[v2,2/3] Documentation: riscv: Add early boot document

Commit Message

Alexandre Ghiti June 21, 2023, 7:22 a.m. UTC

This document describes the constraints and requirements of the early
boot process in a RISC-V kernel.

Signed-off-by: Alexandre Ghiti <[e-mail]>
Reviewed-by: Björn Töpel <[e-mail]>
---
 Documentation/riscv/boot-image-header.rst |   3 -
 Documentation/riscv/boot.rst              | 170 ++++++++++++++++++++++
 Documentation/riscv/index.rst             |   1 +
 3 files changed, 171 insertions(+), 3 deletions(-)
 create mode 100644 Documentation/riscv/boot.rst

Comments

Song Shuai June 21, 2023, 8:15 a.m. UTC | #1

在 2023/6/21 15:22, Alexandre Ghiti 写道:
> This document describes the constraints and requirements of the early
> boot process in a RISC-V kernel.
> 
> Signed-off-by: Alexandre Ghiti <[e-mail]>
> Reviewed-by: Björn Töpel <[e-mail]>
> ---
>   Documentation/riscv/boot-image-header.rst |   3 -
>   Documentation/riscv/boot.rst              | 170 ++++++++++++++++++++++
>   Documentation/riscv/index.rst             |   1 +
>   3 files changed, 171 insertions(+), 3 deletions(-)
>   create mode 100644 Documentation/riscv/boot.rst
> 
> diff --git a/Documentation/riscv/boot-image-header.rst b/Documentation/riscv/boot-image-header.rst
> index d7752533865f..a4a45310c4c4 100644
> --- a/Documentation/riscv/boot-image-header.rst
> +++ b/Documentation/riscv/boot-image-header.rst
> @@ -7,9 +7,6 @@ Boot image header in RISC-V Linux
>   
>   This document only describes the boot image header details for RISC-V Linux.
>   
> -TODO:
> -  Write a complete booting guide.
> -
>   The following 64-byte header is present in decompressed Linux kernel image::
>   
>   	u32 code0;		  /* Executable code */
> diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
> new file mode 100644
> index 000000000000..019ee818686d
> --- /dev/null
> +++ b/Documentation/riscv/boot.rst
> @@ -0,0 +1,170 @@
> +.. SPDX-License-Identifier: GPL-2.0
> +
> +===============================================
> +RISC-V Kernel Boot Requirements and Constraints
> +===============================================
> +
> +:Author: Alexandre Ghiti <[e-mail]>
> +:Date: 23 May 2023
> +
> +This document describes what the RISC-V kernel expects from bootloaders and
> +firmware, but also the constraints that any developer must have in mind when
> +touching the early boot process. For the purposes of this document, the
> +'early boot process' refers to any code that runs before the final virtual
> +mapping is set up.
> +
> +Pre-kernel Requirements and Constraints
> +=======================================
> +
> +The RISC-V kernel expects the following of bootloaders and platform firmware:
> +
> +Register state
> +--------------
> +
> +The RISC-V kernel expects:
> +
> +  * `$a0` to contain the hartid of the current core.
> +  * `$a1` to contain the address of the devicetree in memory.
> +
> +CSR state
> +---------
> +
> +The RISC-V kernel expects:
> +
> +  * `$satp = 0`: the MMU, if present, must be disabled.
> +
> +Reserved memory for resident firmware
> +-------------------------------------
> +
> +The RISC-V kernel must not map any resident memory, or memory protected with
> +PMPs, in the direct mapping, so the firmware must correctly mark those regions
> +as per the devicetree specification and/or the UEFI specification.
> +
> +Kernel location
> +---------------
> +
> +The RISC-V kernel expects to be placed at a PMD boundary (2MB aligned for rv64
> +and 4MB aligned for rv32). Note that the EFI stub will physically relocate the
> +kernel if that's not the case.
> +
> +Hardware description
> +--------------------
> +
> +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> +
> +The devicetree is either passed directly to the kernel from the previous stage
> +using the `$a1` register, or when booting with UEFI, it can be passed using the
> +EFI configuration table.
> +
> +The ACPI tables are passed to the kernel using the EFI configuration table. In
> +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> +"EFI stub and devicetree" tree section below for details about this devicetree.
> +
> +Kernel entrance
> +---------------
> +
> +On SMP systems, there are 2 methods to enter the kernel:
> +
> +- `RISCV_BOOT_SPINWAIT`: the firmware releases all harts in the kernel, one hart
> +  wins a lottery and executes the early boot code while the other harts are
> +  parked waiting for the initialization to finish. This method is mostly used to
> +  support older firmwares without SBI HSM extension and M-mode RISC-V kernel.
> +- `Ordered booting`: the firmware releases only one hart that will execute the
> +  initialization phase and then will start all other harts using the SBI HSM
> +  extension. The ordered booting method is the preferred booting method for
> +  booting the RISC-V kernel because it can support cpu hotplug and kexec.
> +
> +UEFI
> +----
> +
> +UEFI memory map
> +~~~~~~~~~~~~~~~
> +
> +When booting with UEFI, the RISC-V kernel will use only the EFI memory map to
> +populate the system memory.
> +
> +The UEFI firmware must parse the subnodes of the `/reserved-memory` devicetree
> +node and abide by the devicetree specification to convert the attributes of
> +those subnodes (`no-map` and `reusable`) into their correct EFI equivalent
> +(refer to section "3.5.4 /reserved-memory and UEFI" of the devicetree
> +specification v0.4-rc1).
append this note ?

Note that RISC-V edk2 diverges from the devicetree specification to 
declare the !no-map regions as EfiReservedMemoryType instead of 
EfiBootServicesData.
> +
> +RISCV_EFI_BOOT_PROTOCOL
> +~~~~~~~~~~~~~~~~~~~~~~~
> +
> +When booting with UEFI, the EFI stub requires the boot hartid in order to pass
> +it to the RISC-V kernel in `$a1`. The EFI stub retrieves the boot hartid using
> +one of the following methods:
> +
> +- `RISCV_EFI_BOOT_PROTOCOL` (**preferred**).
> +- `boot-hartid` devicetree subnode (**deprecated**).
> +
> +Any new firmware must implement `RISCV_EFI_BOOT_PROTOCOL` as the devicetree
> +based approach is deprecated now.
> +
> +Early Boot Requirements and Constraints
> +=======================================
> +
> +The RISC-V kernel's early boot process operates under the following constraints:
> +
> +EFI stub and devicetree
> +-----------------------
> +
> +When booting with UEFI, the devicetree is supplemented (or created) by the EFI
> +stub with the same parameters as arm64 which are described at the paragraph
> +"UEFI kernel support on ARM" in Documentation/arm/uefi.rst.
> +
> +Virtual mapping installation
> +----------------------------
> +
> +The installation of the virtual mapping is done in 2 steps in the RISC-V kernel:
> +
> +1. :c:func:`setup_vm` installs a temporary kernel mapping in
> +   :c:var:`early_pg_dir` which allows discovery of the system memory. Only the
> +   kernel text/data are mapped at this point. When establishing this mapping, no
> +   allocation can be done (since the system memory is not known yet), so
> +   :c:var:`early_pg_dir` page table is statically allocated (using only one
> +   table for each level).
> +
> +2. :c:func:`setup_vm_final` creates the final kernel mapping in
> +   :c:var:`swapper_pg_dir` and takes advantage of the discovered system memory
> +   to create the linear mapping. When establishing this mapping, the kernel
> +   can allocate memory but cannot access it directly (since the direct mapping
> +   is not present yet), so it uses temporary mappings in the fixmap region to
> +   be able to access the newly allocated page table levels.
> +
> +For :c:func:`virt_to_phys` and :c:func:`phys_to_virt` to be able to correctly
> +convert direct mapping addresses to physical addresses, they need to know the
> +start of the DRAM. This happens after step 1, right before step 2 installs the
> +direct mapping (see :c:func:`setup_bootmem` function in arch/riscv/mm/init.c).
> +Any usage of those macros before the final virtual mapping is installed must
> +be carefully examined.
> +
> +Device-tree mapping via fixmap
> +------------------------------
> +
> +The RISC-V kernel uses the fixmap region to map the devicetree because the
> +devicetree virtual mapping must remain the same between :c:func:`setup_vm` and
> +:c:func:`setup_vm_final` calls since the :c:var:`reserved_mem` array is
> +initialized with virtual addresses established by :c:func:`setup_vm` and used
> +with the mapping established by :c:func:`setup_vm_final`.
> +
> +Pre-MMU execution
> +-----------------
> +
> +A few pieces of code need to run before even the first virtual mapping is
> +established. These are the installation of the first virtual mapping itself,
> +patching of early alternatives and the early parsing of the kernel command line.
> +That code must be very carefully compiled as:
> +
> +- `-fno-pie`: This is needed for relocatable kernels which use `-fPIE`, since
> +  otherwise, any access to a global symbol would go through the GOT which is
> +  only relocated virtually.
> +- `-mcmodel=medany`: Any access to a global symbol must be PC-relative to avoid
> +  any relocations to happen before the MMU is setup.
> +- *all* instrumentation must also be disabled (that includes KASAN, ftrace and
> +  others).
> +
> +As using a symbol from a different compilation unit requires this unit to be
> +compiled with those flags, we advise, as much as possible, not to use external
> +symbols.
> diff --git a/Documentation/riscv/index.rst b/Documentation/riscv/index.rst
> index 175a91db0200..1f66062def6d 100644
> --- a/Documentation/riscv/index.rst
> +++ b/Documentation/riscv/index.rst
> @@ -5,6 +5,7 @@ RISC-V architecture
>   .. toctree::
>       :maxdepth: 1
>   
> +    boot
>       boot-image-header
>       vm-layout
>       hwprobe

