On Sun, Jul 07, 2024 at 09:03:05PM +0200, Oleg Nesterov wrote:
>As I have already said, I can't review this patch.
>
>But I am curious, why set_max_threads() is the only user of the new helper?
>Say, should files_maxfiles_init() use it too or not? Perhaps the changelog
>can explain this?
>
files_maxfiles_init() is called in two places:
* vfs_caches_init()
* page_alloc_init_late()
And in page_alloc_init_late(), it is called after deferred_init_memmap(). This
means the totalram_pages() here return the real free pages at this point. So
it is not necessary to touch it now.
>Thanks,
>
>Oleg.
>
>On 07/07, 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@xxxxxxxxx>
>> >>>---
>> >>> 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@xxxxxxxxx>
>> >CC: David Hildenbrand <david@xxxxxxxxxx>
>> >---
>> > 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.
>>
>> ?
>>
>> Reviewed-by: David Hildenbrand <david@xxxxxxxxxx>
>>
>> --
>> Cheers,
>>
>> David / dhildenb
>>
--
Wei Yang
Help you, Help me
