diff --git a/Documentation/arm/memory.txt b/Documentation/arm/memory.txt
index 4b1c93a8177b..dc6045577a8b 100644
--- a/Documentation/arm/memory.txt
+++ b/Documentation/arm/memory.txt
@@ -1,7 +1,7 @@
 		Kernel Memory Layout on ARM Linux
 
 		Russell King <rmk@arm.linux.org.uk>
-			May 21, 2004 (2.6.6)
+		     November 17, 2005 (2.6.15)
 
 This document describes the virtual memory layout which the Linux
 kernel uses for ARM processors.  It indicates which regions are
@@ -37,6 +37,8 @@ ff000000	ffbfffff	Reserved for future expansion of DMA
 				mapping region.
 
 VMALLOC_END	feffffff	Free for platform use, recommended.
+				VMALLOC_END must be aligned to a 2MB
+				boundary.
 
 VMALLOC_START	VMALLOC_END-1	vmalloc() / ioremap() space.
 				Memory returned by vmalloc/ioremap will
diff --git a/arch/arm/kernel/armksyms.c b/arch/arm/kernel/armksyms.c
index 7a3261f0bf79..9997098009a9 100644
--- a/arch/arm/kernel/armksyms.c
+++ b/arch/arm/kernel/armksyms.c
@@ -120,7 +120,6 @@ EXPORT_SYMBOL(__arch_strncpy_from_user);
 EXPORT_SYMBOL(__get_user_1);
 EXPORT_SYMBOL(__get_user_2);
 EXPORT_SYMBOL(__get_user_4);
-EXPORT_SYMBOL(__get_user_8);
 
 EXPORT_SYMBOL(__put_user_1);
 EXPORT_SYMBOL(__put_user_2);
diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S
index 066597f4345a..f7f183075237 100644
--- a/arch/arm/kernel/entry-common.S
+++ b/arch/arm/kernel/entry-common.S
@@ -48,8 +48,7 @@ work_pending:
 	mov	r0, sp				@ 'regs'
 	mov	r2, why				@ 'syscall'
 	bl	do_notify_resume
-	disable_irq				@ disable interrupts
-	b	no_work_pending
+	b	ret_slow_syscall		@ Check work again
 
 work_resched:
 	bl	schedule
diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c
index a917e3dd3666..765922bcf9e7 100644
--- a/arch/arm/kernel/signal.c
+++ b/arch/arm/kernel/signal.c
@@ -595,23 +595,22 @@ handle_signal(unsigned long sig, struct k_sigaction *ka,
 	 */
 	ret |= !valid_user_regs(regs);
 
-	/*
-	 * Block the signal if we were unsuccessful.
-	 */
 	if (ret != 0) {
-		spin_lock_irq(&tsk->sighand->siglock);
-		sigorsets(&tsk->blocked, &tsk->blocked,
-			  &ka->sa.sa_mask);
-		if (!(ka->sa.sa_flags & SA_NODEFER))
-			sigaddset(&tsk->blocked, sig);
-		recalc_sigpending();
-		spin_unlock_irq(&tsk->sighand->siglock);
+		force_sigsegv(sig, tsk);
+		return;
 	}
 
-	if (ret == 0)
-		return;
+	/*
+	 * Block the signal if we were successful.
+	 */
+	spin_lock_irq(&tsk->sighand->siglock);
+	sigorsets(&tsk->blocked, &tsk->blocked,
+		  &ka->sa.sa_mask);
+	if (!(ka->sa.sa_flags & SA_NODEFER))
+		sigaddset(&tsk->blocked, sig);
+	recalc_sigpending();
+	spin_unlock_irq(&tsk->sighand->siglock);
 
-	force_sigsegv(sig, tsk);
 }
 
 /*
diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S
index 80c8e4c8cefa..9a47770114d4 100644
--- a/arch/arm/kernel/vmlinux.lds.S
+++ b/arch/arm/kernel/vmlinux.lds.S
@@ -172,6 +172,10 @@ SECTIONS
 	.comment 0 : { *(.comment) }
 }
 
