path: root/src/hypervisor.tex

\chapter{Hypervisor Extension, Version 0.4}
\label{hypervisor}

{\bf Warning! This draft specification is likely to change before being
accepted as standard by the RISC-V Foundation.}

This chapter describes the RISC-V hypervisor extension, which virtualizes the
supervisor-level architecture to support the efficient hosting of guest
operating systems atop a type-1 or type-2 hypervisor.
The hypervisor extension changes supervisor mode into
{\em hypervisor-extended supervisor mode} (HS-mode, or {\em hypervisor
mode} for short), where a hypervisor or a hosting-capable operating system
runs.  The hypervisor extension also adds another level of address translation,
from {\em guest physical addresses} to supervisor physical addresses,
to virtualize the
memory and memory-mapped I/O subsystems for a guest operating system.  HS-mode
acts the same as S-mode, but with additional instructions and CSRs that control
the new level of address translation and support hosting a guest OS in virtual
S-mode (VS-mode).
Regular S-mode operating systems can execute without modification either in
HS-mode or as VS-mode guests.

In HS-mode, an OS or hypervisor interacts with the machine through the same
SBI as an OS normally does from S-mode.  An HS-mode hypervisor is expected to
implement the SBI for its VS-mode guest.

The hypervisor extension is enabled by setting bit 7 in the {\tt misa} CSR,
which corresponds to the letter H.  When {\tt misa}[7] is clear, the hart
behaves as though this extension were not implemented, and attempts to use
hypervisor CSRs or instructions raise an illegal instruction exception.
Implementations that include the hypervisor extension are encouraged
not to hardwire {\tt misa}[7], so that the extension may be disabled.

\begin{commentary}
This draft is based on earlier proposals by John Hauser and Paolo Bonzini.
\end{commentary}

\begin{commentary}
The baseline privileged architecture is designed to simplify the use of classic
virtualization techniques, where a guest OS is run at user-level, as
the few privileged instructions can be easily detected and trapped.
The hypervisor extension improves virtualization performance by
reducing the frequency of these traps.

The hypervisor extension has been designed to be efficiently
emulable on platforms that do not implement the extension, by running
the hypervisor in S-mode and trapping into M-mode for hypervisor CSR accesses
and to maintain shadow page tables.  The majority of CSR accesses for
type-2 hypervisors are valid S-mode accesses so need not be trapped.
Hypervisors can support nested virtualization analogously.
\end{commentary}

\section{Privilege Modes}

The current {\em virtualization mode}, denoted V, indicates whether the hart
is currently executing in a guest.
When V=1, the hart is either in virtual S-mode (VS-mode), or in virtual U-mode
(VU-mode) under a guest OS running in VS-mode.
When V=0, the hart is either in M-mode, in HS-mode, or in U-mode under an OS
running in HS-mode.
The virtualization mode also indicates whether two-level address translation
is active (V=1) or inactive (V=0).  Table~\ref{h-operating-modes} lists the
possible operating modes of a RISC-V hart with the hypervisor extension.

\begin{table*}[h!]
\begin{center}
\begin{tabular}{|c|c||l|l|l|}
  \hline
   Virtualization & Privilege & \multirow{2}{*}{Abbreviation} & \multirow{2}{*}{Name} & Two-Level \\
   Mode (V)       & Encoding  &                               &                       & Translation \\ \hline
   0              & 0         & U-mode  & User mode & Off \\
   0              & 1         & HS-mode & Hypervisor-extended supervisor mode & Off \\
   0              & 3         & M-mode  & Machine mode & Off \\
  \hline
   1              & 0         & VU-mode & Virtual user mode & On \\
   1              & 1         & VS-mode & Virtual supervisor mode & On \\
  \hline
 \end{tabular}
\end{center}
\caption{Operating modes with the hypervisor extension.}
\label{h-operating-modes}
\end{table*}

\section{Hypervisor and Virtual Supervisor CSRs}

An OS or hypervisor running in HS-mode uses the supervisor CSRs to interact with the exception,
interrupt, and address-translation subsystems.
Additional CSRs are provided to HS-mode, but not to VS-mode, to manage
two-level address translation and to control the behavior of a VS-mode guest:
{\tt hstatus}, {\tt hedeleg}, {\tt hideleg}, {\tt hcounteren}, and
{\tt hgatp}.

Furthermore, several {\em virtual supervisor} CSRs (VS CSRs) are replicas
of the normal supervisor CSRs.
For example, {\tt vsstatus} is the VS CSR that duplicates the usual
{\tt sstatus} CSR.

When V=1, the VS CSRs substitute for the corresponding supervisor CSRs,
taking over all functions of the usual supervisor CSRs except as specified
otherwise.
Instructions that normally read or modify a supervisor CSR shall instead
access the corresponding VS CSR.
In VS-mode, an attempt to read or write a VS CSR directly by its own
separate CSR address causes an illegal instruction exception.
The VS CSRs can be directly accessed only from M-mode or HS-mode.

While V=1, the normal HS-level supervisor CSRs retain their values but do
not affect the behavior of the machine unless specifically documented to
do so.
Conversely, when V=0, the VS CSRs do not ordinarily affect the behavior of
the machine other than being readable and writable by CSR instructions.

A few standard supervisor CSRs ({\tt scounteren} and, if the N extension
is implemented, {\tt sedeleg} and {\tt sideleg}) have no matching VS CSR.
These supervisor CSRs continue to have their usual function and
accessibility even when V=1, except with VS-mode and VU-mode substituting
for HS-mode and U-mode.
Hypervisor software is expected to manually swap the contents of these
registers as needed.

In this section, we use the term {\em HSXLEN} to refer to the effective XLEN
when executing in HS-mode, and {\em VSXLEN} to refer to the effective
XLEN when executing in VS-mode.

\subsection{Hypervisor Status Register ({\tt hstatus})}

The {\tt hstatus} register is an HSXLEN-bit read/write register
formatted as shown in Figure~\ref{hstatusreg-rv32} when HSXLEN=32 and
Figure~\ref{hstatusreg} when HSXLEN=64.
The {\tt hstatus}
register provides facilities analogous to the {\tt mstatus} register
that track and control the exception behavior of a VS-mode guest.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\begin{tabular}{YcccRcccccWc}
\\
\instbitrange{31}{23} &
\instbit{22} &
\instbit{21} &
\instbit{20} &
\instbitrange{19}{10} &
\instbit{9} &
\instbit{8} &
\instbit{7} &
\instbit{6} &
\instbit{5} &
\instbitrange{4}{1} &
\instbit{0} \\
\hline
\multicolumn{1}{|c|}{\wpri} &
\multicolumn{1}{c|}{VTSR} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{VTVM} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SP2V} &
\multicolumn{1}{c|}{SP2P} &
\multicolumn{1}{c|}{SPV} &
\multicolumn{1}{c|}{STL} &
\multicolumn{1}{c|}{VSBE} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SPRV} \\
\hline
9 & 1 & 1 & 1 & 10 & 1 & 1 & 1 & 1 & 1 & 4 & 1 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Hypervisor-mode status register ({\tt hstatus}) for RV32.}
\label{hstatusreg-rv32}
\end{figure*}

