Merge branch 'for-tip' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/percpu into x86/urgent

54 files changed, 2528 insertions, 282 deletions

diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 08def8deb5f..ecad946920d 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -1882,6 +1882,12 @@ and is between 256 and 4096 characters. It is defined in the file
 			Format: { 0 | 1 }
 			See arch/parisc/kernel/pdc_chassis.c
 
+	percpu_alloc=	[X86] Select which percpu first chunk allocator to use.
+			Allowed values are one of "lpage", "embed" and "4k".
+			See comments in arch/x86/kernel/setup_percpu.c for
+			details on each allocator.  This parameter is primarily
+			for debugging and performance comparison.
+
 	pf.		[PARIDE]
 			See Documentation/blockdev/paride.txt.
 
diff --git a/Documentation/powerpc/dts-bindings/fsl/esdhc.txt b/Documentation/powerpc/dts-bindings/fsl/esdhc.txt
index 5093ddf900d..3ed3797b508 100644
--- a/Documentation/powerpc/dts-bindings/fsl/esdhc.txt
+++ b/Documentation/powerpc/dts-bindings/fsl/esdhc.txt
@@ -10,6 +10,8 @@ Required properties:
   - interrupts : should contain eSDHC interrupt.
   - interrupt-parent : interrupt source phandle.
   - clock-frequency : specifies eSDHC base clock frequency.
+  - sdhci,1-bit-only : (optional) specifies that a controller can
+    only handle 1-bit data transfers.
 
 Example:
 
diff --git a/Documentation/sound/alsa/HD-Audio-Models.txt b/Documentation/sound/alsa/HD-Audio-Models.txt
index de8e10a9410..0d8d23581c4 100644
--- a/Documentation/sound/alsa/HD-Audio-Models.txt
+++ b/Documentation/sound/alsa/HD-Audio-Models.txt
@@ -139,6 +139,7 @@ ALC883/888
   acer		Acer laptops (Travelmate 3012WTMi, Aspire 5600, etc)
   acer-aspire	Acer Aspire 9810
   acer-aspire-4930g Acer Aspire 4930G
+  acer-aspire-6530g Acer Aspire 6530G
   acer-aspire-8930g Acer Aspire 8930G
   medion	Medion Laptops
   medion-md2	Medion MD2
diff --git a/MAINTAINERS b/MAINTAINERS
index dc226e78612..1d4704300c1 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1010,6 +1010,13 @@ W:	http://www.at91.com/
 S:	Maintained
 F:	drivers/mmc/host/at91_mci.c
 
+ATMEL AT91 / AT32 MCI DRIVER
+P:	Nicolas Ferre
+M:	nicolas.ferre@atmel.com
+S:	Maintained
+F:	drivers/mmc/host/atmel-mci.c
+F:	drivers/mmc/host/atmel-mci-regs.h
+
 ATMEL AT91 / AT32 SERIAL DRIVER
 P:	Haavard Skinnemoen
 M:	hskinnemoen@atmel.com
@@ -5094,6 +5101,13 @@ L:	sdhci-devel@lists.ossman.eu
 S:	Maintained
 F:	drivers/mmc/host/sdhci.*
 
+SECURE DIGITAL HOST CONTROLLER INTERFACE (SDHCI) SAMSUNG DRIVER
+P:	Ben Dooks
+M:	ben-linux@fluff.org
+L:	sdhci-devel@lists.ossman.eu
+S:	Maintained
+F:	drivers/mmc/host/sdhci-s3c.c
+
 SECURITY SUBSYSTEM
 P:	James Morris
 M:	jmorris@namei.org
@@ -6216,6 +6230,14 @@ S:	Maintained
 F:	Documentation/i2c/busses/i2c-viapro
 F:	drivers/i2c/busses/i2c-viapro.c
 
