| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <CAKv+Gu_iAzQ6et13aACarqns8-xzQ+YSqj+m3mVGGy=ny8GJBg@mail.gmail.com>
Date: Wed, 26 Feb 2020 08:28:28 +0100
From: Ard Biesheuvel <ardb@...nel.org>
To: Atish Patra <atish.patra@....com>
Cc: Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Albert Ou <aou@...s.berkeley.edu>,
Alexios Zavras <alexios.zavras@...el.com>,
Allison Randal <allison@...utok.net>,
Andrew Morton <akpm@...ux-foundation.org>,
Anup Patel <Anup.Patel@....com>, Arnd Bergmann <arnd@...db.de>,
Borislav Petkov <bp@...e.de>,
Catalin Marinas <catalin.marinas@....com>,
Greentime Hu <greentime.hu@...ive.com>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Ingo Molnar <mingo@...nel.org>,
Kate Stewart <kstewart@...uxfoundation.org>,
Kees Cook <keescook@...omium.org>,
Linus Walleij <linus.walleij@...aro.org>,
linux-efi <linux-efi@...r.kernel.org>,
linux-riscv@...ts.infradead.org, Mao Han <han_mao@...ky.com>,
Mauro Carvalho Chehab <mchehab+samsung@...nel.org>,
Michal Simek <michal.simek@...inx.com>,
Mike Rapoport <rppt@...ux.ibm.com>,
Palmer Dabbelt <palmer@...belt.com>,
Paolo Bonzini <pbonzini@...hat.com>,
Paul Walmsley <paul.walmsley@...ive.com>,
Russell King <linux@...linux.org.uk>,
Thomas Gleixner <tglx@...utronix.de>,
Will Deacon <will@...nel.org>,
Alexander Graf <agraf@...raf.de>,
"leif@...iainc.com" <leif@...iainc.com>,
"Chang, Abner (HPS SW/FW Technologist)" <abner.chang@....com>,
"Schaefer, Daniel (DualStudy)" <daniel.schaefer@....com>
Subject: Re: [RFC PATCH 5/5] RISC-V: Add EFI stub support.
On Wed, 26 Feb 2020 at 02:10, Atish Patra <atish.patra@....com> wrote:
>
> Add a RISC-V architecture specific stub code that actually copies the
> actual kernel image to a valid address and jump to it after boot services
> are terminated. Enable UEFI related kernel configs as well for RISC-V.
>
> Signed-off-by: Atish Patra <atish.patra@....com>
> ---
> arch/riscv/Kconfig | 20 ++++
> arch/riscv/Makefile | 1 +
> arch/riscv/configs/defconfig | 1 +
> drivers/firmware/efi/libstub/Makefile | 8 ++
> drivers/firmware/efi/libstub/riscv-stub.c | 135 ++++++++++++++++++++++
> 5 files changed, 165 insertions(+)
> create mode 100644 drivers/firmware/efi/libstub/riscv-stub.c
>
> diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
> index 42c122170cfd..68b1d565e51d 100644
> --- a/arch/riscv/Kconfig
> +++ b/arch/riscv/Kconfig
> @@ -372,10 +372,30 @@ config CMDLINE_FORCE
>
> endchoice
>
> +config EFI_STUB
> + bool
> +
> +config EFI
> + bool "UEFI runtime support"
> + depends on OF
> + select LIBFDT
> + select UCS2_STRING
> + select EFI_PARAMS_FROM_FDT
> + select EFI_STUB
> + select EFI_GENERIC_ARCH_STUB
> + default y
> + help
> + This option provides support for runtime services provided
> + by UEFI firmware (such as non-volatile variables, realtime
> + clock, and platform reset). A UEFI stub is also provided to
> + allow the kernel to be booted as an EFI application. This
> + is only useful on systems that have UEFI firmware.
