2019-06-03 13:44:46 +08:00
|
|
|
// SPDX-License-Identifier: GPL-2.0-only
|
2014-08-09 05:26:06 +08:00
|
|
|
/*
|
|
|
|
* Kexec bzImage loader
|
|
|
|
*
|
|
|
|
* Copyright (C) 2014 Red Hat Inc.
|
|
|
|
* Authors:
|
|
|
|
* Vivek Goyal <vgoyal@redhat.com>
|
|
|
|
*/
|
|
|
|
|
|
|
|
#define pr_fmt(fmt) "kexec-bzImage64: " fmt
|
|
|
|
|
|
|
|
#include <linux/string.h>
|
|
|
|
#include <linux/printk.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/kexec.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/mm.h>
|
2014-08-09 05:26:11 +08:00
|
|
|
#include <linux/efi.h>
|
x86/setup: Use rng seeds from setup_data
Currently, the only way x86 can get an early boot RNG seed is via EFI,
which is generally always used now for physical machines, but is very
rarely used in VMs, especially VMs that are optimized for starting
"instantaneously", such as Firecracker's MicroVM. For tiny fast booting
VMs, EFI is not something you generally need or want.
Rather, the image loader or firmware should be able to pass a single
random seed, exactly as device tree platforms do with the "rng-seed"
property. Additionally, this is something that bootloaders can append,
with their own seed file management, which is something every other
major OS ecosystem has that Linux does not (yet).
Add SETUP_RNG_SEED, similar to the other eight setup_data entries that
are parsed at boot. It also takes care to zero out the seed immediately
after using, in order to retain forward secrecy. This all takes about 7
trivial lines of code.
Then, on kexec_file_load(), a new fresh seed is generated and passed to
the next kernel, just as is done on device tree architectures when
using kexec. And, importantly, I've tested that QEMU is able to properly
pass SETUP_RNG_SEED as well, making this work for every step of the way.
This code too is pretty straight forward.
Together these measures ensure that VMs and nested kexec()'d kernels
always receive a proper boot time RNG seed at the earliest possible
stage from their parents:
- Host [already has strongly initialized RNG]
- QEMU [passes fresh seed in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- ...
I've verified in several scenarios that this works quite well from a
host kernel to QEMU and down inwards, mixing and matching loaders, with
every layer providing a seed to the next.
[ bp: Massage commit message. ]
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Link: https://lore.kernel.org/r/20220630113300.1892799-1-Jason@zx2c4.com
2022-07-11 01:29:21 +08:00
|
|
|
#include <linux/random.h>
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
#include <asm/bootparam.h>
|
|
|
|
#include <asm/setup.h>
|
2014-08-09 05:26:09 +08:00
|
|
|
#include <asm/crash.h>
|
2014-08-09 05:26:11 +08:00
|
|
|
#include <asm/efi.h>
|
2017-01-27 18:59:46 +08:00
|
|
|
#include <asm/e820/api.h>
|
2014-10-14 06:53:46 +08:00
|
|
|
#include <asm/kexec-bzimage64.h>
|
2014-08-09 05:26:09 +08:00
|
|
|
|
|
|
|
#define MAX_ELFCOREHDR_STR_LEN 30 /* elfcorehdr=0x<64bit-value> */
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Defines lowest physical address for various segments. Not sure where
|
|
|
|
* exactly these limits came from. Current bzimage64 loader in kexec-tools
|
|
|
|
* uses these so I am retaining it. It can be changed over time as we gain
|
|
|
|
* more insight.
|
|
|
|
*/
|
|
|
|
#define MIN_PURGATORY_ADDR 0x3000
|
|
|
|
#define MIN_BOOTPARAM_ADDR 0x3000
|
|
|
|
#define MIN_KERNEL_LOAD_ADDR 0x100000
|
|
|
|
#define MIN_INITRD_LOAD_ADDR 0x1000000
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This is a place holder for all boot loader specific data structure which
|
|
|
|
* gets allocated in one call but gets freed much later during cleanup
|
|
|
|
* time. Right now there is only one field but it can grow as need be.
|
|
|
|
*/
|
|
|
|
struct bzimage64_data {
|
|
|
|
/*
|
|
|
|
* Temporary buffer to hold bootparams buffer. This should be
|
|
|
|
* freed once the bootparam segment has been loaded.
