2014-07-01 16:19:42

by Russell King

Subject: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

ARMv6 and greater introduced a new instruction ("bx") which can be used
to return from function calls. Recent CPUs perform better when the
"bx lr" instruction is used rather than the "mov pc, lr" instruction,
and this sequence is strongly recommended to be used by the ARM
architecture manual (section A.4.1.1).

We provide a new macro "ret" with all its variants for the condition
code which will resolve to the appropriate instruction.

Rather than doing this piecemeal, and miss some instances, change all
the "mov pc" instances to use the new macro, with the exception of
the "movs" instruction and the kprobes code. This allows us to detect
the "mov pc, lr" case and fix it up - and also gives us the possibility
of deploying this for other registers depending on the CPU selection.

Signed-off-by: Russell King <[email protected]>
---
(Resent without the long Cc list)
(Resent with the patch cut in half - sorry, you're going to have to
join the two halves together to test it.)

arch/arm/crypto/aes-armv4.S | 3 ++-
arch/arm/include/asm/assembler.h | 21 +++++++++++++++
arch/arm/include/asm/entry-macro-multi.S | 2 +-
arch/arm/kernel/debug.S | 10 ++++----
arch/arm/kernel/entry-armv.S | 42 +++++++++++++++---------------
arch/arm/kernel/entry-common.S | 13 +++++-----
arch/arm/kernel/entry-header.S | 14 ----------
arch/arm/kernel/fiqasm.S | 4 +--
arch/arm/kernel/head-common.S | 7 ++---
arch/arm/kernel/head-nommu.S | 8 +++---
arch/arm/kernel/head.S | 18 ++++++-------
arch/arm/kernel/hyp-stub.S | 6 ++---
arch/arm/kernel/iwmmxt.S | 10 ++++----
arch/arm/kernel/relocate_kernel.S | 3 ++-
arch/arm/kernel/sleep.S | 2 +-
arch/arm/kvm/init.S | 3 ++-
arch/arm/lib/ashldi3.S | 3 ++-
arch/arm/lib/ashrdi3.S | 3 ++-
arch/arm/lib/backtrace.S | 2 +-
arch/arm/lib/bitops.h | 5 ++--
arch/arm/lib/bswapsdi2.S | 5 ++--
arch/arm/lib/call_with_stack.S | 4 +--
arch/arm/lib/csumpartial.S | 2 +-
arch/arm/lib/csumpartialcopygeneric.S | 5 ++--
arch/arm/lib/delay-loop.S | 18 ++++++-------
arch/arm/lib/div64.S | 13 +++++-----
arch/arm/lib/findbit.S | 10 ++++----
arch/arm/lib/getuser.S | 8 +++---
arch/arm/lib/io-readsb.S | 2 +-
arch/arm/lib/io-readsl.S | 6 ++---
arch/arm/lib/io-readsw-armv3.S | 4 +--
arch/arm/lib/io-readsw-armv4.S | 2 +-
arch/arm/lib/io-writesb.S | 2 +-
arch/arm/lib/io-writesl.S | 10 ++++----
arch/arm/lib/io-writesw-armv3.S | 4 +--
arch/arm/lib/io-writesw-armv4.S | 4 +--
arch/arm/lib/lib1funcs.S | 26 +++++++++----------
arch/arm/lib/lshrdi3.S | 3 ++-
arch/arm/lib/memchr.S | 2 +-
arch/arm/lib/memset.S | 2 +-
arch/arm/lib/memzero.S | 2 +-
arch/arm/lib/muldi3.S | 3 ++-
arch/arm/lib/putuser.S | 10 ++++----
arch/arm/lib/strchr.S | 2 +-
arch/arm/lib/strrchr.S | 2 +-
arch/arm/lib/ucmpdi2.S | 5 ++--
arch/arm/mach-davinci/sleep.S | 2 +-
arch/arm/mach-ep93xx/crunch-bits.S | 6 ++---
arch/arm/mach-imx/suspend-imx6.S | 5 ++--
arch/arm/mach-mvebu/coherency_ll.S | 10 ++++----
arch/arm/mach-mvebu/headsmp-a9.S | 3 ++-
arch/arm/mach-omap2/sleep44xx.S | 3 ++-
arch/arm/mach-omap2/sram242x.S | 6 ++---
arch/arm/mach-omap2/sram243x.S | 6 ++---
arch/arm/mach-pxa/mioa701_bootresume.S | 2 +-
arch/arm/mach-pxa/standby.S | 4 +--
arch/arm/mach-s3c24xx/sleep-s3c2410.S | 2 +-
arch/arm/mach-s3c24xx/sleep-s3c2412.S | 2 +-
arch/arm/mach-shmobile/headsmp.S | 3 ++-
arch/arm/mach-tegra/sleep-tegra20.S | 24 ++++++++---------
arch/arm/mach-tegra/sleep-tegra30.S | 14 +++++-----
arch/arm/mach-tegra/sleep.S | 8 +++---
arch/arm/mm/cache-fa.S | 19 +++++++-------
arch/arm/mm/cache-nop.S | 5 ++--
arch/arm/mm/cache-v4.S | 13 +++++-----
arch/arm/mm/cache-v4wb.S | 15 ++++++-----
arch/arm/mm/cache-v4wt.S | 13 +++++-----
arch/arm/mm/cache-v6.S | 20 +++++++--------
arch/arm/mm/cache-v7.S | 30 +++++++++++-----------
arch/arm/mm/l2c-l2x0-resume.S | 7 ++---
arch/arm/mm/proc-arm1020.S | 34 ++++++++++++------------
arch/arm/mm/proc-arm1020e.S | 34 ++++++++++++------------
arch/arm/mm/proc-arm1022.S | 34 ++++++++++++------------
arch/arm/mm/proc-arm1026.S | 34 ++++++++++++------------
arch/arm/mm/proc-arm720.S | 16 ++++++------
arch/arm/mm/proc-arm740.S | 8 +++---
arch/arm/mm/proc-arm7tdmi.S | 8 +++---
arch/arm/mm/proc-arm920.S | 34 ++++++++++++------------
arch/arm/mm/proc-arm922.S | 34 ++++++++++++------------
arch/arm/mm/proc-arm925.S | 34 ++++++++++++------------
arch/arm/mm/proc-arm926.S | 34 ++++++++++++------------
arch/arm/mm/proc-arm940.S | 24 ++++++++---------
arch/arm/mm/proc-arm946.S | 30 +++++++++++-----------
arch/arm/mm/proc-arm9tdmi.S | 8 +++---
arch/arm/mm/proc-fa526.S | 16 ++++++------
arch/arm/mm/proc-feroceon.S | 44 ++++++++++++++++----------------
arch/arm/mm/proc-mohawk.S | 34 ++++++++++++------------
arch/arm/mm/proc-sa110.S | 16 ++++++------
arch/arm/mm/proc-sa1100.S | 16 ++++++------
arch/arm/mm/proc-v6.S | 16 ++++++------
arch/arm/mm/proc-v7-2level.S | 4 +--
arch/arm/mm/proc-v7-3level.S | 5 ++--
arch/arm/mm/proc-v7.S | 14 +++++-----
arch/arm/mm/proc-v7m.S | 18 ++++++-------
arch/arm/mm/proc-xsc3.S | 32 +++++++++++------------
arch/arm/mm/proc-xscale.S | 34 ++++++++++++------------
arch/arm/mm/tlb-fa.S | 7 ++---
arch/arm/mm/tlb-v4.S | 5 ++--
arch/arm/mm/tlb-v4wb.S | 7 ++---
arch/arm/mm/tlb-v4wbi.S | 7 ++---
arch/arm/mm/tlb-v6.S | 5 ++--
arch/arm/mm/tlb-v7.S | 4 +--
arch/arm/nwfpe/entry.S | 8 +++---
arch/arm/vfp/entry.S | 4 +--
arch/arm/vfp/vfphw.S | 26 +++++++++----------
arch/arm/xen/hypercall.S | 6 ++---
106 files changed, 644 insertions(+), 607 deletions(-)

diff --git a/arch/arm/crypto/aes-armv4.S b/arch/arm/crypto/aes-armv4.S
index 3a14ea8fe97e..ebb9761fb572 100644
--- a/arch/arm/crypto/aes-armv4.S
+++ b/arch/arm/crypto/aes-armv4.S
@@ -35,6 +35,7 @@
@ that is being targetted.

#include <linux/linkage.h>
+#include <asm/assembler.h>

.text

@@ -648,7 +649,7 @@ _armv4_AES_set_encrypt_key:

.Ldone: mov r0,#0
ldmia sp!,{r4-r12,lr}
-.Labrt: mov pc,lr
+.Labrt: ret lr
ENDPROC(private_AES_set_encrypt_key)

.align 5
diff --git a/arch/arm/include/asm/assembler.h b/arch/arm/include/asm/assembler.h
index 906703a5b564..f67fd3afebdf 100644
--- a/arch/arm/include/asm/assembler.h
+++ b/arch/arm/include/asm/assembler.h
@@ -427,4 +427,25 @@ THUMB( orr \reg , \reg , #PSR_T_BIT )
#endif
.endm

+ .irp c,,eq,ne,cs,cc,mi,pl,vs,vc,hi,ls,ge,lt,gt,le,hs,lo
+ .macro ret\c, reg
+#if __LINUX_ARM_ARCH__ < 6
+ mov\c pc, \reg
+#else
+ .ifeqs "\reg", "lr"
+ bx\c \reg
+ .else
+ mov\c pc, \reg
+ .endif
+#endif
+ .endm
+ .endr
+
+ .macro ret.w, reg
+ ret \reg
+#ifdef CONFIG_THUMB2_KERNEL
+ nop
+#endif
+ .endm
+
#endif /* __ASM_ASSEMBLER_H__ */
diff --git a/arch/arm/include/asm/entry-macro-multi.S b/arch/arm/include/asm/entry-macro-multi.S
index 88d61815f0c0..469a2b30fa27 100644
--- a/arch/arm/include/asm/entry-macro-multi.S
+++ b/arch/arm/include/asm/entry-macro-multi.S
@@ -35,5 +35,5 @@
\symbol_name:
mov r8, lr
arch_irq_handler_default
- mov pc, r8
+ ret r8
.endm
diff --git a/arch/arm/kernel/debug.S b/arch/arm/kernel/debug.S
index 14f7c3b14632..78c91b5f97d4 100644
--- a/arch/arm/kernel/debug.S
+++ b/arch/arm/kernel/debug.S
@@ -90,7 +90,7 @@ ENTRY(printascii)
ldrneb r1, [r0], #1
teqne r1, #0
bne 1b
- mov pc, lr
+ ret lr
ENDPROC(printascii)

ENTRY(printch)
@@ -105,7 +105,7 @@ ENTRY(debug_ll_addr)
addruart r2, r3, ip
str r2, [r0]
str r3, [r1]
- mov pc, lr
+ ret lr
ENDPROC(debug_ll_addr)
#endif

@@ -116,7 +116,7 @@ ENTRY(printascii)
mov r0, #0x04 @ SYS_WRITE0
ARM( svc #0x123456 )
THUMB( svc #0xab )
- mov pc, lr
+ ret lr
ENDPROC(printascii)

ENTRY(printch)
@@ -125,14 +125,14 @@ ENTRY(printch)
mov r0, #0x03 @ SYS_WRITEC
ARM( svc #0x123456 )
THUMB( svc #0xab )
- mov pc, lr
+ ret lr
ENDPROC(printch)

ENTRY(debug_ll_addr)
mov r2, #0
str r2, [r0]
str r2, [r1]
- mov pc, lr
+ ret lr
ENDPROC(debug_ll_addr)

#endif
diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S
index 52a949a8077d..36276cdccfbc 100644
--- a/arch/arm/kernel/entry-armv.S
+++ b/arch/arm/kernel/entry-armv.S
@@ -224,7 +224,7 @@ svc_preempt:
1: bl preempt_schedule_irq @ irq en/disable is done inside
ldr r0, [tsk, #TI_FLAGS] @ get new tasks TI_FLAGS
tst r0, #_TIF_NEED_RESCHED
- moveq pc, r8 @ go again
+ reteq r8 @ go again
b 1b
#endif

@@ -490,7 +490,7 @@ ENDPROC(__und_usr)
.pushsection .fixup, "ax"
.align 2
4: str r4, [sp, #S_PC] @ retry current instruction
- mov pc, r9
+ ret r9
.popsection
.pushsection __ex_table,"a"
.long 1b, 4b
@@ -552,7 +552,7 @@ call_fpe:
#endif
tst r0, #0x08000000 @ only CDP/CPRT/LDC/STC have bit 27
tstne r0, #0x04000000 @ bit 26 set on both ARM and Thumb-2
- moveq pc, lr
+ reteq lr
and r8, r0, #0x00000f00 @ mask out CP number
THUMB( lsr r8, r8, #8 )
mov r7, #1
@@ -571,33 +571,33 @@ call_fpe:
THUMB( add pc, r8 )
nop

- movw_pc lr @ CP#0
+ ret.w lr @ CP#0
W(b) do_fpe @ CP#1 (FPE)
W(b) do_fpe @ CP#2 (FPE)
- movw_pc lr @ CP#3
+ ret.w lr @ CP#3
#ifdef CONFIG_CRUNCH
b crunch_task_enable @ CP#4 (MaverickCrunch)
b crunch_task_enable @ CP#5 (MaverickCrunch)
b crunch_task_enable @ CP#6 (MaverickCrunch)
#else
- movw_pc lr @ CP#4
- movw_pc lr @ CP#5
- movw_pc lr @ CP#6
+ ret.w lr @ CP#4
+ ret.w lr @ CP#5
+ ret.w lr @ CP#6
#endif
- movw_pc lr @ CP#7
- movw_pc lr @ CP#8
- movw_pc lr @ CP#9
+ ret.w lr @ CP#7
+ ret.w lr @ CP#8
+ ret.w lr @ CP#9
#ifdef CONFIG_VFP
W(b) do_vfp @ CP#10 (VFP)
W(b) do_vfp @ CP#11 (VFP)
#else
- movw_pc lr @ CP#10 (VFP)
- movw_pc lr @ CP#11 (VFP)
+ ret.w lr @ CP#10 (VFP)
+ ret.w lr @ CP#11 (VFP)
#endif
- movw_pc lr @ CP#12
- movw_pc lr @ CP#13
- movw_pc lr @ CP#14 (Debug)
- movw_pc lr @ CP#15 (Control)
+ ret.w lr @ CP#12
+ ret.w lr @ CP#13
+ ret.w lr @ CP#14 (Debug)
+ ret.w lr @ CP#15 (Control)

#ifdef NEED_CPU_ARCHITECTURE
.align 2
@@ -649,7 +649,7 @@ ENTRY(fp_enter)
.popsection

ENTRY(no_fp)
- mov pc, lr
+ ret lr
ENDPROC(no_fp)

__und_usr_fault_32:
@@ -745,7 +745,7 @@ ENDPROC(__switch_to)
#ifdef CONFIG_ARM_THUMB
bx \reg
#else
- mov pc, \reg
+ ret \reg
#endif
.endm

@@ -837,7 +837,7 @@ kuser_cmpxchg64_fixup:
#if __LINUX_ARM_ARCH__ < 6
bcc kuser_cmpxchg32_fixup
#endif
- mov pc, lr
+ ret lr
.previous

#else
@@ -905,7 +905,7 @@ kuser_cmpxchg32_fixup:
subs r8, r4, r7
rsbcss r8, r8, #(2b - 1b)
strcs r7, [sp, #S_PC]
- mov pc, lr
+ ret lr
.previous

#else
diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S
index 7139d4a7dea7..e52fe5a2d843 100644
--- a/arch/arm/kernel/entry-common.S
+++ b/arch/arm/kernel/entry-common.S
@@ -8,6 +8,7 @@
* published by the Free Software Foundation.
*/

+#include <asm/assembler.h>
#include <asm/unistd.h>
#include <asm/ftrace.h>
#include <asm/unwind.h>
@@ -88,7 +89,7 @@ ENTRY(ret_from_fork)
cmp r5, #0
movne r0, r4
adrne lr, BSYM(1f)
- movne pc, r5
+ retne r5
1: get_thread_info tsk
b ret_slow_syscall
ENDPROC(ret_from_fork)
@@ -290,7 +291,7 @@ ENDPROC(ftrace_graph_caller_old)

.macro mcount_exit
ldmia sp!, {r0-r3, ip, lr}
- mov pc, ip
+ ret ip
.endm

ENTRY(__gnu_mcount_nc)
@@ -298,7 +299,7 @@ UNWIND(.fnstart)
#ifdef CONFIG_DYNAMIC_FTRACE
mov ip, lr
ldmia sp!, {lr}
- mov pc, ip
+ ret ip
#else
__mcount
#endif
@@ -333,12 +334,12 @@ return_to_handler:
bl ftrace_return_to_handler
mov lr, r0 @ r0 has real ret addr
ldmia sp!, {r0-r3}
- mov pc, lr
+ ret lr
#endif

ENTRY(ftrace_stub)
.Lftrace_stub:
- mov pc, lr
+ ret lr
ENDPROC(ftrace_stub)

#endif /* CONFIG_FUNCTION_TRACER */
@@ -561,7 +562,7 @@ sys_mmap2:
streq r5, [sp, #4]
beq sys_mmap_pgoff
mov r0, #-EINVAL
- mov pc, lr
+ ret lr
#else
str r5, [sp, #4]
b sys_mmap_pgoff
diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S
index 5d702f8900b1..8db307d0954b 100644
--- a/arch/arm/kernel/entry-header.S
+++ b/arch/arm/kernel/entry-header.S
@@ -240,12 +240,6 @@
movs pc, lr @ return & move spsr_svc into cpsr
.endm

