RISC-V: Add firmware loading support and default
Subsystems (3)
cover RISC-V: Add firmware loading support and default
This is an RFC as it will break ALL current users! See below for details.
This series consolidates the current RISC-V kernel loading impelementation while also adding support for the -bios option and more advanced kernel image types.
After consolidating the kernel loading we can extend the boot loader to support a -bios option. We can also extend the kernel loading options to support not just ELF files but other standard formats.
Finally we can include the OpenSBI firmware by default for QEMU users.
At the end of this series we are in the good place of no longer requiring users to build firmware to boot a kernel. Instead users can just run QEMu with the -kernel option and everything will work. They can also override the firmware with their own using the -bios option. Using "-bios none" will result in no firmware being loaded (as it is today).
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Unfortunately this series (patch 5 specifically) results in all current Linux boots being broken as users are already loading in their own firmware, which overlaps with the now included default. They will see this error:
rom: requested regions overlap (rom phdr #0: ./images/qemuriscv64/fw_jump.elf. free=0x0000000080008090, addr=0x0000000080000000)
qemu-system-riscv64: rom check and register reset failed
If a current user specieifies "-bios none" their flow will continue working. A user can instead no longer load their firmware and use the default or they can load their firmware with the -bios option.
At the moment the best idea I have to not break all users is to only include the default firmware if the user specifies "-bios opensbi". That is change the default to not loading the firmware. Then we can work on updating documentation and maybe in future change the default to include a firmware and anyone who doesn't want a default firmware can specify "-bios none".
Any other ideas on how to not break everything?
Alistair Francis (5):
hw/riscv: Split out the boot functions
hw/riscv: Add support for loading a firmware
hw/riscv: Extend the kernel loading support
roms: Add OpenSBI version 0.3
hw/riscv: Load OpenSBI as the default firmware
.gitmodules | 3 + Makefile | 3 +- configure | 1 + hw/riscv/Makefile.objs | 1 + hw/riscv/boot.c | 143 ++++++++++++++++++++++++++++ hw/riscv/sifive_e.c | 17 +--- hw/riscv/sifive_u.c | 19 +--- hw/riscv/spike.c | 21 +--- hw/riscv/virt.c | 54 ++--------- include/hw/riscv/boot.h | 30 ++++++ pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes roms/Makefile | 17 ++++ roms/opensbi | 1 + 14 files changed, 216 insertions(+), 94 deletions(-) create mode 100644 hw/riscv/boot.c create mode 100644 include/hw/riscv/boot.h create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbi
Signature · 1 line
2.22.0
RFC 1/5 hw/riscv: Split out the boot functions
Split the common RISC-V boot functions into a seperate file. This allows us to share the common code.
--- hw/riscv/Makefile.objs | 1 + hw/riscv/boot.c | 69 +++++++++++++++++++++++++++++++++++++++++ hw/riscv/sifive_e.c | 17 ++-------- hw/riscv/sifive_u.c | 17 ++-------- hw/riscv/spike.c | 21 +++---------- hw/riscv/virt.c | 51 +++--------------------------- include/hw/riscv/boot.h | 27 ++++++++++++++++ 7 files changed, 110 insertions(+), 93 deletions(-) create mode 100644 hw/riscv/boot.c create mode 100644 include/hw/riscv/boot.h
hw/riscv/Makefile.objs +1 -0
@@ -1,3 +1,4 @@ +obj-y += boot.o obj-$(CONFIG_SPIKE) += riscv_htif.o obj-$(CONFIG_HART) += riscv_hart.o obj-$(CONFIG_SIFIVE_E) += sifive_e.o
hw/riscv/boot.c new file +69 -0
69 lines not on this page · raw message
hw/riscv/sifive_e.c +2 -15
@@ -44,10 +44,10 @@#include "hw/riscv/sifive_prci.h" #include "hw/riscv/sifive_uart.h" #include "hw/riscv/sifive_e.h" +#include "hw/riscv/boot.h" #include "chardev/char.h" #include "sysemu/arch_init.h" #include "exec/address-spaces.h" -#include "elf.h" static const struct MemmapEntry { hwaddr base;
@@ -74,19 +74,6 @@ static const struct MemmapEntry { [SIFIVE_E_DTIM] = { 0x80000000, 0x4000 } }; -static target_ulong load_kernel (const char *kernel_filename) -{ -uint64_t kernel_entry, kernel_high; - -if (load_elf (kernel_filename,NULL ,NULL ,NULL , - &kernel_entry,NULL , &kernel_high, -0 ,EM_RISCV ,1 ,0 ) <0 ) { -error_report ("could not load kernel '%s'" , kernel_filename); -exit (1 ); - } -return kernel_entry; -} -static void sifive_mmio_emulate (MemoryRegion *parent,const char *name,uintptr_t offset,uintptr_t length) {
@@ -131,7 +118,7 @@ static void riscv_sifive_e_init(MachineState *machine) memmap[SIFIVE_E_MROM].base, &address_space_memory);if (machine->kernel_filename) { -load_kernel (machine->kernel_filename); +riscv_load_kernel (machine->kernel_filename); } }
hw/riscv/sifive_u.c +2 -15
@@ -41,11 +41,11 @@#include "hw/riscv/sifive_uart.h" #include "hw/riscv/sifive_prci.h" #include "hw/riscv/sifive_u.h" +#include "hw/riscv/boot.h" #include "chardev/char.h" #include "sysemu/arch_init.h" #include "sysemu/device_tree.h" #include "exec/address-spaces.h" -#include "elf.h" #include <libfdt.h>
@@ -65,19 +65,6 @@ static const struct MemmapEntry {#define GEM_REVISION 0x10070109 -static target_ulong load_kernel (const char *kernel_filename) -{ -uint64_t kernel_entry, kernel_high; - -if (load_elf (kernel_filename,NULL ,NULL ,NULL , - &kernel_entry,NULL , &kernel_high, -0 ,EM_RISCV ,1 ,0 ) <0 ) { -error_report ("could not load kernel '%s'" , kernel_filename); -exit (1 ); - } -return kernel_entry; -} -static void create_fdt (SiFiveUState *s,const struct MemmapEntry *memmap,uint64_t mem_size,const char *cmdline) {
@@ -280,7 +267,7 @@ static void riscv_sifive_u_init(MachineState *machine)create_fdt (s, memmap, machine->ram_size, machine->kernel_cmdline);if (machine->kernel_filename) { -load_kernel (machine->kernel_filename); +riscv_load_kernel (machine->kernel_filename); }/* reset vector */
hw/riscv/spike.c +4 -17
@@ -36,12 +36,12 @@#include "hw/riscv/riscv_hart.h" #include "hw/riscv/sifive_clint.h" #include "hw/riscv/spike.h" +#include "hw/riscv/boot.h" #include "chardev/char.h" #include "sysemu/arch_init.h" #include "sysemu/device_tree.h" #include "sysemu/qtest.h" #include "exec/address-spaces.h" -#include "elf.h" #include <libfdt.h>
@@ -54,19 +54,6 @@ static const struct MemmapEntry { [SPIKE_DRAM] = { 0x80000000, 0x0 }, }; -static target_ulong load_kernel (const char *kernel_filename) -{ -uint64_t kernel_entry, kernel_high; - -if (load_elf_ram_sym (kernel_filename,NULL ,NULL ,NULL , - &kernel_entry,NULL , &kernel_high,0 ,EM_RISCV ,1 ,0 , -NULL , true, htif_symbol_callback) <0 ) { -error_report ("could not load kernel '%s'" , kernel_filename); -exit (1 ); - } -return kernel_entry; -} -static void create_fdt (SpikeState *s,const struct MemmapEntry *memmap,uint64_t mem_size,const char *cmdline) {
@@ -199,7 +186,7 @@ static void spike_board_init(MachineState *machine) mask_rom);if (machine->kernel_filename) { -load_kernel (machine->kernel_filename); +riscv_load_kernel (machine->kernel_filename); }/* reset vector */
@@ -287,7 +274,7 @@ static void spike_v1_10_0_board_init(MachineState *machine) mask_rom);if (machine->kernel_filename) { -load_kernel (machine->kernel_filename); +riscv_load_kernel (machine->kernel_filename); }/* reset vector */
@@ -372,7 +359,7 @@ static void spike_v1_09_1_board_init(MachineState *machine) mask_rom);if (machine->kernel_filename) { -load_kernel (machine->kernel_filename); +riscv_load_kernel (machine->kernel_filename); }/* reset vector */
hw/riscv/virt.c +5 -46
78 lines not on this page · raw message
include/hw/riscv/boot.h new file +27 -0
@@ -0,0 +1,27 @@ +/* +* QEMU RISC-V Boot Helper +* +* Copyright (c) 2017 SiFive, Inc. +* Copyright (c) 2019 Alistair Francis <[email protected]> +* +* This program is free software; you can redistribute it and/or modify it +* under the terms and conditions of the GNU General Public License, +* version 2 or later, as published by the Free Software Foundation. +* +* This program is distributed in the hope 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, see <http://www.gnu.org/licenses/>. +*/ + +#ifndef RISCV_BOOT_H +#define RISCV_BOOT_H + +target_ulong riscv_load_kernel (const char *kernel_filename); +hwaddr riscv_load_initrd (const char *filename,uint64_t mem_size, +uint64_t kernel_entry,hwaddr *start); + +#endif /* RISCV_BOOT_H */
Signature · 1 line
2.22.0
Re: [Qemu-devel] [RFC v1 1/5] hw/riscv: Split out the boot functions
Quoted 69 lines · Alistair Francis
Split the common RISC-V boot functions into a seperate file. This allows us to share the common code.