Sunil V L June 21, 2023, 9:18 a.m. UTC | #2

On Wed, Jun 21, 2023 at 04:15:07PM +0800, Song Shuai wrote:
> 
> 
> 在 2023/6/21 15:22, Alexandre Ghiti 写道:
> > This document describes the constraints and requirements of the early
> > boot process in a RISC-V kernel.
> > 
> > Signed-off-by: Alexandre Ghiti <[e-mail]>
> > Reviewed-by: Björn Töpel <[e-mail]>
> > ---
> >   Documentation/riscv/boot-image-header.rst |   3 -
> >   Documentation/riscv/boot.rst              | 170 ++++++++++++++++++++++
> >   Documentation/riscv/index.rst             |   1 +
> >   3 files changed, 171 insertions(+), 3 deletions(-)
> >   create mode 100644 Documentation/riscv/boot.rst
> > 
> > diff --git a/Documentation/riscv/boot-image-header.rst b/Documentation/riscv/boot-image-header.rst
> > index d7752533865f..a4a45310c4c4 100644
> > --- a/Documentation/riscv/boot-image-header.rst
> > +++ b/Documentation/riscv/boot-image-header.rst
> > @@ -7,9 +7,6 @@ Boot image header in RISC-V Linux
> >   This document only describes the boot image header details for RISC-V Linux.
> > -TODO:
> > -  Write a complete booting guide.
> > -
> >   The following 64-byte header is present in decompressed Linux kernel image::
> >   	u32 code0;		  /* Executable code */
> > diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
> > new file mode 100644
> > index 000000000000..019ee818686d
> > --- /dev/null
> > +++ b/Documentation/riscv/boot.rst
> > @@ -0,0 +1,170 @@
> > +.. SPDX-License-Identifier: GPL-2.0
> > +
> > +===============================================
> > +RISC-V Kernel Boot Requirements and Constraints
> > +===============================================
> > +
> > +:Author: Alexandre Ghiti <[e-mail]>
> > +:Date: 23 May 2023
> > +
> > +This document describes what the RISC-V kernel expects from bootloaders and
> > +firmware, but also the constraints that any developer must have in mind when
> > +touching the early boot process. For the purposes of this document, the
> > +'early boot process' refers to any code that runs before the final virtual
> > +mapping is set up.
> > +
> > +Pre-kernel Requirements and Constraints
> > +=======================================
> > +
> > +The RISC-V kernel expects the following of bootloaders and platform firmware:
> > +
> > +Register state
> > +--------------
> > +
> > +The RISC-V kernel expects:
> > +
> > +  * `$a0` to contain the hartid of the current core.
> > +  * `$a1` to contain the address of the devicetree in memory.
> > +
> > +CSR state
> > +---------
> > +
> > +The RISC-V kernel expects:
> > +
> > +  * `$satp = 0`: the MMU, if present, must be disabled.
> > +
> > +Reserved memory for resident firmware
> > +-------------------------------------
> > +
> > +The RISC-V kernel must not map any resident memory, or memory protected with
> > +PMPs, in the direct mapping, so the firmware must correctly mark those regions
> > +as per the devicetree specification and/or the UEFI specification.
> > +
> > +Kernel location
> > +---------------
> > +
> > +The RISC-V kernel expects to be placed at a PMD boundary (2MB aligned for rv64
> > +and 4MB aligned for rv32). Note that the EFI stub will physically relocate the
> > +kernel if that's not the case.
> > +
> > +Hardware description
> > +--------------------
> > +
> > +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> > +
> > +The devicetree is either passed directly to the kernel from the previous stage
> > +using the `$a1` register, or when booting with UEFI, it can be passed using the
> > +EFI configuration table.
> > +
> > +The ACPI tables are passed to the kernel using the EFI configuration table. In
> > +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> > +"EFI stub and devicetree" tree section below for details about this devicetree.
> > +
> > +Kernel entrance
> > +---------------
> > +
> > +On SMP systems, there are 2 methods to enter the kernel:
> > +
> > +- `RISCV_BOOT_SPINWAIT`: the firmware releases all harts in the kernel, one hart
> > +  wins a lottery and executes the early boot code while the other harts are
> > +  parked waiting for the initialization to finish. This method is mostly used to
> > +  support older firmwares without SBI HSM extension and M-mode RISC-V kernel.
> > +- `Ordered booting`: the firmware releases only one hart that will execute the
> > +  initialization phase and then will start all other harts using the SBI HSM
> > +  extension. The ordered booting method is the preferred booting method for
> > +  booting the RISC-V kernel because it can support cpu hotplug and kexec.
> > +
> > +UEFI
> > +----
> > +
> > +UEFI memory map
> > +~~~~~~~~~~~~~~~
> > +
> > +When booting with UEFI, the RISC-V kernel will use only the EFI memory map to
> > +populate the system memory.
> > +
> > +The UEFI firmware must parse the subnodes of the `/reserved-memory` devicetree
> > +node and abide by the devicetree specification to convert the attributes of
> > +those subnodes (`no-map` and `reusable`) into their correct EFI equivalent
> > +(refer to section "3.5.4 /reserved-memory and UEFI" of the devicetree
> > +specification v0.4-rc1).
> append this note ?
> 
> Note that RISC-V edk2 diverges from the devicetree specification to declare
> the !no-map regions as EfiReservedMemoryType instead of EfiBootServicesData.
> > +
Not required. It will be fixed in EDK2.

Conor Dooley June 21, 2023, 12:15 p.m. UTC | #3

Hey Alex,

On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> This document describes the constraints and requirements of the early
> boot process in a RISC-V kernel.

Thanks a lot for writing this & implementing the feedback :)
I had one or two minor comments about rather long sentences, but there's
little point I think in respinning for that, so I have ignored them.
Other than that, didn't spot anything new on this second reading.
Reviewed-by: Conor Dooley <[e-mail]>

Cheers,
Conor.

Andrew Jones June 21, 2023, 12:19 p.m. UTC | #4

On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> This document describes the constraints and requirements of the early
> boot process in a RISC-V kernel.
> 
> Signed-off-by: Alexandre Ghiti <[e-mail]>
> Reviewed-by: Björn Töpel <[e-mail]>
> ---
>  Documentation/riscv/boot-image-header.rst |   3 -
>  Documentation/riscv/boot.rst              | 170 ++++++++++++++++++++++
>  Documentation/riscv/index.rst             |   1 +
>  3 files changed, 171 insertions(+), 3 deletions(-)
>  create mode 100644 Documentation/riscv/boot.rst
> 
> diff --git a/Documentation/riscv/boot-image-header.rst b/Documentation/riscv/boot-image-header.rst
> index d7752533865f..a4a45310c4c4 100644
> --- a/Documentation/riscv/boot-image-header.rst
> +++ b/Documentation/riscv/boot-image-header.rst
> @@ -7,9 +7,6 @@ Boot image header in RISC-V Linux
>  
>  This document only describes the boot image header details for RISC-V Linux.
>  
> -TODO:
> -  Write a complete booting guide.
> -
>  The following 64-byte header is present in decompressed Linux kernel image::
>  
>  	u32 code0;		  /* Executable code */
> diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
> new file mode 100644
> index 000000000000..019ee818686d
> --- /dev/null
> +++ b/Documentation/riscv/boot.rst
> @@ -0,0 +1,170 @@
> +.. SPDX-License-Identifier: GPL-2.0
> +
> +===============================================
> +RISC-V Kernel Boot Requirements and Constraints
> +===============================================
> +
> +:Author: Alexandre Ghiti <[e-mail]>
> +:Date: 23 May 2023
> +
> +This document describes what the RISC-V kernel expects from bootloaders and
> +firmware, but also the constraints that any developer must have in mind when
> +touching the early boot process. For the purposes of this document, the
> +'early boot process' refers to any code that runs before the final virtual
> +mapping is set up.
> +
> +Pre-kernel Requirements and Constraints
> +=======================================
> +
> +The RISC-V kernel expects the following of bootloaders and platform firmware:
> +
> +Register state
> +--------------
> +
> +The RISC-V kernel expects:
> +
> +  * `$a0` to contain the hartid of the current core.
> +  * `$a1` to contain the address of the devicetree in memory.
> +
> +CSR state
> +---------
> +
> +The RISC-V kernel expects:
> +
> +  * `$satp = 0`: the MMU, if present, must be disabled.
> +
> +Reserved memory for resident firmware
> +-------------------------------------
> +
> +The RISC-V kernel must not map any resident memory, or memory protected with
> +PMPs, in the direct mapping, so the firmware must correctly mark those regions
> +as per the devicetree specification and/or the UEFI specification.
> +
> +Kernel location
> +---------------
> +
> +The RISC-V kernel expects to be placed at a PMD boundary (2MB aligned for rv64
> +and 4MB aligned for rv32). Note that the EFI stub will physically relocate the
> +kernel if that's not the case.
> +
> +Hardware description
> +--------------------
> +
> +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> +
> +The devicetree is either passed directly to the kernel from the previous stage
> +using the `$a1` register, or when booting with UEFI, it can be passed using the
> +EFI configuration table.
> +
> +The ACPI tables are passed to the kernel using the EFI configuration table. In
> +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> +"EFI stub and devicetree" tree section below for details about this devicetree.
                             ^ redundant 'tree' here