+VIA SD/MMC CARD CONTROLLER DRIVER
+P:	Joseph Chan
+M:	JosephChan@via.com.tw
+P:	Harald Welte
+M:	HaraldWelte@viatech.com
+S:	Maintained
+F:	drivers/mmc/host/via-sdmmc.c
+
 VIA UNICHROME(PRO)/CHROME9 FRAMEBUFFER DRIVER
 P:	Joseph Chan
 M:	JosephChan@via.com.tw
diff --git a/arch/alpha/mm/fault.c b/arch/alpha/mm/fault.c
index 4829f96585b..00a31deaa96 100644
--- a/arch/alpha/mm/fault.c
+++ b/arch/alpha/mm/fault.c
@@ -146,7 +146,7 @@ do_page_fault(unsigned long address, unsigned long mmcsr,
 	/* If for any reason at all we couldn't handle the fault,
 	   make sure we exit gracefully rather than endlessly redo
 	   the fault.  */
-	fault = handle_mm_fault(mm, vma, address, cause > 0);
+	fault = handle_mm_fault(mm, vma, address, cause > 0 ? FAULT_FLAG_WRITE : 0);
 	up_read(&mm->mmap_sem);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c
index 0455557a289..6fdcbb70982 100644
--- a/arch/arm/mm/fault.c
+++ b/arch/arm/mm/fault.c
@@ -208,7 +208,7 @@ good_area:
 	 * than endlessly redo the fault.
 	 */
 survive:
-	fault = handle_mm_fault(mm, vma, addr & PAGE_MASK, fsr & (1 << 11));
+	fault = handle_mm_fault(mm, vma, addr & PAGE_MASK, (fsr & (1 << 11)) ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/avr32/mm/fault.c b/arch/avr32/mm/fault.c
index 62d4abbaa65..b61d86d3deb 100644
--- a/arch/avr32/mm/fault.c
+++ b/arch/avr32/mm/fault.c
@@ -133,7 +133,7 @@ good_area:
 	 * fault.
 	 */
 survive:
-	fault = handle_mm_fault(mm, vma, address, writeaccess);
+	fault = handle_mm_fault(mm, vma, address, writeaccess ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/cris/mm/fault.c b/arch/cris/mm/fault.c
index c4c76db90f9..f925115e325 100644
--- a/arch/cris/mm/fault.c
+++ b/arch/cris/mm/fault.c
@@ -163,7 +163,7 @@ do_page_fault(unsigned long address, struct pt_regs *regs,
 	 * the fault.
 	 */
 
-	fault = handle_mm_fault(mm, vma, address, writeaccess & 1);
+	fault = handle_mm_fault(mm, vma, address, (writeaccess & 1) ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/frv/mm/fault.c b/arch/frv/mm/fault.c
index 05093d41d98..30f5d100a81 100644
--- a/arch/frv/mm/fault.c
+++ b/arch/frv/mm/fault.c
@@ -163,7 +163,7 @@ asmlinkage void do_page_fault(int datammu, unsigned long esr0, unsigned long ear
 	 * make sure we exit gracefully rather than endlessly redo
 	 * the fault.
 	 */
-	fault = handle_mm_fault(mm, vma, ear0, write);
+	fault = handle_mm_fault(mm, vma, ear0, write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c
index 23088bed111..19261a99e62 100644
--- a/arch/ia64/mm/fault.c
+++ b/arch/ia64/mm/fault.c
@@ -154,7 +154,7 @@ ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *re
 	 * sure we exit gracefully rather than endlessly redo the
 	 * fault.
 	 */
-	fault = handle_mm_fault(mm, vma, address, (mask & VM_WRITE) != 0);
+	fault = handle_mm_fault(mm, vma, address, (mask & VM_WRITE) ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		/*
 		 * We ran out of memory, or some other thing happened
diff --git a/arch/m32r/mm/fault.c b/arch/m32r/mm/fault.c
index 4a71df4c1b3..7274b47f4c2 100644
--- a/arch/m32r/mm/fault.c
+++ b/arch/m32r/mm/fault.c
@@ -196,7 +196,7 @@ survive:
 	 */
 	addr = (address & PAGE_MASK);
 	set_thread_fault_code(error_code);
-	fault = handle_mm_fault(mm, vma, addr, write);
+	fault = handle_mm_fault(mm, vma, addr, write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/m68k/mm/fault.c b/arch/m68k/mm/fault.c
index f493f03231d..d0e35cf99fc 100644
--- a/arch/m68k/mm/fault.c
+++ b/arch/m68k/mm/fault.c
@@ -155,7 +155,7 @@ good_area:
 	 */
 