-/* those must never be empty */
+/*
+ * These must never be empty
+ * If you have to comment these two assert statements out, your
+ * binutils is too old (for other reasons as well)
+ */
 ASSERT((__proc_info_end - __proc_info_begin), "missing CPU support")
 ASSERT((__arch_info_end - __arch_info_begin), "no machine record defined")
diff --git a/arch/arm/lib/getuser.S b/arch/arm/lib/getuser.S
index d204018070a4..c03ea8e666ba 100644
--- a/arch/arm/lib/getuser.S
+++ b/arch/arm/lib/getuser.S
@@ -54,15 +54,6 @@ __get_user_4:
 	mov	r0, #0
 	mov	pc, lr
 
-	.global	__get_user_8
-__get_user_8:
-5:	ldrt	r2, [r0], #4
-6:	ldrt	r3, [r0]
-	mov	r0, #0
-	mov	pc, lr
-
-__get_user_bad_8:
-	mov	r3, #0
 __get_user_bad:
 	mov	r2, #0
 	mov	r0, #-EFAULT
@@ -73,6 +64,4 @@ __get_user_bad:
 	.long	2b, __get_user_bad
 	.long	3b, __get_user_bad
 	.long	4b, __get_user_bad
-	.long	5b, __get_user_bad_8
-	.long	6b, __get_user_bad_8
 .previous
diff --git a/arch/arm/mm/Makefile b/arch/arm/mm/Makefile
index 59f47d4c2dfe..ffe73ba2bf17 100644
--- a/arch/arm/mm/Makefile
+++ b/arch/arm/mm/Makefile
@@ -51,4 +51,4 @@ obj-$(CONFIG_CPU_ARM1026)	+= proc-arm1026.o
 obj-$(CONFIG_CPU_SA110)		+= proc-sa110.o
 obj-$(CONFIG_CPU_SA1100)	+= proc-sa1100.o
 obj-$(CONFIG_CPU_XSCALE)	+= proc-xscale.o
-obj-$(CONFIG_CPU_V6)		+= proc-v6.o blockops.o
+obj-$(CONFIG_CPU_V6)		+= proc-v6.o
diff --git a/arch/arm/mm/blockops.c b/arch/arm/mm/blockops.c
deleted file mode 100644
index 4f5ee2d08996..000000000000
--- a/arch/arm/mm/blockops.c
+++ /dev/null
@@ -1,185 +0,0 @@
-#include <linux/kernel.h>
-#include <linux/init.h>
-#include <linux/errno.h>
-#include <linux/mm.h>
-
-#include <asm/memory.h>
-#include <asm/ptrace.h>
-#include <asm/cacheflush.h>
-#include <asm/traps.h>
-
-extern struct cpu_cache_fns blk_cache_fns;
-
-#define HARVARD_CACHE
-
-/*
- *	blk_flush_kern_dcache_page(kaddr)
- *
- *	Ensure that the data held in the page kaddr is written back
- *	to the page in question.
- *
- *	- kaddr   - kernel address (guaranteed to be page aligned)
- */
-static void __attribute__((naked))
-blk_flush_kern_dcache_page(void *kaddr)
-{
-	asm(
-	"add	r1, r0, %0							\n\
-	sub	r1, r1, %1							\n\
-1:	.word	0xec401f0e	@ mcrr	p15, 0, r0, r1, c14, 0	@ blocking	\n\
-	mov	r0, #0								\n\
-	mcr	p15, 0, r0, c7, c5, 0						\n\
-	mcr	p15, 0, r0, c7, c10, 4						\n\
-	mov	pc, lr"
-	:
-	: "I" (PAGE_SIZE), "I" (L1_CACHE_BYTES));
-}
-
-/*
- *	blk_dma_inv_range(start,end)
- *
- *	Invalidate the data cache within the specified region; we will
- *	be performing a DMA operation in this region and we want to
- *	purge old data in the cache.
- *
- *	- start   - virtual start address of region
- *	- end     - virtual end address of region
- */
-static void __attribute__((naked))
-blk_dma_inv_range_unified(unsigned long start, unsigned long end)
-{
-	asm(
-	"tst	r0, %0								\n\
-	mcrne	p15, 0, r0, c7, c11, 1		@ clean unified line		\n\
-	tst	r1, %0								\n\
-	mcrne	p15, 0, r1, c7, c15, 1		@ clean & invalidate unified line\n\
-	.word	0xec401f06	@ mcrr	p15, 0, r1, r0, c6, 0	@ blocking	\n\
-	mov	r0, #0								\n\
-	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer		\n\
-	mov	pc, lr"
-	:
-	: "I" (L1_CACHE_BYTES - 1));
-}
-
-static void __attribute__((naked))
-blk_dma_inv_range_harvard(unsigned long start, unsigned long end)
-{
-	asm(
-	"tst	r0, %0								\n\
-	mcrne	p15, 0, r0, c7, c10, 1		@ clean D line			\n\
-	tst	r1, %0								\n\
-	mcrne	p15, 0, r1, c7, c14, 1		@ clean & invalidate D line	\n\
-	.word	0xec401f06	@ mcrr	p15, 0, r1, r0, c6, 0	@ blocking	\n\
-	mov	r0, #0								\n\
-	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer		\n\
-	mov	pc, lr"
-	:
-	: "I" (L1_CACHE_BYTES - 1));
-}
-
-/*
- *	blk_dma_clean_range(start,end)