|
|
|
|
*/
|
|
|
|
void *bootparams_buf;
|
|
|
|
};
|
|
|
|
|
|
|
|
static int setup_initrd(struct boot_params *params,
|
|
|
|
unsigned long initrd_load_addr, unsigned long initrd_len)
|
|
|
|
{
|
|
|
|
params->hdr.ramdisk_image = initrd_load_addr & 0xffffffffUL;
|
|
|
|
params->hdr.ramdisk_size = initrd_len & 0xffffffffUL;
|
|
|
|
|
|
|
|
params->ext_ramdisk_image = initrd_load_addr >> 32;
|
|
|
|
params->ext_ramdisk_size = initrd_len >> 32;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:09 +08:00
|
|
|
static int setup_cmdline(struct kimage *image, struct boot_params *params,
|
2014-08-09 05:26:06 +08:00
|
|
|
unsigned long bootparams_load_addr,
|
|
|
|
unsigned long cmdline_offset, char *cmdline,
|
|
|
|
unsigned long cmdline_len)
|
|
|
|
{
|
|
|
|
char *cmdline_ptr = ((char *)params) + cmdline_offset;
|
2015-07-01 05:57:39 +08:00
|
|
|
unsigned long cmdline_ptr_phys, len = 0;
|
2014-08-09 05:26:06 +08:00
|
|
|
uint32_t cmdline_low_32, cmdline_ext_32;
|
|
|
|
|
2014-08-09 05:26:09 +08:00
|
|
|
if (image->type == KEXEC_TYPE_CRASH) {
|
2015-07-01 05:57:39 +08:00
|
|
|
len = sprintf(cmdline_ptr,
|
2021-02-22 01:49:21 +08:00
|
|
|
"elfcorehdr=0x%lx ", image->elf_load_addr);
|
2014-08-09 05:26:09 +08:00
|
|
|
}
|
2015-07-01 05:57:39 +08:00
|
|
|
memcpy(cmdline_ptr + len, cmdline, cmdline_len);
|
|
|
|
cmdline_len += len;
|
|
|
|
|
2014-08-09 05:26:06 +08:00
|
|
|
cmdline_ptr[cmdline_len - 1] = '\0';
|
|
|
|
|
2014-08-09 05:26:09 +08:00
|
|
|
pr_debug("Final command line is: %s\n", cmdline_ptr);
|
2014-08-09 05:26:06 +08:00
|
|
|
cmdline_ptr_phys = bootparams_load_addr + cmdline_offset;
|
|
|
|
cmdline_low_32 = cmdline_ptr_phys & 0xffffffffUL;
|
|
|
|
cmdline_ext_32 = cmdline_ptr_phys >> 32;
|
|
|
|
|
|
|
|
params->hdr.cmd_line_ptr = cmdline_low_32;
|
|
|
|
if (cmdline_ext_32)
|
|
|
|
params->ext_cmd_line_ptr = cmdline_ext_32;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:11 +08:00
|
|
|
static int setup_e820_entries(struct boot_params *params)
|
2014-08-09 05:26:06 +08:00
|
|
|
{
|
|
|
|
unsigned int nr_e820_entries;
|
|
|
|
|
2017-07-03 01:07:12 +08:00
|
|
|
nr_e820_entries = e820_table_kexec->nr_entries;
|
2014-08-09 05:26:06 +08:00
|
|
|
|
2017-01-29 00:29:08 +08:00
|
|
|
/* TODO: Pass entries more than E820_MAX_ENTRIES_ZEROPAGE in bootparams setup data */
|
|
|
|
if (nr_e820_entries > E820_MAX_ENTRIES_ZEROPAGE)
|
|
|
|
nr_e820_entries = E820_MAX_ENTRIES_ZEROPAGE;
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
params->e820_entries = nr_e820_entries;
|
2017-07-03 01:07:12 +08:00
|
|
|
memcpy(¶ms->e820_table, &e820_table_kexec->entries, nr_e820_entries*sizeof(struct e820_entry));
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
x86/setup: Use rng seeds from setup_data
Currently, the only way x86 can get an early boot RNG seed is via EFI,
which is generally always used now for physical machines, but is very
rarely used in VMs, especially VMs that are optimized for starting
"instantaneously", such as Firecracker's MicroVM. For tiny fast booting
VMs, EFI is not something you generally need or want.
Rather, the image loader or firmware should be able to pass a single
random seed, exactly as device tree platforms do with the "rng-seed"
property. Additionally, this is something that bootloaders can append,
with their own seed file management, which is something every other
major OS ecosystem has that Linux does not (yet).
Add SETUP_RNG_SEED, similar to the other eight setup_data entries that
are parsed at boot. It also takes care to zero out the seed immediately
after using, in order to retain forward secrecy. This all takes about 7
trivial lines of code.
Then, on kexec_file_load(), a new fresh seed is generated and passed to
the next kernel, just as is done on device tree architectures when
using kexec. And, importantly, I've tested that QEMU is able to properly
pass SETUP_RNG_SEED as well, making this work for every step of the way.
This code too is pretty straight forward.
Together these measures ensure that VMs and nested kexec()'d kernels
always receive a proper boot time RNG seed at the earliest possible
stage from their parents:
- Host [already has strongly initialized RNG]
- QEMU [passes fresh seed in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- ...
I've verified in several scenarios that this works quite well from a
host kernel to QEMU and down inwards, mixing and matching loaders, with
every layer providing a seed to the next.
