From: Sascha Hauer <s.hauer@pengutronix.de>
To: "open list:BAREBOX" <barebox@lists.infradead.org>
Cc: Ahmad Fatoum <a.fatoum@pengutronix.de>
Subject: [PATCH v2 03/10] ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+
Date: Thu, 26 Sep 2024 13:17:05 +0200 [thread overview]
Message-ID: <20240926-arm-assembly-memmove-v2-3-0a3313d29a66@pengutronix.de> (raw)
In-Reply-To: <20240926-arm-assembly-memmove-v2-0-0a3313d29a66@pengutronix.de>
Adoption of Linux commit:
| commit 6ebbf2ce437b33022d30badd49dc94d33ecfa498
| Author: Russell King <rmk+kernel@arm.linux.org.uk>
| Date: Mon Jun 30 16:29:12 2014 +0100
|
| ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+
|
| 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.
|
| Reported-by: Will Deacon <will.deacon@arm.com>
| Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1
| Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S
| Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood
| Tested-by: Shawn Guo <shawn.guo@freescale.com>
| Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs
| Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385
| Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci
| Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp
| Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx
| Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen
| Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M
| Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile
| Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Reviewed-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
arch/arm/cpu/cache-armv4.S | 11 ++++++-----
arch/arm/cpu/cache-armv5.S | 13 +++++++------
arch/arm/cpu/cache-armv6.S | 13 +++++++------
arch/arm/cpu/cache-armv7.S | 9 +++++----
arch/arm/cpu/hyp.S | 3 ++-
arch/arm/cpu/setupc_32.S | 7 ++++---
arch/arm/cpu/sm_as.S | 3 ++-
arch/arm/include/asm/assembler.h | 22 ++++++++++++++++++++++
arch/arm/lib32/ashldi3.S | 3 ++-
arch/arm/lib32/ashrdi3.S | 3 ++-
arch/arm/lib32/lshrdi3.S | 3 ++-
arch/arm/lib32/runtime-offset.S | 2 +-
12 files changed, 62 insertions(+), 30 deletions(-)
diff --git a/arch/arm/cpu/cache-armv4.S b/arch/arm/cpu/cache-armv4.S
index 78a098b2fe..024a94c583 100644
--- a/arch/arm/cpu/cache-armv4.S
+++ b/arch/arm/cpu/cache-armv4.S
@@ -2,6 +2,7 @@
#include <linux/linkage.h>
#include <init.h>
+#include <asm/assembler.h>
#define CACHE_DLINESIZE 32
@@ -22,7 +23,7 @@ ENTRY(v4_mmu_cache_on)
mov r0, #0
mcr p15, 0, r0, c8, c7, 0 @ flush I,D TLBs
#endif
- mov pc, r12
+ ret r12
ENDPROC(v4_mmu_cache_on)
__common_mmu_cache_on:
@@ -43,7 +44,7 @@ ENTRY(v4_mmu_cache_off)
mcr p15, 0, r0, c7, c7 @ invalidate whole cache v4
mcr p15, 0, r0, c8, c7 @ invalidate whole TLB v4
#endif
- mov pc, lr
+ ret lr
ENDPROC(v4_mmu_cache_off)
.section .text.v4_mmu_cache_flush
@@ -105,7 +106,7 @@ ENTRY(v4_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)
