linux-kernel - Re: [PATCH v3 1/3] mm/memblock: introduce a new helper memblock_estimated_nr

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20240708003913.dcf7sfklrdpyhom7@master>
Date: Mon, 8 Jul 2024 00:39:13 +0000
From: Wei Yang <richard.weiyang@...il.com>
To: David Hildenbrand <david@...hat.com>
Cc: Wei Yang <richard.weiyang@...il.com>, Mike Rapoport <rppt@...nel.org>,
	akpm@...ux-foundation.org, brauner@...nel.org, oleg@...hat.com,
	mjguzik@...il.com, tandersen@...flix.com, linux-mm@...ck.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 1/3] mm/memblock: introduce a new helper
 memblock_estimated_nr_pages()

On Sun, Jul 07, 2024 at 10:17:27AM +0200, David Hildenbrand wrote:
>On 06.07.24 03:28, Wei Yang wrote:
>> On Fri, Jul 05, 2024 at 12:09:48PM +0300, Mike Rapoport wrote:
>> > On Wed, Jul 03, 2024 at 12:51:49AM +0000, Wei Yang wrote:
>> > > Instead of using raw memblock api, we wrap a new helper for user.
>> > 
>> > The changelog should be more elaborate and explain why this function is
>> > needed.
>> > 
>> > > Signed-off-by: Wei Yang <richard.weiyang@...il.com>
>> > > ---
>> > >   include/linux/memblock.h |  1 +
>> > >   mm/memblock.c            | 19 +++++++++++++++++++
>> > >   2 files changed, 20 insertions(+)
>> > > 
>> > > diff --git a/include/linux/memblock.h b/include/linux/memblock.h
>> > > index 40c62aca36ec..7d1c32b3dc12 100644
>> > > --- a/include/linux/memblock.h
>> > > +++ b/include/linux/memblock.h
>> > > @@ -486,6 +486,7 @@ static inline __init_memblock bool memblock_bottom_up(void)
>> > >   phys_addr_t memblock_phys_mem_size(void);
>> > >   phys_addr_t memblock_reserved_size(void);
>> > > +unsigned long memblock_estimated_nr_pages(void);
>> > >   phys_addr_t memblock_start_of_DRAM(void);
>> > >   phys_addr_t memblock_end_of_DRAM(void);
>> > >   void memblock_enforce_memory_limit(phys_addr_t memory_limit);
>> > > diff --git a/mm/memblock.c b/mm/memblock.c
>> > > index e81fb68f7f88..c1f1aac0459f 100644
>> > > --- a/mm/memblock.c
>> > > +++ b/mm/memblock.c
>> > > @@ -1729,6 +1729,25 @@ phys_addr_t __init_memblock memblock_reserved_size(void)
>> > >   	return memblock.reserved.total_size;
>> > >   }
>> > > +/**
>> > > + * memblock_estimated_nr_pages - return number of pages from memblock point of
>> > > + * view
>> > 
>> > This function returns the estimate for free pages, not the number of pages
>> > in RAM.
>> > 
>> > How about memblock_estimated_nr_free_pages()?
>> > 
>> > > + * some calculation before all pages are freed to buddy system, especially
>> > > + * when CONFIG_DEFERRED_STRUCT_PAGE_INIT is enabled.
>> > 
>> > I'm failing to parse this sentence. The return value here won't depend on
>> > CONFIG_DEFERRED_STRUCT_PAGE_INIT.
>> > 
>> > > + *
>> > > + * At this point, we can get this information from memblock. Since the system
>> > > + * state is not settle down and address alignment, the value is an estimation.
>> > > + *
>> > > + * Return:
>> > > + * An estimated number of pages from memblock point of view.
>> > 
>> >                             ^ free
>> > 
>> > > + */
>> > > +unsigned long __init memblock_estimated_nr_pages(void)
>> > > +{
>> > > +	return PHYS_PFN(memblock_phys_mem_size() - memblock_reserved_size());
>> > > +}
>> > > +
>> > >   /* lowest address */
>> > >   phys_addr_t __init_memblock memblock_start_of_DRAM(void)
>> > >   {
>> > > -- 
>> > > 2.34.1
>> > > 
>> 
>> Thanks for review. Is this one looks better?
>> 
>> Subject: [PATCH] mm/memblock: introduce a new helper memblock_estimated_nr_free_pages()
>> 
>> During bootup, system may need the number of free pages in the whole system
>> to do some calculation before all pages are freed to buddy system. Usually
>> this number is get from totalram_pages(). Since we plan to move the free
>> pages accounting in __free_pages_core(), this value may not represent
>> total free pages at the early stage, especially when
>> CONFIG_DEFERRED_STRUCT_PAGE_INIT is enabled.
>> 
>> Instead of using raw memblock api, let's introduce a new helper for user
>> to get the estimated number of free pages from memblock point of view.
>> 
>> Signed-off-by: Wei Yang <richard.weiyang@...il.com>
>> CC: David Hildenbrand <david@...hat.com>
>> ---
>>   include/linux/memblock.h |  1 +
>>   mm/memblock.c            | 22 ++++++++++++++++++++++
>>   2 files changed, 23 insertions(+)
>> 
>> diff --git a/include/linux/memblock.h b/include/linux/memblock.h
>> index 40c62aca36ec..7d1c32b3dc12 100644
>> --- a/include/linux/memblock.h
>> +++ b/include/linux/memblock.h
>> @@ -486,6 +486,7 @@ static inline __init_memblock bool memblock_bottom_up(void)
>>   phys_addr_t memblock_phys_mem_size(void);
>>   phys_addr_t memblock_reserved_size(void);
>> +unsigned long memblock_estimated_nr_pages(void);
>>   phys_addr_t memblock_start_of_DRAM(void);
>>   phys_addr_t memblock_end_of_DRAM(void);
>>   void memblock_enforce_memory_limit(phys_addr_t memory_limit);
>> diff --git a/mm/memblock.c b/mm/memblock.c
>> index e81fb68f7f88..00decc42e02b 100644
>> --- a/mm/memblock.c
>> +++ b/mm/memblock.c
>> @@ -1729,6 +1729,28 @@ phys_addr_t __init_memblock memblock_reserved_size(void)
>>   	return memblock.reserved.total_size;
>>   }
>> +/**
>> + * memblock_estimated_nr_free_pages - return estimated number of free pages
>> + * from memblock point of view
>> + *
>> + * During bootup, system may need the number of free pages in the whole system
>> + * to do some calculation before all pages are freed to buddy system. Usually
>> + * this number is get from totalram_pages(). Since we plan to move the free
>> + * pages accounting in __free_pages_core(), this value may not represent total
>> + * free pages at the early stage, especially when > + * CONFIG_DEFERRED_STRUCT_PAGE_INIT is enabled.
>
>These historical details should be dropped. "Since we plan ..." will easily
>get outdated soon.
>
>* During bootup, subsystems might need a rough estimate of the number of *
>free pages in the whole system, before precise numbers are available
>* from the buddy. Especially with CONFIG_DEFERRED_STRUCT_PAGE_INIT, the
>* numbers obtained from the buddy might be very imprecise during bootup.
>

Pretty better, thanks :-)

>?
>
>Reviewed-by: David Hildenbrand <david@...hat.com>
>
>-- 
>Cheers,
>
>David / dhildenb

-- 
Wei Yang
Help you, Help me