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\begin{tabular}{MFScccc}
\\
\instbitrange{HSXLEN-1}{34} &
\instbitrange{33}{32} &
\instbitrange{31}{23} &
\instbit{22} &
\instbit{21} &
\instbit{20} &
 \\
\hline
\multicolumn{1}{|c|}{\wpri} &
\multicolumn{1}{c|}{VSXL[1:0]} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{VTSR} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{VTVM} &
 \\
\hline
HSXLEN-34 & 2 & 9 & 1 & 1 & 1 & \\
\end{tabular}
\begin{tabular}{cOcccccFc}
\\
&
\instbitrange{19}{10} &
\instbit{9} &
\instbit{8} &
\instbit{7} &
\instbit{6} &
\instbit{5} &
\instbitrange{4}{1} &
\instbit{0} \\
\hline
 &
\multicolumn{1}{|c|}{\wpri} &
\multicolumn{1}{c|}{SP2V} &
\multicolumn{1}{c|}{SP2P} &
\multicolumn{1}{c|}{SPV} &
\multicolumn{1}{c|}{STL} &
\multicolumn{1}{c|}{VSBE} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SPRV} \\
\hline
 & 10 & 1 & 1 & 1 & 1 & 1 & 4 & 1 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Hypervisor-mode status register ({\tt hstatus}) for RV64.}
\label{hstatusreg}
\end{figure*}

The VSXL field controls the effective XLEN for VS-mode (known as VSXLEN),
which may differ from the XLEN for HS-mode (HSXLEN).
When HSXLEN=32, the VSXL field does not exist, and VSXLEN=32.
When HSXLEN=64, VSXL is a \warl\ field that is encoded the same as the
MXL field of {\tt misa}, shown in Table~\ref{misabase} on
page~\pageref{misabase}.
In particular, the implementation may hardwire VSXL so that VSXLEN=HSXLEN.

If HSXLEN is changed from 32 to a wider width, and if field VSXL is not
hardwired to a forced value, it gets the value corresponding to the
widest supported width not wider than the new HSXLEN.

The {\tt hstatus} fields VTSR and VTVM are defined analogously to the
{\tt mstatus} fields TSR and TVM, but affect the trapping behavior of the SRET
and virtual-memory management instructions only when V=1.

The SPV bit (Supervisor Previous Virtualization Mode) is written by the implementation
whenever a trap is taken into HS-mode.  Just as the SPP bit in {\tt sstatus} is set to the privilege
mode at the time of the trap, the SPV bit in {\tt hstatus} is set to the value of the virtualization
mode V at the time of the trap.  When an SRET instruction is executed when V=0,
V is set to SPV.

When a trap is taken into HS-mode, bits SP2V and SP2P are set to the values
that SPV and the HS-level SPP had before the trap.
When an SRET instruction is executed when V=0, the reverse assignments occur:
after SPV and {\tt sstatus}.SPP have supplied the new virtualization and
privilege modes, they are written with SP2V and SP2P, respectively.

The STL bit (Supervisor Translation Level), which indicates which address-translation level
caused an access-fault or page-fault exception, is also written by the implementation whenever a trap
is taken into HS-mode.
On an access or page fault due to guest physical address translation, STL is
set to 1.
For any other trap into HS-mode, STL is set to 0.

The VSBE bit is a \warl\ field that controls the endianness of explicit
memory accesses made from VS-mode.
If VSBE=0, explicit load and store memory accesses made from VS-mode are
little-endian, and if VSBE=1, they are big-endian.
An implementation may hardwire VSBE to specify always the same endianness
as for HS-mode.

The SPRV bit modifies the privilege with which loads and stores execute
in HS-mode.
When SPRV=0, translation and protection behave as normal.
When SPRV=1, load and store memory addresses are translated and protected, and
endianness is applied, as though the current virtualization mode were set to
{\tt hstatus}.SPV and the current privilege mode were set to the HS-level SPP.
Table~\ref{h-sprv} enumerates the cases.

\begin{table*}[h!]
\begin{center}
\begin{tabular}{|c|c|c||p{4.7in}|}
  \hline
   SPRV & SPV & SPP & Effect \\ \hline \hline
   0    & --  & --  & Normal access; current privilege and virtualization modes apply. \\ \hline
   1    & 0   & 0   & U-level access with HS-level translation and protection only. \\ \hline
   1    & 0   & 1   & HS-level access with HS-level translation and protection only. \\ \hline
   1    & 1   & 0   & VU-level access with two-level translation and protection. The HS-level MXR bit makes any executable page readable.  {\tt vsstatus}.MXR makes readable those pages marked executable at the VS translation level, but only if readable at the guest-physical translation level. \\ \hline
   1    & 1   & 1   & VS-level access with two-level translation and protection. The HS-level MXR bit makes any executable page readable.  {\tt vsstatus}.MXR makes readable those pages marked executable at the VS translation level, but only if readable at the guest-physical translation level.  {\tt vsstatus}.SUM applies instead of the HS-level SUM bit. \\ \hline
 \end{tabular}
\end{center}
\caption{Effect of SPRV on load and store translation and protection.}
\label{h-sprv}
\end{table*}

An MRET or SRET instruction that changes the operating mode to U-mode,
VS-mode, or VU-mode also sets SPRV=0.

\subsection{Hypervisor Trap Delegation Registers ({\tt hedeleg} and {\tt hideleg})}

By default, all traps at any privilege level are handled in M-mode, though
M-mode usually uses the {\tt medeleg} and {\tt mideleg} CSRs to delegate
some traps to HS-mode.  The {\tt hedeleg} and {\tt hideleg} CSRs allow these
traps to be further delegated to a VS-mode guest; their layout is the same
as {\tt medeleg} and {\tt mideleg}.

\begin{figure}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}U}
\instbitrange{HSXLEN-1}{0} \\
\hline
\multicolumn{1}{|c|}{Synchronous Exceptions (\warl)} \\
\hline
HSXLEN \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Hypervisor Exception Delegation Register {\tt hedeleg}.}
\label{hedelegreg}
\end{figure}

\begin{figure}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}U}
\instbitrange{HSXLEN-1}{0} \\
\hline
\multicolumn{1}{|c|}{Interrupts (\warl)} \\
\hline
HSXLEN \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Hypervisor Interrupt Delegation Register {\tt hideleg}.}
\label{hidelegreg}
\end{figure}

The {\tt hedeleg} and {\tt hideleg} registers are only active when V=1.  When
V=1, a trap that has been delegated to HS-mode (using {\tt medeleg} or {\tt
mideleg}) is further delegated to VS-mode if the corresponding {\tt hedeleg} or
{\tt hideleg} bit is set.