- *	- start   - virtual start address of region
- *	- end     - virtual end address of region
- */
-static void __attribute__((naked))
-blk_dma_clean_range(unsigned long start, unsigned long end)
-{
-	asm(
-	".word	0xec401f0c	@ mcrr	p15, 0, r1, r0, c12, 0	@ blocking	\n\
-	mov	r0, #0								\n\
-	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer		\n\
-	mov	pc, lr");
-}
-
-/*
- *	blk_dma_flush_range(start,end)
- *	- start   - virtual start address of region
- *	- end     - virtual end address of region
- */
-static void __attribute__((naked))
-blk_dma_flush_range(unsigned long start, unsigned long end)
-{
-	asm(
-	".word	0xec401f0e	@ mcrr	p15, 0, r1, r0, c14, 0	@ blocking	\n\
-	mov	pc, lr");
-}
-
-static int blockops_trap(struct pt_regs *regs, unsigned int instr)
-{
-	regs->ARM_r4 |= regs->ARM_r2;
-	regs->ARM_pc += 4;
-	return 0;
-}
-
-static char *func[] = {
-	"Prefetch data range",
-	"Clean+Invalidate data range",
-	"Clean data range",
-	"Invalidate data range",
-	"Invalidate instr range"
-};
-
-static struct undef_hook blockops_hook __initdata = {
-	.instr_mask	= 0x0fffffd0,
-	.instr_val	= 0x0c401f00,
-	.cpsr_mask	= PSR_T_BIT,
-	.cpsr_val	= 0,
-	.fn		= blockops_trap,
-};
-
-static int __init blockops_check(void)
-{
-	register unsigned int err asm("r4") = 0;
-	unsigned int err_pos = 1;
-	unsigned int cache_type;
-	int i;
-
-	asm("mrc p15, 0, %0, c0, c0, 1" : "=r" (cache_type));
-
-	printk("Checking V6 block cache operations:\n");
-	register_undef_hook(&blockops_hook);
-
-	__asm__ ("mov	r0, %0\n\t"
-		"mov	r1, %1\n\t"
-		"mov	r2, #1\n\t"
-		".word	0xec401f2c @ mcrr p15, 0, r1, r0, c12, 2\n\t"
-		"mov	r2, #2\n\t"
-		".word	0xec401f0e @ mcrr p15, 0, r1, r0, c14, 0\n\t"
-		"mov	r2, #4\n\t"
-		".word	0xec401f0c @ mcrr p15, 0, r1, r0, c12, 0\n\t"
-		"mov	r2, #8\n\t"
-		".word	0xec401f06 @ mcrr p15, 0, r1, r0, c6, 0\n\t"
-		"mov	r2, #16\n\t"
-		".word	0xec401f05 @ mcrr p15, 0, r1, r0, c5, 0\n\t"
-		:
-		: "r" (PAGE_OFFSET), "r" (PAGE_OFFSET + 128)
-		: "r0", "r1", "r2");
-
-	unregister_undef_hook(&blockops_hook);
-
-	for (i = 0; i < ARRAY_SIZE(func); i++, err_pos <<= 1)
-		printk("%30s: %ssupported\n", func[i], err & err_pos ? "not " : "");
-
-	if ((err & 8) == 0) {
-		printk(" --> Using %s block cache invalidate\n",
-			cache_type & (1 << 24) ? "harvard" : "unified");
-		if (cache_type & (1 << 24))
-			cpu_cache.dma_inv_range = blk_dma_inv_range_harvard;
-		else
-			cpu_cache.dma_inv_range = blk_dma_inv_range_unified;
-	}
-	if ((err & 4) == 0) {
-		printk(" --> Using block cache clean\n");
-		cpu_cache.dma_clean_range        = blk_dma_clean_range;
-	}
-	if ((err & 2) == 0) {
-		printk(" --> Using block cache clean+invalidate\n");
-		cpu_cache.dma_flush_range        = blk_dma_flush_range;
-		cpu_cache.flush_kern_dcache_page = blk_flush_kern_dcache_page;
-	}
-
-	return 0;
-}
-
-__initcall(blockops_check);
diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c
index c168f322ef8c..8b276ee38acf 100644
--- a/arch/arm/mm/init.c
+++ b/arch/arm/mm/init.c
@@ -420,7 +420,8 @@ static void __init bootmem_init(struct meminfo *mi)
  * Set up device the mappings.  Since we clear out the page tables for all
  * mappings above VMALLOC_END, we will remove any debug device mappings.
  * This means you have to be careful how you debug this function, or any
- * called function.  (Do it by code inspection!)
+ * called function.  This means you can't use any function or debugging
+ * method which may touch any device, otherwise the kernel _will_ crash.
  */
 static void __init devicemaps_init(struct machine_desc *mdesc)
 {
@@ -428,6 +429,12 @@ static void __init devicemaps_init(struct machine_desc *mdesc)
 	unsigned long addr;
 	void *vectors;
 