@@ -125,7 +126,7 @@ ENTRY(v4_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)
@@ -143,5 +144,5 @@ ENTRY(v4_dma_flush_range)
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr
diff --git a/arch/arm/cpu/cache-armv5.S b/arch/arm/cpu/cache-armv5.S
index bcb7ebf466..6d9cbba015 100644
--- a/arch/arm/cpu/cache-armv5.S
+++ b/arch/arm/cpu/cache-armv5.S
@@ -2,6 +2,7 @@
#include <linux/linkage.h>
#include <init.h>
+#include <asm/assembler.h>
#define CACHE_DLINESIZE 32
@@ -22,7 +23,7 @@ ENTRY(v5_mmu_cache_on)
mov r0, #0
mcr p15, 0, r0, c8, c7, 0 @ flush I,D TLBs
#endif
- mov pc, r12
+ ret r12
ENDPROC(v5_mmu_cache_on)
__common_mmu_cache_on:
@@ -43,7 +44,7 @@ ENTRY(v5_mmu_cache_off)
mcr p15, 0, r0, c7, c7 @ invalidate whole cache v4
mcr p15, 0, r0, c8, c7 @ invalidate whole TLB v4
#endif
- mov pc, lr
+ ret lr
ENDPROC(v5_mmu_cache_off)
.section .text.v5_mmu_cache_flush
@@ -52,7 +53,7 @@ ENTRY(v5_mmu_cache_flush)
bne 1b
mcr p15, 0, r0, c7, c5, 0 @ flush I cache
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr
ENDPROC(v5_mmu_cache_flush)
/*
@@ -80,7 +81,7 @@ ENTRY(v5_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)
@@ -100,7 +101,7 @@ ENTRY(v5_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)
@@ -118,5 +119,5 @@ ENTRY(v5_dma_flush_range)
cmp r0, r1
blo 1b
mcr p15, 0, r0, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr
diff --git a/arch/arm/cpu/cache-armv6.S b/arch/arm/cpu/cache-armv6.S
index cc720314c0..ab965623a3 100644
--- a/arch/arm/cpu/cache-armv6.S
+++ b/arch/arm/cpu/cache-armv6.S
@@ -2,6 +2,7 @@
#include <linux/linkage.h>
#include <init.h>
+#include <asm/assembler.h>
#define HARVARD_CACHE
#define CACHE_LINE_SIZE 32
@@ -24,7 +25,7 @@ ENTRY(v6_mmu_cache_on)
mov r0, #0
mcr p15, 0, r0, c8, c7, 0 @ flush I,D TLBs
#endif
- mov pc, r12
+ ret r12
ENDPROC(v6_mmu_cache_on)
__common_mmu_cache_on:
@@ -46,7 +47,7 @@ ENTRY(v6_mmu_cache_off)
mcr p15, 0, r0, c7, c7 @ invalidate whole cache v4
mcr p15, 0, r0, c8, c7 @ invalidate whole TLB v4
#endif
- mov pc, lr
+ ret lr
.section .text.v6_mmu_cache_flush
ENTRY(v6_mmu_cache_flush)
@@ -55,7 +56,7 @@ ENTRY(v6_mmu_cache_flush)
mcr p15, 0, r1, c7, c5, 0 @ invalidate I+BTB
mcr p15, 0, r1, c7, c15, 0 @ clean+invalidate unified
mcr p15, 0, r1, c7, c10, 4 @ drain WB
- mov pc, lr
+ ret lr
ENDPROC(v6_mmu_cache_flush)
/*
@@ -95,7 +96,7 @@ ENTRY(v6_dma_inv_range)
blo 1b
mov r0, #0
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr
ENDPROC(v6_dma_inv_range)
/*
@@ -117,7 +118,7 @@ ENTRY(v6_dma_clean_range)
blo 1b
mov r0, #0
mcr p15, 0, r0, c7, c10, 4 @ drain write buffer
- mov pc, lr
+ ret lr
ENDPROC(v6_dma_clean_range)
/*
@@ -139,5 +140,5 @@ 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
ENDPROC(v6_dma_flush_range)
diff --git a/arch/arm/cpu/cache-armv7.S b/arch/arm/cpu/cache-armv7.S
index efd9fe412f..3f6e5e6b73 100644
--- a/arch/arm/cpu/cache-armv7.S
+++ b/arch/arm/cpu/cache-armv7.S
@@ -2,6 +2,7 @@
#include <linux/linkage.h>
