diff mbox series

[v10,4/4] mm: hugetlb_vmemmap: add hugetlb_optimize_vmemmap sysctl

Message ID 20220509062703.64249-5-songmuchun@bytedance.com (mailing list archive)
State New
Headers show
Series add hugetlb_optimize_vmemmap sysctl | expand

Commit Message

Muchun Song May 9, 2022, 6:27 a.m. UTC
We must add hugetlb_free_vmemmap=on (or "off") to the boot cmdline and
reboot the server to enable or disable the feature of optimizing vmemmap
pages associated with HugeTLB pages.  However, rebooting usually takes a
long time.  So add a sysctl to enable or disable the feature at runtime
without rebooting.  Why we need this?  There are 3 use cases.

1) The feature of minimizing overhead of struct page associated with each
HugeTLB is disabled by default without passing "hugetlb_free_vmemmap=on"
to the boot cmdline. When we (ByteDance) deliver the servers to the
users who want to enable this feature, they have to configure the grub
(change boot cmdline) and reboot the servers, whereas rebooting usually
takes a long time (we have thousands of servers).  It's a very bad
experience for the users.  So we need a approach to enable this feature
after rebooting. This is a use case in our practical environment.

2) Some use cases are that HugeTLB pages are allocated 'on the fly'
instead of being pulled from the HugeTLB pool, those workloads would be
affected with this feature enabled.  Those workloads could be identified
by the characteristics of they never explicitly allocating huge pages
with 'nr_hugepages' but only set 'nr_overcommit_hugepages' and then let
the pages be allocated from the buddy allocator at fault time.  We can
confirm it is a real use case from the commit 099730d67417.  For those
workloads, the page fault time could be ~2x slower than before. We
suspect those users want to disable this feature if the system has enabled
this before and they don't think the memory savings benefit is enough to
make up for the performance drop.

3) If the workload which wants vmemmap pages to be optimized and the
workload which wants to set 'nr_overcommit_hugepages' and does not want
the extera overhead at fault time when the overcommitted pages be
allocated from the buddy allocator are deployed in the same server.
The user could enable this feature and set 'nr_hugepages' and
'nr_overcommit_hugepages', then disable the feature.  In this case,
the overcommited HugeTLB pages will not encounter the extra overhead
at fault time.

Signed-off-by: Muchun Song <songmuchun@bytedance.com>
---
 Documentation/admin-guide/sysctl/vm.rst | 39 ++++++++++++++
 include/linux/memory_hotplug.h          |  9 ++++
 mm/hugetlb.c                            |  3 ++
 mm/hugetlb_vmemmap.c                    | 93 +++++++++++++++++++++++++++++----
 mm/memory_hotplug.c                     |  7 +--
 5 files changed, 136 insertions(+), 15 deletions(-)

Comments

Mike Kravetz May 10, 2022, 9:30 p.m. UTC | #1
On 5/8/22 23:27, Muchun Song wrote:
> diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h
> index 029fb7e26504..917112661b5c 100644
> --- a/include/linux/memory_hotplug.h
> +++ b/include/linux/memory_hotplug.h
> @@ -351,4 +351,13 @@ void arch_remove_linear_mapping(u64 start, u64 size);
>  extern bool mhp_supports_memmap_on_memory(unsigned long size);
>  #endif /* CONFIG_MEMORY_HOTPLUG */
>  
> +#ifdef CONFIG_MHP_MEMMAP_ON_MEMORY
> +bool mhp_memmap_on_memory(void);
> +#else
> +static inline bool mhp_memmap_on_memory(void)
> +{
> +	return false;
> +}
> +#endif
> +
>  #endif /* __LINUX_MEMORY_HOTPLUG_H */
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index 8605d7eb7f5c..86158eb9da70 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -1617,6 +1617,9 @@ static DECLARE_WORK(free_hpage_work, free_hpage_workfn);
>  
>  static inline void flush_free_hpage_work(struct hstate *h)
>  {
> +	if (!hugetlb_optimize_vmemmap_enabled())
> +		return;
> +

Hi Muchun,

In v9 I was suggesting that we may be able to eliminate the static_branch_inc/dec from the vmemmap free/alloc paths.  With this patch
I believe hugetlb_optimize_vmemmap_enabled() is really checking
'has hugetlb vmemmap optimization been enabled' OR 'are there still vmemmap
optimized hugetlb pages in the system'.  That may be confusing.

For this specific routine (flush_free_hpage_work) I do not think we need
to worry too much about deciding to call flush_work or not.  This is only
called via set_max_huge_pages which is not a performance sensitive path.

>  	if (hugetlb_optimize_vmemmap_pages(h))
>  		flush_work(&free_hpage_work);
>  }

Here is a patch on top of this patch to show my suggestion for removing
static_branch_inc/dec from the vmemmap free/alloc paths.  It seems simpler
to me, and hugetlb_optimize_vmemmap_enabled would only return true if
hugetlb vmemmap optimization is currently enabled.  I am not insisting
that static_branch_inc/dec be eliminated.  It may not even work.  I have
not tested.  What do you think?

diff --git a/arch/arm64/mm/flush.c b/arch/arm64/mm/flush.c
index fc4f710e9820..2f80751b7c3a 100644
--- a/arch/arm64/mm/flush.c
+++ b/arch/arm64/mm/flush.c
@@ -9,6 +9,7 @@
 #include <linux/export.h>
 #include <linux/mm.h>
 #include <linux/pagemap.h>
+#include <linux/hugetlb.h>
 
