New patch to review for coreboot: dca49b7 ARM: Use local versions of libgcc functions instead of linking against libgcc.
Marc Jones (marc.jones@se-eng.com) just uploaded a new patch set to gerrit, which you can find at http://review.coreboot.org/7322
-gerrit
commit dca49b7d0856343d0a9c4de588b5437a37d8a760 Author: Gabe Black gabeblack@google.com Date: Tue Oct 8 18:24:10 2013 -0700
ARM: Use local versions of libgcc functions instead of linking against libgcc.
The flags used to compile libgcc may make it incompatible with the code it's linked against, and/or the hardware it's going to run on. Rather than try to tease the right libgcc from the compiler, lets just leave it out and use our own implementations of the necessary functions.
Most of these implementations were taken from the Linux kernel, except for uldivmod.S which was taken from a CL originally written for U-Boot by Che-Liang Chiou in December of 2010. It was modified to not use the CLZ instruction on machines that don't have it, anything earlier than ARMv5. The top block was taken from an earlier version of the same CL which didn't use CLZ in that spot. The later block was written from scratch.
BUG=None TEST=Built and booted into the bootblock on nyan. Ran a series of tests which divided and modded a 64 bit value by various 32 bit values which were powers of 2. Confirmed that this function was used and that the returned value was correct. Printed decimal and hex versions of some values and verified that they equaled each other. Built and booted on pit with serial enabled. BRANCH=None
Original-Change-Id: I7527e28af411b7aa7f94579be95a6b352a91a224 Original-Signed-off-by: Gabe Black gabeblack@google.com Original-Reviewed-on: https://chromium-review.googlesource.com/172401 Original-Reviewed-by: David Hendricks dhendrix@chromium.org Original-Commit-Queue: Gabe Black gabeblack@chromium.org Original-Tested-by: Gabe Black gabeblack@chromium.org (cherry picked from commit be8c7a8f3292a7d7651b7c6dafc9a2c53afbd402) Signed-off-by: Marc Jones marc.jones@se-eng.com
Change-Id: Ie3e48f34ebf6fbe20c3dd76ecbcbea7844e9466e --- src/arch/arm/Makefile.inc | 18 +- src/arch/arm/libgcc/Makefile.inc | 26 +++ src/arch/arm/libgcc/ashldi3.S | 57 +++++++ src/arch/arm/libgcc/bpabi.c | 214 ++++++++++++++++++++++++ src/arch/arm/libgcc/bpabi_asm.S | 130 +++++++++++++++ src/arch/arm/libgcc/lib1funcs.S | 344 +++++++++++++++++++++++++++++++++++++++ src/arch/arm/libgcc/lshrdi3.S | 57 +++++++ src/arch/arm/libgcc/macros.h | 47 ++++++ src/arch/arm/libgcc/muldi3.S | 51 ++++++ src/arch/arm/libgcc/ucmpdi2.S | 51 ++++++ src/arch/arm/libgcc/uldivmod.S | 260 +++++++++++++++++++++++++++++ 11 files changed, 1247 insertions(+), 8 deletions(-)
diff --git a/src/arch/arm/Makefile.inc b/src/arch/arm/Makefile.inc index 427e3a4..8044dba 100644 --- a/src/arch/arm/Makefile.inc +++ b/src/arch/arm/Makefile.inc @@ -2,7 +2,7 @@ ## ## This file is part of the coreboot project. ## -## Copyright (C) 2012 The ChromiumOS Authors +## Copyright (C) 2012-2013 The ChromiumOS Authors ## Copyright (C) 2012 Alexandru Gagniuc mr.nuke.me@gmail.com ## Copyright (C) 2009-2010 coresystems GmbH ## Copyright (C) 2009 Ronald G. Minnich @@ -23,6 +23,7 @@ ############################################################################### # Take care of subdirectories ############################################################################### +subdirs-y += libgcc/ subdirs-y += armv4/ armv7/
############################################################################### @@ -58,13 +59,14 @@ bootblock-y += eabi_compat.c bootblock-y += memset.S bootblock-y += memcpy.S bootblock-y += memmove.S +bootblock-y += div0.c
$(objcbfs)/bootblock.debug: $(src)/arch/arm/bootblock.ld $(obj)/ldoptions $$(bootblock-objs) @printf " LINK $(subst $(obj)/,,$(@))\n" ifeq ($(CONFIG_COMPILER_LLVM_CLANG),y) $(LD_bootblock) -m armelf_linux_eabi --gc-sections -static -o $@ -L$(obj) $< -T $(src)/arch/arm/bootblock.ld else - $(CC_bootblock) $(CFLAGS_bootblock) -nostartfiles -Wl,--gc-sections -static -o $@ -L$(obj) -T $(src)/arch/arm/bootblock.ld -Wl,--start-group $(bootblock-objs) $(LIBGCC_FILE_NAME_bootblock) -Wl,--end-group + $(CC_bootblock) $(CFLAGS_bootblock) -nostartfiles -Wl,--gc-sections -static -o $@ -L$(obj) -T $(src)/arch/arm/bootblock.ld -Wl,--start-group $(bootblock-objs) -Wl,--end-group endif
