Re: [PATCH 1/4] mm/memory-hotplug: fix lowmem count overflow when offline pages

[Michal Hocko <mhocko@suse.cz>]

On Wed 22-05-13 12:49:37, Michal Hocko wrote:
> On Wed 22-05-13 17:29:27, Wanpeng Li wrote:
> > Logic memory-remove code fails to correctly account the Total High Memory 
> > when a memory block which contains High Memory is offlined as shown in the
> > example below. The following patch fixes it.
> > 
> > cat /proc/meminfo 
> > MemTotal:        7079452 kB
> > MemFree:         5805976 kB
> > Buffers:           94372 kB
> > Cached:           872000 kB
> > SwapCached:            0 kB
> > Active:           626936 kB
> > Inactive:         519236 kB
> > Active(anon):     180780 kB
> > Inactive(anon):   222944 kB
> > Active(file):     446156 kB
> > Inactive(file):   296292 kB
> > Unevictable:           0 kB
> > Mlocked:               0 kB
> > HighTotal:       7294672 kB
> > HighFree:        5181024 kB
> > LowTotal:       4294752076 kB
> > LowFree:          624952 kB
> 
> Ok, so the HighTotal is higher than MemTotal but it would have been more
> straightforward to show number of HighTotal before hotremove, show how
> much memory has been removed and the number after.
> 
> It is not clear which stable kernels need this fix as well.
> 
> > 
> > Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
> 
> Anyway
> Reviewed-by: Michal Hocko <mhocko@suse.cz>
> 
> with a nit pick bellow
> 
> > ---
> >  mm/page_alloc.c | 4 ++++
> >  1 file changed, 4 insertions(+)
> > 
> > diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> > index 98cbdf6..80474b2 100644
> > --- a/mm/page_alloc.c
> > +++ b/mm/page_alloc.c
> > @@ -6140,6 +6140,10 @@ __offline_isolated_pages(unsigned long start_pfn, unsigned long end_pfn)
> >  		list_del(&page->lru);
> >  		rmv_page_order(page);
> >  		zone->free_area[order].nr_free--;
> > +#ifdef CONFIG_HIGHMEM
> > +		if (PageHighMem(page))
> > +			totalhigh_pages -= 1 << order;
> > +#endif

Hmm, it seems that my old gcc (4.3.4) doesn't like this.
mm/page_alloc.c:6202: error: lvalue required as left operand of assignment

I am not sure this is a compiler bug because PAGEFLAG_FALSE is clearly
returning 0 which should make the whole if a no-op... But maybe we
shouldn't be more clever than compiler.

Andrew, could you fold this follow up fix for
mm-memory-hotplug-fix-lowmem-count-overflow-when-offline-pages.patch

Sorr about the confusion I brought to this.
---
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 717286e..c164e5f 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -6198,8 +6198,10 @@ __offline_isolated_pages(unsigned long start_pfn, unsigned long end_pfn)
 		list_del(&page->lru);
 		rmv_page_order(page);
 		zone->free_area[order].nr_free--;
+#ifdef CONFIG_HIGHMEM
 		if (PageHighMem(page))
 			totalhigh_pages -= 1 << order;
+#endif
 		for (i = 0; i < (1 << order); i++)
 			SetPageReserved((page+i));
 		pfn += (1 << order);

-- 
Michal Hocko
SUSE Labs

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>