> -		pfn = find_smallest_section_pfn(nid, NULL, end_pfn,
> -						pgdat_end_pfn);
> -		if (pfn) {
> -			pgdat->node_start_pfn = pfn;
> -			pgdat->node_spanned_pages = pgdat_end_pfn - pfn;
> -		}
> -	} else if (pgdat_end_pfn == end_pfn) {
> -		/*
> -		 * If the section is biggest section in the pgdat, it need
> -		 * shrink pgdat->node_spanned_pages.
> -		 * In this case, we find second biggest valid mem_section for
> -		 * shrinking zone.
> -		 */
> -		pfn = find_biggest_section_pfn(nid, NULL, pgdat_start_pfn,
> -					       start_pfn);
> -		if (pfn)
> -			pgdat->node_spanned_pages = pfn - pgdat_start_pfn + 1;
> -	}
> -
> -	/*
> -	 * If the section is not biggest or smallest mem_section in the pgdat,
> -	 * it only creates a hole in the pgdat. So in this case, we need not
> -	 * change the pgdat.
> -	 * But perhaps, the pgdat has only hole data. Thus it check the pgdat
> -	 * has only hole or not.
> -	 */
> -	pfn = pgdat_start_pfn;
> -	for (; pfn < pgdat_end_pfn; pfn += PAGES_PER_SUBSECTION) {
> -		if (unlikely(!pfn_valid(pfn)))
> -			continue;
> -
> -		if (pfn_to_nid(pfn) != nid)
> -			continue;
> -
> -		/* Skip range to be removed */
> -		if (pfn >= start_pfn && pfn < end_pfn)
> -			continue;
> +	unsigned long node_start_pfn = 0, node_end_pfn = 0;
> +	struct zone *zone;
>  
> -		/* If we find valid section, we have nothing to do */
> -		return;
> +	for (zone = pgdat->node_zones;
> +	     zone < pgdat->node_zones + MAX_NR_ZONES; zone++) {
> +		unsigned long zone_end_pfn = zone->zone_start_pfn +
> +					     zone->spanned_pages;
> +
> +		/* No need to lock the zones, they can't change. */
> +		if (zone_end_pfn > node_end_pfn)
> +			node_end_pfn = zone_end_pfn;
> +		if (zone->zone_start_pfn < node_start_pfn)
> +			node_start_pfn = zone->zone_start_pfn;
>  	}
>  
> -	/* The pgdat has no valid section */
> -	pgdat->node_start_pfn = 0;
> -	pgdat->node_spanned_pages = 0;
> +	pgdat->node_start_pfn = node_start_pfn;
> +	pgdat->node_spanned_pages = node_end_pfn - node_start_pfn;
>  }
>  
>  static void __remove_zone(struct zone *zone, unsigned long start_pfn,
> @@ -507,7 +465,7 @@ static void __remove_zone(struct zone *z
>  
>  	pgdat_resize_lock(zone->zone_pgdat, &flags);
>  	shrink_zone_span(zone, start_pfn, start_pfn + nr_pages);
> -	shrink_pgdat_span(pgdat, start_pfn, start_pfn + nr_pages);
> +	update_pgdat_span(pgdat);
>  	pgdat_resize_unlock(zone->zone_pgdat, &flags);
>  }
>  
> _

-- 
Michal Hocko
SUSE Labs