Hello Sascha, On 24.10.22 10:55, Sascha Hauer wrote: > On Mon, Oct 24, 2022 at 08:57:13AM +0200, Ahmad Fatoum wrote: >> Recent episode with pointer authentication showed again that for >> platforms without __attribute__((naked)), we are better off writing >> the early header in assembly. We still want to keep the board specific >> entry points in C for ease of use, so we have ENTRY_FUNCTION_WITHSTACK >> generate two symbols: >> >> - A 32-bit stack top value that's placed in .rodata >> >> - An entry point with the normal C code, including stack-using >> prologues >> >> The new common assembly head code will access the stack pointer in a >> position-independent manner and set it up, before continuing with the >> C code. The barebox header is part of the common assembly head code >> ensuring it's not moved around due to compiler code generation. >> >> The common code will need access to board-specific entry point and stack >> top. The former is readily available as the alias __pbl_board_entry. >> The latter is a bit more complicated, as the symbol may not exist for >> boards not using the common header in a multi-image build. >> >> Signed-off-by: Ahmad Fatoum <a.fatoum@xxxxxxxxxxxxxx> >> --- >> Makefile | 3 +-- >> arch/arm/cpu/Makefile | 2 ++ >> arch/arm/cpu/head_64.S | 36 ++++++++++++++++++++++++++++++ >> arch/arm/include/asm/barebox-arm.h | 28 +++++++---------------- >> arch/arm/lib/pbl.lds.S | 9 ++++++++ >> 5 files changed, 56 insertions(+), 22 deletions(-) >> create mode 100644 arch/arm/cpu/head_64.S >> >> diff --git a/Makefile b/Makefile >> index 81ef44122367..a4b9f3c021d7 100644 >> --- a/Makefile >> +++ b/Makefile >> @@ -660,8 +660,7 @@ KBUILD_CFLAGS += $(call cc-option,-fno-stack-check) >> KBUILD_CFLAGS += $(call cc-option,-fcf-protection=none) >> >> # We don't have the necessary infrastructure to benefit from ARMv8.3+ pointer >> -# authentication. On older CPUs, they are interpreted as NOPs and blot the >> -# code and break less portable code that expects a very specific code layout >> +# authentication. On older CPUs, they are interpreted as NOPs bloating the code >> KBUILD_CFLAGS += $(call cc-option,-mbranch-protection=none) >> >> KBUILD_CFLAGS += $(call cc-disable-warning, address-of-packed-member) >> diff --git a/arch/arm/cpu/Makefile b/arch/arm/cpu/Makefile >> index c0993c1abe5d..7674c1464c1f 100644 >> --- a/arch/arm/cpu/Makefile >> +++ b/arch/arm/cpu/Makefile >> @@ -13,6 +13,8 @@ AFLAGS_hyp.pbl.o :=-Wa,-march=armv7-a -Wa,-mcpu=all >> obj-y += start.o entry.o entry_ll$(S64).o >> KASAN_SANITIZE_start.o := n >> >> +pbl-$(CONFIG_CPU_64) += head_64.o >> + >> pbl-$(CONFIG_BOARD_ARM_GENERIC_DT) += board-dt-2nd.o >> pbl-$(CONFIG_BOARD_ARM_GENERIC_DT_AARCH64) += board-dt-2nd-aarch64.o >> >> diff --git a/arch/arm/cpu/head_64.S b/arch/arm/cpu/head_64.S >> new file mode 100644 >> index 000000000000..4ed4ffb05250 >> --- /dev/null >> +++ b/arch/arm/cpu/head_64.S >> @@ -0,0 +1,36 @@ >> +/* SPDX-License-Identifier: GPL-2.0-only */ >> +#include <linux/linkage.h> >> +#include <asm/barebox-arm64.h> >> +#include <asm/image.h> >> + >> +/* Linker will point these at board-specific symbols */ >> +.globl __pbl_board_stack_top >> +.globl __pbl_board_entry >> + >> +.section .text_head_prologue_common, "x" >> +ENTRY(__barebox_arm64_head) >> + nop >> + nop >> + nop > > Why these nops here? Perhaps we decide to place something there in future. For example, `ccmp x18, #0, #0xd, pl` is a fancy nop whose first 2 bytes are the ELF/PE magic 'MZ' when read in little-endian. For now, I left it as normal nops. The .org 20 below is not (only) for alignment, but as a sanity check, if we add too much code here. >> + adr x9, __pbl_board_stack_top >> + ldr w9, [x9] >> + cbz x9, 1f >> + mov sp, x9 >> + b 1f >> + .org 0x20 >> + .asciz "barebox" >> + .word 0xffffffff >> + .word _barebox_image_size /* image size to copy */ >> + .rept 8 >> + .word 0x55555555 >> + .endr >> +1: >> +#ifdef CONFIG_PBL_BREAK >> + brk #17 >> + nop >> +#else >> + nop >> + nop >> +#endif > > Why do you jump over the barebox header? You could put all the code > before it. I want to keep at the same location as ARM32. Cheers, Ahmad > > Sascha > -- Pengutronix e.K. | | Steuerwalder Str. 21 | http://www.pengutronix.de/ | 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
