x86/mm/hotplug: Fix BUG_ON() after hot-remove by not freeing PUD

Since commit:

  af2cf278ef4f ("x86/mm/hotplug: Don't remove PGD entries in remove_pagetable()")

we no longer free PUDs so that we do not have to synchronize
all PGDs on hot-remove/vfree().

But the new 5-level page table patchset reverted that for 4-level
page tables, in the following commit:

  f2a6a7050109: ("x86: Convert the rest of the code to support p4d_t")

This patch restores the damage and disables free_pud() if we are in the
4-level page table case, thus avoiding BUG_ON() after hot-remove.

Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
[ Clarified the changelog and the code comments. ]
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mm@kvack.org
Link: http://lkml.kernel.org/r/20170624180514.3821-1-jglisse@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index 95651dc58e0900d3717a462d4cfe05efffc9374e..0a59daf799f8ed95ad6cef315ff23818e1a110ea 100644 (file)
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -990,7 +990,13 @@ remove_p4d_table(p4d_t *p4d_start, unsigned long addr, unsigned long end,
 
                pud_base = pud_offset(p4d, 0);
                remove_pud_table(pud_base, addr, next, direct);
-               free_pud_table(pud_base, p4d);
+               /*
+                * For 4-level page tables we do not want to free PUDs, but in the
+                * 5-level case we should free them. This code will have to change
+                * to adapt for boot-time switching between 4 and 5 level page tables.
+                */
+               if (CONFIG_PGTABLE_LEVELS == 5)
+                       free_pud_table(pud_base, p4d);
        }
 
        if (direct)