diff --git a/arch/ia64/hp/common/sba_iommu.c b/arch/ia64/hp/common/sba_iommu.c
index 4ce8d1358feecea321af29f6b0be5e8b1dde6bae..c04dd576f3331e544fb6f746abc56288fa62e300 100644
--- a/arch/ia64/hp/common/sba_iommu.c
+++ b/arch/ia64/hp/common/sba_iommu.c
@@ -37,6 +37,7 @@
 #include <linux/crash_dump.h>
 #include <linux/iommu-helper.h>
 #include <linux/dma-mapping.h>
+#include <linux/prefetch.h>
 
 #include <asm/delay.h>		/* ia64_get_itc() */
 #include <asm/io.h>
diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c
index 0799fea4c5886643114bc767a8098b7fd246d34f..20b3593761282746777f2f27304e5ac03bdee1da 100644
--- a/arch/ia64/mm/fault.c
+++ b/arch/ia64/mm/fault.c
@@ -10,6 +10,7 @@
 #include <linux/interrupt.h>
 #include <linux/kprobes.h>
 #include <linux/kdebug.h>
+#include <linux/prefetch.h>
 
 #include <asm/pgtable.h>
 #include <asm/processor.h>
diff --git a/arch/powerpc/lib/sstep.c b/arch/powerpc/lib/sstep.c
index ae5189ab004926072e2f86e04b4b8aa33c0d0867..f73daa6f397007eb0f6e1c728c8ca1768b3c7f0b 100644
--- a/arch/powerpc/lib/sstep.c
+++ b/arch/powerpc/lib/sstep.c
@@ -11,6 +11,7 @@
 #include <linux/kernel.h>
 #include <linux/kprobes.h>
 #include <linux/ptrace.h>
+#include <linux/prefetch.h>
 #include <asm/sstep.h>
 #include <asm/processor.h>
 #include <asm/uaccess.h>
diff --git a/arch/sh/kernel/cpu/sh4/sq.c b/arch/sh/kernel/cpu/sh4/sq.c
index 14726eef1ce0cf62ecfbe1ae45cecb2aad0e278b..f0907995b4c90a7747136ffab7dee04629620838 100644
--- a/arch/sh/kernel/cpu/sh4/sq.c
+++ b/arch/sh/kernel/cpu/sh4/sq.c
@@ -20,6 +20,7 @@
 #include <linux/vmalloc.h>
 #include <linux/mm.h>
 #include <linux/io.h>
+#include <linux/prefetch.h>
 #include <asm/page.h>
 #include <asm/cacheflush.h>
 #include <cpu/sq.h>
diff --git a/arch/x86/include/asm/uaccess.h b/arch/x86/include/asm/uaccess.h
index 99f0ad753f32c482f9f5b222c84cee881696b628..99ddd148a760ff73561d42a1723ff4f0b158210e 100644
--- a/arch/x86/include/asm/uaccess.h
+++ b/arch/x86/include/asm/uaccess.h
@@ -6,7 +6,6 @@
 #include <linux/errno.h>
 #include <linux/compiler.h>
 #include <linux/thread_info.h>
-#include <linux/prefetch.h>
 #include <linux/string.h>
 #include <asm/asm.h>
 #include <asm/page.h>
diff --git a/arch/x86/include/asm/uaccess_32.h b/arch/x86/include/asm/uaccess_32.h
index 088d09fb1615af8081872b04e9b2b3ec15e1fbad..566e803cc6026a11a1391fd25051717cef3f5979 100644
--- a/arch/x86/include/asm/uaccess_32.h
+++ b/arch/x86/include/asm/uaccess_32.h
@@ -6,7 +6,6 @@
  */
 #include <linux/errno.h>
 #include <linux/thread_info.h>
-#include <linux/prefetch.h>
 #include <linux/string.h>
 #include <asm/asm.h>
 #include <asm/page.h>
diff --git a/arch/x86/include/asm/uaccess_64.h b/arch/x86/include/asm/uaccess_64.h
index 316708d5af92d2c5ecd2f7e22e0bc1852b003d7d..1c66d30971adedaac940770d701e141ae464e1f4 100644
--- a/arch/x86/include/asm/uaccess_64.h
+++ b/arch/x86/include/asm/uaccess_64.h
@@ -6,7 +6,6 @@
  */
 #include <linux/compiler.h>
 #include <linux/errno.h>
-#include <linux/prefetch.h>
 #include <linux/lockdep.h>
 #include <asm/alternative.h>
 #include <asm/cpufeature.h>
diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c
index 20e3f8702d1e5701b130d2891e23bf39fbe81663..bcb394dfbb3587f8b4d40cc096b6ff7334ce6b12 100644
--- a/arch/x86/mm/fault.c
+++ b/arch/x86/mm/fault.c
@@ -12,6 +12,7 @@
 #include <linux/mmiotrace.h>		/* kmmio_handler, ...		*/
 #include <linux/perf_event.h>		/* perf_sw_event		*/
 #include <linux/hugetlb.h>		/* hstate_index_to_shift	*/
+#include <linux/prefetch.h>		/* prefetchw			*/
 