When an access-fault or page-fault exception is caused by guest physical
address translation, the trap is not delegated beyond HS-mode, regardless of
the setting of {\tt hedeleg}.

\subsection{Hypervisor Counter-Enable Register ({\tt hcounteren})}

The counter-enable register {\tt hcounteren} is a 32-bit register that
controls the availability of the hardware performance monitoring counters
to the guest virtual machine.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\begin{tabular}{cccMcccccc}
\instbit{31} &
\instbit{30} &
\instbit{29} &
\instbitrange{28}{6} &
\instbit{5} &
\instbit{4} &
\instbit{3} &
\instbit{2} &
\instbit{1} &
\instbit{0} \\
\hline
\multicolumn{1}{|c|}{HPM31} &
\multicolumn{1}{c|}{HPM30} &
\multicolumn{1}{c|}{HPM29} &
\multicolumn{1}{c|}{...} &
\multicolumn{1}{c|}{HPM5} &
\multicolumn{1}{c|}{HPM4} &
\multicolumn{1}{c|}{HPM3} &
\multicolumn{1}{c|}{IR} &
\multicolumn{1}{c|}{TM} &
\multicolumn{1}{c|}{CY} \\
\hline
1 & 1 & 1 & 23 & 1 & 1 & 1 & 1 & 1 & 1 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Hypervisor counter-enable register ({\tt hcounteren}).}
\label{hcounteren}
\end{figure*}

When the CY, TM, IR, or HPM{\em n} bit in the {\tt hcounteren} register
is clear, attempts to read the {\tt cycle}, {\tt time}, {\tt instret}, or
{\tt hpmcounter}{\em n} register while V=1 will cause an illegal
instruction exception.
When one of these bits is set, access to the corresponding register is
permitted when V=1, unless prevented for some other reason.
In VU-mode, a counter is not readable unless the applicable bits are set
in both {\tt hcounteren} and {\tt scounteren}.

{\tt hcounteren} must be implemented.
However, any of the bits may contain a hardwired value of zero,
indicating reads to the corresponding counter will cause an exception
when V=1.
Hence, they are effectively \warl\ fields.

\subsection{Hypervisor Time Delta Registers ({\tt htimedelta}, {\tt htimedeltah})}

The {\tt htimedelta} CSR is a read/write register that contains the delta
between the value of the {\tt time} CSR and the value returned in VS-mode or
VU-mode.
That is, reading the {\tt time} CSR in VS or VU mode returns the sum of the
contents of {\tt htimedelta} and the actual value of {\tt time}.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}J}
\instbitrange{63}{0} \\
\hline
\multicolumn{1}{|c|}{\tt htimedelta} \\
\hline
64 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Hypervisor time delta register, HSXLEN=64.}
\label{hdeltareg}
\end{figure*}

For HSXLEN=32 only, {\tt htimedelta} holds the lower 32 bits of the
delta, and {\tt htimedeltah} holds the upper 32 bits of the delta.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}J}
\instbitrange{31}{0} \\
\hline
\multicolumn{1}{|c|}{\tt htimedelta} \\
\hline
\multicolumn{1}{|c|}{\tt htimedeltah} \\
\hline
32 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Hypervisor time delta registers, HSXLEN=32.}
\label{hdeltahreg}
\end{figure*}

\subsection{Hypervisor Guest Address Translation and Protection Register ({\tt hgatp})}
\label{sec:hgatp}

The {\tt hgatp} register is an HSXLEN-bit read/write register, formatted as
shown in Figure~\ref{rv32hgatp} for HSXLEN=32 and Figure~\ref{rv64hgatp} for
HSXLEN=64, which controls guest physical address translation and protection.
Similar to CSR {\tt satp}, this register holds the physical page number (PPN)
of the guest-physical root page table; a virtual machine identifier (VMID),
which facilitates address-translation fences on a per-virtual-machine basis;
and the MODE field, which selects the address-translation scheme for guest
physical addresses.
When {\tt mstatus}.TVM=1, attempts to read or write {\tt hgatp} while executing
in HS-mode will raise an illegal instruction exception.

\begin{figure}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{cY@{}E@{}K}
\instbit{31} &
\instbitrange{30}{29} &
\instbitrange{28}{22} &
\instbitrange{21}{0} \\
\hline
\multicolumn{1}{|c|}{MODE} &
\multicolumn{1}{c|}{0 (\warl)} &
\multicolumn{1}{c|}{VMID (\warl)} &
\multicolumn{1}{c|}{PPN  (\warl)} \\
\hline
1 & 2 & 7 & 22 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{RV32 Hypervisor guest address translation and protection register
{\tt hgatp}.}
\label{rv32hgatp}
\end{figure}

\begin{figure}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}S@{}Y@{}E@{}K}
\instbitrange{63}{60} &
\instbitrange{59}{58} &
\instbitrange{57}{44} &
\instbitrange{43}{0} \\
\hline
\multicolumn{1}{|c|}{MODE (\warl)} &
\multicolumn{1}{c|}{0 (\warl)} &
\multicolumn{1}{c|}{VMID (\warl)} &
\multicolumn{1}{c|}{PPN  (\warl)} \\
\hline
4 & 2 & 14 & 44 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{RV64 Hypervisor guest address translation and protection register
{\tt hgatp}, for MODE values Bare, Sv39x4, and Sv48x4.}
\label{rv64hgatp}
\end{figure}

Table~\ref{tab:hgatp-mode} shows the encodings of the MODE field for RV32 and
RV64.
When MODE=Bare, guest physical addresses are equal to supervisor physical
addresses, and there is no further memory protection for a guest virtual
machine beyond the physical memory protection scheme described in
Section~\ref{sec:pmp}.
In this case, the remaining fields in {\tt hgatp} have no effect.

For RV32, the only other valid setting for MODE is Sv32x4, which is a
modification of the usual Sv32 paged virtual-memory scheme, extended to support
34-bit guest physical addresses.
For RV64, modes Sv39x4 and Sv48x4 are defined as modifications of the Sv39 and
Sv48 paged virtual-memory schemes.
All of these paged virtual-memory schemes are described in
Section~\ref{sec:guest-addr-translation}.
An additional RV64 scheme, Sv57x4, may be defined in a later version of this
specification.

The remaining MODE settings for RV64 are reserved for future use and may define
different interpretations of the other fields in {\tt hgatp}.

\begin{table}[h]
\begin{center}
\begin{tabular}{|c|c|l|}
\hline
\multicolumn{3}{|c|}{RV32} \\
\hline
Value  & Name & Description \\
\hline
0      & Bare   & No translation or protection. \\
1      & Sv32x4 & Page-based 34-bit virtual addressing (2-bit extension of Sv32). \\
\hline \hline
\multicolumn{3}{|c|}{RV64} \\
\hline
Value  & Name & Description \\
\hline
0      & Bare   & No translation or protection. \\
1--7   & ---    & {\em Reserved} \\
8      & Sv39x4 & Page-based 41-bit virtual addressing (2-bit extension of Sv39). \\
9      & Sv48x4 & Page-based 50-bit virtual addressing (2-bit extension of Sv48). \\
10     & {\em Sv57x4} & {\em Reserved for page-based 59-bit virtual addressing.} \\
11--15 & ---    & {\em Reserved} \\
\hline
\end{tabular}
\end{center}
\caption{Encoding of {\tt hgatp} MODE field.}
\label{tab:hgatp-mode}
\end{table}

RV64 implementations are not required to support all defined RV64 MODE
settings.

