[U-Boot] [PATCH v2] arm: build arch memset/memcpy in Thumb2 mode
Resynchronize memcpy/memset with kernel and build them explicitly in Thumb2 mode (unified syntax). Those assembler files can be built and linked in ARM mode too, however when calling them from Thumb2 built code, the stack got corrupted and the copy did not succeed (the exact details have not been traced back). Hoever, the Linux kernel builds those files in Thumb2 mode. Hence U-Boot should build them in Thumb2 mode too when CONFIG_SYS_THUMB_BUILD is set.
To build the files without warning, some assembler instructions had to be replaced with their UAL compliant variant (thanks Jeroen for this input).
To build the file in Thumb2 mode the implicit-it=always option need to be set to generate Thumb2 compliant IT instructions where needed. We add this option to the general AFLAGS when building for Thumb2.
Signed-off-by: Stefan Agner stefan@agner.ch --- Changes since v1: - Don't set auto-it in AFLAGS - Removed "no-warn-deprecated" warning suppression - Converted non-UAL assembler instructions to their UAL compliant variants
Tested in ARM and Thumb2 mode on Vybrid SoC. Disassembled the memset/ memcpy object files before and after converting the instructions to UAL, the output was identical.
arch/arm/config.mk | 4 +- arch/arm/include/asm/assembler.h | 33 ++++++++++-- arch/arm/lib/memcpy.S | 80 +++++++++++++++++++--------- arch/arm/lib/memset.S | 112 ++++++++++++++++++++------------------- 4 files changed, 142 insertions(+), 87 deletions(-)
diff --git a/arch/arm/config.mk b/arch/arm/config.mk index f0eafd6..14f4e9f 100644 --- a/arch/arm/config.mk +++ b/arch/arm/config.mk @@ -26,7 +26,9 @@ PLATFORM_CPPFLAGS += -D__ARM__
# Choose between ARM/Thumb instruction sets ifeq ($(CONFIG_SYS_THUMB_BUILD),y) -PF_CPPFLAGS_ARM := $(call cc-option, -mthumb -mthumb-interwork,\ +AFLAGS_IMPLICIT_IT := $(call as-option,-Wa$(comma)-mimplicit-it=always) +PF_CPPFLAGS_ARM := $(AFLAGS_IMPLICIT_IT) $(AFLAGS_NOWARN) \ + $(call cc-option, -mthumb -mthumb-interwork,\ $(call cc-option,-marm,)\ $(call cc-option,-mno-thumb-interwork,)\ ) diff --git a/arch/arm/include/asm/assembler.h b/arch/arm/include/asm/assembler.h index 5e4789b..11b80fb 100644 --- a/arch/arm/include/asm/assembler.h +++ b/arch/arm/include/asm/assembler.h @@ -14,12 +14,14 @@ * assembler source. */
+#include <config.h> + /* * Endian independent macros for shifting bytes within registers. */ #ifndef __ARMEB__ -#define pull lsr -#define push lsl +#define lspull lsr +#define lspush lsl #define get_byte_0 lsl #0 #define get_byte_1 lsr #8 #define get_byte_2 lsr #16 @@ -29,8 +31,8 @@ #define put_byte_2 lsl #16 #define put_byte_3 lsl #24 #else -#define pull lsl -#define push lsr +#define lspull lsl +#define lspush lsr #define get_byte_0 lsr #24 #define get_byte_1 lsr #16 #define get_byte_2 lsr #8 @@ -54,7 +56,28 @@ #define PLD(code...) #endif
+ .irp c,,eq,ne,cs,cc,mi,pl,vs,vc,hi,ls,ge,lt,gt,le,hs,lo + .macro ret\c, reg +#if defined(__ARM_ARCH_5E__) || defined(__ARM_ARCH_5TE__) + mov\c pc, \reg +#else + .ifeqs "\reg", "lr" + bx\c \reg + .else + mov\c pc, \reg + .endif +#endif + .endm + .endr + /* - * Cache alligned + * Cache aligned, used for optimized memcpy/memset + * In the kernel this is only enabled for Feroceon CPU's... + * We disable it especially for Thumb builds since those instructions + * are not made in a Thumb ready way... */ +#ifdef CONFIG_SYS_THUMB_BUILD +#define CALGN(code...) +#else #define CALGN(code...) code +#endif diff --git a/arch/arm/lib/memcpy.S b/arch/arm/lib/memcpy.S index f655256..eeaf003 100644 --- a/arch/arm/lib/memcpy.S +++ b/arch/arm/lib/memcpy.S @@ -10,9 +10,14 @@ * published by the Free Software Foundation. */