  survive:
-	fault = handle_mm_fault(mm, vma, address, write);
+	fault = handle_mm_fault(mm, vma, address, write ? FAULT_FLAG_WRITE : 0);
 #ifdef DEBUG
 	printk("handle_mm_fault returns %d\n",fault);
 #endif
diff --git a/arch/microblaze/mm/fault.c b/arch/microblaze/mm/fault.c
index 5e67cd1fab4..956607a63f4 100644
--- a/arch/microblaze/mm/fault.c
+++ b/arch/microblaze/mm/fault.c
@@ -232,7 +232,7 @@ good_area:
 	 * the fault.
 	 */
 survive:
-	fault = handle_mm_fault(mm, vma, address, is_write);
+	fault = handle_mm_fault(mm, vma, address, is_write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c
index 55767ad9f00..6751ce9ede9 100644
--- a/arch/mips/mm/fault.c
+++ b/arch/mips/mm/fault.c
@@ -102,7 +102,7 @@ good_area:
 	 * make sure we exit gracefully rather than endlessly redo
 	 * the fault.
 	 */
-	fault = handle_mm_fault(mm, vma, address, write);
+	fault = handle_mm_fault(mm, vma, address, write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/mn10300/mm/fault.c b/arch/mn10300/mm/fault.c
index 33cf25025da..a62e1e138bc 100644
--- a/arch/mn10300/mm/fault.c
+++ b/arch/mn10300/mm/fault.c
@@ -258,7 +258,7 @@ good_area:
 	 * make sure we exit gracefully rather than endlessly redo
 	 * the fault.
 	 */
-	fault = handle_mm_fault(mm, vma, address, write);
+	fault = handle_mm_fault(mm, vma, address, write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/parisc/mm/fault.c b/arch/parisc/mm/fault.c
index 92c7fa4ecc3..bfb6dd6ab38 100644
--- a/arch/parisc/mm/fault.c
+++ b/arch/parisc/mm/fault.c
@@ -202,7 +202,7 @@ good_area:
 	 * fault.
 	 */
 
-	fault = handle_mm_fault(mm, vma, address, (acc_type & VM_WRITE) != 0);
+	fault = handle_mm_fault(mm, vma, address, (acc_type & VM_WRITE) ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		/*
 		 * We hit a shared mapping outside of the file, or some
diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c
index 5beffc8f481..830bef0a113 100644
--- a/arch/powerpc/mm/fault.c
+++ b/arch/powerpc/mm/fault.c
@@ -302,7 +302,7 @@ good_area:
 	 * the fault.
 	 */
  survive:
-	ret = handle_mm_fault(mm, vma, address, is_write);
+	ret = handle_mm_fault(mm, vma, address, is_write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(ret & VM_FAULT_ERROR)) {
 		if (ret & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/powerpc/platforms/cell/spu_fault.c b/arch/powerpc/platforms/cell/spu_fault.c
index 95d8dadf2d8..d06ba87f1a1 100644
--- a/arch/powerpc/platforms/cell/spu_fault.c
+++ b/arch/powerpc/platforms/cell/spu_fault.c
@@ -70,7 +70,7 @@ int spu_handle_mm_fault(struct mm_struct *mm, unsigned long ea,
 	}
 
 	ret = 0;
-	*flt = handle_mm_fault(mm, vma, ea, is_write);
+	*flt = handle_mm_fault(mm, vma, ea, is_write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(*flt & VM_FAULT_ERROR)) {
 		if (*flt & VM_FAULT_OOM) {
 			ret = -ENOMEM;
diff --git a/arch/s390/lib/uaccess_pt.c b/arch/s390/lib/uaccess_pt.c
index b0b84c35b0a..cb5d59eab0e 100644
--- a/arch/s390/lib/uaccess_pt.c
+++ b/arch/s390/lib/uaccess_pt.c
@@ -66,7 +66,7 @@ static int __handle_fault(struct mm_struct *mm, unsigned long address,
 	}
 