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/Makefile.objs | 1 + hw/riscv/boot.c | 69 +++++++++++++++++++++++++++++++++++++++++ hw/riscv/sifive_e.c | 17 ++-------- hw/riscv/sifive_u.c | 17 ++-------- hw/riscv/spike.c | 21 +++---------- hw/riscv/virt.c | 51 +++--------------------------- include/hw/riscv/boot.h | 27 ++++++++++++++++ 7 files changed, 110 insertions(+), 93 deletions(-) create mode 100644 hw/riscv/boot.c create mode 100644 include/hw/riscv/boot.hhw/riscv/Makefile.objs +1
-0@@ -1,3 +1,4 @@ +obj-y += boot.o obj-$(CONFIG_SPIKE) += riscv_htif.o obj-$(CONFIG_HART) += riscv_hart.o obj-$(CONFIG_SIFIVE_E) += sifive_e.ohw/riscv/boot.c new file +37
-0@@ -0,0 +1,69 @@ +/* +* QEMU RISC-V Boot Helper +* +* Copyright (c) 2017 SiFive, Inc. +* Copyright (c) 2019 Alistair Francis <[email protected]> +* +* This program is free software; you can redistribute it and/or modify it +* under the terms and conditions of the GNU General Public License, +* version 2 or later, as published by the Free Software Foundation. +* +* This program is distributed in the hope 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, see <http://www.gnu.org/licenses/>. +*/ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "qemu/error-report.h" +#include "exec/cpu-defs.h" +#include "hw/loader.h" +#include "hw/riscv/boot.h" +#include "elf.h" + +target_ulong riscv_load_kernel (const char *kernel_filename) +{ +uint64_t kernel_entry, kernel_high; + +if (load_elf (kernel_filename,NULL ,NULL ,NULL , + &kernel_entry,NULL , &kernel_high, +0 ,EM_RISCV ,1 ,0 ) <0 ) { +error_report ("could not load kernel '%s'" , kernel_filename); +exit (1 ); + }
hw/riscv/boot.c line 47
Quoted 10 lines
hw/riscv/boot.c new file +10
-0+ +return kernel_entry; +} + +hwaddr riscv_load_initrd (const char *filename,uint64_t mem_size, +uint64_t kernel_entry,hwaddr *start) +{ +int size; + +/* We want to put the initrd far enough into RAM that when the
nits: /* should be on a separate line
Quoted 9 lines
hw/riscv/boot.c new file +9
-0+* kernel is uncompressed it will not clobber the initrd. However +* on boards without much RAM we must ensure that we still leave +* enough room for a decent sized initrd, and on boards with large +* amounts of RAM we must avoid the initrd being so far up in RAM +* that it is outside lowmem and inaccessible to the kernel. +* So for boards with less than 256MB of RAM we put the initrd +* halfway into RAM, and for boards with 256MB of RAM or more we put +* the initrd at 128MB. +*/
[snip]
Re: [Qemu-devel] [RFC v1 1/5] hw/riscv: Split out the boot functions
Quoted 76 lines · Bin Meng
Quoted 73 lines · Alistair Francis
Split the common RISC-V boot functions into a seperate file. This allows us to share the common code.
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/Makefile.objs | 1 + hw/riscv/boot.c | 69 +++++++++++++++++++++++++++++++++++++++++ hw/riscv/sifive_e.c | 17 ++-------- hw/riscv/sifive_u.c | 17 ++-------- hw/riscv/spike.c | 21 +++---------- hw/riscv/virt.c | 51 +++--------------------------- include/hw/riscv/boot.h | 27 ++++++++++++++++ 7 files changed, 110 insertions(+), 93 deletions(-) create mode 100644 hw/riscv/boot.c create mode 100644 include/hw/riscv/boot.hhw/riscv/Makefile.objs +1
-0@@ -1,3 +1,4 @@ +obj-y += boot.o obj-$(CONFIG_SPIKE) += riscv_htif.o obj-$(CONFIG_HART) += riscv_hart.o obj-$(CONFIG_SIFIVE_E) += sifive_e.ohw/riscv/boot.c new file +41
-0@@ -0,0 +1,69 @@ +/* +* QEMU RISC-V Boot Helper +* +* Copyright (c) 2017 SiFive, Inc. +* Copyright (c) 2019 Alistair Francis <[email protected]> +* +* This program is free software; you can redistribute it and/or modify it +* under the terms and conditions of the GNU General Public License, +* version 2 or later, as published by the Free Software Foundation. +* +* This program is distributed in the hope 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, see <http://www.gnu.org/licenses/>. +*/ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "qemu/error-report.h" +#include "exec/cpu-defs.h" +#include "hw/loader.h" +#include "hw/riscv/boot.h" +#include "elf.h" + +target_ulong riscv_load_kernel (const char *kernel_filename) +{ +uint64_t kernel_entry, kernel_high; + +if (load_elf (kernel_filename,NULL ,NULL ,NULL , + &kernel_entry,NULL , &kernel_high, +0 ,EM_RISCV ,1 ,0 ) <0 ) { +error_report ("could not load kernel '%s'" , kernel_filename); +exit (1 ); + } + +return kernel_entry; +} +
Quoted 8 lines
hw/riscv/boot.c line 47
Quoted 6 lines
hw/riscv/boot.c new file +6
-0+hwaddr riscv_load_initrd (const char *filename,uint64_t mem_size, +uint64_t kernel_entry,hwaddr *start) +{ +int size; + +/* We want to put the initrd far enough into RAM that when the nits: /* should be on a separate line
Fixed
Quoted 11 lines
hw/riscv/boot.c line 56
Quoted 10 lines
hw/riscv/boot.c new file +9
-0+* kernel is uncompressed it will not clobber the initrd. However +* on boards without much RAM we must ensure that we still leave +* enough room for a decent sized initrd, and on boards with large +* amounts of RAM we must avoid the initrd being so far up in RAM +* that it is outside lowmem and inaccessible to the kernel. +* So for boards with less than 256MB of RAM we put the initrd +* halfway into RAM, and for boards with 256MB of RAM or more we put +* the initrd at 128MB. +*/
Quoted 4 lines
[snip]
Reviewed-by Bin Meng <[email protected]>Tested-by Bin Meng <[email protected]>
Thanks!