+#include <linux/linkage.h> #include <asm/assembler.h>
+#ifdef CONFIG_SYS_THUMB_BUILD +#define W(instr) instr.w +#else #define W(instr) instr +#endif
#define LDR1W_SHIFT 0 #define STR1W_SHIFT 0 @@ -30,7 +35,7 @@ .endm
.macro ldr1b ptr reg cond=al abort - ldr\cond()b \reg, [\ptr], #1 + ldrb\cond() \reg, [\ptr], #1 .endm
.macro str1w ptr reg abort @@ -42,7 +47,7 @@ .endm
.macro str1b ptr reg cond=al abort - str\cond()b \reg, [\ptr], #1 + strb\cond() \reg, [\ptr], #1 .endm
.macro enter reg1 reg2 @@ -56,10 +61,12 @@ .text
/* Prototype: void *memcpy(void *dest, const void *src, size_t n); */ - -.globl memcpy -memcpy: - + .syntax unified +#ifdef CONFIG_SYS_THUMB_BUILD + .thumb + .thumb_func +#endif +ENTRY(memcpy) cmp r0, r1 moveq pc, lr
@@ -79,7 +86,7 @@ memcpy:
CALGN( ands ip, r0, #31 ) CALGN( rsb r3, ip, #32 ) - CALGN( sbcnes r4, r3, r2 ) @ C is always set here + CALGN( sbcsne r4, r3, r2 ) @ C is always set here CALGN( bcs 2f ) CALGN( adr r4, 6f ) CALGN( subs r2, r2, r3 ) @ C gets set @@ -178,7 +185,7 @@ memcpy:
CALGN( ands ip, r0, #31 ) CALGN( rsb ip, ip, #32 ) - CALGN( sbcnes r4, ip, r2 ) @ C is always set here + CALGN( sbcsne r4, ip, r2 ) @ C is always set here CALGN( subcc r2, r2, ip ) CALGN( bcc 15f )
@@ -193,24 +200,24 @@ memcpy:
12: PLD( pld [r1, #124] ) 13: ldr4w r1, r4, r5, r6, r7, abort=19f - mov r3, lr, pull #\pull + mov r3, lr, lspull #\pull subs r2, r2, #32 ldr4w r1, r8, r9, ip, lr, abort=19f - orr r3, r3, r4, push #\push - mov r4, r4, pull #\pull - orr r4, r4, r5, push #\push - mov r5, r5, pull #\pull - orr r5, r5, r6, push #\push - mov r6, r6, pull #\pull - orr r6, r6, r7, push #\push - mov r7, r7, pull #\pull - orr r7, r7, r8, push #\push - mov r8, r8, pull #\pull - orr r8, r8, r9, push #\push - mov r9, r9, pull #\pull - orr r9, r9, ip, push #\push - mov ip, ip, pull #\pull - orr ip, ip, lr, push #\push + orr r3, r3, r4, lspush #\push + mov r4, r4, lspull #\pull + orr r4, r4, r5, lspush #\push + mov r5, r5, lspull #\pull + orr r5, r5, r6, lspush #\push + mov r6, r6, lspull #\pull + orr r6, r6, r7, lspush #\push + mov r7, r7, lspull #\pull + orr r7, r7, r8, lspush #\push + mov r8, r8, lspull #\pull + orr r8, r8, r9, lspush #\push + mov r9, r9, lspull #\pull + orr r9, r9, ip, lspush #\push + mov ip, ip, lspull #\pull + orr ip, ip, lr, lspush #\push str8w r0, r3, r4, r5, r6, r7, r8, r9, ip, , abort=19f bge 12b PLD( cmn r2, #96 ) @@ -221,10 +228,10 @@ memcpy: 14: ands ip, r2, #28 beq 16f
-15: mov r3, lr, pull #\pull +15: mov r3, lr, lspull #\pull ldr1w r1, lr, abort=21f subs ip, ip, #4 - orr r3, r3, lr, push #\push + orr r3, r3, lr, lspush #\push str1w r0, r3, abort=21f bgt 15b CALGN( cmp r2, #0 ) @@ -241,3 +248,24 @@ memcpy: 17: forward_copy_shift pull=16 push=16
18: forward_copy_shift pull=24 push=8 + + +/* + * Abort preamble and completion macros. + * If a fixup handler is required then those macros must surround it. + * It is assumed that the fixup code will handle the private part of + * the exit macro. + */ + + .macro copy_abort_preamble +19: ldmfd sp!, {r5 - r9} + b 21f +20: ldmfd sp!, {r5 - r8} +21: + .endm + + .macro copy_abort_end + ldmfd sp!, {r4, pc} + .endm + +ENDPROC(memcpy) diff --git a/arch/arm/lib/memset.S b/arch/arm/lib/memset.S index 0cdf895..7208f20 100644 --- a/arch/arm/lib/memset.S +++ b/arch/arm/lib/memset.S @@ -9,32 +9,25 @@ * * ASM optimised string functions */ +#include <linux/linkage.h> #include <asm/assembler.h>
.text .align 5 - .word 0