[ bp: Massage commit message. ]
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Link: https://lore.kernel.org/r/20220630113300.1892799-1-Jason@zx2c4.com
2022-07-11 01:29:21 +08:00
|
|
|
enum { RNG_SEED_LENGTH = 32 };
|
|
|
|
|
|
|
|
static void
|
|
|
|
setup_rng_seed(struct boot_params *params, unsigned long params_load_addr,
|
|
|
|
unsigned int rng_seed_setup_data_offset)
|
|
|
|
{
|
|
|
|
struct setup_data *sd = (void *)params + rng_seed_setup_data_offset;
|
|
|
|
unsigned long setup_data_phys;
|
|
|
|
|
|
|
|
if (!rng_is_initialized())
|
|
|
|
return;
|
|
|
|
|
|
|
|
sd->type = SETUP_RNG_SEED;
|
|
|
|
sd->len = RNG_SEED_LENGTH;
|
|
|
|
get_random_bytes(sd->data, RNG_SEED_LENGTH);
|
|
|
|
setup_data_phys = params_load_addr + rng_seed_setup_data_offset;
|
|
|
|
sd->next = params->hdr.setup_data;
|
|
|
|
params->hdr.setup_data = setup_data_phys;
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:11 +08:00
|
|
|
#ifdef CONFIG_EFI
|
|
|
|
static int setup_efi_info_memmap(struct boot_params *params,
|
|
|
|
unsigned long params_load_addr,
|
|
|
|
unsigned int efi_map_offset,
|
|
|
|
unsigned int efi_map_sz)
|
|
|
|
{
|
|
|
|
void *efi_map = (void *)params + efi_map_offset;
|
|
|
|
unsigned long efi_map_phys_addr = params_load_addr + efi_map_offset;
|
|
|
|
struct efi_info *ei = ¶ms->efi_info;
|
|
|
|
|
|
|
|
if (!efi_map_sz)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
efi_runtime_map_copy(efi_map, efi_map_sz);
|
|
|
|
|
|
|
|
ei->efi_memmap = efi_map_phys_addr & 0xffffffff;
|
|
|
|
ei->efi_memmap_hi = efi_map_phys_addr >> 32;
|
|
|
|
ei->efi_memmap_size = efi_map_sz;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
prepare_add_efi_setup_data(struct boot_params *params,
|
|
|
|
unsigned long params_load_addr,
|
|
|
|
unsigned int efi_setup_data_offset)
|
|
|
|
{
|
|
|
|
unsigned long setup_data_phys;
|
|
|
|
struct setup_data *sd = (void *)params + efi_setup_data_offset;
|
|
|
|
struct efi_setup_data *esd = (void *)sd + sizeof(struct setup_data);
|
|
|
|
|
2020-01-21 00:23:21 +08:00
|
|
|
esd->fw_vendor = efi_fw_vendor;
|
|
|
|
esd->tables = efi_config_table;
|
2014-08-09 05:26:11 +08:00
|
|
|
esd->smbios = efi.smbios;
|
|
|
|
|
|
|
|
sd->type = SETUP_EFI;
|
|
|
|
sd->len = sizeof(struct efi_setup_data);
|
|
|
|
|
|
|
|
/* Add setup data */
|
|
|
|
setup_data_phys = params_load_addr + efi_setup_data_offset;
|
|
|
|
sd->next = params->hdr.setup_data;
|
|
|
|
params->hdr.setup_data = setup_data_phys;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
setup_efi_state(struct boot_params *params, unsigned long params_load_addr,
|
|
|
|
unsigned int efi_map_offset, unsigned int efi_map_sz,
|
|
|
|
unsigned int efi_setup_data_offset)
|
|
|
|
{
|
|
|
|
struct efi_info *current_ei = &boot_params.efi_info;
|
|
|
|
struct efi_info *ei = ¶ms->efi_info;
|
|
|
|
|
2019-01-18 19:13:08 +08:00
|
|
|
if (!efi_enabled(EFI_RUNTIME_SERVICES))
|
|
|
|
return 0;
|
|
|
|
|
2014-08-09 05:26:11 +08:00
|
|
|
if (!current_ei->efi_memmap_size)
|
|
|
|
return 0;
|
|
|
|
|
2019-08-20 08:17:43 +08:00
|
|
|
params->secure_boot = boot_params.secure_boot;
|
2014-08-09 05:26:11 +08:00
|
|
|
ei->efi_loader_signature = current_ei->efi_loader_signature;
|
|
|
|
ei->efi_systab = current_ei->efi_systab;
|
|
|
|
ei->efi_systab_hi = current_ei->efi_systab_hi;
|
|
|
|
|
|
|
|
ei->efi_memdesc_version = current_ei->efi_memdesc_version;
|
|
|
|
ei->efi_memdesc_size = efi_get_runtime_map_desc_size();
|
|
|
|
|
|
|
|
setup_efi_info_memmap(params, params_load_addr, efi_map_offset,
|
|
|
|
efi_map_sz);
|
|
|
|
prepare_add_efi_setup_data(params, params_load_addr,
|
|
|
|
efi_setup_data_offset);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_EFI */
|
|
|
|
|
2022-06-30 16:36:12 +08:00
|
|
|
static void
|
|
|
|
setup_ima_state(const struct kimage *image, struct boot_params *params,
|
|
|
|
unsigned long params_load_addr,
|
|
|
|
unsigned int ima_setup_data_offset)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_IMA_KEXEC
|
|
|
|