 #include <asm/cacheflush.h>
 #include <asm/cache.h>
@@ -86,7 +87,7 @@ void flush_dcache_page(struct page *page)
 	 * is reused (more details can refer to the comments above
 	 * page_fixed_fake_head()).
 	 */
-	if (hugetlb_optimize_vmemmap_enabled() && PageHuge(page))
+	if (PageHuge(page) && HPageVmemmapOptimized(compound_head(page)))
 		page = compound_head(page);
 
 	if (test_bit(PG_dcache_clean, &page->flags))
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 86158eb9da70..8605d7eb7f5c 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1617,9 +1617,6 @@ static DECLARE_WORK(free_hpage_work, free_hpage_workfn);
 
 static inline void flush_free_hpage_work(struct hstate *h)
 {
-	if (!hugetlb_optimize_vmemmap_enabled())
-		return;
-
 	if (hugetlb_optimize_vmemmap_pages(h))
 		flush_work(&free_hpage_work);
 }
diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c
index fcd9f7872064..8e0890a505b3 100644
--- a/mm/hugetlb_vmemmap.c
+++ b/mm/hugetlb_vmemmap.c
@@ -41,9 +41,9 @@ static void vmemmap_optimize_mode_switch(enum vmemmap_optimize_mode to)
 		return;
 
 	if (to == VMEMMAP_OPTIMIZE_OFF)
-		static_branch_dec(&hugetlb_optimize_vmemmap_key);
+		static_branch_disable(&hugetlb_optimize_vmemmap_key);
 	else
-		static_branch_inc(&hugetlb_optimize_vmemmap_key);
+		static_branch_enable(&hugetlb_optimize_vmemmap_key);
 	WRITE_ONCE(vmemmap_optimize_mode, to);
 }
 
@@ -91,7 +91,6 @@ int hugetlb_vmemmap_alloc(struct hstate *h, struct page *head)
 				  GFP_KERNEL | __GFP_NORETRY | __GFP_THISNODE);
 	if (!ret) {
 		ClearHPageVmemmapOptimized(head);
-		static_branch_dec(&hugetlb_optimize_vmemmap_key);
 	}
 
 	return ret;
@@ -102,14 +101,10 @@ void hugetlb_vmemmap_free(struct hstate *h, struct page *head)
 	unsigned long vmemmap_addr = (unsigned long)head;
 	unsigned long vmemmap_end, vmemmap_reuse, vmemmap_pages;
 
-	vmemmap_pages = hugetlb_optimize_vmemmap_pages(h);
-	if (!vmemmap_pages)
-		return;
-
-	if (READ_ONCE(vmemmap_optimize_mode) == VMEMMAP_OPTIMIZE_OFF)
+	if (!hugetlb_optimize_vmemmap_enabled())
 		return;
 
-	static_branch_inc(&hugetlb_optimize_vmemmap_key);
+	vmemmap_pages = hugetlb_optimize_vmemmap_pages(h);
 
 	vmemmap_addr	+= RESERVE_VMEMMAP_SIZE;
 	vmemmap_end	= vmemmap_addr + (vmemmap_pages << PAGE_SHIFT);
@@ -120,9 +115,7 @@ void hugetlb_vmemmap_free(struct hstate *h, struct page *head)
 	 * to the page which @vmemmap_reuse is mapped to, then free the pages
 	 * which the range [@vmemmap_addr, @vmemmap_end] is mapped to.
 	 */
-	if (vmemmap_remap_free(vmemmap_addr, vmemmap_end, vmemmap_reuse))
-		static_branch_dec(&hugetlb_optimize_vmemmap_key);
-	else
+	if (!vmemmap_remap_free(vmemmap_addr, vmemmap_end, vmemmap_reuse))
 		SetHPageVmemmapOptimized(head);
 }
Mike Kravetz May 11, 2022, 12:39 a.m. UTC | #2
On 5/10/22 14:30, Mike Kravetz wrote:
> On 5/8/22 23:27, Muchun Song wrote:
>> diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h
>> index 029fb7e26504..917112661b5c 100644
>> --- a/include/linux/memory_hotplug.h
>> +++ b/include/linux/memory_hotplug.h
>> @@ -351,4 +351,13 @@ void arch_remove_linear_mapping(u64 start, u64 size);
>>  extern bool mhp_supports_memmap_on_memory(unsigned long size);
>>  #endif /* CONFIG_MEMORY_HOTPLUG */
>>  
>> +#ifdef CONFIG_MHP_MEMMAP_ON_MEMORY
>> +bool mhp_memmap_on_memory(void);
>> +#else
>> +static inline bool mhp_memmap_on_memory(void)
>> +{
>> +	return false;
>> +}
>> +#endif
>> +
>>  #endif /* __LINUX_MEMORY_HOTPLUG_H */
>> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
>> index 8605d7eb7f5c..86158eb9da70 100644
>> --- a/mm/hugetlb.c
>> +++ b/mm/hugetlb.c
>> @@ -1617,6 +1617,9 @@ static DECLARE_WORK(free_hpage_work, free_hpage_workfn);
>>  
>>  static inline void flush_free_hpage_work(struct hstate *h)
>>  {
>> +	if (!hugetlb_optimize_vmemmap_enabled())
>> +		return;
>> +
> 
> Hi Muchun,
> 
> In v9 I was suggesting that we may be able to eliminate the static_branch_inc/dec from the vmemmap free/alloc paths.  With this patch
> I believe hugetlb_optimize_vmemmap_enabled() is really checking
> 'has hugetlb vmemmap optimization been enabled' OR 'are there still vmemmap
> optimized hugetlb pages in the system'.  That may be confusing.
> 

Sorry, I forgot about the use of hugetlb_optimize_vmemmap_enabled in
page_fixed_fake_head.  We need to know if there are any vmemmap optimized
hugetlb pages in the system in this performance sensitive path.  So,
static_branch_inc/dec is indeed a good idea.