> +
> +Kernel entrance
> +---------------
> +
> +On SMP systems, there are 2 methods to enter the kernel:
> +
> +- `RISCV_BOOT_SPINWAIT`: the firmware releases all harts in the kernel, one hart
> +  wins a lottery and executes the early boot code while the other harts are
> +  parked waiting for the initialization to finish. This method is mostly used to
> +  support older firmwares without SBI HSM extension and M-mode RISC-V kernel.
> +- `Ordered booting`: the firmware releases only one hart that will execute the
> +  initialization phase and then will start all other harts using the SBI HSM
> +  extension. The ordered booting method is the preferred booting method for
> +  booting the RISC-V kernel because it can support cpu hotplug and kexec.
> +
> +UEFI
> +----
> +
> +UEFI memory map
> +~~~~~~~~~~~~~~~
> +
> +When booting with UEFI, the RISC-V kernel will use only the EFI memory map to
> +populate the system memory.
> +
> +The UEFI firmware must parse the subnodes of the `/reserved-memory` devicetree
> +node and abide by the devicetree specification to convert the attributes of
> +those subnodes (`no-map` and `reusable`) into their correct EFI equivalent
> +(refer to section "3.5.4 /reserved-memory and UEFI" of the devicetree
> +specification v0.4-rc1).
> +
> +RISCV_EFI_BOOT_PROTOCOL
> +~~~~~~~~~~~~~~~~~~~~~~~
> +
> +When booting with UEFI, the EFI stub requires the boot hartid in order to pass
> +it to the RISC-V kernel in `$a1`. The EFI stub retrieves the boot hartid using
> +one of the following methods:
> +
> +- `RISCV_EFI_BOOT_PROTOCOL` (**preferred**).
> +- `boot-hartid` devicetree subnode (**deprecated**).
> +
> +Any new firmware must implement `RISCV_EFI_BOOT_PROTOCOL` as the devicetree
> +based approach is deprecated now.
> +
> +Early Boot Requirements and Constraints
> +=======================================
> +
> +The RISC-V kernel's early boot process operates under the following constraints:
> +
> +EFI stub and devicetree
> +-----------------------
> +
> +When booting with UEFI, the devicetree is supplemented (or created) by the EFI
> +stub with the same parameters as arm64 which are described at the paragraph
> +"UEFI kernel support on ARM" in Documentation/arm/uefi.rst.
> +
> +Virtual mapping installation
> +----------------------------
> +
> +The installation of the virtual mapping is done in 2 steps in the RISC-V kernel:
> +
> +1. :c:func:`setup_vm` installs a temporary kernel mapping in
> +   :c:var:`early_pg_dir` which allows discovery of the system memory. Only the
> +   kernel text/data are mapped at this point. When establishing this mapping, no
> +   allocation can be done (since the system memory is not known yet), so
> +   :c:var:`early_pg_dir` page table is statically allocated (using only one
> +   table for each level).
> +
> +2. :c:func:`setup_vm_final` creates the final kernel mapping in
> +   :c:var:`swapper_pg_dir` and takes advantage of the discovered system memory
> +   to create the linear mapping. When establishing this mapping, the kernel
> +   can allocate memory but cannot access it directly (since the direct mapping
> +   is not present yet), so it uses temporary mappings in the fixmap region to
> +   be able to access the newly allocated page table levels.
> +
> +For :c:func:`virt_to_phys` and :c:func:`phys_to_virt` to be able to correctly
> +convert direct mapping addresses to physical addresses, they need to know the
> +start of the DRAM. This happens after step 1, right before step 2 installs the
> +direct mapping (see :c:func:`setup_bootmem` function in arch/riscv/mm/init.c).
> +Any usage of those macros before the final virtual mapping is installed must
> +be carefully examined.
> +
> +Device-tree mapping via fixmap
> +------------------------------
> +
> +The RISC-V kernel uses the fixmap region to map the devicetree because the
> +devicetree virtual mapping must remain the same between :c:func:`setup_vm` and
> +:c:func:`setup_vm_final` calls since the :c:var:`reserved_mem` array is
> +initialized with virtual addresses established by :c:func:`setup_vm` and used
> +with the mapping established by :c:func:`setup_vm_final`.
> +
> +Pre-MMU execution
> +-----------------
> +
> +A few pieces of code need to run before even the first virtual mapping is
> +established. These are the installation of the first virtual mapping itself,
> +patching of early alternatives and the early parsing of the kernel command line.
> +That code must be very carefully compiled as:
> +
> +- `-fno-pie`: This is needed for relocatable kernels which use `-fPIE`, since
> +  otherwise, any access to a global symbol would go through the GOT which is
> +  only relocated virtually.
> +- `-mcmodel=medany`: Any access to a global symbol must be PC-relative to avoid
> +  any relocations to happen before the MMU is setup.
> +- *all* instrumentation must also be disabled (that includes KASAN, ftrace and
> +  others).
> +
> +As using a symbol from a different compilation unit requires this unit to be
> +compiled with those flags, we advise, as much as possible, not to use external
> +symbols.
> diff --git a/Documentation/riscv/index.rst b/Documentation/riscv/index.rst
> index 175a91db0200..1f66062def6d 100644
> --- a/Documentation/riscv/index.rst
> +++ b/Documentation/riscv/index.rst
> @@ -5,6 +5,7 @@ RISC-V architecture
>  .. toctree::
>      :maxdepth: 1
>  
> +    boot
>      boot-image-header
>      vm-layout
>      hwprobe
> -- 
> 2.39.2
>

Otherwise looks good to me.

Thanks,
drew

Alexandre Ghiti June 21, 2023, 12:26 p.m. UTC | #5

On Wed, Jun 21, 2023 at 2:15 PM Conor Dooley <[e-mail]> wrote:
>
> Hey Alex,
>
> On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> > This document describes the constraints and requirements of the early
> > boot process in a RISC-V kernel.
>
> Thanks a lot for writing this & implementing the feedback :)
> I had one or two minor comments about rather long sentences, but there's

Arf, sorry, which one? So that I can fix that in case of a v3.

> little point I think in respinning for that, so I have ignored them.
> Other than that, didn't spot anything new on this second reading.
> Reviewed-by: Conor Dooley <[e-mail]>
>
> Cheers,
> Conor.

Sunil V L June 21, 2023, 1:23 p.m. UTC | #6

On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> This document describes the constraints and requirements of the early
> boot process in a RISC-V kernel.
> 
> Signed-off-by: Alexandre Ghiti <[e-mail]>
> Reviewed-by: Björn Töpel <[e-mail]>
> ---

Reviewed-by: Sunil V L <[e-mail]>

Thanks!
Sunil

Conor Dooley June 22, 2023, 2:19 p.m. UTC | #7

On Wed, Jun 21, 2023 at 02:26:02PM +0200, Alexandre Ghiti wrote:
> On Wed, Jun 21, 2023 at 2:15 PM Conor Dooley <[e-mail]> wrote:
> >
> > Hey Alex,
> >
> > On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> > > This document describes the constraints and requirements of the early
> > > boot process in a RISC-V kernel.
> >
> > Thanks a lot for writing this & implementing the feedback :)
> > I had one or two minor comments about rather long sentences, but there's
> 
> Arf, sorry, which one? So that I can fix that in case of a v3.

> > little point I think in respinning for that, so I have ignored them.

Nah, not worth the effort ;)

> > Other than that, didn't spot anything new on this second reading.
> > Reviewed-by: Conor Dooley <[e-mail]>
> >
> > Cheers,
> > Conor.