 survive:
-	fault = handle_mm_fault(mm, vma, address, write_access);
+	fault = handle_mm_fault(mm, vma, address, write_access ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
index 220a152c836..74eb26bf197 100644
--- a/arch/s390/mm/fault.c
+++ b/arch/s390/mm/fault.c
@@ -352,7 +352,7 @@ good_area:
 	 * make sure we exit gracefully rather than endlessly redo
 	 * the fault.
 	 */
-	fault = handle_mm_fault(mm, vma, address, write);
+	fault = handle_mm_fault(mm, vma, address, write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM) {
 			up_read(&mm->mmap_sem);
diff --git a/arch/sh/mm/fault_32.c b/arch/sh/mm/fault_32.c
index 2c50f80fc33..cc8ddbdf3d7 100644
--- a/arch/sh/mm/fault_32.c
+++ b/arch/sh/mm/fault_32.c
@@ -133,7 +133,7 @@ good_area:
 	 * the fault.
 	 */
 survive:
-	fault = handle_mm_fault(mm, vma, address, writeaccess);
+	fault = handle_mm_fault(mm, vma, address, writeaccess ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/sh/mm/tlbflush_64.c b/arch/sh/mm/tlbflush_64.c
index 7876997ba19..fcbb6e135ce 100644
--- a/arch/sh/mm/tlbflush_64.c
+++ b/arch/sh/mm/tlbflush_64.c
@@ -187,7 +187,7 @@ good_area:
 	 * the fault.
 	 */
 survive:
-	fault = handle_mm_fault(mm, vma, address, writeaccess);
+	fault = handle_mm_fault(mm, vma, address, writeaccess ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/sparc/mm/fault_32.c b/arch/sparc/mm/fault_32.c
index 12e447fc854..a5e30c642ee 100644
--- a/arch/sparc/mm/fault_32.c
+++ b/arch/sparc/mm/fault_32.c
@@ -241,7 +241,7 @@ good_area:
 	 * make sure we exit gracefully rather than endlessly redo
 	 * the fault.
 	 */
-	fault = handle_mm_fault(mm, vma, address, write);
+	fault = handle_mm_fault(mm, vma, address, write ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
@@ -484,7 +484,7 @@ good_area:
 		if(!(vma->vm_flags & (VM_READ | VM_EXEC)))
 			goto bad_area;
 	}
-	switch (handle_mm_fault(mm, vma, address, write)) {
+	switch (handle_mm_fault(mm, vma, address, write ? FAULT_FLAG_WRITE : 0)) {
 	case VM_FAULT_SIGBUS:
 	case VM_FAULT_OOM:
 		goto do_sigbus;
diff --git a/arch/sparc/mm/fault_64.c b/arch/sparc/mm/fault_64.c
index 4ab8993b086..e5620b27c8b 100644
--- a/arch/sparc/mm/fault_64.c
+++ b/arch/sparc/mm/fault_64.c
@@ -398,7 +398,7 @@ good_area:
 			goto bad_area;
 	}
 
-	fault = handle_mm_fault(mm, vma, address, (fault_code & FAULT_CODE_WRITE));
+	fault = handle_mm_fault(mm, vma, address, (fault_code & FAULT_CODE_WRITE) ? FAULT_FLAG_WRITE : 0);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		if (fault & VM_FAULT_OOM)
 			goto out_of_memory;
diff --git a/arch/um/kernel/trap.c b/arch/um/kernel/trap.c
index 7384d8accfe..637c6505dc0 100644
--- a/arch/um/kernel/trap.c
+++ b/arch/um/kernel/trap.c
@@ -65,7 +65,7 @@ good_area:
 	do {
 		int fault;
 
