Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-13 Thread Michal Hocko 
On Fri 13-10-17 22:42:46, Michael Ellerman wrote:
> Vlastimil Babka  writes:
> > On 10/11/2017 08:51 AM, Michal Hocko wrote:
> >> On Wed 11-10-17 13:37:50, Michael Ellerman wrote:
> >>> Michal Hocko  writes:
>  On Tue 10-10-17 23:05:08, Michael Ellerman wrote:
> > Michal Hocko  writes:
> >> From: Michal Hocko 
> >>
> >> Memory offlining can fail just too eagerly under a heavy memory 
> >> pressure.
> ...
> >
> > This breaks offline for me.
> >
> > Prior to this commit:
> >   /sys/devices/system/memory/memory0# time echo 0 > online
> >   -bash: echo: write error: Device or resource busy
> >
> > Well, that means offline didn't actually work for that block even before
> > this patch, right? Is it even a movable_node block? I guess not?
> 
> Correct. It should fail.
> 
> > After:
> >   /sys/devices/system/memory/memory0# time echo 0 > online
> >   -bash: echo: write error: Device or resource busy
> >   
> >   real  2m0.009s
> >   user  0m0.000s
> >   sys   1m25.035s
> >
> > There's no way that block can be removed, it contains the kernel text,
> > so it should instantly fail - which it used to.
> >
> > Ah, right. So your complain is really about that the failure is not
> > instant anymore for blocks that can't be offlined.
> 
> Yes. Previously it failed instantly, now it doesn't fail, and loops
> infinitely (once the 2 minute limit is removed).

Yeah it failed only because the migration code retried few times and we
bailed out which is wrong as well. I will send two patches as a reply to
this email.

> >> This is really strange! As you write in other email the page is
> >> reserved. That means that some of the earlier checks 
> >>if (zone_idx(zone) == ZONE_MOVABLE)
> >>return false;
> >>mt = get_pageblock_migratetype(page);
> >>if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
> >
> > The MIGRATE_MOVABLE check is indeed bogus, because that doesn't
> > guarantee there are no unmovable pages in the block (CMA block OTOH
> > should be a guarantee).
> 
> OK I'll try that and get back to you.

Thanks!
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-13 Thread Michael Ellerman 
Vlastimil Babka  writes:
> On 10/11/2017 08:51 AM, Michal Hocko wrote:
>> On Wed 11-10-17 13:37:50, Michael Ellerman wrote:
>>> Michal Hocko  writes:
 On Tue 10-10-17 23:05:08, Michael Ellerman wrote:
> Michal Hocko  writes:
>> From: Michal Hocko 
>>
>> Memory offlining can fail just too eagerly under a heavy memory pressure.
...
>
> This breaks offline for me.
>
> Prior to this commit:
>   /sys/devices/system/memory/memory0# time echo 0 > online
>   -bash: echo: write error: Device or resource busy
>
> Well, that means offline didn't actually work for that block even before
> this patch, right? Is it even a movable_node block? I guess not?

Correct. It should fail.

> After:
>   /sys/devices/system/memory/memory0# time echo 0 > online
>   -bash: echo: write error: Device or resource busy
>   
>   real2m0.009s
>   user0m0.000s
>   sys 1m25.035s
>
> There's no way that block can be removed, it contains the kernel text,
> so it should instantly fail - which it used to.
>
> Ah, right. So your complain is really about that the failure is not
> instant anymore for blocks that can't be offlined.

Yes. Previously it failed instantly, now it doesn't fail, and loops
infinitely (once the 2 minute limit is removed).

>> This is really strange! As you write in other email the page is
>> reserved. That means that some of the earlier checks 
>>  if (zone_idx(zone) == ZONE_MOVABLE)
>>  return false;
>>  mt = get_pageblock_migratetype(page);
>>  if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
>
> The MIGRATE_MOVABLE check is indeed bogus, because that doesn't
> guarantee there are no unmovable pages in the block (CMA block OTOH
> should be a guarantee).

OK I'll try that and get back to you.

cheers


>> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
>> index 3badcedf96a7..5b4d85ae445c 100644
>> --- a/mm/page_alloc.c
>> +++ b/mm/page_alloc.c
>> @@ -7355,9 +7355,6 @@ bool has_unmovable_pages(struct zone *zone, struct 
>> page *page, int count,
>>   */
>>  if (zone_idx(zone) == ZONE_MOVABLE)
>>  return false;
>> -mt = get_pageblock_migratetype(page);
>> -if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
>> -return false;
>>  
>>  pfn = page_to_pfn(page);
>>  for (found = 0, iter = 0; iter < pageblock_nr_pages; iter++) {
>> @@ -7368,6 +7365,9 @@ bool has_unmovable_pages(struct zone *zone, struct 
>> page *page, int count,
>>  
>>  page = pfn_to_page(check);
>>  
>> +if (PageReserved(page))
>> +return true;
>> +
>>  /*
>>   * Hugepages are not in LRU lists, but they're movable.
>>   * We need not scan over tail pages bacause we don't
>>

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-11 Thread Michal Hocko 
On Wed 11-10-17 19:35:04, Anshuman Khandual wrote:
[...]
> >   $ grep __init_begin /proc/kallsyms
> >   c0d7 T __init_begin
> >   $ ./page-types -r -a 0x0,0xd7
> >flagspage-count   MB  symbolic-flags 
> > long-symbolic-flags
> >   0x0001   215   13  
> > __r___ reserved
> >total   215   13
> 
> Hey Michael,
> 
> What tool is this 'page-types' ?

tools/vm/page-types.c

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-11 Thread Anshuman Khandual 
On 10/11/2017 10:49 AM, Michael Ellerman wrote:
> Michael Ellerman  writes:
>> Michal Hocko  writes:
>>> On Tue 10-10-17 23:05:08, Michael Ellerman wrote:
 Michal Hocko  writes:
> From: Michal Hocko 
> Memory offlining can fail just too eagerly under a heavy memory pressure.
>
> [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> mapping:88ff926c9f38 index:0x3
> [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> [ 5410.336811] page dumped because: isolation failed
> [ 5410.336813] page->mem_cgroup:8801cd662000
> [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
>
> Isolation has failed here because the page is not on LRU. Most probably
> because it was on the pcp LRU cache or it has been removed from the LRU
> already but it hasn't been freed yet. In both cases the page doesn't look
> non-migrable so retrying more makes sense.
 This breaks offline for me.

 Prior to this commit:
   /sys/devices/system/memory/memory0# time echo 0 > online
   -bash: echo: write error: Device or resource busy
   
   real 0m0.001s
   user 0m0.000s
   sys  0m0.001s

 After:
   /sys/devices/system/memory/memory0# time echo 0 > online
   -bash: echo: write error: Device or resource busy
   
   real 2m0.009s
   user 0m0.000s
   sys  1m25.035s

 There's no way that block can be removed, it contains the kernel text,
 so it should instantly fail - which it used to.
>>> OK, that means that start_isolate_page_range should have failed but it
>>> hasn't for some reason. I strongly suspect has_unmovable_pages is doing
>>> something wrong. Is the kernel text marked somehow? E.g. PageReserved?
>> I'm not sure how the text is marked, will have to dig into that.
> Yeah it's reserved:
> 
>   $ grep __init_begin /proc/kallsyms
>   c0d7 T __init_begin
>   $ ./page-types -r -a 0x0,0xd7
>flags  page-count   MB  symbolic-flags 
> long-symbolic-flags
>   0x0001 215   13  
> __r___ reserved
>total 215   13

Hey Michael,

What tool is this 'page-types' ?