Palmer Dabbelt June 22, 2023, 2:39 p.m. UTC | #8

On Wed, 21 Jun 2023 00:22:33 PDT (-0700), [e-mail] wrote:
> This document describes the constraints and requirements of the early
> boot process in a RISC-V kernel.
>
> Signed-off-by: Alexandre Ghiti <[e-mail]>
> Reviewed-by: Björn Töpel <[e-mail]>
> ---
>  Documentation/riscv/boot-image-header.rst |   3 -
>  Documentation/riscv/boot.rst              | 170 ++++++++++++++++++++++
>  Documentation/riscv/index.rst             |   1 +
>  3 files changed, 171 insertions(+), 3 deletions(-)
>  create mode 100644 Documentation/riscv/boot.rst
>
> diff --git a/Documentation/riscv/boot-image-header.rst b/Documentation/riscv/boot-image-header.rst
> index d7752533865f..a4a45310c4c4 100644
> --- a/Documentation/riscv/boot-image-header.rst
> +++ b/Documentation/riscv/boot-image-header.rst
> @@ -7,9 +7,6 @@ Boot image header in RISC-V Linux
>
>  This document only describes the boot image header details for RISC-V Linux.
>
> -TODO:
> -  Write a complete booting guide.
> -
>  The following 64-byte header is present in decompressed Linux kernel image::
>
>  	u32 code0;		  /* Executable code */
> diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
> new file mode 100644
> index 000000000000..019ee818686d
> --- /dev/null
> +++ b/Documentation/riscv/boot.rst
> @@ -0,0 +1,170 @@
> +.. SPDX-License-Identifier: GPL-2.0
> +
> +===============================================
> +RISC-V Kernel Boot Requirements and Constraints
> +===============================================
> +
> +:Author: Alexandre Ghiti <[e-mail]>
> +:Date: 23 May 2023
> +
> +This document describes what the RISC-V kernel expects from bootloaders and
> +firmware, but also the constraints that any developer must have in mind when
> +touching the early boot process. For the purposes of this document, the
> +'early boot process' refers to any code that runs before the final virtual
> +mapping is set up.
> +
> +Pre-kernel Requirements and Constraints
> +=======================================
> +
> +The RISC-V kernel expects the following of bootloaders and platform firmware:
> +
> +Register state
> +--------------
> +
> +The RISC-V kernel expects:
> +
> +  * `$a0` to contain the hartid of the current core.
> +  * `$a1` to contain the address of the devicetree in memory.
> +
> +CSR state
> +---------
> +
> +The RISC-V kernel expects:
> +
> +  * `$satp = 0`: the MMU, if present, must be disabled.
> +
> +Reserved memory for resident firmware
> +-------------------------------------
> +
> +The RISC-V kernel must not map any resident memory, or memory protected with
> +PMPs, in the direct mapping, so the firmware must correctly mark those regions
> +as per the devicetree specification and/or the UEFI specification.
> +
> +Kernel location
> +---------------
> +
> +The RISC-V kernel expects to be placed at a PMD boundary (2MB aligned for rv64
> +and 4MB aligned for rv32). Note that the EFI stub will physically relocate the
> +kernel if that's not the case.
> +
> +Hardware description
> +--------------------
> +
> +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> +
> +The devicetree is either passed directly to the kernel from the previous stage
> +using the `$a1` register, or when booting with UEFI, it can be passed using the
> +EFI configuration table.
> +
> +The ACPI tables are passed to the kernel using the EFI configuration table. In
> +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> +"EFI stub and devicetree" tree section below for details about this devicetree.
> +
> +Kernel entrance
> +---------------
> +
> +On SMP systems, there are 2 methods to enter the kernel:
> +
> +- `RISCV_BOOT_SPINWAIT`: the firmware releases all harts in the kernel, one hart
> +  wins a lottery and executes the early boot code while the other harts are
> +  parked waiting for the initialization to finish. This method is mostly used to
> +  support older firmwares without SBI HSM extension and M-mode RISC-V kernel.
> +- `Ordered booting`: the firmware releases only one hart that will execute the
> +  initialization phase and then will start all other harts using the SBI HSM
> +  extension. The ordered booting method is the preferred booting method for
> +  booting the RISC-V kernel because it can support cpu hotplug and kexec.
> +
> +UEFI
> +----
> +
> +UEFI memory map
> +~~~~~~~~~~~~~~~
> +
> +When booting with UEFI, the RISC-V kernel will use only the EFI memory map to
> +populate the system memory.
> +
> +The UEFI firmware must parse the subnodes of the `/reserved-memory` devicetree
> +node and abide by the devicetree specification to convert the attributes of
> +those subnodes (`no-map` and `reusable`) into their correct EFI equivalent
> +(refer to section "3.5.4 /reserved-memory and UEFI" of the devicetree
> +specification v0.4-rc1).
> +
> +RISCV_EFI_BOOT_PROTOCOL
> +~~~~~~~~~~~~~~~~~~~~~~~
> +
> +When booting with UEFI, the EFI stub requires the boot hartid in order to pass
> +it to the RISC-V kernel in `$a1`. The EFI stub retrieves the boot hartid using
> +one of the following methods:
> +
> +- `RISCV_EFI_BOOT_PROTOCOL` (**preferred**).
> +- `boot-hartid` devicetree subnode (**deprecated**).
> +
> +Any new firmware must implement `RISCV_EFI_BOOT_PROTOCOL` as the devicetree
> +based approach is deprecated now.
> +
> +Early Boot Requirements and Constraints
> +=======================================
> +
> +The RISC-V kernel's early boot process operates under the following constraints:
> +
> +EFI stub and devicetree
> +-----------------------
> +
> +When booting with UEFI, the devicetree is supplemented (or created) by the EFI
> +stub with the same parameters as arm64 which are described at the paragraph
> +"UEFI kernel support on ARM" in Documentation/arm/uefi.rst.
> +
> +Virtual mapping installation
> +----------------------------
> +
> +The installation of the virtual mapping is done in 2 steps in the RISC-V kernel:
> +
> +1. :c:func:`setup_vm` installs a temporary kernel mapping in
> +   :c:var:`early_pg_dir` which allows discovery of the system memory. Only the
> +   kernel text/data are mapped at this point. When establishing this mapping, no
> +   allocation can be done (since the system memory is not known yet), so
> +   :c:var:`early_pg_dir` page table is statically allocated (using only one
> +   table for each level).
> +
> +2. :c:func:`setup_vm_final` creates the final kernel mapping in
> +   :c:var:`swapper_pg_dir` and takes advantage of the discovered system memory
> +   to create the linear mapping. When establishing this mapping, the kernel
> +   can allocate memory but cannot access it directly (since the direct mapping
> +   is not present yet), so it uses temporary mappings in the fixmap region to
> +   be able to access the newly allocated page table levels.
> +
> +For :c:func:`virt_to_phys` and :c:func:`phys_to_virt` to be able to correctly
> +convert direct mapping addresses to physical addresses, they need to know the
> +start of the DRAM. This happens after step 1, right before step 2 installs the
> +direct mapping (see :c:func:`setup_bootmem` function in arch/riscv/mm/init.c).
> +Any usage of those macros before the final virtual mapping is installed must
> +be carefully examined.
> +
> +Device-tree mapping via fixmap
> +------------------------------
> +
> +The RISC-V kernel uses the fixmap region to map the devicetree because the
> +devicetree virtual mapping must remain the same between :c:func:`setup_vm` and
> +:c:func:`setup_vm_final` calls since the :c:var:`reserved_mem` array is
> +initialized with virtual addresses established by :c:func:`setup_vm` and used
> +with the mapping established by :c:func:`setup_vm_final`.
> +
> +Pre-MMU execution
> +-----------------
> +
> +A few pieces of code need to run before even the first virtual mapping is
> +established. These are the installation of the first virtual mapping itself,
> +patching of early alternatives and the early parsing of the kernel command line.
> +That code must be very carefully compiled as:
> +
> +- `-fno-pie`: This is needed for relocatable kernels which use `-fPIE`, since
> +  otherwise, any access to a global symbol would go through the GOT which is
> +  only relocated virtually.
> +- `-mcmodel=medany`: Any access to a global symbol must be PC-relative to avoid
> +  any relocations to happen before the MMU is setup.
> +- *all* instrumentation must also be disabled (that includes KASAN, ftrace and
> +  others).
> +
> +As using a symbol from a different compilation unit requires this unit to be
> +compiled with those flags, we advise, as much as possible, not to use external
> +symbols.
> diff --git a/Documentation/riscv/index.rst b/Documentation/riscv/index.rst
> index 175a91db0200..1f66062def6d 100644
> --- a/Documentation/riscv/index.rst
> +++ b/Documentation/riscv/index.rst
> @@ -5,6 +5,7 @@ RISC-V architecture
>  .. toctree::
>      :maxdepth: 1
>
> +    boot
>      boot-image-header
>      vm-layout
>      hwprobe

Reviewed-by: Palmer Dabbelt <[e-mail]>
Acked-by: Palmer Dabbelt <[e-mail]>

in case someone wants to pick this up along with the others.