-		fault = handle_mm_fault(mm, vma, address, is_write);
+		fault = handle_mm_fault(mm, vma, address, is_write ? FAULT_FLAG_WRITE : 0);
 		if (unlikely(fault & VM_FAULT_ERROR)) {
 			if (fault & VM_FAULT_OOM) {
 				goto out_of_memory;
diff --git a/arch/x86/crypto/aesni-intel_asm.S b/arch/x86/crypto/aesni-intel_asm.S
index caba9960170..eb0566e8331 100644
--- a/arch/x86/crypto/aesni-intel_asm.S
+++ b/arch/x86/crypto/aesni-intel_asm.S
@@ -845,7 +845,7 @@ ENTRY(aesni_cbc_enc)
  */
 ENTRY(aesni_cbc_dec)
 	cmp $16, LEN
-	jb .Lcbc_dec_ret
+	jb .Lcbc_dec_just_ret
 	mov 480(KEYP), KLEN
 	add $240, KEYP
 	movups (IVP), IV
@@ -891,6 +891,7 @@ ENTRY(aesni_cbc_dec)
 	add $16, OUTP
 	cmp $16, LEN
 	jge .Lcbc_dec_loop1
-	movups IV, (IVP)
 .Lcbc_dec_ret:
+	movups IV, (IVP)
+.Lcbc_dec_just_ret:
 	ret
diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c
index 4e663398f77..c580c5ec1ca 100644
--- a/arch/x86/crypto/aesni-intel_glue.c
+++ b/arch/x86/crypto/aesni-intel_glue.c
@@ -198,6 +198,7 @@ static int ecb_encrypt(struct blkcipher_desc *desc,
 
 	blkcipher_walk_init(&walk, dst, src, nbytes);
 	err = blkcipher_walk_virt(desc, &walk);
+	desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
 
 	kernel_fpu_begin();
 	while ((nbytes = walk.nbytes)) {
@@ -221,6 +222,7 @@ static int ecb_decrypt(struct blkcipher_desc *desc,
 
 	blkcipher_walk_init(&walk, dst, src, nbytes);
 	err = blkcipher_walk_virt(desc, &walk);
+	desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
 
 	kernel_fpu_begin();
 	while ((nbytes = walk.nbytes)) {
@@ -266,6 +268,7 @@ static int cbc_encrypt(struct blkcipher_desc *desc,
 
 	blkcipher_walk_init(&walk, dst, src, nbytes);
 	err = blkcipher_walk_virt(desc, &walk);
+	desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
 
 	kernel_fpu_begin();
 	while ((nbytes = walk.nbytes)) {
@@ -289,6 +292,7 @@ static int cbc_decrypt(struct blkcipher_desc *desc,
 
 	blkcipher_walk_init(&walk, dst, src, nbytes);
 	err = blkcipher_walk_virt(desc, &walk);
+	desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
 
 	kernel_fpu_begin();
 	while ((nbytes = walk.nbytes)) {
diff --git a/arch/x86/crypto/fpu.c b/arch/x86/crypto/fpu.c
index 5f9781a3815..daef6cd2b45 100644
--- a/arch/x86/crypto/fpu.c
+++ b/arch/x86/crypto/fpu.c
@@ -48,7 +48,7 @@ static int crypto_fpu_encrypt(struct blkcipher_desc *desc_in,
 	struct blkcipher_desc desc = {
 		.tfm = child,
 		.info = desc_in->info,
-		.flags = desc_in->flags,
+		.flags = desc_in->flags & ~CRYPTO_TFM_REQ_MAY_SLEEP,
 	};
 
 	kernel_fpu_begin();
@@ -67,7 +67,7 @@ static int crypto_fpu_decrypt(struct blkcipher_desc *desc_in,
 	struct blkcipher_desc desc = {
 		.tfm = child,
 		.info = desc_in->info,
-		.flags = desc_in->flags,
+		.flags = desc_in->flags & ~CRYPTO_TFM_REQ_MAY_SLEEP,
 	};
 
 	kernel_fpu_begin();
diff --git a/arch/x86/include/asm/percpu.h b/arch/x86/include/asm/percpu.h
index 02ecb30982a..103f1ddb0d8 100644
--- a/arch/x86/include/asm/percpu.h
+++ b/arch/x86/include/asm/percpu.h
@@ -42,6 +42,7 @@
 
