Re: Linux 4.18-rc7

[Linus Torvalds]

On Wed, Aug 1, 2018 at 10:15 AM Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>
> I'm still unhappy about the vma_init() ones, and I have not decided
> how to go with those. Either the memset() in vma_init(), or just
> reverting the (imho unnecessary) commit 2c4541e24c55. Kirill, Andrew,
> comments?

Ugh. Adding a memset looks simple, but screws up some places that have
other initialization. It also requires adding a new include of
<linux/string.h>, or we'd need to uninline vma_init() and put it
somewhere else.

But just reverting commit 2c4541e24c55 ("mm: use vma_init() to
initialize VMAs on stack and data segments") entirely isn't good
either, because some of the cases aren't about the TLB flush
interface, and call down to "real" VM functions. The 'pseudo_vma' use
of remove_inode_hugepages() and hugetlbfs_fallocate() in particular is
odd, but using vma_init() looks good there. And those places had the
memset() already.

So I'm inclined to simply mark the TLB-related vma_init() cases
special, and use something like this:

  #define TLB_FLUSH_VMA(mm,flags) { .vm_mm = (mm), .vm_flags = (flags) }

to make it very obvious when we're doing that vma initialization for
flush_tlb_range(). It's done as an initializer, exactly so that the
only valid syntax is to do somethin glike this:

        struct vm_area_struct vma = TLB_FLUSH_VMA(mm, VM_EXEC);

That leaves vma_init() users to be just the actual real allocation
path, and a few very specific specual vmas (the hugetlbfs and
mempolicy pseudo-vma, and a couple of "gate" vmas).

Suggested patch attached. Comments?

                 Linus
 arch/arm/mach-rpc/ecard.c    |  5 +----
 arch/arm64/include/asm/tlb.h |  4 +---
 arch/arm64/mm/hugetlbpage.c  | 10 ++++------
 arch/ia64/include/asm/tlb.h  |  7 +++----
 include/linux/mm.h           |  3 +++
 5 files changed, 12 insertions(+), 17 deletions(-)

diff --git a/arch/arm/mach-rpc/ecard.c b/arch/arm/mach-rpc/ecard.c
index 8db62cc54a6a..04b2f22c2739 100644
--- a/arch/arm/mach-rpc/ecard.c
+++ b/arch/arm/mach-rpc/ecard.c
@@ -212,7 +212,7 @@ static DEFINE_MUTEX(ecard_mutex);
  */
 static void ecard_init_pgtables(struct mm_struct *mm)
 {
-	struct vm_area_struct vma;
+	struct vm_area_struct vma = TLB_FLUSH_VMA(mm, VM_EXEC);
 
 	/* We want to set up the page tables for the following mapping:
 	 *  Virtual	Physical
@@ -237,9 +237,6 @@ static void ecard_init_pgtables(struct mm_struct *mm)
 
 	memcpy(dst_pgd, src_pgd, sizeof(pgd_t) * (EASI_SIZE / PGDIR_SIZE));
 
-	vma_init(&vma, mm);
-	vma.vm_flags = VM_EXEC;
-
 	flush_tlb_range(&vma, IO_START, IO_START + IO_SIZE);
 	flush_tlb_range(&vma, EASI_START, EASI_START + EASI_SIZE);
 }
diff --git a/arch/arm64/include/asm/tlb.h b/arch/arm64/include/asm/tlb.h
index d87f2d646caa..0ad1cf233470 100644
--- a/arch/arm64/include/asm/tlb.h
+++ b/arch/arm64/include/asm/tlb.h
@@ -37,9 +37,7 @@ static inline void __tlb_remove_table(void *_table)
 
 static inline void tlb_flush(struct mmu_gather *tlb)
 {
-	struct vm_area_struct vma;
-
-	vma_init(&vma, tlb->mm);
+	struct vm_area_struct vma = TLB_FLUSH_VMA(tlb->mm, 0);
 
 	/*
 	 * The ASID allocator will either invalidate the ASID or mark
diff --git a/arch/arm64/mm/hugetlbpage.c b/arch/arm64/mm/hugetlbpage.c
index 1854e49aa18a..192b3ba07075 100644
--- a/arch/arm64/mm/hugetlbpage.c
+++ b/arch/arm64/mm/hugetlbpage.c
@@ -108,13 +108,10 @@ static pte_t get_clear_flush(struct mm_struct *mm,
 			     unsigned long pgsize,
 			     unsigned long ncontig)
 {
-	struct vm_area_struct vma;
 	pte_t orig_pte = huge_ptep_get(ptep);
 	bool valid = pte_valid(orig_pte);
 	unsigned long i, saddr = addr;
 
-	vma_init(&vma, mm);
-
 	for (i = 0; i < ncontig; i++, addr += pgsize, ptep++) {
 		pte_t pte = ptep_get_and_clear(mm, addr, ptep);
 
@@ -127,8 +124,10 @@ static pte_t get_clear_flush(struct mm_struct *mm,
 			orig_pte = pte_mkdirty(orig_pte);
 	}
 
-	if (valid)
+	if (valid) {
+		struct vm_area_struct vma = TLB_FLUSH_VMA(mm, 0);
 		flush_tlb_range(&vma, saddr, addr);
+	}
 	return orig_pte;
 }
 
@@ -147,10 +146,9 @@ static void clear_flush(struct mm_struct *mm,
 			     unsigned long pgsize,
 			     unsigned long ncontig)
 {
-	struct vm_area_struct vma;
+	struct vm_area_struct vma = TLB_FLUSH_VMA(mm, 0);
 	unsigned long i, saddr = addr;
 
-	vma_init(&vma, mm);
 	for (i = 0; i < ncontig; i++, addr += pgsize, ptep++)
 		pte_clear(mm, addr, ptep);
 
diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h
index db89e7306081..516355a774bf 100644
--- a/arch/ia64/include/asm/tlb.h
+++ b/arch/ia64/include/asm/tlb.h
@@ -115,12 +115,11 @@ ia64_tlb_flush_mmu_tlbonly(struct mmu_gather *tlb, unsigned long start, unsigned
 		flush_tlb_all();
 	} else {
 		/*
-		 * XXX fix me: flush_tlb_range() should take an mm pointer instead of a
-		 * vma pointer.
+		 * flush_tlb_range() takes a vma instead of a mm pointer because
+		 * some architectures want the vm_flags for ITLB/DTLB flush.
 		 */
-		struct vm_area_struct vma;
+		struct vm_area_struct vma = TLB_FLUSH_VMA(tlb->mm, 0);
 
-		vma_init(&vma, tlb->mm);
 		/* flush the address range from the tlb: */
 		flush_tlb_range(&vma, start, end);
 		/* now flush the virt. page-table area mapping the address range: */
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 7ba6d356d18f..68a5121694ef 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -466,6 +466,9 @@ static inline void vma_set_anonymous(struct vm_area_struct *vma)
 	vma->vm_ops = NULL;
 }
 
+/* flush_tlb_range() takes a vma, not a mm, and can care about flags */
+#define TLB_FLUSH_VMA(mm,flags) { .vm_mm = (mm), .vm_flags = (flags) }
+
 struct mmu_gather;
 struct inode;