Andrew Jones June 22, 2023, 5:23 p.m. UTC | #9

On Wed, Jun 21, 2023 at 02:19:50PM +0200, Andrew Jones wrote:
> On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
...
> > +Hardware description
> > +--------------------
> > +
> > +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> > +
> > +The devicetree is either passed directly to the kernel from the previous stage
> > +using the `$a1` register, or when booting with UEFI, it can be passed using the
> > +EFI configuration table.
> > +
> > +The ACPI tables are passed to the kernel using the EFI configuration table. In
> > +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> > +"EFI stub and devicetree" tree section below for details about this devicetree.
>                              ^ redundant 'tree' here
>

Can whoever applies this drop this 'tree' I pointed while applying?

Thanks,
drew

Alexandre Ghiti June 22, 2023, 5:32 p.m. UTC | #10

On Thu, Jun 22, 2023 at 7:24 PM Andrew Jones <[e-mail]> wrote:
>
> On Wed, Jun 21, 2023 at 02:19:50PM +0200, Andrew Jones wrote:
> > On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> ...
> > > +Hardware description
> > > +--------------------
> > > +
> > > +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> > > +
> > > +The devicetree is either passed directly to the kernel from the previous stage
> > > +using the `$a1` register, or when booting with UEFI, it can be passed using the
> > > +EFI configuration table.
> > > +
> > > +The ACPI tables are passed to the kernel using the EFI configuration table. In
> > > +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> > > +"EFI stub and devicetree" tree section below for details about this devicetree.
> >                              ^ redundant 'tree' here
> >
>
> Can whoever applies this drop this 'tree' I pointed while applying?

Yes, sorry: @Conor Dooley give me the overly long lines you wanted me
to break and I'll send a v3, it won't take much time and I'll fix this
too!

>
> Thanks,
> drew

Conor Dooley June 22, 2023, 5:41 p.m. UTC | #11

On Thu, Jun 22, 2023 at 07:32:23PM +0200, Alexandre Ghiti wrote:
> On Thu, Jun 22, 2023 at 7:24 PM Andrew Jones <[e-mail]> wrote:
> >
> > On Wed, Jun 21, 2023 at 02:19:50PM +0200, Andrew Jones wrote:
> > > On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> > ...
> > > > +Hardware description
> > > > +--------------------
> > > > +
> > > > +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> > > > +
> > > > +The devicetree is either passed directly to the kernel from the previous stage
> > > > +using the `$a1` register, or when booting with UEFI, it can be passed using the
> > > > +EFI configuration table.
> > > > +
> > > > +The ACPI tables are passed to the kernel using the EFI configuration table. In
> > > > +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> > > > +"EFI stub and devicetree" tree section below for details about this devicetree.
> > >                              ^ redundant 'tree' here
> > >
> >
> > Can whoever applies this drop this 'tree' I pointed while applying?
> 
> Yes, sorry: @Conor Dooley give me the overly long lines you wanted me
> to break and I'll send a v3, it won't take much time and I'll fix this
> too!

Look, I said it was pure pedantry on my part, and not something that
helps with understanding. You've been warned!

diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
index 019ee818686d..a52793fd9f6c 100644
--- a/Documentation/riscv/boot.rst
+++ b/Documentation/riscv/boot.rst
@@ -143,11 +143,12 @@ be carefully examined.
 Device-tree mapping via fixmap
 ------------------------------
 
-The RISC-V kernel uses the fixmap region to map the devicetree because the
-devicetree virtual mapping must remain the same between :c:func:`setup_vm` and
-:c:func:`setup_vm_final` calls since the :c:var:`reserved_mem` array is
-initialized with virtual addresses established by :c:func:`setup_vm` and used
-with the mapping established by :c:func:`setup_vm_final`.
+
+As the :c:var:`reserved_mem` array is initialized with virtual addresses
+established by :c:func:`setup_vm`, and used with the mapping established by
+:c:func:`setup_vm_final`, the RISC-V kernel uses the fixmap region to map the
+devicetree. This ensures that the same virtual mapping is used by both
+functions.
 
 Pre-MMU execution
 -----------------

Alexandre Ghiti June 23, 2023, 8:32 a.m. UTC | #12

On Thu, Jun 22, 2023 at 7:41 PM Conor Dooley <[e-mail]> wrote:
>
> On Thu, Jun 22, 2023 at 07:32:23PM +0200, Alexandre Ghiti wrote:
> > On Thu, Jun 22, 2023 at 7:24 PM Andrew Jones <[e-mail]> wrote:
> > >
> > > On Wed, Jun 21, 2023 at 02:19:50PM +0200, Andrew Jones wrote:
> > > > On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> > > ...
> > > > > +Hardware description
> > > > > +--------------------
> > > > > +
> > > > > +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> > > > > +
> > > > > +The devicetree is either passed directly to the kernel from the previous stage
> > > > > +using the `$a1` register, or when booting with UEFI, it can be passed using the
> > > > > +EFI configuration table.
> > > > > +
> > > > > +The ACPI tables are passed to the kernel using the EFI configuration table. In
> > > > > +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> > > > > +"EFI stub and devicetree" tree section below for details about this devicetree.
> > > >                              ^ redundant 'tree' here
> > > >
> > >
> > > Can whoever applies this drop this 'tree' I pointed while applying?
> >
> > Yes, sorry: @Conor Dooley give me the overly long lines you wanted me
> > to break and I'll send a v3, it won't take much time and I'll fix this
> > too!
>
> Look, I said it was pure pedantry on my part, and not something that
> helps with understanding. You've been warned!
>
> diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
> index 019ee818686d..a52793fd9f6c 100644
> --- a/Documentation/riscv/boot.rst
> +++ b/Documentation/riscv/boot.rst
> @@ -143,11 +143,12 @@ be carefully examined.
>  Device-tree mapping via fixmap
>  ------------------------------
>
> -The RISC-V kernel uses the fixmap region to map the devicetree because the
> -devicetree virtual mapping must remain the same between :c:func:`setup_vm` and
> -:c:func:`setup_vm_final` calls since the :c:var:`reserved_mem` array is
> -initialized with virtual addresses established by :c:func:`setup_vm` and used
> -with the mapping established by :c:func:`setup_vm_final`.
> +
> +As the :c:var:`reserved_mem` array is initialized with virtual addresses
> +established by :c:func:`setup_vm`, and used with the mapping established by
> +:c:func:`setup_vm_final`, the RISC-V kernel uses the fixmap region to map the
> +devicetree. This ensures that the same virtual mapping is used by both
> +functions.
>

Oh my, so pedantic...:) I don't like the last sentence though, I'll
change it for:

"This ensures that the devicetree remains accessible by both virtual mappings."

Thanks!