- @
- @ 32-bit wide "mov pc, reg"
- @
- .macro movw_pc, reg
- mov pc, \reg
- .endm
#else /* CONFIG_THUMB2_KERNEL */
.macro svc_exit, rpsr, irq = 0
.if \irq != 0
@@ -304,14 +298,6 @@
movs pc, lr @ return & move spsr_svc into cpsr
.endm
#endif /* ifdef CONFIG_CPU_V7M / else */
-
- @
- @ 32-bit wide "mov pc, reg"
- @
- .macro movw_pc, reg
- mov pc, \reg
- nop
- .endm
#endif /* !CONFIG_THUMB2_KERNEL */

/*
diff --git a/arch/arm/kernel/fiqasm.S b/arch/arm/kernel/fiqasm.S
index 207f9d652010..8dd26e1a9bd6 100644
--- a/arch/arm/kernel/fiqasm.S
+++ b/arch/arm/kernel/fiqasm.S
@@ -32,7 +32,7 @@ ENTRY(__set_fiq_regs)
ldr lr, [r0]
msr cpsr_c, r1 @ return to SVC mode
mov r0, r0 @ avoid hazard prior to ARMv4
- mov pc, lr
+ ret lr
ENDPROC(__set_fiq_regs)

ENTRY(__get_fiq_regs)
@@ -45,5 +45,5 @@ ENTRY(__get_fiq_regs)
str lr, [r0]
msr cpsr_c, r1 @ return to SVC mode
mov r0, r0 @ avoid hazard prior to ARMv4
- mov pc, lr
+ ret lr
ENDPROC(__get_fiq_regs)
diff --git a/arch/arm/kernel/head-common.S b/arch/arm/kernel/head-common.S
index 572a38335c96..8733012d231f 100644
--- a/arch/arm/kernel/head-common.S
+++ b/arch/arm/kernel/head-common.S
@@ -10,6 +10,7 @@
* published by the Free Software Foundation.
*
*/
+#include <asm/assembler.h>

#define ATAG_CORE 0x54410001
#define ATAG_CORE_SIZE ((2*4 + 3*4) >> 2)
@@ -61,10 +62,10 @@ __vet_atags:
cmp r5, r6
bne 1f

-2: mov pc, lr @ atag/dtb pointer is ok
+2: ret lr @ atag/dtb pointer is ok

1: mov r2, #0
- mov pc, lr
+ ret lr
ENDPROC(__vet_atags)

/*
@@ -162,7 +163,7 @@ __lookup_processor_type:
cmp r5, r6
blo 1b
mov r5, #0 @ unknown processor
-2: mov pc, lr
+2: ret lr
ENDPROC(__lookup_processor_type)

/*
diff --git a/arch/arm/kernel/head-nommu.S b/arch/arm/kernel/head-nommu.S
index 716249cc2ee1..cc176b67c134 100644
--- a/arch/arm/kernel/head-nommu.S
+++ b/arch/arm/kernel/head-nommu.S
@@ -82,7 +82,7 @@ ENTRY(stext)
adr lr, BSYM(1f) @ return (PIC) address
ARM( add pc, r10, #PROCINFO_INITFUNC )
THUMB( add r12, r10, #PROCINFO_INITFUNC )
- THUMB( mov pc, r12 )
+ THUMB( ret r12 )
1: b __after_proc_init
ENDPROC(stext)

@@ -119,7 +119,7 @@ ENTRY(secondary_startup)
mov r13, r12 @ __secondary_switched address
ARM( add pc, r10, #PROCINFO_INITFUNC )
THUMB( add r12, r10, #PROCINFO_INITFUNC )
- THUMB( mov pc, r12 )
+ THUMB( ret r12 )
ENDPROC(secondary_startup)

ENTRY(__secondary_switched)
@@ -164,7 +164,7 @@ __after_proc_init:
#endif
mcr p15, 0, r0, c1, c0, 0 @ write control reg
#endif /* CONFIG_CPU_CP15 */
- mov pc, r13
+ ret r13
ENDPROC(__after_proc_init)
.ltorg

@@ -254,7 +254,7 @@ ENTRY(__setup_mpu)
orr r0, r0, #CR_M @ Set SCTRL.M (MPU on)
mcr p15, 0, r0, c1, c0, 0 @ Enable MPU
isb
- mov pc,lr
+ ret lr
ENDPROC(__setup_mpu)
#endif
#include "head-common.S"
diff --git a/arch/arm/kernel/head.S b/arch/arm/kernel/head.S
index 2c35f0ff2fdc..664eee8c4a26 100644
--- a/arch/arm/kernel/head.S
+++ b/arch/arm/kernel/head.S
@@ -140,7 +140,7 @@ ENTRY(stext)
mov r8, r4 @ set TTBR1 to swapper_pg_dir
ARM( add pc, r10, #PROCINFO_INITFUNC )
THUMB( add r12, r10, #PROCINFO_INITFUNC )
- THUMB( mov pc, r12 )
+ THUMB( ret r12 )
1: b __enable_mmu
ENDPROC(stext)
.ltorg
@@ -335,7 +335,7 @@ __create_page_tables:
sub r4, r4, #0x1000 @ point to the PGD table
mov r4, r4, lsr #ARCH_PGD_SHIFT
#endif
- mov pc, lr
+ ret lr
ENDPROC(__create_page_tables)
.ltorg
.align
@@ -383,7 +383,7 @@ ENTRY(secondary_startup)
ARM( add pc, r10, #PROCINFO_INITFUNC ) @ initialise processor
@ (return control reg)
THUMB( add r12, r10, #PROCINFO_INITFUNC )
- THUMB( mov pc, r12 )
+ THUMB( ret r12 )
ENDPROC(secondary_startup)

/*
@@ -468,7 +468,7 @@ ENTRY(__turn_mmu_on)
instr_sync
mov r3, r3
mov r3, r13
- mov pc, r3
+ ret r3
__turn_mmu_on_end:
ENDPROC(__turn_mmu_on)
.popsection
@@ -487,7 +487,7 @@ __fixup_smp:
orr r4, r4, #0x0000b000
orr r4, r4, #0x00000020 @ val 0x4100b020
teq r3, r4 @ ARM 11MPCore?
- moveq pc, lr @ yes, assume SMP
+ reteq lr @ yes, assume SMP

mrc p15, 0, r0, c0, c0, 5 @ read MPIDR
and r0, r0, #0xc0000000 @ multiprocessing extensions and
@@ -500,7 +500,7 @@ __fixup_smp:
orr r4, r4, #0x0000c000
orr r4, r4, #0x00000090
teq r3, r4 @ Check for ARM Cortex-A9
- movne pc, lr @ Not ARM Cortex-A9,
+ retne lr @ Not ARM Cortex-A9,

@ If a future SoC *does* use 0x0 as the PERIPH_BASE, then the
@ below address check will need to be #ifdef'd or equivalent
@@ -512,7 +512,7 @@ __fixup_smp:
ARM_BE8(rev r0, r0) @ byteswap if big endian
and r0, r0, #0x3 @ number of CPUs
teq r0, #0x0 @ is 1?
- movne pc, lr
+ retne lr

__fixup_smp_on_up:
adr r0, 1f
@@ -539,7 +539,7 @@ smp_on_up:
.text
__do_fixup_smp_on_up:
cmp r4, r5
- movhs pc, lr
+ reths lr
ldmia r4!, {r0, r6}
ARM( str r6, [r0, r3] )
THUMB( add r0, r0, r3 )
@@ -672,7 +672,7 @@ ARM_BE8(rev16 ip, ip)
2: cmp r4, r5
ldrcc r7, [r4], #4 @ use branch for delay slot
bcc 1b
- mov pc, lr
+ ret lr
#endif
ENDPROC(__fixup_a_pv_table)

diff --git a/arch/arm/kernel/hyp-stub.S b/arch/arm/kernel/hyp-stub.S
index 797b1a6a4906..56ce6290c831 100644
--- a/arch/arm/kernel/hyp-stub.S
+++ b/arch/arm/kernel/hyp-stub.S
@@ -99,7 +99,7 @@ ENTRY(__hyp_stub_install_secondary)
* immediately.
*/
compare_cpu_mode_with_primary r4, r5, r6, r7
- movne pc, lr
+ retne lr

/*
* Once we have given up on one CPU, we do not try to install the
@@ -111,7 +111,7 @@ ENTRY(__hyp_stub_install_secondary)
*/

cmp r4, #HYP_MODE
- movne pc, lr @ give up if the CPU is not in HYP mode
+ retne lr @ give up if the CPU is not in HYP mode

/*
* Configure HSCTLR to set correct exception endianness/instruction set
@@ -201,7 +201,7 @@ ENDPROC(__hyp_get_vectors)
@ fall through
ENTRY(__hyp_set_vectors)
__HVC(0)
- mov pc, lr
+ ret lr
ENDPROC(__hyp_set_vectors)

#ifndef ZIMAGE
diff --git a/arch/arm/kernel/iwmmxt.S b/arch/arm/kernel/iwmmxt.S
index a5599cfc43cb..0960be7953f0 100644
--- a/arch/arm/kernel/iwmmxt.S
+++ b/arch/arm/kernel/iwmmxt.S
@@ -179,7 +179,7 @@ concan_load:
get_thread_info r10
#endif
4: dec_preempt_count r10, r3
- mov pc, lr
+ ret lr

/*
* Back up Concan regs to save area and disable access to them
@@ -265,7 +265,7 @@ ENTRY(iwmmxt_task_copy)
mov r3, lr @ preserve return address
bl concan_dump
msr cpsr_c, ip @ restore interrupt mode
- mov pc, r3
+ ret r3

/*
* Restore Concan state from given memory address
@@ -301,7 +301,7 @@ ENTRY(iwmmxt_task_restore)
mov r3, lr @ preserve return address
bl concan_load
msr cpsr_c, ip @ restore interrupt mode
- mov pc, r3
+ ret r3

/*
* Concan handling on task switch
@@ -323,7 +323,7 @@ ENTRY(iwmmxt_task_switch)
add r3, r0, #TI_IWMMXT_STATE @ get next task Concan save area
ldr r2, [r2] @ get current Concan owner
teq r2, r3 @ next task owns it?
- movne pc, lr @ no: leave Concan disabled
+ retne lr @ no: leave Concan disabled

1: @ flip Concan access
XSC(eor r1, r1, #0x3)
@@ -350,7 +350,7 @@ ENTRY(iwmmxt_task_release)
eors r0, r0, r1 @ if equal...
streq r0, [r3] @ then clear ownership
msr cpsr_c, r2 @ restore interrupts
- mov pc, lr
+ ret lr

.data
concan_owner:
diff --git a/arch/arm/kernel/relocate_kernel.S b/arch/arm/kernel/relocate_kernel.S
index 95858966d84e..35e72585ec1d 100644
--- a/arch/arm/kernel/relocate_kernel.S
+++ b/arch/arm/kernel/relocate_kernel.S
@@ -3,6 +3,7 @@
*/

#include <linux/linkage.h>
+#include <asm/assembler.h>
#include <asm/kexec.h>

.align 3 /* not needed for this code, but keeps fncpy() happy */
@@ -59,7 +60,7 @@ ENTRY(relocate_new_kernel)
mov r0,#0
ldr r1,kexec_mach_type
ldr r2,kexec_boot_atags
- ARM( mov pc, lr )
+ ARM( ret lr )
THUMB( bx lr )

.align
diff --git a/arch/arm/kernel/sleep.S b/arch/arm/kernel/sleep.S
index 1b880db2a033..e1e60e5a7a27 100644
--- a/arch/arm/kernel/sleep.S
+++ b/arch/arm/kernel/sleep.S
@@ -107,7 +107,7 @@ ENTRY(cpu_resume_mmu)
instr_sync
mov r0, r0
mov r0, r0
- mov pc, r3 @ jump to virtual address
+ ret r3 @ jump to virtual address
ENDPROC(cpu_resume_mmu)
.popsection
cpu_resume_after_mmu:
diff --git a/arch/arm/kvm/init.S b/arch/arm/kvm/init.S
index 1b9844d369cc..b2d229f09c07 100644
--- a/arch/arm/kvm/init.S
+++ b/arch/arm/kvm/init.S
@@ -17,6 +17,7 @@
*/

#include <linux/linkage.h>
+#include <asm/assembler.h>
#include <asm/unified.h>
#include <asm/asm-offsets.h>
#include <asm/kvm_asm.h>
@@ -134,7 +135,7 @@ phase2:
ldr r0, =TRAMPOLINE_VA
adr r1, target
bfi r0, r1, #0, #PAGE_SHIFT
- mov pc, r0
+ ret r0

target: @ We're now in the trampoline code, switch page tables
mcrr p15, 4, r2, r3, c2
diff --git a/arch/arm/lib/ashldi3.S b/arch/arm/lib/ashldi3.S
index 638deb13da1c..b05e95840651 100644
--- a/arch/arm/lib/ashldi3.S
+++ b/arch/arm/lib/ashldi3.S
@@ -27,6 +27,7 @@ Boston, MA 02110-1301, USA. */


#include <linux/linkage.h>
+#include <asm/assembler.h>

#ifdef __ARMEB__
#define al r1
@@ -47,7 +48,7 @@ ENTRY(__aeabi_llsl)
THUMB( lsrmi r3, al, ip )
THUMB( orrmi ah, ah, r3 )
mov al, al, lsl r2
- mov pc, lr
+ ret lr

ENDPROC(__ashldi3)
ENDPROC(__aeabi_llsl)
diff --git a/arch/arm/lib/ashrdi3.S b/arch/arm/lib/ashrdi3.S
index 015e8aa5a1d1..275d7d2341a4 100644
--- a/arch/arm/lib/ashrdi3.S
+++ b/arch/arm/lib/ashrdi3.S
@@ -27,6 +27,7 @@ Boston, MA 02110-1301, USA. */


#include <linux/linkage.h>
+#include <asm/assembler.h>

#ifdef __ARMEB__
#define al r1
@@ -47,7 +48,7 @@ ENTRY(__aeabi_lasr)
THUMB( lslmi r3, ah, ip )
THUMB( orrmi al, al, r3 )
mov ah, ah, asr r2
- mov pc, lr
+ ret lr

ENDPROC(__ashrdi3)
ENDPROC(__aeabi_lasr)
diff --git a/arch/arm/lib/backtrace.S b/arch/arm/lib/backtrace.S
index 4102be617fce..fab5a50503ae 100644
--- a/arch/arm/lib/backtrace.S
+++ b/arch/arm/lib/backtrace.S
@@ -25,7 +25,7 @@
ENTRY(c_backtrace)

#if !defined(CONFIG_FRAME_POINTER) || !defined(CONFIG_PRINTK)
- mov pc, lr
+ ret lr
ENDPROC(c_backtrace)
#else
stmfd sp!, {r4 - r8, lr} @ Save an extra register so we have a location...
diff --git a/arch/arm/lib/bitops.h b/arch/arm/lib/bitops.h
index 9f12ed1eea86..7d807cfd8ef5 100644
--- a/arch/arm/lib/bitops.h
+++ b/arch/arm/lib/bitops.h
@@ -1,3 +1,4 @@
+#include <asm/assembler.h>
#include <asm/unwind.h>

#if __LINUX_ARM_ARCH__ >= 6
@@ -70,7 +71,7 @@ UNWIND( .fnstart )
\instr r2, r2, r3
str r2, [r1, r0, lsl #2]
restore_irqs ip
- mov pc, lr
+ ret lr
UNWIND( .fnend )
ENDPROC(\name )
.endm
@@ -98,7 +99,7 @@ UNWIND( .fnstart )
\store r2, [r1]
moveq r0, #0
restore_irqs ip
- mov pc, lr
+ ret lr
UNWIND( .fnend )
ENDPROC(\name )
.endm
diff --git a/arch/arm/lib/bswapsdi2.S b/arch/arm/lib/bswapsdi2.S
index 9fcdd154eff9..07cda737bb11 100644
--- a/arch/arm/lib/bswapsdi2.S
+++ b/arch/arm/lib/bswapsdi2.S
@@ -1,4 +1,5 @@
#include <linux/linkage.h>
+#include <asm/assembler.h>

#if __LINUX_ARM_ARCH__ >= 6
ENTRY(__bswapsi2)
@@ -18,7 +19,7 @@ ENTRY(__bswapsi2)
mov r3, r3, lsr #8
bic r3, r3, #0xff00
eor r0, r3, r0, ror #8
- mov pc, lr
+ ret lr
ENDPROC(__bswapsi2)

ENTRY(__bswapdi2)
@@ -31,6 +32,6 @@ ENTRY(__bswapdi2)
bic r1, r1, #0xff00
eor r1, r1, r0, ror #8
eor r0, r3, ip, ror #8
- mov pc, lr
+ ret lr
ENDPROC(__bswapdi2)
#endif
diff --git a/arch/arm/lib/call_with_stack.S b/arch/arm/lib/call_with_stack.S
index 916c80f13ae7..ed1a421813cb 100644
--- a/arch/arm/lib/call_with_stack.S
+++ b/arch/arm/lib/call_with_stack.S
@@ -36,9 +36,9 @@ ENTRY(call_with_stack)
mov r0, r1

adr lr, BSYM(1f)
- mov pc, r2
+ ret r2

1: ldr lr, [sp]
ldr sp, [sp, #4]
- mov pc, lr
+ ret lr
ENDPROC(call_with_stack)
diff --git a/arch/arm/lib/csumpartial.S b/arch/arm/lib/csumpartial.S
index 31d3cb34740d..984e0f29d548 100644
--- a/arch/arm/lib/csumpartial.S
+++ b/arch/arm/lib/csumpartial.S
@@ -97,7 +97,7 @@ td3 .req lr
#endif
#endif
adcnes sum, sum, td0 @ update checksum
- mov pc, lr
+ ret lr