 #include <asm/traps.h>			/* dotraplinkage, ...		*/
 #include <asm/pgalloc.h>		/* pgd_*(), ...			*/
diff --git a/drivers/misc/sgi-gru/grufault.c b/drivers/misc/sgi-gru/grufault.c
index 38657cdaf54d32b247c89e105611f0b96a267c75..c4acac74725c4de50ef0954a258a851c05a03930 100644
--- a/drivers/misc/sgi-gru/grufault.c
+++ b/drivers/misc/sgi-gru/grufault.c
@@ -33,6 +33,7 @@
 #include <linux/io.h>
 #include <linux/uaccess.h>
 #include <linux/security.h>
+#include <linux/prefetch.h>
 #include <asm/pgtable.h>
 #include "gru.h"
 #include "grutables.h"
diff --git a/drivers/misc/sgi-gru/grumain.c b/drivers/misc/sgi-gru/grumain.c
index f8538bbd0bfae72c11128d773f00aadc91f0c1ba..ae16c8cb4f3e94b5629be31a0f7073cbb05bb18c 100644
--- a/drivers/misc/sgi-gru/grumain.c
+++ b/drivers/misc/sgi-gru/grumain.c
@@ -28,6 +28,7 @@
 #include <linux/device.h>
 #include <linux/list.h>
 #include <linux/err.h>
+#include <linux/prefetch.h>
 #include <asm/uv/uv_hub.h>
 #include "gru.h"
 #include "grutables.h"
diff --git a/drivers/net/acenic.c b/drivers/net/acenic.c
index ee648fe5d96fc48b65c59c1a2965aea7902cbf63..01560bb67a7afabe48058cab783c979ce24a9aa6 100644
--- a/drivers/net/acenic.c
+++ b/drivers/net/acenic.c
@@ -68,6 +68,7 @@
 #include <linux/sockios.h>
 #include <linux/firmware.h>
 #include <linux/slab.h>
+#include <linux/prefetch.h>
 
 #if defined(CONFIG_VLAN_8021Q) || defined(CONFIG_VLAN_8021Q_MODULE)
 #include <linux/if_vlan.h>
diff --git a/drivers/net/ehea/ehea_main.c b/drivers/net/ehea/ehea_main.c
index cf79cf759e13608146f897e957128b927bb7479e..2c60435f2beb4a30b92acb47b16716891fd522d2 100644
--- a/drivers/net/ehea/ehea_main.c
+++ b/drivers/net/ehea/ehea_main.c
@@ -41,6 +41,7 @@
 #include <linux/memory.h>
 #include <asm/kexec.h>
 #include <linux/mutex.h>
+#include <linux/prefetch.h>
 
 #include <net/ip.h>
 
diff --git a/drivers/staging/pohmelfs/inode.c b/drivers/staging/pohmelfs/inode.c
index c93ef207b0b442120400621ce36dd7e873ec966f..c0f0ac7c1cdb8f8cb6dd1aae58a8318b1ae4fb87 100644
--- a/drivers/staging/pohmelfs/inode.c
+++ b/drivers/staging/pohmelfs/inode.c
@@ -29,6 +29,7 @@
 #include <linux/slab.h>
 #include <linux/statfs.h>
 #include <linux/writeback.h>
+#include <linux/prefetch.h>
 