> +
> endmenu
>
> menu "Power management options"
>
> source "kernel/power/Kconfig"
> +source "drivers/firmware/Kconfig"
>
> endmenu
> diff --git a/arch/riscv/Makefile b/arch/riscv/Makefile
> index b9009a2fbaf5..0afaa89ba9ad 100644
> --- a/arch/riscv/Makefile
> +++ b/arch/riscv/Makefile
> @@ -78,6 +78,7 @@ head-y := arch/riscv/kernel/head.o
> core-y += arch/riscv/
>
> libs-y += arch/riscv/lib/
> +core-$(CONFIG_EFI_STUB) += $(objtree)/drivers/firmware/efi/libstub/lib.a
>
> PHONY += vdso_install
> vdso_install:
> diff --git a/arch/riscv/configs/defconfig b/arch/riscv/configs/defconfig
> index e2ff95cb3390..0a5d3578f51e 100644
> --- a/arch/riscv/configs/defconfig
> +++ b/arch/riscv/configs/defconfig
> @@ -125,3 +125,4 @@ CONFIG_DEBUG_BLOCK_EXT_DEVT=y
> # CONFIG_FTRACE is not set
> # CONFIG_RUNTIME_TESTING_MENU is not set
> CONFIG_MEMTEST=y
> +CONFIG_EFI=y
> diff --git a/drivers/firmware/efi/libstub/Makefile b/drivers/firmware/efi/libstub/Makefile
> index 2c5b76787126..38facb61745b 100644
> --- a/drivers/firmware/efi/libstub/Makefile
> +++ b/drivers/firmware/efi/libstub/Makefile
> @@ -21,6 +21,8 @@ cflags-$(CONFIG_ARM64) := $(subst $(CC_FLAGS_FTRACE),,$(KBUILD_CFLAGS)) \
> cflags-$(CONFIG_ARM) := $(subst $(CC_FLAGS_FTRACE),,$(KBUILD_CFLAGS)) \
> -fno-builtin -fpic \
> $(call cc-option,-mno-single-pic-base)
> +cflags-$(CONFIG_RISCV) := $(subst $(CC_FLAGS_FTRACE),,$(KBUILD_CFLAGS)) \
> + -fpic
>
> cflags-$(CONFIG_EFI_GENERIC_ARCH_STUB) += -I$(srctree)/scripts/dtc/libfdt
>
> @@ -55,6 +57,7 @@ lib-$(CONFIG_EFI_GENERIC_ARCH_STUB) += efi-stub.o fdt.o string.o \
> lib-$(CONFIG_ARM) += arm32-stub.o
> lib-$(CONFIG_ARM64) += arm64-stub.o
> lib-$(CONFIG_X86) += x86-stub.o
> +lib-$(CONFIG_RISCV) += riscv-stub.o
> CFLAGS_arm32-stub.o := -DTEXT_OFFSET=$(TEXT_OFFSET)
> CFLAGS_arm64-stub.o := -DTEXT_OFFSET=$(TEXT_OFFSET)
>
> @@ -79,6 +82,11 @@ STUBCOPY_FLAGS-$(CONFIG_ARM64) += --prefix-alloc-sections=.init \
> --prefix-symbols=__efistub_
> STUBCOPY_RELOC-$(CONFIG_ARM64) := R_AARCH64_ABS
>
> +STUBCOPY_FLAGS-$(CONFIG_RISCV) += --prefix-alloc-sections=.init \
> + --prefix-symbols=__efistub_
> +STUBCOPY_RELOC-$(CONFIG_RISCV) := R_RISCV_HI20
> +
> +
> $(obj)/%.stub.o: $(obj)/%.o FORCE
> $(call if_changed,stubcopy)
>
> diff --git a/drivers/firmware/efi/libstub/riscv-stub.c b/drivers/firmware/efi/libstub/riscv-stub.c
> new file mode 100644
> index 000000000000..3935b29ea93a
> --- /dev/null
> +++ b/drivers/firmware/efi/libstub/riscv-stub.c
> @@ -0,0 +1,135 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (C) 2013, 2014 Linaro Ltd; <roy.franz@...aro.org>
> + * Copyright (C) 2020 Western Digital Corporation or its affiliates.
> + *
> + * This file implements the EFI boot stub for the RISC-V kernel.
> + * Adapted from ARM64 version at drivers/firmware/efi/libstub/arm64-stub.c.
> + */
> +
> +#include <linux/efi.h>
> +#include <linux/libfdt.h>
> +#include <linux/libfdt_env.h>
> +#include <asm/efi.h>
> +#include <asm/sections.h>
> +
> +#include "efistub.h"
> +/*
> + * RISCV requires the kernel image to placed TEXT_OFFSET bytes beyond a 2 MB
> + * aligned base for 64 bit and 4MB for 32 bit.
> + */
> +#if IS_ENABLED(CONFIG_64BIT)
You can use #ifdef here
> +#define MIN_KIMG_ALIGN SZ_2M
> +#else
> +#define MIN_KIMG_ALIGN SZ_4M
> +#endif
> +/*
> + * TEXT_OFFSET ensures that we don't overwrite the firmware that probably sits
> + * at the beginning of the DRAM.
> + */
Ugh. Really? On an EFI system, that memory should be reserved in some
way, we shouldn't be able to stomp on it like that.