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-11 Thread Michal Hocko 
On Wed 11-10-17 13:17:13, Vlastimil Babka wrote:
> On 10/11/2017 10:13 AM, Michal Hocko wrote:
> > On Wed 11-10-17 10:04:39, Vlastimil Babka wrote:
> >> On 10/11/2017 08:51 AM, Michal Hocko wrote:
> > [...]
> >>> This is really strange! As you write in other email the page is
> >>> reserved. That means that some of the earlier checks 
> >>>   if (zone_idx(zone) == ZONE_MOVABLE)
> >>>   return false;
> >>>   mt = get_pageblock_migratetype(page);
> >>>   if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
> >>
> >> The MIGRATE_MOVABLE check is indeed bogus, because that doesn't
> >> guarantee there are no unmovable pages in the block (CMA block OTOH
> >> should be a guarantee).
> > 
> > OK, thanks for confirmation. I will remove the MIGRATE_MOVABLE check
> > here. Do you think it is worth removing CMA check as well? This is
> > merely an optimization AFAIU because we do not have to check the full
> > pageblockworth of pfns.
> 
> Actually, we should remove the CMA part as well. It's true that
> MIGRATE_CMA does guarantee that the *buddy allocator* won't allocate
> non-MOVABLE pages from the pageblock. But if the memory got allocated as
> an actual CMA allocation (alloc_contig...) it will almost certainly not
> be movable.

That was my suspicious. Thanks!
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-11 Thread Vlastimil Babka 
On 10/11/2017 10:13 AM, Michal Hocko wrote:
> On Wed 11-10-17 10:04:39, Vlastimil Babka wrote:
>> On 10/11/2017 08:51 AM, Michal Hocko wrote:
> [...]
>>> This is really strange! As you write in other email the page is
>>> reserved. That means that some of the earlier checks 
>>> if (zone_idx(zone) == ZONE_MOVABLE)
>>> return false;
>>> mt = get_pageblock_migratetype(page);
>>> if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
>>
>> The MIGRATE_MOVABLE check is indeed bogus, because that doesn't
>> guarantee there are no unmovable pages in the block (CMA block OTOH
>> should be a guarantee).
> 
> OK, thanks for confirmation. I will remove the MIGRATE_MOVABLE check
> here. Do you think it is worth removing CMA check as well? This is
> merely an optimization AFAIU because we do not have to check the full
> pageblockworth of pfns.

Actually, we should remove the CMA part as well. It's true that
MIGRATE_CMA does guarantee that the *buddy allocator* won't allocate
non-MOVABLE pages from the pageblock. But if the memory got allocated as
an actual CMA allocation (alloc_contig...) it will almost certainly not
be movable.

> Anyway, let's way for Michael to confirm it really helps. If yes I will
> post a full patch and ask Andrew to add it as a prerequisite for this
> patch when sending to Linus to prevent the regression.
>

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-11 Thread Michal Hocko 
On Wed 11-10-17 10:04:39, Vlastimil Babka wrote:
> On 10/11/2017 08:51 AM, Michal Hocko wrote:
[...]
> > This is really strange! As you write in other email the page is
> > reserved. That means that some of the earlier checks 
> > if (zone_idx(zone) == ZONE_MOVABLE)
> > return false;
> > mt = get_pageblock_migratetype(page);
> > if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
> 
> The MIGRATE_MOVABLE check is indeed bogus, because that doesn't
> guarantee there are no unmovable pages in the block (CMA block OTOH
> should be a guarantee).

OK, thanks for confirmation. I will remove the MIGRATE_MOVABLE check
here. Do you think it is worth removing CMA check as well? This is
merely an optimization AFAIU because we do not have to check the full
pageblockworth of pfns.

Anyway, let's way for Michael to confirm it really helps. If yes I will
post a full patch and ask Andrew to add it as a prerequisite for this
patch when sending to Linus to prevent the regression.
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-11 Thread Vlastimil Babka 
On 10/11/2017 08:51 AM, Michal Hocko wrote:
> On Wed 11-10-17 13:37:50, Michael Ellerman wrote:
>> Michal Hocko  writes:
>>
>>> On Tue 10-10-17 23:05:08, Michael Ellerman wrote:
 Michal Hocko  writes:

> From: Michal Hocko 
>
> Memory offlining can fail just too eagerly under a heavy memory pressure.
>
> [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> mapping:88ff926c9f38 index:0x3
> [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> [ 5410.336811] page dumped because: isolation failed
> [ 5410.336813] page->mem_cgroup:8801cd662000
> [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
>
> Isolation has failed here because the page is not on LRU. Most probably
> because it was on the pcp LRU cache or it has been removed from the LRU
> already but it hasn't been freed yet. In both cases the page doesn't look
> non-migrable so retrying more makes sense.

 This breaks offline for me.

 Prior to this commit:
   /sys/devices/system/memory/memory0# time echo 0 > online
   -bash: echo: write error: Device or resource busy

Well, that means offline didn't actually work for that block even before
this patch, right? Is it even a movable_node block? I guess not?

   real 0m0.001s
   user 0m0.000s
   sys  0m0.001s

 After:
   /sys/devices/system/memory/memory0# time echo 0 > online
   -bash: echo: write error: Device or resource busy
   
   real 2m0.009s
   user 0m0.000s
   sys  1m25.035s


 There's no way that block can be removed, it contains the kernel text,
 so it should instantly fail - which it used to.

Ah, right. So your complain is really about that the failure is not
instant anymore for blocks that can't be offlined.

>>> OK, that means that start_isolate_page_range should have failed but it
>>> hasn't for some reason. I strongly suspect has_unmovable_pages is doing
>>> something wrong. Is the kernel text marked somehow? E.g. PageReserved?
>>
>> I'm not sure how the text is marked, will have to dig into that.
>>
>>> In other words, does the diff below helps?
>>
>> No that doesn't help.
> 
> This is really strange! As you write in other email the page is
> reserved. That means that some of the earlier checks 
>   if (zone_idx(zone) == ZONE_MOVABLE)
>   return false;
>   mt = get_pageblock_migratetype(page);
>   if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))

The MIGRATE_MOVABLE check is indeed bogus, because that doesn't
guarantee there are no unmovable pages in the block (CMA block OTOH
should be a guarantee).