Alistair
RFC 2/5 hw/riscv: Add support for loading a firmware
In reply to Alistair FrancisAdd support for loading a firmware file for the virt machine and the SiFive U. This can be run with the following command:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel vmlinux
--- hw/riscv/boot.c | 41 +++++++++++++++++++++++++++++++++++++++-- hw/riscv/sifive_e.c | 2 +- hw/riscv/sifive_u.c | 6 +++++- hw/riscv/spike.c | 6 +++--- hw/riscv/virt.c | 7 ++++++- include/hw/riscv/boot.h | 4 +++- 6 files changed, 57 insertions(+), 9 deletions(-)
hw/riscv/boot.c +39 -2
@@ -23,13 +23,50 @@#include "exec/cpu-defs.h" #include "hw/loader.h" #include "hw/riscv/boot.h" +#include "hw/boards.h" #include "elf.h" -target_ulong riscv_load_kernel (const char *kernel_filename) +#if defined(TARGET_RISCV32 ) +# define KERNEL_BOOT_ADDRESS 0x80400000 +#else +# define KERNEL_BOOT_ADDRESS 0x80200000 +#endif + +static uint64_t kernel_translate (void *opaque,uint64_t addr) +{ +MachineState *machine = opaque; + +/* +* If the user specified a firmware move the kernel to the offset +* start address. +*/ +if (machine->firmware) { +return (addr &0x7fffffff ) +KERNEL_BOOT_ADDRESS ; + }else { +return addr; + } +} + +target_ulong riscv_load_firmware (const char *firmware_filename) +{ +uint64_t firmware_entry, firmware_start, firmware_end; + +if (load_elf (firmware_filename,NULL ,NULL ,NULL , + &firmware_entry, &firmware_start, &firmware_end, +0 ,EM_RISCV ,1 ,0 ) <0 ) { +error_report ("could not load firmware '%s'" , firmware_filename); +exit (1 ); + } + +return firmware_entry; +} + +target_ulong riscv_load_kernel (MachineState *machine, +const char *kernel_filename) {uint64_t kernel_entry, kernel_high; -if (load_elf (kernel_filename,NULL ,NULL ,NULL , +if (load_elf (kernel_filename,NULL , kernel_translate, machine, &kernel_entry,NULL , &kernel_high,0 ,EM_RISCV ,1 ,0 ) <0 ) {error_report ("could not load kernel '%s'" , kernel_filename);
hw/riscv/sifive_e.c +1 -1
@@ -118,7 +118,7 @@ static void riscv_sifive_e_init(MachineState *machine) memmap[SIFIVE_E_MROM].base, &address_space_memory);if (machine->kernel_filename) { -riscv_load_kernel (machine->kernel_filename); +riscv_load_kernel (machine, machine->kernel_filename); } }
hw/riscv/sifive_u.c +5 -1
@@ -266,8 +266,12 @@ static void riscv_sifive_u_init(MachineState *machine)/* create device tree */ create_fdt (s, memmap, machine->ram_size, machine->kernel_cmdline); +if (machine->firmware) { +riscv_load_firmware (machine->firmware); + } +if (machine->kernel_filename) { -riscv_load_kernel (machine->kernel_filename); +riscv_load_kernel (machine, machine->kernel_filename); }/* reset vector */
hw/riscv/spike.c +3 -3
@@ -186,7 +186,7 @@ static void spike_board_init(MachineState *machine) mask_rom);if (machine->kernel_filename) { -riscv_load_kernel (machine->kernel_filename); +riscv_load_kernel (machine, machine->kernel_filename); }/* reset vector */
@@ -274,7 +274,7 @@ static void spike_v1_10_0_board_init(MachineState *machine) mask_rom);if (machine->kernel_filename) { -riscv_load_kernel (machine->kernel_filename); +riscv_load_kernel (machine, machine->kernel_filename); }/* reset vector */
@@ -359,7 +359,7 @@ static void spike_v1_09_1_board_init(MachineState *machine) mask_rom);if (machine->kernel_filename) { -riscv_load_kernel (machine->kernel_filename); +riscv_load_kernel (machine, machine->kernel_filename); }/* reset vector */
hw/riscv/virt.c +6 -1
@@ -380,8 +380,13 @@ static void riscv_virt_board_init(MachineState *machine)memory_region_add_subregion (system_memory, memmap[VIRT_MROM ].base, mask_rom); +if (machine->firmware) { +riscv_load_firmware (machine->firmware); + } +if (machine->kernel_filename) { -uint64_t kernel_entry =riscv_load_kernel (machine->kernel_filename); +uint64_t kernel_entry =riscv_load_kernel (machine, + machine->kernel_filename);if (machine->initrd_filename) {hwaddr start;
include/hw/riscv/boot.h +3 -1
@@ -20,7 +20,9 @@#ifndef RISCV_BOOT_H #define RISCV_BOOT_H -target_ulong riscv_load_kernel (const char *kernel_filename); +target_ulong riscv_load_firmware (const char *firmware_filename); +target_ulong riscv_load_kernel (MachineState *machine, +const char *kernel_filename);hwaddr riscv_load_initrd (const char *filename,uint64_t mem_size,uint64_t kernel_entry,hwaddr *start);
Signature · 1 line
2.22.0
Re: [Qemu-devel] [RFC v1 2/5] hw/riscv: Add support for loading a firmware
Quoted 32 lines · Alistair Francis
Add support for loading a firmware file for the virt machine and the SiFive U. This can be run with the following command:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel vmlinux
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/boot.c | 41 +++++++++++++++++++++++++++++++++++++++-- hw/riscv/sifive_e.c | 2 +- hw/riscv/sifive_u.c | 6 +++++- hw/riscv/spike.c | 6 +++--- hw/riscv/virt.c | 7 ++++++- include/hw/riscv/boot.h | 4 +++- 6 files changed, 57 insertions(+), 9 deletions(-)hw/riscv/boot.c +5
-1@@ -23,13 +23,50 @@#include "exec/cpu-defs.h" #include "hw/loader.h" #include "hw/riscv/boot.h" +#include "hw/boards.h" #include "elf.h" -target_ulong riscv_load_kernel (const char *kernel_filename) +#if defined(TARGET_RISCV32 ) +# define KERNEL_BOOT_ADDRESS 0x80400000 +#else +# define KERNEL_BOOT_ADDRESS 0x80200000
hw/riscv/boot.c line 42
Quoted 10 lines
hw/riscv/boot.c +10
-0+#endif + +static uint64_t kernel_translate (void *opaque,uint64_t addr) +{ +MachineState *machine = opaque; + +/* +* If the user specified a firmware move the kernel to the offset +* start address. +*/
Why?
hw/riscv/boot.c line 44
Quoted 2 lines
hw/riscv/boot.c +2
-0+if (machine->firmware) { +return (addr &0x7fffffff ) +KERNEL_BOOT_ADDRESS ;
So with both "-bios" and "-kernel", the kernel address will be moved to another address other than 0x80200000 (for 64-bit). This does not look good to me.
hw/riscv/boot.c line 49
Quoted 5 lines
hw/riscv/boot.c +5
-0+ }else { +return addr; + } +} +
[snip]
Regards,
Bin
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 2/5] hw/riscv: Add support for loading a firmware
I was actually just writing up the same thing. Shifting all the addresses in the ELF file by 2 or 4MB is somewhat surprising behavior, especially because this will apply to segments that are mapped even at much higher addresses. If you want a segment aligned to a 1GB superpage boundary you now need to place it slightly below so that it will be bumped up to the right place. Further, ANDing all addresses with 0x7fffffff makes it impossible to map anything beyond the first 2GB of RAM.
Jonathan
Quoted 66 lines · Bin Meng
Quoted 32 lines · Alistair Francis
Add support for loading a firmware file for the virt machine and the SiFive U. This can be run with the following command:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel vmlinux
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/boot.c | 41 +++++++++++++++++++++++++++++++++++++++-- hw/riscv/sifive_e.c | 2 +- hw/riscv/sifive_u.c | 6 +++++- hw/riscv/spike.c | 6 +++--- hw/riscv/virt.c | 7 ++++++- include/hw/riscv/boot.h | 4 +++- 6 files changed, 57 insertions(+), 9 deletions(-)hw/riscv/boot.c +5
-1@@ -23,13 +23,50 @@#include "exec/cpu-defs.h" #include "hw/loader.h" #include "hw/riscv/boot.h" +#include "hw/boards.h" #include "elf.h" -target_ulong riscv_load_kernel (const char *kernel_filename) +#if defined(TARGET_RISCV32 ) +# define KERNEL_BOOT_ADDRESS 0x80400000 +#else +# define KERNEL_BOOT_ADDRESS 0x80200000hw/riscv/boot.c line 42
Quoted 10 lines
hw/riscv/boot.c +10
-0+#endif + +static uint64_t kernel_translate (void *opaque,uint64_t addr) +{ +MachineState *machine = opaque; + +/* +* If the user specified a firmware move the kernel to the offset +* start address. +*/ Why?
hw/riscv/boot.c line 44
Quoted 2 lines
hw/riscv/boot.c +2
-0+if (machine->firmware) { +return (addr &0x7fffffff ) +KERNEL_BOOT_ADDRESS ;So with both "-bios" and "-kernel", the kernel address will be moved to another address other than
0x80200000(for 64-bit). This does not look good to me.hw/riscv/boot.c line 49
Quoted 5 lines
hw/riscv/boot.c +5
-0+ }else { +return addr; + } +} +[snip]
Regards,
Bin
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 2/5] hw/riscv: Add support for loading a firmware
Hi,
Quoted 3 lines · Jonathan Behrens
I was actually just writing up the same thing. Shifting all the addresses in the ELF file by 2 or 4MB is somewhat surprising behavior, especially because this will apply to segments that are mapped even at much higher addresses. If you want a segment aligned to a 1GB superpage boundary you now need to place it slightly below so that it will be bumped up to the right place. Further, ANDing all addresses with
0x7fffffffmakes it impossible to map anything beyond the first 2GB of RAM.
Yes, current kernel_translate() logic is tightly coupled to the kernel entry VA, and if we link kernel at some other address it will just fail.
Quoted 58 lines
Jonathan
Quoted 55 lines · Bin Meng
Quoted 32 lines · Alistair Francis
Add support for loading a firmware file for the virt machine and the SiFive U. This can be run with the following command:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel vmlinux
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/boot.c | 41 +++++++++++++++++++++++++++++++++++++++-- hw/riscv/sifive_e.c | 2 +- hw/riscv/sifive_u.c | 6 +++++- hw/riscv/spike.c | 6 +++--- hw/riscv/virt.c | 7 ++++++- include/hw/riscv/boot.h | 4 +++- 6 files changed, 57 insertions(+), 9 deletions(-)hw/riscv/boot.c +5
-1@@ -23,13 +23,50 @@#include "exec/cpu-defs.h" #include "hw/loader.h" #include "hw/riscv/boot.h" +#include "hw/boards.h" #include "elf.h" -target_ulong riscv_load_kernel (const char *kernel_filename) +#if defined(TARGET_RISCV32 ) +# define KERNEL_BOOT_ADDRESS 0x80400000 +#else +# define KERNEL_BOOT_ADDRESS 0x80200000hw/riscv/boot.c line 42
Quoted 10 lines
hw/riscv/boot.c +10
-0+#endif + +static uint64_t kernel_translate (void *opaque,uint64_t addr) +{ +MachineState *machine = opaque; + +/* +* If the user specified a firmware move the kernel to the offset +* start address. +*/ Why?
hw/riscv/boot.c line 44
Quoted 2 lines
hw/riscv/boot.c +2
-0+if (machine->firmware) { +return (addr &0x7fffffff ) +KERNEL_BOOT_ADDRESS ;So with both "-bios" and "-kernel", the kernel address will be moved to another address other than
0x80200000(for 64-bit). This does not look good to me.