 #include "netfs.h"
 
diff --git a/drivers/usb/gadget/goku_udc.c b/drivers/usb/gadget/goku_udc.c
index 48a760220baf992c48541f2cae822943690bae5c..bf6e11c758d5ec84b99e944a97e69c27b59a698f 100644
--- a/drivers/usb/gadget/goku_udc.c
+++ b/drivers/usb/gadget/goku_udc.c
@@ -38,6 +38,7 @@
 #include <linux/device.h>
 #include <linux/usb/ch9.h>
 #include <linux/usb/gadget.h>
+#include <linux/prefetch.h>
 
 #include <asm/byteorder.h>
 #include <asm/io.h>
diff --git a/drivers/usb/gadget/imx_udc.c b/drivers/usb/gadget/imx_udc.c
index 5408186afc35c41dee81ff8afb7bcc55163a68a4..ade40066decf59463447b1c89a600536024f6f3f 100644
--- a/drivers/usb/gadget/imx_udc.c
+++ b/drivers/usb/gadget/imx_udc.c
@@ -30,6 +30,7 @@
 #include <linux/delay.h>
 #include <linux/timer.h>
 #include <linux/slab.h>
+#include <linux/prefetch.h>
 
 #include <linux/usb/ch9.h>
 #include <linux/usb/gadget.h>
diff --git a/drivers/usb/gadget/omap_udc.c b/drivers/usb/gadget/omap_udc.c
index cb5cd422f3f513a163d606fde97f0f2f646f48a8..82fd2493533228e56f60164f64855d43468c5c9e 100644
--- a/drivers/usb/gadget/omap_udc.c
+++ b/drivers/usb/gadget/omap_udc.c
@@ -44,6 +44,7 @@
 #include <linux/usb/otg.h>
 #include <linux/dma-mapping.h>
 #include <linux/clk.h>
+#include <linux/prefetch.h>
 
 #include <asm/byteorder.h>
 #include <asm/io.h>
diff --git a/drivers/usb/gadget/pxa25x_udc.c b/drivers/usb/gadget/pxa25x_udc.c
index 444b60aa15e951dc489da9444b793031dd0d77de..365c02fc25fcf0f89668b79ce2a72933ada22a40 100644
--- a/drivers/usb/gadget/pxa25x_udc.c
+++ b/drivers/usb/gadget/pxa25x_udc.c
@@ -46,6 +46,7 @@
 #include <linux/seq_file.h>
 #include <linux/debugfs.h>
 #include <linux/io.h>
+#include <linux/prefetch.h>
 
 #include <asm/byteorder.h>
 #include <asm/dma.h>
diff --git a/drivers/usb/gadget/pxa27x_udc.c b/drivers/usb/gadget/pxa27x_udc.c
index 78a39a41547d0fb19c99254222c870ea51f1c3cf..57607696735cfb934ea8f2f38921bf0a44e03299 100644
--- a/drivers/usb/gadget/pxa27x_udc.c
+++ b/drivers/usb/gadget/pxa27x_udc.c
@@ -32,6 +32,7 @@
 #include <linux/irq.h>
 #include <linux/gpio.h>
 #include <linux/slab.h>
+#include <linux/prefetch.h>
 
 #include <asm/byteorder.h>
 #include <mach/hardware.h>
diff --git a/drivers/usb/host/isp1362-hcd.c b/drivers/usb/host/isp1362-hcd.c
index f97570a847ca073ff16c3eee7ae9845f5515e04c..9c37dad3e8168989eaf919cf8993fafb499d6126 100644
--- a/drivers/usb/host/isp1362-hcd.c
+++ b/drivers/usb/host/isp1362-hcd.c
@@ -81,6 +81,7 @@
 #include <linux/pm.h>
 #include <linux/io.h>
 #include <linux/bitmap.h>
+#include <linux/prefetch.h>
 
