mm: check if section present during memory block registering
authorYinghai Lu <yinghai@kernel.org>
Fri, 4 Sep 2015 22:42:39 +0000 (15:42 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Fri, 4 Sep 2015 23:54:41 +0000 (16:54 -0700)
Tony Luck found on his setup, if memory block size 512M will cause crash
during booting.

  BUG: unable to handle kernel paging request at ffffea0074000020
  IP: get_nid_for_pfn+0x17/0x40
  PGD 128ffcb067 PUD 128ffc9067 PMD 0
  Oops: 0000 [#1] SMP
  Modules linked in:
  CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.2.0-rc8 #1
  ...
  Call Trace:
     ? register_mem_sect_under_node+0x66/0xe0
     register_one_node+0x17b/0x240
     ? pci_iommu_alloc+0x6e/0x6e
     topology_init+0x3c/0x95
     do_one_initcall+0xcd/0x1f0

The system has non continuous RAM address:
 BIOS-e820: [mem 0x0000001300000000-0x0000001cffffffff] usable
 BIOS-e820: [mem 0x0000001d70000000-0x0000001ec7ffefff] usable
 BIOS-e820: [mem 0x0000001f00000000-0x0000002bffffffff] usable
 BIOS-e820: [mem 0x0000002c18000000-0x0000002d6fffefff] usable
 BIOS-e820: [mem 0x0000002e00000000-0x00000039ffffffff] usable

So there are start sections in memory block not present.  For example:

    memory block : [0x2c18000000, 0x2c20000000) 512M

first three sections are not present.

The current register_mem_sect_under_node() assume first section is
present, but memory block section number range [start_section_nr,
end_section_nr] would include not present section.

For arch that support vmemmap, we don't setup memmap for struct page
area within not present sections area.

So skip the pfn range that belong to absent section.

[akpm@linux-foundation.org: simplification]
[rientjes@google.com: more simplification]
Fixes: bdee237c0343 ("x86: mm: Use 2GB memory block size on large memory x86-64 systems")
Fixes: 982792c782ef ("x86, mm: probe memory block size for generic x86 64bit")
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: David Rientjes <rientjes@google.com>
Reported-by: Tony Luck <tony.luck@intel.com>
Tested-by: Tony Luck <tony.luck@intel.com>
Cc: Greg KH <greg@kroah.com>
Cc: Ingo Molnar <mingo@elte.hu>
Tested-by: David Rientjes <rientjes@google.com>
Cc: <stable@vger.kernel.org> [3.15+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
drivers/base/node.c

index 31df474d72f4a275ba0c87fb79a26c538fc661ca..560751bad2947105a0d06e7650bf906c3e4f0de7 100644 (file)
@@ -392,6 +392,16 @@ int register_mem_sect_under_node(struct memory_block *mem_blk, int nid)
        for (pfn = sect_start_pfn; pfn <= sect_end_pfn; pfn++) {
                int page_nid;
 
+               /*
+                * memory block could have several absent sections from start.
+                * skip pfn range from absent section
+                */
+               if (!pfn_present(pfn)) {
+                       pfn = round_down(pfn + PAGES_PER_SECTION,
+                                        PAGES_PER_SECTION) - 1;
+                       continue;
+               }
+
                page_nid = get_nid_for_pfn(pfn);
                if (page_nid < 0)
                        continue;