Please disregard my attempt below at removing static_branch_inc/dec.

I still find the name hugetlb_optimize_vmemmap_enabled a bit confusing as
it tests two conditions (enabled and pages in use).

You have already 'open coded' just the check for enabled in the routine
hugetlb_vmemmap_free with:

	READ_ONCE(vmemmap_optimize_mode) == VMEMMAP_OPTIMIZE_OFF

How about having hugetlb_optimize_vmemmap_enabled() just check
vmemmap_optimize_mode in a manner like above?  Then rename
hugetlb_optimize_vmemmap_enabled to something like:
hugetlb_optimized_vmemmap_possible().  Sorry, I can think if a great name.
Muchun Song May 11, 2022, 9:45 a.m. UTC | #3
On Tue, May 10, 2022 at 05:39:40PM -0700, Mike Kravetz wrote:
> On 5/10/22 14:30, Mike Kravetz wrote:
> > On 5/8/22 23:27, Muchun Song wrote:
> >> diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h
> >> index 029fb7e26504..917112661b5c 100644
> >> --- a/include/linux/memory_hotplug.h
> >> +++ b/include/linux/memory_hotplug.h
> >> @@ -351,4 +351,13 @@ void arch_remove_linear_mapping(u64 start, u64 size);
> >>  extern bool mhp_supports_memmap_on_memory(unsigned long size);
> >>  #endif /* CONFIG_MEMORY_HOTPLUG */
> >>  
> >> +#ifdef CONFIG_MHP_MEMMAP_ON_MEMORY
> >> +bool mhp_memmap_on_memory(void);
> >> +#else
> >> +static inline bool mhp_memmap_on_memory(void)
> >> +{
> >> +	return false;
> >> +}
> >> +#endif
> >> +
> >>  #endif /* __LINUX_MEMORY_HOTPLUG_H */
> >> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> >> index 8605d7eb7f5c..86158eb9da70 100644
> >> --- a/mm/hugetlb.c
> >> +++ b/mm/hugetlb.c
> >> @@ -1617,6 +1617,9 @@ static DECLARE_WORK(free_hpage_work, free_hpage_workfn);
> >>  
> >>  static inline void flush_free_hpage_work(struct hstate *h)
> >>  {
> >> +	if (!hugetlb_optimize_vmemmap_enabled())
> >> +		return;
> >> +
> > 
> > Hi Muchun,
> > 
> > In v9 I was suggesting that we may be able to eliminate the static_branch_inc/dec from the vmemmap free/alloc paths.  With this patch
> > I believe hugetlb_optimize_vmemmap_enabled() is really checking
> > 'has hugetlb vmemmap optimization been enabled' OR 'are there still vmemmap
> > optimized hugetlb pages in the system'.  That may be confusing.
> > 
> 
> Sorry, I forgot about the use of hugetlb_optimize_vmemmap_enabled in
> page_fixed_fake_head.  We need to know if there are any vmemmap optimized
> hugetlb pages in the system in this performance sensitive path.  So,
> static_branch_inc/dec is indeed a good idea.
>

Agree.

> Please disregard my attempt below at removing static_branch_inc/dec.
> 
> I still find the name hugetlb_optimize_vmemmap_enabled a bit confusing as
> it tests two conditions (enabled and pages in use).
>

Right. It tests two conditions.

> You have already 'open coded' just the check for enabled in the routine
> hugetlb_vmemmap_free with:
> 
> 	READ_ONCE(vmemmap_optimize_mode) == VMEMMAP_OPTIMIZE_OFF
> 
> How about having hugetlb_optimize_vmemmap_enabled() just check
> vmemmap_optimize_mode in a manner like above?  Then rename

I'm wondering is it necessary to do this? vmemmap_optimize_mode
is a internal state in hugetlb_vmemmap.c, at leaset now there is
no outside users who care about this.  Open-coding may be not
an issue (I guess)?  If one day someone cares it, maybe it it
the time to do this and rename hugetlb_optimize_vmemmap_enabled()?
I'm not against doing this, just expressing some of my thoughts.
What do you think, Mike?

> hugetlb_optimize_vmemmap_enabled to something like:
> hugetlb_optimized_vmemmap_possible().  Sorry, I can think if a great name.
> 

At least I cannot come up with an appropriate name.
hugetlb_optimize_vmemmap_may_enabled()? It's not easy to come
up with a good name.