 #else /* ...!ASSEMBLY */
 
+#include <linux/kernel.h>
 #include <linux/stringify.h>
 
 #ifdef CONFIG_SMP
@@ -155,6 +156,15 @@ do {							\
 /* We can use this directly for local CPU (faster). */
 DECLARE_PER_CPU(unsigned long, this_cpu_off);
 
+#ifdef CONFIG_NEED_MULTIPLE_NODES
+void *pcpu_lpage_remapped(void *kaddr);
+#else
+static inline void *pcpu_lpage_remapped(void *kaddr)
+{
+	return NULL;
+}
+#endif
+
 #endif /* !__ASSEMBLY__ */
 
 #ifdef CONFIG_SMP
diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c
index 9c3f0823e6a..29a3eef7cf4 100644
--- a/arch/x86/kernel/setup_percpu.c
+++ b/arch/x86/kernel/setup_percpu.c
@@ -124,7 +124,7 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size,
 }
 
 /*
- * Remap allocator
+ * Large page remap allocator
  *
  * This allocator uses PMD page as unit.  A PMD page is allocated for
  * each cpu and each is remapped into vmalloc area using PMD mapping.
@@ -137,105 +137,185 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size,
  * better than only using 4k mappings while still being NUMA friendly.
  */
 #ifdef CONFIG_NEED_MULTIPLE_NODES
-static size_t pcpur_size __initdata;
-static void **pcpur_ptrs __initdata;
+struct pcpul_ent {
+	unsigned int	cpu;
+	void		*ptr;
+};
+
+static size_t pcpul_size;
+static struct pcpul_ent *pcpul_map;
+static struct vm_struct pcpul_vm;
 
-static struct page * __init pcpur_get_page(unsigned int cpu, int pageno)
+static struct page * __init pcpul_get_page(unsigned int cpu, int pageno)
 {
 	size_t off = (size_t)pageno << PAGE_SHIFT;
 
-	if (off >= pcpur_size)
+	if (off >= pcpul_size)
 		return NULL;
 
-	return virt_to_page(pcpur_ptrs[cpu] + off);
+	return virt_to_page(pcpul_map[cpu].ptr + off);
 }
 
-static ssize_t __init setup_pcpu_remap(size_t static_size)
+static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen)
 {
-	static struct vm_struct vm;
-	size_t ptrs_size, dyn_size;
+	size_t map_size, dyn_size;
 	unsigned int cpu;
+	int i, j;
 	ssize_t ret;
 
-	/*
-	 * If large page isn't supported, there's no benefit in doing
-	 * this.  Also, on non-NUMA, embedding is better.
-	 *
-	 * NOTE: disabled for now.
-	 */
-	if (true || !cpu_has_pse || !pcpu_need_numa())
+	if (!chosen) {
+		size_t vm_size = VMALLOC_END - VMALLOC_START;
+		size_t tot_size = num_possible_cpus() * PMD_SIZE;
+
+		/* on non-NUMA, embedding is better */
+		if (!pcpu_need_numa())
+			return -EINVAL;
+
+		/* don't consume more than 20% of vmalloc area */
+		if (tot_size > vm_size / 5) {
+			pr_info("PERCPU: too large chunk size %zuMB for "
+				"large page remap\n", tot_size >> 20);
+			return -EINVAL;
+		}
+	}
+
+	/* need PSE */
+	if (!cpu_has_pse) {
+		pr_warning("PERCPU: lpage allocator requires PSE\n");
 		return -EINVAL;
+	}
 
 	/*
 	 * Currently supports only single page.  Supporting multiple
 	 * pages won't be too difficult if it ever becomes necessary.
 	 */
-	pcpur_size = PFN_ALIGN(static_size + PERCPU_MODULE_RESERVE +
+	pcpul_size = PFN_ALIGN(static_size + PERCPU_MODULE_RESERVE +
 			       PERCPU_DYNAMIC_RESERVE);
-	if (pcpur_size > PMD_SIZE) {
+	if (pcpul_size > PMD_SIZE) {
 		pr_warning("PERCPU: static data is larger than large page, "
 			   "can't use large page\n");