> +#define TEXT_OFFSET MIN_KIMG_ALIGN
> +
> +typedef __attribute__((noreturn)) void (*jump_kernel_func)(unsigned int,
> + unsigned long);
> +
> +efi_status_t check_platform_features(void)
> +{
> + return EFI_SUCCESS;
> +}
> +
> +u64 get_boot_hartid_from_fdt(unsigned long fdt)
static
> +{
> + int chosen_node, len;
> + const fdt64_t *prop;
> + uint64_t hartid = U64_MAX;
> +
> + chosen_node = fdt_path_offset((void *)fdt, "/chosen");
> + if (chosen_node < 0)
> + return hartid;
Just return U64_MAX here
> + prop = fdt_getprop((void *)fdt, chosen_node, "efi-boot-hartid", &len);
Please call this 'boot-hartid' not 'efi-boot-hartid' as the hartid
value is independent of whether you boot via EFI or not.
> + if (!prop || len != sizeof(u64))
> + return hartid;
> +
Return U64_MAX
> + hartid = fdt64_to_cpu(*prop);
> +
and just return the swabbed value, so you can get rid of the local var.
> + return hartid;
> +}
> +
> +/*
> + * Jump to real kernel here with following constraints.
> + * 1. MMU should be disabled.
> + * 2. a0 should contain hartid
> + * 3. a1 should DT address
> + */
> +void __noreturn efi_enter_kernel(unsigned long entrypoint, unsigned long fdt)
This prototype has changed, and now includes the size of the fdt in param 3.
> +{
> + unsigned long kernel_entry = entrypoint + _start_kernel - _start;
stext_offset ? It has a terrible name though, and I'll probably
propose to change it at some point, for all arches. But you can still
use it here.
> + jump_kernel_func jump_kernel = (void (*)(unsigned int, unsigned long))kernel_entry;
> + u64 hartid = get_boot_hartid_from_fdt(fdt);
> +
> + if (hartid == U64_MAX)
> + /* We can not use panic or BUG at this point */
> + __asm__ __volatile__ ("ebreak");
> + /* Disable MMU */
> + csr_write(CSR_SATP, 0);
> + jump_kernel(hartid, fdt);
> +}
> +
> +efi_status_t handle_kernel_image(unsigned long *image_addr,
> + unsigned long *image_size,
> + unsigned long *reserve_addr,
> + unsigned long *reserve_size,
> + unsigned long dram_base,
> + efi_loaded_image_t *image)
> +{
> + efi_status_t status;
> + unsigned long kernel_size, kernel_memsize = 0;
> + unsigned long preferred_offset;
> +
> + /*
> + * The preferred offset of the kernel Image is TEXT_OFFSET bytes beyond
> + * a KIMG_ALIGN aligned base.
> + */
> + preferred_offset = round_up(dram_base, MIN_KIMG_ALIGN) + TEXT_OFFSET;
> +
> + kernel_size = _edata - _start;
> + kernel_memsize = kernel_size + (_end - _edata);
> +
> + /*
> + * Try a straight allocation at the preferred offset.
> + * This will work around the issue where, if dram_base == 0x0,
> + * efi_low_alloc() refuses to allocate at 0x0 (to prevent the
> + * address of the allocation to be mistaken for a FAIL return
> + * value or a NULL pointer). It will also ensure that, on
> + * platforms where the [dram_base, dram_base + TEXT_OFFSET)
> + * interval is partially occupied by the firmware (like on APM
> + * Mustang), we can still place the kernel at the address
> + * 'dram_base + TEXT_OFFSET'.
Better drop this entire last sentence (unless it is relevant, but then
rework it to drop the APM Mustang reference)
> + */
> + if (*image_addr == preferred_offset)
> + return EFI_SUCCESS;
> +
> + *image_addr = *reserve_addr = preferred_offset;
> + *reserve_size = round_up(kernel_memsize, EFI_ALLOC_ALIGN);
> +
> + status = efi_bs_call(allocate_pages, EFI_ALLOCATE_ADDRESS,
> + EFI_LOADER_DATA,
> + *reserve_size / EFI_PAGE_SIZE,
> + (efi_physical_addr_t *)reserve_addr);
> +
> + if (status != EFI_SUCCESS) {
> + *reserve_size = kernel_memsize + TEXT_OFFSET;
> + status = efi_low_alloc(*reserve_size, MIN_KIMG_ALIGN,
> + reserve_addr);
> +
> + if (status != EFI_SUCCESS) {
> + pr_efi_err("Failed to relocate kernel\n");
> + *reserve_size = 0;
> + return status;
> + }
> + *image_addr = *reserve_addr + TEXT_OFFSET;
> + }
> + memcpy((void *)*image_addr, image->image_base, kernel_size);
> +
> + return EFI_SUCCESS;
> +}
> --
> 2.24.0
>
Powered by blists - more mailing lists