endif # CONFIG_ARCH_BOOTBLOCK_ARM @@ -89,7 +91,7 @@ $(objcbfs)/romstage.debug: $$(romstage-objs) $(src)/arch/arm/romstage.ld $(obj)/ ifeq ($(CONFIG_COMPILER_LLVM_CLANG),y) $(LD_romstage) -nostdlib -nostartfiles --gc-sections -static -o $@ -L$(obj) $(romstage-objs) -T $(src)/arch/arm/romstage.ld else - $(CC_romstage) $(CFLAGS_romstage) -nostartfiles -Wl,--gc-sections -static -o $@ -L$(obj) -T $(src)/arch/arm/romstage.ld -Wl,--start-group $(romstage-objs) $(LIBGCC_FILE_NAME_romstage) -Wl,--end-group + $(CC_romstage) $(CFLAGS_romstage) -nostartfiles -Wl,--gc-sections -static -o $@ -L$(obj) -T $(src)/arch/arm/romstage.ld -Wl,--start-group $(romstage-objs) -Wl,--end-group endif
endif # CONFIG_ARCH_ROMSTAGE_ARM @@ -111,20 +113,20 @@ ramstage-y += memcpy.S ramstage-y += memmove.S ramstage-srcs += $(wildcard src/mainboard/$(MAINBOARDDIR)/mainboard.c)
-$(objcbfs)/ramstage.debug: $$(ramstage-objs) $(LIBGCC_FILE_NAME_ramstage) $(src)/arch/arm/ramstage.ld $(obj)/ldoptions +$(objcbfs)/ramstage.debug: $$(ramstage-objs) $(src)/arch/arm/ramstage.ld $(obj)/ldoptions @printf " CC $(subst $(obj)/,,$(@))\n" ifeq ($(CONFIG_COMPILER_LLVM_CLANG),y) $(LD_ramstage) -m armelf_linux_eabi --gc-sections -o $@ -L$(obj) $< -T $(src)/arch/arm/ramstage.ld else - $(CC_ramstage) $(CFLAGS_ramstage) -nostdlib -Wl,--gc-sections -static -o $@ -L$(obj) -Wl,--start-group $(ramstage-objs) $(LIBGCC_FILE_NAME_ramstage) -Wl,--end-group -T $(src)/arch/arm/ramstage.ld + $(CC_ramstage) $(CFLAGS_ramstage) -nostdlib -Wl,--gc-sections -static -o $@ -L$(obj) -Wl,--start-group $(ramstage-objs) -Wl,--end-group -T $(src)/arch/arm/ramstage.ld endif
-$(objgenerated)/ramstage.o: $(stages_o) $$(ramstage-objs) $(LIBGCC_FILE_NAME_ramstage) +$(objgenerated)/ramstage.o: $(stages_o) $$(ramstage-objs) @printf " CC $(subst $(obj)/,,$(@))\n" ifeq ($(CONFIG_COMPILER_LLVM_CLANG),y) - $(LD_ramstage) -m -m armelf_linux_eabi --gc-sections -r -o $@ --wrap __divdi3 --wrap __udivdi3 --wrap __moddi3 --wrap __umoddi3 --wrap __uidiv --start-group $(ramstage-objs) $(LIBGCC_FILE_NAME_ramstage) --end-group + $(LD_ramstage) -m -m armelf_linux_eabi --gc-sections -r -o $@ --wrap __divdi3 --wrap __udivdi3 --wrap __moddi3 --wrap __umoddi3 --wrap __uidiv --start-group $(ramstage-objs) --end-group else - $(CC_ramstage) $(CFLAGS_ramstage) $(CPPFLAGS_ramstage) -nostdlib -Wl,--gc-sections -r -o $@ -Wl,--start-group $(stages_o) $(ramstage-objs) $(LIBGCC_FILE_NAME_ramstage) -Wl,--end-group + $(CC_ramstage) $(CFLAGS_ramstage) $(CPPFLAGS_ramstage) -nostdlib -Wl,--gc-sections -r -o $@ -Wl,--start-group $(stages_o) $(ramstage-objs) -Wl,--end-group endif
endif # CONFIG_ARCH_RAMSTAGE_ARM diff --git a/src/arch/arm/libgcc/Makefile.inc b/src/arch/arm/libgcc/Makefile.inc new file mode 100644 index 0000000..d914afd --- /dev/null +++ b/src/arch/arm/libgcc/Makefile.inc @@ -0,0 +1,26 @@ +################################################################################ +## +## This file is part of the coreboot project. +## +## Copyright (C) 2013 The ChromiumOS Authors +## +## This program is free software; you can redistribute it and/or modify +## it under the terms of the GNU General Public License as published by +## the Free Software Foundation; version 2 of the License. +## +## This program is distributed in the hope that it will be useful, +## but WITHOUT ANY WARRANTY; without even the implied warranty of +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +## GNU General Public License for more details. +## +## You should have received a copy of the GNU General Public License +## along with this program; if not, write to the Free Software +## Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA +## +################################################################################ + +libgcc_files = ashldi3.S lib1funcs.S lshrdi3.S muldi3.S ucmpdi2.S uldivmod.S + +bootblock-y += $(libgcc_files) +romstage-y += $(libgcc_files) +ramstage-y += $(libgcc_files) diff --git a/src/arch/arm/libgcc/ashldi3.S b/src/arch/arm/libgcc/ashldi3.S new file mode 100644 index 0000000..4a9b97c --- /dev/null +++ b/src/arch/arm/libgcc/ashldi3.S @@ -0,0 +1,57 @@ +/* Copyright 1995, 