 #include <asm/irq.h>
 #include <asm/system.h>
diff --git a/drivers/usb/host/sl811-hcd.c b/drivers/usb/host/sl811-hcd.c
index 18b7099a8125c9e0cbcfe9858259dfe85482826d..fafccc2fd33140386e83413022a9a8c661b9e4be 100644
--- a/drivers/usb/host/sl811-hcd.c
+++ b/drivers/usb/host/sl811-hcd.c
@@ -47,6 +47,7 @@
 #include <linux/usb/sl811.h>
 #include <linux/usb/hcd.h>
 #include <linux/platform_device.h>
+#include <linux/prefetch.h>
 
 #include <asm/io.h>
 #include <asm/irq.h>
diff --git a/drivers/video/udlfb.c b/drivers/video/udlfb.c
index 68041d9dc2602e263474e7892c24b20c30e151b0..695066b5b2e6d628204695727c3afde94a0f6de7 100644
--- a/drivers/video/udlfb.c
+++ b/drivers/video/udlfb.c
@@ -27,6 +27,7 @@
 #include <linux/fb.h>
 #include <linux/vmalloc.h>
 #include <linux/slab.h>
+#include <linux/prefetch.h>
 #include <linux/delay.h>
 #include <video/udlfb.h>
 #include "edid.h"
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index ba41da59e31b1d93348622d7b47655234f5ff930..96fcfa522dab72f837d991d5afaaacd4e73d8f74 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -10,6 +10,7 @@
 #include <linux/swap.h>
 #include <linux/writeback.h>
 #include <linux/pagevec.h>
+#include <linux/prefetch.h>
 #include "extent_io.h"
 #include "extent_map.h"
 #include "compat.h"
diff --git a/fs/dcache.c b/fs/dcache.c
index 22a0ef41bad1b1cd17fc6ce84023d52f65096ba5..18b2a1f10ed89cab1562f696adebf391bf2a8763 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -35,6 +35,7 @@
 #include <linux/hardirq.h>
 #include <linux/bit_spinlock.h>
 #include <linux/rculist_bl.h>
+#include <linux/prefetch.h>
 #include "internal.h"
 
 /*
diff --git a/fs/logfs/dev_bdev.c b/fs/logfs/dev_bdev.c
index 1adc8d455f0ea2d66237436184766c5ed120688f..df0de27c273349c22ee6b8770b589663801c68f5 100644
--- a/fs/logfs/dev_bdev.c
+++ b/fs/logfs/dev_bdev.c
@@ -10,6 +10,7 @@
 #include <linux/blkdev.h>
 #include <linux/buffer_head.h>
 #include <linux/gfp.h>
+#include <linux/prefetch.h>
 
 #define PAGE_OFS(ofs) ((ofs) & (PAGE_SIZE-1))
 
diff --git a/include/asm-generic/xor.h b/include/asm-generic/xor.h
index aaab875e1a351f5596a22af645f567c2d2b5ae66..6028fb862254b8dfaeec0dd93de3a8fdf052ca5c 100644
--- a/include/asm-generic/xor.h
+++ b/include/asm-generic/xor.h
@@ -13,7 +13,7 @@
  * Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
-#include <asm/processor.h>
+#include <linux/prefetch.h>
 
 static void
 xor_8regs_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
diff --git a/kernel/rcutiny.c b/kernel/rcutiny.c
index 421abfd3641d2f011fe7b33b85a36a6b7cf8faf1..7bbac7d0f5abc1d4737c0664c6779d9db1ca5b85 100644
--- a/kernel/rcutiny.c
+++ b/kernel/rcutiny.c
@@ -35,6 +35,7 @@
 #include <linux/init.h>
 #include <linux/time.h>
 #include <linux/cpu.h>
+#include <linux/prefetch.h>
 
 /* Controls for rcu_kthread() kthread, replacing RCU_SOFTIRQ used previously. */
 static struct task_struct *rcu_kthread_task;
diff --git a/kernel/rcutree.c b/kernel/rcutree.c
index e486f7c3ffb8b239a31eddb37f9ed0251629bdeb..f07d2f03181a9b6c296a09e5e2869ea7c5844580 100644
--- a/kernel/rcutree.c
+++ b/kernel/rcutree.c
@@ -49,6 +49,7 @@
 #include <linux/kernel_stat.h>
 #include <linux/wait.h>
 #include <linux/kthread.h>
+#include <linux/prefetch.h>
 