#include <init.h>
+#include <asm/assembler.h>
.section .text.v7_mmu_cache_on
ENTRY(v7_mmu_cache_on)
@@ -140,7 +141,7 @@ iflush:
mcr p15, 0, r12, c7, c5, 0 @ invalidate I+BTB
dsb
isb
- mov pc, lr
+ ret lr
ENDPROC(__v7_mmu_cache_flush_invalidate)
/*
@@ -182,7 +183,7 @@ ENTRY(v7_dma_inv_range)
cmp r0, r1
blo 1b
dsb
- mov pc, lr
+ ret lr
ENDPROC(v7_dma_inv_range)
/*
@@ -201,7 +202,7 @@ ENTRY(v7_dma_clean_range)
cmp r0, r1
blo 1b
dsb
- mov pc, lr
+ ret lr
ENDPROC(v7_dma_clean_range)
/*
@@ -220,5 +221,5 @@ ENTRY(v7_dma_flush_range)
cmp r0, r1
blo 1b
dsb
- mov pc, lr
+ ret lr
ENDPROC(v7_dma_flush_range)
diff --git a/arch/arm/cpu/hyp.S b/arch/arm/cpu/hyp.S
index b5e4807877..016bcd79c0 100644
--- a/arch/arm/cpu/hyp.S
+++ b/arch/arm/cpu/hyp.S
@@ -4,6 +4,7 @@
#include <asm/system.h>
#include <asm/opcodes-virt.h>
#include <init.h>
+#include <asm/assembler.h>
.arch_extension sec
.arch_extension virt
@@ -80,7 +81,7 @@ THUMB( orr r12, r12, #PSR_T_BIT )
__ERET
1: msr cpsr_c, r12
2:
- mov pc, r2
+ ret r2
ENDPROC(armv7_hyp_install)
ENTRY(armv7_switch_to_hyp)
diff --git a/arch/arm/cpu/setupc_32.S b/arch/arm/cpu/setupc_32.S
index eafc9b52c6..d3449d9646 100644
--- a/arch/arm/cpu/setupc_32.S
+++ b/arch/arm/cpu/setupc_32.S
@@ -2,6 +2,7 @@
#include <linux/linkage.h>
#include <asm/sections.h>
+#include <asm/assembler.h>
.section .text.setupc
@@ -32,7 +33,7 @@ ENTRY(setup_c)
bl sync_caches_for_execution
sub lr, r5, r4 /* adjust return address to new location */
pop {r4, r5}
- mov pc, lr
+ ret lr
ENDPROC(setup_c)
/*
@@ -76,13 +77,13 @@ ENTRY(relocate_to_adr)
ldr r0,=1f
sub r0, r0, r8
add r0, r0, r6
- mov pc, r0 /* jump to relocated address */
+ ret r0 /* jump to relocated address */
1:
bl relocate_to_current_adr /* relocate binary */
mov lr, r7
pop {r3, r4, r5, r6, r7, r8}
- mov pc, lr
+ ret lr
ENDPROC(relocate_to_adr)
diff --git a/arch/arm/cpu/sm_as.S b/arch/arm/cpu/sm_as.S
index f55ac8661c..32007147d4 100644
--- a/arch/arm/cpu/sm_as.S
+++ b/arch/arm/cpu/sm_as.S
@@ -5,6 +5,7 @@
#include <asm-generic/memory_layout.h>
#include <asm/secure.h>
#include <asm/system.h>
+#include <asm/assembler.h>
.arch_extension sec
.arch_extension virt
@@ -147,7 +148,7 @@ secure_monitor:
hyp_trap:
mrs lr, elr_hyp @ for older asm: .byte 0x00, 0xe3, 0x0e, 0xe1
- mov pc, lr @ do no switch modes, but
+ ret lr @ do no switch modes, but
@ return to caller
ENTRY(psci_cpu_entry)
diff --git a/arch/arm/include/asm/assembler.h b/arch/arm/include/asm/assembler.h
index 4e7ad57170..e8f5625a0a 100644
--- a/arch/arm/include/asm/assembler.h
+++ b/arch/arm/include/asm/assembler.h
@@ -340,4 +340,26 @@
blx\c \dst
.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_BAREBOX
+ nop
+#endif
+ .endm
+
#endif
diff --git a/arch/arm/lib32/ashldi3.S b/arch/arm/lib32/ashldi3.S
index b62e06f602..dccb732078 100644
--- a/arch/arm/lib32/ashldi3.S
+++ b/arch/arm/lib32/ashldi3.S
@@ -23,6 +23,7 @@ General Public License for more details.