-1: subs r2, r2, #4 @ 1 do we have enough - blt 5f @ 1 bytes to align with? - cmp r3, #2 @ 1 - strltb r1, [r0], #1 @ 1 - strleb r1, [r0], #1 @ 1 - strb r1, [r0], #1 @ 1 - add r2, r2, r3 @ 1 (r2 = r2 - (4 - r3)) -/* - * The pointer is now aligned and the length is adjusted. Try doing the - * memset again. - */ - -.globl memset -memset: + .syntax unified +#ifdef CONFIG_SYS_THUMB_BUILD + .thumb + .thumb_func +#endif +ENTRY(memset) ands r3, r0, #3 @ 1 unaligned? - bne 1b @ 1 + mov ip, r0 @ preserve r0 as return value + bne 6f @ 1 /* - * we know that the pointer in r0 is aligned to a word boundary. + * we know that the pointer in ip is aligned to a word boundary. */ - orr r1, r1, r1, lsl #8 +1: orr r1, r1, r1, lsl #8 orr r1, r1, r1, lsl #16 mov r3, r1 cmp r2, #16 @@ -43,29 +36,28 @@ memset: #if ! CALGN(1)+0
/* - * We need an extra register for this loop - save the return address and - * use the LR + * We need 2 extra registers for this loop - use r8 and the LR */ - str lr, [sp, #-4]! - mov ip, r1 + stmfd sp!, {r8, lr} + mov r8, r1 mov lr, r1
2: subs r2, r2, #64 - stmgeia r0!, {r1, r3, ip, lr} @ 64 bytes at a time. - stmgeia r0!, {r1, r3, ip, lr} - stmgeia r0!, {r1, r3, ip, lr} - stmgeia r0!, {r1, r3, ip, lr} + stmiage ip!, {r1, r3, r8, lr} @ 64 bytes at a time. + stmiage ip!, {r1, r3, r8, lr} + stmiage ip!, {r1, r3, r8, lr} + stmiage ip!, {r1, r3, r8, lr} bgt 2b - ldmeqfd sp!, {pc} @ Now <64 bytes to go. + ldmfdeq sp!, {r8, pc} @ Now <64 bytes to go. /* * No need to correct the count; we're only testing bits from now on */ tst r2, #32 - stmneia r0!, {r1, r3, ip, lr} - stmneia r0!, {r1, r3, ip, lr} + stmiane ip!, {r1, r3, r8, lr} + stmiane ip!, {r1, r3, r8, lr} tst r2, #16 - stmneia r0!, {r1, r3, ip, lr} - ldr lr, [sp], #4 + stmiane ip!, {r1, r3, r8, lr} + ldmfd sp!, {r8, lr}
#else
@@ -74,53 +66,63 @@ memset: * whole cache lines at once. */
- stmfd sp!, {r4-r7, lr} + stmfd sp!, {r4-r8, lr} mov r4, r1 mov r5, r1 mov r6, r1 mov r7, r1 - mov ip, r1 + mov r8, r1 mov lr, r1
cmp r2, #96 - tstgt r0, #31 + tstgt ip, #31 ble 3f
- and ip, r0, #31 - rsb ip, ip, #32 - sub r2, r2, ip - movs ip, ip, lsl #(32 - 4) - stmcsia r0!, {r4, r5, r6, r7} - stmmiia r0!, {r4, r5} - tst ip, #(1 << 30) - mov ip, r1 - strne r1, [r0], #4 + and r8, ip, #31 + rsb r8, r8, #32 + sub r2, r2, r8 + movs r8, r8, lsl #(32 - 4) + stmiacs ip!, {r4, r5, r6, r7} + stmiami ip!, {r4, r5} + tst r8, #(1 << 30) + mov r8, r1 + strne r1, [ip], #4
3: subs r2, r2, #64 - stmgeia r0!, {r1, r3-r7, ip, lr} - stmgeia r0!, {r1, r3-r7, ip, lr} + stmiage ip!, {r1, r3-r8, lr} + stmiage ip!, {r1, r3-r8, lr} bgt 3b - ldmeqfd sp!, {r4-r7, pc} + ldmfdeq sp!, {r4-r8, pc}
tst r2, #32 - stmneia r0!, {r1, r3-r7, ip, lr} + stmiane ip!, {r1, r3-r8, lr} tst r2, #16 - stmneia r0!, {r4-r7} - ldmfd sp!, {r4-r7, lr} + stmiane ip!, {r4-r7} + ldmfd sp!, {r4-r8, lr}
#endif
4: tst r2, #8 - stmneia r0!, {r1, r3} + stmiane ip!, {r1, r3} tst r2, #4 - strne r1, [r0], #4 + strne r1, [ip], #4 /* * When we get here, we've got less than 4 bytes to zero. We * may have an unaligned pointer as well. */ 5: tst r2, #2 - strneb r1, [r0], #1 - strneb r1, [r0], #1 + strbne r1, [ip], #1 + strbne r1, [ip], #1 tst r2, #1 - strneb r1, [r0], #1 - mov pc, lr + strbne r1, [ip], #1 + ret lr + +6: subs r2, r2, #4 @ 1 do we have enough + blt 5b @ 1 bytes to align with? + cmp r3, #2 @ 1 + strblt r1, [ip], #1 @ 1 + strble r1, [ip], #1 @ 1 + strb r1, [ip], #1 @ 1 + add r2, r2, r3 @ 1 (r2 = r2 - (4 - r3)) + b 1b +ENDPROC(memset)
On 2014-11-21 16:34, Stefan Agner wrote:
Resynchronize memcpy/memset with kernel and build them explicitly in Thumb2 mode (unified syntax). Those assembler files can be built and linked in ARM mode too, however when calling them from Thumb2 built code, the stack got corrupted and the copy did not succeed (the exact details have not been traced back). Hoever, the Linux kernel builds those files in Thumb2 mode. Hence U-Boot should build them in Thumb2 mode too when CONFIG_SYS_THUMB_BUILD is set.