1996, 1998, 1999, 2000, 2003, 2004, 2005 + Free Software Foundation, Inc. + +This file is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 2, or (at your option) any +later version. + +In addition to the permissions in the GNU General Public License, the +Free Software Foundation gives you unlimited permission to link the +compiled version of this file into combinations with other programs, +and to distribute those combinations without any restriction coming +from the use of this file. (The General Public License restrictions +do apply in other respects; for example, they cover modification of +the file, and distribution when not linked into a combine +executable.) + +This file is distributed in the hope that it will be useful, but +WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING. If not, write to +the Free Software Foundation, 51 Franklin Street, Fifth Floor, +Boston, MA 02110-1301, USA. */ + + +#if defined __GNUC__ + +#include "macros.h" + +#ifdef __ARMEB__ +#define al r1 +#define ah r0 +#else +#define al r0 +#define ah r1 +#endif + +ENTRY(__ashldi3) +ENTRY(__aeabi_llsl) + + subs r3, r2, #32 + rsb ip, r2, #32 + movmi ah, ah, lsl r2 + movpl ah, al, lsl r3 + ARM( orrmi ah, ah, al, lsr ip ) + THUMB( lsrmi r3, al, ip ) + THUMB( orrmi ah, ah, r3 ) + mov al, al, lsl r2 + mov pc, lr + +ENDPROC(__ashldi3) +ENDPROC(__aeabi_llsl) + +#endif diff --git a/src/arch/arm/libgcc/bpabi.c b/src/arch/arm/libgcc/bpabi.c new file mode 100644 index 0000000..8cb7ddf --- /dev/null +++ b/src/arch/arm/libgcc/bpabi.c @@ -0,0 +1,214 @@ +/* Miscellaneous BPABI functions. + + Copyright (C) 2003, 2004 Free Software Foundation, Inc. + Contributed by CodeSourcery, LLC. + + This file is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + In addition to the permissions in the GNU General Public License, the + Free Software Foundation gives you unlimited permission to link the + compiled version of this file into combinations with other programs, + and to distribute those combinations without any restriction coming + from the use of this file. (The General Public License restrictions + do apply in other respects; for example, they cover modification of + the file, and distribution when not linked into a combine + executable.) + + This file is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. If not, write to + the Free Software Foundation, 51 Franklin Street, Fifth Floor, + Boston, MA 02110-1301, USA. */ + +#if defined __GNUC__ + +#include <stdint.h> + +uint64_t __udivmoddi4(uint64_t n, uint64_t d, uint64_t *rp); +extern int64_t __divdi3(int64_t, int64_t); +extern uint64_t __udivdi3(uint64_t, uint64_t); +extern int64_t __gnu_ldivmod_helper(int64_t, int64_t, int64_t *); +extern uint64_t __gnu_uldivmod_helper(uint64_t, uint64_t, uint64_t *); + +typedef union +{ + struct { + int32_t low; + int32_t high; + } s; + int64_t ll; +} DWunion; + +uint64_t +__udivmoddi4(uint64_t n, uint64_t d, uint64_t *rp) +{ + const DWunion nn = {.ll = n}; + const DWunion dd = {.ll = d}; + DWunion rr; + uint32_t d0, d1, n0, n1, n2; + uint32_t q0, q1; + uint32_t b, bm; + + d0 = dd.s.low; + d1 = dd.s.high; + n0 = nn.s.low; + n1 = nn.s.high; + + if (d1 == 0) { + if (d0 > n1) { + /* 0q = nn / 0D */ + udiv_qrnnd(q0, n0, n1, n0, d0); + q1 = 0; + /* Remainder in n0. */ + } else { + /* qq = NN / 0d */ + if (d0 == 0) + d0 = 1 / d0; /* Divide intentionally by zero. */ + + udiv_qrnnd(q1, n1, 0, n1, d0); + udiv_qrnnd(q0, n0, n1, n0, d0); + + /* Remainder in n0. */ + } + + if (rp != 0) { + rr.s.low = n0; + rr.s.high = 0; + *rp = rr.ll; + } + } else { + if (d1 > n1) { + /* 00 = nn / DD */ + q0 = 0; + q1 = 0; + + /* Remainder in n1n0. */ + if (rp != 0) { + rr.s.low = n0; + rr.s.high = n1; + *rp = rr.ll; + } + } else { + /* 0q = NN / dd */ + + count_leading_zeros(bm, d1); + if (bm == 0) { + /* From (n1 >= d1) /\ (the most significant + bit of d1 is set), conclude (the most + significant bit of n1 is set) /\ (the + quotient digit q0 = 0 or 1). + + This special case is necessary, not an + optimization. */ + + /* The condition on the next line takes + advantage of that n1 >= d1 (true due to + program flow). */ + if (n1 > d1 || n0 >= d0) { + q0 = 1; + sub_ddmmss(n1, n0, n1, n0, d1, d0); + } else + q0 = 0; + + q1 = 0; + + if (rp != 0) { + rr.s.low = n0; + rr.s.high = n1; + *rp = rr.ll; + } + } else { + uint32_t m1, m0; + /* Normalize. */ + + b = 32 - bm; + + d1 = (d1 << bm) | (d0 >> b); + d0 = d0 << bm; + n2 = n1 >> b; + n1 = (n1 << bm) | (n0 >> b); + n0 = n0 << bm; + + udiv_qrnnd(q0, n1, n2, n1, d1); + umul_ppmm(m1, m0, q0, d0); + + if (m1 > n1 || (m1 == n1 && m0 > n0)) { + q0--; + sub_ddmmss(m1, m0, m1, m0, d1, d0); + } + + q1 = 0; + + /* Remainder in (n1n0 - m1m0) >> bm. */ + if (rp != 0) { + sub_ddmmss(n1, n0, n1, n0, m1, m0); + rr.s.low = (n1 << b) | (n0 >> bm); + rr.s.high = n1 >> bm; + *rp = rr.ll; + } + } + } + } + + const DWunion ww = {{.low = q0, .high = q1}}; + return ww.ll; +} + +int64_t +__divdi3(int64_t u, int64_t v) +{ + int32_t c = 0; + DWunion uu = {.ll = u}; + DWunion vv = {.ll = v}; + int64_t w; + + if (uu.s.high < 0) { + c = ~c; + uu.ll = -uu.ll; + } + if (vv.s.high < 0) { + c = ~c; + vv.ll = -vv.ll; + } + + w = __udivmoddi4(uu.ll, vv.ll, (uint64_t *)0); + if (c) + w = -w; + + return w; +} + +int64_t +__gnu_ldivmod_helper (int64_t a, int64_t b, int64_t *remainder) +{ + int64_t quotient; + + quotient = __divdi3(a, b); + *remainder = a - b * quotient; + return quotient; +} + +uint64_t +__udivdi3(uint64_t n, uint64_t d) +{ + return __udivmoddi4(n, d, (uint64_t *)0); +} + +uint64_t +__gnu_uldivmod_helper(uint64_t a, uint64_t b, uint64_t *remainder) +{ + uint64_t quotient; + + quotient = __udivdi3(a, b); + *remainder = a - b * quotient; + return quotient; +} + +#endif diff --git a/src/arch/arm/libgcc/bpabi_asm.S b/src/arch/arm/libgcc/bpabi_asm.S new file mode 100644 index 0000000..a3942df --- /dev/null +++ b/src/arch/arm/libgcc/bpabi_asm.S @@ -0,0 +1,130 @@ +/* Miscellaneous BPABI functions. + + Copyright (C) 2003, 2004, 2007, 2008 Free Software Foundation, Inc. + Contributed by CodeSourcery, LLC. + + This file is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + In addition to the permissions in the GNU General Public License, the + Free Software Foundation gives you unlimited permission to link the + compiled version of this file into combinations with other programs, + and to distribute those combinations without any restriction coming + from the use of this file. (The General Public License restrictions + do apply in other respects; for example, they cover modification of + the file, and distribution when not linked into a combine + executable.) + + This file is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. If not, write to + the Free Software Foundation, 51 Franklin Street, Fifth Floor, + Boston, MA 02110-1301, USA. */ + + +#include "macros.h" + +#ifdef __ARMEB__ +#define xxh r0 +#define xxl r1 +#define yyh r2 +#define yyl r3 +#else +#define xxh r1 +#define xxl r0 +#define yyh r3 +#define yyl r2 +#endif + +#if defined __thumb2__ + +.macro do_it cond + it \cond +.endm +#define do_push push +#define do_pop pop + +#else + +.macro do_it cond +.endm +#define do_push stmfd sp!, +#define do_pop ldmfd sp!, + +#endif + +ENTRY(__aeabi_lcmp) + + cmp xxh, yyh + do_it lt + movlt r0, #-1 + do_it gt + movgt r0, #1 + do_it ne + movne pc, lr + subs r0, xxl, yyl + do_it lo + movlo r0, #-1 + do_it hi + movhi r0, #1 + mov pc, lr + +ENDPROC(__aeabi_lcmp) + +ENTRY(__aeabi_ulcmp) + + cmp xxh, yyh + do_it lo + movlo r0, #-1 + do_it hi + movhi r0, #1 + do_it ne + movne pc, lr + cmp xxl, yyl + do_it lo + movlo r0, #-1 + do_it hi + movhi r0, #1 + do_it eq + moveq r0, #0 + mov pc, lr + +ENDPROC(__aeabi_ulcmp) + +ENTRY(__aeabi_ldivmod) + sub sp, sp, #8 +#if defined(__thumb2__) + mov ip, sp + push {ip, lr} +#else + do_push {sp, lr} +#endif + bl __gnu_ldivmod_helper + ldr lr, [sp, #4] + add sp, sp, #8 + do_pop {r2, r3} + mov pc, lr + +ENDPROC(__aeabi_ldivmod) + +ENTRY(__aeabi_uldivmod) + sub sp, sp, #8 +#if defined(__thumb2__) + mov ip, sp + push {ip, lr} +#else + do_push {sp, lr} +#endif + bl __gnu_uldivmod_helper + ldr lr, [sp, #4] + add sp, sp, #8 + do_pop {r2, r3} + mov pc, lr + +ENDPROC(__aeabi_uldivmod) diff --git a/src/arch/arm/libgcc/lib1funcs.S b/src/arch/arm/libgcc/lib1funcs.S new file mode 100644 index 0000000..bdef583 --- /dev/null +++ b/src/arch/arm/libgcc/lib1funcs.S @@ -0,0 +1,344 @@ +/* + * linux/arch/arm/lib/lib1funcs.S: Optimized ARM division