[RFC v2 PATCH 4/7] thp, s390: thp splitting backend for s390

[Gerald Schaefer <gerald.schaefer@xxxxxxxxxx>]

This patch is part of the architecture backend for thp on s390.
It provides the functions related to thp splitting, including
serialization against gup. Unlike other archs, pmdp_splitting_flush()
cannot use a tlb flushing operation to serialize against gup on s390,
because that wouldn't be stopped by the disabled IRQs. So instead,
smp_call_function() is called with an empty function, which will have
the expected effect.

Signed-off-by: Gerald Schaefer <gerald.schaefer@xxxxxxxxxx>
---
 arch/s390/include/asm/pgtable.h | 13 +++++++++++++
 arch/s390/mm/gup.c              | 11 ++++++++++-
 arch/s390/mm/pgtable.c          | 18 ++++++++++++++++++
 3 files changed, 41 insertions(+), 1 deletion(-)

diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 6bd7d74..c3f0775 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -347,6 +347,8 @@ extern struct page *vmemmap;
 
 #define _SEGMENT_ENTRY_LARGE	0x400	/* STE-format control, large page   */
 #define _SEGMENT_ENTRY_CO	0x100	/* change-recording override   */
+#define _SEGMENT_ENTRY_SPLIT_BIT 0	/* THP splitting bit number */
+#define _SEGMENT_ENTRY_SPLIT	(1UL << _SEGMENT_ENTRY_SPLIT_BIT)
 
 /* Page status table bits for virtualization */
 #define RCP_ACC_BITS	0xf000000000000000UL
@@ -506,6 +508,10 @@ static inline int pmd_bad(pmd_t pmd)
 	return (pmd_val(pmd) & mask) != _SEGMENT_ENTRY;
 }
 
+#define __HAVE_ARCH_PMDP_SPLITTING_FLUSH
+extern void pmdp_splitting_flush(struct vm_area_struct *vma,
+				 unsigned long addr, pmd_t *pmdp);
+
 static inline int pte_none(pte_t pte)
 {
 	return (pte_val(pte) & _PAGE_INVALID) && !(pte_val(pte) & _PAGE_SWT);
@@ -1159,6 +1165,13 @@ static inline pmd_t *pmd_offset(pud_t *pud, unsigned long address)
 #define pte_offset_map(pmd, address) pte_offset_kernel(pmd, address)
 #define pte_unmap(pte) do { } while (0)
 
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+static inline int pmd_trans_splitting(pmd_t pmd)
+{
+	return pmd_val(pmd) & _SEGMENT_ENTRY_SPLIT;
+}
+#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
+
 /*
  * 31 bit swap entry format:
  * A page-table entry has some bits we have to treat in a special way.
diff --git a/arch/s390/mm/gup.c b/arch/s390/mm/gup.c
index 65cb06e..d34fd6e 100644
--- a/arch/s390/mm/gup.c
+++ b/arch/s390/mm/gup.c
@@ -115,7 +115,16 @@ static inline int gup_pmd_range(pud_t *pudp, pud_t pud, unsigned long addr,
 		pmd = *pmdp;
 		barrier();
 		next = pmd_addr_end(addr, end);
-		if (pmd_none(pmd))
+		/*
+		 * The pmd_trans_splitting() check below explains why
+		 * pmdp_splitting_flush() has to serialize with
+		 * smp_call_function() against our disabled IRQs, to stop
+		 * this gup-fast code from running while we set the
+		 * splitting bit in the pmd. Returning zero will take
+		 * the slow path that will call wait_split_huge_page()
+		 * if the pmd is still in splitting state.
+		 */
+		if (pmd_none(pmd) || pmd_trans_splitting(pmd))
 			return 0;
 		if (unlikely(pmd_huge(pmd))) {
 			if (!gup_huge_pmd(pmdp, pmd, addr, next,
diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c
index 18df31d..5a73522 100644
--- a/arch/s390/mm/pgtable.c
+++ b/arch/s390/mm/pgtable.c
@@ -866,3 +866,21 @@ bool kernel_page_present(struct page *page)
 	return cc == 0;
 }
 #endif /* CONFIG_HIBERNATION && CONFIG_DEBUG_PAGEALLOC */
+
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+static void pmdp_splitting_flush_sync(void *arg)
+{
+	/* Simply deliver the interrupt */
+}
+
+void pmdp_splitting_flush(struct vm_area_struct *vma, unsigned long address,
+			  pmd_t *pmdp)
+{
+	VM_BUG_ON(address & ~HPAGE_PMD_MASK);
+	if (!test_and_set_bit(_SEGMENT_ENTRY_SPLIT_BIT,
+			      (unsigned long *) pmdp)) {
+		/* need to serialize against gup-fast (IRQ disabled) */
+		smp_call_function(pmdp_splitting_flush_sync, NULL, 1);
+	}
+}
+#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
-- 
1.7.11.5

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/