struct setup_data *sd = (void *)params + ima_setup_data_offset;
|
|
|
|
unsigned long setup_data_phys;
|
|
|
|
struct ima_setup_data *ima;
|
|
|
|
|
|
|
|
if (!image->ima_buffer_size)
|
|
|
|
return;
|
|
|
|
|
|
|
|
sd->type = SETUP_IMA;
|
|
|
|
sd->len = sizeof(*ima);
|
|
|
|
|
|
|
|
ima = (void *)sd + sizeof(struct setup_data);
|
|
|
|
ima->addr = image->ima_buffer_addr;
|
|
|
|
ima->size = image->ima_buffer_size;
|
|
|
|
|
|
|
|
/* Add setup data */
|
|
|
|
setup_data_phys = params_load_addr + ima_setup_data_offset;
|
|
|
|
sd->next = params->hdr.setup_data;
|
|
|
|
params->hdr.setup_data = setup_data_phys;
|
|
|
|
#endif /* CONFIG_IMA_KEXEC */
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:11 +08:00
|
|
|
static int
|
|
|
|
setup_boot_parameters(struct kimage *image, struct boot_params *params,
|
|
|
|
unsigned long params_load_addr,
|
|
|
|
unsigned int efi_map_offset, unsigned int efi_map_sz,
|
2022-06-30 16:36:12 +08:00
|
|
|
unsigned int setup_data_offset)
|
2014-08-09 05:26:06 +08:00
|
|
|
{
|
|
|
|
unsigned int nr_e820_entries;
|
|
|
|
unsigned long long mem_k, start, end;
|
2014-08-09 05:26:09 +08:00
|
|
|
int i, ret = 0;
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
/* Get subarch from existing bootparams */
|
|
|
|
params->hdr.hardware_subarch = boot_params.hdr.hardware_subarch;
|
|
|
|
|
|
|
|
/* Copying screen_info will do? */
|
2020-10-14 17:24:28 +08:00
|
|
|
memcpy(¶ms->screen_info, &screen_info, sizeof(struct screen_info));
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
/* Fill in memsize later */
|
|
|
|
params->screen_info.ext_mem_k = 0;
|
|
|
|
params->alt_mem_k = 0;
|
|
|
|
|
2019-02-05 01:38:52 +08:00
|
|
|
/* Always fill in RSDP: it is either 0 or a valid value */
|
|
|
|
params->acpi_rsdp_addr = boot_params.acpi_rsdp_addr;
|
|
|
|
|
2014-08-09 05:26:06 +08:00
|
|
|
/* Default APM info */
|
|
|
|
memset(¶ms->apm_bios_info, 0, sizeof(params->apm_bios_info));
|
|
|
|
|
|
|
|
/* Default drive info */
|
|
|
|
memset(¶ms->hd0_info, 0, sizeof(params->hd0_info));
|
|
|
|
memset(¶ms->hd1_info, 0, sizeof(params->hd1_info));
|
|
|
|
|
2014-08-09 05:26:09 +08:00
|
|
|
if (image->type == KEXEC_TYPE_CRASH) {
|
|
|
|
ret = crash_setup_memmap_entries(image, params);
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
} else
|
2014-08-09 05:26:11 +08:00
|
|
|
setup_e820_entries(params);
|
2014-08-09 05:26:09 +08:00
|
|
|
|
2014-08-09 05:26:06 +08:00
|
|
|
nr_e820_entries = params->e820_entries;
|
|
|
|
|
|
|
|
for (i = 0; i < nr_e820_entries; i++) {
|
2017-01-29 00:09:33 +08:00
|
|
|
if (params->e820_table[i].type != E820_TYPE_RAM)
|
2014-08-09 05:26:06 +08:00
|
|
|
continue;
|
2017-01-27 20:54:38 +08:00
|
|
|
start = params->e820_table[i].addr;
|
|
|
|
end = params->e820_table[i].addr + params->e820_table[i].size - 1;
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
if ((start <= 0x100000) && end > 0x100000) {
|
|
|
|
mem_k = (end >> 10) - (0x100000 >> 10);
|
|
|
|
params->screen_info.ext_mem_k = mem_k;
|
|
|
|
params->alt_mem_k = mem_k;
|
|
|
|
if (mem_k > 0xfc00)
|
|
|
|
params->screen_info.ext_mem_k = 0xfc00; /* 64M*/
|
|
|
|
if (mem_k > 0xffffffff)
|
|
|
|
params->alt_mem_k = 0xffffffff;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:11 +08:00
|
|
|
#ifdef CONFIG_EFI
|
|
|
|
/* Setup EFI state */
|
|
|
|
setup_efi_state(params, params_load_addr, efi_map_offset, efi_map_sz,
|
2022-06-30 16:36:12 +08:00
|
|
|
setup_data_offset);
|
|
|
|
setup_data_offset += sizeof(struct setup_data) +
|
|
|
|
sizeof(struct efi_setup_data);
|
2014-08-09 05:26:11 +08:00
|
|
|
#endif
|
2022-06-30 16:36:12 +08:00
|
|
|
|
x86/setup: Use rng seeds from setup_data
Currently, the only way x86 can get an early boot RNG seed is via EFI,
which is generally always used now for physical machines, but is very
rarely used in VMs, especially VMs that are optimized for starting
"instantaneously", such as Firecracker's MicroVM. For tiny fast booting
VMs, EFI is not something you generally need or want.