A write to {\tt hgatp} with an unsupported MODE value is not ignored as it is
for {\tt satp}.
Instead, the fields of {\tt hgatp} are {\warl} in the normal way, when so
indicated.

As explained in Section~\ref{sec:guest-addr-translation}, for the paged
virtual-memory schemes (Sv32x4, Sv39x4, and Sv48x4), the root page table is
16~KiB and must be aligned to a 16-KiB boundary.
In these modes, the lowest two bits of the physical page number (PPN) in
{\tt hgatp} always read as zeros.
An implementation that supports only the defined paged virtual-memory schemes
and/or Bare may hardwire PPN[1:0] to zero.

The number of supervisor physical address bits is implementation-defined; any
unimplemented address bits are hardwired to zero in {\tt hgatp}.PPN.
The number of VMID bits is also implementation-defined and may be zero.
The number of implemented VMID bits, termed {\mbox {\em VMIDLEN}}, may be
determined by writing one to every bit position in the VMID field, then reading
back the value in {\tt hgatp} to see which bit positions in the VMID field hold
a one.
The least-significant bits of VMID are implemented first:
that is, if VMIDLEN~$>$~0, VMID[VMIDLEN-1:0] is writable.
The maximal value of VMIDLEN, termed VMIDMAX, is 7 for Sv32x4 or 14 for Sv39x4
and Sv48x4.

Note that writing {\tt hgatp} does not imply any ordering constraints between
page-table updates and subsequent guest physical address translations.
If the new virtual machine's guest physical page tables have been modified, it
may be necessary to execute an HFENCE.GVMA instruction
(see Section~\ref{sec:hfence.vma}) before or after writing {\tt hgatp}.

\subsection{Virtual Supervisor Status Register ({\tt vsstatus})}

The {\tt vsstatus} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt sstatus}, formatted as
shown in Figure~\ref{vsstatusreg} when VSXLEN=32 and
Figure~\ref{vsstatusreg} when VSXLEN=64.
When V=1, {\tt vsstatus} substitutes for the usual {\tt sstatus}, so
instructions that normally read or modify {\tt sstatus} actually access
{\tt vsstatus} instead.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\scalebox{0.95}{
\begin{tabular}{cWcccccWccccWcc}
\\
\instbit{31} &
\instbitrange{30}{20} &
\instbit{19} &
\instbit{18} &
\instbit{17} &
\instbitrange{16}{15} &
\instbitrange{14}{13} &
\instbitrange{12}{9} &
\instbit{8} &
\instbit{7} &
\instbit{6} &
\instbit{5} &
\instbitrange{4}{2} &
\instbit{1} &
\instbit{0} \\
\hline
\multicolumn{1}{|c|}{SD} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{MXR} &
\multicolumn{1}{c|}{SUM} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{XS[1:0]} &
\multicolumn{1}{c|}{FS[1:0]} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SPP} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{UBE} &
\multicolumn{1}{c|}{SPIE} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SIE}  &
\multicolumn{1}{c|}{\wpri} \\
\hline
1 & 11 & 1 & 1 & 1 & 2 & 2 & 4 & 1 & 1 & 1 & 1 & 3 & 1 & 1 \\
\end{tabular}}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor status register ({\tt vsstatus}) for RV32.}
\label{vsstatusreg-rv32}
\end{figure*}

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\begin{tabular}{cMFScccc}
\\
\instbit{VSXLEN-1} &
\instbitrange{VSXLEN-2}{34} &
\instbitrange{33}{32} &
\instbitrange{31}{20} &
\instbit{19} &
\instbit{18} &
\instbit{17} &
 \\
\hline
\multicolumn{1}{|c|}{SD} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{UXL[1:0]} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{MXR} &
\multicolumn{1}{c|}{SUM} &
\multicolumn{1}{c|}{\wpri} &
 \\
\hline
1 & VSXLEN-35 & 2 & 12 & 1 & 1 & 1 & \\
\end{tabular}
\begin{tabular}{cWWFccccWcc}
\\
&
\instbitrange{16}{15} &
\instbitrange{14}{13} &
\instbitrange{12}{9} &
\instbit{8} &
\instbit{7} &
\instbit{6} &
\instbit{5} &
\instbitrange{4}{2} &
\instbit{1} &
\instbit{0} \\
\hline
 &
\multicolumn{1}{|c|}{XS[1:0]} &
\multicolumn{1}{c|}{FS[1:0]} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SPP} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{UBE} &
\multicolumn{1}{c|}{SPIE} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SIE} &
\multicolumn{1}{c|}{\wpri} \\
\hline
 & 2 & 2 & 4 & 1 & 1 & 1 & 1 & 3 & 1 & 1 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor status register ({\tt vsstatus}) for RV64.}
\label{vsstatusreg}
\end{figure*}

The UXL field controls the effective XLEN for VU-mode, which may differ
from the XLEN for VS-mode (VSXLEN).
When VSXLEN=32, the UXL field does not exist, and VU-mode XLEN=32.
When VSXLEN=64, UXL is a \warl\ field that is encoded the same as the MXL
field of {\tt misa}, shown in Table~\ref{misabase} on
page~\pageref{misabase}.
In particular, the implementation may hardwire field UXL so that VU-mode
XLEN=VSXLEN.

If VSXLEN is changed from 32 to a wider width, and if field UXL is not
hardwired to a forced value, it gets the value corresponding to the
widest supported width not wider than the new VSXLEN.

When V=1, both {\tt vsstatus}.FS and the HS-level {\tt sstatus}.FS are in
effect.  Attempts
to execute a floating-point instruction when either field is 0 (Off) raise an
illegal-instruction exception.  Modifying the floating-point state when V=1
causes both fields to be set to 3 (Dirty).

\begin{commentary}
For a hypervisor to benefit from the extension context status, it must
have its own copy in the HS-level {\tt sstatus}, maintained independently
of a guest OS running in VS-mode.
While a version of the extension context status obviously must exist in
{\tt vsstatus} for VS-mode, a hypervisor cannot rely on this version
being maintained correctly, given that VS-level software can change
{\tt vsstatus}.FS arbitrarily.
If the HS-level {\tt sstatus}.FS were not independently active and
maintained by the hardware in parallel with {\tt vsstatus}.FS while V=1,
hypervisors would always be forced to conservatively swap all
floating-point state when context-switching between virtual machines.
\end{commentary}

Read-only fields SD and XS summarize the extension context status as it
is visible to VS-mode only.
For example, the value of the HS-level {\tt sstatus}.FS does not affect
{\tt vsstatus}.SD.