+	/*
+	 * Allocate the vector page early.
+	 */
+	vectors = alloc_bootmem_low_pages(PAGE_SIZE);
+	BUG_ON(!vectors);
+
 	for (addr = VMALLOC_END; addr; addr += PGDIR_SIZE)
 		pmd_clear(pmd_off_k(addr));
 
@@ -461,12 +468,6 @@ static void __init devicemaps_init(struct machine_desc *mdesc)
 	create_mapping(&map);
 #endif
 
-	flush_cache_all();
-	local_flush_tlb_all();
-
-	vectors = alloc_bootmem_low_pages(PAGE_SIZE);
-	BUG_ON(!vectors);
-
 	/*
 	 * Create a mapping for the machine vectors at the high-vectors
 	 * location (0xffff0000).  If we aren't using high-vectors, also
@@ -491,12 +492,13 @@ static void __init devicemaps_init(struct machine_desc *mdesc)
 		mdesc->map_io();
 
 	/*
-	 * Finally flush the tlb again - this ensures that we're in a
-	 * consistent state wrt the writebuffer if the writebuffer needs
-	 * draining.  After this point, we can start to touch devices
-	 * again.
+	 * Finally flush the caches and tlb to ensure that we're in a
+	 * consistent state wrt the writebuffer.  This also ensures that
+	 * any write-allocated cache lines in the vector page are written
+	 * back.  After this point, we can start to touch devices again.
 	 */
 	local_flush_tlb_all();
+	flush_cache_all();
 }
 