Rather, the image loader or firmware should be able to pass a single
random seed, exactly as device tree platforms do with the "rng-seed"
property. Additionally, this is something that bootloaders can append,
with their own seed file management, which is something every other
major OS ecosystem has that Linux does not (yet).
Add SETUP_RNG_SEED, similar to the other eight setup_data entries that
are parsed at boot. It also takes care to zero out the seed immediately
after using, in order to retain forward secrecy. This all takes about 7
trivial lines of code.
Then, on kexec_file_load(), a new fresh seed is generated and passed to
the next kernel, just as is done on device tree architectures when
using kexec. And, importantly, I've tested that QEMU is able to properly
pass SETUP_RNG_SEED as well, making this work for every step of the way.
This code too is pretty straight forward.
Together these measures ensure that VMs and nested kexec()'d kernels
always receive a proper boot time RNG seed at the earliest possible
stage from their parents:
- Host [already has strongly initialized RNG]
- QEMU [passes fresh seed in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- ...
I've verified in several scenarios that this works quite well from a
host kernel to QEMU and down inwards, mixing and matching loaders, with
every layer providing a seed to the next.
[ bp: Massage commit message. ]
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Link: https://lore.kernel.org/r/20220630113300.1892799-1-Jason@zx2c4.com
2022-07-11 01:29:21 +08:00
|
|
|
if (IS_ENABLED(CONFIG_IMA_KEXEC)) {
|
|
|
|
/* Setup IMA log buffer state */
|
|
|
|
setup_ima_state(image, params, params_load_addr,
|
|
|
|
setup_data_offset);
|
|
|
|
setup_data_offset += sizeof(struct setup_data) +
|
|
|
|
sizeof(struct ima_setup_data);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Setup RNG seed */
|
|
|
|
setup_rng_seed(params, params_load_addr, setup_data_offset);
|
2022-06-30 16:36:12 +08:00
|
|
|
|
2014-08-09 05:26:06 +08:00
|
|
|
/* Setup EDD info */
|
|
|
|
memcpy(params->eddbuf, boot_params.eddbuf,
|
|
|
|
EDDMAXNR * sizeof(struct edd_info));
|
|
|
|
params->eddbuf_entries = boot_params.eddbuf_entries;
|
|
|
|
|
|
|
|
memcpy(params->edd_mbr_sig_buffer, boot_params.edd_mbr_sig_buffer,
|
|
|
|
EDD_MBR_SIG_MAX * sizeof(unsigned int));
|
|
|
|
|
2014-08-09 05:26:09 +08:00
|
|
|
return ret;
|
2014-08-09 05:26:06 +08:00
|
|
|
}
|
|
|
|
|
2014-10-14 06:53:46 +08:00
|
|
|
static int bzImage64_probe(const char *buf, unsigned long len)
|
2014-08-09 05:26:06 +08:00
|
|
|
{
|
|
|
|
int ret = -ENOEXEC;
|
|
|
|
struct setup_header *header;
|
|
|
|
|
2016-02-24 07:34:30 +08:00
|
|
|
/* kernel should be at least two sectors long */
|
2014-08-09 05:26:06 +08:00
|
|
|
if (len < 2 * 512) {
|
|
|
|
pr_err("File is too short to be a bzImage\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
header = (struct setup_header *)(buf + offsetof(struct boot_params, hdr));
|
|
|
|
if (memcmp((char *)&header->header, "HdrS", 4) != 0) {
|
|
|
|
pr_err("Not a bzImage\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (header->boot_flag != 0xAA55) {
|
|
|
|
pr_err("No x86 boot sector present\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (header->version < 0x020C) {
|
|
|
|
pr_err("Must be at least protocol version 2.12\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!(header->loadflags & LOADED_HIGH)) {
|
|
|
|
pr_err("zImage not a bzImage\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!(header->xloadflags & XLF_KERNEL_64)) {
|
|
|
|
pr_err("Not a bzImage64. XLF_KERNEL_64 is not set.\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!(header->xloadflags & XLF_CAN_BE_LOADED_ABOVE_4G)) {
|
|
|
|
pr_err("XLF_CAN_BE_LOADED_ABOVE_4G is not set.\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:11 +08:00
|
|
|
/*
|
|
|
|
* Can't handle 32bit EFI as it does not allow loading kernel
|
|
|
|
* above 4G. This should be handled by 32bit bzImage loader
|
|
|
|
*/
|
|
|
|
if (efi_enabled(EFI_RUNTIME_SERVICES) && !efi_enabled(EFI_64BIT)) {
|
|
|
|
pr_debug("EFI is 32 bit. Can't load kernel above 4G.\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2019-05-24 15:38:09 +08:00
|
|
|
if (!(header->xloadflags & XLF_5LEVEL) && pgtable_l5_enabled()) {
|
|
|
|
pr_err("bzImage cannot handle 5-level paging mode.\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:06 +08:00
|
|
|
/* I've got a bzImage */
