* [-next] lots of messages due to "mm, memory_hotplug: be more verbose for memory offline failures"
@ 2018-12-17 15:59 Heiko Carstens
2018-12-17 16:03 ` Michal Hocko
0 siblings, 1 reply; 5+ messages in thread
From: Heiko Carstens @ 2018-12-17 15:59 UTC (permalink / raw)
To: Michal Hocko
Cc: Andrew Morton, Oscar Salvador, Anshuman Khandual,
Stephen Rothwell, linux-mm, linux-kernel, linux-next, linux-s390
Hi Michal,
with linux-next as of today on s390 I see tons of messages like
[ 20.536664] page dumped because: has_unmovable_pages
[ 20.536792] page:000003d081ff4080 count:1 mapcount:0 mapping:000000008ff88600 index:0x0 compound_mapcount: 0
[ 20.536794] flags: 0x3fffe0000010200(slab|head)
[ 20.536795] raw: 03fffe0000010200 0000000000000100 0000000000000200 000000008ff88600
[ 20.536796] raw: 0000000000000000 0020004100000000 ffffffff00000001 0000000000000000
[ 20.536797] page dumped because: has_unmovable_pages
[ 20.536814] page:000003d0823b0000 count:1 mapcount:0 mapping:0000000000000000 index:0x0
[ 20.536815] flags: 0x7fffe0000000000()
[ 20.536817] raw: 07fffe0000000000 0000000000000100 0000000000000200 0000000000000000
[ 20.536818] raw: 0000000000000000 0000000000000000 ffffffff00000001 0000000000000000
bisect points to b323c049a999 ("mm, memory_hotplug: be more verbose for memory offline failures")
which is the first commit with which the messages appear.
Note: there is _no_ memory hotplug involved when these messages appear.
I don't know if it helps, but this is the contents of /proc/zoneinfo:
Node 0, zone DMA
per-node stats
nr_inactive_anon 8
nr_active_anon 8389
nr_inactive_file 43418
nr_active_file 22655
nr_unevictable 0
nr_slab_reclaimable 8192
nr_slab_unreclaimable 11368
nr_isolated_anon 0
nr_isolated_file 0
workingset_nodes 0
workingset_refault 0
workingset_activate 0
workingset_restore 0
workingset_nodereclaim 0
nr_anon_pages 7088
nr_mapped 16328
nr_file_pages 66132
nr_dirty 0
nr_writeback 0
nr_writeback_temp 0
nr_shmem 55
nr_shmem_hugepages 0
nr_shmem_pmdmapped 0
nr_anon_transparent_hugepages 4
nr_unstable 0
nr_vmscan_write 0
nr_vmscan_immediate_reclaim 0
nr_dirtied 20723
nr_written 18227
nr_kernel_misc_reclaimable 0
pages free 519834
min 1899
low 2419
high 2939
spanned 524288
present 524288
managed 520562
protection: (0, 3988, 3988)
nr_free_pages 519834
nr_zone_inactive_anon 0
nr_zone_active_anon 0
nr_zone_inactive_file 0
nr_zone_active_file 0
nr_zone_unevictable 0
nr_zone_write_pending 0
nr_mlock 0
nr_page_table_pages 0
nr_kernel_stack 0
nr_bounce 0
nr_zspages 0
nr_free_cma 0
numa_hit 40
numa_miss 0
numa_foreign 0
numa_interleave 12
numa_local 40
numa_other 0
pagesets
cpu: 0
count: 336
high: 378
batch: 63
vm stats threshold: 40
cpu: 1
count: 60
high: 378
batch: 63
vm stats threshold: 40
cpu: 2
count: 60
high: 378
batch: 63
vm stats threshold: 40
cpu: 3
count: 0
high: 378
batch: 63
vm stats threshold: 40
cpu: 4
count: 62
high: 378
batch: 63
vm stats threshold: 40
cpu: 5
count: 0
high: 378
batch: 63
vm stats threshold: 40
cpu: 6
count: 59
high: 378
batch: 63