routines + * + * Author: Nicolas Pitre nico@fluxnic.net + * - contributed to gcc-3.4 on Sep 30, 2003 + * - adapted for the Linux kernel on Oct 2, 2003 + */ + +/* Copyright 1995, 1996, 1998, 1999, 2000, 2003 Free Software Foundation, Inc. + +This file is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 2, or (at your option) any +later version. + +In addition to the permissions in the GNU General Public License, the +Free Software Foundation gives you unlimited permission to link the +compiled version of this file into combinations with other programs, +and to distribute those combinations without any restriction coming +from the use of this file. (The General Public License restrictions +do apply in other respects; for example, they cover modification of +the file, and distribution when not linked into a combine +executable.) + +This file is distributed in the hope that it will be useful, but +WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING. If not, write to +the Free Software Foundation, 59 Temple Place - Suite 330, +Boston, MA 02111-1307, USA. */ + + +#if defined __GNUC__ + +#include "macros.h" + +.macro ARM_DIV_BODY dividend, divisor, result, curbit + +#if __COREBOOT_ARM_ARCH__ >= 5 + + clz \curbit, \divisor + clz \result, \dividend + sub \result, \curbit, \result + mov \curbit, #1 + mov \divisor, \divisor, lsl \result + mov \curbit, \curbit, lsl \result + mov \result, #0 + +#else + + @ Initially shift the divisor left 3 bits if possible, + @ set curbit accordingly. This allows for curbit to be located + @ at the left end of each 4 bit nibbles in the division loop + @ to save one loop in most cases. + tst \divisor, #0xe0000000 + moveq \divisor, \divisor, lsl #3 + moveq \curbit, #8 + movne \curbit, #1 + + @ Unless the divisor is very big, shift it up in multiples of + @ four bits, since this is the amount of unwinding in the main + @ division loop. Continue shifting until the divisor is + @ larger than the dividend. +1: cmp \divisor, #0x10000000 + cmplo \divisor, \dividend + movlo \divisor, \divisor, lsl #4 + movlo \curbit, \curbit, lsl #4 + blo 1b + + @ For very big divisors, we must shift it a bit at a time, or + @ we will be in danger of overflowing. +1: cmp \divisor, #0x80000000 + cmplo \divisor, \dividend + movlo \divisor, \divisor, lsl #1 + movlo \curbit, \curbit, lsl #1 + blo 1b + + mov \result, #0 + +#endif + + @ Division loop +1: cmp \dividend, \divisor + subhs \dividend, \dividend, \divisor + orrhs \result, \result, \curbit + cmp \dividend, \divisor, lsr #1 + subhs \dividend, \dividend, \divisor, lsr #1 + orrhs \result, \result, \curbit, lsr #1 + cmp \dividend, \divisor, lsr #2 + subhs \dividend, \dividend, \divisor, lsr #2 + orrhs \result, \result, \curbit, lsr #2 + cmp \dividend, \divisor, lsr #3 + subhs \dividend, \dividend, \divisor, lsr #3 + orrhs \result, \result, \curbit, lsr #3 + cmp \dividend, #0 @ Early termination? + movnes \curbit, \curbit, lsr #4 @ No, any more bits to do? + movne \divisor, \divisor, lsr #4 + bne 1b + +.endm + + +.macro ARM_DIV2_ORDER divisor, order + +#if __COREBOOT_ARM_ARCH__ >= 5 + + clz \order, \divisor + rsb \order, \order, #31 + +#else + + cmp \divisor, #(1 << 16) + movhs \divisor, \divisor, lsr #16 + movhs \order, #16 + movlo \order, #0 + + cmp \divisor, #(1 << 8) + movhs \divisor, \divisor, lsr #8 + addhs \order, \order, #8 + + cmp \divisor, #(1 << 4) + movhs \divisor, \divisor, lsr #4 + addhs \order, \order, #4 + + cmp \divisor, #(1 << 2) + addhi \order, \order, #3 + addls \order, \order, \divisor, lsr #1 + +#endif + +.endm + + +.macro ARM_MOD_BODY dividend, divisor, order, spare + +#if __COREBOOT_ARM_ARCH__ >= 5 + + clz \order, \divisor + clz \spare, \dividend + sub \order, \order, \spare + mov \divisor, \divisor, lsl \order + +#else + + mov \order, #0 + + @ Unless the divisor is very big, shift it up in multiples of + @ four bits, since this is the amount of unwinding in the main + @ division loop. Continue shifting until the divisor is + @ larger than the dividend. +1: cmp \divisor, #0x10000000 + cmplo \divisor, \dividend + movlo \divisor, \divisor, lsl #4 + addlo \order, \order, #4 + blo 1b + + @ For very big divisors, we must shift it a bit at a time, or + @ we will be in danger