>   return false;
> has bailed out early. I would be quite surprised if the kernel text was
> sitting in the zone movable. The migrate type check is more fishy
> AFAICS. I can imagine that the kernel text can share the movable or CMA
> mt block. I am not really familiar with this function but it looks
> suspicious. So does it help to remove this check?
> --- 
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 3badcedf96a7..5b4d85ae445c 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -7355,9 +7355,6 @@ bool has_unmovable_pages(struct zone *zone, struct page 
> *page, int count,
>*/
>   if (zone_idx(zone) == ZONE_MOVABLE)
>   return false;
> - mt = get_pageblock_migratetype(page);
> - if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
> - return false;
>  
>   pfn = page_to_pfn(page);
>   for (found = 0, iter = 0; iter < pageblock_nr_pages; iter++) {
> @@ -7368,6 +7365,9 @@ bool has_unmovable_pages(struct zone *zone, struct page 
> *page, int count,
>  
>   page = pfn_to_page(check);
>  
> + if (PageReserved(page))
> + return true;
> +
>   /*
>* Hugepages are not in LRU lists, but they're movable.
>* We need not scan over tail pages bacause we don't
>

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-10 Thread Michal Hocko 
On Wed 11-10-17 13:37:50, Michael Ellerman wrote:
> Michal Hocko  writes:
> 
> > On Tue 10-10-17 23:05:08, Michael Ellerman wrote:
> >> Michal Hocko  writes:
> >> 
> >> > From: Michal Hocko 
> >> >
> >> > Memory offlining can fail just too eagerly under a heavy memory pressure.
> >> >
> >> > [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> >> > mapping:88ff926c9f38 index:0x3
> >> > [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> >> > [ 5410.336811] page dumped because: isolation failed
> >> > [ 5410.336813] page->mem_cgroup:8801cd662000
> >> > [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] 
> >> > failed
> >> >
> >> > Isolation has failed here because the page is not on LRU. Most probably
> >> > because it was on the pcp LRU cache or it has been removed from the LRU
> >> > already but it hasn't been freed yet. In both cases the page doesn't look
> >> > non-migrable so retrying more makes sense.
> >> 
> >> This breaks offline for me.
> >> 
> >> Prior to this commit:
> >>   /sys/devices/system/memory/memory0# time echo 0 > online
> >>   -bash: echo: write error: Device or resource busy
> >>   
> >>   real 0m0.001s
> >>   user 0m0.000s
> >>   sys  0m0.001s
> >> 
> >> After:
> >>   /sys/devices/system/memory/memory0# time echo 0 > online
> >>   -bash: echo: write error: Device or resource busy
> >>   
> >>   real 2m0.009s
> >>   user 0m0.000s
> >>   sys  1m25.035s
> >> 
> >> 
> >> There's no way that block can be removed, it contains the kernel text,
> >> so it should instantly fail - which it used to.
> >
> > OK, that means that start_isolate_page_range should have failed but it
> > hasn't for some reason. I strongly suspect has_unmovable_pages is doing
> > something wrong. Is the kernel text marked somehow? E.g. PageReserved?
> 
> I'm not sure how the text is marked, will have to dig into that.
> 
> > In other words, does the diff below helps?
> 
> No that doesn't help.

This is really strange! As you write in other email the page is
reserved. That means that some of the earlier checks 
if (zone_idx(zone) == ZONE_MOVABLE)
return false;
mt = get_pageblock_migratetype(page);
if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
return false;
has bailed out early. I would be quite surprised if the kernel text was
sitting in the zone movable. The migrate type check is more fishy
AFAICS. I can imagine that the kernel text can share the movable or CMA
mt block. I am not really familiar with this function but it looks
suspicious. So does it help to remove this check?
--- 
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 3badcedf96a7..5b4d85ae445c 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -7355,9 +7355,6 @@ bool has_unmovable_pages(struct zone *zone, struct page 
*page, int count,
 */
if (zone_idx(zone) == ZONE_MOVABLE)
return false;
-   mt = get_pageblock_migratetype(page);
-   if (mt == MIGRATE_MOVABLE || is_migrate_cma(mt))
-   return false;
 
pfn = page_to_pfn(page);
for (found = 0, iter = 0; iter < pageblock_nr_pages; iter++) {
@@ -7368,6 +7365,9 @@ bool has_unmovable_pages(struct zone *zone, struct page 
*page, int count,
 
page = pfn_to_page(check);
 
+   if (PageReserved(page))
+   return true;
+
/*
 * Hugepages are not in LRU lists, but they're movable.
 * We need not scan over tail pages bacause we don't

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-10 Thread Michael Ellerman 
Michael Ellerman  writes:
> Michal Hocko  writes:
>> On Tue 10-10-17 23:05:08, Michael Ellerman wrote:
>>> Michal Hocko  writes:
>>> > From: Michal Hocko 
>>> > Memory offlining can fail just too eagerly under a heavy memory pressure.
>>> >
>>> > [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
>>> > mapping:88ff926c9f38 index:0x3
>>> > [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
>>> > [ 5410.336811] page dumped because: isolation failed
>>> > [ 5410.336813] page->mem_cgroup:8801cd662000
>>> > [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
>>> >
>>> > Isolation has failed here because the page is not on LRU. Most probably
>>> > because it was on the pcp LRU cache or it has been removed from the LRU
>>> > already but it hasn't been freed yet. In both cases the page doesn't look
>>> > non-migrable so retrying more makes sense.
>>> 
>>> This breaks offline for me.
>>> 
>>> Prior to this commit:
>>>   /sys/devices/system/memory/memory0# time echo 0 > online
>>>   -bash: echo: write error: Device or resource busy
>>>   
>>>   real  0m0.001s
>>>   user  0m0.000s
>>>   sys   0m0.001s
>>> 
>>> After:
>>>   /sys/devices/system/memory/memory0# time echo 0 > online
>>>   -bash: echo: write error: Device or resource busy
>>>   
>>>   real  2m0.009s
>>>   user  0m0.000s
>>>   sys   1m25.035s
>>> 
>>> There's no way that block can be removed, it contains the kernel text,
>>> so it should instantly fail - which it used to.
>>
>> OK, that means that start_isolate_page_range should have failed but it
>> hasn't for some reason. I strongly suspect has_unmovable_pages is doing
>> something wrong. Is the kernel text marked somehow? E.g. PageReserved?
>
> I'm not sure how the text is marked, will have to dig into that.

Yeah it's reserved:

  $ grep __init_begin /proc/kallsyms
  c0d7 T __init_begin
  $ ./page-types -r -a 0x0,0xd7
   flagspage-count   MB  symbolic-flags 
long-symbolic-flags
  0x0001   215   13  
__r___ reserved
   total   215   13


I added some printks, we're getting EBUSY from do_migrate_range(pfn, end_pfn).

So we seem to just have an infinite loop:

  repeat:
/* start memory hot removal */
ret = -EINTR;
if (signal_pending(current))
goto failed_removal;
  
cond_resched();
lru_add_drain_all_cpuslocked();
drain_all_pages(zone);
  
pfn = scan_movable_pages(start_pfn, end_pfn);
if (pfn) { /* We have movable pages */
ret = do_migrate_range(pfn, end_pfn);
printk_ratelimited("memory-hotplug: migrate range returned 
%ld\n", ret);
goto repeat;
}


eg:

  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  __offline_pages: 354031 callbacks suppressed
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  memory-hotplug: migrate range returned -16
  __offline_pages: 355794 callbacks suppressed


cheers

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-10 Thread Michael Ellerman 
Michal Hocko  writes:

> On Tue 10-10-17 23:05:08, Michael Ellerman wrote:
>> Michal Hocko  writes:
>> 
>> > From: Michal Hocko 
>> >
>> > Memory offlining can fail just too eagerly under a heavy memory pressure.
>> >
>> > [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
>> > mapping:88ff926c9f38 index:0x3
>> > [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
>> > [ 5410.336811] page dumped because: isolation failed
>> > [ 5410.336813] page->mem_cgroup:8801cd662000
>> > [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
>> >
>> > Isolation has failed here because the page is not on LRU. Most probably
>> > because it was on the pcp LRU cache or it has been removed from the LRU
>> > already but it hasn't been freed yet. In both cases the page doesn't look
>> > non-migrable so retrying more makes sense.
>> 
>> This breaks offline for me.
>> 
>> Prior to this commit:
>>   /sys/devices/system/memory/memory0# time echo 0 > online
>>   -bash: echo: write error: Device or resource busy
>>   
>>   real   0m0.001s
>>   user   0m0.000s
>>   sys0m0.001s
>> 
>> After:
>>   /sys/devices/system/memory/memory0# time echo 0 > online
>>   -bash: echo: write error: Device or resource busy
>>   
>>   real   2m0.009s
>>   user   0m0.000s
>>   sys1m25.035s
>> 
>> 
>> There's no way that block can be removed, it contains the kernel text,
>> so it should instantly fail - which it used to.
>
> OK, that means that start_isolate_page_range should have failed but it
> hasn't for some reason. I strongly suspect has_unmovable_pages is doing
> something wrong. Is the kernel text marked somehow? E.g. PageReserved?

I'm not sure how the text is marked, will have to dig into that.

> In other words, does the diff below helps?

No that doesn't help.

> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 3badcedf96a7..00d042052501 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -7368,6 +7368,9 @@ bool has_unmovable_pages(struct zone *zone, struct page 
> *page, int count,
>  
>   page = pfn_to_page(check);
>  
> + if (PageReserved(page))
> + return true;
> +
>   /*
>* Hugepages are not in LRU lists, but they're movable.
>* We need not scan over tail pages bacause we don't
>
>
>> With commit 3aa2823fdf66 ("mm, memory_hotplug: remove timeout from
>> __offline_memory") also applied, it appears to just get stuck forever,
>> and I get lots of:
>> 
>>   [ 1232.112953] INFO: task kworker/3:0:4609 blocked for more than 120 
>> seconds.
>>   [ 1232.113067]   Not tainted 4.14.0-rc4-gcc6-next-20171009-g49827b9 #1
>>   [ 1232.113183] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables 
>> this message.
>>   [ 1232.113319] kworker/3:0 D11984  4609  2 0x0800
>>   [ 1232.113416] Workqueue: memcg_kmem_cache memcg_kmem_cache_create_func
>>   [ 1232.113531] Call Trace:
>>   [ 1232.113579] [c000fb2db7a0] [c000fb2db900] 0xc000fb2db900 
>> (unreliable)
>>   [ 1232.113717] [c000fb2db970] [c001c964] 
>> __switch_to+0x304/0x6e0
>>   [ 1232.113840] [c000fb2dba10] [c0a408c0] __schedule+0x2e0/0xa80
>>   [ 1232.113978] [c000fb2dbae0] [c0a410a8] schedule+0x48/0xc0
>>   [ 1232.114113] [c000fb2dbb10] [c0a44d88] 
>> rwsem_down_read_failed+0x128/0x1b0
>>   [ 1232.114269] [c000fb2dbb70] [c01696a8] 
>> __percpu_down_read+0x108/0x110
>>   [ 1232.114426] [c000fb2dbba0] [c032e498] 
>> get_online_mems+0x68/0x80
>>   [ 1232.115487] [c000fb2dbbc0] [c02c82ec] 
>> memcg_create_kmem_cache+0x4c/0x190
>>   [ 1232.115651] [c000fb2dbc60] [c03483b8] 
>> memcg_kmem_cache_create_func+0x38/0xf0
>>   [ 1232.115809] [c000fb2dbc90] [c0121594] 
>> process_one_work+0x2b4/0x590
>>   [ 1232.115964] [c000fb2dbd20] [c0121908] 
>> worker_thread+0x98/0x5d0
>>   [ 1232.116095] [c000fb2dbdc0] [c012a134] kthread+0x164/0x1b0
>>   [ 1232.116229] [c000fb2dbe30] [c000bae0] 
>> ret_from_kernel_thread+0x5c/0x7c
>
> I do not see how this is related to the offline path.

It's blocked doing get_online_mems(). So it's unrelated to the offline,
but it can't proceed until the offline finishes, which it never does,
IIUIC.

cheers

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-10 Thread Michal Hocko 
On Tue 10-10-17 23:05:08, Michael Ellerman wrote:
> Michal Hocko  writes:
> 
> > From: Michal Hocko 
> >
> > Memory offlining can fail just too eagerly under a heavy memory pressure.
> >
> > [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> > mapping:88ff926c9f38 index:0x3
> > [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> > [ 5410.336811] page dumped because: isolation failed
> > [ 5410.336813] page->mem_cgroup:8801cd662000
> > [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
> >
> > Isolation has failed here because the page is not on LRU. Most probably
> > because it was on the pcp LRU cache or it has been removed from the LRU
> > already but it hasn't been freed yet. In both cases the page doesn't look
> > non-migrable so retrying more makes sense.
> 
> This breaks offline for me.
> 
> Prior to this commit:
>   /sys/devices/system/memory/memory0# time echo 0 > online
>   -bash: echo: write error: Device or resource busy
>   
>   real0m0.001s
>   user0m0.000s
>   sys 0m0.001s
> 
> After:
>   /sys/devices/system/memory/memory0# time echo 0 > online
>   -bash: echo: write error: Device or resource busy
>   
>   real2m0.009s
>   user0m0.000s
>   sys 1m25.035s
> 
> 
> There's no way that block can be removed, it contains the kernel text,
> so it should instantly fail - which it used to.

OK, that means that start_isolate_page_range should have failed but it
hasn't for some reason. I strongly suspect has_unmovable_pages is doing
something wrong. Is the kernel text marked somehow? E.g. PageReserved?
In other words, does the diff below helps?

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 3badcedf96a7..00d042052501 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -7368,6 +7368,9 @@ bool has_unmovable_pages(struct zone *zone, struct page 
*page, int count,
 
page = pfn_to_page(check);
 
+   if (PageReserved(page))
+   return true;
+
/*
 * Hugepages are not in LRU lists, but they're movable.
 * We need not scan over tail pages bacause we don't