When V=0, {\tt vsstatus} does not directly affect the behavior of the machine,
unless the MPRV feature in the {\tt mstatus} register or the SPRV feature
in the {\tt hstatus} register is used to execute a load or store
{\em as though} V=1.

\subsection{Virtual Supervisor Interrupt Registers ({\tt vsip} and {\tt vsie})}

The {\tt vsip} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt sip}, formatted as shown
in Figure~\ref{vsipreg}.
The {\tt vsip} register indicates pending VS-level interrupts.

When V=1, {\tt vsip} substitutes for the usual {\tt sip}, so instructions
that normally read or modify {\tt sip} actually access {\tt vsip}
instead.
However, pending HS-level interrupts continue to be indicated in the
HS-level {\tt sip} register, not in {\tt vsip}.

\note{Need to describe how {\tt vsip}.SEIP interacts with PLIC.
Current thinking is that {\tt vsip}.SEIP should purely be a
read-write storage bit to emulate the PLIC for VS-mode; the PLIC should not be
wired into {\tt vsip}.SEIP.

A future revision of the PLIC is expected to provide direct support
for VS-level interrupts to reduce virtualization overhead.}

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\begin{tabular}{KcFcFcc}
\instbitrange{VSXLEN-1}{10} &
\instbit{9} &
\instbitrange{8}{6} &
\instbit{5} &
\instbitrange{4}{2} &
\instbit{1} &
\instbit{0} \\
\hline
\multicolumn{1}{|c|}{\wpri} &
\multicolumn{1}{c|}{SEIP} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{STIP} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SSIP} &
\multicolumn{1}{c|}{\wpri} \\
\hline
VSXLEN-10 & 1 & 3 & 1 & 3 & 1 & 1 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor interrupt-pending register ({\tt vsip}).}
\label{vsipreg}
\end{figure*}

The {\tt vsie} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt sie}, formatted as shown in
Figure~\ref{vsiereg}.
The {\tt vsie} register contains interrupt enable bits for VS-level
interrupts.

When V=1, {\tt vsie} substitutes for the usual {\tt sie}, so instructions
that normally read or modify {\tt sie} actually access {\tt vsie} instead.
However, the enables for HS-level interrupts continue to be provided by
the HS-level {\tt sie} register, not by {\tt vsie}.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\begin{tabular}{KcFcFcc}
\instbitrange{VSXLEN-1}{10} &
\instbit{9} &
\instbitrange{8}{6} &
\instbit{5} &
\instbitrange{4}{2} &
\instbit{1} &
\instbit{0} \\
\hline
\multicolumn{1}{|c|}{\wpri} &
\multicolumn{1}{c|}{SEIE} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{STIE} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SSIE} &
\multicolumn{1}{c|}{\wpri} \\
\hline
VSXLEN-10 & 1 & 3 & 1 & 3 & 1 & 1 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor interrupt-enable register ({\tt vsie}).}
\label{vsiereg}
\end{figure*}

\subsection{Virtual Supervisor Trap Vector Base Address Register ({\tt vstvec})}

The {\tt vstvec} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt stvec}, formatted as shown
in Figure~\ref{vstvecreg}.
When V=1, {\tt vstvec} substitutes for the usual {\tt stvec}, so
instructions that normally read or modify {\tt stvec} actually access
{\tt vstvec} instead.
When V=0, {\tt vstvec} does not directly affect the behavior of the
machine.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{J@{}R}
\instbitrange{VSXLEN-1}{2} &
\instbitrange{1}{0} \\
\hline
\multicolumn{1}{|c|}{BASE[VSXLEN-1:2] (\warl)} &
\multicolumn{1}{c|}{MODE (\warl)} \\
\hline
VSXLEN-2 & 2 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor trap vector base address register ({\tt vstvec}).}
\label{vstvecreg}
\end{figure*}

\subsection{Virtual Supervisor Scratch Register ({\tt vsscratch})}

The {\tt vsscratch} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt sscratch}, formatted as
shown in Figure~\ref{vsscratchreg}.
When V=1, {\tt vsscratch} substitutes for the usual {\tt sscratch}, so
instructions that normally read or modify {\tt sscratch} actually access
{\tt vsscratch} instead.
The contents of {\tt vsscratch} never directly affect the behavior of
the machine.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}J}
\instbitrange{VSXLEN-1}{0} \\
\hline
\multicolumn{1}{|c|}{\tt vsscratch} \\
\hline
VSXLEN \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor scratch register ({\tt vsscratch}).}
\label{vsscratchreg}
\end{figure*}

\subsection{Virtual Supervisor Exception Program Counter ({\tt vsepc})}

The {\tt vsepc} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt sepc}, formatted as shown
in Figure~\ref{vsepcreg}.
When V=1, {\tt vsepc} substitutes for the usual {\tt sepc}, so
instructions that normally read or modify {\tt sepc} actually access
{\tt vsepc} instead.
When V=0, {\tt vsepc} does not directly affect the behavior of the
machine.

{\tt vsepc} is a \warl\ register that must be able to hold the same set of
values that {\tt sepc} can hold.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}J}
\instbitrange{VSXLEN-1}{0} \\
\hline
\multicolumn{1}{|c|}{\tt vsepc} \\
\hline
VSXLEN \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor exception program counter ({\tt vsepc}).}
\label{vsepcreg}
\end{figure*}

\subsection{Virtual Supervisor Cause Register ({\tt vscause})}

The {\tt vscause} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt scause}, formatted as shown
in Figure~\ref{vscausereg}.
When V=1, {\tt vscause} substitutes for the usual {\tt scause}, so
instructions that normally read or modify {\tt scause} actually access
{\tt vscause} instead.
When V=0, {\tt vscause} does not directly affect the behavior of the
machine.

{\tt vscause} is a \wlrl\ register that must be able to hold the same set of
values that {\tt scause} can hold.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{c@{}U}
\instbit{VSXLEN-1} &
\instbitrange{VSXLEN-2}{0} \\
\hline
\multicolumn{1}{|c|}{Interrupt} &
\multicolumn{1}{c|}{Exception Code (\wlrl)} \\
\hline
1 & VSXLEN-1 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor cause register ({\tt vscause}).}
\label{vscausereg}
\end{figure*}

\subsection{Virtual Supervisor Trap Value Register ({\tt vstval})}

The {\tt vstval} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt stval}, formatted as shown
in Figure~\ref{vstvalreg}.
When V=1, {\tt vstval} substitutes for the usual {\tt stval}, so
instructions that normally read or modify {\tt stval} actually access
{\tt vstval} instead.
When V=0, {\tt vstval} does not directly affect the behavior of the
machine.