Thanks.
Muchun Song May 11, 2022, 10:57 a.m. UTC | #4
On Wed, May 11, 2022 at 05:45:57PM +0800, Muchun Song wrote:
> On Tue, May 10, 2022 at 05:39:40PM -0700, Mike Kravetz wrote:
> > On 5/10/22 14:30, Mike Kravetz wrote:
> > > On 5/8/22 23:27, Muchun Song wrote:
> > >> diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h
> > >> index 029fb7e26504..917112661b5c 100644
> > >> --- a/include/linux/memory_hotplug.h
> > >> +++ b/include/linux/memory_hotplug.h
> > >> @@ -351,4 +351,13 @@ void arch_remove_linear_mapping(u64 start, u64 size);
> > >>  extern bool mhp_supports_memmap_on_memory(unsigned long size);
> > >>  #endif /* CONFIG_MEMORY_HOTPLUG */
> > >>  
> > >> +#ifdef CONFIG_MHP_MEMMAP_ON_MEMORY
> > >> +bool mhp_memmap_on_memory(void);
> > >> +#else
> > >> +static inline bool mhp_memmap_on_memory(void)
> > >> +{
> > >> +	return false;
> > >> +}
> > >> +#endif
> > >> +
> > >>  #endif /* __LINUX_MEMORY_HOTPLUG_H */
> > >> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> > >> index 8605d7eb7f5c..86158eb9da70 100644
> > >> --- a/mm/hugetlb.c
> > >> +++ b/mm/hugetlb.c
> > >> @@ -1617,6 +1617,9 @@ static DECLARE_WORK(free_hpage_work, free_hpage_workfn);
> > >>  
> > >>  static inline void flush_free_hpage_work(struct hstate *h)
> > >>  {
> > >> +	if (!hugetlb_optimize_vmemmap_enabled())
> > >> +		return;
> > >> +
> > > 
> > > Hi Muchun,
> > > 
> > > In v9 I was suggesting that we may be able to eliminate the static_branch_inc/dec from the vmemmap free/alloc paths.  With this patch
> > > I believe hugetlb_optimize_vmemmap_enabled() is really checking
> > > 'has hugetlb vmemmap optimization been enabled' OR 'are there still vmemmap
> > > optimized hugetlb pages in the system'.  That may be confusing.
> > > 
> > 
> > Sorry, I forgot about the use of hugetlb_optimize_vmemmap_enabled in
> > page_fixed_fake_head.  We need to know if there are any vmemmap optimized
> > hugetlb pages in the system in this performance sensitive path.  So,
> > static_branch_inc/dec is indeed a good idea.
> >
> 
> Agree.
> 
> > Please disregard my attempt below at removing static_branch_inc/dec.
> > 
> > I still find the name hugetlb_optimize_vmemmap_enabled a bit confusing as
> > it tests two conditions (enabled and pages in use).
> >
> 
> Right. It tests two conditions.
> 
> > You have already 'open coded' just the check for enabled in the routine
> > hugetlb_vmemmap_free with:
> > 
> > 	READ_ONCE(vmemmap_optimize_mode) == VMEMMAP_OPTIMIZE_OFF
> > 
> > How about having hugetlb_optimize_vmemmap_enabled() just check
> > vmemmap_optimize_mode in a manner like above?  Then rename
> 
> I'm wondering is it necessary to do this? vmemmap_optimize_mode
> is a internal state in hugetlb_vmemmap.c, at leaset now there is
> no outside users who care about this.  Open-coding may be not
> an issue (I guess)?  If one day someone cares it, maybe it it
> the time to do this and rename hugetlb_optimize_vmemmap_enabled()?
> I'm not against doing this, just expressing some of my thoughts.
> What do you think, Mike?
> 
> > hugetlb_optimize_vmemmap_enabled to something like:
> > hugetlb_optimized_vmemmap_possible().  Sorry, I can think if a great name.
> > 
> 
> At least I cannot come up with an appropriate name.
> hugetlb_optimize_vmemmap_may_enabled()? It's not easy to come
> up with a good name.
>

Instead of renaming, how about remove hugetlb_optimize_vmemmap_enabled()
directly?  I found there are only two places (mm/memory_hotplug.c and
arch/arm64/mm/flush.c) except include/linux/page-flags.h where use this
helper.

In arch/arm64/mm/flush.c, we could replace it with

  if (PageHuge(page) && HPageVmemmapOptimized(compound_head(page)))

In mm/memory_hotplug.c, I have a plan to remove it as well (I'll
post them out after this patch merged).