ENTRY(csum_partial)
stmfd sp!, {buf, lr}
diff --git a/arch/arm/lib/csumpartialcopygeneric.S b/arch/arm/lib/csumpartialcopygeneric.S
index d6e742d24007..10b45909610c 100644
--- a/arch/arm/lib/csumpartialcopygeneric.S
+++ b/arch/arm/lib/csumpartialcopygeneric.S
@@ -7,6 +7,7 @@
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*/
+#include <asm/assembler.h>

/*
* unsigned int
@@ -40,7 +41,7 @@ sum .req r3
adcs sum, sum, ip, put_byte_1 @ update checksum
strb ip, [dst], #1
tst dst, #2
- moveq pc, lr @ dst is now 32bit aligned
+ reteq lr @ dst is now 32bit aligned

.Ldst_16bit: load2b r8, ip
sub len, len, #2
@@ -48,7 +49,7 @@ sum .req r3
strb r8, [dst], #1
adcs sum, sum, ip, put_byte_1
strb ip, [dst], #1
- mov pc, lr @ dst is now 32bit aligned
+ ret lr @ dst is now 32bit aligned

/*
* Handle 0 to 7 bytes, with any alignment of source and
diff --git a/arch/arm/lib/delay-loop.S b/arch/arm/lib/delay-loop.S
index bc1033b897b4..518bf6e93f78 100644
--- a/arch/arm/lib/delay-loop.S
+++ b/arch/arm/lib/delay-loop.S
@@ -35,7 +35,7 @@ ENTRY(__loop_const_udelay) @ 0 <= r0 <= 0x7fffff06
mul r0, r2, r0 @ max = 2^32-1
add r0, r0, r1, lsr #32-6
movs r0, r0, lsr #6
- moveq pc, lr
+ reteq lr

/*
* loops = r0 * HZ * loops_per_jiffy / 1000000
@@ -46,23 +46,23 @@ ENTRY(__loop_const_udelay) @ 0 <= r0 <= 0x7fffff06
ENTRY(__loop_delay)
subs r0, r0, #1
#if 0
- movls pc, lr
+ retls lr
subs r0, r0, #1
- movls pc, lr
+ retls lr
subs r0, r0, #1
- movls pc, lr
+ retls lr
subs r0, r0, #1
- movls pc, lr
+ retls lr
subs r0, r0, #1
- movls pc, lr
+ retls lr
subs r0, r0, #1
- movls pc, lr
+ retls lr
subs r0, r0, #1
- movls pc, lr
+ retls lr
subs r0, r0, #1
#endif
bhi __loop_delay
- mov pc, lr
+ ret lr
ENDPROC(__loop_udelay)
ENDPROC(__loop_const_udelay)
ENDPROC(__loop_delay)
diff --git a/arch/arm/lib/div64.S b/arch/arm/lib/div64.S
index e55c4842c290..a9eafe4981eb 100644
--- a/arch/arm/lib/div64.S
+++ b/arch/arm/lib/div64.S
@@ -13,6 +13,7 @@
*/

#include <linux/linkage.h>
+#include <asm/assembler.h>
#include <asm/unwind.h>

#ifdef __ARMEB__
@@ -97,7 +98,7 @@ UNWIND(.fnstart)
mov yl, #0
cmpeq xl, r4
movlo xh, xl
- movlo pc, lr
+ retlo lr

@ The division loop for lower bit positions.
@ Here we shift remainer bits leftwards rather than moving the
@@ -111,14 +112,14 @@ UNWIND(.fnstart)
subcs xh, xh, r4
movs ip, ip, lsr #1
bne 4b
- mov pc, lr
+ ret lr

@ The top part of remainder became zero. If carry is set
@ (the 33th bit) this is a false positive so resume the loop.
@ Otherwise, if lower part is also null then we are done.
6: bcs 5b
cmp xl, #0
- moveq pc, lr
+ reteq lr

@ We still have remainer bits in the low part. Bring them up.

@@ -144,7 +145,7 @@ UNWIND(.fnstart)
movs ip, ip, lsr #1
mov xh, #1
bne 4b
- mov pc, lr
+ ret lr

8: @ Division by a power of 2: determine what that divisor order is
@ then simply shift values around
@@ -184,13 +185,13 @@ UNWIND(.fnstart)
THUMB( orr yl, yl, xh )
mov xh, xl, lsl ip
mov xh, xh, lsr ip
- mov pc, lr
+ ret lr

@ eq -> division by 1: obvious enough...
9: moveq yl, xl
moveq yh, xh
moveq xh, #0
- moveq pc, lr
+ reteq lr
UNWIND(.fnend)

UNWIND(.fnstart)
diff --git a/arch/arm/lib/findbit.S b/arch/arm/lib/findbit.S
index 64f6bc1a9132..7848780e8834 100644
--- a/arch/arm/lib/findbit.S
+++ b/arch/arm/lib/findbit.S
@@ -35,7 +35,7 @@ ENTRY(_find_first_zero_bit_le)
2: cmp r2, r1 @ any more?
blo 1b
3: mov r0, r1 @ no free bits
- mov pc, lr
+ ret lr
ENDPROC(_find_first_zero_bit_le)

/*
@@ -76,7 +76,7 @@ ENTRY(_find_first_bit_le)
2: cmp r2, r1 @ any more?
blo 1b
3: mov r0, r1 @ no free bits
- mov pc, lr
+ ret lr
ENDPROC(_find_first_bit_le)

/*
@@ -114,7 +114,7 @@ ENTRY(_find_first_zero_bit_be)
2: cmp r2, r1 @ any more?
blo 1b
3: mov r0, r1 @ no free bits
- mov pc, lr
+ ret lr
ENDPROC(_find_first_zero_bit_be)

ENTRY(_find_next_zero_bit_be)
@@ -148,7 +148,7 @@ ENTRY(_find_first_bit_be)
2: cmp r2, r1 @ any more?
blo 1b
3: mov r0, r1 @ no free bits
- mov pc, lr
+ ret lr
ENDPROC(_find_first_bit_be)

ENTRY(_find_next_bit_be)
@@ -192,5 +192,5 @@ ENDPROC(_find_next_bit_be)
#endif
cmp r1, r0 @ Clamp to maxbit
movlo r0, r1
- mov pc, lr
+ ret lr

diff --git a/arch/arm/lib/getuser.S b/arch/arm/lib/getuser.S
index 9b06bb41fca6..0f958e3d8180 100644
--- a/arch/arm/lib/getuser.S
+++ b/arch/arm/lib/getuser.S
@@ -36,7 +36,7 @@ ENTRY(__get_user_1)
check_uaccess r0, 1, r1, r2, __get_user_bad
1: TUSER(ldrb) r2, [r0]
mov r0, #0
- mov pc, lr
+ ret lr
ENDPROC(__get_user_1)

ENTRY(__get_user_2)
@@ -56,20 +56,20 @@ rb .req r0
orr r2, rb, r2, lsl #8
#endif
mov r0, #0
- mov pc, lr
+ ret lr
ENDPROC(__get_user_2)

ENTRY(__get_user_4)
check_uaccess r0, 4, r1, r2, __get_user_bad
4: TUSER(ldr) r2, [r0]
mov r0, #0
- mov pc, lr
+ ret lr
ENDPROC(__get_user_4)

__get_user_bad:
mov r2, #0
mov r0, #-EFAULT
- mov pc, lr
+ ret lr
ENDPROC(__get_user_bad)

.pushsection __ex_table, "a"
diff --git a/arch/arm/lib/io-readsb.S b/arch/arm/lib/io-readsb.S
index 9f4238987fe9..c31b2f3153f1 100644
--- a/arch/arm/lib/io-readsb.S
+++ b/arch/arm/lib/io-readsb.S
@@ -25,7 +25,7 @@

ENTRY(__raw_readsb)
teq r2, #0 @ do we have to check for the zero len?
- moveq pc, lr
+ reteq lr
ands ip, r1, #3
bne .Linsb_align

diff --git a/arch/arm/lib/io-readsl.S b/arch/arm/lib/io-readsl.S
index 7a7430950c79..2ed86fa5465f 100644
--- a/arch/arm/lib/io-readsl.S
+++ b/arch/arm/lib/io-readsl.S
@@ -12,7 +12,7 @@

ENTRY(__raw_readsl)
teq r2, #0 @ do we have to check for the zero len?
- moveq pc, lr
+ reteq lr
ands ip, r1, #3
bne 3f

@@ -33,7 +33,7 @@ ENTRY(__raw_readsl)
stmcsia r1!, {r3, ip}
ldrne r3, [r0, #0]
strne r3, [r1, #0]
- mov pc, lr
+ ret lr

3: ldr r3, [r0]
cmp ip, #2
@@ -75,5 +75,5 @@ ENTRY(__raw_readsl)
strb r3, [r1, #1]
8: mov r3, ip, get_byte_0
strb r3, [r1, #0]
- mov pc, lr
+ ret lr
ENDPROC(__raw_readsl)
diff --git a/arch/arm/lib/io-readsw-armv3.S b/arch/arm/lib/io-readsw-armv3.S
index 88487c8c4f23..413da9914529 100644
--- a/arch/arm/lib/io-readsw-armv3.S
+++ b/arch/arm/lib/io-readsw-armv3.S
@@ -27,11 +27,11 @@
strb r3, [r1], #1

subs r2, r2, #1
- moveq pc, lr
+ reteq lr

ENTRY(__raw_readsw)
teq r2, #0 @ do we have to check for the zero len?
- moveq pc, lr
+ reteq lr
tst r1, #3
bne .Linsw_align

diff --git a/arch/arm/lib/io-readsw-armv4.S b/arch/arm/lib/io-readsw-armv4.S
index 1f393d42593d..d9a45e9692ae 100644
--- a/arch/arm/lib/io-readsw-armv4.S
+++ b/arch/arm/lib/io-readsw-armv4.S
@@ -26,7 +26,7 @@

ENTRY(__raw_readsw)
teq r2, #0
- moveq pc, lr
+ reteq lr
tst r1, #3
bne .Linsw_align

diff --git a/arch/arm/lib/io-writesb.S b/arch/arm/lib/io-writesb.S
index 68b92f4acaeb..a46bbc9b168b 100644
--- a/arch/arm/lib/io-writesb.S
+++ b/arch/arm/lib/io-writesb.S
@@ -45,7 +45,7 @@

ENTRY(__raw_writesb)
teq r2, #0 @ do we have to check for the zero len?
- moveq pc, lr
+ reteq lr
ands ip, r1, #3
bne .Loutsb_align

diff --git a/arch/arm/lib/io-writesl.S b/arch/arm/lib/io-writesl.S
index d0d104a0dd11..4ea2435988c1 100644
--- a/arch/arm/lib/io-writesl.S
+++ b/arch/arm/lib/io-writesl.S
@@ -12,7 +12,7 @@

ENTRY(__raw_writesl)
teq r2, #0 @ do we have to check for the zero len?
- moveq pc, lr
+ reteq lr
ands ip, r1, #3
bne 3f

@@ -33,7 +33,7 @@ ENTRY(__raw_writesl)
ldrne r3, [r1, #0]
strcs ip, [r0, #0]
strne r3, [r0, #0]
- mov pc, lr
+ ret lr

3: bic r1, r1, #3
ldr r3, [r1], #4
@@ -47,7 +47,7 @@ ENTRY(__raw_writesl)
orr ip, ip, r3, lspush #16
str ip, [r0]
bne 4b
- mov pc, lr
+ ret lr

5: mov ip, r3, lspull #8
ldr r3, [r1], #4
@@ -55,7 +55,7 @@ ENTRY(__raw_writesl)
orr ip, ip, r3, lspush #24
str ip, [r0]
bne 5b
- mov pc, lr
+ ret lr

6: mov ip, r3, lspull #24
ldr r3, [r1], #4
@@ -63,5 +63,5 @@ ENTRY(__raw_writesl)
orr ip, ip, r3, lspush #8
str ip, [r0]
bne 6b
- mov pc, lr
+ ret lr
ENDPROC(__raw_writesl)
diff --git a/arch/arm/lib/io-writesw-armv3.S b/arch/arm/lib/io-writesw-armv3.S
index 49b800419e32..121789eb6802 100644
--- a/arch/arm/lib/io-writesw-armv3.S
+++ b/arch/arm/lib/io-writesw-armv3.S
@@ -28,11 +28,11 @@
orr r3, r3, r3, lsl #16
str r3, [r0]
subs r2, r2, #1
- moveq pc, lr
+ reteq lr

ENTRY(__raw_writesw)
teq r2, #0 @ do we have to check for the zero len?
- moveq pc, lr
+ reteq lr
tst r1, #3
bne .Loutsw_align

diff --git a/arch/arm/lib/io-writesw-armv4.S b/arch/arm/lib/io-writesw-armv4.S
index ff4f71b579ee..269f90c51ad2 100644
--- a/arch/arm/lib/io-writesw-armv4.S
+++ b/arch/arm/lib/io-writesw-armv4.S
@@ -31,7 +31,7 @@

ENTRY(__raw_writesw)
teq r2, #0
- moveq pc, lr
+ reteq lr
ands r3, r1, #3
bne .Loutsw_align

@@ -96,5 +96,5 @@ ENTRY(__raw_writesw)
tst r2, #1
3: movne ip, r3, lsr #8
strneh ip, [r0]
- mov pc, lr
+ ret lr
ENDPROC(__raw_writesw)
diff --git a/arch/arm/lib/lib1funcs.S b/arch/arm/lib/lib1funcs.S
index c562f649734c..947567ff67f9 100644
--- a/arch/arm/lib/lib1funcs.S
+++ b/arch/arm/lib/lib1funcs.S
@@ -210,7 +210,7 @@ ENTRY(__aeabi_uidiv)
UNWIND(.fnstart)

subs r2, r1, #1
- moveq pc, lr
+ reteq lr
bcc Ldiv0
cmp r0, r1
bls 11f
@@ -220,16 +220,16 @@ UNWIND(.fnstart)
ARM_DIV_BODY r0, r1, r2, r3

mov r0, r2
- mov pc, lr
+ ret lr

11: moveq r0, #1
movne r0, #0
- mov pc, lr
+ ret lr

12: ARM_DIV2_ORDER r1, r2

mov r0, r0, lsr r2
- mov pc, lr
+ ret lr

UNWIND(.fnend)
ENDPROC(__udivsi3)
@@ -244,11 +244,11 @@ UNWIND(.fnstart)
moveq r0, #0
tsthi r1, r2 @ see if divisor is power of 2
andeq r0, r0, r2
- movls pc, lr
+ retls lr

ARM_MOD_BODY r0, r1, r2, r3

- mov pc, lr
+ ret lr

UNWIND(.fnend)
ENDPROC(__umodsi3)
@@ -274,23 +274,23 @@ UNWIND(.fnstart)

cmp ip, #0
rsbmi r0, r0, #0
- mov pc, lr
+ ret lr

10: teq ip, r0 @ same sign ?
rsbmi r0, r0, #0
- mov pc, lr
+ ret lr

11: movlo r0, #0
moveq r0, ip, asr #31
orreq r0, r0, #1
- mov pc, lr
+ ret lr

12: ARM_DIV2_ORDER r1, r2

cmp ip, #0
mov r0, r3, lsr r2
rsbmi r0, r0, #0
- mov pc, lr
+ ret lr

UNWIND(.fnend)
ENDPROC(__divsi3)
@@ -315,7 +315,7 @@ UNWIND(.fnstart)

10: cmp ip, #0
rsbmi r0, r0, #0
- mov pc, lr
+ ret lr

UNWIND(.fnend)
ENDPROC(__modsi3)
@@ -331,7 +331,7 @@ UNWIND(.save {r0, r1, ip, lr} )
ldmfd sp!, {r1, r2, ip, lr}
mul r3, r0, r2
sub r1, r1, r3
- mov pc, lr
+ ret lr

UNWIND(.fnend)
ENDPROC(__aeabi_uidivmod)
@@ -344,7 +344,7 @@ UNWIND(.save {r0, r1, ip, lr} )
ldmfd sp!, {r1, r2, ip, lr}
mul r3, r0, r2
sub r1, r1, r3
- mov pc, lr
+ ret lr

UNWIND(.fnend)
ENDPROC(__aeabi_idivmod)
diff --git a/arch/arm/lib/lshrdi3.S b/arch/arm/lib/lshrdi3.S
index f83d449141f7..922dcd88b02b 100644
--- a/arch/arm/lib/lshrdi3.S
+++ b/arch/arm/lib/lshrdi3.S
@@ -27,6 +27,7 @@ Boston, MA 02110-1301, USA. */


#include <linux/linkage.h>
+#include <asm/assembler.h>

#ifdef __ARMEB__
#define al r1
@@ -47,7 +48,7 @@ ENTRY(__aeabi_llsr)
THUMB( lslmi r3, ah, ip )
THUMB( orrmi al, al, r3 )
mov ah, ah, lsr r2
- mov pc, lr
+ ret lr