|
|
|
|
pr_debug("It's a relocatable bzImage64\n");
|
|
|
|
ret = 0;
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2014-10-14 06:53:46 +08:00
|
|
|
static void *bzImage64_load(struct kimage *image, char *kernel,
|
|
|
|
unsigned long kernel_len, char *initrd,
|
|
|
|
unsigned long initrd_len, char *cmdline,
|
|
|
|
unsigned long cmdline_len)
|
2014-08-09 05:26:06 +08:00
|
|
|
{
|
|
|
|
|
|
|
|
struct setup_header *header;
|
|
|
|
int setup_sects, kern16_size, ret = 0;
|
2016-11-29 20:45:48 +08:00
|
|
|
unsigned long setup_header_size, params_cmdline_sz;
|
2014-08-09 05:26:06 +08:00
|
|
|
struct boot_params *params;
|
|
|
|
unsigned long bootparam_load_addr, kernel_load_addr, initrd_load_addr;
|
|
|
|
struct bzimage64_data *ldata;
|
|
|
|
struct kexec_entry64_regs regs64;
|
|
|
|
void *stack;
|
|
|
|
unsigned int setup_hdr_offset = offsetof(struct boot_params, hdr);
|
2014-08-09 05:26:11 +08:00
|
|
|
unsigned int efi_map_offset, efi_map_sz, efi_setup_data_offset;
|
2016-11-29 20:45:48 +08:00
|
|
|
struct kexec_buf kbuf = { .image = image, .buf_max = ULONG_MAX,
|
|
|
|
.top_down = true };
|
2018-04-14 06:36:43 +08:00
|
|
|
struct kexec_buf pbuf = { .image = image, .buf_min = MIN_PURGATORY_ADDR,
|
|
|
|
.buf_max = ULONG_MAX, .top_down = true };
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
header = (struct setup_header *)(kernel + setup_hdr_offset);
|
|
|
|
setup_sects = header->setup_sects;
|
|
|
|
if (setup_sects == 0)
|
|
|
|
setup_sects = 4;
|
|
|
|
|
|
|
|
kern16_size = (setup_sects + 1) * 512;
|
|
|
|
if (kernel_len < kern16_size) {
|
|
|
|
pr_err("bzImage truncated\n");
|
|
|
|
return ERR_PTR(-ENOEXEC);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (cmdline_len > header->cmdline_size) {
|
|
|
|
pr_err("Kernel command line too long\n");
|
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:09 +08:00
|
|
|
/*
|
|
|
|
* In case of crash dump, we will append elfcorehdr=<addr> to
|
|
|
|
* command line. Make sure it does not overflow
|
|
|
|
*/
|
|
|
|
if (cmdline_len + MAX_ELFCOREHDR_STR_LEN > header->cmdline_size) {
|
|
|
|
pr_debug("Appending elfcorehdr=<addr> to command line exceeds maximum allowed length\n");
|
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Allocate and load backup region */
|
|
|
|
if (image->type == KEXEC_TYPE_CRASH) {
|
|
|
|
ret = crash_load_segments(image);
|
|
|
|
if (ret)
|
|
|
|
return ERR_PTR(ret);
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:06 +08:00
|
|
|
/*
|
|
|
|
* Load purgatory. For 64bit entry point, purgatory code can be
|
|
|
|
* anywhere.
|
|
|
|
*/
|
2018-04-14 06:36:43 +08:00
|
|
|
ret = kexec_load_purgatory(image, &pbuf);
|
2014-08-09 05:26:06 +08:00
|
|
|
if (ret) {
|
|
|
|
pr_err("Loading purgatory failed\n");
|
|
|
|
return ERR_PTR(ret);
|
|
|
|
}
|
|
|
|
|
2018-04-14 06:36:43 +08:00
|
|
|
pr_debug("Loaded purgatory at 0x%lx\n", pbuf.mem);
|
2014-08-09 05:26:06 +08:00
|
|
|
|
2014-08-09 05:26:11 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Load Bootparams and cmdline and space for efi stuff.
|
|
|
|
*
|
|
|
|
* Allocate memory together for multiple data structures so
|
|
|
|
* that they all can go in single area/segment and we don't
|
|
|
|
* have to create separate segment for each. Keeps things
|
|
|
|
* little bit simple
|
|
|
|
*/
|
|
|
|
efi_map_sz = efi_get_runtime_map_size();
|
2014-08-09 05:26:09 +08:00
|
|
|
params_cmdline_sz = sizeof(struct boot_params) + cmdline_len +
|
|
|
|
MAX_ELFCOREHDR_STR_LEN;
|
2014-08-09 05:26:11 +08:00
|
|
|
params_cmdline_sz = ALIGN(params_cmdline_sz, 16);
|
2018-04-21 05:56:10 +08:00
|
|
|
kbuf.bufsz = params_cmdline_sz + ALIGN(efi_map_sz, 16) +
|
2014-08-09 05:26:11 +08:00
|
|
|
sizeof(struct setup_data) +
|
x86/setup: Use rng seeds from setup_data
Currently, the only way x86 can get an early boot RNG seed is via EFI,
which is generally always used now for physical machines, but is very
rarely used in VMs, especially VMs that are optimized for starting
"instantaneously", such as Firecracker's MicroVM. For tiny fast booting
VMs, EFI is not something you generally need or want.
Rather, the image loader or firmware should be able to pass a single
random seed, exactly as device tree platforms do with the "rng-seed"
property. Additionally, this is something that bootloaders can append,
with their own seed file management, which is something every other
major OS ecosystem has that Linux does not (yet).