*/
#include <linux/linkage.h>
+#include <asm/assembler.h>
#ifdef __ARMEB__
#define al r1
@@ -44,7 +45,7 @@ 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/lib32/ashrdi3.S b/arch/arm/lib32/ashrdi3.S
index db849b65fc..3db06281e5 100644
--- a/arch/arm/lib32/ashrdi3.S
+++ b/arch/arm/lib32/ashrdi3.S
@@ -23,6 +23,7 @@ General Public License for more details.
*/
#include <linux/linkage.h>
+#include <asm/assembler.h>
#ifdef __ARMEB__
#define al r1
@@ -44,7 +45,7 @@ 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/lib32/lshrdi3.S b/arch/arm/lib32/lshrdi3.S
index e77e96c7bc..5af522482c 100644
--- a/arch/arm/lib32/lshrdi3.S
+++ b/arch/arm/lib32/lshrdi3.S
@@ -23,6 +23,7 @@ General Public License for more details.
*/
#include <linux/linkage.h>
+#include <asm/assembler.h>
#ifdef __ARMEB__
#define al r1
@@ -44,7 +45,7 @@ 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/lib32/runtime-offset.S b/arch/arm/lib32/runtime-offset.S
index ac104de119..d9ba864b3b 100644
--- a/arch/arm/lib32/runtime-offset.S
+++ b/arch/arm/lib32/runtime-offset.S
@@ -14,7 +14,7 @@ ENTRY(get_runtime_offset)
ldr r1, linkadr
subs r0, r0, r1
THUMB( adds r0, r0, #1)
- mov pc, lr
+ ret lr
linkadr:
.word get_runtime_offset
--
2.39.5
next prev parent reply other threads:[~2024-09-26 11:23 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-09-26 11:17 [PATCH v2 00/10] ARM: add assembler optimized memmove Sascha Hauer
2024-09-26 11:17 ` [PATCH v2 01/10] ARM: Use optimized reads[bwl] and writes[bwl] functions Sascha Hauer
2024-09-26 11:17 ` [PATCH v2 02/10] ARM: rename logical shift macros push pull into lspush lspull Sascha Hauer
2024-09-26 11:17 ` Sascha Hauer [this message]
2024-09-26 11:17 ` [PATCH v2 04/10] ARM: update lib1funcs.S from Linux Sascha Hauer
2024-09-26 11:17 ` [PATCH v2 05/10] ARM: update findbit.S " Sascha Hauer
2024-09-26 11:17 ` [PATCH v2 06/10] ARM: update io-* " Sascha Hauer
2024-09-26 11:17 ` [PATCH v2 07/10] ARM: always assume the unified syntax for assembly code Sascha Hauer
2024-09-26 11:17 ` [PATCH v2 08/10] ARM: update memcpy.S and memset.S from Linux Sascha Hauer
2024-09-26 11:17 ` [PATCH v2 09/10] lib/string.c: export non optimized memmove as __default_memmove Sascha Hauer
2024-09-26 11:17 ` [PATCH v2 10/10] ARM: add optimized memmove Sascha Hauer
2024-09-27 5:12 ` Marco Felsch
2024-09-27 10:04 ` Sascha Hauer
2024-09-27 10:39 ` [PATCH v2 00/10] ARM: add assembler " Sascha Hauer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20240926-arm-assembly-memmove-v2-3-0a3313d29a66@pengutronix.de \
--to=s.hauer@pengutronix.de \
--cc=a.fatoum@pengutronix.de \
--cc=barebox@lists.infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox