[Devel] [PATCH RHEL7 COMMIT] arch: x86: don't charge kernel page tables to kmemcg

[Konstantin Khorenko]

The commit is pushed to "branch-rh7-3.10.0-327.18.2.vz7.14.x-ovz" and will appear at https://src.openvz.org/scm/ovz/vzkernel.git
after rh7-3.10.0-327.18.2.vz7.14.9
------>
commit 00bc193297d25db75b6c1bdb9855a993d33bd866
Author: Vladimir Davydov <vdavydov at virtuozzo.com>
Date:   Tue May 31 13:14:11 2016 +0400

    arch: x86: don't charge kernel page tables to kmemcg
    
    Kernel page tables (e.g. those used for vmalloc) can be shared among
    processes from different cgroups so accounting them to a paritcular one
    can pin other cgroups for indefinitely long.
    
    Signed-off-by: Vladimir Davydov <vdavydov at virtuozzo.com>
---
 arch/x86/include/asm/pgalloc.h | 14 +++++++++++---
 arch/x86/mm/pgtable.c          | 14 +++++++++-----
 2 files changed, 20 insertions(+), 8 deletions(-)

diff --git a/arch/x86/include/asm/pgalloc.h b/arch/x86/include/asm/pgalloc.h
index 58e4567..6515de6 100644
--- a/arch/x86/include/asm/pgalloc.h
+++ b/arch/x86/include/asm/pgalloc.h
@@ -4,6 +4,7 @@
 #include <linux/threads.h>
 #include <linux/mm.h>		/* for struct page */
 #include <linux/pagemap.h>
+#include <linux/sched.h>	/* for init_mm */
 
 static inline int  __paravirt_pgd_alloc(struct mm_struct *mm) { return 0; }
 
@@ -81,7 +82,11 @@ static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd,
 static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr)
 {
 	struct page *page;
-	page = alloc_pages(GFP_KERNEL_ACCOUNT | __GFP_REPEAT | __GFP_ZERO, 0);
+	gfp_t gfp = GFP_KERNEL_ACCOUNT | __GFP_REPEAT | __GFP_ZERO;
+
+	if (mm == &init_mm)
+		gfp &= ~__GFP_ACCOUNT;
+	page = alloc_pages(gfp, 0);
 	if (!page)
 		return NULL;
 	if (!pgtable_pmd_page_ctor(page)) {
@@ -125,8 +130,11 @@ static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pud_t *pud)
 
 static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)
 {
-	return (pud_t *)__get_free_page(GFP_KERNEL_ACCOUNT|__GFP_REPEAT|
-					__GFP_ZERO);
+	gfp_t gfp = GFP_KERNEL_ACCOUNT | __GFP_REPEAT;
+
+	if (mm == &init_mm)
+		gfp &= ~__GFP_ACCOUNT;
+	return (pud_t *)get_zeroed_page(gfp);
 }
 
 static inline void pud_free(struct mm_struct *mm, pud_t *pud)
diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c
index f86258a..7328ae8 100644
--- a/arch/x86/mm/pgtable.c
+++ b/arch/x86/mm/pgtable.c
@@ -5,7 +5,7 @@
 #include <asm/tlb.h>
 #include <asm/fixmap.h>
 
-#define PGALLOC_GFP GFP_KERNEL_ACCOUNT | __GFP_NOTRACK | __GFP_REPEAT | __GFP_ZERO
+#define PGALLOC_GFP (GFP_KERNEL_ACCOUNT | __GFP_NOTRACK | __GFP_REPEAT | __GFP_ZERO)
 
 #ifdef CONFIG_HIGHPTE
 #define PGALLOC_USER_GFP __GFP_HIGHMEM
@@ -17,7 +17,7 @@ gfp_t __userpte_alloc_gfp = PGALLOC_GFP | PGALLOC_USER_GFP;
 
 pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
 {
-	return (pte_t *)__get_free_page(PGALLOC_GFP);
+	return (pte_t *)__get_free_page(PGALLOC_GFP & ~__GFP_ACCOUNT);
 }
 
 pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address)
@@ -203,13 +203,17 @@ static void free_pmds(pmd_t *pmds[])
 		}
 }
 
-static int preallocate_pmds(pmd_t *pmds[])
+static int preallocate_pmds(struct mm_struct *mm, pmd_t *pmds[])
 {
 	int i;
 	bool failed = false;
+	gfp_t gfp = PGALLOC_GFP;
+
+	if (mm == &init_mm)
+		gfp &= ~__GFP_ACCOUNT;
 
 	for(i = 0; i < PREALLOCATED_PMDS; i++) {
-		pmd_t *pmd = (pmd_t *)__get_free_page(PGALLOC_GFP);
+		pmd_t *pmd = (pmd_t *)__get_free_page(gfp);
 		if (!pmd)
 			failed = true;
 		if (pmd && !pgtable_pmd_page_ctor(virt_to_page(pmd))) {
@@ -287,7 +291,7 @@ pgd_t *pgd_alloc(struct mm_struct *mm)
 
 	mm->pgd = pgd;
 
-	if (preallocate_pmds(pmds) != 0)
+	if (preallocate_pmds(mm, pmds) != 0)
 		goto out_free_pgd;
 
 	if (paravirt_pgd_alloc(mm) != 0)