Finally, there is no outside users of it, we could remove it and squash
it into page_fixed_fake_head(). What do you think this, Mike?
Mike Kravetz May 11, 2022, 5:53 p.m. UTC | #5
On 5/11/22 03:57, Muchun Song wrote:
> On Wed, May 11, 2022 at 05:45:57PM +0800, Muchun Song wrote:
>> On Tue, May 10, 2022 at 05:39:40PM -0700, Mike Kravetz wrote:
>>> On 5/10/22 14:30, Mike Kravetz wrote:
>>>> On 5/8/22 23:27, Muchun Song wrote:
>>>>> diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h
>>>>> index 029fb7e26504..917112661b5c 100644
>>>>> --- a/include/linux/memory_hotplug.h
>>>>> +++ b/include/linux/memory_hotplug.h
>>>>> @@ -351,4 +351,13 @@ void arch_remove_linear_mapping(u64 start, u64 size);
>>>>>  extern bool mhp_supports_memmap_on_memory(unsigned long size);
>>>>>  #endif /* CONFIG_MEMORY_HOTPLUG */
>>>>>  
>>>>> +#ifdef CONFIG_MHP_MEMMAP_ON_MEMORY
>>>>> +bool mhp_memmap_on_memory(void);
>>>>> +#else
>>>>> +static inline bool mhp_memmap_on_memory(void)
>>>>> +{
>>>>> +	return false;
>>>>> +}
>>>>> +#endif
>>>>> +
>>>>>  #endif /* __LINUX_MEMORY_HOTPLUG_H */
>>>>> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
>>>>> index 8605d7eb7f5c..86158eb9da70 100644
>>>>> --- a/mm/hugetlb.c
>>>>> +++ b/mm/hugetlb.c
>>>>> @@ -1617,6 +1617,9 @@ static DECLARE_WORK(free_hpage_work, free_hpage_workfn);
>>>>>  
>>>>>  static inline void flush_free_hpage_work(struct hstate *h)
>>>>>  {
>>>>> +	if (!hugetlb_optimize_vmemmap_enabled())
>>>>> +		return;
>>>>> +
>>>>
>>>> Hi Muchun,
>>>>
>>>> In v9 I was suggesting that we may be able to eliminate the static_branch_inc/dec from the vmemmap free/alloc paths.  With this patch
>>>> I believe hugetlb_optimize_vmemmap_enabled() is really checking
>>>> 'has hugetlb vmemmap optimization been enabled' OR 'are there still vmemmap
>>>> optimized hugetlb pages in the system'.  That may be confusing.
>>>>
>>>
>>> Sorry, I forgot about the use of hugetlb_optimize_vmemmap_enabled in
>>> page_fixed_fake_head.  We need to know if there are any vmemmap optimized
>>> hugetlb pages in the system in this performance sensitive path.  So,
>>> static_branch_inc/dec is indeed a good idea.
>>>
>>
>> Agree.
>>
>>> Please disregard my attempt below at removing static_branch_inc/dec.
>>>
>>> I still find the name hugetlb_optimize_vmemmap_enabled a bit confusing as
>>> it tests two conditions (enabled and pages in use).
>>>
>>
>> Right. It tests two conditions.
>>
>>> You have already 'open coded' just the check for enabled in the routine
>>> hugetlb_vmemmap_free with:
>>>
>>> 	READ_ONCE(vmemmap_optimize_mode) == VMEMMAP_OPTIMIZE_OFF
>>>
>>> How about having hugetlb_optimize_vmemmap_enabled() just check
>>> vmemmap_optimize_mode in a manner like above?  Then rename
>>
>> I'm wondering is it necessary to do this? vmemmap_optimize_mode
>> is a internal state in hugetlb_vmemmap.c, at leaset now there is
>> no outside users who care about this.  Open-coding may be not
>> an issue (I guess)?  If one day someone cares it, maybe it it
>> the time to do this and rename hugetlb_optimize_vmemmap_enabled()?
>> I'm not against doing this, just expressing some of my thoughts.
>> What do you think, Mike?
>>
>>> hugetlb_optimize_vmemmap_enabled to something like:
>>> hugetlb_optimized_vmemmap_possible().  Sorry, I can think if a great name.
>>>
>>
>> At least I cannot come up with an appropriate name.
>> hugetlb_optimize_vmemmap_may_enabled()? It's not easy to come
>> up with a good name.
>>
> 
> Instead of renaming, how about remove hugetlb_optimize_vmemmap_enabled()
> directly?  I found there are only two places (mm/memory_hotplug.c and
> arch/arm64/mm/flush.c) except include/linux/page-flags.h where use this
> helper.
> 
> In arch/arm64/mm/flush.c, we could replace it with
> 
>   if (PageHuge(page) && HPageVmemmapOptimized(compound_head(page)))
> 
> In mm/memory_hotplug.c, I have a plan to remove it as well (I'll
> post them out after this patch merged).
> 
> Finally, there is no outside users of it, we could remove it and squash
> it into page_fixed_fake_head(). What do you think this, Mike?

That sounds good.

Sorry for all the questions/suggestions around
hugetlb_optimize_vmemmap_enabled.  It just caused me a little confusion
as it is providing information on two conditions.  I wanted to prevent it
from causing confusion for others reading the code in the future.

This patch as written is fine with plans for a follow up to remove
hugetlb_optimize_vmemmap_enabled.

Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Muchun Song May 12, 2022, 3:34 a.m. UTC | #6
On Wed, May 11, 2022 at 10:53:07AM -0700, Mike Kravetz wrote:
> On 5/11/22 03:57, Muchun Song wrote:
> > On Wed, May 11, 2022 at 05:45:57PM +0800, Muchun Song wrote:
> >> On Tue, May 10, 2022 at 05:39:40PM -0700, Mike Kravetz wrote:
> >>> On 5/10/22 14:30, Mike Kravetz wrote:
> >>>> On 5/8/22 23:27, Muchun Song wrote:
> >>>>> diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h
> >>>>> index 029fb7e26504..917112661b5c 100644
> >>>>> --- a/include/linux/memory_hotplug.h
> >>>>> +++ b/include/linux/memory_hotplug.h
> >>>>> @@ -351,4 +351,13 @@ void arch_remove_linear_mapping(u64 start, u64 size);
> >>>>>  extern bool mhp_supports_memmap_on_memory(unsigned long size);
> >>>>>  #endif /* CONFIG_MEMORY_HOTPLUG */
> >>>>>  
> >>>>> +#ifdef CONFIG_MHP_MEMMAP_ON_MEMORY
> >>>>> +bool mhp_memmap_on_memory(void);
> >>>>> +#else
> >>>>> +static inline bool mhp_memmap_on_memory(void)
> >>>>> +{
> >>>>> +	return false;
> >>>>> +}
> >>>>> +#endif
> >>>>> +
> >>>>>  #endif /* __LINUX_MEMORY_HOTPLUG_H */
> >>>>> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> >>>>> index 8605d7eb7f5c..86158eb9da70 100644
> >>>>> --- a/mm/hugetlb.c
> >>>>> +++ b/mm/hugetlb.c
> >>>>> @@ -1617,6 +1617,9 @@ static DECLARE_WORK(free_hpage_work, free_hpage_workfn);
> >>>>>  
> >>>>>  static inline void flush_free_hpage_work(struct hstate *h)
> >>>>>  {
> >>>>> +	if (!hugetlb_optimize_vmemmap_enabled())
> >>>>> +		return;
> >>>>> +
> >>>>
> >>>> Hi Muchun,
> >>>>
> >>>> In v9 I was suggesting that we may be able to eliminate the static_branch_inc/dec from the vmemmap free/alloc paths.  With this patch
> >>>> I believe hugetlb_optimize_vmemmap_enabled() is really checking
> >>>> 'has hugetlb vmemmap optimization been enabled' OR 'are there still vmemmap
> >>>> optimized hugetlb pages in the system'.  That may be confusing.
> >>>>
> >>>
> >>> Sorry, I forgot about the use of hugetlb_optimize_vmemmap_enabled in
> >>> page_fixed_fake_head.  We need to know if there are any vmemmap optimized
> >>> hugetlb pages in the system in this performance sensitive path.  So,
> >>> static_branch_inc/dec is indeed a good idea.
> >>>
> >>
> >> Agree.
> >>
> >>> Please disregard my attempt below at removing static_branch_inc/dec.
> >>>
> >>> I still find the name hugetlb_optimize_vmemmap_enabled a bit confusing as
> >>> it tests two conditions (enabled and pages in use).
> >>>
> >>
> >> Right. It tests two conditions.
> >>
> >>> You have already 'open coded' just the check for enabled in the routine
> >>> hugetlb_vmemmap_free with:
> >>>
> >>> 	READ_ONCE(vmemmap_optimize_mode) == VMEMMAP_OPTIMIZE_OFF
> >>>
> >>> How about having hugetlb_optimize_vmemmap_enabled() just check
> >>> vmemmap_optimize_mode in a manner like above?  Then rename
> >>
> >> I'm wondering is it necessary to do this? vmemmap_optimize_mode
> >> is a internal state in hugetlb_vmemmap.c, at leaset now there is
> >> no outside users who care about this.  Open-coding may be not
> >> an issue (I guess)?  If one day someone cares it, maybe it it
> >> the time to do this and rename hugetlb_optimize_vmemmap_enabled()?
> >> I'm not against doing this, just expressing some of my thoughts.
> >> What do you think, Mike?
> >>
> >>> hugetlb_optimize_vmemmap_enabled to something like:
> >>> hugetlb_optimized_vmemmap_possible().  Sorry, I can think if a great name.
> >>>
> >>
> >> At least I cannot come up with an appropriate name.
> >> hugetlb_optimize_vmemmap_may_enabled()? It's not easy to come
> >> up with a good name.
> >>
> > 
> > Instead of renaming, how about remove hugetlb_optimize_vmemmap_enabled()
> > directly?  I found there are only two places (mm/memory_hotplug.c and
> > arch/arm64/mm/flush.c) except include/linux/page-flags.h where use this
> > helper.
> > 
> > In arch/arm64/mm/flush.c, we could replace it with
> > 
> >   if (PageHuge(page) && HPageVmemmapOptimized(compound_head(page)))
> > 
> > In mm/memory_hotplug.c, I have a plan to remove it as well (I'll
> > post them out after this patch merged).
> > 
> > Finally, there is no outside users of it, we could remove it and squash
> > it into page_fixed_fake_head(). What do you think this, Mike?
> 
> That sounds good.
> 
> Sorry for all the questions/suggestions around
> hugetlb_optimize_vmemmap_enabled.  It just caused me a little confusion
> as it is providing information on two conditions.  I wanted to prevent it
> from causing confusion for others reading the code in the future.
>

Sorry for the confusing. I'll post the subsequent patches ASAP.

> This patch as written is fine with plans for a follow up to remove
> hugetlb_optimize_vmemmap_enabled.
> 
> Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>

Thanks Mike.
diff mbox series

Patch

diff --git a/Documentation/admin-guide/sysctl/vm.rst b/Documentation/admin-guide/sysctl/vm.rst
index 747e325ebcd0..5c9aa171a0d3 100644
--- a/Documentation/admin-guide/sysctl/vm.rst
+++ b/Documentation/admin-guide/sysctl/vm.rst
@@ -562,6 +562,45 @@  Change the minimum size of the hugepage pool.
 See Documentation/admin-guide/mm/hugetlbpage.rst
 
