[PATCH] powerpc/mm: Make sure a local_irq_disable prevent a parallel THP split

From: "Aneesh Kumar K.V" <redacted>

We have generic code like the one in get_futex_key that assume that
a local_irq_disable prevents a parallel THP split. Support that by
adding a dummy smp call function after setting _PAGE_SPLITTING. Code
paths like get_user_pages_fast still need to check for _PAGE_SPLITTING
after disabling IRQ which indicate that a parallel THP splitting is
ongoing. Now if they don't find _PAGE_SPLITTING set, then we can be
sure that parallel split will now block in pmdp_splitting flush
until we enables IRQ

Signed-off-by: Aneesh Kumar K.V <redacted>
---
 arch/powerpc/mm/pgtable_64.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/arch/powerpc/mm/pgtable_64.c b/arch/powerpc/mm/pgtable_64.c
index 62bf5e8e78da..f6ce1f111f5b 100644
--- a/arch/powerpc/mm/pgtable_64.c
+++ b/arch/powerpc/mm/pgtable_64.c

@@ -647,6 +647,11 @@ void pmdp_splitting_flush(struct vm_area_struct *vma,
 		if (old & _PAGE_HASHPTE)
 			hpte_do_hugepage_flush(vma->vm_mm, address, pmdp);
 	}
+	/*
+	 * This ensures that generic code that rely on IRQ disabling
+	 * to prevent a parallel THP split work as expected.
+	 */
+	kick_all_cpus_sync();
 }
 
 /*

-- 
1.8.3.2

On 03/15/2014 06:47 AM, Aneesh Kumar K.V wrote:

From: "Aneesh Kumar K.V" <redacted>

We have generic code like the one in get_futex_key that assume that
a local_irq_disable prevents a parallel THP split. Support that by
adding a dummy smp call function after setting _PAGE_SPLITTING. Code
paths like get_user_pages_fast still need to check for _PAGE_SPLITTING
after disabling IRQ which indicate that a parallel THP splitting is
ongoing. Now if they don't find _PAGE_SPLITTING set, then we can be
sure that parallel split will now block in pmdp_splitting flush
until we enables IRQ

Signed-off-by: Aneesh Kumar K.V <redacted>

Acked-by: Rik van Riel <redacted>

-- 
All rights reversed