ENDPROC(__lshrdi3)
ENDPROC(__aeabi_llsr)
diff --git a/arch/arm/lib/memchr.S b/arch/arm/lib/memchr.S
index 1da86991d700..74a5bed6d999 100644
--- a/arch/arm/lib/memchr.S
+++ b/arch/arm/lib/memchr.S
@@ -22,5 +22,5 @@ ENTRY(memchr)
bne 1b
sub r0, r0, #1
2: movne r0, #0
- mov pc, lr
+ ret lr
ENDPROC(memchr)
diff --git a/arch/arm/lib/memset.S b/arch/arm/lib/memset.S
index 94b0650ea98f..671455c854fa 100644
--- a/arch/arm/lib/memset.S
+++ b/arch/arm/lib/memset.S
@@ -110,7 +110,7 @@ ENTRY(memset)
strneb r1, [ip], #1
tst r2, #1
strneb r1, [ip], #1
- mov pc, lr
+ ret lr

6: subs r2, r2, #4 @ 1 do we have enough
blt 5b @ 1 bytes to align with?
diff --git a/arch/arm/lib/memzero.S b/arch/arm/lib/memzero.S
index 3fbdef5f802a..385ccb306fa2 100644
--- a/arch/arm/lib/memzero.S
+++ b/arch/arm/lib/memzero.S
@@ -121,5 +121,5 @@ ENTRY(__memzero)
strneb r2, [r0], #1 @ 1
tst r1, #1 @ 1 a byte left over
strneb r2, [r0], #1 @ 1
- mov pc, lr @ 1
+ ret lr @ 1
ENDPROC(__memzero)
diff --git a/arch/arm/lib/muldi3.S b/arch/arm/lib/muldi3.S
index 36c91b4957e2..204305956925 100644
--- a/arch/arm/lib/muldi3.S
+++ b/arch/arm/lib/muldi3.S
@@ -11,6 +11,7 @@
*/

#include <linux/linkage.h>
+#include <asm/assembler.h>

#ifdef __ARMEB__
#define xh r0
@@ -41,7 +42,7 @@ ENTRY(__aeabi_lmul)
adc xh, xh, yh, lsr #16
adds xl, xl, ip, lsl #16
adc xh, xh, ip, lsr #16
- mov pc, lr
+ ret lr

ENDPROC(__muldi3)
ENDPROC(__aeabi_lmul)
diff --git a/arch/arm/lib/putuser.S b/arch/arm/lib/putuser.S
index 3d73dcb959b0..38d660d3705f 100644
--- a/arch/arm/lib/putuser.S
+++ b/arch/arm/lib/putuser.S
@@ -36,7 +36,7 @@ ENTRY(__put_user_1)
check_uaccess r0, 1, r1, ip, __put_user_bad
1: TUSER(strb) r2, [r0]
mov r0, #0
- mov pc, lr
+ ret lr
ENDPROC(__put_user_1)

ENTRY(__put_user_2)
@@ -60,14 +60,14 @@ ENTRY(__put_user_2)
#endif
#endif /* CONFIG_THUMB2_KERNEL */
mov r0, #0
- mov pc, lr
+ ret lr
ENDPROC(__put_user_2)

ENTRY(__put_user_4)
check_uaccess r0, 4, r1, ip, __put_user_bad
4: TUSER(str) r2, [r0]
mov r0, #0
- mov pc, lr
+ ret lr
ENDPROC(__put_user_4)

ENTRY(__put_user_8)
@@ -80,12 +80,12 @@ ENTRY(__put_user_8)
6: TUSER(str) r3, [r0]
#endif
mov r0, #0
- mov pc, lr
+ ret lr
ENDPROC(__put_user_8)

__put_user_bad:
mov r0, #-EFAULT
- mov pc, lr
+ ret lr
ENDPROC(__put_user_bad)

.pushsection __ex_table, "a"
diff --git a/arch/arm/lib/strchr.S b/arch/arm/lib/strchr.S
index d8f2a1c1aea4..013d64c71e8d 100644
--- a/arch/arm/lib/strchr.S
+++ b/arch/arm/lib/strchr.S
@@ -23,5 +23,5 @@ ENTRY(strchr)
teq r2, r1
movne r0, #0
subeq r0, r0, #1
- mov pc, lr
+ ret lr
ENDPROC(strchr)
diff --git a/arch/arm/lib/strrchr.S b/arch/arm/lib/strrchr.S
index 302f20cd2423..3cec1c7482c4 100644
--- a/arch/arm/lib/strrchr.S
+++ b/arch/arm/lib/strrchr.S
@@ -22,5 +22,5 @@ ENTRY(strrchr)
teq r2, #0
bne 1b
mov r0, r3
- mov pc, lr
+ ret lr
ENDPROC(strrchr)
diff --git a/arch/arm/lib/ucmpdi2.S b/arch/arm/lib/ucmpdi2.S
index f0df6a91db04..ad4a6309141a 100644
--- a/arch/arm/lib/ucmpdi2.S
+++ b/arch/arm/lib/ucmpdi2.S
@@ -11,6 +11,7 @@
*/

#include <linux/linkage.h>
+#include <asm/assembler.h>

#ifdef __ARMEB__
#define xh r0
@@ -31,7 +32,7 @@ ENTRY(__ucmpdi2)
movlo r0, #0
moveq r0, #1
movhi r0, #2
- mov pc, lr
+ ret lr

ENDPROC(__ucmpdi2)

@@ -44,7 +45,7 @@ ENTRY(__aeabi_ulcmp)
movlo r0, #-1
moveq r0, #0
movhi r0, #1
- mov pc, lr
+ ret lr

ENDPROC(__aeabi_ulcmp)

diff --git a/arch/arm/mach-davinci/sleep.S b/arch/arm/mach-davinci/sleep.S
index d4e9316ecacb..a5336a5e2739 100644
--- a/arch/arm/mach-davinci/sleep.S
+++ b/arch/arm/mach-davinci/sleep.S
@@ -213,7 +213,7 @@ ddr2clk_stop_done:
cmp ip, r0
bne ddr2clk_stop_done

- mov pc, lr
+ ret lr
ENDPROC(davinci_ddr_psc_config)

CACHE_FLUSH:
diff --git a/arch/arm/mach-ep93xx/crunch-bits.S b/arch/arm/mach-ep93xx/crunch-bits.S
index e96923a3017b..ee0be2af5c61 100644
--- a/arch/arm/mach-ep93xx/crunch-bits.S
+++ b/arch/arm/mach-ep93xx/crunch-bits.S
@@ -198,7 +198,7 @@ crunch_load:
get_thread_info r10
#endif
2: dec_preempt_count r10, r3
- mov pc, lr
+ ret lr

/*
* Back up crunch regs to save area and disable access to them
@@ -277,7 +277,7 @@ ENTRY(crunch_task_copy)
mov r3, lr @ preserve return address
bl crunch_save
msr cpsr_c, ip @ restore interrupt mode
- mov pc, r3
+ ret r3

/*
* Restore crunch state from given memory address
@@ -310,4 +310,4 @@ ENTRY(crunch_task_restore)
mov r3, lr @ preserve return address
bl crunch_load
msr cpsr_c, ip @ restore interrupt mode
- mov pc, r3
+ ret r3
diff --git a/arch/arm/mach-imx/suspend-imx6.S b/arch/arm/mach-imx/suspend-imx6.S
index fe123b079c05..74b50f1982db 100644
--- a/arch/arm/mach-imx/suspend-imx6.S
+++ b/arch/arm/mach-imx/suspend-imx6.S
@@ -10,6 +10,7 @@
*/

#include <linux/linkage.h>
+#include <asm/assembler.h>
#include <asm/asm-offsets.h>
#include <asm/hardware/cache-l2x0.h>
#include "hardware.h"
@@ -301,7 +302,7 @@ rbc_loop:
resume_mmdc

/* return to suspend finish */
- mov pc, lr
+ ret lr

resume:
/* invalidate L1 I-cache first */
@@ -325,7 +326,7 @@ resume:
mov r5, #0x1
resume_mmdc

- mov pc, lr
+ ret lr
ENDPROC(imx6_suspend)

/*
diff --git a/arch/arm/mach-mvebu/coherency_ll.S b/arch/arm/mach-mvebu/coherency_ll.S
index 510c29e079ca..f5d881b5d0f7 100644
--- a/arch/arm/mach-mvebu/coherency_ll.S
+++ b/arch/arm/mach-mvebu/coherency_ll.S
@@ -46,7 +46,7 @@ ENTRY(ll_get_coherency_base)
ldr r1, =coherency_base
ldr r1, [r1]
2:
- mov pc, lr
+ ret lr
ENDPROC(ll_get_coherency_base)

/*
@@ -63,7 +63,7 @@ ENTRY(ll_get_coherency_cpumask)
mov r2, #(1 << 24)
lsl r3, r2, r3
ARM_BE8(rev r3, r3)
- mov pc, lr
+ ret lr
ENDPROC(ll_get_coherency_cpumask)

/*
@@ -94,7 +94,7 @@ ENTRY(ll_add_cpu_to_smp_group)
strex r1, r2, [r0]
cmp r1, #0
bne 1b
- mov pc, lr
+ ret lr
ENDPROC(ll_add_cpu_to_smp_group)

ENTRY(ll_enable_coherency)
@@ -118,7 +118,7 @@ ENTRY(ll_enable_coherency)
bne 1b
dsb
mov r0, #0
- mov pc, lr
+ ret lr
ENDPROC(ll_enable_coherency)

ENTRY(ll_disable_coherency)
@@ -141,7 +141,7 @@ ENTRY(ll_disable_coherency)
cmp r1, #0
bne 1b
dsb
- mov pc, lr
+ ret lr
ENDPROC(ll_disable_coherency)

.align 2
diff --git a/arch/arm/mach-mvebu/headsmp-a9.S b/arch/arm/mach-mvebu/headsmp-a9.S
index 5925366bc03c..7c91ddb6f1f7 100644
--- a/arch/arm/mach-mvebu/headsmp-a9.S
+++ b/arch/arm/mach-mvebu/headsmp-a9.S
@@ -14,6 +14,7 @@

#include <linux/linkage.h>
#include <linux/init.h>
+#include <asm/assembler.h>

__CPUINIT
#define CPU_RESUME_ADDR_REG 0xf10182d4
@@ -24,7 +25,7 @@
armada_375_smp_cpu1_enable_code_start:
ldr r0, [pc, #4]
ldr r1, [r0]
- mov pc, r1
+ ret r1
.word CPU_RESUME_ADDR_REG
armada_375_smp_cpu1_enable_code_end:

diff --git a/arch/arm/mach-omap2/sleep44xx.S b/arch/arm/mach-omap2/sleep44xx.S
index 9086ce03ae12..b84a0122d823 100644
--- a/arch/arm/mach-omap2/sleep44xx.S
+++ b/arch/arm/mach-omap2/sleep44xx.S
@@ -10,6 +10,7 @@
*/

#include <linux/linkage.h>
+#include <asm/assembler.h>
#include <asm/smp_scu.h>
#include <asm/memory.h>
#include <asm/hardware/cache-l2x0.h>
@@ -334,7 +335,7 @@ ENDPROC(omap4_cpu_resume)

#ifndef CONFIG_OMAP4_ERRATA_I688
ENTRY(omap_bus_sync)
- mov pc, lr
+ ret lr
ENDPROC(omap_bus_sync)
#endif

diff --git a/arch/arm/mach-omap2/sram242x.S b/arch/arm/mach-omap2/sram242x.S
index 680a7c56cc3e..2c88ff2d0236 100644
--- a/arch/arm/mach-omap2/sram242x.S
+++ b/arch/arm/mach-omap2/sram242x.S
@@ -101,7 +101,7 @@ i_dll_wait:
i_dll_delay:
subs r4, r4, #0x1
bne i_dll_delay
- mov pc, lr
+ ret lr

/*
* shift up or down voltage, use R9 as input to tell level.
@@ -125,7 +125,7 @@ volt_delay:
ldr r7, [r3] @ get timer value
cmp r5, r7 @ time up?
bhi volt_delay @ not yet->branch
- mov pc, lr @ back to caller.
+ ret lr @ back to caller.

omap242x_sdi_cm_clksel2_pll:
.word OMAP2420_CM_REGADDR(PLL_MOD, CM_CLKSEL2)
@@ -220,7 +220,7 @@ volt_delay_c:
ldr r7, [r10] @ get timer value
cmp r8, r7 @ time up?
bhi volt_delay_c @ not yet->branch
- mov pc, lr @ back to caller
+ ret lr @ back to caller

omap242x_srs_cm_clksel2_pll:
.word OMAP2420_CM_REGADDR(PLL_MOD, CM_CLKSEL2)
diff --git a/arch/arm/mach-omap2/sram243x.S b/arch/arm/mach-omap2/sram243x.S
index a1e9edd673f4..d5deb9761fc7 100644
--- a/arch/arm/mach-omap2/sram243x.S
+++ b/arch/arm/mach-omap2/sram243x.S
@@ -101,7 +101,7 @@ i_dll_wait:
i_dll_delay:
subs r4, r4, #0x1
bne i_dll_delay
- mov pc, lr
+ ret lr

/*
* shift up or down voltage, use R9 as input to tell level.
@@ -125,7 +125,7 @@ volt_delay:
ldr r7, [r3] @ get timer value
cmp r5, r7 @ time up?
bhi volt_delay @ not yet->branch
- mov pc, lr @ back to caller.
+ ret lr @ back to caller.

omap243x_sdi_cm_clksel2_pll:
.word OMAP2430_CM_REGADDR(PLL_MOD, CM_CLKSEL2)
@@ -220,7 +220,7 @@ volt_delay_c:
ldr r7, [r10] @ get timer value
cmp r8, r7 @ time up?
bhi volt_delay_c @ not yet->branch
- mov pc, lr @ back to caller
+ ret lr @ back to caller

omap243x_srs_cm_clksel2_pll:
.word OMAP2430_CM_REGADDR(PLL_MOD, CM_CLKSEL2)
diff --git a/arch/arm/mach-pxa/mioa701_bootresume.S b/arch/arm/mach-pxa/mioa701_bootresume.S
index 324d25a48c85..81591491ab94 100644
--- a/arch/arm/mach-pxa/mioa701_bootresume.S
+++ b/arch/arm/mach-pxa/mioa701_bootresume.S
@@ -29,7 +29,7 @@ ENTRY(mioa701_jumpaddr)
str r1, [r0] @ Early disable resume for next boot
ldr r0, mioa701_jumpaddr @ (Murphy's Law)
ldr r0, [r0]
- mov pc, r0
+ ret r0
2:

ENTRY(mioa701_bootstrap_lg)
diff --git a/arch/arm/mach-pxa/standby.S b/arch/arm/mach-pxa/standby.S
index 29f5f5c180b7..eab1645bb4ad 100644
--- a/arch/arm/mach-pxa/standby.S
+++ b/arch/arm/mach-pxa/standby.S
@@ -29,7 +29,7 @@ ENTRY(pxa_cpu_standby)
.align 5
1: mcr p14, 0, r2, c7, c0, 0 @ put the system into Standby
str r1, [r0] @ make sure PSSR_PH/STS are clear
- mov pc, lr
+ ret lr

#endif

@@ -108,7 +108,7 @@ ENTRY(pm_enter_standby_start)
bic r0, r0, #0x20000000
str r0, [r1, #PXA3_DMCIER]

- mov pc, lr
+ ret lr
ENTRY(pm_enter_standby_end)

#endif
diff --git a/arch/arm/mach-s3c24xx/sleep-s3c2410.S b/arch/arm/mach-s3c24xx/sleep-s3c2410.S
index c9b91223697c..875ba8911127 100644
--- a/arch/arm/mach-s3c24xx/sleep-s3c2410.S
+++ b/arch/arm/mach-s3c24xx/sleep-s3c2410.S
@@ -66,4 +66,4 @@ s3c2410_do_sleep:
streq r8, [r5] @ SDRAM power-down config
streq r9, [r6] @ CPU sleep
1: beq 1b
- mov pc, r14
+ ret lr
diff --git a/arch/arm/mach-s3c24xx/sleep-s3c2412.S b/arch/arm/mach-s3c24xx/sleep-s3c2412.S
index 5adaceb7da13..6bf5b4d8743c 100644
--- a/arch/arm/mach-s3c24xx/sleep-s3c2412.S
+++ b/arch/arm/mach-s3c24xx/sleep-s3c2412.S
@@ -65,4 +65,4 @@ s3c2412_sleep_enter1:
strne r9, [r3]
bne s3c2412_sleep_enter1

- mov pc, r14
+ ret lr
diff --git a/arch/arm/mach-shmobile/headsmp.S b/arch/arm/mach-shmobile/headsmp.S
index e5be5c88644b..293007579b8e 100644
--- a/arch/arm/mach-shmobile/headsmp.S
+++ b/arch/arm/mach-shmobile/headsmp.S
@@ -12,6 +12,7 @@
*/
#include <linux/linkage.h>
#include <linux/init.h>
+#include <asm/assembler.h>
#include <asm/memory.h>

ENTRY(shmobile_invalidate_start)
@@ -75,7 +76,7 @@ shmobile_smp_boot_next:

shmobile_smp_boot_found:
ldr r0, [r7, r1, lsl #2]
- mov pc, r9
+ ret r9
ENDPROC(shmobile_smp_boot)