>  Pre-MMU execution
>  -----------------

Alexandre Ghiti June 23, 2023, 8:48 a.m. UTC | #13

On 21/06/2023 14:19, Andrew Jones wrote:
> On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
>> This document describes the constraints and requirements of the early
>> boot process in a RISC-V kernel.
>>
>> Signed-off-by: Alexandre Ghiti <[e-mail]>
>> Reviewed-by: Björn Töpel <[e-mail]>
>> ---
>>   Documentation/riscv/boot-image-header.rst |   3 -
>>   Documentation/riscv/boot.rst              | 170 ++++++++++++++++++++++
>>   Documentation/riscv/index.rst             |   1 +
>>   3 files changed, 171 insertions(+), 3 deletions(-)
>>   create mode 100644 Documentation/riscv/boot.rst
>>
>> diff --git a/Documentation/riscv/boot-image-header.rst b/Documentation/riscv/boot-image-header.rst
>> index d7752533865f..a4a45310c4c4 100644
>> --- a/Documentation/riscv/boot-image-header.rst
>> +++ b/Documentation/riscv/boot-image-header.rst
>> @@ -7,9 +7,6 @@ Boot image header in RISC-V Linux
>>   
>>   This document only describes the boot image header details for RISC-V Linux.
>>   
>> -TODO:
>> -  Write a complete booting guide.
>> -
>>   The following 64-byte header is present in decompressed Linux kernel image::
>>   
>>   	u32 code0;		  /* Executable code */
>> diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
>> new file mode 100644
>> index 000000000000..019ee818686d
>> --- /dev/null
>> +++ b/Documentation/riscv/boot.rst
>> @@ -0,0 +1,170 @@
>> +.. SPDX-License-Identifier: GPL-2.0
>> +
>> +===============================================
>> +RISC-V Kernel Boot Requirements and Constraints
>> +===============================================
>> +
>> +:Author: Alexandre Ghiti <[e-mail]>
>> +:Date: 23 May 2023
>> +
>> +This document describes what the RISC-V kernel expects from bootloaders and
>> +firmware, but also the constraints that any developer must have in mind when
>> +touching the early boot process. For the purposes of this document, the
>> +'early boot process' refers to any code that runs before the final virtual
>> +mapping is set up.
>> +
>> +Pre-kernel Requirements and Constraints
>> +=======================================
>> +
>> +The RISC-V kernel expects the following of bootloaders and platform firmware:
>> +
>> +Register state
>> +--------------
>> +
>> +The RISC-V kernel expects:
>> +
>> +  * `$a0` to contain the hartid of the current core.
>> +  * `$a1` to contain the address of the devicetree in memory.
>> +
>> +CSR state
>> +---------
>> +
>> +The RISC-V kernel expects:
>> +
>> +  * `$satp = 0`: the MMU, if present, must be disabled.
>> +
>> +Reserved memory for resident firmware
>> +-------------------------------------
>> +
>> +The RISC-V kernel must not map any resident memory, or memory protected with
>> +PMPs, in the direct mapping, so the firmware must correctly mark those regions
>> +as per the devicetree specification and/or the UEFI specification.
>> +
>> +Kernel location
>> +---------------
>> +
>> +The RISC-V kernel expects to be placed at a PMD boundary (2MB aligned for rv64
>> +and 4MB aligned for rv32). Note that the EFI stub will physically relocate the
>> +kernel if that's not the case.
>> +
>> +Hardware description
>> +--------------------
>> +
>> +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
>> +
>> +The devicetree is either passed directly to the kernel from the previous stage
>> +using the `$a1` register, or when booting with UEFI, it can be passed using the
>> +EFI configuration table.
>> +
>> +The ACPI tables are passed to the kernel using the EFI configuration table. In
>> +this case, a tiny devicetree is still created by the EFI stub. Please refer to
>> +"EFI stub and devicetree" tree section below for details about this devicetree.
>                               ^ redundant 'tree' here
>
>> +
>> +Kernel entrance
>> +---------------
>> +
>> +On SMP systems, there are 2 methods to enter the kernel:
>> +
>> +- `RISCV_BOOT_SPINWAIT`: the firmware releases all harts in the kernel, one hart
>> +  wins a lottery and executes the early boot code while the other harts are
>> +  parked waiting for the initialization to finish. This method is mostly used to
>> +  support older firmwares without SBI HSM extension and M-mode RISC-V kernel.
>> +- `Ordered booting`: the firmware releases only one hart that will execute the
>> +  initialization phase and then will start all other harts using the SBI HSM
>> +  extension. The ordered booting method is the preferred booting method for
>> +  booting the RISC-V kernel because it can support cpu hotplug and kexec.
>> +
>> +UEFI
>> +----
>> +
>> +UEFI memory map
>> +~~~~~~~~~~~~~~~
>> +
>> +When booting with UEFI, the RISC-V kernel will use only the EFI memory map to
>> +populate the system memory.
>> +
>> +The UEFI firmware must parse the subnodes of the `/reserved-memory` devicetree
>> +node and abide by the devicetree specification to convert the attributes of
>> +those subnodes (`no-map` and `reusable`) into their correct EFI equivalent
>> +(refer to section "3.5.4 /reserved-memory and UEFI" of the devicetree
>> +specification v0.4-rc1).
>> +
>> +RISCV_EFI_BOOT_PROTOCOL
>> +~~~~~~~~~~~~~~~~~~~~~~~
>> +
>> +When booting with UEFI, the EFI stub requires the boot hartid in order to pass
>> +it to the RISC-V kernel in `$a1`. The EFI stub retrieves the boot hartid using
>> +one of the following methods:
>> +
>> +- `RISCV_EFI_BOOT_PROTOCOL` (**preferred**).
>> +- `boot-hartid` devicetree subnode (**deprecated**).
>> +
>> +Any new firmware must implement `RISCV_EFI_BOOT_PROTOCOL` as the devicetree
>> +based approach is deprecated now.
>> +
>> +Early Boot Requirements and Constraints
>> +=======================================
>> +
>> +The RISC-V kernel's early boot process operates under the following constraints:
>> +
>> +EFI stub and devicetree
>> +-----------------------
>> +
>> +When booting with UEFI, the devicetree is supplemented (or created) by the EFI
>> +stub with the same parameters as arm64 which are described at the paragraph
>> +"UEFI kernel support on ARM" in Documentation/arm/uefi.rst.
>> +
>> +Virtual mapping installation
>> +----------------------------
>> +
>> +The installation of the virtual mapping is done in 2 steps in the RISC-V kernel:
>> +
>> +1. :c:func:`setup_vm` installs a temporary kernel mapping in
>> +   :c:var:`early_pg_dir` which allows discovery of the system memory. Only the
>> +   kernel text/data are mapped at this point. When establishing this mapping, no
>> +   allocation can be done (since the system memory is not known yet), so
>> +   :c:var:`early_pg_dir` page table is statically allocated (using only one
>> +   table for each level).
>> +
>> +2. :c:func:`setup_vm_final` creates the final kernel mapping in
>> +   :c:var:`swapper_pg_dir` and takes advantage of the discovered system memory
>> +   to create the linear mapping. When establishing this mapping, the kernel
>> +   can allocate memory but cannot access it directly (since the direct mapping
>> +   is not present yet), so it uses temporary mappings in the fixmap region to
>> +   be able to access the newly allocated page table levels.
>> +
>> +For :c:func:`virt_to_phys` and :c:func:`phys_to_virt` to be able to correctly
>> +convert direct mapping addresses to physical addresses, they need to know the
>> +start of the DRAM. This happens after step 1, right before step 2 installs the
>> +direct mapping (see :c:func:`setup_bootmem` function in arch/riscv/mm/init.c).
>> +Any usage of those macros before the final virtual mapping is installed must
>> +be carefully examined.
>> +
>> +Device-tree mapping via fixmap
>> +------------------------------
>> +
>> +The RISC-V kernel uses the fixmap region to map the devicetree because the
>> +devicetree virtual mapping must remain the same between :c:func:`setup_vm` and
>> +:c:func:`setup_vm_final` calls since the :c:var:`reserved_mem` array is
>> +initialized with virtual addresses established by :c:func:`setup_vm` and used
>> +with the mapping established by :c:func:`setup_vm_final`.
>> +
>> +Pre-MMU execution
>> +-----------------
>> +
>> +A few pieces of code need to run before even the first virtual mapping is
>> +established. These are the installation of the first virtual mapping itself,
>> +patching of early alternatives and the early parsing of the kernel command line.
>> +That code must be very carefully compiled as:
>> +
>> +- `-fno-pie`: This is needed for relocatable kernels which use `-fPIE`, since
>> +  otherwise, any access to a global symbol would go through the GOT which is
>> +  only relocated virtually.
>> +- `-mcmodel=medany`: Any access to a global symbol must be PC-relative to avoid
>> +  any relocations to happen before the MMU is setup.
>> +- *all* instrumentation must also be disabled (that includes KASAN, ftrace and
>> +  others).
>> +
>> +As using a symbol from a different compilation unit requires this unit to be
>> +compiled with those flags, we advise, as much as possible, not to use external
>> +symbols.
>> diff --git a/Documentation/riscv/index.rst b/Documentation/riscv/index.rst
>> index 175a91db0200..1f66062def6d 100644
>> --- a/Documentation/riscv/index.rst
>> +++ b/Documentation/riscv/index.rst
>> @@ -5,6 +5,7 @@ RISC-V architecture
>>   .. toctree::
>>       :maxdepth: 1
>>   
>> +    boot
>>       boot-image-header
>>       vm-layout
>>       hwprobe
>> -- 
>> 2.39.2
>>
> Otherwise looks good to me.


I assume I can add your RB here, don't waste time answering if I'm right :)

Thanks!