vm stats threshold: 40
cpu: 7
count: 0
high: 378
batch: 63
vm stats threshold: 40
node_unreclaimable: 0
start_pfn: 0
Node 0, zone Normal
pages free 912587
min 3732
low 4754
high 5776
spanned 1048576
present 1048576
managed 1022150
protection: (0, 0, 0)
nr_free_pages 912587
nr_zone_inactive_anon 8
nr_zone_active_anon 8389
nr_zone_inactive_file 43418
nr_zone_active_file 22655
nr_zone_unevictable 0
nr_zone_write_pending 0
nr_mlock 0
nr_page_table_pages 548
nr_kernel_stack 3072
nr_bounce 0
nr_zspages 0
nr_free_cma 1024
numa_hit 3115288
numa_miss 0
numa_foreign 0
numa_interleave 6865
numa_local 3115288
numa_other 0
pagesets
cpu: 0
count: 86
high: 90
batch: 15
vm stats threshold: 48
cpu: 1
count: 80
high: 90
batch: 15
vm stats threshold: 48
cpu: 2
count: 76
high: 90
batch: 15
vm stats threshold: 48
cpu: 3
count: 53
high: 90
batch: 15
vm stats threshold: 48
cpu: 4
count: 81
high: 90
batch: 15
vm stats threshold: 48
cpu: 5
count: 18
high: 90
batch: 15
vm stats threshold: 48
cpu: 6
count: 73
high: 90
batch: 15
vm stats threshold: 48
cpu: 7
count: 63
high: 90
batch: 15
vm stats threshold: 48
node_unreclaimable: 0
start_pfn: 524288
Node 0, zone Movable
pages free 0
min 0
low 0
high 0
spanned 0
present 0
managed 0
protection: (0, 0, 0)
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [-next] lots of messages due to "mm, memory_hotplug: be more verbose for memory offline failures"
2018-12-17 15:59 [-next] lots of messages due to "mm, memory_hotplug: be more verbose for memory offline failures" Heiko Carstens
@ 2018-12-17 16:03 ` Michal Hocko
2018-12-17 16:39 ` Michal Hocko
0 siblings, 1 reply; 5+ messages in thread
From: Michal Hocko @ 2018-12-17 16:03 UTC (permalink / raw)
To: Heiko Carstens
Cc: Andrew Morton, Oscar Salvador, Anshuman Khandual,
Stephen Rothwell, linux-mm, linux-kernel, linux-next, linux-s390
On Mon 17-12-18 16:59:22, Heiko Carstens wrote:
> Hi Michal,
>
> with linux-next as of today on s390 I see tons of messages like
>
> [ 20.536664] page dumped because: has_unmovable_pages
> [ 20.536792] page:000003d081ff4080 count:1 mapcount:0 mapping:000000008ff88600 index:0x0 compound_mapcount: 0
> [ 20.536794] flags: 0x3fffe0000010200(slab|head)
> [ 20.536795] raw: 03fffe0000010200 0000000000000100 0000000000000200 000000008ff88600
> [ 20.536796] raw: 0000000000000000 0020004100000000 ffffffff00000001 0000000000000000
> [ 20.536797] page dumped because: has_unmovable_pages
> [ 20.536814] page:000003d0823b0000 count:1 mapcount:0 mapping:0000000000000000 index:0x0
> [ 20.536815] flags: 0x7fffe0000000000()
> [ 20.536817] raw: 07fffe0000000000 0000000000000100 0000000000000200 0000000000000000
> [ 20.536818] raw: 0000000000000000 0000000000000000 ffffffff00000001 0000000000000000
>
> bisect points to b323c049a999 ("mm, memory_hotplug: be more verbose for memory offline failures")
> which is the first commit with which the messages appear.
I would bet this is CMA allocator. How much is tons? Maybe we want a
rate limit or the other user is not really interested in them at all?