Add SETUP_RNG_SEED, similar to the other eight setup_data entries that
are parsed at boot. It also takes care to zero out the seed immediately
after using, in order to retain forward secrecy. This all takes about 7
trivial lines of code.
Then, on kexec_file_load(), a new fresh seed is generated and passed to
the next kernel, just as is done on device tree architectures when
using kexec. And, importantly, I've tested that QEMU is able to properly
pass SETUP_RNG_SEED as well, making this work for every step of the way.
This code too is pretty straight forward.
Together these measures ensure that VMs and nested kexec()'d kernels
always receive a proper boot time RNG seed at the earliest possible
stage from their parents:
- Host [already has strongly initialized RNG]
- QEMU [passes fresh seed in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- Linux [uses parent's seed and gathers entropy of its own]
- kexec [passes this in SETUP_RNG_SEED field]
- ...
I've verified in several scenarios that this works quite well from a
host kernel to QEMU and down inwards, mixing and matching loaders, with
every layer providing a seed to the next.
[ bp: Massage commit message. ]
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Link: https://lore.kernel.org/r/20220630113300.1892799-1-Jason@zx2c4.com
2022-07-11 01:29:21 +08:00
|
|
|
sizeof(struct efi_setup_data) +
|
|
|
|
sizeof(struct setup_data) +
|
|
|
|
RNG_SEED_LENGTH;
|
2014-08-09 05:26:11 +08:00
|
|
|
|
2022-06-30 16:36:12 +08:00
|
|
|
if (IS_ENABLED(CONFIG_IMA_KEXEC))
|
|
|
|
kbuf.bufsz += sizeof(struct setup_data) +
|
|
|
|
sizeof(struct ima_setup_data);
|
|
|
|
|
2016-11-29 20:45:48 +08:00
|
|
|
params = kzalloc(kbuf.bufsz, GFP_KERNEL);
|
2014-08-09 05:26:06 +08:00
|
|
|
if (!params)
|
|
|
|
return ERR_PTR(-ENOMEM);
|
2014-08-09 05:26:11 +08:00
|
|
|
efi_map_offset = params_cmdline_sz;
|
2018-04-21 05:56:10 +08:00
|
|
|
efi_setup_data_offset = efi_map_offset + ALIGN(efi_map_sz, 16);
|
2014-08-09 05:26:06 +08:00
|
|
|
|
2019-06-08 02:54:32 +08:00
|
|
|
/* Copy setup header onto bootparams. Documentation/x86/boot.rst */
|
2014-08-09 05:26:06 +08:00
|
|
|
setup_header_size = 0x0202 + kernel[0x0201] - setup_hdr_offset;
|
|
|
|
|
|
|
|
/* Is there a limit on setup header size? */
|
|
|
|
memcpy(¶ms->hdr, (kernel + setup_hdr_offset), setup_header_size);
|
|
|
|
|
2016-11-29 20:45:48 +08:00
|
|
|
kbuf.buffer = params;
|
|
|
|
kbuf.memsz = kbuf.bufsz;
|
|
|
|
kbuf.buf_align = 16;
|
|
|
|
kbuf.buf_min = MIN_BOOTPARAM_ADDR;
|
|
|
|
ret = kexec_add_buffer(&kbuf);
|
2014-08-09 05:26:06 +08:00
|
|
|
if (ret)
|
|
|
|
goto out_free_params;
|
2016-11-29 20:45:48 +08:00
|
|
|
bootparam_load_addr = kbuf.mem;
|
2014-08-09 05:26:11 +08:00
|
|
|
pr_debug("Loaded boot_param, command line and misc at 0x%lx bufsz=0x%lx memsz=0x%lx\n",
|
2016-11-29 20:45:48 +08:00
|
|
|
bootparam_load_addr, kbuf.bufsz, kbuf.bufsz);
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
/* Load kernel */
|
2016-11-29 20:45:48 +08:00
|
|
|
kbuf.buffer = kernel + kern16_size;
|
|
|
|
kbuf.bufsz = kernel_len - kern16_size;
|
|
|
|
kbuf.memsz = PAGE_ALIGN(header->init_size);
|
|
|
|
kbuf.buf_align = header->kernel_alignment;
|
|
|
|
kbuf.buf_min = MIN_KERNEL_LOAD_ADDR;
|
2018-12-28 09:12:47 +08:00
|
|
|
kbuf.mem = KEXEC_BUF_MEM_UNKNOWN;
|
2016-11-29 20:45:48 +08:00
|
|
|
ret = kexec_add_buffer(&kbuf);
|
2014-08-09 05:26:06 +08:00
|
|
|
if (ret)
|
|
|
|
goto out_free_params;
|
2016-11-29 20:45:48 +08:00
|
|
|
kernel_load_addr = kbuf.mem;
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
pr_debug("Loaded 64bit kernel at 0x%lx bufsz=0x%lx memsz=0x%lx\n",
|
2016-11-29 20:45:48 +08:00
|
|
|
kernel_load_addr, kbuf.bufsz, kbuf.memsz);