To build the files without warning, some assembler instructions had to be replaced with their UAL compliant variant (thanks Jeroen for this input).
To build the file in Thumb2 mode the implicit-it=always option need to be set to generate Thumb2 compliant IT instructions where needed. We add this option to the general AFLAGS when building for Thumb2.
Signed-off-by: Stefan Agner stefan@agner.ch
Changes since v1:
- Don't set auto-it in AFLAGS
- Removed "no-warn-deprecated" warning suppression
- Converted non-UAL assembler instructions to their UAL compliant variants
Tested in ARM and Thumb2 mode on Vybrid SoC. Disassembled the memset/ memcpy object files before and after converting the instructions to UAL, the output was identical.
Any thoughts on that? Would be nice to have it in 2015.01 as the current state breaks Thumb2 for our board.
-- Stefan
Hi Stefan,
On 21 November 2014 at 08:34, Stefan Agner stefan@agner.ch wrote:
Resynchronize memcpy/memset with kernel and build them explicitly in Thumb2 mode (unified syntax). Those assembler files can be built and linked in ARM mode too, however when calling them from Thumb2 built code, the stack got corrupted and the copy did not succeed (the exact details have not been traced back). Hoever, the Linux kernel builds those files in Thumb2 mode. Hence U-Boot should build them in Thumb2 mode too when CONFIG_SYS_THUMB_BUILD is set.
To build the files without warning, some assembler instructions had to be replaced with their UAL compliant variant (thanks Jeroen for this input).
To build the file in Thumb2 mode the implicit-it=always option need to be set to generate Thumb2 compliant IT instructions where needed. We add this option to the general AFLAGS when building for Thumb2.
Signed-off-by: Stefan Agner stefan@agner.ch
Changes since v1:
- Don't set auto-it in AFLAGS
- Removed "no-warn-deprecated" warning suppression
- Converted non-UAL assembler instructions to their UAL compliant variants
Reviewed-by: Simon Glass sjg@chromium.org
Tested on pit and found that it fixed the problem with Thumb mode. Still works fine in ARM mode.
Tested-by: Simon Glass sjg@chromium.org
A few nits: - typo 'Hoever' in commit message - you could mention the Linux commit/release you synced to
Tested in ARM and Thumb2 mode on Vybrid SoC. Disassembled the memset/ memcpy object files before and after converting the instructions to UAL, the output was identical.
arch/arm/config.mk | 4 +- arch/arm/include/asm/assembler.h | 33 ++++++++++-- arch/arm/lib/memcpy.S | 80 +++++++++++++++++++--------- arch/arm/lib/memset.S | 112 ++++++++++++++++++++------------------- 4 files changed, 142 insertions(+), 87 deletions(-)
Regards, Simon
On 2014-11-30 20:33, Simon Glass wrote:
Hi Stefan,
On 21 November 2014 at 08:34, Stefan Agner stefan@agner.ch wrote:
Resynchronize memcpy/memset with kernel and build them explicitly in Thumb2 mode (unified syntax). Those assembler files can be built and linked in ARM mode too, however when calling them from Thumb2 built code, the stack got corrupted and the copy did not succeed (the exact details have not been traced back). Hoever, the Linux kernel builds those files in Thumb2 mode. Hence U-Boot should build them in Thumb2 mode too when CONFIG_SYS_THUMB_BUILD is set.
To build the files without warning, some assembler instructions had to be replaced with their UAL compliant variant (thanks Jeroen for this input).
To build the file in Thumb2 mode the implicit-it=always option need to be set to generate Thumb2 compliant IT instructions where needed. We add this option to the general AFLAGS when building for Thumb2.
Signed-off-by: Stefan Agner stefan@agner.ch
Changes since v1:
- Don't set auto-it in AFLAGS
- Removed "no-warn-deprecated" warning suppression
- Converted non-UAL assembler instructions to their UAL compliant variants
Reviewed-by: Simon Glass sjg@chromium.org
Tested on pit and found that it fixed the problem with Thumb mode. Still works fine in ARM mode.
