by Jiang Liu

[permalink] [raw]

Subject: Re: [RFC PATCH 2/12] memory-hogplug : check memory offline in offline_pages

On 06/27/2012 01:44 PM, Yasuaki Ishimatsu wrote:
> When offline_pages() is called to offlined memory, the function fails since
> all memory has been offlined. In this case, the function should succeed.
> The patch adds the check function into offline_pages().
>
> CC: Len Brown <[email protected]>
> CC: Benjamin Herrenschmidt <[email protected]>
> CC: Paul Mackerras <[email protected]>
> CC: Christoph Lameter <[email protected]>
> Cc: Minchan Kim <[email protected]>
> CC: Andrew Morton <[email protected]>
> CC: KOSAKI Motohiro <[email protected]>
> CC: Wen Congyang <[email protected]>
> Signed-off-by: Yasuaki Ishimatsu <[email protected]>
>
> ---
> drivers/base/memory.c | 20 ++++++++++++++++++++
> include/linux/memory.h | 1 +
> mm/memory_hotplug.c | 5 +++++
> 3 files changed, 26 insertions(+)
>
> Index: linux-3.5-rc4/drivers/base/memory.c
> ===================================================================
> --- linux-3.5-rc4.orig/drivers/base/memory.c 2012-06-26 13:28:16.726211752 +0900
> +++ linux-3.5-rc4/drivers/base/memory.c 2012-06-26 13:34:22.423639904 +0900
> @@ -70,6 +70,26 @@ void unregister_memory_isolate_notifier(
> }
> EXPORT_SYMBOL(unregister_memory_isolate_notifier);
>
> +bool memory_is_offline(unsigned long start_pfn, unsigned long end_pfn)
> +{
> + struct memory_block *mem;
> + struct mem_section *section;
> + unsigned long pfn, section_nr;
> +
> + for (pfn = start_pfn; pfn < end_pfn; pfn += PAGES_PER_SECTION) {
> + section_nr = pfn_to_section_nr(pfn);
> + section = __nr_to_section(section_nr);
> + mem = find_memory_block(section);
Seems find_memory_block_hinted() is more efficient than find_memory_block() here.

> + if (!mem)
> + continue;
> + if (mem->state == MEM_OFFLINE)
> + continue;
> + return false;
> + }
> +
> + return true;
> +}
> +
> /*
> * register_memory - Setup a sysfs device for a memory block
> */
> Index: linux-3.5-rc4/include/linux/memory.h
> ===================================================================
> --- linux-3.5-rc4.orig/include/linux/memory.h 2012-06-25 04:53:04.000000000 +0900
> +++ linux-3.5-rc4/include/linux/memory.h 2012-06-26 13:34:22.424639891 +0900
> @@ -120,6 +120,7 @@ extern int memory_isolate_notify(unsigne
> extern struct memory_block *find_memory_block_hinted(struct mem_section *,
> struct memory_block *);
> extern struct memory_block *find_memory_block(struct mem_section *);
> +extern bool memory_is_offline(unsigned long start_pfn, unsigned long end_pfn);
> #define CONFIG_MEM_BLOCK_SIZE (PAGES_PER_SECTION<<PAGE_SHIFT)
> enum mem_add_context { BOOT, HOTPLUG };
> #endif /* CONFIG_MEMORY_HOTPLUG_SPARSE */
> Index: linux-3.5-rc4/mm/memory_hotplug.c
> ===================================================================
> --- linux-3.5-rc4.orig/mm/memory_hotplug.c 2012-06-26 13:28:16.743211538 +0900
> +++ linux-3.5-rc4/mm/memory_hotplug.c 2012-06-26 13:48:38.264940468 +0900
> @@ -887,6 +887,11 @@ static int __ref offline_pages(unsigned
>
> lock_memory_hotplug();
>
> + if (memory_is_offline(start_pfn, end_pfn)) {
> + ret = 0;
> + goto out;
> + }
> +
> zone = page_zone(pfn_to_page(start_pfn));
> node = zone_to_nid(zone);
> nr_pages = end_pfn - start_pfn;
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>