So why not simply return KERNEL_BOOT_ADDRESS in kernel_translate()?
Regards,
Bin
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 2/5] hw/riscv: Add support for loading a firmware
Quoted 11 lines · Bin Meng
Hi,
Quoted 3 lines · Jonathan Behrens
I was actually just writing up the same thing. Shifting all the addresses in the ELF file by 2 or 4MB is somewhat surprising behavior, especially because this will apply to segments that are mapped even at much higher addresses. If you want a segment aligned to a 1GB superpage boundary you now need to place it slightly below so that it will be bumped up to the right place. Further, ANDing all addresses with
0x7fffffffmakes it impossible to map anything beyond the first 2GB of RAM.Yes, current
kernel_translate()logic is tightly coupled to the kernel entry VA, and if we link kernel at some other address it will just fail.
I thought this was required but it looks like it isn't. I have remove the kernel_translate() function.
Quoted 43 lines
Quoted 43 lines
Jonathan
Quoted 39 lines · Bin Meng
Quoted 36 lines · Alistair Francis
Add support for loading a firmware file for the virt machine and the SiFive U. This can be run with the following command:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel vmlinux
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/boot.c | 41 +++++++++++++++++++++++++++++++++++++++-- hw/riscv/sifive_e.c | 2 +- hw/riscv/sifive_u.c | 6 +++++- hw/riscv/spike.c | 6 +++--- hw/riscv/virt.c | 7 ++++++- include/hw/riscv/boot.h | 4 +++- 6 files changed, 57 insertions(+), 9 deletions(-)hw/riscv/boot.c +9
-1@@ -23,13 +23,50 @@#include "exec/cpu-defs.h" #include "hw/loader.h" #include "hw/riscv/boot.h" +#include "hw/boards.h" #include "elf.h" -target_ulong riscv_load_kernel (const char *kernel_filename) +#if defined(TARGET_RISCV32 ) +# define KERNEL_BOOT_ADDRESS 0x80400000 +#else +# define KERNEL_BOOT_ADDRESS 0x80200000 +#endif + +static uint64_t kernel_translate (void *opaque,uint64_t addr) +{
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
hw/riscv/boot.c line 42
Quoted 6 lines
hw/riscv/boot.c +6
-0+MachineState *machine = opaque; + +/* +* If the user specified a firmware move the kernel to the offset +* start address. +*/ Why?
Removed.
Quoted 10 lines
Quoted 8 lines
Quoted 8 lines
hw/riscv/boot.c line 44
Quoted 3 lines
hw/riscv/boot.c +2
-0+if (machine->firmware) { +return (addr &0x7fffffff ) +KERNEL_BOOT_ADDRESS ;So with both "-bios" and "-kernel", the kernel address will be moved to another address other than
0x80200000(for 64-bit). This does not look good to me.So why not simply return
KERNEL_BOOT_ADDRESSinkernel_translate()?
That's what I am doing now.
Alistair
Quoted 3 lines
Regards,
Bin
RFC 3/5 hw/riscv: Extend the kernel loading support
In reply to Alistair FrancisExtend the RISC-V kernel loader to support uImage and Image files. A Linux kernel can now be booted with:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel Image
--- hw/riscv/boot.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-)
hw/riscv/boot.c +14 -5
@@ -67,13 +67,22 @@ target_ulong riscv_load_kernel(MachineState *machine,uint64_t kernel_entry, kernel_high;if (load_elf (kernel_filename,NULL , kernel_translate, machine, - &kernel_entry,NULL , &kernel_high, -0 ,EM_RISCV ,1 ,0 ) <0 ) { -error_report ("could not load kernel '%s'" , kernel_filename); -exit (1 ); + &kernel_entry,NULL , &kernel_high,0 ,EM_RISCV ,1 ,0 ) >0 ) { +return kernel_entry; + } + +if (load_uimage_as (kernel_filename, &kernel_entry,NULL ,NULL , + kernel_translate, machine,NULL ) >0 ) { +return kernel_entry; + } + +if (load_image_targphys_as (kernel_filename,KERNEL_BOOT_ADDRESS , + ram_size,NULL ) >0 ) { +return kernel_entry; } -return kernel_entry; +error_report ("could not load kernel '%s'" , kernel_filename); +exit (1 ); } hwaddr riscv_load_initrd(const char *filename, uint64_t mem_size,
Signature · 1 line
2.22.0
Re: [Qemu-devel] [RFC v1 3/5] hw/riscv: Extend the kernel loading support
Quoted 19 lines · Alistair Francis
Extend the RISC-V kernel loader to support uImage and Image files. A Linux kernel can now be booted with:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel Image
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/boot.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-)hw/riscv/boot.c +0
-0@@ -67,13 +67,22 @@ target_ulong riscv_load_kernel(MachineState *machine,uint64_t kernel_entry, kernel_high; if (load_elf(kernel_filename, NULL, kernel_translate, machine,
hw/riscv/boot.c target_ulong riscv_load_kernel(MachineState *machine, · line 75
Quoted 10 lines
hw/riscv/boot.c +6
-4- &kernel_entry, NULL, &kernel_high, - 0, EM_RISCV, 1, 0) < 0) { -error_report ("could not load kernel '%s'" , kernel_filename); -exit (1 ); + &kernel_entry, NULL, &kernel_high, 0, EM_RISCV, 1, 0) > 0) { +return kernel_entry; + } + +if (load_uimage_as (kernel_filename, &kernel_entry,NULL ,NULL , + kernel_translate, machine,NULL ) >0 ) {
We should not set the 'kernel_translate' here for uImage.
In fact, the whole kernel_translate() is not necessary.
hw/riscv/boot.c target_ulong riscv_load_kernel(MachineState *machine, · line 73
Quoted 3 lines
hw/riscv/boot.c +3
-0+return kernel_entry; + } +
Regards,
Bin
Re: [Qemu-devel] [RFC v1 3/5] hw/riscv: Extend the kernel loading support
Quoted 35 lines · Bin Meng
Quoted 19 lines · Alistair Francis
Extend the RISC-V kernel loader to support uImage and Image files. A Linux kernel can now be booted with:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel Image
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/boot.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-)hw/riscv/boot.c +0
-0@@ -67,13 +67,22 @@ target_ulong riscv_load_kernel(MachineState *machine,uint64_t kernel_entry, kernel_high; if (load_elf(kernel_filename, NULL, kernel_translate, machine,hw/riscv/boot.c target_ulong riscv_load_kernel(MachineState *machine, · line 75
Quoted 10 lines
hw/riscv/boot.c +6
-4- &kernel_entry, NULL, &kernel_high, - 0, EM_RISCV, 1, 0) < 0) { -error_report ("could not load kernel '%s'" , kernel_filename); -exit (1 ); + &kernel_entry, NULL, &kernel_high, 0, EM_RISCV, 1, 0) > 0) { +return kernel_entry; + } + +if (load_uimage_as (kernel_filename, &kernel_entry,NULL ,NULL , + kernel_translate, machine,NULL ) >0 ) {We should not set the '
kernel_translate' here for uImage.
Quoted 1 line
In fact, the whole
kernel_translate()is not necessary.
I have removed the kernel_translate() function. I tested loading uImage files though and they don't seem to work as the image is loaded at the wrong address.
I have removed uImage loading support from this series. We can look at it in the future if we decide we want it.
Alistair
Quoted 7 lines
hw/riscv/boot.c target_ulong riscv_load_kernel(MachineState *machine, · line 73
Quoted 4 lines
hw/riscv/boot.c +3
-0+return kernel_entry; + } +Regards,
Bin
Re: [Qemu-devel] [RFC v1 3/5] hw/riscv: Extend the kernel loading support
Quoted 37 lines · Alistair Francis
Quoted 35 lines · Bin Meng
Quoted 19 lines · Alistair Francis
Extend the RISC-V kernel loader to support uImage and Image files. A Linux kernel can now be booted with:
qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel Image
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/boot.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-)hw/riscv/boot.c +0
-0@@ -67,13 +67,22 @@ target_ulong riscv_load_kernel(MachineState *machine,uint64_t kernel_entry, kernel_high; if (load_elf(kernel_filename, NULL, kernel_translate, machine,hw/riscv/boot.c target_ulong riscv_load_kernel(MachineState *machine, · line 75
Quoted 10 lines
hw/riscv/boot.c +6
-4- &kernel_entry, NULL, &kernel_high, - 0, EM_RISCV, 1, 0) < 0) { -error_report ("could not load kernel '%s'" , kernel_filename); -exit (1 ); + &kernel_entry, NULL, &kernel_high, 0, EM_RISCV, 1, 0) > 0) { +return kernel_entry; + } + +if (load_uimage_as (kernel_filename, &kernel_entry,NULL ,NULL , + kernel_translate, machine,NULL ) >0 ) {We should not set the '
kernel_translate' here for uImage.