> With commit 3aa2823fdf66 ("mm, memory_hotplug: remove timeout from
> __offline_memory") also applied, it appears to just get stuck forever,
> and I get lots of:
> 
>   [ 1232.112953] INFO: task kworker/3:0:4609 blocked for more than 120 
> seconds.
>   [ 1232.113067]   Not tainted 4.14.0-rc4-gcc6-next-20171009-g49827b9 #1
>   [ 1232.113183] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables 
> this message.
>   [ 1232.113319] kworker/3:0 D11984  4609  2 0x0800
>   [ 1232.113416] Workqueue: memcg_kmem_cache memcg_kmem_cache_create_func
>   [ 1232.113531] Call Trace:
>   [ 1232.113579] [c000fb2db7a0] [c000fb2db900] 0xc000fb2db900 
> (unreliable)
>   [ 1232.113717] [c000fb2db970] [c001c964] __switch_to+0x304/0x6e0
>   [ 1232.113840] [c000fb2dba10] [c0a408c0] __schedule+0x2e0/0xa80
>   [ 1232.113978] [c000fb2dbae0] [c0a410a8] schedule+0x48/0xc0
>   [ 1232.114113] [c000fb2dbb10] [c0a44d88] 
> rwsem_down_read_failed+0x128/0x1b0
>   [ 1232.114269] [c000fb2dbb70] [c01696a8] 
> __percpu_down_read+0x108/0x110
>   [ 1232.114426] [c000fb2dbba0] [c032e498] 
> get_online_mems+0x68/0x80
>   [ 1232.115487] [c000fb2dbbc0] [c02c82ec] 
> memcg_create_kmem_cache+0x4c/0x190
>   [ 1232.115651] [c000fb2dbc60] [c03483b8] 
> memcg_kmem_cache_create_func+0x38/0xf0
>   [ 1232.115809] [c000fb2dbc90] [c0121594] 
> process_one_work+0x2b4/0x590
>   [ 1232.115964] [c000fb2dbd20] [c0121908] 
> worker_thread+0x98/0x5d0
>   [ 1232.116095] [c000fb2dbdc0] [c012a134] kthread+0x164/0x1b0
>   [ 1232.116229] [c000fb2dbe30] [c000bae0] 
> ret_from_kernel_thread+0x5c/0x7c

I do not see how this is related to the offline path.
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-10-10 Thread Michael Ellerman 
Michal Hocko  writes:

> From: Michal Hocko 
>
> Memory offlining can fail just too eagerly under a heavy memory pressure.
>
> [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> mapping:88ff926c9f38 index:0x3
> [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> [ 5410.336811] page dumped because: isolation failed
> [ 5410.336813] page->mem_cgroup:8801cd662000
> [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
>
> Isolation has failed here because the page is not on LRU. Most probably
> because it was on the pcp LRU cache or it has been removed from the LRU
> already but it hasn't been freed yet. In both cases the page doesn't look
> non-migrable so retrying more makes sense.

This breaks offline for me.

Prior to this commit:
  /sys/devices/system/memory/memory0# time echo 0 > online
  -bash: echo: write error: Device or resource busy
  
  real  0m0.001s
  user  0m0.000s
  sys   0m0.001s

After:
  /sys/devices/system/memory/memory0# time echo 0 > online
  -bash: echo: write error: Device or resource busy
  
  real  2m0.009s
  user  0m0.000s
  sys   1m25.035s


There's no way that block can be removed, it contains the kernel text,
so it should instantly fail - which it used to.


With commit 3aa2823fdf66 ("mm, memory_hotplug: remove timeout from
__offline_memory") also applied, it appears to just get stuck forever,
and I get lots of:

  [ 1232.112953] INFO: task kworker/3:0:4609 blocked for more than 120 seconds.
  [ 1232.113067]   Not tainted 4.14.0-rc4-gcc6-next-20171009-g49827b9 #1
  [ 1232.113183] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables 
this message.
  [ 1232.113319] kworker/3:0 D11984  4609  2 0x0800
  [ 1232.113416] Workqueue: memcg_kmem_cache memcg_kmem_cache_create_func
  [ 1232.113531] Call Trace:
  [ 1232.113579] [c000fb2db7a0] [c000fb2db900] 0xc000fb2db900 
(unreliable)
  [ 1232.113717] [c000fb2db970] [c001c964] __switch_to+0x304/0x6e0
  [ 1232.113840] [c000fb2dba10] [c0a408c0] __schedule+0x2e0/0xa80
  [ 1232.113978] [c000fb2dbae0] [c0a410a8] schedule+0x48/0xc0
  [ 1232.114113] [c000fb2dbb10] [c0a44d88] 
rwsem_down_read_failed+0x128/0x1b0
  [ 1232.114269] [c000fb2dbb70] [c01696a8] 
__percpu_down_read+0x108/0x110
  [ 1232.114426] [c000fb2dbba0] [c032e498] get_online_mems+0x68/0x80
  [ 1232.115487] [c000fb2dbbc0] [c02c82ec] 
memcg_create_kmem_cache+0x4c/0x190
  [ 1232.115651] [c000fb2dbc60] [c03483b8] 
memcg_kmem_cache_create_func+0x38/0xf0
  [ 1232.115809] [c000fb2dbc90] [c0121594] 
process_one_work+0x2b4/0x590
  [ 1232.115964] [c000fb2dbd20] [c0121908] worker_thread+0x98/0x5d0
  [ 1232.116095] [c000fb2dbdc0] [c012a134] kthread+0x164/0x1b0
  [ 1232.116229] [c000fb2dbe30] [c000bae0] 
ret_from_kernel_thread+0x5c/0x7c


cheers

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-13 Thread Michal Hocko 
On Wed 13-09-17 14:19:19, Vlastimil Babka wrote:
> On 09/13/2017 02:14 PM, Michal Hocko wrote:
>  Do you think that the changelog should be more clear about this?
> >>>
> >>> It certainly wouldn't hurt :)
> >>
> >> So what do you think about the following wording:
> > 
> > Ups, wrong patch
> > 
> > 
> > From 8639496a834b4a7c24972ec23b17e50f0d6a304c Mon Sep 17 00:00:00 2001
> > From: Michal Hocko 
> > Date: Mon, 14 Aug 2017 10:46:12 +0200
> > Subject: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early
> > 
> > Memory offlining can fail just too eagerly under a heavy memory pressure.
> > 
> > [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> > mapping:88ff926c9f38 index:0x3
> > [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> > [ 5410.336811] page dumped because: isolation failed
> > [ 5410.336813] page->mem_cgroup:8801cd662000
> > [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
> > 
> > Isolation has failed here because the page is not on LRU. Most probably
> > because it was on the pcp LRU cache or it has been removed from the LRU
> > already but it hasn't been freed yet. In both cases the page doesn't look
> > non-migrable so retrying more makes sense.
> > 
> > __offline_pages seems rather cluttered when it comes to the retry
> > logic. We have 5 retries at maximum and a timeout. We could argue
> > whether the timeout makes sense but failing just because of a race when
> > somebody isoltes a page from LRU or puts it on a pcp LRU lists is just
> > wrong. It only takes it to race with a process which unmaps some pages
> > and remove them from the LRU list and we can fail the whole offline
> > because of something that is a temporary condition and actually not
> > harmful for the offline.
> > 
> > Please note that unmovable pages should be already excluded during
> > start_isolate_page_range. We could argue that has_unmovable_pages is
> > racy and MIGRATE_MOVABLE check doesn't provide any hard guarantee either
> > but kernel zones (aka < ZONE_MOVABLE) will very likely detect unmovable
> > pages in most cases and movable zone shouldn't contain unmovable pages
> > at all. Some of those pages might be pinned but not for ever because
> > that would be a bug on its own. In any case the context is still
> > interruptible and so the userspace can easily bail out when the
> > operation takes too long. This is certainly better behavior than a
> > hardcoded retry loop which is racy.
> > 
> > Fix this by removing the max retry count and only rely on the timeout
> > resp. interruption by a signal from the userspace. Also retry rather
> > than fail when check_pages_isolated sees some !free pages because those
> > could be a result of the race as well.
> > 
> > Signed-off-by: Michal Hocko 
> 
> Yeah, that's better, thanks.
> 
> Acked-by: Vlastimil Babka 

Thanks. I will give it a day and repost the series. If somebody still
have some concerns please speak up.