{\tt vstval} is a \warl\ register that must be able to hold the same set of
values that {\tt stval} can hold.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}J}
\instbitrange{VSXLEN-1}{0} \\
\hline
\multicolumn{1}{|c|}{\tt vstval} \\
\hline
VSXLEN \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Virtual supervisor trap value register ({\tt vstval}).}
\label{vstvalreg}
\end{figure*}

\subsection{Virtual Supervisor Address Translation and Protection Register ({\tt vsatp})}

The {\tt vsatp} register is a VSXLEN-bit read/write register that is
VS-mode's version of supervisor register {\tt satp}, formatted as shown
in Figure~\ref{rv32vsatpreg} for VSXLEN=32 and Figure~\ref{rv64vsatpreg}
for VSXLEN=64.
When V=1, {\tt vsatp} substitutes for the usual {\tt satp}, so
instructions that normally read or modify {\tt satp} actually access
{\tt vsatp} instead.
{\tt vsatp} controls VS-level address translation, the first stage of
two-level translation for guest virtual addresses (see
Section~\ref{sec:two-level-translation}).

When V=0, a write to {\tt vsatp} with an unsupported MODE value is not
ignored as it is for {\tt satp}.
Instead, the fields of {\tt vsatp} are {\warl} in the normal way.

When V=0, {\tt vsatp} does not directly affect the behavior of the machine,
unless the MPRV feature in the {\tt mstatus} register or the SPRV feature
in the {\tt hstatus} register is used to execute a load or store
{\em as though} V=1.

\begin{figure}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{c@{}E@{}K}
\instbit{31} &
\instbitrange{30}{22} &
\instbitrange{21}{0} \\
\hline
\multicolumn{1}{|c|}{MODE (\warl)} &
\multicolumn{1}{c|}{ASID (\warl)} &
\multicolumn{1}{c|}{PPN  (\warl)} \\
\hline
1 & 9 & 22 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{RV32 virtual supervisor address translation and protection register {\tt vsatp}.}
\label{rv32vsatpreg}
\end{figure}

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}S@{}T@{}U}
\instbitrange{63}{60} &
\instbitrange{59}{44} &
\instbitrange{43}{0} \\
\hline
\multicolumn{1}{|c|}{MODE (\warl)} &
\multicolumn{1}{c|}{ASID (\warl)} &
\multicolumn{1}{c|}{PPN  (\warl)} \\
\hline
4 & 16 & 44 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{RV64 virtual supervisor address translation and protection register {\tt vsatp}, for MODE
values Bare, Sv39, and Sv48.}
\label{rv64vsatpreg}
\end{figure*}

\section{Hypervisor Instructions}

The hypervisor extension adds two new privileged fence instructions.

\subsection{Hypervisor Memory-Management Fence Instructions}
\label{sec:hfence.vma}

\vspace{-0.2in}
\begin{center}
\begin{tabular}{O@{}R@{}R@{}F@{}R@{}S}
\\
\instbitrange{31}{25} &
\instbitrange{24}{20} &
\instbitrange{19}{15} &
\instbitrange{14}{12} &
\instbitrange{11}{7} &
\instbitrange{6}{0} \\
\hline
\multicolumn{1}{|c|}{funct7} &
\multicolumn{1}{c|}{rs2} &
\multicolumn{1}{c|}{rs1} &
\multicolumn{1}{c|}{funct3} &
\multicolumn{1}{c|}{rd} &
\multicolumn{1}{c|}{opcode} \\
\hline
7 & 5 & 5 & 3 & 5 & 7 \\
HFENCE.GVMA & vmid & gaddr & PRIV & 0 & SYSTEM \\
HFENCE.VVMA & asid & vaddr & PRIV & 0 & SYSTEM \\
\end{tabular}
\end{center}

The hypervisor memory-management fence instructions, HFENCE.GVMA and
HFENCE.VVMA, are valid only in HS-mode when {\tt mstatus}.TVM=0, or in M-mode
(irrespective of {\tt mstatus}.TVM).
These instructions perform a function similar to SFENCE.VMA
(Section~\ref{sec:sfence.vma}), except applying to the guest-physical
memory-management data structures controlled by CSR {\tt hgatp} (HFENCE.GVMA)
or the VS-level memory-management data structures controlled by CSR {\tt vsatp}
(HFENCE.VVMA).
Instruction SFENCE.VMA applies only to the memory-management data structures
controlled by the current {\tt satp} (either the HS-level {\tt satp} when
V=0 or {\tt vsatp} when V=1).

If an HFENCE.VVMA instruction executes without trapping, its effect is much the
same as temporarily entering VS-mode and executing SFENCE.VMA.
Executing an HFENCE.VVMA guarantees that any previous stores already visible
to the current hart are ordered before all subsequent implicit reads by that
hart of the VS-level memory-management data structures, when those implicit
reads are for instructions that
\begin{compactitem}
\item
are subsequent to the HFENCE.VVMA, and
\item
execute when {\tt hgatp}.VMID has the same setting as it did when HFENCE.VVMA
executed.
\end{compactitem}
Implicit reads need not be ordered when {\tt hgatp}.VMID is different than at
the time HFENCE.VVMA executed.
If operand {\em rs1}$\neq${\tt x0}, it specifies a single guest virtual
address, and if operand {\em rs2}$\neq${\tt x0}, it specifies a single guest
address-space identifier
(ASID).

\begin{commentary}
An HFENCE.VVMA instruction applies only to a single virtual machine, identified
by the setting of {\tt hgatp}.VMID when HFENCE.VVMA executes.
\end{commentary}

When {\em rs2}$\neq${\tt x0}, bits XLEN-1:ASIDMAX of the value held in {\em
rs2} are reserved for future use and should be zeroed by software and ignored
by current implementations.
Furthermore, if ASIDLEN~$<$~ASIDMAX, the implementation shall ignore bits
ASIDMAX-1:ASIDLEN of the value held in {\em rs2}.

\begin{commentary}
Simpler implementations of HFENCE.VVMA can ignore the guest virtual address in
{\em rs1} and the guest ASID value in {\em rs2}, as well as {\tt hgatp}.VMID,
and always perform a global fence for the VS-level memory management of all
virtual machines, or even a global fence for all memory-management data
structures.
\end{commentary}

Executing an HFENCE.GVMA instruction guarantees that any previous stores
already visible to the current hart are ordered before all subsequent implicit
reads by that hart of guest-physical memory-management data structures done for instructions
that follow the HFENCE.GVMA.
If operand {\em rs1}$\neq${\tt x0}, it specifies a single guest physical
address, shifted right by 2~bits, and if operand {\em rs2}$\neq${\tt x0}, it
specifies a single virtual machine identifier (VMID).

\begin{commentary}
For HFENCE.GVMA, a guest physical address specified in {\em rs1} is shifted
right by 2~bits to accommodate addresses wider than the current XLEN.
For RV32, the hypervisor extension permits guest physical addresses as wide as
34 bits, and {\em rs1} specifies bits 33:2 of such an address.
This shift-by-2 encoding of guest physical addresses matches the encoding of
physical addresses in PMP address registers (Section~\ref{sec:pmp}) and in page
table entries (Sections \ref{sec:sv32}, \ref{sec:sv39}, and~\ref{sec:sv48}).
\end{commentary}

When {\em rs2}$\neq${\tt x0}, bits XLEN-1:VMIDMAX of the value held in {\em
rs2} are reserved for future use and should be zeroed by software and ignored
by current implementations.
Furthermore, if VMIDLEN~$<$~VMIDMAX, the implementation shall ignore bits
VMIDMAX-1:VMIDLEN of the value held in {\em rs2}.

\begin{commentary}
Simpler implementations of HFENCE.GVMA can ignore the guest physical address in
{\em rs1} and the VMID value in {\em rs2} and always perform a global fence for
the guest-physical memory management of all virtual machines, or even a global
fence for all memory-management data structures.
\end{commentary}

\section{Machine-Level CSRs}

The hypervisor extension augments machine status registers {\tt mstatus}
and (for RV32 only) {\tt mstatush}.