ENTRY(shmobile_smp_sleep)
diff --git a/arch/arm/mach-tegra/sleep-tegra20.S b/arch/arm/mach-tegra/sleep-tegra20.S
index aaaf3abd2688..be4bc5f853f5 100644
--- a/arch/arm/mach-tegra/sleep-tegra20.S
+++ b/arch/arm/mach-tegra/sleep-tegra20.S
@@ -78,7 +78,7 @@ ENTRY(tegra20_hotplug_shutdown)
/* Put this CPU down */
cpu_id r0
bl tegra20_cpu_shutdown
- mov pc, lr @ should never get here
+ ret lr @ should never get here
ENDPROC(tegra20_hotplug_shutdown)

/*
@@ -96,7 +96,7 @@ ENDPROC(tegra20_hotplug_shutdown)
*/
ENTRY(tegra20_cpu_shutdown)
cmp r0, #0
- moveq pc, lr @ must not be called for CPU 0
+ reteq lr @ must not be called for CPU 0
mov32 r1, TEGRA_PMC_VIRT + PMC_SCRATCH41
mov r12, #CPU_RESETTABLE
str r12, [r1]
@@ -117,7 +117,7 @@ ENTRY(tegra20_cpu_shutdown)
cpu_id r3
cmp r3, r0
beq .
- mov pc, lr
+ ret lr
ENDPROC(tegra20_cpu_shutdown)
#endif

@@ -164,7 +164,7 @@ ENTRY(tegra_pen_lock)
cmpeq r12, r0 @ !turn == cpu?
beq 1b @ while !turn == cpu && flag[!cpu] == 1

- mov pc, lr @ locked
+ ret lr @ locked
ENDPROC(tegra_pen_lock)

ENTRY(tegra_pen_unlock)
@@ -176,7 +176,7 @@ ENTRY(tegra_pen_unlock)
addne r2, r3, #PMC_SCRATCH39
mov r12, #0
str r12, [r2]
- mov pc, lr
+ ret lr
ENDPROC(tegra_pen_unlock)

/*
@@ -189,7 +189,7 @@ ENTRY(tegra20_cpu_clear_resettable)
mov32 r1, TEGRA_PMC_VIRT + PMC_SCRATCH41
mov r12, #CPU_NOT_RESETTABLE
str r12, [r1]
- mov pc, lr
+ ret lr
ENDPROC(tegra20_cpu_clear_resettable)

/*
@@ -202,7 +202,7 @@ ENTRY(tegra20_cpu_set_resettable_soon)
mov32 r1, TEGRA_PMC_VIRT + PMC_SCRATCH41
mov r12, #CPU_RESETTABLE_SOON
str r12, [r1]
- mov pc, lr
+ ret lr
ENDPROC(tegra20_cpu_set_resettable_soon)

/*
@@ -217,7 +217,7 @@ ENTRY(tegra20_cpu_is_resettable_soon)
cmp r12, #CPU_RESETTABLE_SOON
moveq r0, #1
movne r0, #0
- mov pc, lr
+ ret lr
ENDPROC(tegra20_cpu_is_resettable_soon)

/*
@@ -239,7 +239,7 @@ ENTRY(tegra20_sleep_core_finish)
mov32 r1, TEGRA_IRAM_LPx_RESUME_AREA
add r0, r0, r1

- mov pc, r3
+ ret r3
ENDPROC(tegra20_sleep_core_finish)

/*
@@ -402,7 +402,7 @@ exit_selfrefresh_loop:

mov32 r0, TEGRA_PMC_BASE
ldr r0, [r0, #PMC_SCRATCH41]
- mov pc, r0 @ jump to tegra_resume
+ ret r0 @ jump to tegra_resume
ENDPROC(tegra20_lp1_reset)

/*
@@ -455,7 +455,7 @@ tegra20_switch_cpu_to_clk32k:
mov r0, #0 /* brust policy = 32KHz */
str r0, [r5, #CLK_RESET_SCLK_BURST]

- mov pc, lr
+ ret lr

/*
* tegra20_enter_sleep
@@ -535,7 +535,7 @@ padsave_done:
adr r2, tegra20_sclk_save
str r0, [r2]
dsb
- mov pc, lr
+ ret lr

tegra20_sdram_pad_address:
.word TEGRA_APB_MISC_BASE + APB_MISC_XM2CFGCPADCTRL
diff --git a/arch/arm/mach-tegra/sleep-tegra30.S b/arch/arm/mach-tegra/sleep-tegra30.S
index b16d4a57fa59..09cad9b071de 100644
--- a/arch/arm/mach-tegra/sleep-tegra30.S
+++ b/arch/arm/mach-tegra/sleep-tegra30.S
@@ -142,7 +142,7 @@ ENTRY(tegra30_hotplug_shutdown)
/* Powergate this CPU */
mov r0, #TEGRA30_POWER_HOTPLUG_SHUTDOWN
bl tegra30_cpu_shutdown
- mov pc, lr @ should never get here
+ ret lr @ should never get here
ENDPROC(tegra30_hotplug_shutdown)

/*
@@ -161,7 +161,7 @@ ENTRY(tegra30_cpu_shutdown)
bne _no_cpu0_chk @ It's not Tegra30

cmp r3, #0
- moveq pc, lr @ Must never be called for CPU 0
+ reteq lr @ Must never be called for CPU 0
_no_cpu0_chk:

ldr r12, =TEGRA_FLOW_CTRL_VIRT
@@ -266,7 +266,7 @@ ENTRY(tegra30_sleep_core_finish)
mov32 r1, TEGRA_IRAM_LPx_RESUME_AREA
add r0, r0, r1

- mov pc, r3
+ ret r3
ENDPROC(tegra30_sleep_core_finish)

/*
@@ -285,7 +285,7 @@ ENTRY(tegra30_sleep_cpu_secondary_finish)
mov r0, #0 @ power mode flags (!hotplug)
bl tegra30_cpu_shutdown
mov r0, #1 @ never return here
- mov pc, r7
+ ret r7
ENDPROC(tegra30_sleep_cpu_secondary_finish)

/*
@@ -529,7 +529,7 @@ __no_dual_emc_chanl:

mov32 r0, TEGRA_PMC_BASE
ldr r0, [r0, #PMC_SCRATCH41]
- mov pc, r0 @ jump to tegra_resume
+ ret r0 @ jump to tegra_resume
ENDPROC(tegra30_lp1_reset)

.align L1_CACHE_SHIFT
@@ -659,7 +659,7 @@ _no_pll_in_iddq:
mov r0, #0 /* brust policy = 32KHz */
str r0, [r5, #CLK_RESET_SCLK_BURST]

- mov pc, lr
+ ret lr

/*
* tegra30_enter_sleep
@@ -819,7 +819,7 @@ pmc_io_dpd_skip:

dsb

- mov pc, lr
+ ret lr

.ltorg
/* dummy symbol for end of IRAM */
diff --git a/arch/arm/mach-tegra/sleep.S b/arch/arm/mach-tegra/sleep.S
index 8d06213fbc47..f024a5109e8e 100644
--- a/arch/arm/mach-tegra/sleep.S
+++ b/arch/arm/mach-tegra/sleep.S
@@ -87,7 +87,7 @@ ENTRY(tegra_init_l2_for_a15)
mcrne p15, 0x1, r0, c9, c0, 2
_exit_init_l2_a15:

- mov pc, lr
+ ret lr
ENDPROC(tegra_init_l2_for_a15)

/*
@@ -111,7 +111,7 @@ ENTRY(tegra_sleep_cpu_finish)
add r3, r3, r0
mov r0, r1

- mov pc, r3
+ ret r3
ENDPROC(tegra_sleep_cpu_finish)

/*
@@ -139,7 +139,7 @@ ENTRY(tegra_shut_off_mmu)
moveq r3, #0
streq r3, [r2, #L2X0_CTRL]
#endif
- mov pc, r0
+ ret r0
ENDPROC(tegra_shut_off_mmu)
.popsection

@@ -156,6 +156,6 @@ ENTRY(tegra_switch_cpu_to_pllp)
str r0, [r5, #CLK_RESET_CCLK_BURST]
mov r0, #0
str r0, [r5, #CLK_RESET_CCLK_DIVIDER]
- mov pc, lr
+ ret lr
ENDPROC(tegra_switch_cpu_to_pllp)
#endif
diff --git a/arch/arm/mm/cache-fa.S b/arch/arm/mm/cache-fa.S
index e505befe51b5..2f0c58836ae7 100644
--- a/arch/arm/mm/cache-fa.S
+++ b/arch/arm/mm/cache-fa.S
@@ -15,6 +15,7 @@
*/
#include <linux/linkage.h>
#include <linux/init.h>
+#include <asm/assembler.h>
#include <asm/memory.h>
#include <asm/page.h>

@@ -45,7 +46,7 @@
ENTRY(fa_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
- mov pc, lr
+ ret lr
ENDPROC(fa_flush_icache_all)

/*
@@ -71,7 +72,7 @@ __flush_whole_cache:
mcrne p15, 0, ip, c7, c5, 6 @ invalidate BTB
mcrne p15, 0, ip, c7, c10, 4 @ drain write buffer
mcrne p15, 0, ip, c7, c5, 4 @ prefetch flush
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -99,7 +100,7 @@ ENTRY(fa_flush_user_cache_range)
mcrne p15, 0, ip, c7, c5, 6 @ invalidate BTB
mcrne p15, 0, ip, c7, c10, 4 @ data write barrier
mcrne p15, 0, ip, c7, c5, 4 @ prefetch flush
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -135,7 +136,7 @@ ENTRY(fa_coherent_user_range)
mcr p15, 0, r0, c7, c5, 6 @ invalidate BTB
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
mcr p15, 0, r0, c7, c5, 4 @ prefetch flush
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -155,7 +156,7 @@ ENTRY(fa_flush_kern_dcache_area)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* dma_inv_range(start, end)
@@ -181,7 +182,7 @@ fa_dma_inv_range:
blo 1b
mov r0, #0
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -199,7 +200,7 @@ fa_dma_clean_range:
blo 1b
mov r0, #0
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start,end)
@@ -214,7 +215,7 @@ ENTRY(fa_dma_flush_range)
blo 1b
mov r0, #0
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -237,7 +238,7 @@ ENDPROC(fa_dma_map_area)
* - dir - DMA direction
*/
ENTRY(fa_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(fa_dma_unmap_area)

.globl fa_flush_kern_cache_louis
diff --git a/arch/arm/mm/cache-nop.S b/arch/arm/mm/cache-nop.S
index 8e12ddca0031..f1cc9861031f 100644
--- a/arch/arm/mm/cache-nop.S
+++ b/arch/arm/mm/cache-nop.S
@@ -5,11 +5,12 @@
*/
#include <linux/linkage.h>
#include <linux/init.h>
+#include <asm/assembler.h>

#include "proc-macros.S"

ENTRY(nop_flush_icache_all)
- mov pc, lr
+ ret lr
ENDPROC(nop_flush_icache_all)

.globl nop_flush_kern_cache_all
@@ -29,7 +30,7 @@ ENDPROC(nop_flush_icache_all)

ENTRY(nop_coherent_user_range)
mov r0, 0
- mov pc, lr
+ ret lr
ENDPROC(nop_coherent_user_range)

.globl nop_flush_kern_dcache_area
diff --git a/arch/arm/mm/cache-v4.S b/arch/arm/mm/cache-v4.S
index a7ba68f59f0c..91e3adf155cb 100644
--- a/arch/arm/mm/cache-v4.S
+++ b/arch/arm/mm/cache-v4.S
@@ -9,6 +9,7 @@
*/
#include <linux/linkage.h>
#include <linux/init.h>
+#include <asm/assembler.h>
#include <asm/page.h>
#include "proc-macros.S"

@@ -18,7 +19,7 @@
* Unconditionally clean and invalidate the entire icache.
*/
ENTRY(v4_flush_icache_all)
- mov pc, lr
+ ret lr
ENDPROC(v4_flush_icache_all)

/*
@@ -40,7 +41,7 @@ ENTRY(v4_flush_kern_cache_all)
#ifdef CONFIG_CPU_CP15
mov r0, #0
mcr p15, 0, r0, c7, c7, 0 @ flush ID cache
- mov pc, lr
+ ret lr
#else
/* FALLTHROUGH */
#endif
@@ -59,7 +60,7 @@ ENTRY(v4_flush_user_cache_range)
#ifdef CONFIG_CPU_CP15
mov ip, #0
mcr p15, 0, ip, c7, c7, 0 @ flush ID cache
- mov pc, lr
+ ret lr
#else
/* FALLTHROUGH */
#endif
@@ -89,7 +90,7 @@ ENTRY(v4_coherent_kern_range)
*/
ENTRY(v4_coherent_user_range)
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -116,7 +117,7 @@ ENTRY(v4_dma_flush_range)
mov r0, #0
mcr p15, 0, r0, c7, c7, 0 @ flush ID cache
#endif
- mov pc, lr
+ ret lr

/*
* dma_unmap_area(start, size, dir)
@@ -136,7 +137,7 @@ ENTRY(v4_dma_unmap_area)
* - dir - DMA direction
*/
ENTRY(v4_dma_map_area)
- mov pc, lr
+ ret lr
ENDPROC(v4_dma_unmap_area)
ENDPROC(v4_dma_map_area)

diff --git a/arch/arm/mm/cache-v4wb.S b/arch/arm/mm/cache-v4wb.S
index cd4945321407..2522f8c8fbb1 100644
--- a/arch/arm/mm/cache-v4wb.S
+++ b/arch/arm/mm/cache-v4wb.S
@@ -9,6 +9,7 @@
*/
#include <linux/linkage.h>
#include <linux/init.h>
+#include <asm/assembler.h>
#include <asm/memory.h>
#include <asm/page.h>
#include "proc-macros.S"
@@ -58,7 +59,7 @@ flush_base:
ENTRY(v4wb_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
- mov pc, lr
+ ret lr
ENDPROC(v4wb_flush_icache_all)

/*
@@ -94,7 +95,7 @@ __flush_whole_cache:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -122,7 +123,7 @@ ENTRY(v4wb_flush_user_cache_range)
blo 1b
tst r2, #VM_EXEC
mcrne p15, 0, ip, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -170,7 +171,7 @@ ENTRY(v4wb_coherent_user_range)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr


/*
@@ -195,7 +196,7 @@ v4wb_dma_inv_range:
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -212,7 +213,7 @@ v4wb_dma_clean_range:
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -248,7 +249,7 @@ ENDPROC(v4wb_dma_map_area)
* - dir - DMA direction
*/
ENTRY(v4wb_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(v4wb_dma_unmap_area)

.globl v4wb_flush_kern_cache_louis
diff --git a/arch/arm/mm/cache-v4wt.S b/arch/arm/mm/cache-v4wt.S
index 11e5e5838bc5..a0982ce49007 100644
--- a/arch/arm/mm/cache-v4wt.S
+++ b/arch/arm/mm/cache-v4wt.S
@@ -13,6 +13,7 @@
*/
#include <linux/linkage.h>
#include <linux/init.h>
+#include <asm/assembler.h>
#include <asm/page.h>
#include "proc-macros.S"

@@ -48,7 +49,7 @@
ENTRY(v4wt_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
- mov pc, lr
+ ret lr
ENDPROC(v4wt_flush_icache_all)

/*
@@ -71,7 +72,7 @@ __flush_whole_cache:
tst r2, #VM_EXEC
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
mcr p15, 0, ip, c7, c6, 0 @ invalidate D cache
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -94,7 +95,7 @@ ENTRY(v4wt_flush_user_cache_range)
add r0, r0, #CACHE_DLINESIZE
cmp r0, r1
blo 1b
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -126,7 +127,7 @@ ENTRY(v4wt_coherent_user_range)
cmp r0, r1
blo 1b
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -160,7 +161,7 @@ v4wt_dma_inv_range:
add r0, r0, #CACHE_DLINESIZE
cmp r0, r1
blo 1b
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -192,7 +193,7 @@ ENTRY(v4wt_dma_unmap_area)
* - dir - DMA direction
*/
ENTRY(v4wt_dma_map_area)
- mov pc, lr
+ ret lr
ENDPROC(v4wt_dma_unmap_area)
ENDPROC(v4wt_dma_map_area)

diff --git a/arch/arm/mm/cache-v6.S b/arch/arm/mm/cache-v6.S
index d8fd4d4bd3d4..24659952c278 100644
--- a/arch/arm/mm/cache-v6.S
+++ b/arch/arm/mm/cache-v6.S
@@ -51,7 +51,7 @@ ENTRY(v6_flush_icache_all)
#else
mcr p15, 0, r0, c7, c5, 0 @ invalidate I-cache
#endif
- mov pc, lr
+ ret lr
ENDPROC(v6_flush_icache_all)

/*
@@ -73,7 +73,7 @@ ENTRY(v6_flush_kern_cache_all)
#else
mcr p15, 0, r0, c7, c15, 0 @ Cache clean+invalidate
#endif
- mov pc, lr
+ ret lr

/*
* v6_flush_cache_all()
@@ -98,7 +98,7 @@ ENTRY(v6_flush_user_cache_all)
* - we have a VIPT cache.
*/
ENTRY(v6_flush_user_cache_range)
- mov pc, lr
+ ret lr

/*
* v6_coherent_kern_range(start,end)
@@ -150,7 +150,7 @@ ENTRY(v6_coherent_user_range)
#else
mcr p15, 0, r0, c7, c5, 6 @ invalidate BTB
#endif
- mov pc, lr
+ ret lr