>
> Thanks,
> drew
>
> _______________________________________________
> linux-riscv mailing list
> [e-mail]
> http://lists.infradead.org/mailman/listinfo/linux-riscv

Andrew Jones June 23, 2023, 9:24 a.m. UTC | #14

On Fri, Jun 23, 2023 at 10:48:55AM +0200, Alexandre Ghiti wrote:
> 
> On 21/06/2023 14:19, Andrew Jones wrote:
> > On Wed, Jun 21, 2023 at 09:22:33AM +0200, Alexandre Ghiti wrote:
> > > This document describes the constraints and requirements of the early
> > > boot process in a RISC-V kernel.
> > > 
> > > Signed-off-by: Alexandre Ghiti <[e-mail]>
> > > Reviewed-by: Björn Töpel <[e-mail]>
> > > ---
> > >   Documentation/riscv/boot-image-header.rst |   3 -
> > >   Documentation/riscv/boot.rst              | 170 ++++++++++++++++++++++
> > >   Documentation/riscv/index.rst             |   1 +
> > >   3 files changed, 171 insertions(+), 3 deletions(-)
> > >   create mode 100644 Documentation/riscv/boot.rst
> > > 
> > > diff --git a/Documentation/riscv/boot-image-header.rst b/Documentation/riscv/boot-image-header.rst
> > > index d7752533865f..a4a45310c4c4 100644
> > > --- a/Documentation/riscv/boot-image-header.rst
> > > +++ b/Documentation/riscv/boot-image-header.rst
> > > @@ -7,9 +7,6 @@ Boot image header in RISC-V Linux
> > >   This document only describes the boot image header details for RISC-V Linux.
> > > -TODO:
> > > -  Write a complete booting guide.
> > > -
> > >   The following 64-byte header is present in decompressed Linux kernel image::
> > >   	u32 code0;		  /* Executable code */
> > > diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
> > > new file mode 100644
> > > index 000000000000..019ee818686d
> > > --- /dev/null
> > > +++ b/Documentation/riscv/boot.rst
> > > @@ -0,0 +1,170 @@
> > > +.. SPDX-License-Identifier: GPL-2.0
> > > +
> > > +===============================================
> > > +RISC-V Kernel Boot Requirements and Constraints
> > > +===============================================
> > > +
> > > +:Author: Alexandre Ghiti <[e-mail]>
> > > +:Date: 23 May 2023
> > > +
> > > +This document describes what the RISC-V kernel expects from bootloaders and
> > > +firmware, but also the constraints that any developer must have in mind when
> > > +touching the early boot process. For the purposes of this document, the
> > > +'early boot process' refers to any code that runs before the final virtual
> > > +mapping is set up.
> > > +
> > > +Pre-kernel Requirements and Constraints
> > > +=======================================
> > > +
> > > +The RISC-V kernel expects the following of bootloaders and platform firmware:
> > > +
> > > +Register state
> > > +--------------
> > > +
> > > +The RISC-V kernel expects:
> > > +
> > > +  * `$a0` to contain the hartid of the current core.
> > > +  * `$a1` to contain the address of the devicetree in memory.
> > > +
> > > +CSR state
> > > +---------
> > > +
> > > +The RISC-V kernel expects:
> > > +
> > > +  * `$satp = 0`: the MMU, if present, must be disabled.
> > > +
> > > +Reserved memory for resident firmware
> > > +-------------------------------------
> > > +
> > > +The RISC-V kernel must not map any resident memory, or memory protected with
> > > +PMPs, in the direct mapping, so the firmware must correctly mark those regions
> > > +as per the devicetree specification and/or the UEFI specification.
> > > +
> > > +Kernel location
> > > +---------------
> > > +
> > > +The RISC-V kernel expects to be placed at a PMD boundary (2MB aligned for rv64
> > > +and 4MB aligned for rv32). Note that the EFI stub will physically relocate the
> > > +kernel if that's not the case.
> > > +
> > > +Hardware description
> > > +--------------------
> > > +
> > > +The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
> > > +
> > > +The devicetree is either passed directly to the kernel from the previous stage
> > > +using the `$a1` register, or when booting with UEFI, it can be passed using the
> > > +EFI configuration table.
> > > +
> > > +The ACPI tables are passed to the kernel using the EFI configuration table. In
> > > +this case, a tiny devicetree is still created by the EFI stub. Please refer to
> > > +"EFI stub and devicetree" tree section below for details about this devicetree.
> >                               ^ redundant 'tree' here
> > 
> > > +
> > > +Kernel entrance
> > > +---------------
> > > +
> > > +On SMP systems, there are 2 methods to enter the kernel:
> > > +
> > > +- `RISCV_BOOT_SPINWAIT`: the firmware releases all harts in the kernel, one hart
> > > +  wins a lottery and executes the early boot code while the other harts are
> > > +  parked waiting for the initialization to finish. This method is mostly used to
> > > +  support older firmwares without SBI HSM extension and M-mode RISC-V kernel.
> > > +- `Ordered booting`: the firmware releases only one hart that will execute the
> > > +  initialization phase and then will start all other harts using the SBI HSM
> > > +  extension. The ordered booting method is the preferred booting method for
> > > +  booting the RISC-V kernel because it can support cpu hotplug and kexec.
> > > +
> > > +UEFI
> > > +----
> > > +
> > > +UEFI memory map
> > > +~~~~~~~~~~~~~~~
> > > +
> > > +When booting with UEFI, the RISC-V kernel will use only the EFI memory map to
> > > +populate the system memory.
> > > +
> > > +The UEFI firmware must parse the subnodes of the `/reserved-memory` devicetree
> > > +node and abide by the devicetree specification to convert the attributes of
> > > +those subnodes (`no-map` and `reusable`) into their correct EFI equivalent
> > > +(refer to section "3.5.4 /reserved-memory and UEFI" of the devicetree
> > > +specification v0.4-rc1).
> > > +
> > > +RISCV_EFI_BOOT_PROTOCOL
> > > +~~~~~~~~~~~~~~~~~~~~~~~
> > > +
> > > +When booting with UEFI, the EFI stub requires the boot hartid in order to pass
> > > +it to the RISC-V kernel in `$a1`. The EFI stub retrieves the boot hartid using
> > > +one of the following methods:
> > > +
> > > +- `RISCV_EFI_BOOT_PROTOCOL` (**preferred**).
> > > +- `boot-hartid` devicetree subnode (**deprecated**).
> > > +
> > > +Any new firmware must implement `RISCV_EFI_BOOT_PROTOCOL` as the devicetree
> > > +based approach is deprecated now.
> > > +
> > > +Early Boot Requirements and Constraints
> > > +=======================================
> > > +
> > > +The RISC-V kernel's early boot process operates under the following constraints:
> > > +
> > > +EFI stub and devicetree
> > > +-----------------------
> > > +
> > > +When booting with UEFI, the devicetree is supplemented (or created) by the EFI
> > > +stub with the same parameters as arm64 which are described at the paragraph
> > > +"UEFI kernel support on ARM" in Documentation/arm/uefi.rst.
> > > +
> > > +Virtual mapping installation
> > > +----------------------------
> > > +
> > > +The installation of the virtual mapping is done in 2 steps in the RISC-V kernel:
> > > +
> > > +1. :c:func:`setup_vm` installs a temporary kernel mapping in
> > > +   :c:var:`early_pg_dir` which allows discovery of the system memory. Only the
> > > +   kernel text/data are mapped at this point. When establishing this mapping, no
> > > +   allocation can be done (since the system memory is not known yet), so
> > > +   :c:var:`early_pg_dir` page table is statically allocated (using only one
> > > +   table for each level).
> > > +
> > > +2. :c:func:`setup_vm_final` creates the final kernel mapping in
> > > +   :c:var:`swapper_pg_dir` and takes advantage of the discovered system memory
> > > +   to create the linear mapping. When establishing this mapping, the kernel
> > > +   can allocate memory but cannot access it directly (since the direct mapping
> > > +   is not present yet), so it uses temporary mappings in the fixmap region to
> > > +   be able to access the newly allocated page table levels.
> > > +
> > > +For :c:func:`virt_to_phys` and :c:func:`phys_to_virt` to be able to correctly
> > > +convert direct mapping addresses to physical addresses, they need to know the
> > > +start of the DRAM. This happens after step 1, right before step 2 installs the
> > > +direct mapping (see :c:func:`setup_bootmem` function in arch/riscv/mm/init.c).
> > > +Any usage of those macros before the final virtual mapping is installed must
> > > +be carefully examined.
> > > +
> > > +Device-tree mapping via fixmap
> > > +------------------------------
> > > +
> > > +The RISC-V kernel uses the fixmap region to map the devicetree because the
> > > +devicetree virtual mapping must remain the same between :c:func:`setup_vm` and
> > > +:c:func:`setup_vm_final` calls since the :c:var:`reserved_mem` array is
> > > +initialized with virtual addresses established by :c:func:`setup_vm` and used
> > > +with the mapping established by :c:func:`setup_vm_final`.
> > > +
> > > +Pre-MMU execution
> > > +-----------------
> > > +
> > > +A few pieces of code need to run before even the first virtual mapping is
> > > +established. These are the installation of the first virtual mapping itself,
> > > +patching of early alternatives and the early parsing of the kernel command line.
> > > +That code must be very carefully compiled as:
> > > +
> > > +- `-fno-pie`: This is needed for relocatable kernels which use `-fPIE`, since
> > > +  otherwise, any access to a global symbol would go through the GOT which is
> > > +  only relocated virtually.
> > > +- `-mcmodel=medany`: Any access to a global symbol must be PC-relative to avoid
> > > +  any relocations to happen before the MMU is setup.
> > > +- *all* instrumentation must also be disabled (that includes KASAN, ftrace and
> > > +  others).
> > > +
> > > +As using a symbol from a different compilation unit requires this unit to be
> > > +compiled with those flags, we advise, as much as possible, not to use external
> > > +symbols.
> > > diff --git a/Documentation/riscv/index.rst b/Documentation/riscv/index.rst
> > > index 175a91db0200..1f66062def6d 100644
> > > --- a/Documentation/riscv/index.rst
> > > +++ b/Documentation/riscv/index.rst
> > > @@ -5,6 +5,7 @@ RISC-V architecture
> > >   .. toctree::
> > >       :maxdepth: 1
> > > +    boot
> > >       boot-image-header
> > >       vm-layout
> > >       hwprobe
> > > -- 
> > > 2.39.2
> > > 
> > Otherwise looks good to me.
> 
> 
> I assume I can add your RB here, don't waste time answering if I'm right :)