--
Michal Hocko
SUSE Labs
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [-next] lots of messages due to "mm, memory_hotplug: be more verbose for memory offline failures"
2018-12-17 16:03 ` Michal Hocko
@ 2018-12-17 16:39 ` Michal Hocko
2018-12-18 7:55 ` Heiko Carstens
0 siblings, 1 reply; 5+ messages in thread
From: Michal Hocko @ 2018-12-17 16:39 UTC (permalink / raw)
To: Heiko Carstens
Cc: Andrew Morton, Oscar Salvador, Anshuman Khandual,
Stephen Rothwell, linux-mm, linux-kernel, linux-next, linux-s390
On Mon 17-12-18 17:03:50, Michal Hocko wrote:
> On Mon 17-12-18 16:59:22, Heiko Carstens wrote:
> > Hi Michal,
> >
> > with linux-next as of today on s390 I see tons of messages like
> >
> > [ 20.536664] page dumped because: has_unmovable_pages
> > [ 20.536792] page:000003d081ff4080 count:1 mapcount:0 mapping:000000008ff88600 index:0x0 compound_mapcount: 0
> > [ 20.536794] flags: 0x3fffe0000010200(slab|head)
> > [ 20.536795] raw: 03fffe0000010200 0000000000000100 0000000000000200 000000008ff88600
> > [ 20.536796] raw: 0000000000000000 0020004100000000 ffffffff00000001 0000000000000000
> > [ 20.536797] page dumped because: has_unmovable_pages
> > [ 20.536814] page:000003d0823b0000 count:1 mapcount:0 mapping:0000000000000000 index:0x0
> > [ 20.536815] flags: 0x7fffe0000000000()
> > [ 20.536817] raw: 07fffe0000000000 0000000000000100 0000000000000200 0000000000000000
> > [ 20.536818] raw: 0000000000000000 0000000000000000 ffffffff00000001 0000000000000000
> >
> > bisect points to b323c049a999 ("mm, memory_hotplug: be more verbose for memory offline failures")
> > which is the first commit with which the messages appear.
>
> I would bet this is CMA allocator. How much is tons? Maybe we want a
> rate limit or the other user is not really interested in them at all?
In other words, this should silence those messages.
diff --git a/include/linux/page-isolation.h b/include/linux/page-isolation.h
index 4ae347cbc36d..4eb26d278046 100644
--- a/include/linux/page-isolation.h
+++ b/include/linux/page-isolation.h
@@ -30,8 +30,11 @@ static inline bool is_migrate_isolate(int migratetype)
}
#endif
+#define SKIP_HWPOISON 0x1
+#define REPORT_FAILURE 0x2
+
bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
- int migratetype, bool skip_hwpoisoned_pages);
+ int migratetype, int flags);
void set_pageblock_migratetype(struct page *page, int migratetype);
int move_freepages_block(struct zone *zone, struct page *page,
int migratetype, int *num_movable);
@@ -44,10 +47,14 @@ int move_freepages_block(struct zone *zone, struct page *page,
* For isolating all pages in the range finally, the caller have to
* free all pages in the range. test_page_isolated() can be used for
* test it.
+ *
+ * The following flags are allowed (they can be combined in a bit mask)
+ * SKIP_HWPOISON - ignore hwpoison pages
+ * REPORT_FAILURE - report details about the failure to isolate the range
*/
int
start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
- unsigned migratetype, bool skip_hwpoisoned_pages);
+ unsigned migratetype, int flags);
/*
* Changes MIGRATE_ISOLATE to MIGRATE_MOVABLE.
diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index c82193db4be6..8537429d33a6 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -1226,7 +1226,7 @@ static bool is_pageblock_removable_nolock(struct page *page)
if (!zone_spans_pfn(zone, pfn))
return false;
- return !has_unmovable_pages(zone, page, 0, MIGRATE_MOVABLE, true);
+ return !has_unmovable_pages(zone, page, 0, MIGRATE_MOVABLE, SKIP_HWPOISON);
}
/* Checks if this range of memory is likely to be hot-removable. */
@@ -1577,7 +1577,8 @@ static int __ref __offline_pages(unsigned long start_pfn,
/* set above range as isolated */
ret = start_isolate_page_range(start_pfn, end_pfn,
- MIGRATE_MOVABLE, true);
+ MIGRATE_MOVABLE,
+ SKIP_HWPOISON | REPORT_FAILURE);
if (ret) {
mem_hotplug_done();
reason = "failure to isolate range";
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index ec2c7916dc2d..ee4043419791 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -7754,8 +7754,7 @@ void *__init alloc_large_system_hash(const char *tablename,
* race condition. So you can't expect this function should be exact.