\subsection{Machine Status Registers ({\tt mstatus} and {\tt mstatush})}

The hypervisor extension adds two fields, MPV and MTL, to the
machine-level {\tt mstatus} or {\tt mstatush} CSR, and modifies the
behavior of several existing {\tt mstatus} fields.
Figure~\ref{hypervisor-mstatus} shows the modified {\tt mstatus} register
when the hypervisor extension is provided and MXLEN=64.
When MXLEN=32, the hypervisor extension adds fields not to {\tt mstatus}
but to {\tt mstatush}, which must exist.
Figure~\ref{hypervisor-mstatush} shows the {\tt mstatush} register when
the hypervisor extension is provided and MXLEN=32.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\begin{tabular}{cMccccFFc}
\\
\instbit{MXLEN-1} &
\instbitrange{MXLEN-2}{40} &
\instbit{39} &
\instbit{38} &
\instbit{37} &
\instbit{36} &
\instbitrange{35}{34} &
\instbitrange{33}{32} &
 \\
\hline
\multicolumn{1}{|c|}{SD} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{MPV} &
\multicolumn{1}{c|}{MTL} &
\multicolumn{1}{c|}{MBE} &
\multicolumn{1}{c|}{SBE} &
\multicolumn{1}{c|}{SXL[1:0]} &
\multicolumn{1}{c|}{UXL[1:0]} &
 \\
\hline
1 & MXLEN-41 & 1 & 1 & 1 & 1 & 2 & 2 & \\
\end{tabular}
\begin{tabular}{cEccccccWWc}
\\
&
\instbitrange{31}{23} &
\instbit{22} &
\instbit{21} &
\instbit{20} &
\instbit{19} &
\instbit{18} &
\instbit{17} &
\instbitrange{16}{15} &
\instbitrange{14}{13} &
 \\
\hline
 &
\multicolumn{1}{|c|}{\wpri} &
\multicolumn{1}{c|}{TSR} &
\multicolumn{1}{c|}{TW} &
\multicolumn{1}{c|}{TVM} &
\multicolumn{1}{c|}{MXR} &
\multicolumn{1}{c|}{SUM} &
\multicolumn{1}{c|}{MPRV} &
\multicolumn{1}{c|}{XS[1:0]} &
\multicolumn{1}{c|}{FS[1:0]} &
 \\
\hline
 & 9 & 1 & 1 & 1 & 1 & 1 & 1 & 2 & 2 & \\
\end{tabular}
\begin{tabular}{cFWcccccccccc}
\\
&
\instbitrange{12}{11} &
\instbitrange{10}{9} &
\instbit{8} &
\instbit{7} &
\instbit{6} &
\instbit{5} &
\instbit{4} &
\instbit{3} &
\instbit{2} &
\instbit{1} &
\instbit{0} \\
\hline
 &
\multicolumn{1}{|c|}{MPP[1:0]} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SPP} &
\multicolumn{1}{c|}{MPIE} &
\multicolumn{1}{c|}{UBE} &
\multicolumn{1}{c|}{SPIE} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{MIE} &
\multicolumn{1}{c|}{\wpri} &
\multicolumn{1}{c|}{SIE} &
\multicolumn{1}{c|}{\wpri} \\
\hline
 & 2 & 2 & 1 & 1 & 1 & 1 & 1 & 1 & 1 & 1 & 1 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Machine-mode status register ({\tt mstatus}) for RV64 when the hypervisor extension is provided.}
\label{hypervisor-mstatus}
\end{figure*}

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\setlength{\tabcolsep}{4pt}
\begin{tabular}{LccccF}
\\
\instbitrange{31}{8} &
\instbit{7} &
\instbit{6} &
\instbit{5} &
\instbit{4} &
\instbitrange{3}{0} \\
\hline
\multicolumn{1}{|c|}{\wpri} &
\multicolumn{1}{c|}{MPV} &
\multicolumn{1}{c|}{MTL} &
\multicolumn{1}{c|}{MBE} &
\multicolumn{1}{c|}{SBE} &
\multicolumn{1}{c|}{\wpri} \\
\hline
24 & 1 & 1 & 1 & 1 & 4 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Additional machine-mode status register ({\tt mstatush}) for RV32 when the hypervisor extension is provided.
The format of {\tt mstatus} is unchanged for RV32.}
\label{hypervisor-mstatush}
\end{figure*}

The MPV bit (Machine Previous Virtualization Mode) is written by the implementation
whenever a trap is taken into M-mode.  Just as the MPP bit is set to the privilege
mode at the time of the trap, the MPV bit is set to the value of the virtualization
mode V at the time of the trap.  When an MRET instruction is executed, the
virtualization mode V is set to MPV, unless MPP=3, in which case V remains 0.

The MTL bit (Machine Translation Level), which indicates which address-translation level
caused an access-fault or page-fault exception, is also written by the implementation whenever a trap
is taken into M-mode.
On an access or page fault due to guest physical address translation, MTL is
set to 1.
For any other trap into M-mode, MTL is set to 0.

The TSR and TVM fields of {\tt mstatus} affect execution only in HS-mode,
not in VS-mode.
The TW field affects execution in all modes except M-mode.