Quoted 8 lines
Quoted 1 line
In fact, the whole
kernel_translate()is not necessary.I have removed the
kernel_translate()function. I tested loading uImage files though and they don't seem to work as the image is loaded at the wrong address.I have removed uImage loading support from this series. We can look at it in the future if we decide we want it.
Ah, my mistake, it looks like my uImage wasn't built correctly. I'll keep this in.
Alistair
Quoted 10 lines
Alistair
Quoted 7 lines
hw/riscv/boot.c target_ulong riscv_load_kernel(MachineState *machine, · line 73
Quoted 4 lines
hw/riscv/boot.c +3
-0+return kernel_entry; + } +Regards,
Bin
RFC 4/5 roms: Add OpenSBI version 0.3
In reply to Alistair FrancisAdd OpenSBI version 0.3 as a git submodule and as a prebult binary.
--- .gitmodules | 3 +++ Makefile | 3 ++- configure | 1 + pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes roms/Makefile | 17 +++++++++++++++++ roms/opensbi | 1 + 7 files changed, 24 insertions(+), 1 deletion(-) create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbi
.gitmodules +3 -0
@@ -55,3 +55,6 @@ [submodule "slirp"] path = slirp url = https://git.qemu.org/git/libslirp.git +[submodule "roms/opensbi"] + path = roms/opensbi + url = https://github.com/riscv/opensbi.git
Makefile +2 -1
@@ -761,7 +761,8 @@ palcode-clipper \ u-boot.e500 u-boot-sam460-20100605.bin \ qemu_vga.ndrv \ edk2-licenses.txt \ -hppa-firmware.img +hppa-firmware.img \ +opensbi-riscv32-fw_jump.elf opensbi-riscv64-fw_jump.elf DESCS=50-edk2-i386-secure.json 50-edk2-x86_64-secure.json \ 60-edk2-aarch64.json 60-edk2-arm.json 60-edk2-i386.json 60-edk2-x86_64.json
configure +1 -0
@@ -7936,6 +7936,7 @@ for bios_file in \ $source_path/pc-bios/*.rom \ $source_path/pc-bios/*.dtb \ $source_path/pc-bios/*.img \ + $source_path/pc-bios/*.elf \ $source_path/pc-bios/openbios-* \ $source_path/pc-bios/u-boot.* \ $source_path/pc-bios/edk2-*.fd.bz2 \
pc-bios/opensbi-riscv32-fw_jump.elf new file binary +0 -0
pc-bios/opensbi-riscv64-fw_jump.elf new file binary +0 -0
roms/Makefile +17 -0
@@ -37,6 +37,8 @@ find-cross-prefix = $(subst gcc,,$(notdir $(call find-cross-gcc,$(1)))) powerpc64_cross_prefix := $(call find-cross-prefix,powerpc64) powerpc_cross_prefix := $(call find-cross-prefix,powerpc) x86_64_cross_prefix := $(call find-cross-prefix,x86_64) +riscv32_cross_prefix := $(call find-cross-prefix,riscv32) +riscv64_cross_prefix := $(call find-cross-prefix,riscv64) # tag our seabios builds SEABIOS_EXTRAVERSION="-prebuilt.qemu.org"
@@ -62,6 +64,8 @@ default: @echo " u-boot.e500 -- update u-boot.e500" @echo " u-boot.sam460 -- update u-boot.sam460" @echo " efi -- update UEFI (edk2) platform firmware" + @echo " opensbi32 -- update OpenSBI for 32-bit" + @echo " opensbi64 -- update OpenSBI for 64-bit" @echo " clean -- delete the files generated by the previous" \ "build targets"
@@ -162,6 +166,18 @@ skiboot: efi: edk2-basetools $(MAKE) -f Makefile.edk2 +opensbi32: + $(MAKE) -C opensbi \ + CROSS_COMPILE=$(riscv32_cross_prefix) \ + PLATFORM="qemu/virt" + cp opensbi/build/platform/qemu/virt/firmware/fw_jump.elf ../pc-bios/opensbi-riscv32-fw_jump.elf + +opensbi64: + $(MAKE) -C opensbi \ + CROSS_COMPILE=$(riscv64_cross_prefix) \ + PLATFORM="qemu/virt" + cp opensbi/build/platform/qemu/virt/firmware/fw_jump.elf ../pc-bios/opensbi-riscv64-fw_jump.elf + clean: rm -rf seabios/.config seabios/out seabios/builds $(MAKE) -C sgabios clean
@@ -173,3 +189,4 @@ clean: $(MAKE) -C u-boot-sam460ex distclean $(MAKE) -C skiboot clean $(MAKE) -f Makefile.edk2 clean + $(MAKE) -C opensbi clean
roms/opensbi new file +1 -0
@@ -0,0 +1 @@ +Subproject commit ca20ac0cd4c099006d4eea4d9ac7bd7b58e2ae0f
Signature · 1 line
2.22.0
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 4/5] roms: Add OpenSBI version 0.3
Quoted 4 lines · Alistair Francis
Add OpenSBI version 0.3 as a git submodule and as a prebult binary.
Signed-off-by Alistair Francis <[email protected]>
Quoted 13 lines
--- .gitmodules | 3 +++ Makefile | 3 ++- configure | 1 + pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes roms/Makefile | 17 +++++++++++++++++ roms/opensbi | 1 + 7 files changed, 24 insertions(+), 1 deletion(-) create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbi
The OpenSBI firmwares are platform specific so we should have machine directory under pc-bios/ directory
So for virt machine we will have:
pc-bios/riscv32/virt/fw_jump.elf
pc-bios/riscv64/virt/fw_jump.elf
Regards,
Anup
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 4/5] roms: Add OpenSBI version 0.3
Quoted 21 lines · Anup Patel
Quoted 4 lines · Alistair Francis
Add OpenSBI version 0.3 as a git submodule and as a prebult binary.
Signed-off-by Alistair Francis <[email protected]>Quoted 13 lines
--- .gitmodules | 3 +++ Makefile | 3 ++- configure | 1 + pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes roms/Makefile | 17 +++++++++++++++++ roms/opensbi | 1 + 7 files changed, 24 insertions(+), 1 deletion(-) create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbi
Quoted 6 lines
The OpenSBI firmwares are platform specific so we should have machine directory under pc-bios/ directory
So for virt machine we will have:
pc-bios/riscv32/virt/fw_jump.elf
pc-bios/riscv64/virt/fw_jump.elf
And we should only integrate plain binary image for "bios" images here.
pc-bios/riscv32/virt/fw_jump.bin
pc-bios/riscv64/virt/fw_jump.bin
Regards,
Bin
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 4/5] roms: Add OpenSBI version 0.3
Quoted 23 lines · Bin Meng
Quoted 21 lines · Anup Patel
Quoted 17 lines · Alistair Francis
Add OpenSBI version 0.3 as a git submodule and as a prebult binary.
Signed-off-by Alistair Francis <[email protected]>--- .gitmodules | 3 +++ Makefile | 3 ++- configure | 1 + pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes roms/Makefile | 17 +++++++++++++++++ roms/opensbi | 1 + 7 files changed, 24 insertions(+), 1 deletion(-) create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbi
Quoted 6 lines
Quoted 6 lines
The OpenSBI firmwares are platform specific so we should have machine directory under pc-bios/ directory
So for virt machine we will have:
pc-bios/riscv32/virt/fw_jump.elf
pc-bios/riscv64/virt/fw_jump.elf
I have updated the names to indicate the machine. The pc-bios directly appears to be flat (at least for binaries) so I don't want to add subdirectories.
Quoted 5 lines
And we should only integrate plain binary image for "bios" images here.
pc-bios/riscv32/virt/fw_jump.bin
pc-bios/riscv64/virt/fw_jump.bin
Yep, fixed.
Alistair
Quoted 3 lines
Regards,
Bin
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 4/5] roms: Add OpenSBI version 0.3
Quoted 28 lines · Alistair Francis
Quoted 26 lines · Bin Meng
Quoted 24 lines · Anup Patel
Quoted 4 lines · Alistair Francis
Add OpenSBI version 0.3 as a git submodule and as a prebult binary.
Signed-off-by Alistair Francis <[email protected]>Quoted 13 lines
--- .gitmodules | 3 +++ Makefile | 3 ++- configure | 1 + pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes roms/Makefile | 17 +++++++++++++++++ roms/opensbi | 1 + 7 files changed, 24 insertions(+), 1 deletion(-) create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbiThe OpenSBI firmwares are platform specific so we should have machine directory under pc-bios/ directory
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
So for virt machine we will have:
pc-bios/riscv32/virt/fw_jump.elf
pc-bios/riscv64/virt/fw_jump.elfI have updated the names to indicate the machine. The pc-bios directly appears to be flat (at least for binaries) so I don't want to add subdirectories.