2012-06-30 15:58:27

by Jiang Liu

[permalink] [raw]

Subject: Re: [RFC PATCH 10/12] memory-hotplug : free memmap of sparse-vmemmap

On 06/27/2012 01:56 PM, Yasuaki Ishimatsu wrote:
> I don't think that all pages of virtual mapping in removed memory can be
> freed, since page which type is MIX_SECTION_INFO is difficult to free.
> So, the patch only frees page which type is SECTION_INFO at first.
>
> CC: Len Brown <[email protected]>
> CC: Benjamin Herrenschmidt <[email protected]>
> CC: Paul Mackerras <[email protected]>
> CC: Christoph Lameter <[email protected]>
> Cc: Minchan Kim <[email protected]>
> CC: Andrew Morton <[email protected]>
> CC: KOSAKI Motohiro <[email protected]>
> CC: Wen Congyang <[email protected]>
> Signed-off-by: Yasuaki Ishimatsu <[email protected]>
>
> ---
> arch/x86/mm/init_64.c | 89 ++++++++++++++++++++++++++++++++++++++++++++++++++
> include/linux/mm.h | 2 +
> mm/memory_hotplug.c | 5 ++
> mm/sparse.c | 5 +-
> 4 files changed, 99 insertions(+), 2 deletions(-)
>
> Index: linux-3.5-rc4/include/linux/mm.h
> ===================================================================
> --- linux-3.5-rc4.orig/include/linux/mm.h 2012-06-27 09:11:13.790150442 +0900
> +++ linux-3.5-rc4/include/linux/mm.h 2012-06-27 09:11:16.433117400 +0900
> @@ -1588,6 +1588,8 @@ int vmemmap_populate(struct page *start_
> void vmemmap_populate_print_last(void);
> void register_page_bootmem_memmap(unsigned long section_nr, struct page *map,
> unsigned long size);
> +void vmemmap_kfree(struct page *memmpa, unsigned long nr_pages);
> +void vmemmap_free_bootmem(struct page *memmpa, unsigned long nr_pages);
>
> enum mf_flags {
> MF_COUNT_INCREASED = 1 << 0,
> Index: linux-3.5-rc4/mm/sparse.c
> ===================================================================
> --- linux-3.5-rc4.orig/mm/sparse.c 2012-06-27 09:06:35.317631878 +0900
> +++ linux-3.5-rc4/mm/sparse.c 2012-06-27 09:11:16.434117388 +0900
> @@ -614,12 +614,13 @@ static inline struct page *kmalloc_secti
> /* This will make the necessary allocations eventually. */
> return sparse_mem_map_populate(pnum, nid);
> }
> -static void __kfree_section_memmap(struct page *memmap, unsigned long nr_pages)
> +static void __kfree_section_memmap(struct page *page, unsigned long nr_pages)
> {
> - return; /* XXX: Not implemented yet */
> + vmemmap_kfree(page, nr_pages);
> }
> static void free_map_bootmem(struct page *page, unsigned long nr_pages)
> {
> + vmemmap_free_bootmem(page, nr_pages);
> }
> #else
> static struct page *__kmalloc_section_memmap(unsigned long nr_pages)
> Index: linux-3.5-rc4/arch/x86/mm/init_64.c
> ===================================================================
> --- linux-3.5-rc4.orig/arch/x86/mm/init_64.c 2012-06-27 09:11:13.791150430 +0900
> +++ linux-3.5-rc4/arch/x86/mm/init_64.c 2012-06-27 09:11:59.254581998 +0900
> @@ -978,6 +978,95 @@ vmemmap_populate(struct page *start_page
> return 0;
> }
>
> +unsigned long find_and_clear_pte_page(unsigned long addr, unsigned long end,
> + struct page *page)
I think the third parameter should be "struct page **pp" instead of "struct page *page".
And "page = pte_page(*pte)" should be "*pp = pte_page(*pte)".
Otherwise the found page pointer can't be returned to the caller and vmemmap_kfree()
just sees random value in variable "page".