 /*
diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c
index 0f128c28fee4..10901398e4a2 100644
--- a/arch/arm/mm/ioremap.c
+++ b/arch/arm/mm/ioremap.c
@@ -130,8 +130,7 @@ remap_area_pages(unsigned long start, unsigned long phys_addr,
  * mapping.  See include/asm-arm/proc-armv/pgtable.h for more information.
  */
 void __iomem *
-__ioremap(unsigned long phys_addr, size_t size, unsigned long flags,
-	  unsigned long align)
+__ioremap(unsigned long phys_addr, size_t size, unsigned long flags)
 {
 	void * addr;
 	struct vm_struct * area;
diff --git a/drivers/mtd/maps/ipaq-flash.c b/drivers/mtd/maps/ipaq-flash.c
index 35097c9bbf50..b8ccb0a95789 100644
--- a/drivers/mtd/maps/ipaq-flash.c
+++ b/drivers/mtd/maps/ipaq-flash.c
@@ -246,7 +246,7 @@ int __init ipaq_mtd_init(void)
 			ipaq_map[i].size = h3xxx_max_flash_size;
 			ipaq_map[i].set_vpp = h3xxx_set_vpp;
 			ipaq_map[i].phys = cs_phys[i];
-			ipaq_map[i].virt = __ioremap(cs_phys[i], 0x04000000, 0, 1);
+			ipaq_map[i].virt = ioremap(cs_phys[i], 0x04000000);
 			if (machine_is_h3100 () || machine_is_h1900())
 				ipaq_map[i].bankwidth = 2;
 		}
@@ -280,7 +280,7 @@ int __init ipaq_mtd_init(void)
 		nb_parts = ARRAY_SIZE(jornada_partitions);
 		ipaq_map[0].size = jornada_max_flash_size;
 		ipaq_map[0].set_vpp = jornada56x_set_vpp;
-		ipaq_map[0].virt = (__u32)__ioremap(0x0, 0x04000000, 0, 1);
+		ipaq_map[0].virt = (__u32)ioremap(0x0, 0x04000000);
 	}
 #endif
 #ifdef CONFIG_SA1100_JORNADA720
@@ -442,7 +442,7 @@ static int __init h1900_special_case(void)
 	ipaq_map[0].size = 0x80000;
 	ipaq_map[0].set_vpp = h3xxx_set_vpp;
 	ipaq_map[0].phys = 0x0;
-	ipaq_map[0].virt = __ioremap(0x0, 0x04000000, 0, 1);
+	ipaq_map[0].virt = ioremap(0x0, 0x04000000);
 	ipaq_map[0].bankwidth = 2;
 
 	printk(KERN_NOTICE "iPAQ flash: probing %d-bit flash bus, window=%lx with JEDEC.\n", ipaq_map[0].bankwidth*8, ipaq_map[0].virt);
diff --git a/drivers/mtd/nand/h1910.c b/drivers/mtd/nand/h1910.c
index 041e4b3358fb..f68f7a99a630 100644
--- a/drivers/mtd/nand/h1910.c
+++ b/drivers/mtd/nand/h1910.c
@@ -112,7 +112,7 @@ static int __init h1910_init (void)
 	if (!machine_is_h1900())
 		return -ENODEV;
 
-	nandaddr = __ioremap(0x08000000, 0x1000, 0, 1);
+	nandaddr = ioremap(0x08000000, 0x1000);
 	if (!nandaddr) {
 		printk("Failed to ioremap nand flash.\n");
 		return -ENOMEM;
diff --git a/include/asm-arm/arch-ixp4xx/io.h b/include/asm-arm/arch-ixp4xx/io.h
index 688f7f90d93e..942b622455bc 100644
--- a/include/asm-arm/arch-ixp4xx/io.h
+++ b/include/asm-arm/arch-ixp4xx/io.h
@@ -59,11 +59,10 @@ extern int ixp4xx_pci_write(u32 addr, u32 cmd, u32 data);
  * fallback to the default.
  */
 static inline void __iomem *
-__ixp4xx_ioremap(unsigned long addr, size_t size, unsigned long flags, unsigned long align)
+__ixp4xx_ioremap(unsigned long addr, size_t size, unsigned long flags)
 {
-	extern void __iomem * __ioremap(unsigned long, size_t, unsigned long, unsigned long);
 	if((addr < 0x48000000) || (addr > 0x4fffffff))
-		return __ioremap(addr, size, flags, align);
+		return __ioremap(addr, size, flags);
 
 	return (void *)addr;
 }
@@ -71,13 +70,11 @@ __ixp4xx_ioremap(unsigned long addr, size_t size, unsigned long flags, unsigned
 static inline void
 __ixp4xx_iounmap(void __iomem *addr)
 {
-	extern void __iounmap(void __iomem *addr);
-
 	if ((u32)addr >= VMALLOC_START)
 		__iounmap(addr);
 }
 
-#define __arch_ioremap(a, s, f, x)	__ixp4xx_ioremap(a, s, f, x)
+#define __arch_ioremap(a, s, f)		__ixp4xx_ioremap(a, s, f)
 #define	__arch_iounmap(a)		__ixp4xx_iounmap(a)
 