of overflowing. +1: cmp \divisor, #0x80000000 + cmplo \divisor, \dividend + movlo \divisor, \divisor, lsl #1 + addlo \order, \order, #1 + blo 1b + +#endif + + @ Perform all needed substractions to keep only the reminder. + @ Do comparisons in batch of 4 first. + subs \order, \order, #3 @ yes, 3 is intended here + blt 2f + +1: cmp \dividend, \divisor + subhs \dividend, \dividend, \divisor + cmp \dividend, \divisor, lsr #1 + subhs \dividend, \dividend, \divisor, lsr #1 + cmp \dividend, \divisor, lsr #2 + subhs \dividend, \dividend, \divisor, lsr #2 + cmp \dividend, \divisor, lsr #3 + subhs \dividend, \dividend, \divisor, lsr #3 + cmp \dividend, #1 + mov \divisor, \divisor, lsr #4 + subges \order, \order, #4 + bge 1b + + tst \order, #3 + teqne \dividend, #0 + beq 5f + + @ Either 1, 2 or 3 comparison/substractions are left. +2: cmn \order, #2 + blt 4f + beq 3f + cmp \dividend, \divisor + subhs \dividend, \dividend, \divisor + mov \divisor, \divisor, lsr #1 +3: cmp \dividend, \divisor + subhs \dividend, \dividend, \divisor + mov \divisor, \divisor, lsr #1 +4: cmp \dividend, \divisor + subhs \dividend, \dividend, \divisor +5: +.endm + + +ENTRY(__udivsi3) +ENTRY(__aeabi_uidiv) + + subs r2, r1, #1 + moveq pc, lr + bcc Ldiv0 + cmp r0, r1 + bls 11f + tst r1, r2 + beq 12f + + ARM_DIV_BODY r0, r1, r2, r3 + + mov r0, r2 + mov pc, lr + +11: moveq r0, #1 + movne r0, #0 + mov pc, lr + +12: ARM_DIV2_ORDER r1, r2 + + mov r0, r0, lsr r2 + mov pc, lr + +ENDPROC(__udivsi3) +ENDPROC(__aeabi_uidiv) + +ENTRY(__umodsi3) + + subs r2, r1, #1 @ compare divisor with 1 + bcc Ldiv0 + cmpne r0, r1 @ compare dividend with divisor + moveq r0, #0 + tsthi r1, r2 @ see if divisor is power of 2 + andeq r0, r0, r2 + movls pc, lr + + ARM_MOD_BODY r0, r1, r2, r3 + + mov pc, lr + +ENDPROC(__umodsi3) + +ENTRY(__divsi3) +ENTRY(__aeabi_idiv) + + cmp r1, #0 + eor ip, r0, r1 @ save the sign of the result. + beq Ldiv0 + rsbmi r1, r1, #0 @ loops below use unsigned. + subs r2, r1, #1 @ division by 1 or -1 ? + beq 10f + movs r3, r0 + rsbmi r3, r0, #0 @ positive dividend value + cmp r3, r1 + bls 11f + tst r1, r2 @ divisor is power of 2 ? + beq 12f + + ARM_DIV_BODY r3, r1, r0, r2 + + cmp ip, #0 + rsbmi r0, r0, #0 + mov pc, lr + +10: teq ip, r0 @ same sign ? + rsbmi r0, r0, #0 + mov pc, lr + +11: movlo r0, #0 + moveq r0, ip, asr #31 + orreq r0, r0, #1 + mov pc, lr + +12: ARM_DIV2_ORDER r1, r2 + + cmp ip, #0 + mov r0, r3, lsr r2 + rsbmi r0, r0, #0 + mov pc, lr + +ENDPROC(__divsi3) +ENDPROC(__aeabi_idiv) + +ENTRY(__modsi3) + + cmp r1, #0 + beq Ldiv0 + rsbmi r1, r1, #0 @ loops below use unsigned. + movs ip, r0 @ preserve sign of dividend + rsbmi r0, r0, #0 @ if negative make positive + subs r2, r1, #1 @ compare divisor with 1 + cmpne r0, r1 @ compare dividend with divisor + moveq r0, #0 + tsthi r1, r2 @ see if divisor is power of 2 + andeq r0, r0, r2 + bls 10f + + ARM_MOD_BODY r0, r1, r2, r3 + +10: cmp ip, #0 + rsbmi r0, r0, #0 + mov pc, lr + +ENDPROC(__modsi3) + +ENTRY(__aeabi_uidivmod) + + stmfd sp!, {r0, r1, ip, lr} + bl __aeabi_uidiv + ldmfd sp!, {r1, r2, ip, lr} + mul r3, r0, r2 + sub r1, r1, r3 + mov pc, lr + +ENDPROC(__aeabi_uidivmod) + +ENTRY(__aeabi_idivmod) + stmfd sp!, {r0, r1, ip, lr} + bl __aeabi_idiv + ldmfd sp!, {r1, r2, ip, lr} + mul r3, r0, r2 + sub r1, r1, r3 + mov pc, lr + +ENDPROC(__aeabi_idivmod) + + +Ldiv0: + str lr, [sp, #-8]! + bl __div0 + mov r0, #0 @ About as wrong as it could be. + ldr pc, [sp], #8 +ENDPROC(Ldiv0) + +#endif diff --git a/src/arch/arm/libgcc/lshrdi3.S b/src/arch/arm/libgcc/lshrdi3.S new file mode 100644 index 0000000..cdf9095 --- /dev/null +++ b/src/arch/arm/libgcc/lshrdi3.S @@ -0,0 +1,57 @@ +/* Copyright 1995, 1996, 1998, 1999, 2000, 2003, 2004, 2005 + Free Software Foundation, Inc. + +This file is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 2, or (at your option) any +later version. + +In addition to the permissions in the GNU General Public License, the +Free Software Foundation gives you unlimited permission to link the +compiled version of this file into combinations with other programs, +and to distribute those combinations without any restriction coming +from the use of this file. (The General Public License restrictions +do apply in other respects; for