/*
* Fault handling for the cache operation above. If the virtual address in r0
@@ -158,7 +158,7 @@ ENTRY(v6_coherent_user_range)
*/
9001:
mov r0, #-EFAULT
- mov pc, lr
+ ret lr
UNWIND(.fnend )
ENDPROC(v6_coherent_user_range)
ENDPROC(v6_coherent_kern_range)
@@ -188,7 +188,7 @@ ENTRY(v6_flush_kern_dcache_area)
mov r0, #0
mcr p15, 0, r0, c7, c10, 4
#endif
- mov pc, lr
+ ret lr


/*
@@ -239,7 +239,7 @@ v6_dma_inv_range:
blo 1b
mov r0, #0
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* v6_dma_clean_range(start,end)
@@ -262,7 +262,7 @@ v6_dma_clean_range:
blo 1b
mov r0, #0
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* v6_dma_flush_range(start,end)
@@ -290,7 +290,7 @@ ENTRY(v6_dma_flush_range)
blo 1b
mov r0, #0
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -323,7 +323,7 @@ ENTRY(v6_dma_unmap_area)
teq r2, #DMA_TO_DEVICE
bne v6_dma_inv_range
#endif
- mov pc, lr
+ ret lr
ENDPROC(v6_dma_unmap_area)

.globl v6_flush_kern_cache_louis
diff --git a/arch/arm/mm/cache-v7.S b/arch/arm/mm/cache-v7.S
index 615c99e38ba1..b966656d2c2d 100644
--- a/arch/arm/mm/cache-v7.S
+++ b/arch/arm/mm/cache-v7.S
@@ -61,7 +61,7 @@ ENTRY(v7_invalidate_l1)
bgt 1b
dsb st
isb
- mov pc, lr
+ ret lr
ENDPROC(v7_invalidate_l1)

/*
@@ -76,7 +76,7 @@ ENTRY(v7_flush_icache_all)
mov r0, #0
ALT_SMP(mcr p15, 0, r0, c7, c1, 0) @ invalidate I-cache inner shareable
ALT_UP(mcr p15, 0, r0, c7, c5, 0) @ I+BTB cache invalidate
- mov pc, lr
+ ret lr
ENDPROC(v7_flush_icache_all)

/*
@@ -94,7 +94,7 @@ ENTRY(v7_flush_dcache_louis)
ALT_UP(ands r3, r0, #(7 << 27)) @ extract LoUU from clidr
#ifdef CONFIG_ARM_ERRATA_643719
ALT_SMP(mrceq p15, 0, r2, c0, c0, 0) @ read main ID register
- ALT_UP(moveq pc, lr) @ LoUU is zero, so nothing to do
+ ALT_UP(reteq lr) @ LoUU is zero, so nothing to do
ldreq r1, =0x410fc090 @ ID of ARM Cortex A9 r0p?
biceq r2, r2, #0x0000000f @ clear minor revision number
teqeq r2, r1 @ test for errata affected core and if so...
@@ -102,7 +102,7 @@ ENTRY(v7_flush_dcache_louis)
#endif
ALT_SMP(mov r3, r3, lsr #20) @ r3 = LoUIS * 2
ALT_UP(mov r3, r3, lsr #26) @ r3 = LoUU * 2
- moveq pc, lr @ return if level == 0
+ reteq lr @ return if level == 0
mov r10, #0 @ r10 (starting level) = 0
b flush_levels @ start flushing cache levels
ENDPROC(v7_flush_dcache_louis)
@@ -168,7 +168,7 @@ finished:
mcr p15, 2, r10, c0, c0, 0 @ select current cache level in cssr
dsb st
isb
- mov pc, lr
+ ret lr
ENDPROC(v7_flush_dcache_all)

/*
@@ -191,7 +191,7 @@ ENTRY(v7_flush_kern_cache_all)
ALT_UP(mcr p15, 0, r0, c7, c5, 0) @ I+BTB cache invalidate
ARM( ldmfd sp!, {r4-r5, r7, r9-r11, lr} )
THUMB( ldmfd sp!, {r4-r7, r9-r11, lr} )
- mov pc, lr
+ ret lr
ENDPROC(v7_flush_kern_cache_all)

/*
@@ -209,7 +209,7 @@ ENTRY(v7_flush_kern_cache_louis)
ALT_UP(mcr p15, 0, r0, c7, c5, 0) @ I+BTB cache invalidate
ARM( ldmfd sp!, {r4-r5, r7, r9-r11, lr} )
THUMB( ldmfd sp!, {r4-r7, r9-r11, lr} )
- mov pc, lr
+ ret lr
ENDPROC(v7_flush_kern_cache_louis)

/*
@@ -235,7 +235,7 @@ ENTRY(v7_flush_user_cache_all)
* - we have a VIPT cache.
*/
ENTRY(v7_flush_user_cache_range)
- mov pc, lr
+ ret lr
ENDPROC(v7_flush_user_cache_all)
ENDPROC(v7_flush_user_cache_range)

@@ -296,7 +296,7 @@ ENTRY(v7_coherent_user_range)
ALT_UP(mcr p15, 0, r0, c7, c5, 6) @ invalidate BTB
dsb ishst
isb
- mov pc, lr
+ ret lr

/*
* Fault handling for the cache operation above. If the virtual address in r0
@@ -307,7 +307,7 @@ ENTRY(v7_coherent_user_range)
dsb
#endif
mov r0, #-EFAULT
- mov pc, lr
+ ret lr
UNWIND(.fnend )
ENDPROC(v7_coherent_kern_range)
ENDPROC(v7_coherent_user_range)
@@ -336,7 +336,7 @@ ENTRY(v7_flush_kern_dcache_area)
cmp r0, r1
blo 1b
dsb st
- mov pc, lr
+ ret lr
ENDPROC(v7_flush_kern_dcache_area)

/*
@@ -369,7 +369,7 @@ v7_dma_inv_range:
cmp r0, r1
blo 1b
dsb st
- mov pc, lr
+ ret lr
ENDPROC(v7_dma_inv_range)

/*
@@ -391,7 +391,7 @@ v7_dma_clean_range:
cmp r0, r1
blo 1b
dsb st
- mov pc, lr
+ ret lr
ENDPROC(v7_dma_clean_range)

/*
@@ -413,7 +413,7 @@ ENTRY(v7_dma_flush_range)
cmp r0, r1
blo 1b
dsb st
- mov pc, lr
+ ret lr
ENDPROC(v7_dma_flush_range)

/*
@@ -439,7 +439,7 @@ ENTRY(v7_dma_unmap_area)
add r1, r1, r0
teq r2, #DMA_TO_DEVICE
bne v7_dma_inv_range
- mov pc, lr
+ ret lr
ENDPROC(v7_dma_unmap_area)

__INITDATA
diff --git a/arch/arm/mm/l2c-l2x0-resume.S b/arch/arm/mm/l2c-l2x0-resume.S
index 99b05f21a59a..fda415e4ca8f 100644
--- a/arch/arm/mm/l2c-l2x0-resume.S
+++ b/arch/arm/mm/l2c-l2x0-resume.S
@@ -6,6 +6,7 @@
* This code can only be used to if you are running in the secure world.
*/
#include <linux/linkage.h>
+#include <asm/assembler.h>
#include <asm/hardware/cache-l2x0.h>

.text
@@ -27,7 +28,7 @@ ENTRY(l2c310_early_resume)

@ Check that the address has been initialised
teq r1, #0
- moveq pc, lr
+ reteq lr

@ The prefetch and power control registers are revision dependent
@ and can be written whether or not the L2 cache is enabled
@@ -41,7 +42,7 @@ ENTRY(l2c310_early_resume)
@ Don't setup the L2 cache if it is already enabled
ldr r0, [r1, #L2X0_CTRL]
tst r0, #L2X0_CTRL_EN
- movne pc, lr
+ retne lr

str r3, [r1, #L310_TAG_LATENCY_CTRL]
str r4, [r1, #L310_DATA_LATENCY_CTRL]
@@ -51,7 +52,7 @@ ENTRY(l2c310_early_resume)
str r2, [r1, #L2X0_AUX_CTRL]
mov r9, #L2X0_CTRL_EN
str r9, [r1, #L2X0_CTRL]
- mov pc, lr
+ ret lr
ENDPROC(l2c310_early_resume)

.align
diff --git a/arch/arm/mm/proc-arm1020.S b/arch/arm/mm/proc-arm1020.S
index d1a2d05971e0..86ee5d47ce3c 100644
--- a/arch/arm/mm/proc-arm1020.S
+++ b/arch/arm/mm/proc-arm1020.S
@@ -73,7 +73,7 @@
* cpu_arm1020_proc_init()
*/
ENTRY(cpu_arm1020_proc_init)
- mov pc, lr
+ ret lr

/*
* cpu_arm1020_proc_fin()
@@ -83,7 +83,7 @@ ENTRY(cpu_arm1020_proc_fin)
bic r0, r0, #0x1000 @ ...i............
bic r0, r0, #0x000e @ ............wca.
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* cpu_arm1020_reset(loc)
@@ -107,7 +107,7 @@ ENTRY(cpu_arm1020_reset)
bic ip, ip, #0x000f @ ............wcam
bic ip, ip, #0x1100 @ ...i...s........
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm1020_reset)
.popsection

@@ -117,7 +117,7 @@ ENDPROC(cpu_arm1020_reset)
.align 5
ENTRY(cpu_arm1020_do_idle)
mcr p15, 0, r0, c7, c0, 4 @ Wait for interrupt
- mov pc, lr
+ ret lr

/* ================================= CACHE ================================ */

@@ -133,7 +133,7 @@ ENTRY(arm1020_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
#endif
- mov pc, lr
+ ret lr
ENDPROC(arm1020_flush_icache_all)

/*
@@ -169,7 +169,7 @@ __flush_whole_cache:
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
#endif
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -200,7 +200,7 @@ ENTRY(arm1020_flush_user_cache_range)
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
#endif
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -242,7 +242,7 @@ ENTRY(arm1020_coherent_user_range)
blo 1b
mcr p15, 0, ip, c7, c10, 4 @ drain WB
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -264,7 +264,7 @@ ENTRY(arm1020_flush_kern_dcache_area)
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_inv_range(start, end)
@@ -297,7 +297,7 @@ arm1020_dma_inv_range:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -320,7 +320,7 @@ arm1020_dma_clean_range:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -342,7 +342,7 @@ ENTRY(arm1020_dma_flush_range)
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -365,7 +365,7 @@ ENDPROC(arm1020_dma_map_area)
* - dir - DMA direction
*/
ENTRY(arm1020_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(arm1020_dma_unmap_area)

.globl arm1020_flush_kern_cache_louis
@@ -384,7 +384,7 @@ ENTRY(cpu_arm1020_dcache_clean_area)
subs r1, r1, #CACHE_DLINESIZE
bhi 1b
#endif
- mov pc, lr
+ ret lr

/* =============================== PageTable ============================== */

@@ -423,7 +423,7 @@ ENTRY(cpu_arm1020_switch_mm)
mcr p15, 0, r0, c2, c0, 0 @ load page table pointer
mcr p15, 0, r1, c8, c7, 0 @ invalidate I & D TLBs
#endif /* CONFIG_MMU */
- mov pc, lr
+ ret lr

/*
* cpu_arm1020_set_pte(ptep, pte)
@@ -441,7 +441,7 @@ ENTRY(cpu_arm1020_set_pte_ext)
#endif
mcr p15, 0, r0, c7, c10, 4 @ drain WB
#endif /* CONFIG_MMU */
- mov pc, lr
+ ret lr

.type __arm1020_setup, #function
__arm1020_setup:
@@ -460,7 +460,7 @@ __arm1020_setup:
#ifdef CONFIG_CPU_CACHE_ROUND_ROBIN
orr r0, r0, #0x4000 @ .R.. .... .... ....
#endif
- mov pc, lr
+ ret lr
.size __arm1020_setup, . - __arm1020_setup

/*
diff --git a/arch/arm/mm/proc-arm1020e.S b/arch/arm/mm/proc-arm1020e.S
index 9d89405c3d03..a6331d78601f 100644
--- a/arch/arm/mm/proc-arm1020e.S
+++ b/arch/arm/mm/proc-arm1020e.S
@@ -73,7 +73,7 @@
* cpu_arm1020e_proc_init()
*/
ENTRY(cpu_arm1020e_proc_init)
- mov pc, lr
+ ret lr

/*
* cpu_arm1020e_proc_fin()
@@ -83,7 +83,7 @@ ENTRY(cpu_arm1020e_proc_fin)
bic r0, r0, #0x1000 @ ...i............
bic r0, r0, #0x000e @ ............wca.
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* cpu_arm1020e_reset(loc)
@@ -107,7 +107,7 @@ ENTRY(cpu_arm1020e_reset)
bic ip, ip, #0x000f @ ............wcam
bic ip, ip, #0x1100 @ ...i...s........
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm1020e_reset)
.popsection

@@ -117,7 +117,7 @@ ENDPROC(cpu_arm1020e_reset)
.align 5
ENTRY(cpu_arm1020e_do_idle)
mcr p15, 0, r0, c7, c0, 4 @ Wait for interrupt
- mov pc, lr
+ ret lr

/* ================================= CACHE ================================ */

@@ -133,7 +133,7 @@ ENTRY(arm1020e_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
#endif
- mov pc, lr
+ ret lr
ENDPROC(arm1020e_flush_icache_all)

/*
@@ -168,7 +168,7 @@ __flush_whole_cache:
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
#endif
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -197,7 +197,7 @@ ENTRY(arm1020e_flush_user_cache_range)
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
#endif
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -236,7 +236,7 @@ ENTRY(arm1020e_coherent_user_range)
blo 1b
mcr p15, 0, ip, c7, c10, 4 @ drain WB
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -257,7 +257,7 @@ ENTRY(arm1020e_flush_kern_dcache_area)
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_inv_range(start, end)
@@ -286,7 +286,7 @@ arm1020e_dma_inv_range:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -308,7 +308,7 @@ arm1020e_dma_clean_range:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -328,7 +328,7 @@ ENTRY(arm1020e_dma_flush_range)
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -351,7 +351,7 @@ ENDPROC(arm1020e_dma_map_area)
* - dir - DMA direction
*/
ENTRY(arm1020e_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(arm1020e_dma_unmap_area)

.globl arm1020e_flush_kern_cache_louis
@@ -369,7 +369,7 @@ ENTRY(cpu_arm1020e_dcache_clean_area)
subs r1, r1, #CACHE_DLINESIZE
bhi 1b
#endif
- mov pc, lr
+ ret lr

/* =============================== PageTable ============================== */

@@ -407,7 +407,7 @@ ENTRY(cpu_arm1020e_switch_mm)
mcr p15, 0, r0, c2, c0, 0 @ load page table pointer
mcr p15, 0, r1, c8, c7, 0 @ invalidate I & D TLBs
#endif
- mov pc, lr
+ ret lr

/*
* cpu_arm1020e_set_pte(ptep, pte)
@@ -423,7 +423,7 @@ ENTRY(cpu_arm1020e_set_pte_ext)
mcr p15, 0, r0, c7, c10, 1 @ clean D entry
#endif
#endif /* CONFIG_MMU */
- mov pc, lr
+ ret lr

.type __arm1020e_setup, #function
__arm1020e_setup:
@@ -441,7 +441,7 @@ __arm1020e_setup:
#ifdef CONFIG_CPU_CACHE_ROUND_ROBIN
orr r0, r0, #0x4000 @ .R.. .... .... ....
#endif
- mov pc, lr
+ ret lr
.size __arm1020e_setup, . - __arm1020e_setup

/*
diff --git a/arch/arm/mm/proc-arm1022.S b/arch/arm/mm/proc-arm1022.S
index 6f01a0ae3b30..a126b7a59928 100644
--- a/arch/arm/mm/proc-arm1022.S
+++ b/arch/arm/mm/proc-arm1022.S
@@ -62,7 +62,7 @@
* cpu_arm1022_proc_init()
*/
ENTRY(cpu_arm1022_proc_init)
- mov pc, lr
+ ret lr

/*
* cpu_arm1022_proc_fin()
@@ -72,7 +72,7 @@ ENTRY(cpu_arm1022_proc_fin)
bic r0, r0, #0x1000 @ ...i............
bic r0, r0, #0x000e @ ............wca.
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* cpu_arm1022_reset(loc)
@@ -96,7 +96,7 @@ ENTRY(cpu_arm1022_reset)
bic ip, ip, #0x000f @ ............wcam
bic ip, ip, #0x1100 @ ...i...s........
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm1022_reset)
.popsection

@@ -106,7 +106,7 @@ ENDPROC(cpu_arm1022_reset)
.align 5
ENTRY(cpu_arm1022_do_idle)
mcr p15, 0, r0, c7, c0, 4 @ Wait for interrupt
- mov pc, lr
+ ret lr

/* ================================= CACHE ================================ */

@@ -122,7 +122,7 @@ ENTRY(arm1022_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
#endif
- mov pc, lr
+ ret lr
ENDPROC(arm1022_flush_icache_all)

/*
@@ -156,7 +156,7 @@ __flush_whole_cache:
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
#endif
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -185,7 +185,7 @@ ENTRY(arm1022_flush_user_cache_range)
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
#endif
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -225,7 +225,7 @@ ENTRY(arm1022_coherent_user_range)
blo 1b
mcr p15, 0, ip, c7, c10, 4 @ drain WB
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -246,7 +246,7 @@ ENTRY(arm1022_flush_kern_dcache_area)
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_inv_range(start, end)
@@ -275,7 +275,7 @@ arm1022_dma_inv_range:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -297,7 +297,7 @@ arm1022_dma_clean_range:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -317,7 +317,7 @@ ENTRY(arm1022_dma_flush_range)
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -340,7 +340,7 @@ ENDPROC(arm1022_dma_map_area)
* - dir - DMA direction
*/
ENTRY(arm1022_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(arm1022_dma_unmap_area)