 #define	writeb(v, p)			__ixp4xx_writeb(v, p)
diff --git a/include/asm-arm/io.h b/include/asm-arm/io.h
index 2e6799632f12..ae69db4a1010 100644
--- a/include/asm-arm/io.h
+++ b/include/asm-arm/io.h
@@ -54,6 +54,12 @@ extern void __raw_readsl(void __iomem *addr, void *data, int longlen);
 #define __raw_readw(a)		(__chk_io_ptr(a), *(volatile unsigned short __force *)(a))
 #define __raw_readl(a)		(__chk_io_ptr(a), *(volatile unsigned int __force   *)(a))
 
+/*
+ * Architecture ioremap implementation.
+ */
+extern void __iomem * __ioremap(unsigned long, size_t, unsigned long);
+extern void __iounmap(void __iomem *addr);
+
 /*
  * Bad read/write accesses...
  */
@@ -256,18 +262,15 @@ out:
  * ioremap takes a PCI memory address, as specified in
  * Documentation/IO-mapping.txt.
  */
-extern void __iomem * __ioremap(unsigned long, size_t, unsigned long, unsigned long);
-extern void __iounmap(void __iomem *addr);
-
 #ifndef __arch_ioremap
-#define ioremap(cookie,size)		__ioremap(cookie,size,0,1)
-#define ioremap_nocache(cookie,size)	__ioremap(cookie,size,0,1)
-#define ioremap_cached(cookie,size)	__ioremap(cookie,size,L_PTE_CACHEABLE,1)
+#define ioremap(cookie,size)		__ioremap(cookie,size,0)
+#define ioremap_nocache(cookie,size)	__ioremap(cookie,size,0)
+#define ioremap_cached(cookie,size)	__ioremap(cookie,size,L_PTE_CACHEABLE)
 #define iounmap(cookie)			__iounmap(cookie)
 #else
-#define ioremap(cookie,size)		__arch_ioremap((cookie),(size),0,1)
-#define ioremap_nocache(cookie,size)	__arch_ioremap((cookie),(size),0,1)
-#define ioremap_cached(cookie,size)	__arch_ioremap((cookie),(size),L_PTE_CACHEABLE,1)
+#define ioremap(cookie,size)		__arch_ioremap((cookie),(size),0)
+#define ioremap_nocache(cookie,size)	__arch_ioremap((cookie),(size),0)
+#define ioremap_cached(cookie,size)	__arch_ioremap((cookie),(size),L_PTE_CACHEABLE)
 #define iounmap(cookie)			__arch_iounmap(cookie)
 #endif
 
diff --git a/include/asm-arm/uaccess.h b/include/asm-arm/uaccess.h
index a2fdad0138b3..064f0f5e8e2b 100644
--- a/include/asm-arm/uaccess.h
+++ b/include/asm-arm/uaccess.h
@@ -100,7 +100,6 @@ static inline void set_fs (mm_segment_t fs)
 extern int __get_user_1(void *);
 extern int __get_user_2(void *);
 extern int __get_user_4(void *);
-extern int __get_user_8(void *);
 extern int __get_user_bad(void);
 
 #define __get_user_x(__r2,__p,__e,__s,__i...)				\
@@ -114,7 +113,7 @@ extern int __get_user_bad(void);
 #define get_user(x,p)							\
 	({								\
 		const register typeof(*(p)) __user *__p asm("r0") = (p);\
-		register typeof(*(p)) __r2 asm("r2");			\
+		register unsigned int __r2 asm("r2");			\
 		register int __e asm("r0");				\
 		switch (sizeof(*(__p))) {				\
 		case 1:							\
@@ -126,12 +125,9 @@ extern int __get_user_bad(void);
 		case 4:							\
 	       		__get_user_x(__r2, __p, __e, 4, "lr");		\
 			break;						\
-		case 8:							\
-			__get_user_x(__r2, __p, __e, 8, "lr");		\
-	       		break;						\
 		default: __e = __get_user_bad(); break;			\
 		}							\
-		x = __r2;						\
+		x = (typeof(*(p))) __r2;				\
 		__e;							\
 	})