*/
bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
- int migratetype,
- bool skip_hwpoisoned_pages)
+ int migratetype, int flags)
{
unsigned long pfn, iter, found;
@@ -7818,7 +7817,7 @@ bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
* The HWPoisoned page may be not in buddy system, and
* page_count() is not 0.
*/
- if (skip_hwpoisoned_pages && PageHWPoison(page))
+ if ((flags & SKIP_HWPOISON) && PageHWPoison(page))
continue;
if (__PageMovable(page))
@@ -7845,7 +7844,8 @@ bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
return false;
unmovable:
WARN_ON_ONCE(zone_idx(zone) == ZONE_MOVABLE);
- dump_page(pfn_to_page(pfn+iter), "unmovable page");
+ if (flags & REPORT_FAILURE)
+ dump_page(pfn_to_page(pfn+iter), "unmovable page");
return true;
}
@@ -7972,8 +7972,7 @@ int alloc_contig_range(unsigned long start, unsigned long end,
*/
ret = start_isolate_page_range(pfn_max_align_down(start),
- pfn_max_align_up(end), migratetype,
- false);
+ pfn_max_align_up(end), migratetype, 0);
if (ret)
return ret;
diff --git a/mm/page_isolation.c b/mm/page_isolation.c
index 43e085608846..ce323e56b34d 100644
--- a/mm/page_isolation.c
+++ b/mm/page_isolation.c
@@ -15,8 +15,7 @@
#define CREATE_TRACE_POINTS
#include <trace/events/page_isolation.h>
-static int set_migratetype_isolate(struct page *page, int migratetype,
- bool skip_hwpoisoned_pages)
+static int set_migratetype_isolate(struct page *page, int migratetype, int isol_flags)
{
struct zone *zone;
unsigned long flags, pfn;
@@ -60,8 +59,7 @@ static int set_migratetype_isolate(struct page *page, int migratetype,
* FIXME: Now, memory hotplug doesn't call shrink_slab() by itself.
* We just check MOVABLE pages.
*/
- if (!has_unmovable_pages(zone, page, arg.pages_found, migratetype,
- skip_hwpoisoned_pages))
+ if (!has_unmovable_pages(zone, page, arg.pages_found, migratetype, flags))
ret = 0;
/*
@@ -185,7 +183,7 @@ __first_valid_page(unsigned long pfn, unsigned long nr_pages)
* prevents two threads from simultaneously working on overlapping ranges.
*/
int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
- unsigned migratetype, bool skip_hwpoisoned_pages)
+ unsigned migratetype, int flags)
{
unsigned long pfn;
unsigned long undo_pfn;
@@ -199,7 +197,7 @@ int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
pfn += pageblock_nr_pages) {
page = __first_valid_page(pfn, pageblock_nr_pages);
if (page &&
- set_migratetype_isolate(page, migratetype, skip_hwpoisoned_pages)) {
+ set_migratetype_isolate(page, migratetype, flags)) {
undo_pfn = pfn;
goto undo;
}
--
Michal Hocko
SUSE Labs
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [-next] lots of messages due to "mm, memory_hotplug: be more verbose for memory offline failures"
2018-12-17 16:39 ` Michal Hocko
@ 2018-12-18 7:55 ` Heiko Carstens
2018-12-18 9:09 ` Michal Hocko
0 siblings, 1 reply; 5+ messages in thread
From: Heiko Carstens @ 2018-12-18 7:55 UTC (permalink / raw)
To: Michal Hocko
Cc: Andrew Morton, Oscar Salvador, Anshuman Khandual,
Stephen Rothwell, linux-mm, linux-kernel, linux-next, linux-s390
On Mon, Dec 17, 2018 at 05:39:49PM +0100, Michal Hocko wrote:
> On Mon 17-12-18 17:03:50, Michal Hocko wrote:
> > On Mon 17-12-18 16:59:22, Heiko Carstens wrote:
> > > Hi Michal,
> > >
> > > with linux-next as of today on s390 I see tons of messages like
> > >
> > > [ 20.536664] page dumped because: has_unmovable_pages
> > > [ 20.536792] page:000003d081ff4080 count:1 mapcount:0 mapping:000000008ff88600 index:0x0 compound_mapcount: 0
> > > [ 20.536794] flags: 0x3fffe0000010200(slab|head)
> > > [ 20.536795] raw: 03fffe0000010200 0000000000000100 0000000000000200 000000008ff88600
> > > [ 20.536796] raw: 0000000000000000 0020004100000000 ffffffff00000001 0000000000000000
> > > [ 20.536797] page dumped because: has_unmovable_pages
> > > [ 20.536814] page:000003d0823b0000 count:1 mapcount:0 mapping:0000000000000000 index:0x0
> > > [ 20.536815] flags: 0x7fffe0000000000()
> > > [ 20.536817] raw: 07fffe0000000000 0000000000000100 0000000000000200 0000000000000000
> > > [ 20.536818] raw: 0000000000000000 0000000000000000 ffffffff00000001 0000000000000000
> > >
> > > bisect points to b323c049a999 ("mm, memory_hotplug: be more verbose for memory offline failures")
> > > which is the first commit with which the messages appear.