 #include "rcutree.h"
 
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 3f8bce264df66f712e9a44092f871d9f90eafe22..d49df7840541f5e4796f651afa0d6122ef08edbc 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -54,6 +54,7 @@
 #include <trace/events/kmem.h>
 #include <linux/ftrace_event.h>
 #include <linux/memcontrol.h>
+#include <linux/prefetch.h>
 
 #include <asm/tlbflush.h>
 #include <asm/div64.h>
diff --git a/mm/prio_tree.c b/mm/prio_tree.c
index 603ae98d9694dfbf88561abb266021b1bf370fb1..799dcfd7cd8c5499b78bdc1b0b6f976ce79465bc 100644
--- a/mm/prio_tree.c
+++ b/mm/prio_tree.c
@@ -13,6 +13,7 @@
 
 #include <linux/mm.h>
 #include <linux/prio_tree.h>
+#include <linux/prefetch.h>
 
 /*
  * See lib/prio_tree.c for details on the general radix priority search tree
diff --git a/mm/slab.c b/mm/slab.c
index 46a9c163a92f222aeab9adc3a7ed56511e89a23b..bcfa4987c8ae2c9fc273f1baf75d4461ed743052 100644
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -115,6 +115,7 @@
 #include	<linux/debugobjects.h>
 #include	<linux/kmemcheck.h>
 #include	<linux/memory.h>
+#include	<linux/prefetch.h>
 
 #include	<asm/cacheflush.h>
 #include	<asm/tlbflush.h>
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 8bfd45050a61494f919e90dfb1538ea8310f4f20..c9177202c8ce2ce2f9bfe9045f2a40a8105de8e8 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -42,6 +42,7 @@
 #include <linux/delayacct.h>
 #include <linux/sysctl.h>
 #include <linux/oom.h>
+#include <linux/prefetch.h>
 
 #include <asm/tlbflush.h>
 #include <asm/div64.h>
diff --git a/net/core/dst.c b/net/core/dst.c
index 91104d35de7d9d6f7d276e8ae02b833e5ddc4338..0a3920bf361372acd9934ca9ddbff57d4d3e1216 100644
--- a/net/core/dst.c
+++ b/net/core/dst.c
@@ -19,6 +19,7 @@
 #include <linux/types.h>
 #include <net/net_namespace.h>
 #include <linux/sched.h>
+#include <linux/prefetch.h>
 
 #include <net/dst.h>
 
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index aeeece72b72fa25d29dfd3162a2f9c3a8ef6fcbd..6ed9e27d8202c9d4bf2ec43537a83a444c7d8822 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -156,6 +156,7 @@
 #include <linux/wait.h>
 #include <linux/etherdevice.h>
 #include <linux/kthread.h>
+#include <linux/prefetch.h>
 #include <net/net_namespace.h>
 #include <net/checksum.h>
 #include <net/ipv6.h>
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 7ebeed0a877cf4bb244d488346e745d45da93959..960ea899c864ecb2d8457fa5cd6e2a486fe80cf4 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -57,6 +57,7 @@
 #include <linux/init.h>
 #include <linux/scatterlist.h>
 #include <linux/errqueue.h>
+#include <linux/prefetch.h>
 
 #include <net/protocol.h>
 #include <net/dst.h>
diff --git a/tools/perf/util/include/linux/list.h b/tools/perf/util/include/linux/list.h
index 356c7e467b83fd59af5cd191a577197ee88349b8..99358d61e9a5828fcc2e276a9c37fc8cc3610321 100644
--- a/tools/perf/util/include/linux/list.h
+++ b/tools/perf/util/include/linux/list.h
@@ -23,5 +23,5 @@ static inline void list_del_range(struct list_head *begin,
  * @head: the head for your list.
  */
 #define list_for_each_from(pos, head) \
-	for (; prefetch(pos->next), pos != (head); pos = pos->next)
+	for (; pos != (head); pos = pos->next)
 #endif