Should we include pre-built OpenSBI "bios" for "sifive_u" machine too?
Quoted 8 lines
Quoted 5 lines
And we should only integrate plain binary image for "bios" images here.
pc-bios/riscv32/virt/fw_jump.bin
pc-bios/riscv64/virt/fw_jump.binYep, fixed.
Regards,
Bin
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 4/5] roms: Add OpenSBI version 0.3
Quoted 34 lines · Bin Meng
Quoted 32 lines · Alistair Francis
Quoted 29 lines · Bin Meng
Quoted 27 lines · Anup Patel
Quoted 4 lines · Alistair Francis
Add OpenSBI version 0.3 as a git submodule and as a prebult binary.
Signed-off-by Alistair Francis <[email protected]>Quoted 13 lines
--- .gitmodules | 3 +++ Makefile | 3 ++- configure | 1 + pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes roms/Makefile | 17 +++++++++++++++++ roms/opensbi | 1 + 7 files changed, 24 insertions(+), 1 deletion(-) create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbiThe OpenSBI firmwares are platform specific so we should have machine directory under pc-bios/ directory
So for virt machine we will have:
pc-bios/riscv32/virt/fw_jump.elf
pc-bios/riscv64/virt/fw_jump.elf
Quoted 6 lines
Quoted 4 lines
I have updated the names to indicate the machine. The pc-bios directly appears to be flat (at least for binaries) so I don't want to add subdirectories.
Should we include pre-built OpenSBI "bios" for "
sifive_u" machine too?
Yep, I am doing that now.
Alistair
Quoted 12 lines
Quoted 9 lines
Quoted 6 lines
And we should only integrate plain binary image for "bios" images here.
pc-bios/riscv32/virt/fw_jump.bin
pc-bios/riscv64/virt/fw_jump.binYep, fixed.
Regards,
Bin
RFC 5/5 hw/riscv: Load OpenSBI as the default firmware
In reply to Alistair FrancisIf the user hasn't specified a firmware to load (with -bios) or specified no bios (with -bios none) then load OpenSBI by default. This allows users to boot a RISC-V kernel with just -kernel.
--- hw/riscv/boot.c | 28 ++++++++++++++++++++++++++++ hw/riscv/sifive_u.c | 4 +--- hw/riscv/virt.c | 4 +--- include/hw/riscv/boot.h | 1 + 4 files changed, 31 insertions(+), 6 deletions(-)
hw/riscv/boot.c +28 -0
@@ -18,6 +18,7 @@*/ #include "qemu/osdep.h" +#include "qemu-common.h" #include "qemu/units.h" #include "qemu/error-report.h" #include "exec/cpu-defs.h"
@@ -32,6 +33,12 @@# define KERNEL_BOOT_ADDRESS 0x80200000#endif +#if defined(TARGET_RISCV32 ) +# define BIOS_FILENAME "opensbi-riscv32-fw_jump.elf" +#else +# define BIOS_FILENAME "opensbi-riscv64-fw_jump.elf" +#endif +static uint64_t kernel_translate (void *opaque,uint64_t addr) {MachineState *machine = opaque;
@@ -47,6 +54,27 @@ static uint64_t kernel_translate(void *opaque, uint64_t addr) } } +void riscv_find_and_load_firmware (MachineState *machine) +{ +char *firmware_filename; + +if (!machine->firmware) { +/* The user didn't specify a firmware, default to OpenSBI */ + firmware_filename =qemu_find_file (QEMU_FILE_TYPE_BIOS ,BIOS_FILENAME ); + }else { + firmware_filename = machine->firmware; + } + +if (strcmp (firmware_filename,"none" )) { +/* If not "none" load the firmware */ +riscv_load_firmware (firmware_filename); + } + +if (!machine->firmware) { +g_free (firmware_filename); + } +} +target_ulong riscv_load_firmware (const char *firmware_filename) {uint64_t firmware_entry, firmware_start, firmware_end;
hw/riscv/sifive_u.c +1 -3
@@ -266,9 +266,7 @@ static void riscv_sifive_u_init(MachineState *machine)/* create device tree */ create_fdt (s, memmap, machine->ram_size, machine->kernel_cmdline); -if (machine->firmware) { -riscv_load_firmware (machine->firmware); - } +riscv_find_and_load_firmware (machine);if (machine->kernel_filename) {riscv_load_kernel (machine, machine->kernel_filename);
hw/riscv/virt.c +1 -3
@@ -380,9 +380,7 @@ static void riscv_virt_board_init(MachineState *machine)memory_region_add_subregion (system_memory, memmap[VIRT_MROM ].base, mask_rom); -if (machine->firmware) { -riscv_load_firmware (machine->firmware); - } +riscv_find_and_load_firmware (machine);if (machine->kernel_filename) { uint64_t kernel_entry = riscv_load_kernel(machine,
include/hw/riscv/boot.h +1 -0
@@ -20,6 +20,7 @@#ifndef RISCV_BOOT_H #define RISCV_BOOT_H +void riscv_find_and_load_firmware (MachineState *machine);target_ulong riscv_load_firmware (const char *firmware_filename);target_ulong riscv_load_kernel (MachineState *machine,const char *kernel_filename);
Signature · 1 line
2.22.0
Re: [Qemu-devel] [Qemu-riscv] [RFC v1 5/5] hw/riscv: Load OpenSBI as the default firmware
Quoted 26 lines · Alistair Francis
If the user hasn't specified a firmware to load (with -bios) or specified no bios (with -bios none) then load OpenSBI by default. This allows users to boot a RISC-V kernel with just -kernel.
Signed-off-by Alistair Francis <[email protected]>--- hw/riscv/boot.c | 28 ++++++++++++++++++++++++++++ hw/riscv/sifive_u.c | 4 +--- hw/riscv/virt.c | 4 +--- include/hw/riscv/boot.h | 1 + 4 files changed, 31 insertions(+), 6 deletions(-)hw/riscv/boot.c +1
-0@@ -18,6 +18,7 @@*/ #include "qemu/osdep.h" +#include "qemu-common.h" #include "qemu/units.h" #include "qemu/error-report.h" #include "exec/cpu-defs.h"
hw/riscv/boot.c line 40
Quoted 8 lines
hw/riscv/boot.c +5
-0@@ -32,6 +33,12 @@# define KERNEL_BOOT_ADDRESS 0x80200000#endif +#if defined(TARGET_RISCV32 ) +# define BIOS_FILENAME "opensbi-riscv32-fw_jump.elf" +#else +# define BIOS_FILENAME "opensbi-riscv64-fw_jump.elf" +#endif
Based on my comment on PATCH4, BIOS_FILENAME should be derived from QEMU machine name and TARGET_RISCVx
Agree ??
Regards,
Anup
Quoted 78 lines
hw/riscv/boot.c +22
-0+static uint64_t kernel_translate (void *opaque,uint64_t addr) {MachineState *machine = opaque;@@ -47,6 +54,27 @@ static uint64_t kernel_translate(void *opaque, uint64_t addr) } } +void riscv_find_and_load_firmware (MachineState *machine) +{ +char *firmware_filename; + +if (!machine->firmware) { +/* The user didn't specify a firmware, default to OpenSBI */ + firmware_filename =qemu_find_file (QEMU_FILE_TYPE_BIOS ,BIOS_FILENAME ); + }else { + firmware_filename = machine->firmware; + } + +if (strcmp (firmware_filename,"none" )) { +/* If not "none" load the firmware */ +riscv_load_firmware (firmware_filename); + } + +if (!machine->firmware) { +g_free (firmware_filename); + } +} +target_ulong riscv_load_firmware (const char *firmware_filename) {uint64_t firmware_entry, firmware_start, firmware_end;hw/riscv/sifive_u.c +1
-3@@ -266,9 +266,7 @@ static void riscv_sifive_u_init(MachineState *machine)/* create device tree */ create_fdt (s, memmap, machine->ram_size, machine->kernel_cmdline); -if (machine->firmware) { -riscv_load_firmware (machine->firmware); - } +riscv_find_and_load_firmware (machine);if (machine->kernel_filename) {riscv_load_kernel (machine, machine->kernel_filename);hw/riscv/virt.c +1
-3@@ -380,9 +380,7 @@ static void riscv_virt_board_init(MachineState *machine)memory_region_add_subregion (system_memory, memmap[VIRT_MROM ].base, mask_rom); -if (machine->firmware) { -riscv_load_firmware (machine->firmware); - } +riscv_find_and_load_firmware (machine);if (machine->kernel_filename) { uint64_t kernel_entry = riscv_load_kernel(machine,include/hw/riscv/boot.h +1
-1@@ -20,6 +20,7 @@#ifndef RISCV_BOOT_H #define RISCV_BOOT_H +void riscv_find_and_load_firmware (MachineState *machine);target_ulong riscv_load_firmware (const char *firmware_filename);target_ulong riscv_load_kernel (MachineState *machine,const char *kernel_filename); --2.22.0
Quoted 11 lines · Alistair Francis
This is an RFC as it will break ALL current users! See below for details.