Yup, when I read it the first time I wasn't sure I was familiar enough
with everything to give it an r-b, but I just read it again and grepped
a few things that I wasn't already familiar with, so

Reviewed-by: Andrew Jones <[e-mail]>

Thanks,
drew

Patch

diff --git a/Documentation/riscv/boot-image-header.rst b/Documentation/riscv/boot-image-header.rst
index d7752533865f..a4a45310c4c4 100644
--- a/Documentation/riscv/boot-image-header.rst
+++ b/Documentation/riscv/boot-image-header.rst
@@ -7,9 +7,6 @@  Boot image header in RISC-V Linux
 
 This document only describes the boot image header details for RISC-V Linux.
 
-TODO:
-  Write a complete booting guide.
-
 The following 64-byte header is present in decompressed Linux kernel image::
 
 	u32 code0;		  /* Executable code */
diff --git a/Documentation/riscv/boot.rst b/Documentation/riscv/boot.rst
new file mode 100644
index 000000000000..019ee818686d
--- /dev/null
+++ b/Documentation/riscv/boot.rst
@@ -0,0 +1,170 @@ 
+.. SPDX-License-Identifier: GPL-2.0
+
+===============================================
+RISC-V Kernel Boot Requirements and Constraints
+===============================================
+
+:Author: Alexandre Ghiti <[e-mail]>
+:Date: 23 May 2023
+
+This document describes what the RISC-V kernel expects from bootloaders and
+firmware, but also the constraints that any developer must have in mind when
+touching the early boot process. For the purposes of this document, the
+'early boot process' refers to any code that runs before the final virtual
+mapping is set up.
+
+Pre-kernel Requirements and Constraints
+=======================================
+
+The RISC-V kernel expects the following of bootloaders and platform firmware:
+
+Register state
+--------------
+
+The RISC-V kernel expects:
+
+  * `$a0` to contain the hartid of the current core.
+  * `$a1` to contain the address of the devicetree in memory.
+
+CSR state
+---------
+
+The RISC-V kernel expects:
+
+  * `$satp = 0`: the MMU, if present, must be disabled.
+
+Reserved memory for resident firmware
+-------------------------------------
+
+The RISC-V kernel must not map any resident memory, or memory protected with
+PMPs, in the direct mapping, so the firmware must correctly mark those regions
+as per the devicetree specification and/or the UEFI specification.
+
+Kernel location
+---------------
+
+The RISC-V kernel expects to be placed at a PMD boundary (2MB aligned for rv64
+and 4MB aligned for rv32). Note that the EFI stub will physically relocate the
+kernel if that's not the case.
+
+Hardware description
+--------------------
+
+The firmware can pass either a devicetree or ACPI tables to the RISC-V kernel.
+
+The devicetree is either passed directly to the kernel from the previous stage
+using the `$a1` register, or when booting with UEFI, it can be passed using the
+EFI configuration table.
+
+The ACPI tables are passed to the kernel using the EFI configuration table. In
+this case, a tiny devicetree is still created by the EFI stub. Please refer to
+"EFI stub and devicetree" tree section below for details about this devicetree.
+
+Kernel entrance
+---------------
+
+On SMP systems, there are 2 methods to enter the kernel:
+
+- `RISCV_BOOT_SPINWAIT`: the firmware releases all harts in the kernel, one hart
+  wins a lottery and executes the early boot code while the other harts are
+  parked waiting for the initialization to finish. This method is mostly used to
+  support older firmwares without SBI HSM extension and M-mode RISC-V kernel.
+- `Ordered booting`: the firmware releases only one hart that will execute the
+  initialization phase and then will start all other harts using the SBI HSM
+  extension. The ordered booting method is the preferred booting method for
+  booting the RISC-V kernel because it can support cpu hotplug and kexec.
+
+UEFI
+----
+
+UEFI memory map
+~~~~~~~~~~~~~~~
+
+When booting with UEFI, the RISC-V kernel will use only the EFI memory map to
+populate the system memory.
+
+The UEFI firmware must parse the subnodes of the `/reserved-memory` devicetree
+node and abide by the devicetree specification to convert the attributes of
+those subnodes (`no-map` and `reusable`) into their correct EFI equivalent
+(refer to section "3.5.4 /reserved-memory and UEFI" of the devicetree
+specification v0.4-rc1).
+
+RISCV_EFI_BOOT_PROTOCOL
+~~~~~~~~~~~~~~~~~~~~~~~
+
+When booting with UEFI, the EFI stub requires the boot hartid in order to pass
+it to the RISC-V kernel in `$a1`. The EFI stub retrieves the boot hartid using
+one of the following methods:
+
+- `RISCV_EFI_BOOT_PROTOCOL` (**preferred**).
+- `boot-hartid` devicetree subnode (**deprecated**).
+
+Any new firmware must implement `RISCV_EFI_BOOT_PROTOCOL` as the devicetree
+based approach is deprecated now.
+
+Early Boot Requirements and Constraints
+=======================================
+
+The RISC-V kernel's early boot process operates under the following constraints:
+
+EFI stub and devicetree
+-----------------------
+
+When booting with UEFI, the devicetree is supplemented (or created) by the EFI
+stub with the same parameters as arm64 which are described at the paragraph
+"UEFI kernel support on ARM" in Documentation/arm/uefi.rst.
+
+Virtual mapping installation
+----------------------------
+
+The installation of the virtual mapping is done in 2 steps in the RISC-V kernel:
+
+1. :c:func:`setup_vm` installs a temporary kernel mapping in
+   :c:var:`early_pg_dir` which allows discovery of the system memory. Only the
+   kernel text/data are mapped at this point. When establishing this mapping, no
+   allocation can be done (since the system memory is not known yet), so
+   :c:var:`early_pg_dir` page table is statically allocated (using only one
+   table for each level).
+
+2. :c:func:`setup_vm_final` creates the final kernel mapping in
+   :c:var:`swapper_pg_dir` and takes advantage of the discovered system memory
+   to create the linear mapping. When establishing this mapping, the kernel
+   can allocate memory but cannot access it directly (since the direct mapping
+   is not present yet), so it uses temporary mappings in the fixmap region to
+   be able to access the newly allocated page table levels.
+
+For :c:func:`virt_to_phys` and :c:func:`phys_to_virt` to be able to correctly
+convert direct mapping addresses to physical addresses, they need to know the
+start of the DRAM. This happens after step 1, right before step 2 installs the
+direct mapping (see :c:func:`setup_bootmem` function in arch/riscv/mm/init.c).
+Any usage of those macros before the final virtual mapping is installed must
+be carefully examined.
+
+Device-tree mapping via fixmap
+------------------------------
+
+The RISC-V kernel uses the fixmap region to map the devicetree because the
+devicetree virtual mapping must remain the same between :c:func:`setup_vm` and
+:c:func:`setup_vm_final` calls since the :c:var:`reserved_mem` array is
+initialized with virtual addresses established by :c:func:`setup_vm` and used
+with the mapping established by :c:func:`setup_vm_final`.
+
+Pre-MMU execution
+-----------------
+
+A few pieces of code need to run before even the first virtual mapping is
+established. These are the installation of the first virtual mapping itself,
+patching of early alternatives and the early parsing of the kernel command line.
+That code must be very carefully compiled as:
+
+- `-fno-pie`: This is needed for relocatable kernels which use `-fPIE`, since
+  otherwise, any access to a global symbol would go through the GOT which is
+  only relocated virtually.
+- `-mcmodel=medany`: Any access to a global symbol must be PC-relative to avoid
+  any relocations to happen before the MMU is setup.
+- *all* instrumentation must also be disabled (that includes KASAN, ftrace and
+  others).
+
+As using a symbol from a different compilation unit requires this unit to be
+compiled with those flags, we advise, as much as possible, not to use external
+symbols.
diff --git a/Documentation/riscv/index.rst b/Documentation/riscv/index.rst
index 175a91db0200..1f66062def6d 100644
--- a/Documentation/riscv/index.rst
+++ b/Documentation/riscv/index.rst
@@ -5,6 +5,7 @@  RISC-V architecture
 .. toctree::
     :maxdepth: 1
 
+    boot
     boot-image-header
     vm-layout
     hwprobe