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-13 Thread Vlastimil Babka 
On 09/13/2017 02:14 PM, Michal Hocko wrote:
 Do you think that the changelog should be more clear about this?
>>>
>>> It certainly wouldn't hurt :)
>>
>> So what do you think about the following wording:
> 
> Ups, wrong patch
> 
> 
> From 8639496a834b4a7c24972ec23b17e50f0d6a304c Mon Sep 17 00:00:00 2001
> From: Michal Hocko 
> Date: Mon, 14 Aug 2017 10:46:12 +0200
> Subject: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early
> 
> Memory offlining can fail just too eagerly under a heavy memory pressure.
> 
> [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> mapping:88ff926c9f38 index:0x3
> [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> [ 5410.336811] page dumped because: isolation failed
> [ 5410.336813] page->mem_cgroup:8801cd662000
> [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
> 
> Isolation has failed here because the page is not on LRU. Most probably
> because it was on the pcp LRU cache or it has been removed from the LRU
> already but it hasn't been freed yet. In both cases the page doesn't look
> non-migrable so retrying more makes sense.
> 
> __offline_pages seems rather cluttered when it comes to the retry
> logic. We have 5 retries at maximum and a timeout. We could argue
> whether the timeout makes sense but failing just because of a race when
> somebody isoltes a page from LRU or puts it on a pcp LRU lists is just
> wrong. It only takes it to race with a process which unmaps some pages
> and remove them from the LRU list and we can fail the whole offline
> because of something that is a temporary condition and actually not
> harmful for the offline.
> 
> Please note that unmovable pages should be already excluded during
> start_isolate_page_range. We could argue that has_unmovable_pages is
> racy and MIGRATE_MOVABLE check doesn't provide any hard guarantee either
> but kernel zones (aka < ZONE_MOVABLE) will very likely detect unmovable
> pages in most cases and movable zone shouldn't contain unmovable pages
> at all. Some of those pages might be pinned but not for ever because
> that would be a bug on its own. In any case the context is still
> interruptible and so the userspace can easily bail out when the
> operation takes too long. This is certainly better behavior than a
> hardcoded retry loop which is racy.
> 
> Fix this by removing the max retry count and only rely on the timeout
> resp. interruption by a signal from the userspace. Also retry rather
> than fail when check_pages_isolated sees some !free pages because those
> could be a result of the race as well.
> 
> Signed-off-by: Michal Hocko 

Yeah, that's better, thanks.

Acked-by: Vlastimil Babka

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-13 Thread Michal Hocko 
On Wed 13-09-17 14:10:01, Michal Hocko wrote:
> On Wed 13-09-17 13:41:20, Vlastimil Babka wrote:
> > On 09/11/2017 10:17 AM, Michal Hocko wrote:
> [...]
> > > Yes, we should be able to distinguish the two and hopefully we can teach
> > > the migration code to distinguish between EBUSY (likely permanent) and
> > > EGAIN (temporal) failure. This sound like something we should aim for
> > > longterm I guess. Anyway as I've said in other email. If somebody really
> > > wants to have a guaratee of a bounded retry then it is trivial to set up
> > > an alarm and send a signal itself to bail out.
> > 
> > Sure, I would just be careful about not breaking existing userspace
> > (udev?) when offline triggered via ACPI from some management interface
> > (or whatever the exact mechanism is).
> 
> The thing is that there is absolutely no timing guarantee even with
> retry limit in place. We are doing allocations, potentially bouncing on
> locks which can be taken elsewhere etc... So if somebody really depend
> on this then it is pretty much broken already.
> 
> > > Do you think that the changelog should be more clear about this?
> > 
> > It certainly wouldn't hurt :)
> 
> So what do you think about the following wording:

Ups, wrong patch


>From 8639496a834b4a7c24972ec23b17e50f0d6a304c Mon Sep 17 00:00:00 2001
From: Michal Hocko 
Date: Mon, 14 Aug 2017 10:46:12 +0200
Subject: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

Memory offlining can fail just too eagerly under a heavy memory pressure.

[ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
mapping:88ff926c9f38 index:0x3
[ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
[ 5410.336811] page dumped because: isolation failed
[ 5410.336813] page->mem_cgroup:8801cd662000
[ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed

Isolation has failed here because the page is not on LRU. Most probably
because it was on the pcp LRU cache or it has been removed from the LRU
already but it hasn't been freed yet. In both cases the page doesn't look
non-migrable so retrying more makes sense.

__offline_pages seems rather cluttered when it comes to the retry
logic. We have 5 retries at maximum and a timeout. We could argue
whether the timeout makes sense but failing just because of a race when
somebody isoltes a page from LRU or puts it on a pcp LRU lists is just
wrong. It only takes it to race with a process which unmaps some pages
and remove them from the LRU list and we can fail the whole offline
because of something that is a temporary condition and actually not
harmful for the offline.

Please note that unmovable pages should be already excluded during
start_isolate_page_range. We could argue that has_unmovable_pages is
racy and MIGRATE_MOVABLE check doesn't provide any hard guarantee either
but kernel zones (aka < ZONE_MOVABLE) will very likely detect unmovable
pages in most cases and movable zone shouldn't contain unmovable pages
at all. Some of those pages might be pinned but not for ever because
that would be a bug on its own. In any case the context is still
interruptible and so the userspace can easily bail out when the
operation takes too long. This is certainly better behavior than a
hardcoded retry loop which is racy.

Fix this by removing the max retry count and only rely on the timeout
resp. interruption by a signal from the userspace. Also retry rather
than fail when check_pages_isolated sees some !free pages because those
could be a result of the race as well.

Signed-off-by: Michal Hocko 

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-13 Thread Michal Hocko 
On Wed 13-09-17 13:41:20, Vlastimil Babka wrote:
> On 09/11/2017 10:17 AM, Michal Hocko wrote:
[...]
> > Yes, we should be able to distinguish the two and hopefully we can teach
> > the migration code to distinguish between EBUSY (likely permanent) and
> > EGAIN (temporal) failure. This sound like something we should aim for
> > longterm I guess. Anyway as I've said in other email. If somebody really
> > wants to have a guaratee of a bounded retry then it is trivial to set up
> > an alarm and send a signal itself to bail out.
> 
> Sure, I would just be careful about not breaking existing userspace
> (udev?) when offline triggered via ACPI from some management interface
> (or whatever the exact mechanism is).

The thing is that there is absolutely no timing guarantee even with
retry limit in place. We are doing allocations, potentially bouncing on
locks which can be taken elsewhere etc... So if somebody really depend
on this then it is pretty much broken already.

> > Do you think that the changelog should be more clear about this?
> 
> It certainly wouldn't hurt :)

So what do you think about the following wording:

commit 23c4ded55c2ba880165a9f5b8a67694361fb6bc7
Author: Michal Hocko 
Date:   Mon Aug 28 13:13:06 2017 +0200

mm, memory_hotplug: remove timeout from __offline_memory

We have a hardcoded 120s timeout after which the memory offline fails
basically since the hot remove has been introduced. This is essentially
a policy implemented in the kernel. Moreover there is no way to adjust
the timeout and so we are sometimes facing memory offline failures if
the system is under a heavy memory pressure or very intensive CPU
workload on large machines.

It is not very clear what purpose the timeout actually serves. The
offline operation is interruptible by a signal so if userspace wants
some timeout based termination this can be done trivially by sending a
signal.

If there is a strong usecase to do this from the kernel then we should
do it properly and have a it tunable from the userspace with the timeout
disabled by default along with the explanation who uses it and for what
purporse.

Acked-by: Vlastimil Babka 
Signed-off-by: Michal Hocko 
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-13 Thread Vlastimil Babka 
On 09/11/2017 10:17 AM, Michal Hocko wrote:
> On Fri 08-09-17 19:26:06, Vlastimil Babka wrote:
>> On 09/04/2017 10:21 AM, Michal Hocko wrote:
>>> From: Michal Hocko 
>>>
>>> Fix this by removing the max retry count and only rely on the timeout
>>> resp. interruption by a signal from the userspace. Also retry rather
>>> than fail when check_pages_isolated sees some !free pages because those
>>> could be a result of the race as well.
>>>
>>> Signed-off-by: Michal Hocko 
>>
>> Even within a movable node where has_unmovable_pages() is a non-issue, you 
>> could
>> have pinned movable pages where the pinning is not temporary.
> 
> Who would pin those pages? Such a page would be unreclaimable as well
> and thus a memory leak and I would argue it would be a bug.

I don't know who exactly, but generally it's a problem for CMA and a
reason why there was some effort from PeterZ to introduce an API for
long-term pinning.

>> So after this
>> patch, this will really keep retrying forever. I'm not saying it's wrong, 
>> just
>> pointing it out, since the changelog seems to assume there would be only
>> temporary failures possible and thus unbound retries are always correct.
>> The obvious problem if we wanted to avoid this, is how to recognize
>> non-temporary failures...
> 
> Yes, we should be able to distinguish the two and hopefully we can teach
> the migration code to distinguish between EBUSY (likely permanent) and
> EGAIN (temporal) failure. This sound like something we should aim for
> longterm I guess. Anyway as I've said in other email. If somebody really
> wants to have a guaratee of a bounded retry then it is trivial to set up
> an alarm and send a signal itself to bail out.

Sure, I would just be careful about not breaking existing userspace
(udev?) when offline triggered via ACPI from some management interface
(or whatever the exact mechanism is).

> Do you think that the changelog should be more clear about this?

It certainly wouldn't hurt :)

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-11 Thread Michal Hocko 
On Fri 08-09-17 19:26:06, Vlastimil Babka wrote:
> On 09/04/2017 10:21 AM, Michal Hocko wrote:
> > From: Michal Hocko 
> > 
> > Memory offlining can fail just too eagerly under a heavy memory pressure.
> > 
> > [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> > mapping:88ff926c9f38 index:0x3
> > [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> > [ 5410.336811] page dumped because: isolation failed
> > [ 5410.336813] page->mem_cgroup:8801cd662000
> > [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
> > 
> > Isolation has failed here because the page is not on LRU. Most probably
> > because it was on the pcp LRU cache or it has been removed from the LRU
> > already but it hasn't been freed yet. In both cases the page doesn't look
> > non-migrable so retrying more makes sense.
> > 
> > __offline_pages seems rather cluttered when it comes to the retry
> > logic. We have 5 retries at maximum and a timeout. We could argue
> > whether the timeout makes sense but failing just because of a race when
> > somebody isoltes a page from LRU or puts it on a pcp LRU lists is just
> > wrong. It only takes it to race with a process which unmaps some pages
> > and remove them from the LRU list and we can fail the whole offline
> > because of something that is a temporary condition and actually not
> > harmful for the offline. Please note that unmovable pages should be
> > already excluded during start_isolate_page_range.
> 
> Hmm, the has_unmovable_pages() check doesn't offer any strict guarantees due 
> to
> races, per its comment. Also at the very quick glance, I see a check where it
> assumes that MIGRATE_MOVABLE pageblock will have no unmovable pages. There is 
> no
> such guarantee even without races.

Yes, you are right that there are races possible but practically
speaking non-movable memblocks (in !MOVABLE_ZONE) would be very likely
to have reliably unmovable pages and so has_unmovable_pages would bail
out. And ZONE_MOVABLE memblocks with permanently pinned pages sound like
a bug to me.

> > Fix this by removing the max retry count and only rely on the timeout
> > resp. interruption by a signal from the userspace. Also retry rather
> > than fail when check_pages_isolated sees some !free pages because those
> > could be a result of the race as well.
> > 
> > Signed-off-by: Michal Hocko 
> 
> Even within a movable node where has_unmovable_pages() is a non-issue, you 
> could
> have pinned movable pages where the pinning is not temporary.

Who would pin those pages? Such a page would be unreclaimable as well
and thus a memory leak and I would argue it would be a bug.

> So after this
> patch, this will really keep retrying forever. I'm not saying it's wrong, just
> pointing it out, since the changelog seems to assume there would be only
> temporary failures possible and thus unbound retries are always correct.
> The obvious problem if we wanted to avoid this, is how to recognize
> non-temporary failures...

Yes, we should be able to distinguish the two and hopefully we can teach
the migration code to distinguish between EBUSY (likely permanent) and
EGAIN (temporal) failure. This sound like something we should aim for
longterm I guess. Anyway as I've said in other email. If somebody really
wants to have a guaratee of a bounded retry then it is trivial to set up
an alarm and send a signal itself to bail out.

Do you think that the changelog should be more clear about this?
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-08 Thread Vlastimil Babka 
On 09/04/2017 10:21 AM, Michal Hocko wrote:
> From: Michal Hocko 
> 
> Memory offlining can fail just too eagerly under a heavy memory pressure.
> 
> [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> mapping:88ff926c9f38 index:0x3
> [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> [ 5410.336811] page dumped because: isolation failed
> [ 5410.336813] page->mem_cgroup:8801cd662000
> [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
> 
> Isolation has failed here because the page is not on LRU. Most probably
> because it was on the pcp LRU cache or it has been removed from the LRU
> already but it hasn't been freed yet. In both cases the page doesn't look
> non-migrable so retrying more makes sense.
> 
> __offline_pages seems rather cluttered when it comes to the retry
> logic. We have 5 retries at maximum and a timeout. We could argue
> whether the timeout makes sense but failing just because of a race when
> somebody isoltes a page from LRU or puts it on a pcp LRU lists is just
> wrong. It only takes it to race with a process which unmaps some pages
> and remove them from the LRU list and we can fail the whole offline
> because of something that is a temporary condition and actually not
> harmful for the offline. Please note that unmovable pages should be
> already excluded during start_isolate_page_range.

Hmm, the has_unmovable_pages() check doesn't offer any strict guarantees due to
races, per its comment. Also at the very quick glance, I see a check where it
assumes that MIGRATE_MOVABLE pageblock will have no unmovable pages. There is no
such guarantee even without races.

> Fix this by removing the max retry count and only rely on the timeout
> resp. interruption by a signal from the userspace. Also retry rather
> than fail when check_pages_isolated sees some !free pages because those
> could be a result of the race as well.
> 
> Signed-off-by: Michal Hocko 

Even within a movable node where has_unmovable_pages() is a non-issue, you could
have pinned movable pages where the pinning is not temporary. So after this
patch, this will really keep retrying forever. I'm not saying it's wrong, just
pointing it out, since the changelog seems to assume there would be only
temporary failures possible and thus unbound retries are always correct.
The obvious problem if we wanted to avoid this, is how to recognize
non-temporary failures...

> ---
>  mm/memory_hotplug.c | 40 ++--
>  1 file changed, 10 insertions(+), 30 deletions(-)
> 
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index 459bbc182d10..c9dcbe6d2ac6 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -1597,7 +1597,7 @@ static int __ref __offline_pages(unsigned long 
> start_pfn,
>  {
>   unsigned long pfn, nr_pages, expire;
>   long offlined_pages;
> - int ret, drain, retry_max, node;
> + int ret, node;
>   unsigned long flags;
>   unsigned long valid_start, valid_end;
>   struct zone *zone;
> @@ -1634,43 +1634,25 @@ static int __ref __offline_pages(unsigned long 
> start_pfn,
>  
>   pfn = start_pfn;
>   expire = jiffies + timeout;
> - drain = 0;
> - retry_max = 5;
>  repeat:
>   /* start memory hot removal */
> - ret = -EAGAIN;
> + ret = -EBUSY;
>   if (time_after(jiffies, expire))
>   goto failed_removal;
>   ret = -EINTR;
>   if (signal_pending(current))
>   goto failed_removal;
> - ret = 0;
> - if (drain) {
> - lru_add_drain_all_cpuslocked();
> - cond_resched();
> - drain_all_pages(zone);
> - }
> +
> + cond_resched();
> + lru_add_drain_all_cpuslocked();
> + drain_all_pages(zone);
>  
>   pfn = scan_movable_pages(start_pfn, end_pfn);
>   if (pfn) { /* We have movable pages */
>   ret = do_migrate_range(pfn, end_pfn);
> - if (!ret) {
> - drain = 1;
> - goto repeat;
> - } else {
> - if (ret < 0)
> - if (--retry_max == 0)
> - goto failed_removal;
> - yield();
> - drain = 1;
> - goto repeat;
> - }
> + goto repeat;
>   }
> - /* drain all zone's lru pagevec, this is asynchronous... */
> - lru_add_drain_all_cpuslocked();
> - yield();
> - /* drain pcp pages, this is synchronous. */
> - drain_all_pages(zone);
> +
>   /*
>* dissolve free hugepages in the memory block before doing offlining
>* actually in order to make hugetlbfs's object counting consistent.
> @@ -1680,10 +1662,8 @@ static int __ref __offline_pages(unsigned long 
> start_pfn,
>   goto failed_removal;
>   /* check again */
>   offlined_pages = check_pages_isolated(start_pfn, end_pfn);
> - if (offlined_

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-05 Thread Michal Hocko 
On Tue 05-09-17 11:59:36, Anshuman Khandual wrote:
[...]
> > @@ -1634,43 +1634,25 @@ static int __ref __offline_pages(unsigned long 
> > start_pfn,
> >  
> > pfn = start_pfn;
> > expire = jiffies + timeout;
> > -   drain = 0;
> > -   retry_max = 5;
> >  repeat:
> > /* start memory hot removal */
> > -   ret = -EAGAIN;
> > +   ret = -EBUSY;
> > if (time_after(jiffies, expire))
> > goto failed_removal;
> > ret = -EINTR;
> > if (signal_pending(current))
> > goto failed_removal;
> > -   ret = 0;
> > -   if (drain) {
> > -   lru_add_drain_all_cpuslocked();
> > -   cond_resched();
> > -   drain_all_pages(zone);
> > -   }
> 
> Why we had this condition before that only when we fail in migration
> later in do_migrate_range function, drain the lru lists in the next
> attempt. Why not from the first attempt itself ? Just being curious.
 
I can only guess but draining used to invoke IPIs and that is really
costly so an optimistic attempt could try without draining and do that
only if the migration fails. Now that we have it all done in WQ context
there shouldn't be any reason to optimize for draining.
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 1/2] mm, memory_hotplug: do not fail offlining too early

2017-09-04 Thread Anshuman Khandual 
On 09/04/2017 01:51 PM, Michal Hocko wrote:
> From: Michal Hocko 
> 
> Memory offlining can fail just too eagerly under a heavy memory pressure.
> 
> [ 5410.336792] page:ea22a646bd00 count:255 mapcount:252 
> mapping:88ff926c9f38 index:0x3
> [ 5410.336809] flags: 0x9855fe40010048(uptodate|active|mappedtodisk)
> [ 5410.336811] page dumped because: isolation failed
> [ 5410.336813] page->mem_cgroup:8801cd662000
> [ 5420.655030] memory offlining [mem 0x18b58000-0x18b5] failed
> 
> Isolation has failed here because the page is not on LRU. Most probably
> because it was on the pcp LRU cache or it has been removed from the LRU
> already but it hasn't been freed yet. In both cases the page doesn't look
> non-migrable so retrying more makes sense.
> 
> __offline_pages seems rather cluttered when it comes to the retry
> logic. We have 5 retries at maximum and a timeout. We could argue
> whether the timeout makes sense but failing just because of a race when
> somebody isoltes a page from LRU or puts it on a pcp LRU lists is just
> wrong. It only takes it to race with a process which unmaps some pages
> and remove them from the LRU list and we can fail the whole offline
> because of something that is a temporary condition and actually not
> harmful for the offline. Please note that unmovable pages should be
> already excluded during start_isolate_page_range.
> 
> Fix this by removing the max retry count and only rely on the timeout
> resp. interruption by a signal from the userspace. Also retry rather
> than fail when check_pages_isolated sees some !free pages because those
> could be a result of the race as well.
> 
> Signed-off-by: Michal Hocko 
> ---
>  mm/memory_hotplug.c | 40 ++--
>  1 file changed, 10 insertions(+), 30 deletions(-)
> 
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index 459bbc182d10..c9dcbe6d2ac6 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -1597,7 +1597,7 @@ static int __ref __offline_pages(unsigned long 
> start_pfn,
>  {
>   unsigned long pfn, nr_pages, expire;
>   long offlined_pages;
> - int ret, drain, retry_max, node;
> + int ret, node;
>   unsigned long flags;
>   unsigned long valid_start, valid_end;
>   struct zone *zone;
> @@ -1634,43 +1634,25 @@ static int __ref __offline_pages(unsigned long 
> start_pfn,
>  
>   pfn = start_pfn;
>   expire = jiffies + timeout;
> - drain = 0;
> - retry_max = 5;
>  repeat:
>   /* start memory hot removal */
> - ret = -EAGAIN;
> + ret = -EBUSY;
>   if (time_after(jiffies, expire))
>   goto failed_removal;
>   ret = -EINTR;
>   if (signal_pending(current))
>   goto failed_removal;
> - ret = 0;
> - if (drain) {
> - lru_add_drain_all_cpuslocked();
> - cond_resched();
> - drain_all_pages(zone);
> - }

Why we had this condition before that only when we fail in migration
later in do_migrate_range function, drain the lru lists in the next
attempt. Why not from the first attempt itself ? Just being curious.