Tested-by: Simon Glass sjg@chromium.org
A few nits:
- typo 'Hoever' in commit message
- you could mention the Linux commit/release you synced to
Thanks for testing Simon.
The Linux commit/release I synced to was some 3.18-rcX. However, the relevant files did not change since 3.17, so we can add that as the synced version.
-- Stefan
Tested in ARM and Thumb2 mode on Vybrid SoC. Disassembled the memset/ memcpy object files before and after converting the instructions to UAL, the output was identical.
arch/arm/config.mk | 4 +- arch/arm/include/asm/assembler.h | 33 ++++++++++-- arch/arm/lib/memcpy.S | 80 +++++++++++++++++++--------- arch/arm/lib/memset.S | 112 ++++++++++++++++++++------------------- 4 files changed, 142 insertions(+), 87 deletions(-)
Regards, Simon
I tested these changes on several compilers and binutils. 4.3.3, 4.6.3, 4.7.2, 4.8.2 and 4.9.2 gcc with binutils 2.19, 2.21-24 in both THUMB and ARM on the Vybrid. It seem we must have an image <= 224kB which is the IRAM bank0 size - 0x8000 of u-boot base; u-boot since v2014.04 will not boot for me as the QSPI seemed to take it over this boundary.
The attached patches are meant to be attached as they aren't meant to be taken by patchworks, etc but just to show the small changes I made to minimize the image to less than 224kB.
[build script] # File sizes similar because of 'mkimage' padding+hdr for 'imx'. # 2nd size is for the 'thumb2' builds. set -e
## (Sourcery CodeBench Lite 2012.03-56) 4.6.3 ## (Sourcery CodeBench Lite 2012.03-56) 2.21.53.20110905 # 227092 Dec 1 15:20 u-boot.imx # 173844 Dec 1 15:36 u-boot.imx export PATH=$PATH:$HOME/x-tools/arm-2012.03/bin
## 4.7.4 20130913 (release) [ARM/embedded-4_7-branch revision 202601] ## GNU assembler (GNU Binutils) 2.23.2 # 222996 Dec 1 15:19 u-boot.imx # 173844 Dec 1 15:35 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-2013.q3/bin
## (crosstool-NG hg+default-9838aecd6340) 4.8.2 ## (GNU Binutils) 2.23.2 # 222996 Dec 1 15:18 u-boot.imx # 173844 Dec 1 15:34 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-none-eabi/bin
## (Sourcery G++ Lite 2009q1-161) 4.3.3 ## (Sourcery G++ Lite 2009q1-161) 2.19.51.20090205 # 218900 Dec 1 15:17 u-boot.imx - no CONFIG_SYS_LONGHELP # GCC does not produce working THUMB binaries. #export PATH=$PATH:/opt/arm-2009q1/bin NAME=arm-none-eabi-
## (crosstool-NG 1.20.0) 4.9.2 ## (crosstool-NG 1.20.0) 2.24 # 222996 Dec 1 15:21 u-boot.imx # 173844 Dec 1 15:32 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-none-linux-gnueabi-vybrid-4.9.2/bin/
## (crosstool-NG hg+default-2a773bd38e71) 4.7.3 ## (crosstool-NG hg+default-2a773bd38e71) 2.22 # 222996 Dec 1 15:22 u-boot.imx # 173844 Dec 1 15:31 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-none-linux-gnueabi-4.7.2/bin
## (crosstool-NG hg+-c7fa97debb36) 4.6.3 ## (crosstool-NG hg+-c7fa97debb36) 2.21.1 # 227092 Dec 1 15:24 u-boot.imx # 173844 Dec 1 15:30 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-none-linux-gnueabi.gcc4.6.3/bin #NAME=arm-none-linux-gnueabi-
## (Timesys 20130708) 4.7.3 ## (Timesys 20130708) 2.23.2 # 227092 Dec 1 15:25 u-boot.imx # 173844 Dec 1 15:29 u-boot.imx #export PATH=$PATH:$HOME/x-tools/armv7l-timesys-linux-gnueabi/bin #NAME=armv7l-timesys-linux-gnueabi-
## (crosstool-NG linaro-1.13.1-4.8-2013.10 - Linaro GCC 2013.10) 4.8.2 20131014 ## (crosstool-NG linaro-1.13.1-4.8-2013.10 - Linaro GCC 2013.10) 2.23.2.20130610 Linaro 2013.10-4 # 222996 Dec 1 15:26 u-boot.imx # 173844 Dec 1 15:28 u-boot.imx #export PATH=$PATH:$HOME/x-tools/gcc-linaro-arm-linux-gnueabihf-4.8-2013.10_linux/bin #NAME=arm-linux-gnueabihf-
make CROSS_COMPILE=$NAME vf610twr_config make CROSS_COMPILE=$NAME $* u-boot.imx cp u-boot.imx /srv/tftp
[end script]
I applied Stefan's memcpy/thumb patch, reverted c6150aaf and applied 'fs/ext4/ext4fs.c, fs/fs.c fs/fat/fat_write.c: Adjust 64bit math methods' by Tom Rini trini@ti.com (it has a little whitespace fat in it). All of the images booted on a Vybrid Tower (Rev H) from an SD card. Only issue is that I think the 'AFLAGS_NOWARN' can be removed as per below...