> >
> > I would bet this is CMA allocator. How much is tons? Maybe we want a
> > rate limit or the other user is not really interested in them at all?
Yes, the system in question has a 4NB CMA area. "tons" translates to several hundred.
> In other words, this should silence those messages.
Yes, with the patch below applied the messages don't appear anymore.
> diff --git a/include/linux/page-isolation.h b/include/linux/page-isolation.h
> index 4ae347cbc36d..4eb26d278046 100644
> --- a/include/linux/page-isolation.h
> +++ b/include/linux/page-isolation.h
> @@ -30,8 +30,11 @@ static inline bool is_migrate_isolate(int migratetype)
> }
> #endif
>
> +#define SKIP_HWPOISON 0x1
> +#define REPORT_FAILURE 0x2
> +
> bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
> - int migratetype, bool skip_hwpoisoned_pages);
> + int migratetype, int flags);
> void set_pageblock_migratetype(struct page *page, int migratetype);
> int move_freepages_block(struct zone *zone, struct page *page,
> int migratetype, int *num_movable);
> @@ -44,10 +47,14 @@ int move_freepages_block(struct zone *zone, struct page *page,
> * For isolating all pages in the range finally, the caller have to
> * free all pages in the range. test_page_isolated() can be used for
> * test it.
> + *
> + * The following flags are allowed (they can be combined in a bit mask)
> + * SKIP_HWPOISON - ignore hwpoison pages
> + * REPORT_FAILURE - report details about the failure to isolate the range
> */
> int
> start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
> - unsigned migratetype, bool skip_hwpoisoned_pages);
> + unsigned migratetype, int flags);
>
> /*
> * Changes MIGRATE_ISOLATE to MIGRATE_MOVABLE.
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index c82193db4be6..8537429d33a6 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -1226,7 +1226,7 @@ static bool is_pageblock_removable_nolock(struct page *page)
> if (!zone_spans_pfn(zone, pfn))
> return false;
>
> - return !has_unmovable_pages(zone, page, 0, MIGRATE_MOVABLE, true);
> + return !has_unmovable_pages(zone, page, 0, MIGRATE_MOVABLE, SKIP_HWPOISON);
> }
>
> /* Checks if this range of memory is likely to be hot-removable. */
> @@ -1577,7 +1577,8 @@ static int __ref __offline_pages(unsigned long start_pfn,
>
> /* set above range as isolated */
> ret = start_isolate_page_range(start_pfn, end_pfn,
> - MIGRATE_MOVABLE, true);
> + MIGRATE_MOVABLE,
> + SKIP_HWPOISON | REPORT_FAILURE);
> if (ret) {
> mem_hotplug_done();
> reason = "failure to isolate range";
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index ec2c7916dc2d..ee4043419791 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -7754,8 +7754,7 @@ void *__init alloc_large_system_hash(const char *tablename,
> * race condition. So you can't expect this function should be exact.
> */
> bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
> - int migratetype,
> - bool skip_hwpoisoned_pages)
> + int migratetype, int flags)
> {
> unsigned long pfn, iter, found;
>
> @@ -7818,7 +7817,7 @@ bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
> * The HWPoisoned page may be not in buddy system, and
> * page_count() is not 0.
> */
> - if (skip_hwpoisoned_pages && PageHWPoison(page))
> + if ((flags & SKIP_HWPOISON) && PageHWPoison(page))
> continue;
>
> if (__PageMovable(page))
> @@ -7845,7 +7844,8 @@ bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
> return false;
> unmovable:
> WARN_ON_ONCE(zone_idx(zone) == ZONE_MOVABLE);
> - dump_page(pfn_to_page(pfn+iter), "unmovable page");
> + if (flags & REPORT_FAILURE)
> + dump_page(pfn_to_page(pfn+iter), "unmovable page");
> return true;
> }
>
> @@ -7972,8 +7972,7 @@ int alloc_contig_range(unsigned long start, unsigned long end,
> */
>
> ret = start_isolate_page_range(pfn_max_align_down(start),
> - pfn_max_align_up(end), migratetype,
> - false);
> + pfn_max_align_up(end), migratetype, 0);
> if (ret)
> return ret;
>
> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
> index 43e085608846..ce323e56b34d 100644
> --- a/mm/page_isolation.c
> +++ b/mm/page_isolation.c
> @@ -15,8 +15,7 @@
> #define CREATE_TRACE_POINTS
> #include <trace/events/page_isolation.h>
>
> -static int set_migratetype_isolate(struct page *page, int migratetype,
> - bool skip_hwpoisoned_pages)
> +static int set_migratetype_isolate(struct page *page, int migratetype, int isol_flags)
> {
> struct zone *zone;
> unsigned long flags, pfn;
> @@ -60,8 +59,7 @@ static int set_migratetype_isolate(struct page *page, int migratetype,
> * FIXME: Now, memory hotplug doesn't call shrink_slab() by itself.
> * We just check MOVABLE pages.
> */
> - if (!has_unmovable_pages(zone, page, arg.pages_found, migratetype,
> - skip_hwpoisoned_pages))
> + if (!has_unmovable_pages(zone, page, arg.pages_found, migratetype, flags))
> ret = 0;
>
> /*
> @@ -185,7 +183,7 @@ __first_valid_page(unsigned long pfn, unsigned long nr_pages)
> * prevents two threads from simultaneously working on overlapping ranges.
> */
> int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
> - unsigned migratetype, bool skip_hwpoisoned_pages)
> + unsigned migratetype, int flags)
> {
> unsigned long pfn;
> unsigned long undo_pfn;
> @@ -199,7 +197,7 @@ int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
> pfn += pageblock_nr_pages) {
> page = __first_valid_page(pfn, pageblock_nr_pages);
> if (page &&
> - set_migratetype_isolate(page, migratetype, skip_hwpoisoned_pages)) {
> + set_migratetype_isolate(page, migratetype, flags)) {
> undo_pfn = pfn;
> goto undo;
> }
> --
> Michal Hocko
> SUSE Labs
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [-next] lots of messages due to "mm, memory_hotplug: be more verbose for memory offline failures"
2018-12-18 7:55 ` Heiko Carstens
@ 2018-12-18 9:09 ` Michal Hocko
0 siblings, 0 replies; 5+ messages in thread
From: Michal Hocko @ 2018-12-18 9:09 UTC (permalink / raw)
To: Heiko Carstens
Cc: Andrew Morton, Oscar Salvador, Anshuman Khandual,
Stephen Rothwell, linux-mm, linux-kernel, linux-next, linux-s390
On Tue 18-12-18 08:55:38, Heiko Carstens wrote:
> On Mon, Dec 17, 2018 at 05:39:49PM +0100, Michal Hocko wrote:
> > On Mon 17-12-18 17:03:50, Michal Hocko wrote:
> > > On Mon 17-12-18 16:59:22, Heiko Carstens wrote:
> > > > Hi Michal,
> > > >
> > > > with linux-next as of today on s390 I see tons of messages like
> > > >
> > > > [ 20.536664] page dumped because: has_unmovable_pages
> > > > [ 20.536792] page:000003d081ff4080 count:1 mapcount:0 mapping:000000008ff88600 index:0x0 compound_mapcount: 0
> > > > [ 20.536794] flags: 0x3fffe0000010200(slab|head)
> > > > [ 20.536795] raw: 03fffe0000010200 0000000000000100 0000000000000200 000000008ff88600
> > > > [ 20.536796] raw: 0000000000000000 0020004100000000 ffffffff00000001 0000000000000000
> > > > [ 20.536797] page dumped because: has_unmovable_pages
> > > > [ 20.536814] page:000003d0823b0000 count:1 mapcount:0 mapping:0000000000000000 index:0x0
> > > > [ 20.536815] flags: 0x7fffe0000000000()
> > > > [ 20.536817] raw: 07fffe0000000000 0000000000000100 0000000000000200 0000000000000000
> > > > [ 20.536818] raw: 0000000000000000 0000000000000000 ffffffff00000001 0000000000000000
> > > >
> > > > bisect points to b323c049a999 ("mm, memory_hotplug: be more verbose for memory offline failures")
> > > > which is the first commit with which the messages appear.
> > >
> > > I would bet this is CMA allocator. How much is tons? Maybe we want a
> > > rate limit or the other user is not really interested in them at all?
>
> Yes, the system in question has a 4NB CMA area. "tons" translates to several hundred.
OK, I guess these messages on their own without a wider context are not
that helpful. It is still surprising to see slab pages or non-movable
pages in the CMA area. The later might be an CMA allocation I guess but
slab pages shouldn't be there at all AFAIU.
> > In other words, this should silence those messages.
>
> Yes, with the patch below applied the messages don't appear anymore.
OK, I will post an official patch. Even if CMA allocator decides to
report failures it can simply add the flag.
Thanks!
> > diff --git a/include/linux/page-isolation.h b/include/linux/page-isolation.h
> > index 4ae347cbc36d..4eb26d278046 100644
> > --- a/include/linux/page-isolation.h
> > +++ b/include/linux/page-isolation.h
> > @@ -30,8 +30,11 @@ static inline bool is_migrate_isolate(int migratetype)
> > }
> > #endif
> >
> > +#define SKIP_HWPOISON 0x1
> > +#define REPORT_FAILURE 0x2
> > +
> > bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
> > - int migratetype, bool skip_hwpoisoned_pages);
> > + int migratetype, int flags);
> > void set_pageblock_migratetype(struct page *page, int migratetype);
> > int move_freepages_block(struct zone *zone, struct page *page,
> > int migratetype, int *num_movable);
> > @@ -44,10 +47,14 @@ int move_freepages_block(struct zone *zone, struct page *page,
> > * For isolating all pages in the range finally, the caller have to
> > * free all pages in the range. test_page_isolated() can be used for
> > * test it.
> > + *
> > + * The following flags are allowed (they can be combined in a bit mask)
> > + * SKIP_HWPOISON - ignore hwpoison pages
> > + * REPORT_FAILURE - report details about the failure to isolate the range
> > */
> > int
> > start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
> > - unsigned migratetype, bool skip_hwpoisoned_pages);
> > + unsigned migratetype, int flags);
> >
> > /*
> > * Changes MIGRATE_ISOLATE to MIGRATE_MOVABLE.
> > diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> > index c82193db4be6..8537429d33a6 100644
> > --- a/mm/memory_hotplug.c
> > +++ b/mm/memory_hotplug.c
> > @@ -1226,7 +1226,7 @@ static bool is_pageblock_removable_nolock(struct page *page)
> > if (!zone_spans_pfn(zone, pfn))
> > return false;
> >
> > - return !has_unmovable_pages(zone, page, 0, MIGRATE_MOVABLE, true);
> > + return !has_unmovable_pages(zone, page, 0, MIGRATE_MOVABLE, SKIP_HWPOISON);
> > }
> >
> > /* Checks if this range of memory is likely to be hot-removable. */
> > @@ -1577,7 +1577,8 @@ static int __ref __offline_pages(unsigned long start_pfn,
> >
> > /* set above range as isolated */
> > ret = start_isolate_page_range(start_pfn, end_pfn,
> > - MIGRATE_MOVABLE, true);
> > + MIGRATE_MOVABLE,
> > + SKIP_HWPOISON | REPORT_FAILURE);
> > if (ret) {
> > mem_hotplug_done();
> > reason = "failure to isolate range";
> > diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> > index ec2c7916dc2d..ee4043419791 100644
> > --- a/mm/page_alloc.c
> > +++ b/mm/page_alloc.c
> > @@ -7754,8 +7754,7 @@ void *__init alloc_large_system_hash(const char *tablename,
> > * race condition. So you can't expect this function should be exact.
