Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux

ARM: 7877/1: use built-in byte swap function

Enable the compiler intrinsic for byte swapping on arch ARM. This
allows the compiler to detect and be able to optimize out byte
swappings, and has a very modest benefit on vmlinux size (Linaro gcc
4.8):

text data bss dec hex filename
2840310 123932 61960 3026202 2e2d1a vmlinux-lart #orig
2840152 123932 61960 3026044 2e2c7c vmlinux-lart #builtin-bswap

6473120 314840 5616016 12403976 bd4508 vmlinux-mxs #orig
6472586 314848 5616016 12403450 bd42fa vmlinux-mxs #builtin-bswap

7419872 318372 379556 8117800 7bde28 vmlinux-imx_v6_v7 #orig
7419170 318364 379556 8117090 7bdb62 vmlinux-imx_v6_v7 #builtin-bswap

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Reviewed-by: Nicolas Pitre <nico@linaro.org>
Acked-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

authored by

Kim Phillips and committed by
Russell King
017f161a 03aa6580

+53 -5
+1
arch/arm/Kconfig
··· 6 6 select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST 7 7 select ARCH_HAVE_CUSTOM_GPIO_H 8 8 select ARCH_MIGHT_HAVE_PC_PARPORT 9 + select ARCH_USE_BUILTIN_BSWAP 9 10 select ARCH_USE_CMPXCHG_LOCKREF 10 11 select ARCH_WANT_IPC_PARSE_VERSION 11 12 select BUILDTIME_EXTABLE_SORT if MMU
+11 -4
arch/arm/boot/compressed/Makefile
··· 108 108 109 109 targets := vmlinux vmlinux.lds \ 110 110 piggy.$(suffix_y) piggy.$(suffix_y).o \ 111 - lib1funcs.o lib1funcs.S ashldi3.o ashldi3.S \ 112 - font.o font.c head.o misc.o $(OBJS) 111 + lib1funcs.o lib1funcs.S ashldi3.o ashldi3.S bswapsdi2.o \ 112 + bswapsdi2.S font.o font.c head.o misc.o $(OBJS) 113 113 114 114 # Make sure files are removed during clean 115 115 extra-y += piggy.gzip piggy.lzo piggy.lzma piggy.xzkern piggy.lz4 \ 116 - lib1funcs.S ashldi3.S $(libfdt) $(libfdt_hdrs) \ 116 + lib1funcs.S ashldi3.S bswapsdi2.S $(libfdt) $(libfdt_hdrs) \ 117 117 hyp-stub.S 118 118 119 119 ifeq ($(CONFIG_FUNCTION_TRACER),y) ··· 156 156 $(obj)/ashldi3.S: $(srctree)/arch/$(SRCARCH)/lib/ashldi3.S 157 157 $(call cmd,shipped) 158 158 159 + # For __bswapsi2, __bswapdi2 160 + bswapsdi2 = $(obj)/bswapsdi2.o 161 + 162 + $(obj)/bswapsdi2.S: $(srctree)/arch/$(SRCARCH)/lib/bswapsdi2.S 163 + $(call cmd,shipped) 164 + 159 165 # We need to prevent any GOTOFF relocs being used with references 160 166 # to symbols in the .bss section since we cannot relocate them 161 167 # independently from the rest at run time. This can be achieved by ··· 183 177 fi 184 178 185 179 $(obj)/vmlinux: $(obj)/vmlinux.lds $(obj)/$(HEAD) $(obj)/piggy.$(suffix_y).o \ 186 - $(addprefix $(obj)/, $(OBJS)) $(lib1funcs) $(ashldi3) FORCE 180 + $(addprefix $(obj)/, $(OBJS)) $(lib1funcs) $(ashldi3) \ 181 + $(bswapsdi2) FORCE 187 182 @$(check_for_multiple_zreladdr) 188 183 $(call if_changed,ld) 189 184 @$(check_for_bad_syms)
+4
arch/arm/kernel/armksyms.c
··· 35 35 extern void __udivsi3(void); 36 36 extern void __umodsi3(void); 37 37 extern void __do_div64(void); 38 + extern void __bswapsi2(void); 39 + extern void __bswapdi2(void); 38 40 39 41 extern void __aeabi_idiv(void); 40 42 extern void __aeabi_idivmod(void); ··· 116 114 EXPORT_SYMBOL(__udivsi3); 117 115 EXPORT_SYMBOL(__umodsi3); 118 116 EXPORT_SYMBOL(__do_div64); 117 + EXPORT_SYMBOL(__bswapsi2); 118 + EXPORT_SYMBOL(__bswapdi2); 119 119 120 120 #ifdef CONFIG_AEABI 121 121 EXPORT_SYMBOL(__aeabi_idiv);
+1 -1
arch/arm/lib/Makefile
··· 13 13 ashldi3.o ashrdi3.o lshrdi3.o muldi3.o \ 14 14 ucmpdi2.o lib1funcs.o div64.o \ 15 15 io-readsb.o io-writesb.o io-readsl.o io-writesl.o \ 16 - call_with_stack.o 16 + call_with_stack.o bswapsdi2.o 17 17 18 18 mmu-y := clear_user.o copy_page.o getuser.o putuser.o 19 19
+36
arch/arm/lib/bswapsdi2.S
··· 1 + #include <linux/linkage.h> 2 + 3 + #if __LINUX_ARM_ARCH__ >= 6 4 + ENTRY(__bswapsi2) 5 + rev r0, r0 6 + bx lr 7 + ENDPROC(__bswapsi2) 8 + 9 + ENTRY(__bswapdi2) 10 + rev r3, r0 11 + rev r0, r1 12 + mov r1, r3 13 + bx lr 14 + ENDPROC(__bswapdi2) 15 + #else 16 + ENTRY(__bswapsi2) 17 + eor r3, r0, r0, ror #16 18 + mov r3, r3, lsr #8 19 + bic r3, r3, #0xff00 20 + eor r0, r3, r0, ror #8 21 + mov pc, lr 22 + ENDPROC(__bswapsi2) 23 + 24 + ENTRY(__bswapdi2) 25 + mov ip, r1 26 + eor r3, ip, ip, ror #16 27 + eor r1, r0, r0, ror #16 28 + mov r1, r1, lsr #8 29 + mov r3, r3, lsr #8 30 + bic r3, r3, #0xff00 31 + bic r1, r1, #0xff00 32 + eor r1, r1, r0, ror #8 33 + eor r0, r3, ip, ror #8 34 + mov pc, lr 35 + ENDPROC(__bswapdi2) 36 + #endif