.globl arm1022_flush_kern_cache_louis
@@ -358,7 +358,7 @@ ENTRY(cpu_arm1022_dcache_clean_area)
subs r1, r1, #CACHE_DLINESIZE
bhi 1b
#endif
- mov pc, lr
+ ret lr

/* =============================== PageTable ============================== */

@@ -389,7 +389,7 @@ ENTRY(cpu_arm1022_switch_mm)
mcr p15, 0, r0, c2, c0, 0 @ load page table pointer
mcr p15, 0, r1, c8, c7, 0 @ invalidate I & D TLBs
#endif
- mov pc, lr
+ ret lr

/*
* cpu_arm1022_set_pte_ext(ptep, pte, ext)
@@ -405,7 +405,7 @@ ENTRY(cpu_arm1022_set_pte_ext)
mcr p15, 0, r0, c7, c10, 1 @ clean D entry
#endif
#endif /* CONFIG_MMU */
- mov pc, lr
+ ret lr

.type __arm1022_setup, #function
__arm1022_setup:
@@ -423,7 +423,7 @@ __arm1022_setup:
#ifdef CONFIG_CPU_CACHE_ROUND_ROBIN
orr r0, r0, #0x4000 @ .R..............
#endif
- mov pc, lr
+ ret lr
.size __arm1022_setup, . - __arm1022_setup

/*
diff --git a/arch/arm/mm/proc-arm1026.S b/arch/arm/mm/proc-arm1026.S
index 4799a24b43e6..fc294067e977 100644
--- a/arch/arm/mm/proc-arm1026.S
+++ b/arch/arm/mm/proc-arm1026.S
@@ -62,7 +62,7 @@
* cpu_arm1026_proc_init()
*/
ENTRY(cpu_arm1026_proc_init)
- mov pc, lr
+ ret lr

/*
* cpu_arm1026_proc_fin()
@@ -72,7 +72,7 @@ ENTRY(cpu_arm1026_proc_fin)
bic r0, r0, #0x1000 @ ...i............
bic r0, r0, #0x000e @ ............wca.
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* cpu_arm1026_reset(loc)
@@ -96,7 +96,7 @@ ENTRY(cpu_arm1026_reset)
bic ip, ip, #0x000f @ ............wcam
bic ip, ip, #0x1100 @ ...i...s........
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm1026_reset)
.popsection

@@ -106,7 +106,7 @@ ENDPROC(cpu_arm1026_reset)
.align 5
ENTRY(cpu_arm1026_do_idle)
mcr p15, 0, r0, c7, c0, 4 @ Wait for interrupt
- mov pc, lr
+ ret lr

/* ================================= CACHE ================================ */

@@ -122,7 +122,7 @@ ENTRY(arm1026_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
#endif
- mov pc, lr
+ ret lr
ENDPROC(arm1026_flush_icache_all)

/*
@@ -151,7 +151,7 @@ __flush_whole_cache:
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
#endif
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -180,7 +180,7 @@ ENTRY(arm1026_flush_user_cache_range)
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
#endif
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -219,7 +219,7 @@ ENTRY(arm1026_coherent_user_range)
blo 1b
mcr p15, 0, ip, c7, c10, 4 @ drain WB
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -240,7 +240,7 @@ ENTRY(arm1026_flush_kern_dcache_area)
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_inv_range(start, end)
@@ -269,7 +269,7 @@ arm1026_dma_inv_range:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -291,7 +291,7 @@ arm1026_dma_clean_range:
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -311,7 +311,7 @@ ENTRY(arm1026_dma_flush_range)
blo 1b
#endif
mcr p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -334,7 +334,7 @@ ENDPROC(arm1026_dma_map_area)
* - dir - DMA direction
*/
ENTRY(arm1026_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(arm1026_dma_unmap_area)

.globl arm1026_flush_kern_cache_louis
@@ -352,7 +352,7 @@ ENTRY(cpu_arm1026_dcache_clean_area)
subs r1, r1, #CACHE_DLINESIZE
bhi 1b
#endif
- mov pc, lr
+ ret lr

/* =============================== PageTable ============================== */

@@ -378,7 +378,7 @@ ENTRY(cpu_arm1026_switch_mm)
mcr p15, 0, r0, c2, c0, 0 @ load page table pointer
mcr p15, 0, r1, c8, c7, 0 @ invalidate I & D TLBs
#endif
- mov pc, lr
+ ret lr

/*
* cpu_arm1026_set_pte_ext(ptep, pte, ext)
@@ -394,7 +394,7 @@ ENTRY(cpu_arm1026_set_pte_ext)
mcr p15, 0, r0, c7, c10, 1 @ clean D entry
#endif
#endif /* CONFIG_MMU */
- mov pc, lr
+ ret lr

.type __arm1026_setup, #function
__arm1026_setup:
@@ -417,7 +417,7 @@ __arm1026_setup:
#ifdef CONFIG_CPU_CACHE_ROUND_ROBIN
orr r0, r0, #0x4000 @ .R.. .... .... ....
#endif
- mov pc, lr
+ ret lr
.size __arm1026_setup, . - __arm1026_setup

/*
diff --git a/arch/arm/mm/proc-arm720.S b/arch/arm/mm/proc-arm720.S
index d42c37f9f5bc..2baa66b3ac9b 100644
--- a/arch/arm/mm/proc-arm720.S
+++ b/arch/arm/mm/proc-arm720.S
@@ -51,14 +51,14 @@
*/
ENTRY(cpu_arm720_dcache_clean_area)
ENTRY(cpu_arm720_proc_init)
- mov pc, lr
+ ret lr

ENTRY(cpu_arm720_proc_fin)
mrc p15, 0, r0, c1, c0, 0
bic r0, r0, #0x1000 @ ...i............
bic r0, r0, #0x000e @ ............wca.
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* Function: arm720_proc_do_idle(void)
@@ -66,7 +66,7 @@ ENTRY(cpu_arm720_proc_fin)
* Purpose : put the processor in proper idle mode
*/
ENTRY(cpu_arm720_do_idle)
- mov pc, lr
+ ret lr

/*
* Function: arm720_switch_mm(unsigned long pgd_phys)
@@ -81,7 +81,7 @@ ENTRY(cpu_arm720_switch_mm)
mcr p15, 0, r0, c2, c0, 0 @ update page table ptr
mcr p15, 0, r1, c8, c7, 0 @ flush TLB (v4)
#endif
- mov pc, lr
+ ret lr

/*
* Function: arm720_set_pte_ext(pte_t *ptep, pte_t pte, unsigned int ext)
@@ -94,7 +94,7 @@ ENTRY(cpu_arm720_set_pte_ext)
#ifdef CONFIG_MMU
armv3_set_pte_ext wc_disable=0
#endif
- mov pc, lr
+ ret lr

/*
* Function: arm720_reset
@@ -112,7 +112,7 @@ ENTRY(cpu_arm720_reset)
bic ip, ip, #0x000f @ ............wcam
bic ip, ip, #0x2100 @ ..v....s........
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm720_reset)
.popsection

@@ -128,7 +128,7 @@ __arm710_setup:
bic r0, r0, r5
ldr r5, arm710_cr1_set
orr r0, r0, r5
- mov pc, lr @ __ret (head.S)
+ ret lr @ __ret (head.S)
.size __arm710_setup, . - __arm710_setup

/*
@@ -156,7 +156,7 @@ __arm720_setup:
mrc p15, 0, r0, c1, c0 @ get control register
bic r0, r0, r5
orr r0, r0, r6
- mov pc, lr @ __ret (head.S)
+ ret lr @ __ret (head.S)
.size __arm720_setup, . - __arm720_setup

/*
diff --git a/arch/arm/mm/proc-arm740.S b/arch/arm/mm/proc-arm740.S
index 9b0ae90cbf17..ac1ea6b3bce4 100644
--- a/arch/arm/mm/proc-arm740.S
+++ b/arch/arm/mm/proc-arm740.S
@@ -32,7 +32,7 @@ ENTRY(cpu_arm740_proc_init)
ENTRY(cpu_arm740_do_idle)
ENTRY(cpu_arm740_dcache_clean_area)
ENTRY(cpu_arm740_switch_mm)
- mov pc, lr
+ ret lr

/*
* cpu_arm740_proc_fin()
@@ -42,7 +42,7 @@ ENTRY(cpu_arm740_proc_fin)
bic r0, r0, #0x3f000000 @ bank/f/lock/s
bic r0, r0, #0x0000000c @ w-buffer/cache
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* cpu_arm740_reset(loc)
@@ -56,7 +56,7 @@ ENTRY(cpu_arm740_reset)
mrc p15, 0, ip, c1, c0, 0 @ get ctrl register
bic ip, ip, #0x0000000c @ ............wc..
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm740_reset)
.popsection

@@ -115,7 +115,7 @@ __arm740_setup:
@ need some benchmark
orr r0, r0, #0x0000000d @ MPU/Cache/WB

- mov pc, lr
+ ret lr

.size __arm740_setup, . - __arm740_setup

diff --git a/arch/arm/mm/proc-arm7tdmi.S b/arch/arm/mm/proc-arm7tdmi.S
index f6cc3f63ce39..bf6ba4bc30ff 100644
--- a/arch/arm/mm/proc-arm7tdmi.S
+++ b/arch/arm/mm/proc-arm7tdmi.S
@@ -32,13 +32,13 @@ ENTRY(cpu_arm7tdmi_proc_init)
ENTRY(cpu_arm7tdmi_do_idle)
ENTRY(cpu_arm7tdmi_dcache_clean_area)
ENTRY(cpu_arm7tdmi_switch_mm)
- mov pc, lr
+ ret lr

/*
* cpu_arm7tdmi_proc_fin()
*/
ENTRY(cpu_arm7tdmi_proc_fin)
- mov pc, lr
+ ret lr

/*
* Function: cpu_arm7tdmi_reset(loc)
@@ -47,13 +47,13 @@ ENTRY(cpu_arm7tdmi_proc_fin)
*/
.pushsection .idmap.text, "ax"
ENTRY(cpu_arm7tdmi_reset)
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm7tdmi_reset)
.popsection

.type __arm7tdmi_setup, #function
__arm7tdmi_setup:
- mov pc, lr
+ ret lr
.size __arm7tdmi_setup, . - __arm7tdmi_setup

__INITDATA
diff --git a/arch/arm/mm/proc-arm920.S b/arch/arm/mm/proc-arm920.S
index 549557df6d57..22bf8dde4f84 100644
--- a/arch/arm/mm/proc-arm920.S
+++ b/arch/arm/mm/proc-arm920.S
@@ -63,7 +63,7 @@
* cpu_arm920_proc_init()
*/
ENTRY(cpu_arm920_proc_init)
- mov pc, lr
+ ret lr

/*
* cpu_arm920_proc_fin()
@@ -73,7 +73,7 @@ ENTRY(cpu_arm920_proc_fin)
bic r0, r0, #0x1000 @ ...i............
bic r0, r0, #0x000e @ ............wca.
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* cpu_arm920_reset(loc)
@@ -97,7 +97,7 @@ ENTRY(cpu_arm920_reset)
bic ip, ip, #0x000f @ ............wcam
bic ip, ip, #0x1100 @ ...i...s........
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm920_reset)
.popsection

@@ -107,7 +107,7 @@ ENDPROC(cpu_arm920_reset)
.align 5
ENTRY(cpu_arm920_do_idle)
mcr p15, 0, r0, c7, c0, 4 @ Wait for interrupt
- mov pc, lr
+ ret lr


#ifndef CONFIG_CPU_DCACHE_WRITETHROUGH
@@ -120,7 +120,7 @@ ENTRY(cpu_arm920_do_idle)
ENTRY(arm920_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
- mov pc, lr
+ ret lr
ENDPROC(arm920_flush_icache_all)

/*
@@ -151,7 +151,7 @@ __flush_whole_cache:
tst r2, #VM_EXEC
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -177,7 +177,7 @@ ENTRY(arm920_flush_user_cache_range)
blo 1b
tst r2, #VM_EXEC
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -211,7 +211,7 @@ ENTRY(arm920_coherent_user_range)
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -231,7 +231,7 @@ ENTRY(arm920_flush_kern_dcache_area)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_inv_range(start, end)
@@ -257,7 +257,7 @@ arm920_dma_inv_range:
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -276,7 +276,7 @@ arm920_dma_clean_range:
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -293,7 +293,7 @@ ENTRY(arm920_dma_flush_range)
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -316,7 +316,7 @@ ENDPROC(arm920_dma_map_area)
* - dir - DMA direction
*/
ENTRY(arm920_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(arm920_dma_unmap_area)

.globl arm920_flush_kern_cache_louis
@@ -332,7 +332,7 @@ ENTRY(cpu_arm920_dcache_clean_area)
add r0, r0, #CACHE_DLINESIZE
subs r1, r1, #CACHE_DLINESIZE
bhi 1b
- mov pc, lr
+ ret lr

/* =============================== PageTable ============================== */

@@ -367,7 +367,7 @@ ENTRY(cpu_arm920_switch_mm)
mcr p15, 0, r0, c2, c0, 0 @ load page table pointer
mcr p15, 0, ip, c8, c7, 0 @ invalidate I & D TLBs
#endif
- mov pc, lr
+ ret lr

/*
* cpu_arm920_set_pte(ptep, pte, ext)
@@ -382,7 +382,7 @@ ENTRY(cpu_arm920_set_pte_ext)
mcr p15, 0, r0, c7, c10, 1 @ clean D entry
mcr p15, 0, r0, c7, c10, 4 @ drain WB
#endif
- mov pc, lr
+ ret lr

/* Suspend/resume support: taken from arch/arm/plat-s3c24xx/sleep.S */
.globl cpu_arm920_suspend_size
@@ -423,7 +423,7 @@ __arm920_setup:
mrc p15, 0, r0, c1, c0 @ get control register v4
bic r0, r0, r5
orr r0, r0, r6
- mov pc, lr
+ ret lr
.size __arm920_setup, . - __arm920_setup

/*
diff --git a/arch/arm/mm/proc-arm922.S b/arch/arm/mm/proc-arm922.S
index 2a758b06c6f6..0c6d5ac5a6d4 100644
--- a/arch/arm/mm/proc-arm922.S
+++ b/arch/arm/mm/proc-arm922.S
@@ -65,7 +65,7 @@
* cpu_arm922_proc_init()
*/
ENTRY(cpu_arm922_proc_init)
- mov pc, lr
+ ret lr

/*
* cpu_arm922_proc_fin()
@@ -75,7 +75,7 @@ ENTRY(cpu_arm922_proc_fin)
bic r0, r0, #0x1000 @ ...i............
bic r0, r0, #0x000e @ ............wca.
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* cpu_arm922_reset(loc)
@@ -99,7 +99,7 @@ ENTRY(cpu_arm922_reset)
bic ip, ip, #0x000f @ ............wcam
bic ip, ip, #0x1100 @ ...i...s........
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0
ENDPROC(cpu_arm922_reset)
.popsection

@@ -109,7 +109,7 @@ ENDPROC(cpu_arm922_reset)
.align 5
ENTRY(cpu_arm922_do_idle)
mcr p15, 0, r0, c7, c0, 4 @ Wait for interrupt
- mov pc, lr
+ ret lr


#ifndef CONFIG_CPU_DCACHE_WRITETHROUGH
@@ -122,7 +122,7 @@ ENTRY(cpu_arm922_do_idle)
ENTRY(arm922_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
- mov pc, lr
+ ret lr
ENDPROC(arm922_flush_icache_all)

/*
@@ -153,7 +153,7 @@ __flush_whole_cache:
tst r2, #VM_EXEC
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -179,7 +179,7 @@ ENTRY(arm922_flush_user_cache_range)
blo 1b
tst r2, #VM_EXEC
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -213,7 +213,7 @@ ENTRY(arm922_coherent_user_range)
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -233,7 +233,7 @@ ENTRY(arm922_flush_kern_dcache_area)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_inv_range(start, end)
@@ -259,7 +259,7 @@ arm922_dma_inv_range:
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -278,7 +278,7 @@ arm922_dma_clean_range:
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -295,7 +295,7 @@ ENTRY(arm922_dma_flush_range)
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -318,7 +318,7 @@ ENDPROC(arm922_dma_map_area)
* - dir - DMA direction
*/
ENTRY(arm922_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(arm922_dma_unmap_area)

.globl arm922_flush_kern_cache_louis
@@ -336,7 +336,7 @@ ENTRY(cpu_arm922_dcache_clean_area)
subs r1, r1, #CACHE_DLINESIZE
bhi 1b
#endif
- mov pc, lr
+ ret lr

/* =============================== PageTable ============================== */

@@ -371,7 +371,7 @@ ENTRY(cpu_arm922_switch_mm)
mcr p15, 0, r0, c2, c0, 0 @ load page table pointer
mcr p15, 0, ip, c8, c7, 0 @ invalidate I & D TLBs
#endif
- mov pc, lr
+ ret lr

/*
* cpu_arm922_set_pte_ext(ptep, pte, ext)
@@ -386,7 +386,7 @@ ENTRY(cpu_arm922_set_pte_ext)
mcr p15, 0, r0, c7, c10, 1 @ clean D entry
mcr p15, 0, r0, c7, c10, 4 @ drain WB
#endif /* CONFIG_MMU */
- mov pc, lr
+ ret lr