> +{
> + pgd_t *pgd;
> + pud_t *pud;
> + pmd_t *pmd;
> + pte_t *pte;
> + unsigned long next;
> +
> + page = NULL;
> +
> + pgd = pgd_offset_k(addr);
> + if (pgd_none(*pgd))
> + return PAGE_SIZE;
> +
> + pud = pud_offset(pgd, addr);
> + if (pud_none(*pud))
> + return PAGE_SIZE;
> +
> + if (!cpu_has_pse) {
> + next = (addr + PAGE_SIZE) & PAGE_MASK;
> + pmd = pmd_offset(pud, addr);
> + if (pmd_none(*pmd))
> + return next;
> +
> + pte = pte_offset_kernel(pmd, addr);
> + if (pte_none(*pte))
> + return next;
> +
> + page = pte_page(*pte);
> + pte_clear(&init_mm, addr, pte);
> + } else {
> + next = pmd_addr_end(addr, end);
> +
> + pmd = pmd_offset(pud, addr);
> + if (pmd_none(*pmd))
> + return next;
> +
> + page = pmd_page(*pmd);
> + pmd_clear(pmd);
> + }
> +
> + return next;
> +}
> +
> +void __meminit
> +vmemmap_kfree(struct page *memmap, unsigned long nr_pages)
> +{
> + unsigned long addr = (unsigned long)memmap;
> + unsigned long end = (unsigned long)(memmap + nr_pages);
> + unsigned long next;
> + unsigned int order;
> + struct page *page;
> +
> + for (; addr < end; addr = next) {
> + next = find_and_clear_pte_page(addr, end, page);
> + if (!page)
> + continue;
> +
> + if (is_vmalloc_addr(page))
> + vfree(page);
> + else {
> + order = next - addr;
> + free_pages((unsigned long)page,
> + get_order(sizeof(struct page) * order));
> + }
> + }
> +}
> +
> +void __meminit
> +vmemmap_free_bootmem(struct page *memmap, unsigned long nr_pages)
> +{
> + unsigned long addr = (unsigned long)memmap;
> + unsigned long end = (unsigned long)(memmap + nr_pages);
> + unsigned long next;
> + struct page *page;
> + unsigned long magic;
> +
> + for (; addr < end; addr = next) {
> + next = find_and_clear_pte_page(addr, end, page);
> + if (!page)
> + continue;
> +
> + magic = (unsigned long) page->lru.next;
> + if (magic == SECTION_INFO)
> + put_page_bootmem(page);
> + }
> +}
> +
> void __meminit
> register_page_bootmem_memmap(unsigned long section_nr, struct page *start_page,
> unsigned long size)
> Index: linux-3.5-rc4/mm/memory_hotplug.c
> ===================================================================
> --- linux-3.5-rc4.orig/mm/memory_hotplug.c 2012-06-27 09:11:13.789150454 +0900
> +++ linux-3.5-rc4/mm/memory_hotplug.c 2012-06-27 09:11:16.436117363 +0900
> @@ -303,6 +303,8 @@ static int __meminit __add_section(int n
> #ifdef CONFIG_SPARSEMEM_VMEMMAP
> static int __remove_section(struct zone *zone, struct mem_section *ms)
> {
> + unsigned long flags;
> + struct pglist_data *pgdat = zone->zone_pgdat;
> int ret;
>
> if (!valid_section(ms))
> @@ -310,6 +312,9 @@ static int __remove_section(struct zone
>
> ret = unregister_memory_section(ms);
>
> + pgdat_resize_lock(pgdat, &flags);
> + sparse_remove_one_section(zone, ms);
> + pgdat_resize_unlock(pgdat, &flags);
> return ret;
> }
> #else
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>