 
+hugetlb_optimize_vmemmap
+========================
+
+This knob is not available when memory_hotplug.memmap_on_memory (kernel parameter)
+is configured or the size of 'struct page' (a structure defined in
+include/linux/mm_types.h) is not power of two (an unusual system config could
+result in this).
+
+Enable (set to 1) or disable (set to 0) the feature of optimizing vmemmap pages
+associated with each HugeTLB page.
+
+Once enabled, the vmemmap pages of subsequent allocation of HugeTLB pages from
+buddy allocator will be optimized (7 pages per 2MB HugeTLB page and 4095 pages
+per 1GB HugeTLB page), whereas already allocated HugeTLB pages will not be
+optimized.  When those optimized HugeTLB pages are freed from the HugeTLB pool
+to the buddy allocator, the vmemmap pages representing that range needs to be
+remapped again and the vmemmap pages discarded earlier need to be rellocated
+again.  If your use case is that HugeTLB pages are allocated 'on the fly' (e.g.
+never explicitly allocating HugeTLB pages with 'nr_hugepages' but only set
+'nr_overcommit_hugepages', those overcommitted HugeTLB pages are allocated 'on
+the fly') instead of being pulled from the HugeTLB pool, you should weigh the
+benefits of memory savings against the more overhead (~2x slower than before)
+of allocation or freeing HugeTLB pages between the HugeTLB pool and the buddy
+allocator.  Another behavior to note is that if the system is under heavy memory
+pressure, it could prevent the user from freeing HugeTLB pages from the HugeTLB
+pool to the buddy allocator since the allocation of vmemmap pages could be
+failed, you have to retry later if your system encounter this situation.
+
+Once disabled, the vmemmap pages of subsequent allocation of HugeTLB pages from
+buddy allocator will not be optimized meaning the extra overhead at allocation
+time from buddy allocator disappears, whereas already optimized HugeTLB pages
+will not be affected.  If you want to make sure there are no optimized HugeTLB
+pages, you can set "nr_hugepages" to 0 first and then disable this.  Note that
+writing 0 to nr_hugepages will make any "in use" HugeTLB pages become surplus
+pages.  So, those surplus pages are still optimized until they are no longer
+in use.  You would need to wait for those surplus pages to be released before
+there are no optimized pages in the system.
+
+
 nr_hugepages_mempolicy
 ======================
 
diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h
index 029fb7e26504..917112661b5c 100644
--- a/include/linux/memory_hotplug.h
+++ b/include/linux/memory_hotplug.h
@@ -351,4 +351,13 @@  void arch_remove_linear_mapping(u64 start, u64 size);
 extern bool mhp_supports_memmap_on_memory(unsigned long size);
 #endif /* CONFIG_MEMORY_HOTPLUG */
 
+#ifdef CONFIG_MHP_MEMMAP_ON_MEMORY
+bool mhp_memmap_on_memory(void);
+#else
+static inline bool mhp_memmap_on_memory(void)
+{
+	return false;
+}
+#endif
+
 #endif /* __LINUX_MEMORY_HOTPLUG_H */
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 8605d7eb7f5c..86158eb9da70 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1617,6 +1617,9 @@  static DECLARE_WORK(free_hpage_work, free_hpage_workfn);
 
 static inline void flush_free_hpage_work(struct hstate *h)
 {
+	if (!hugetlb_optimize_vmemmap_enabled())
+		return;
+
 	if (hugetlb_optimize_vmemmap_pages(h))
 		flush_work(&free_hpage_work);
 }
diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c
index cc4ec752ec16..fcd9f7872064 100644
--- a/mm/hugetlb_vmemmap.c
+++ b/mm/hugetlb_vmemmap.c
@@ -10,6 +10,7 @@ 
  */
 #define pr_fmt(fmt)	"HugeTLB: " fmt
 
+#include <linux/memory_hotplug.h>
 #include "hugetlb_vmemmap.h"
 
 /*
@@ -22,21 +23,40 @@ 
 #define RESERVE_VMEMMAP_NR		1U
 #define RESERVE_VMEMMAP_SIZE		(RESERVE_VMEMMAP_NR << PAGE_SHIFT)
 
+enum vmemmap_optimize_mode {
+	VMEMMAP_OPTIMIZE_OFF,
+	VMEMMAP_OPTIMIZE_ON,
+};
+
 DEFINE_STATIC_KEY_MAYBE(CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON,
 			hugetlb_optimize_vmemmap_key);
 EXPORT_SYMBOL(hugetlb_optimize_vmemmap_key);
 
+static enum vmemmap_optimize_mode vmemmap_optimize_mode =
+	IS_ENABLED(CONFIG_HUGETLB_PAGE_FREE_VMEMMAP_DEFAULT_ON);
+
+static void vmemmap_optimize_mode_switch(enum vmemmap_optimize_mode to)
+{
+	if (vmemmap_optimize_mode == to)
+		return;
+
+	if (to == VMEMMAP_OPTIMIZE_OFF)
+		static_branch_dec(&hugetlb_optimize_vmemmap_key);
+	else
+		static_branch_inc(&hugetlb_optimize_vmemmap_key);
+	WRITE_ONCE(vmemmap_optimize_mode, to);
+}
+
 static int __init hugetlb_vmemmap_early_param(char *buf)
 {
 	bool enable;
+	enum vmemmap_optimize_mode mode;
 
 	if (kstrtobool(buf, &enable))
 		return -EINVAL;
 
