u-boot/arch/mips/cpu/start.S
Paul Burton 703ec9ddf9 MIPS: Stop building position independent code
U-Boot has up until now built with -fpic for the MIPS architecture,
producing position independent code which uses indirection through a
global offset table, making relocation fairly straightforward as it
simply involves patching up GOT entries.

Using -fpic does however have some downsides. The biggest of these is
that generated code is bloated in various ways. For example, function
calls are indirected through the GOT & the t9 register:

  8f998064   lw     t9,-32668(gp)
  0320f809   jalr   t9

Without -fpic the call is simply:

  0f803f01   jal    be00fc04 <puts>

This is more compact & faster (due to the lack of the load & the
dependency the jump has on its result). It is also easier to read &
debug because the disassembly shows what function is being called,
rather than just an offset from gp which would then have to be looked up
in the ELF to discover the target function.

Another disadvantage of -fpic is that each function begins with a
sequence to calculate the value of the gp register, for example:

  3c1c0004   lui    gp,0x4
  279c3384   addiu  gp,gp,13188
  0399e021   addu   gp,gp,t9

Without using -fpic this sequence no longer appears at the start of each
function, reducing code size considerably.

This patch switches U-Boot from building with -fpic to building with
-fno-pic, in order to gain the benefits described above. The cost of
this is an extra step during the build process to extract relocation
data from the ELF & write it into a new .rel section in a compact
format, plus the added complexity of dealing with multiple types of
relocation rather than the single type that applied to the GOT. The
benefit is smaller, cleaner, more debuggable code. The relocate_code()
function is reimplemented in C to handle the new relocation scheme,
which also makes it easier to read & debug.

Taking maltael_defconfig as an example the size of u-boot.bin built
using the Codescape MIPS 2016.05-06 toolchain (gcc 4.9.2, binutils
2.24.90) shrinks from 254KiB to 224KiB.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Cc: u-boot@lists.denx.de
Reviewed-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Tested-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
2017-07-25 20:44:00 +02:00