example, they cover modification of +the file, and distribution when not linked into a combine +executable.) + +This file is distributed in the hope that it will be useful, but +WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING. If not, write to +the Free Software Foundation, 51 Franklin Street, Fifth Floor, +Boston, MA 02110-1301, USA. */ + + +#if defined __GNUC__ + +#include "macros.h" + +#ifdef __ARMEB__ +#define al r1 +#define ah r0 +#else +#define al r0 +#define ah r1 +#endif + +ENTRY(__lshrdi3) +ENTRY(__aeabi_llsr) + + subs r3, r2, #32 + rsb ip, r2, #32 + movmi al, al, lsr r2 + movpl al, ah, lsr r3 + ARM( orrmi al, al, ah, lsl ip ) + THUMB( lslmi r3, ah, ip ) + THUMB( orrmi al, al, r3 ) + mov ah, ah, lsr r2 + mov pc, lr + +ENDPROC(__lshrdi3) +ENDPROC(__aeabi_llsr) + +#endif diff --git a/src/arch/arm/libgcc/macros.h b/src/arch/arm/libgcc/macros.h new file mode 100644 index 0000000..1e9e96c --- /dev/null +++ b/src/arch/arm/libgcc/macros.h @@ -0,0 +1,47 @@ +/* + * This file is part of the coreboot project. + * + * Copyright 2013 Google Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#ifndef __ARCH_ARM_LIBGCC_H__ +#define __ARCH_ARM_LIBGCC_H__ + +#if defined __arm__ +# define ARM(x...) x +# define THUMB(x...) +#elif defined __thumb__ +# define ARM(x...) +# define THUMB(x...) x +#else +# error Not in ARM or thumb mode! +#endif + +#define ALIGN .align 0 + +#define ENDPROC(name) \ + .type name, %function; \ + END(name) + +#define ENTRY(name) \ + .globl name; \ + ALIGN; \ + name: + +#define END(name) \ + .size name, .-name + +#endif /* __ARCH_ARM_LIBGCC_H__ */ diff --git a/src/arch/arm/libgcc/muldi3.S b/src/arch/arm/libgcc/muldi3.S new file mode 100644 index 0000000..faa92b5 --- /dev/null +++ b/src/arch/arm/libgcc/muldi3.S @@ -0,0 +1,51 @@ +/* + * linux/arch/arm/lib/muldi3.S + * + * Author: Nicolas Pitre + * Created: Oct 19, 2005 + * Copyright: Monta Vista Software, Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#if defined __GNUC__ + +#include "macros.h" + +#ifdef __ARMEB__ +#define xh r0 +#define xl r1 +#define yh r2 +#define yl r3 +#else +#define xl r0 +#define xh r1 +#define yl r2 +#define yh r3 +#endif + +ENTRY(__muldi3) +ENTRY(__aeabi_lmul) + + mul xh, yl, xh + mla xh, xl, yh, xh + mov ip, xl, lsr #16 + mov yh, yl, lsr #16 + bic xl, xl, ip, lsl #16 + bic yl, yl, yh, lsl #16 + mla xh, yh, ip, xh + mul yh, xl, yh + mul xl, yl, xl + mul ip, yl, ip + adds xl, xl, yh, lsl #16 + adc xh, xh, yh, lsr #16 + adds xl, xl, ip, lsl #16 + adc xh, xh, ip, lsr #16 + mov pc, lr + +ENDPROC(__muldi3) +ENDPROC(__aeabi_lmul) + +#endif diff --git a/src/arch/arm/libgcc/ucmpdi2.S b/src/arch/arm/libgcc/ucmpdi2.S new file mode 100644 index 0000000..7190b9f --- /dev/null +++ b/src/arch/arm/libgcc/ucmpdi2.S @@ -0,0 +1,51 @@ +/* + * linux/arch/arm/lib/ucmpdi2.S + * + * Author: Nicolas Pitre + * Created: Oct 19, 2005 + * Copyright: Monta Vista Software, Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#if defined __GNUC__ + +#include "macros.h" + +#ifdef __ARMEB__ +#define xh r0 +#define xl r1 +#define yh r2 +#define yl r3 +#else +#define xl r0 +#define xh r1 +#define yl r2 +#define yh r3 +#endif + +ENTRY(__ucmpdi2) + + cmp xh, yh + cmpeq xl, yl + movlo r0, #0 + moveq r0, #1 + movhi r0, #2 + mov pc, lr + +ENDPROC(__ucmpdi2) + +ENTRY(__aeabi_ulcmp) + + cmp xh, yh + cmpeq xl, yl + movlo r0, #-1 + moveq r0, #0 + movhi r0, #1 + mov pc, lr + +ENDPROC(__aeabi_ulcmp) + +#endif diff --git a/src/arch/arm/libgcc/uldivmod.S b/src/arch/arm/libgcc/uldivmod.S new file mode 100644 index 0000000..e4704c2 --- /dev/null +++ b/src/arch/arm/libgcc/uldivmod.S @@ -0,0 +1,260 @@ +/* + * Copyright 2010, Google Inc. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following disclaimer + * in the documentation and/or other materials provided with the + * distribution. + * Neither the name of Google Inc. nor the names of its + * contributors may be used to endorse or promote products derived from + * this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL") version 2 as published by the Free + * Software Foundation. + */ + +/* + * A, Q = r0 + (r1 << 32) + * B, R = r2 + (r3 << 32) + * A / B = Q ... R + */ + + .text + .global __aeabi_uldivmod + .type __aeabi_uldivmod, function + .align 0 + +A_0 .req r0 +A_1 .req r1 +B_0 .req r2 +B_1 .req r3 +C_0 .req r4 +C_1 .req r5 +D_0 .req r6 +D_1 .req r7 + +Q_0 .req r0 +Q_1 .req r1 +R_0 .req r2 +R_1 .req r3 + +__aeabi_uldivmod: + stmfd sp!, {r4, r5, r6, r7, lr} + @ Test if B == 0 + orrs ip, B_0, B_1 @ Z set -> B == 0 + beq L_div_by_0 + @ Test if B is power of 2: (B & (B - 1)) == 0 + subs C_0, B_0, #1 + sbc C_1, B_1, #0 + tst C_0, B_0 + tsteq B_1, C_1 + beq L_pow2 + @ Test if A_1 == B_1 == 0 + orrs ip, A_1, B_1 + beq L_div_32_32 + +L_div_64_64: +/* CLZ only exists in ARM architecture version 5 and above. */ +#if __COREBOOT_ARM_ARCH__ >= 5 + mov C_0, #1 + mov C_1, #0 + @ D_0 = clz A + teq A_1, #0 + clz D_0, A_1 + clzeq ip, A_0 + addeq D_0, D_0, ip + @ D_1 = clz B + teq B_1, #0 + clz D_1, B_1 + clzeq ip, B_0 + addeq D_1, D_1, ip + @ if clz B - clz A > 0 + subs D_0, D_1, D_0 + bls L_done_shift + @ B <<= (clz B - clz A) + subs D_1, D_0, #32 + rsb ip, D_0, #32 + movmi B_1, B_1, lsl D_0 + orrmi B_1, B_1, B_0, lsr ip + movpl B_1, B_0, lsl D_1 + mov B_0, B_0, lsl D_0 + @ C = 1 << (clz B - clz A) + movmi C_1, C_1, lsl D_0 + orrmi C_1, C_1, C_0, lsr ip + movpl C_1, C_0, lsl D_1 + mov C_0, C_0, lsl D_0 +L_done_shift: + mov D_0, #0 + mov D_1, #0 + @ C: current bit; D: result +#else + @ C: current bit; D: result + mov C_0, #1 + mov C_1, #0 + mov D_0, #0 + mov D_1, #0 +L_lsl_4: + cmp B_1, #0x10000000 + cmpcc B_1, A_1 + cmpeq B_0, A_0 + bcs L_lsl_1 + @ B <<= 4 + mov B_1, B_1, lsl #4 + orr B_1, B_1, B_0, lsr #28 + mov B_0, B_0, lsl #4 + @ C <<= 4 + mov C_1, C_1, lsl #4 + orr C_1, C_1, C_0, lsr #28 + mov C_0, C_0, lsl #4 + b L_lsl_4 +L_lsl_1: + cmp B_1, #0x80000000 + cmpcc B_1, A_1 + cmpeq B_0, A_0 + bcs L_subtract + @ B <<= 1 + mov B_1, B_1, lsl #1 + orr B_1, B_1, B_0, lsr #31 + mov B_0, B_0, lsl #1 + @ C <<= 1 + mov C_1, C_1, lsl #1 + orr C_1, C_1, C_0, lsr #31 + mov C_0, C_0, lsl #1 + b L_lsl_1 +#endif +L_subtract: + @ if A >= B + cmp A_1, B_1 + cmpeq A_0, B_0 + bcc L_update + @ A -= B + subs A_0, A_0, B_0 + sbc A_1, A_1, B_1 + @ D |= C + orr D_0, D_0, C_0 + orr D_1, D_1, C_1 +L_update: + @ if A == 0: break + orrs ip, A_1, A_0 + beq L_exit + @ C >>= 1 + movs C_1, C_1, lsr #1 + movs C_0, C_0, rrx + @ if C == 0: break + orrs ip, C_1, C_0 + beq L_exit + @ B >>= 1 + movs B_1, B_1, lsr #1 + mov B_0, B_0, rrx + b L_subtract +L_exit: + @ Note: A, B & Q, R are aliases + mov R_0, A_0 + mov R_1, A_1 + mov Q_0, D_0 + mov Q_1, D_1 + ldmfd sp!, {r4, r5, r6, r7, pc} + +L_div_32_32: + @ Note: A_0 & r0 are aliases + @ Q_1 r1 + mov r1, B_0 + bl __aeabi_uidivmod + mov R_0, r1 + mov R_1, #0 + mov Q_1, #0 + ldmfd sp!, {r4, r5, r6, r7, pc} + +L_pow2: +/* CLZ only exists in ARM architecture version 5 and above. */ +#if __COREBOOT_ARM_ARCH__ >= 5 + @ Note: A, B and Q, R are aliases + @ R = A & (B - 1) + and C_0, A_0, C_0 + and C_1, A_1, C_1 + @ Q = A >> log2(B) + @ Note: B must not be 0 here! + clz D_0, B_0 + add D_1, D_0, #1 + rsbs D_0, D_0, #31 + bpl L_1 + clz D_0, B_1 + rsb D_0, D_0, #31 + mov A_0, A_1, lsr D_0 + add D_0, D_0, #32 +L_1: + movpl A_0, A_0, lsr D_0 + orrpl A_0, A_0, A_1, lsl D_1 + mov A_1, A_1, lsr D_0 + @ Mov back C to R + mov R_0, C_0 + mov R_1, C_1 + ldmfd sp!, {r4, r5, r6, r7, pc} +#else + @ Note: A, B and Q, R are aliases + @ R = A & (B - 1) + and C_0, A_0, C_0 + and C_1, A_1, C_1 + @ Q = A >> log2(B) + @ Note: B must not be 0 here! + @ Count the leading zeroes in B. + mov D_0, #0 + orrs B_0, B_0, B_0 + @ If B is greater than 1 << 31, divide A and B by 1 << 32. + moveq A_0, A_1 + moveq A_1, #0 + moveq B_0, B_1 + @ Count the remaining leading zeroes in B. + movs B_1, B_0, lsl #16 + addeq D_0, #16 + moveq B_0, B_0, lsr #16 + tst B_0, #0xff + addeq D_0, #8 + moveq B_0, B_0, lsr #8 + tst B_0, #0xf + addeq D_0, #4 + moveq B_0, B_0, lsr #4 + tst B_0, #0x3 + addeq D_0, #2 + moveq B_0, B_0, lsr #2 + tst B_0, #0x1 + addeq D_0, #1 + @ Shift A to the right by the appropriate amount. + rsb D_1, D_0, #32 + mov Q_0, A_0, lsr D_0 + orr Q_0, A_1, lsl D_1 + mov Q_1, A_1, lsr D_0 + @ Move C to R + mov R_0, C_0 + mov R_1, C_1 + ldmfd sp!, {r4, r5, r6, r7, pc} +#endif + +L_div_by_0: + bl __div0 + @ As wrong as it could be + mov Q_0, #0 + mov Q_1, #0 + mov R_0, #0 + mov R_1, #0 + ldmfd sp!, {r4, r5, r6, r7, pc}
-
Marc Jones