On 21 Nov 2014, stefan@agner.ch wrote:
Resynchronize memcpy/memset with kernel and build them explicitly in Thumb2 mode (unified syntax). Those assembler files can be built and linked in ARM mode too, however when calling them from Thumb2 built code, the stack got corrupted and the copy did not succeed (the exact details have not been traced back). Hoever, the Linux kernel builds those files in Thumb2 mode. Hence U-Boot should build them in Thumb2 mode too when CONFIG_SYS_THUMB_BUILD is set.
To build the files without warning, some assembler instructions had to be replaced with their UAL compliant variant (thanks Jeroen for this input).
To build the file in Thumb2 mode the implicit-it=always option need to be set to generate Thumb2 compliant IT instructions where needed. We add this option to the general AFLAGS when building for Thumb2.
Signed-off-by: Stefan Agner stefan@agner.ch
Changes since v1: - Don't set auto-it in AFLAGS - Removed "no-warn-deprecated" warning suppression - Converted non-UAL assembler instructions to their UAL compliant variants
Tested in ARM and Thumb2 mode on Vybrid SoC. Disassembled the memset/ memcpy object files before and after converting the instructions to UAL, the output was identical.
arch/arm/config.mk | 4 +- arch/arm/include/asm/assembler.h | 33 ++++++++++-- arch/arm/lib/memcpy.S | 80 +++++++++++++++++++--------- arch/arm/lib/memset.S | 112 ++++++++++++++++++++------------------- 4 files changed, 142 insertions(+), 87 deletions(-)
diff --git a/arch/arm/config.mk b/arch/arm/config.mk index f0eafd6..14f4e9f 100644 --- a/arch/arm/config.mk +++ b/arch/arm/config.mk @@ -26,7 +26,9 @@ PLATFORM_CPPFLAGS += -D__ARM__
# Choose between ARM/Thumb instruction sets ifeq ($(CONFIG_SYS_THUMB_BUILD),y) -PF_CPPFLAGS_ARM := $(call cc-option, -mthumb -mthumb-interwork,\ +AFLAGS_IMPLICIT_IT := $(call as-option,-Wa$(comma)-mimplicit-it=always) +PF_CPPFLAGS_ARM := $(AFLAGS_IMPLICIT_IT) $(AFLAGS_NOWARN) \ + $(call cc-option, -mthumb
I don't think we need the '$(AFLAGS_NOWARN)' as you fixed that? I removed it during my tests and didn't see anything. Especially all builds produced output. I guess this is copy/paste from a Linux Makefile as I don't even see where it is set in U-Boot.
-mthumb-interwork,\ $(call cc-option,-marm,)\ $(call cc-option,-mno-thumb-interwork,)\ ) diff --git a/arch/arm/include/asm/assembler.h b/arch/arm/include/asm/assembler.h index 5e4789b..11b80fb 100644 --- a/arch/arm/include/asm/assembler.h +++ b/arch/arm/include/asm/assembler.h @@ -14,12 +14,14 @@ * assembler source. */
[snip]
Tested-by: Bill Pringlemeir bpringlemeir@nbsps.com
On 2014-12-03 16:44, Bill Pringlemeir wrote:
I tested these changes on several compilers and binutils. 4.3.3, 4.6.3, 4.7.2, 4.8.2 and 4.9.2 gcc with binutils 2.19, 2.21-24 in both THUMB and ARM on the Vybrid. It seem we must have an image <= 224kB which is the IRAM bank0 size - 0x8000 of u-boot base; u-boot since v2014.04 will not boot for me as the QSPI seemed to take it over this boundary.
The attached patches are meant to be attached as they aren't meant to be taken by patchworks, etc but just to show the small changes I made to minimize the image to less than 224kB.