.type __arm922_setup, #function
__arm922_setup:
@@ -401,7 +401,7 @@ __arm922_setup:
mrc p15, 0, r0, c1, c0 @ get control register v4
bic r0, r0, r5
orr r0, r0, r6
- mov pc, lr
+ ret lr
.size __arm922_setup, . - __arm922_setup

/*
diff --git a/arch/arm/mm/proc-arm925.S b/arch/arm/mm/proc-arm925.S
index ba0d58e1a2a2..c32d073282ea 100644
--- a/arch/arm/mm/proc-arm925.S
+++ b/arch/arm/mm/proc-arm925.S
@@ -86,7 +86,7 @@
* cpu_arm925_proc_init()
*/
ENTRY(cpu_arm925_proc_init)
- mov pc, lr
+ ret lr

/*
* cpu_arm925_proc_fin()
@@ -96,7 +96,7 @@ ENTRY(cpu_arm925_proc_fin)
bic r0, r0, #0x1000 @ ...i............
bic r0, r0, #0x000e @ ............wca.
mcr p15, 0, r0, c1, c0, 0 @ disable caches
- mov pc, lr
+ ret lr

/*
* cpu_arm925_reset(loc)
@@ -129,7 +129,7 @@ ENDPROC(cpu_arm925_reset)
bic ip, ip, #0x000f @ ............wcam
bic ip, ip, #0x1100 @ ...i...s........
mcr p15, 0, ip, c1, c0, 0 @ ctrl register
- mov pc, r0
+ ret r0

/*
* cpu_arm925_do_idle()
@@ -145,7 +145,7 @@ ENTRY(cpu_arm925_do_idle)
mcr p15, 0, r2, c1, c0, 0 @ Disable I cache
mcr p15, 0, r0, c7, c0, 4 @ Wait for interrupt
mcr p15, 0, r1, c1, c0, 0 @ Restore ICache enable
- mov pc, lr
+ ret lr

/*
* flush_icache_all()
@@ -155,7 +155,7 @@ ENTRY(cpu_arm925_do_idle)
ENTRY(arm925_flush_icache_all)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
- mov pc, lr
+ ret lr
ENDPROC(arm925_flush_icache_all)

/*
@@ -188,7 +188,7 @@ __flush_whole_cache:
tst r2, #VM_EXEC
mcrne p15, 0, ip, c7, c5, 0 @ invalidate I cache
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* flush_user_cache_range(start, end, flags)
@@ -225,7 +225,7 @@ ENTRY(arm925_flush_user_cache_range)
blo 1b
tst r2, #VM_EXEC
mcrne p15, 0, ip, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* coherent_kern_range(start, end)
@@ -259,7 +259,7 @@ ENTRY(arm925_coherent_user_range)
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
mov r0, #0
- mov pc, lr
+ ret lr

/*
* flush_kern_dcache_area(void *addr, size_t size)
@@ -279,7 +279,7 @@ ENTRY(arm925_flush_kern_dcache_area)
mov r0, #0
mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_inv_range(start, end)
@@ -307,7 +307,7 @@ arm925_dma_inv_range:
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_clean_range(start, end)
@@ -328,7 +328,7 @@ arm925_dma_clean_range:
blo 1b
#endif
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_flush_range(start, end)
@@ -350,7 +350,7 @@ ENTRY(arm925_dma_flush_range)
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/*
* dma_map_area(start, size, dir)
@@ -373,7 +373,7 @@ ENDPROC(arm925_dma_map_area)
* - dir - DMA direction
*/
ENTRY(arm925_dma_unmap_area)
- mov pc, lr
+ ret lr
ENDPROC(arm925_dma_unmap_area)

.globl arm925_flush_kern_cache_louis
@@ -390,7 +390,7 @@ ENTRY(cpu_arm925_dcache_clean_area)
bhi 1b
#endif
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr

/* =============================== PageTable ============================== */

@@ -419,7 +419,7 @@ ENTRY(cpu_arm925_switch_mm)
mcr p15, 0, r0, c2, c0, 0 @ load page table pointer
mcr p15, 0, ip, c8, c7, 0 @ invalidate I & D TLBs
#endif
- mov pc, lr
+ ret lr

/*
* cpu_arm925_set_pte_ext(ptep, pte, ext)
@@ -436,7 +436,7 @@ ENTRY(cpu_arm925_set_pte_ext)
#endif
mcr p15, 0, r0, c7, c10, 4 @ drain WB
#endif /* CONFIG_MMU */
- mov pc, lr
+ ret lr

.type __arm925_setup, #function
__arm925_setup:
@@ -469,7 +469,7 @@ __arm925_setup:
#ifdef CONFIG_CPU_CACHE_ROUND_ROBIN
orr r0, r0, #0x4000 @ .1.. .... .... ....
#endif
- mov pc, lr
+ ret lr
.size __arm925_setup, . - __arm925_setup

/*
diff --git a/arch/arm/mm/proc-arm926.S b/arch/arm/mm/proc-arm926.S
index 0f098f407c9f..252b2503038d 100644

2014-07-01 16:42:42

by Måns Rullgård

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

Russell King <rmk+kernel-lFZ/[email protected]> writes:

> ARMv6 and greater introduced a new instruction ("bx") which can be used
> to return from function calls. Recent CPUs perform better when the
> "bx lr" instruction is used rather than the "mov pc, lr" instruction,
> and this sequence is strongly recommended to be used by the ARM
> architecture manual (section A.4.1.1).
>
> We provide a new macro "ret" with all its variants for the condition
> code which will resolve to the appropriate instruction.

When the source register is not "lr" the name "ret" is a misnomer since
only the "bx lr" instruction is predicted as a function return. The
"bx" instruction with other source registers uses the normal prediction
mechanisms, leaving the return stack alone, and should not be used for
function returns. Any code currently using another register to return
from a function should probably be modified to use lr instead, unless
there are special reasons for doing otherwise. If code jumping to an
address in a non-lr register is not a return, using the "ret" name will
make for some rather confusing reading.

I would suggest either using a more neutral name than "ret" or adding an
alias to be used for non-return jumps so as to make the intent clearer.

--
M?ns Rullg?rd
[email protected]

2014-07-01 17:13:54

by Russell King - ARM Linux

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

On Tue, Jul 01, 2014 at 05:42:42PM +0100, M?ns Rullg?rd wrote:
> Russell King <[email protected]> writes:
>
> > ARMv6 and greater introduced a new instruction ("bx") which can be used
> > to return from function calls. Recent CPUs perform better when the
> > "bx lr" instruction is used rather than the "mov pc, lr" instruction,
> > and this sequence is strongly recommended to be used by the ARM
> > architecture manual (section A.4.1.1).
> >
> > We provide a new macro "ret" with all its variants for the condition
> > code which will resolve to the appropriate instruction.
>
> When the source register is not "lr" the name "ret" is a misnomer since
> only the "bx lr" instruction is predicted as a function return. The
> "bx" instruction with other source registers uses the normal prediction
> mechanisms, leaving the return stack alone, and should not be used for
> function returns. Any code currently using another register to return
> from a function should probably be modified to use lr instead, unless
> there are special reasons for doing otherwise. If code jumping to an
> address in a non-lr register is not a return, using the "ret" name will
> make for some rather confusing reading.
>
> I would suggest either using a more neutral name than "ret" or adding an
> alias to be used for non-return jumps so as to make the intent clearer.

If you read the patch, the branches which are changed are those which
do indeed return in some way. There are those which do this having
moved lr to a different register.

As you point out, "bx lr" /may/ be treated specially (I've actually been
discussing this with Will Deacon over the last couple of days, who has
also been talking to the hardware people in ARM, and Will is happy with
this patch as in its current form.) This is why I've changed all
"mov pc, reg" instructions which return in some way to use this macro,
and left others (those which are used to call some function and return
back to the same point) alone.

I have thought about introducing a "call" macro for those other sites,
but as there are soo few of them, I've left them as-is for the time
being (this patch is already large enough.)

If there are any in the patch which you have specific concerns about,
please specifically point them out those which give you a concern.

Thanks.

--
FTTC broadband for 0.8mile line: now at 9.7Mbps down 460kbps up... slowly
improving, and getting towards what was expected from it.

2014-07-01 17:24:43

by Måns Rullgård

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

Russell King - ARM Linux <[email protected]> writes:

> On Tue, Jul 01, 2014 at 05:42:42PM +0100, M?ns Rullg?rd wrote:
>> Russell King <[email protected]> writes:
>>
>> > ARMv6 and greater introduced a new instruction ("bx") which can be used
>> > to return from function calls. Recent CPUs perform better when the
>> > "bx lr" instruction is used rather than the "mov pc, lr" instruction,
>> > and this sequence is strongly recommended to be used by the ARM
>> > architecture manual (section A.4.1.1).
>> >
>> > We provide a new macro "ret" with all its variants for the condition
>> > code which will resolve to the appropriate instruction.
>>
>> When the source register is not "lr" the name "ret" is a misnomer since
>> only the "bx lr" instruction is predicted as a function return. The
>> "bx" instruction with other source registers uses the normal prediction
>> mechanisms, leaving the return stack alone, and should not be used for
>> function returns. Any code currently using another register to return
>> from a function should probably be modified to use lr instead, unless
>> there are special reasons for doing otherwise. If code jumping to an
>> address in a non-lr register is not a return, using the "ret" name will
>> make for some rather confusing reading.
>>
>> I would suggest either using a more neutral name than "ret" or adding an
>> alias to be used for non-return jumps so as to make the intent clearer.
>
> If you read the patch, the branches which are changed are those which
> do indeed return in some way. There are those which do this having
> moved lr to a different register.

The patch is huge, and it wasn't obvious (to me) from the diff context
what the non-lr jumps were doing.

> As you point out, "bx lr" /may/ be treated specially (I've actually been

Most, if not all, Cortex-A cores do this according the public TRMs.
They also do the same thing for "mov pc, lr" so there will probably be
no performance gain from this change. It's still a good idea though,
since we don't know what future cores will do.

> discussing this with Will Deacon over the last couple of days, who has
> also been talking to the hardware people in ARM, and Will is happy with
> this patch as in its current form.) This is why I've changed all
> "mov pc, reg" instructions which return in some way to use this macro,
> and left others (those which are used to call some function and return
> back to the same point) alone.

In that case the patch should be fine. Your patch description didn't
make it clear that only actual returns were being changed.

--
M?ns Rullg?rd
[email protected]

2014-07-01 17:36:22

by Will Deacon

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

Hi Mans,

On Tue, Jul 01, 2014 at 06:24:43PM +0100, M?ns Rullg?rd wrote:
> Russell King - ARM Linux <[email protected]> writes:
> > As you point out, "bx lr" /may/ be treated specially (I've actually been
>
> Most, if not all, Cortex-A cores do this according the public TRMs.
> They also do the same thing for "mov pc, lr" so there will probably be
> no performance gain from this change. It's still a good idea though,
> since we don't know what future cores will do.

Funnily enough, that's not actually true (and is more or less what prompted
this patch after discussion with Russell). There are cores out there that
don't predict mov pc, lr at all (let alone do anything with the return
stack).

> > discussing this with Will Deacon over the last couple of days, who has
> > also been talking to the hardware people in ARM, and Will is happy with
> > this patch as in its current form.) This is why I've changed all
> > "mov pc, reg" instructions which return in some way to use this macro,
> > and left others (those which are used to call some function and return
> > back to the same point) alone.
>
> In that case the patch should be fine. Your patch description didn't
> make it clear that only actual returns were being changed.

I'm led to believe that some predictors require lr in order to update the
return stack, whilst others don't. That part is all horribly
micro-architectural, so the current patch is doing the right thing by
sticking to the ARM ARM but enabling us to hook into other registers later
on if we choose.

Cheers,

Will

2014-07-01 19:26:18

by Stephen Warren

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

On 07/01/2014 10:19 AM, Russell King wrote:
> ARMv6 and greater introduced a new instruction ("bx") which can be used
> to return from function calls. Recent CPUs perform better when the
> "bx lr" instruction is used rather than the "mov pc, lr" instruction,
> and this sequence is strongly recommended to be used by the ARM
> architecture manual (section A.4.1.1).
>
> We provide a new macro "ret" with all its variants for the condition
> code which will resolve to the appropriate instruction.
>
> Rather than doing this piecemeal, and miss some instances, change all
> the "mov pc" instances to use the new macro, with the exception of
> the "movs" instruction and the kprobes code. This allows us to detect
> the "mov pc, lr" case and fix it up - and also gives us the possibility
> of deploying this for other registers depending on the CPU selection.

Tested-by: Stephen Warren <[email protected]>

(On an NVIDIA Tegra Jetson TK1 board, both CPU hotplug and system sleep
were tested, which are the use-cases that actually use the edited
assembly files)

2014-07-01 20:04:25

by Måns Rullgård

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

Will Deacon <[email protected]> writes:

> Hi Mans,
>
> On Tue, Jul 01, 2014 at 06:24:43PM +0100, M?ns Rullg?rd wrote:
>> Russell King - ARM Linux <[email protected]> writes:
>> > As you point out, "bx lr" /may/ be treated specially (I've actually been
>>
>> Most, if not all, Cortex-A cores do this according the public TRMs.
>> They also do the same thing for "mov pc, lr" so there will probably be
>> no performance gain from this change. It's still a good idea though,
>> since we don't know what future cores will do.
>
> Funnily enough, that's not actually true (and is more or less what prompted
> this patch after discussion with Russell). There are cores out there that
> don't predict mov pc, lr at all (let alone do anything with the return
> stack).

Even ones where the TRM says they do? I suppose the only way to know
for sure is to measure it.

>> > discussing this with Will Deacon over the last couple of days, who has
>> > also been talking to the hardware people in ARM, and Will is happy with
>> > this patch as in its current form.) This is why I've changed all
>> > "mov pc, reg" instructions which return in some way to use this macro,
>> > and left others (those which are used to call some function and return
>> > back to the same point) alone.
>>
>> In that case the patch should be fine. Your patch description didn't
>> make it clear that only actual returns were being changed.
>
> I'm led to believe that some predictors require lr in order to update the
> return stack, whilst others don't. That part is all horribly
> micro-architectural, so the current patch is doing the right thing by
> sticking to the ARM ARM but enabling us to hook into other registers later
> on if we choose.

Agreed.

--
M?ns Rullg?rd
[email protected]

2014-07-01 22:42:06

by Robert Jarzmik

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

Russell King <[email protected]> writes:

> ARMv6 and greater introduced a new instruction ("bx") which can be used
> to return from function calls. Recent CPUs perform better when the
> "bx lr" instruction is used rather than the "mov pc, lr" instruction,
> and this sequence is strongly recommended to be used by the ARM
> architecture manual (section A.4.1.1).

For ARMv5 - XScale, and more specificaly the mioa701_bootresume.S path (suspend
to RAM) :
Tested-by: Robert Jarzmik <[email protected]>

Cheers.

--
Robert

2014-07-02 03:17:25

by Nicolas Pitre

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

On Tue, 1 Jul 2014, Will Deacon wrote:

> Hi Mans,
>
> On Tue, Jul 01, 2014 at 06:24:43PM +0100, M?ns Rullg?rd wrote:
> > Russell King - ARM Linux <[email protected]> writes:
> > > As you point out, "bx lr" /may/ be treated specially (I've actually been
> >
> > Most, if not all, Cortex-A cores do this according the public TRMs.
> > They also do the same thing for "mov pc, lr" so there will probably be
> > no performance gain from this change. It's still a good idea though,
> > since we don't know what future cores will do.
>
> Funnily enough, that's not actually true (and is more or less what prompted
> this patch after discussion with Russell). There are cores out there that
> don't predict mov pc, lr at all (let alone do anything with the return
> stack).
>
> > > discussing this with Will Deacon over the last couple of days, who has
> > > also been talking to the hardware people in ARM, and Will is happy with
> > > this patch as in its current form.) This is why I've changed all
> > > "mov pc, reg" instructions which return in some way to use this macro,
> > > and left others (those which are used to call some function and return
> > > back to the same point) alone.
> >
> > In that case the patch should be fine. Your patch description didn't
> > make it clear that only actual returns were being changed.
>
> I'm led to believe that some predictors require lr in order to update the
> return stack, whilst others don't. That part is all horribly
> micro-architectural, so the current patch is doing the right thing by
> sticking to the ARM ARM but enabling us to hook into other registers later
> on if we choose.

May I suggest to have a patch with only the macro definition in it and
all this discussion in the commit log please? The usage sites should be
done in a separate patch to make it clearer.


Nicolas

2014-07-05 01:06:39

by Tomasz Figa

Subject: Re: [PATCH] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ (part1)

Hi Russell,

On 01.07.2014 18:19, Russell King wrote:
> ARMv6 and greater introduced a new instruction ("bx") which can be used
> to return from function calls. Recent CPUs perform better when the
> "bx lr" instruction is used rather than the "mov pc, lr" instruction,
> and this sequence is strongly recommended to be used by the ARM
> architecture manual (section A.4.1.1).
>
> We provide a new macro "ret" with all its variants for the condition
> code which will resolve to the appropriate instruction.
>
> Rather than doing this piecemeal, and miss some instances, change all
> the "mov pc" instances to use the new macro, with the exception of
> the "movs" instruction and the kprobes code. This allows us to detect
> the "mov pc, lr" case and fix it up - and also gives us the possibility
> of deploying this for other registers depending on the CPU selection.
>
> Signed-off-by: Russell King <[email protected]>

Build- and boot-tested on mach-s3c64xx.

Tested-by: Tomasz Figa <[email protected]>

Best regards,
Tomasz