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
/* Load initrd high */
|
|
|
|
if (initrd) {
|
2016-11-29 20:45:48 +08:00
|
|
|
kbuf.buffer = initrd;
|
|
|
|
kbuf.bufsz = kbuf.memsz = initrd_len;
|
|
|
|
kbuf.buf_align = PAGE_SIZE;
|
|
|
|
kbuf.buf_min = MIN_INITRD_LOAD_ADDR;
|
2018-12-28 09:12:47 +08:00
|
|
|
kbuf.mem = KEXEC_BUF_MEM_UNKNOWN;
|
2016-11-29 20:45:48 +08:00
|
|
|
ret = kexec_add_buffer(&kbuf);
|
2014-08-09 05:26:06 +08:00
|
|
|
if (ret)
|
|
|
|
goto out_free_params;
|
2016-11-29 20:45:48 +08:00
|
|
|
initrd_load_addr = kbuf.mem;
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
pr_debug("Loaded initrd at 0x%lx bufsz=0x%lx memsz=0x%lx\n",
|
|
|
|
initrd_load_addr, initrd_len, initrd_len);
|
|
|
|
|
|
|
|
setup_initrd(params, initrd_load_addr, initrd_len);
|
|
|
|
}
|
|
|
|
|
2014-08-09 05:26:09 +08:00
|
|
|
setup_cmdline(image, params, bootparam_load_addr,
|
|
|
|
sizeof(struct boot_params), cmdline, cmdline_len);
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
/* bootloader info. Do we need a separate ID for kexec kernel loader? */
|
|
|
|
params->hdr.type_of_loader = 0x0D << 4;
|
|
|
|
params->hdr.loadflags = 0;
|
|
|
|
|
|
|
|
/* Setup purgatory regs for entry */
|
|
|
|
ret = kexec_purgatory_get_set_symbol(image, "entry64_regs", ®s64,
|
|
|
|
sizeof(regs64), 1);
|
|
|
|
if (ret)
|
|
|
|
goto out_free_params;
|
|
|
|
|
|
|
|
regs64.rbx = 0; /* Bootstrap Processor */
|
|
|
|
regs64.rsi = bootparam_load_addr;
|
|
|
|
regs64.rip = kernel_load_addr + 0x200;
|
|
|
|
stack = kexec_purgatory_get_symbol_addr(image, "stack_end");
|
|
|
|
if (IS_ERR(stack)) {
|
|
|
|
pr_err("Could not find address of symbol stack_end\n");
|
|
|
|
ret = -EINVAL;
|
|
|
|
goto out_free_params;
|
|
|
|
}
|
|
|
|
|
|
|
|
regs64.rsp = (unsigned long)stack;
|
|
|
|
ret = kexec_purgatory_get_set_symbol(image, "entry64_regs", ®s64,
|
|
|
|
sizeof(regs64), 0);
|
|
|
|
if (ret)
|
|
|
|
goto out_free_params;
|
|
|
|
|
2014-08-09 05:26:11 +08:00
|
|
|
ret = setup_boot_parameters(image, params, bootparam_load_addr,
|
|
|
|
efi_map_offset, efi_map_sz,
|
|
|
|
efi_setup_data_offset);
|
2014-08-09 05:26:09 +08:00
|
|
|
if (ret)
|
|
|
|
goto out_free_params;
|
2014-08-09 05:26:06 +08:00
|
|
|
|
|
|
|
/* Allocate loader specific data */
|
|
|
|
ldata = kzalloc(sizeof(struct bzimage64_data), GFP_KERNEL);
|
|
|
|
if (!ldata) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out_free_params;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Store pointer to params so that it could be freed after loading
|
|
|
|
* params segment has been loaded and contents have been copied
|
|
|
|
* somewhere else.
|
|
|
|
*/
|
|
|
|
ldata->bootparams_buf = params;
|
|
|
|
return ldata;
|
|
|
|
|
|
|
|
out_free_params:
|
|
|
|
kfree(params);
|
|
|
|
return ERR_PTR(ret);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This cleanup function is called after various segments have been loaded */
|
2014-10-14 06:53:46 +08:00
|
|
|
static int bzImage64_cleanup(void *loader_data)
|
2014-08-09 05:26:06 +08:00
|
|
|
{
|
|
|
|
struct bzimage64_data *ldata = loader_data;
|
|
|
|
|
|
|
|
if (!ldata)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
kfree(ldata->bootparams_buf);
|
|
|
|
ldata->bootparams_buf = NULL;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2018-04-14 06:35:49 +08:00
|
|
|
const struct kexec_file_ops kexec_bzImage64_ops = {
|
2014-08-09 05:26:06 +08:00
|
|
|
.probe = bzImage64_probe,
|
|
|
|
.load = bzImage64_load,
|
|
|
|
.cleanup = bzImage64_cleanup,
|
2014-08-09 05:26:13 +08:00
|
|
|
#ifdef CONFIG_KEXEC_BZIMAGE_VERIFY_SIG
|
2022-07-14 21:40:25 +08:00
|
|
|
.verify_sig = kexec_kernel_verify_pe_sig,
|
2014-08-09 05:26:13 +08:00
|
|
|
#endif
|
2014-08-09 05:26:06 +08:00
|
|
|
};
|