[build script] # File sizes similar because of 'mkimage' padding+hdr for 'imx'. # 2nd size is for the 'thumb2' builds. set -e
## (Sourcery CodeBench Lite 2012.03-56) 4.6.3 ## (Sourcery CodeBench Lite 2012.03-56) 2.21.53.20110905 # 227092 Dec 1 15:20 u-boot.imx # 173844 Dec 1 15:36 u-boot.imx export PATH=$PATH:$HOME/x-tools/arm-2012.03/bin
## 4.7.4 20130913 (release) [ARM/embedded-4_7-branch revision 202601] ## GNU assembler (GNU Binutils) 2.23.2 # 222996 Dec 1 15:19 u-boot.imx # 173844 Dec 1 15:35 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-2013.q3/bin
## (crosstool-NG hg+default-9838aecd6340) 4.8.2 ## (GNU Binutils) 2.23.2 # 222996 Dec 1 15:18 u-boot.imx # 173844 Dec 1 15:34 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-none-eabi/bin
## (Sourcery G++ Lite 2009q1-161) 4.3.3 ## (Sourcery G++ Lite 2009q1-161) 2.19.51.20090205 # 218900 Dec 1 15:17 u-boot.imx - no CONFIG_SYS_LONGHELP # GCC does not produce working THUMB binaries. #export PATH=$PATH:/opt/arm-2009q1/bin NAME=arm-none-eabi-
## (crosstool-NG 1.20.0) 4.9.2 ## (crosstool-NG 1.20.0) 2.24 # 222996 Dec 1 15:21 u-boot.imx # 173844 Dec 1 15:32 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-none-linux-gnueabi-vybrid-4.9.2/bin/
## (crosstool-NG hg+default-2a773bd38e71) 4.7.3 ## (crosstool-NG hg+default-2a773bd38e71) 2.22 # 222996 Dec 1 15:22 u-boot.imx # 173844 Dec 1 15:31 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-none-linux-gnueabi-4.7.2/bin
## (crosstool-NG hg+-c7fa97debb36) 4.6.3 ## (crosstool-NG hg+-c7fa97debb36) 2.21.1 # 227092 Dec 1 15:24 u-boot.imx # 173844 Dec 1 15:30 u-boot.imx #export PATH=$PATH:$HOME/x-tools/arm-none-linux-gnueabi.gcc4.6.3/bin #NAME=arm-none-linux-gnueabi-
## (Timesys 20130708) 4.7.3 ## (Timesys 20130708) 2.23.2 # 227092 Dec 1 15:25 u-boot.imx # 173844 Dec 1 15:29 u-boot.imx #export PATH=$PATH:$HOME/x-tools/armv7l-timesys-linux-gnueabi/bin #NAME=armv7l-timesys-linux-gnueabi-
## (crosstool-NG linaro-1.13.1-4.8-2013.10 - Linaro GCC 2013.10) 4.8.2 20131014 ## (crosstool-NG linaro-1.13.1-4.8-2013.10 - Linaro GCC 2013.10) 2.23.2.20130610 Linaro 2013.10-4 # 222996 Dec 1 15:26 u-boot.imx # 173844 Dec 1 15:28 u-boot.imx #export PATH=$PATH:$HOME/x-tools/gcc-linaro-arm-linux-gnueabihf-4.8-2013.10_linux/bin #NAME=arm-linux-gnueabihf-
make CROSS_COMPILE=$NAME vf610twr_config make CROSS_COMPILE=$NAME $* u-boot.imx cp u-boot.imx /srv/tftp
[end script]
I applied Stefan's memcpy/thumb patch, reverted c6150aaf and applied 'fs/ext4/ext4fs.c, fs/fs.c fs/fat/fat_write.c: Adjust 64bit math methods' by Tom Rini trini@ti.com (it has a little whitespace fat in it). All of the images booted on a Vybrid Tower (Rev H) from an SD card. Only issue is that I think the 'AFLAGS_NOWARN' can be removed as per below...
On 21 Nov 2014, stefan@agner.ch wrote:
Resynchronize memcpy/memset with kernel and build them explicitly in Thumb2 mode (unified syntax). Those assembler files can be built and linked in ARM mode too, however when calling them from Thumb2 built code, the stack got corrupted and the copy did not succeed (the exact details have not been traced back). Hoever, the Linux kernel builds those files in Thumb2 mode. Hence U-Boot should build them in Thumb2 mode too when CONFIG_SYS_THUMB_BUILD is set.
To build the files without warning, some assembler instructions had to be replaced with their UAL compliant variant (thanks Jeroen for this input).
To build the file in Thumb2 mode the implicit-it=always option need to be set to generate Thumb2 compliant IT instructions where needed. We add this option to the general AFLAGS when building for Thumb2.
Signed-off-by: Stefan Agner stefan@agner.ch
Changes since v1: - Don't set auto-it in AFLAGS - Removed "no-warn-deprecated" warning suppression - Converted non-UAL assembler instructions to their UAL compliant variants
Tested in ARM and Thumb2 mode on Vybrid SoC. Disassembled the memset/ memcpy object files before and after converting the instructions to UAL, the output was identical.