-	if (enable)
-		static_branch_enable(&hugetlb_optimize_vmemmap_key);
-	else
-		static_branch_disable(&hugetlb_optimize_vmemmap_key);
+	mode = enable ? VMEMMAP_OPTIMIZE_ON : VMEMMAP_OPTIMIZE_OFF;
+	vmemmap_optimize_mode_switch(mode);
 
 	return 0;
 }
@@ -69,8 +89,10 @@  int hugetlb_vmemmap_alloc(struct hstate *h, struct page *head)
 	 */
 	ret = vmemmap_remap_alloc(vmemmap_addr, vmemmap_end, vmemmap_reuse,
 				  GFP_KERNEL | __GFP_NORETRY | __GFP_THISNODE);
-	if (!ret)
+	if (!ret) {
 		ClearHPageVmemmapOptimized(head);
+		static_branch_dec(&hugetlb_optimize_vmemmap_key);
+	}
 
 	return ret;
 }
@@ -84,6 +106,11 @@  void hugetlb_vmemmap_free(struct hstate *h, struct page *head)
 	if (!vmemmap_pages)
 		return;
 
+	if (READ_ONCE(vmemmap_optimize_mode) == VMEMMAP_OPTIMIZE_OFF)
+		return;
+
+	static_branch_inc(&hugetlb_optimize_vmemmap_key);
+
 	vmemmap_addr	+= RESERVE_VMEMMAP_SIZE;
 	vmemmap_end	= vmemmap_addr + (vmemmap_pages << PAGE_SHIFT);
 	vmemmap_reuse	= vmemmap_addr - PAGE_SIZE;
@@ -93,7 +120,9 @@  void hugetlb_vmemmap_free(struct hstate *h, struct page *head)
 	 * to the page which @vmemmap_reuse is mapped to, then free the pages
 	 * which the range [@vmemmap_addr, @vmemmap_end] is mapped to.
 	 */
-	if (!vmemmap_remap_free(vmemmap_addr, vmemmap_end, vmemmap_reuse))
+	if (vmemmap_remap_free(vmemmap_addr, vmemmap_end, vmemmap_reuse))
+		static_branch_dec(&hugetlb_optimize_vmemmap_key);
+	else
 		SetHPageVmemmapOptimized(head);
 }
 
@@ -110,9 +139,6 @@  void __init hugetlb_vmemmap_init(struct hstate *h)
 	BUILD_BUG_ON(__NR_USED_SUBPAGE >=
 		     RESERVE_VMEMMAP_SIZE / sizeof(struct page));
 
-	if (!hugetlb_optimize_vmemmap_enabled())
-		return;
-
 	if (!is_power_of_2(sizeof(struct page))) {
 		pr_warn_once("cannot optimize vmemmap pages because \"struct page\" crosses page boundaries\n");
 		static_branch_disable(&hugetlb_optimize_vmemmap_key);
@@ -134,3 +160,52 @@  void __init hugetlb_vmemmap_init(struct hstate *h)
 	pr_info("can optimize %d vmemmap pages for %s\n",
 		h->optimize_vmemmap_pages, h->name);
 }
+
+#ifdef CONFIG_PROC_SYSCTL
+static int hugetlb_optimize_vmemmap_handler(struct ctl_table *table, int write,
+					    void *buffer, size_t *length,
+					    loff_t *ppos)
+{
+	int ret;
+	enum vmemmap_optimize_mode mode;
+	static DEFINE_MUTEX(sysctl_mutex);
+
+	if (write && !capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	mutex_lock(&sysctl_mutex);
+	mode = vmemmap_optimize_mode;
+	table->data = &mode;
+	ret = proc_dointvec_minmax(table, write, buffer, length, ppos);
+	if (write && !ret)
+		vmemmap_optimize_mode_switch(mode);
+	mutex_unlock(&sysctl_mutex);
+
+	return ret;
+}
+
+static struct ctl_table hugetlb_vmemmap_sysctls[] = {
+	{
+		.procname	= "hugetlb_optimize_vmemmap",
+		.maxlen		= sizeof(enum vmemmap_optimize_mode),
+		.mode		= 0644,
+		.proc_handler	= hugetlb_optimize_vmemmap_handler,
+		.extra1		= SYSCTL_ZERO,
+		.extra2		= SYSCTL_ONE,
+	},
+	{ }
+};
+
+static __init int hugetlb_vmemmap_sysctls_init(void)
+{
+	/*
+	 * If "memory_hotplug.memmap_on_memory" is enabled or "struct page"
+	 * crosses page boundaries, the vmemmap pages cannot be optimized.
+	 */
+	if (!mhp_memmap_on_memory() && is_power_of_2(sizeof(struct page)))
+		register_sysctl_init("vm", hugetlb_vmemmap_sysctls);
+
+	return 0;
+}
+late_initcall(hugetlb_vmemmap_sysctls_init);
+#endif /* CONFIG_PROC_SYSCTL */
diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index a6101ae402f9..c72070cdd055 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -63,15 +63,10 @@  static bool memmap_on_memory __ro_after_init;
 module_param_cb(memmap_on_memory, &memmap_on_memory_ops, &memmap_on_memory, 0444);
 MODULE_PARM_DESC(memmap_on_memory, "Enable memmap on memory for memory hotplug");
 
-static inline bool mhp_memmap_on_memory(void)
+bool mhp_memmap_on_memory(void)
 {
 	return memmap_on_memory;
 }
-#else
-static inline bool mhp_memmap_on_memory(void)
-{
-	return false;
-}
 #endif
 
 enum {