> > */
> > bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
> > - int migratetype,
> > - bool skip_hwpoisoned_pages)
> > + int migratetype, int flags)
> > {
> > unsigned long pfn, iter, found;
> >
> > @@ -7818,7 +7817,7 @@ bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
> > * The HWPoisoned page may be not in buddy system, and
> > * page_count() is not 0.
> > */
> > - if (skip_hwpoisoned_pages && PageHWPoison(page))
> > + if ((flags & SKIP_HWPOISON) && PageHWPoison(page))
> > continue;
> >
> > if (__PageMovable(page))
> > @@ -7845,7 +7844,8 @@ bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
> > return false;
> > unmovable:
> > WARN_ON_ONCE(zone_idx(zone) == ZONE_MOVABLE);
> > - dump_page(pfn_to_page(pfn+iter), "unmovable page");
> > + if (flags & REPORT_FAILURE)
> > + dump_page(pfn_to_page(pfn+iter), "unmovable page");
> > return true;
> > }
> >
> > @@ -7972,8 +7972,7 @@ int alloc_contig_range(unsigned long start, unsigned long end,
> > */
> >
> > ret = start_isolate_page_range(pfn_max_align_down(start),
> > - pfn_max_align_up(end), migratetype,
> > - false);
> > + pfn_max_align_up(end), migratetype, 0);
> > if (ret)
> > return ret;
> >
> > diff --git a/mm/page_isolation.c b/mm/page_isolation.c
> > index 43e085608846..ce323e56b34d 100644
> > --- a/mm/page_isolation.c
> > +++ b/mm/page_isolation.c
> > @@ -15,8 +15,7 @@
> > #define CREATE_TRACE_POINTS
> > #include <trace/events/page_isolation.h>
> >
> > -static int set_migratetype_isolate(struct page *page, int migratetype,
> > - bool skip_hwpoisoned_pages)
> > +static int set_migratetype_isolate(struct page *page, int migratetype, int isol_flags)
> > {
> > struct zone *zone;
> > unsigned long flags, pfn;
> > @@ -60,8 +59,7 @@ static int set_migratetype_isolate(struct page *page, int migratetype,
> > * FIXME: Now, memory hotplug doesn't call shrink_slab() by itself.
> > * We just check MOVABLE pages.
> > */
> > - if (!has_unmovable_pages(zone, page, arg.pages_found, migratetype,
> > - skip_hwpoisoned_pages))
> > + if (!has_unmovable_pages(zone, page, arg.pages_found, migratetype, flags))
> > ret = 0;
> >
> > /*
> > @@ -185,7 +183,7 @@ __first_valid_page(unsigned long pfn, unsigned long nr_pages)
> > * prevents two threads from simultaneously working on overlapping ranges.
> > */
> > int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
> > - unsigned migratetype, bool skip_hwpoisoned_pages)
> > + unsigned migratetype, int flags)
> > {
> > unsigned long pfn;
> > unsigned long undo_pfn;
> > @@ -199,7 +197,7 @@ int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
> > pfn += pageblock_nr_pages) {
> > page = __first_valid_page(pfn, pageblock_nr_pages);
> > if (page &&
> > - set_migratetype_isolate(page, migratetype, skip_hwpoisoned_pages)) {
> > + set_migratetype_isolate(page, migratetype, flags)) {
> > undo_pfn = pfn;
> > goto undo;
> > }
> > --
> > Michal Hocko
> > SUSE Labs
> >
--
Michal Hocko
SUSE Labs
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2018-12-18 9:09 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-12-17 15:59 [-next] lots of messages due to "mm, memory_hotplug: be more verbose for memory offline failures" Heiko Carstens
2018-12-17 16:03 ` Michal Hocko
2018-12-17 16:39 ` Michal Hocko
2018-12-18 7:55 ` Heiko Carstens
2018-12-18 9:09 ` Michal Hocko
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).