arch/arm/config.mk | 4 +- arch/arm/include/asm/assembler.h | 33 ++++++++++-- arch/arm/lib/memcpy.S | 80 +++++++++++++++++++--------- arch/arm/lib/memset.S | 112 ++++++++++++++++++++------------------- 4 files changed, 142 insertions(+), 87 deletions(-)
diff --git a/arch/arm/config.mk b/arch/arm/config.mk index f0eafd6..14f4e9f 100644 --- a/arch/arm/config.mk +++ b/arch/arm/config.mk @@ -26,7 +26,9 @@ PLATFORM_CPPFLAGS += -D__ARM__
# Choose between ARM/Thumb instruction sets ifeq ($(CONFIG_SYS_THUMB_BUILD),y) -PF_CPPFLAGS_ARM := $(call cc-option, -mthumb -mthumb-interwork,\ +AFLAGS_IMPLICIT_IT := $(call as-option,-Wa$(comma)-mimplicit-it=always) +PF_CPPFLAGS_ARM := $(AFLAGS_IMPLICIT_IT) $(AFLAGS_NOWARN) \ + $(call cc-option, -mthumb
I don't think we need the '$(AFLAGS_NOWARN)' as you fixed that? I removed it during my tests and didn't see anything. Especially all builds produced output. I guess this is copy/paste from a Linux Makefile as I don't even see where it is set in U-Boot.
You are right, I removed the assignation of that but did not remove it's inclucion into the PF_CPPFLAGS_ARM variable...
Can this be fixed by the merger or should I create a new revision?
-- Stefan
-mthumb-interwork,\ $(call cc-option,-marm,)\ $(call cc-option,-mno-thumb-interwork,)\ ) diff --git a/arch/arm/include/asm/assembler.h b/arch/arm/include/asm/assembler.h index 5e4789b..11b80fb 100644 --- a/arch/arm/include/asm/assembler.h +++ b/arch/arm/include/asm/assembler.h @@ -14,12 +14,14 @@ * assembler source. */
[snip]
Tested-by: Bill Pringlemeir bpringlemeir@nbsps.com
Hi Stefan,
Am 03.12.2014 um 18:04 schrieb Stefan Agner:
Can this be fixed by the merger or should I create a new revision?
It looks as if this was neither applied nor respun? I have some more patches to make CONFIG_USE_PRIVATE_LIBGCC build for Thumb that I would like to rebase on it.
In particular, I have pieced together a tiny asm/unified.h file with THUMB() and ARM() macros and doing the .syntax unified in that central place, and resyncing the various arch/arm/lib/_*.S files from Linux. https://github.com/afaerber/u-boot/commits/stm32
Is there a particular reason you are doing the .syntax inline here?
And looking beyond this patch, assuming this is for your Vybrid's M4, do you already have a concept for how to go about vectors.S, start.S and the like for ARMv7-M? :) I've stubbed them out for now but haven't managed to successfully link yet...
Regards, Andreas
Hi Andreas,
On 2014-12-15 15:24, Andreas Färber wrote:
Hi Stefan,
Am 03.12.2014 um 18:04 schrieb Stefan Agner:
Can this be fixed by the merger or should I create a new revision?
It looks as if this was neither applied nor respun? I have some more patches to make CONFIG_USE_PRIVATE_LIBGCC build for Thumb that I would like to rebase on it.
Not sure what the expectation is... Probably I should just do a respin since there are now three minor nits...
In particular, I have pieced together a tiny asm/unified.h file with THUMB() and ARM() macros and doing the .syntax unified in that central place, and resyncing the various arch/arm/lib/_*.S files from Linux. https://github.com/afaerber/u-boot/commits/stm32
Is there a particular reason you are doing the .syntax inline here?
No particular reason other than laziness :-)
And looking beyond this patch, assuming this is for your Vybrid's M4, do you already have a concept for how to go about vectors.S, start.S and the like for ARMv7-M? :) I've stubbed them out for now but haven't managed to successfully link yet...
Actually no, I don't use U-Boot on the M4 (so far). For Linux on the Vybrid M4, I use a special boot loader which runs on the A5 Linux. On Vybrid (the variant we use at least) boots always from the A5 first, hence I have a running Linux. Or alternatively, I could run the M4 from U-Boot running on the A5. So far, I had no need to run U-Boot on the M4.
The reason I was interested in Thumb2 was the size: Due to lack of 5V on the Colibri module we need to load the boot loader using Vybrid's serial loader over UART which is somewhat slow. So the 25% reduction of size means 25% less download time...
-- Stefan
Hello Stefan,
On Mon, 15 Dec 2014 22:06:11 +0100, Stefan Agner stefan@agner.ch wrote:
Hi Andreas,
On 2014-12-15 15:24, Andreas Färber wrote:
Hi Stefan,
Am 03.12.2014 um 18:04 schrieb Stefan Agner:
Can this be fixed by the merger or should I create a new revision?
It looks as if this was neither applied nor respun? I have some more patches to make CONFIG_USE_PRIVATE_LIBGCC build for Thumb that I would like to rebase on it.
Not sure what the expectation is... Probably I should just do a respin since there are now three minor nits...
Agreed, and assuming the three nits are cared for, I'd take it as a bugfix.
Amicalement,
participants (5)
-
Albert ARIBAUD -
Andreas Färber -
Bill Pringlemeir -
Simon Glass -
Stefan Agner