This series consolidates the current RISC-V kernel loading impelementation while also adding support for the -bios option and more advanced kernel image types.
After consolidating the kernel loading we can extend the boot loader to support a -bios option. We can also extend the kernel loading options to support not just ELF files but other standard formats.
Quoted 5 lines
Finally we can include the OpenSBI firmware by default for QEMU users.
At the end of this series we are in the good place of no longer requiring users to build firmware to boot a kernel. Instead users can just run QEMu with the -kernel option and everything will work. They can
This is great. I like booting kernel directly for testing without bothering the firmware stuff.
Quoted 35 lines
also override the firmware with their own using the -bios option. Using "-bios none" will result in no firmware being loaded (as it is today).
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Unfortunately this series (patch 5 specifically) results in all current Linux boots being broken as users are already loading in their own firmware, which overlaps with the now included default. They will see this error:
rom: requested regions overlap (rom phdr #0: ./images/qemuriscv64/fw_jump.elf. free=0x0000000080008090, addr=0x0000000080000000)
qemu-system-riscv64: rom check and register reset failedIf a current user specieifies "-bios none" their flow will continue working. A user can instead no longer load their firmware and use the default or they can load their firmware with the -bios option.
At the moment the best idea I have to not break all users is to only include the default firmware if the user specifies "-bios opensbi". That is change the default to not loading the firmware. Then we can work on updating documentation and maybe in future change the default to include a firmware and anyone who doesn't want a default firmware can specify "-bios none".
Any other ideas on how to not break everything?
Alistair Francis (5):
hw/riscv: Split out the boot functions
hw/riscv: Add support for loading a firmware
hw/riscv: Extend the kernel loading support
roms: Add OpenSBI version 0.3
hw/riscv: Load OpenSBI as the default firmware
Quoted 12 lines
.gitmodules | 3 + Makefile | 3 +- configure | 1 + hw/riscv/Makefile.objs | 1 + hw/riscv/boot.c | 143 ++++++++++++++++++++++++++++ hw/riscv/sifive_e.c | 17 +--- hw/riscv/sifive_u.c | 19 +--- hw/riscv/spike.c | 21 +--- hw/riscv/virt.c | 54 ++--------- include/hw/riscv/boot.h | 30 ++++++ pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes
Since we are considering adding "bios" images, I prefer to add the pure binary images instead of ELF images here.
Quoted 8 lines
roms/Makefile | 17 ++++ roms/opensbi | 1 + 14 files changed, 216 insertions(+), 94 deletions(-) create mode 100644 hw/riscv/boot.c create mode 100644 include/hw/riscv/boot.h create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbi
Regards,
Bin
Quoted 14 lines · Bin Meng
Quoted 11 lines · Alistair Francis
This is an RFC as it will break ALL current users! See below for details.
This series consolidates the current RISC-V kernel loading impelementation while also adding support for the -bios option and more advanced kernel image types.
After consolidating the kernel loading we can extend the boot loader to support a -bios option. We can also extend the kernel loading options to support not just ELF files but other standard formats.
Quoted 8 lines
Quoted 5 lines
Finally we can include the OpenSBI firmware by default for QEMU users.
At the end of this series we are in the good place of no longer requiring users to build firmware to boot a kernel. Instead users can just run QEMu with the -kernel option and everything will work. They can
This is great. I like booting kernel directly for testing without bothering the firmware stuff.
That's the goal! Now we just need to not break everything.
Quoted 49 lines
Quoted 36 lines
also override the firmware with their own using the -bios option. Using "-bios none" will result in no firmware being loaded (as it is today).
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Unfortunately this series (patch 5 specifically) results in all current Linux boots being broken as users are already loading in their own firmware, which overlaps with the now included default. They will see this error:
rom: requested regions overlap (rom phdr #0: ./images/qemuriscv64/fw_jump.elf. free=0x0000000080008090, addr=0x0000000080000000)
qemu-system-riscv64: rom check and register reset failedIf a current user specieifies "-bios none" their flow will continue working. A user can instead no longer load their firmware and use the default or they can load their firmware with the -bios option.
At the moment the best idea I have to not break all users is to only include the default firmware if the user specifies "-bios opensbi". That is change the default to not loading the firmware. Then we can work on updating documentation and maybe in future change the default to include a firmware and anyone who doesn't want a default firmware can specify "-bios none".
Any other ideas on how to not break everything?
Alistair Francis (5):
hw/riscv: Split out the boot functions
hw/riscv: Add support for loading a firmware
hw/riscv: Extend the kernel loading support
roms: Add OpenSBI version 0.3
hw/riscv: Load OpenSBI as the default firmwareQuoted 12 lines
.gitmodules | 3 + Makefile | 3 +- configure | 1 + hw/riscv/Makefile.objs | 1 + hw/riscv/boot.c | 143 ++++++++++++++++++++++++++++ hw/riscv/sifive_e.c | 17 +--- hw/riscv/sifive_u.c | 19 +--- hw/riscv/spike.c | 21 +--- hw/riscv/virt.c | 54 ++--------- include/hw/riscv/boot.h | 30 ++++++ pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes
Quoted 2 lines
Since we are considering adding "bios" images, I prefer to add the pure binary images instead of ELF images here.
I didn't think about that. Can we just boot them in QEMU like we do with the ELFs?
Alistair
Quoted 12 lines
Quoted 9 lines
roms/Makefile | 17 ++++ roms/opensbi | 1 + 14 files changed, 216 insertions(+), 94 deletions(-) create mode 100644 hw/riscv/boot.c create mode 100644 include/hw/riscv/boot.h create mode 100644 pc-bios/opensbi-riscv32-fw_jump.elf create mode 100644 pc-bios/opensbi-riscv64-fw_jump.elf create mode 160000 roms/opensbiRegards,
Bin
Quoted 76 lines · Alistair Francis
Quoted 14 lines · Bin Meng
Quoted 11 lines · Alistair Francis
This is an RFC as it will break ALL current users! See below for details.
This series consolidates the current RISC-V kernel loading impelementation while also adding support for the -bios option and more advanced kernel image types.
After consolidating the kernel loading we can extend the boot loader to support a -bios option. We can also extend the kernel loading options to support not just ELF files but other standard formats.
Quoted 8 lines
Quoted 5 lines
Finally we can include the OpenSBI firmware by default for QEMU users.
At the end of this series we are in the good place of no longer requiring users to build firmware to boot a kernel. Instead users can just run QEMu with the -kernel option and everything will work. They can
This is great. I like booting kernel directly for testing without bothering the firmware stuff.
That's the goal! Now we just need to not break everything.
Quoted 49 lines
Quoted 48 lines
also override the firmware with their own using the -bios option. Using "-bios none" will result in no firmware being loaded (as it is today).
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Unfortunately this series (patch 5 specifically) results in all current Linux boots being broken as users are already loading in their own firmware, which overlaps with the now included default. They will see this error:
rom: requested regions overlap (rom phdr #0: ./images/qemuriscv64/fw_jump.elf. free=0x0000000080008090, addr=0x0000000080000000)
qemu-system-riscv64: rom check and register reset failedIf a current user specieifies "-bios none" their flow will continue working. A user can instead no longer load their firmware and use the default or they can load their firmware with the -bios option.
At the moment the best idea I have to not break all users is to only include the default firmware if the user specifies "-bios opensbi". That is change the default to not loading the firmware. Then we can work on updating documentation and maybe in future change the default to include a firmware and anyone who doesn't want a default firmware can specify "-bios none".
Any other ideas on how to not break everything?
Alistair Francis (5):
hw/riscv: Split out the boot functions
hw/riscv: Add support for loading a firmware
hw/riscv: Extend the kernel loading support
roms: Add OpenSBI version 0.3
hw/riscv: Load OpenSBI as the default firmware.gitmodules | 3 + Makefile | 3 +- configure | 1 + hw/riscv/Makefile.objs | 1 + hw/riscv/boot.c | 143 ++++++++++++++++++++++++++++ hw/riscv/sifive_e.c | 17 +--- hw/riscv/sifive_u.c | 19 +--- hw/riscv/spike.c | 21 +--- hw/riscv/virt.c | 54 ++--------- include/hw/riscv/boot.h | 30 ++++++ pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes
Quoted 5 lines
Quoted 2 lines
Since we are considering adding "bios" images, I prefer to add the pure binary images instead of ELF images here.
I didn't think about that. Can we just boot them in QEMU like we do with the ELFs?
Yes, use load_image_targphys() instead of load_elf().
Regards,
Bin
Quoted 78 lines · Bin Meng
Quoted 76 lines · Alistair Francis
Quoted 14 lines · Bin Meng
Quoted 11 lines · Alistair Francis
This is an RFC as it will break ALL current users! See below for details.