281 lines
5.5 KiB
ArmAsm

/*
* Startup Code for MIPS32 CPU-core
*
* Copyright (c) 2003 Wolfgang Denk <wd@denx.de>
*
* SPDX-License-Identifier: GPL-2.0+
*/
#include <asm-offsets.h>
#include <config.h>
#include <asm/asm.h>
#include <asm/regdef.h>
#include <asm/mipsregs.h>
#ifndef CONFIG_SYS_INIT_SP_ADDR
#define CONFIG_SYS_INIT_SP_ADDR (CONFIG_SYS_SDRAM_BASE + \
CONFIG_SYS_INIT_SP_OFFSET)
#endif
#ifdef CONFIG_32BIT
# define MIPS_RELOC 3
# define STATUS_SET 0
#endif
#ifdef CONFIG_64BIT
# ifdef CONFIG_SYS_LITTLE_ENDIAN
# define MIPS64_R_INFO(ssym, r_type3, r_type2, r_type) \
(((r_type) << 24) | ((r_type2) << 16) | ((r_type3) << 8) | (ssym))
# else
# define MIPS64_R_INFO(ssym, r_type3, r_type2, r_type) \
((r_type) | ((r_type2) << 8) | ((r_type3) << 16) | (ssym) << 24)
# endif
# define MIPS_RELOC MIPS64_R_INFO(0x00, 0x00, 0x12, 0x03)
# define STATUS_SET ST0_KX
#endif
.set noreorder
.macro init_wr sel
MTC0 zero, CP0_WATCHLO,\sel
mtc0 t1, CP0_WATCHHI,\sel
mfc0 t0, CP0_WATCHHI,\sel
bgez t0, wr_done
nop
.endm
.macro uhi_mips_exception
move k0, t9 # preserve t9 in k0
move k1, a0 # preserve a0 in k1
li t9, 15 # UHI exception operation
li a0, 0 # Use hard register context
sdbbp 1 # Invoke UHI operation
.endm
.macro setup_stack_gd
li t0, -16
PTR_LI t1, CONFIG_SYS_INIT_SP_ADDR
and sp, t1, t0 # force 16 byte alignment
PTR_SUBU \
sp, sp, GD_SIZE # reserve space for gd
and sp, sp, t0 # force 16 byte alignment
move k0, sp # save gd pointer
#ifdef CONFIG_SYS_MALLOC_F_LEN
li t2, CONFIG_SYS_MALLOC_F_LEN
PTR_SUBU \
sp, sp, t2 # reserve space for early malloc
and sp, sp, t0 # force 16 byte alignment
#endif
move fp, sp
/* Clear gd */
move t0, k0
1:
PTR_S zero, 0(t0)
blt t0, t1, 1b
PTR_ADDIU t0, PTRSIZE
#ifdef CONFIG_SYS_MALLOC_F_LEN
PTR_S sp, GD_MALLOC_BASE(k0) # gd->malloc_base offset
#endif
.endm
ENTRY(_start)
/* U-Boot entry point */
b reset
mtc0 zero, CP0_COUNT # clear cp0 count for most accurate boot timing
#if defined(CONFIG_SYS_XWAY_EBU_BOOTCFG)
/*
* Almost all Lantiq XWAY SoC devices have an external bus unit (EBU) to
* access external NOR flashes. If the board boots from NOR flash the
* internal BootROM does a blind read at address 0xB0000010 to read the
* initial configuration for that EBU in order to access the flash
* device with correct parameters. This config option is board-specific.
*/
.org 0x10
.word CONFIG_SYS_XWAY_EBU_BOOTCFG
.word 0x0
#endif
#if defined(CONFIG_MALTA)
/*
* Linux expects the Board ID here.
*/
.org 0x10
.word 0x00000420 # 0x420 (Malta Board with CoreLV)
.word 0x00000000
#endif
#if defined(CONFIG_ROM_EXCEPTION_VECTORS)
/*
* Exception vector entry points. When running from ROM, an exception
* cannot be handled. Halt execution and transfer control to debugger,
* if one is attached.
*/
.org 0x200
/* TLB refill, 32 bit task */
uhi_mips_exception
.org 0x280
/* XTLB refill, 64 bit task */
uhi_mips_exception
.org 0x300
/* Cache error exception */
uhi_mips_exception
.org 0x380
/* General exception */
uhi_mips_exception
.org 0x400
/* Catch interrupt exceptions */
uhi_mips_exception
.org 0x480
/* EJTAG debug exception */
1: b 1b
nop
.org 0x500
#endif
reset:
#if __mips_isa_rev >= 6
mfc0 t0, CP0_CONFIG, 5
and t0, t0, MIPS_CONF5_VP
beqz t0, 1f
nop
b 2f
mfc0 t0, CP0_GLOBALNUMBER
#endif
#ifdef CONFIG_ARCH_BMIPS
1: mfc0 t0, CP0_DIAGNOSTIC, 3
and t0, t0, (1 << 31)
#else
1: mfc0 t0, CP0_EBASE
and t0, t0, EBASE_CPUNUM
#endif
/* Hang if this isn't the first CPU in the system */
2: beqz t0, 4f
nop
3: wait
b 3b
nop
/* Init CP0 Status */
4: mfc0 t0, CP0_STATUS
and t0, ST0_IMPL
or t0, ST0_BEV | ST0_ERL | STATUS_SET
mtc0 t0, CP0_STATUS
/*
* Check whether CP0 Config1 is implemented. If not continue
* with legacy Watch register initialization.
*/
mfc0 t0, CP0_CONFIG
bgez t0, wr_legacy
nop
/*
* Check WR bit in CP0 Config1 to determine if Watch registers
* are implemented.
*/
mfc0 t0, CP0_CONFIG, 1
andi t0, (1 << 3)
beqz t0, wr_done
nop
/* Clear Watch Status bits and disable watch exceptions */
li t1, 0x7 # Clear I, R and W conditions
init_wr 0
init_wr 1
init_wr 2
init_wr 3
init_wr 4
init_wr 5
init_wr 6
init_wr 7
b wr_done
nop
wr_legacy:
MTC0 zero, CP0_WATCHLO
mtc0 zero, CP0_WATCHHI
wr_done:
/* Clear WP, IV and SW interrupts */
mtc0 zero, CP0_CAUSE
/* Clear timer interrupt (CP0_COUNT cleared on branch to 'reset') */
mtc0 zero, CP0_COMPARE
#ifndef CONFIG_SKIP_LOWLEVEL_INIT
mfc0 t0, CP0_CONFIG
and t0, t0, MIPS_CONF_IMPL
or t0, t0, CONF_CM_UNCACHED
mtc0 t0, CP0_CONFIG
ehb
#endif
#ifdef CONFIG_MIPS_CM
PTR_LA t9, mips_cm_map
jalr t9
nop
#endif
#ifdef CONFIG_MIPS_INIT_STACK_IN_SRAM
/* Set up initial stack and global data */
setup_stack_gd
# ifdef CONFIG_DEBUG_UART
/* Earliest point to set up debug uart */
PTR_LA t9, debug_uart_init
jalr t9
nop
# endif
#endif
#ifndef CONFIG_SKIP_LOWLEVEL_INIT
# ifdef CONFIG_SYS_MIPS_CACHE_INIT_RAM_LOAD
/* Initialize any external memory */
PTR_LA t9, lowlevel_init
jalr t9
nop
# endif
/* Initialize caches... */
PTR_LA t9, mips_cache_reset
jalr t9
nop
# ifndef CONFIG_SYS_MIPS_CACHE_INIT_RAM_LOAD
/* Initialize any external memory */
PTR_LA t9, lowlevel_init
jalr t9
nop
# endif
#endif
#ifndef CONFIG_MIPS_INIT_STACK_IN_SRAM
/* Set up initial stack and global data */
setup_stack_gd
# ifdef CONFIG_DEBUG_UART
/* Earliest point to set up debug uart */
PTR_LA t9, debug_uart_init
jalr t9
nop
# endif
#endif
move a0, zero # a0 <-- boot_flags = 0
PTR_LA t9, board_init_f
jr t9
move ra, zero
END(_start)