This series consolidates the current RISC-V kernel loading impelementation while also adding support for the -bios option and more advanced kernel image types.
After consolidating the kernel loading we can extend the boot loader to support a -bios option. We can also extend the kernel loading options to support not just ELF files but other standard formats.
Quoted 8 lines
Quoted 5 lines
Finally we can include the OpenSBI firmware by default for QEMU users.
At the end of this series we are in the good place of no longer requiring users to build firmware to boot a kernel. Instead users can just run QEMu with the -kernel option and everything will work. They can
This is great. I like booting kernel directly for testing without bothering the firmware stuff.
That's the goal! Now we just need to not break everything.
Quoted 49 lines
Quoted 48 lines
also override the firmware with their own using the -bios option. Using "-bios none" will result in no firmware being loaded (as it is today).
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Unfortunately this series (patch 5 specifically) results in all current Linux boots being broken as users are already loading in their own firmware, which overlaps with the now included default. They will see this error:
rom: requested regions overlap (rom phdr #0: ./images/qemuriscv64/fw_jump.elf. free=0x0000000080008090, addr=0x0000000080000000)
qemu-system-riscv64: rom check and register reset failedIf a current user specieifies "-bios none" their flow will continue working. A user can instead no longer load their firmware and use the default or they can load their firmware with the -bios option.
At the moment the best idea I have to not break all users is to only include the default firmware if the user specifies "-bios opensbi". That is change the default to not loading the firmware. Then we can work on updating documentation and maybe in future change the default to include a firmware and anyone who doesn't want a default firmware can specify "-bios none".
Any other ideas on how to not break everything?
Alistair Francis (5):
hw/riscv: Split out the boot functions
hw/riscv: Add support for loading a firmware
hw/riscv: Extend the kernel loading support
roms: Add OpenSBI version 0.3
hw/riscv: Load OpenSBI as the default firmware.gitmodules | 3 + Makefile | 3 +- configure | 1 + hw/riscv/Makefile.objs | 1 + hw/riscv/boot.c | 143 ++++++++++++++++++++++++++++ hw/riscv/sifive_e.c | 17 +--- hw/riscv/sifive_u.c | 19 +--- hw/riscv/spike.c | 21 +--- hw/riscv/virt.c | 54 ++--------- include/hw/riscv/boot.h | 30 ++++++ pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytes
Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
Since we are considering adding "bios" images, I prefer to add the pure binary images instead of ELF images here.
I didn't think about that. Can we just boot them in QEMU like we do with the ELFs?
Yes, use
load_image_targphys()instead ofload_elf().
Ah, that is obvious. I'll update it to use the bin files then.
Alistair
Quoted 3 lines
Regards,
Bin
Quoted 9 lines · Alistair Francis
Quoted 8 lines · Bin Meng
Quoted 7 lines · Alistair Francis
Quoted 5 lines · Bin Meng
Quoted 2 lines
pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytesSince we are considering adding "bios" images, I prefer to add the pure binary images instead of ELF images here.
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
I didn't think about that. Can we just boot them in QEMU like we do with the ELFs?
Yes, use
load_image_targphys()instead ofload_elf().Ah, that is obvious. I'll update it to use the bin files then.
I'm unclear on the advantages of using one format over the other, but one question comes to mind: once this is in, we will probably want to have OpenSBI packaged separately in distributions, the same way it already happens for SeaBIOS, SLOF and edk2-based firmwares.
Will using either of the formats prevent that from happening?
Signature · 1 line
Andrea Bolognani / Red Hat / Virtualization
Quoted 11 lines · Andrea Bolognani
Quoted 9 lines · Alistair Francis
Quoted 8 lines · Bin Meng
Quoted 7 lines · Alistair Francis
Quoted 5 lines · Bin Meng
Quoted 2 lines
pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytesSince we are considering adding "bios" images, I prefer to add the pure binary images instead of ELF images here.
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
I didn't think about that. Can we just boot them in QEMU like we do with the ELFs?
Yes, use
load_image_targphys()instead ofload_elf().Ah, that is obvious. I'll update it to use the bin files then.
I'm unclear on the advantages of using one format over the other,
The main one that I see is that everyone else is already using .bin and no one else is using .elf.
Quoted 5 lines
but one question comes to mind: once this is in, we will probably want to have OpenSBI packaged separately in distributions, the same way it already happens for SeaBIOS, SLOF and edk2-based firmwares.
Will using either of the formats prevent that from happening?
Both options allow this.
OE-Core already packages OpenSBI by default, Fedora and Debian are moving to OpenSBI for RISC-V targets as well.
Any distro that supports the RISC-V toolchain (which is all upstreamed) can build OpenSBI.
Alistair
Quoted 4 lines
--
Andrea Bolognani / Red Hat / Virtualization
Quoted 29 lines · Alistair Francis
Quoted 11 lines · Andrea Bolognani
Quoted 9 lines · Alistair Francis
Quoted 8 lines · Bin Meng
Quoted 7 lines · Alistair Francis
Quoted 5 lines · Bin Meng
Quoted 2 lines
pc-bios/opensbi-riscv32-fw_jump.elf | Bin 0 -> 197988 bytes pc-bios/opensbi-riscv64-fw_jump.elf | Bin 0 -> 200192 bytesSince we are considering adding "bios" images, I prefer to add the pure binary images instead of ELF images here.
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
I didn't think about that. Can we just boot them in QEMU like we do with the ELFs?
Yes, use
load_image_targphys()instead ofload_elf().Ah, that is obvious. I'll update it to use the bin files then.
I'm unclear on the advantages of using one format over the other,
The main one that I see is that everyone else is already using .bin and no one else is using .elf.
Quoted 4 lines
but one question comes to mind: once this is in, we will probably want to have OpenSBI packaged separately in distributions, the same way it already happens for SeaBIOS, SLOF and edk2-based firmwares.
Quoted 6 lines
Quoted 1 line
Will using either of the formats prevent that from happening?
Both options allow this.
OE-Core already packages OpenSBI by default, Fedora and Debian are moving to OpenSBI for RISC-V targets as well.
Fedora uses OpenSBI for the last 2 or 3 months now. I don't plan to update BBL builds. OpenSBI packages in Fedora/RISCV isn't finalized, but it does ship *.elf and *.bin files.
Quoted 11 lines
Any distro that supports the RISC-V toolchain (which is all upstreamed) can build OpenSBI.
Alistair
Quoted 4 lines
--
Andrea Bolognani / Red Hat / Virtualization
Quoted 13 lines · David Abdurachmanov
Quoted 12 lines · Alistair Francis
Quoted 5 lines · Andrea Bolognani
but one question comes to mind: once this is in, we will probably want to have OpenSBI packaged separately in distributions, the same way it already happens for SeaBIOS, SLOF and edk2-based firmwares.
Will using either of the formats prevent that from happening?
Both options allow this.
OE-Core already packages OpenSBI by default, Fedora and Debian are moving to OpenSBI for RISC-V targets as well.
Quoted 6 lines
Quoted 2 lines
Any distro that supports the RISC-V toolchain (which is all upstreamed) can build OpenSBI.
Fedora uses OpenSBI for the last 2 or 3 months now. I don't plan to update BBL builds. OpenSBI packages in Fedora/RISCV isn't finalized, but it does ship *.elf and *.bin files.
Sounds good to me, thanks for confirming!
Signature · 1 line
Andrea Bolognani / Red Hat / Virtualization
Quoted 5 lines · Andrea Bolognani
Quoted 4 lines · David Abdurachmanov
Quoted 3 lines · Alistair Francis
OE-Core already packages OpenSBI by default, Fedora and Debian are moving to OpenSBI for RISC-V targets as well.
Quoted 8 lines
Quoted 6 lines
Quoted 2 lines
Any distro that supports the RISC-V toolchain (which is all upstreamed) can build OpenSBI.
Fedora uses OpenSBI for the last 2 or 3 months now. I don't plan to update BBL builds. OpenSBI packages in Fedora/RISCV isn't finalized, but it does ship *.elf and *.bin files.
Sounds good to me, thanks for confirming!
From further off-list discussion with David, I have learned that recent Fedora images include an OpenSBI build with embedded U-Boot payload, such that you only need to have that single file on the host and pass it to QEMU via -kernel[1] for RISC-V guest boot to work. I played with it over the past few days, and it works very nicely.
I think this is the result that we want to ultimately reach: a single RISC-V "firmware" binary installed on the host through an appropriate distro package, shared among guests, with everything else that is guest-specific being contained in the corresponding disk image.
This is what other architectures are already doing, with SeaBIOS and OVMF on x86_64, AAVMF on aarch64 and SLOF on ppc64 all being handled this way: RISC-V should, where it makes sense, follow suit.
QEMU also recently introduced a JSON-based specification that can be used to advertise guest firmwares and libvirt already supports it, which makes firmware configuration either extremely convenient or entirely automatic for the user: the OpenSBI support should also be advertised this way.
[1] I guess that'd be -bios after these patches?
Signature · 1 line
Andrea Bolognani / Red Hat / Virtualization