The hypervisor extension changes the behavior of the the Modify Privilege
field, MPRV, of {\tt mstatus}.
When MPRV=0, translation and protection behave as normal.  When
MPRV=1, loads and stores are translated and protected as though the current
privilege mode were set to MPP and the current virtualization mode were set to
MPV.  Table~\ref{h-mprv} enumerates the cases.

\begin{table*}[h!]
\begin{center}
\begin{tabular}{|c|c|c||p{4.5in}|}
  \hline
   MPRV & MPV & MPP & Effect \\ \hline \hline
   0    & --  & --  & Normal access; current privilege and virtualization modes apply. \\ \hline
   1    & 0   & 0   & U-level access with HS-level translation and protection only. \\ \hline
   1    & 0   & 1   & HS-level access with HS-level translation and protection only.  \\ \hline
   1    & --  & 3   & M-level access with no translation. \\ \hline
   1    & 1   & 0   & VU-level access with two-level translation and protection. The HS-level MXR bit makes any executable page readable.  {\tt vsstatus}.MXR makes readable those pages marked executable at the VS translation level, but only if readable at the guest-physical translation level. \\ \hline
   1    & 1   & 1   & VS-level access with two-level translation and protection. The HS-level MXR bit makes any executable page readable.  {\tt vsstatus}.MXR makes readable those pages marked executable at the VS translation level, but only if readable at the guest-physical translation level.  {\tt vsstatus}.SUM applies instead of the HS-level SUM bit. \\ \hline
 \end{tabular}
\end{center}
\caption{Effect of MPRV on load and store translation and protection.  When MPRV=1, MPP$\neq$3, and {\tt hstatus}.SPRV=1, the effective privilege is further modified: {\tt hstatus}.SPV applies instead of MPV, and the HS-level SPP applies instead of MPP.}
\label{h-mprv}
\end{table*}

The {\tt mstatus} register is a superset of the HS-level {\tt sstatus}
register but is not a superset of {\tt vsstatus}.

\section{Two-Level Address Translation}
\label{sec:two-level-translation}

Whenever the current virtualization mode V is 1 (and assuming
{\tt mstatus}.MPRV=0), two-level address translation and protection is in
effect.
For any virtual memory access, the original virtual address is first converted
by VS-level address translation, as controlled by the {\tt vsatp}
register, into a {\em guest physical address}.
The guest physical address is then converted by guest physical address
translation, as controlled by the {\tt hgatp} register, into a supervisor
physical address.
Although there is no option to disable two-level address translation when V=1,
either level of translation can be effectively disabled by zeroing the
corresponding {\tt vsatp} or {\tt hgatp} register.

The {\tt vsstatus} field MXR, which makes execute-only pages readable, only
overrides VS-level page protection.
Setting MXR at VS-level does not override guest-physical page protections.
Setting MXR at HS-level, however, overrides both VS-level and guest-physical
execute-only permissions.

When V=1, memory accesses that would normally bypass address translation are
subject to guest physical address translation alone.
This includes memory accesses made in support of VS-level address translation,
such as reads and writes of VS-level page tables.

Machine-level physical memory protection applies to supervisor physical
addresses and is in effect regardless of virtualization mode.

\subsection{Guest Physical Address Translation}
\label{sec:guest-addr-translation}

The mapping of guest physical addresses to supervisor physical addresses is
controlled by CSR {\tt hgatp} (Section~\ref{sec:hgatp}).

When the address translation scheme selected by the MODE field of {\tt hgatp}
is Bare, guest physical addresses are equal to supervisor physical addresses
without modification, and no memory protection applies in the trivial
translation of guest physical addresses to supervisor physical addresses.

When {\tt hgatp}.MODE specifies a translation scheme of Sv32x4, Sv39x4, or
Sv48x4, guest physical address translation is a variation on the usual
page-based virtual address translation scheme of Sv32, Sv39, or Sv48,
respectively.
In each case, the size of the incoming address is widened by 2~bits (to 34, 41,
or 50 bits).
To accommodate the 2~extra bits, the root page table (only) is expanded by a
factor of four to be 16~KiB instead of the usual 4~KiB.
Matching its larger size, the root page table also must be aligned to a 16~KiB
boundary instead of the usual 4~KiB page boundary.
Except as noted, all other aspects of Sv32, Sv39, or Sv48 are adopted unchanged
for guest physical address translation.
Non-root page tables and all page table entries (PTEs) have the same formats as
documented in Sections \ref{sec:sv32}, \ref{sec:sv39}, and~\ref{sec:sv48}.

For Sv32x4, an incoming guest physical address is partitioned into a virtual
page number (VPN) and page offset as shown in Figure~\ref{sv32x4va}.
This partitioning is identical to that for an Sv32 virtual address as depicted
in Figure~\ref{sv32va} (page~\pageref{sv32va}), except with 2 more bits at the
high end in VPN[1].
(Note that the fields of a partitioned guest physical address also correspond
one-for-one with the structure that Sv32 assigns to a physical address,
depicted in Figure~\ref{rv32va}.)

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}E@{}O@{}E}
\instbitrange{33}{22} &
\instbitrange{21}{12} &
\instbitrange{11}{0} \\
\hline
\multicolumn{1}{|c|}{VPN[1]} &
\multicolumn{1}{c|}{VPN[0]} &
\multicolumn{1}{c|}{page offset} \\
\hline
12 & 10 & 12 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Sv32x4 virtual address (guest physical address).}
\label{sv32x4va}
\end{figure*}

For Sv39x4, an incoming guest physical address is partitioned as shown in
Figure~\ref{sv39x4va}.
This partitioning is identical to that for an Sv39 virtual address as depicted
in Figure~\ref{sv39va} (page~\pageref{sv39va}), except with 2 more bits at the
high end in VPN[2].
Address bits 63:41 must all be zeros, or else a page-fault exception occurs,
attributed to guest physical address translation.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}E@{}O@{}O@{}O}
\instbitrange{40}{30} &
\instbitrange{29}{21} &
\instbitrange{20}{12} &
\instbitrange{11}{0} \\
\hline
\multicolumn{1}{|c|}{VPN[2]} &
\multicolumn{1}{c|}{VPN[1]} &
\multicolumn{1}{c|}{VPN[0]} &
\multicolumn{1}{c|}{page offset} \\
\hline
11 & 9 & 9 & 12 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Sv39x4 virtual address (guest physical address).}
\label{sv39x4va}
\end{figure*}

For Sv48x4, an incoming guest physical address is partitioned as shown in
Figure~\ref{sv48x4va}.
This partitioning is identical to that for an Sv48 virtual address as depicted
in Figure~\ref{sv48va} (page~\pageref{sv48va}), except with 2 more bits at the
high end in VPN[3].
Address bits 63:50 must all be zeros, or else a page-fault exception occurs,
attributed to guest physical address translation.

\begin{figure*}[h!]
{\footnotesize
\begin{center}
\begin{tabular}{@{}E@{}O@{}O@{}O@{}O}
\instbitrange{49}{39} &
\instbitrange{38}{30} &
\instbitrange{29}{21} &
\instbitrange{20}{12} &
\instbitrange{11}{0} \\
\hline
\multicolumn{1}{|c|}{VPN[3]} &
\multicolumn{1}{c|}{VPN[2]} &
\multicolumn{1}{c|}{VPN[1]} &
\multicolumn{1}{c|}{VPN[0]} &
\multicolumn{1}{c|}{page offset} \\
\hline
11 & 9 & 9 & 9 & 12 \\
\end{tabular}
\end{center}
}
\vspace{-0.1in}
\caption{Sv48x4 virtual address (guest physical address).}
\label{sv48x4va}
\end{figure*}

\begin{commentary}
The page-based guest physical address translation scheme for RV32, Sv32x4, is
defined to support a 34-bit guest physical address so that an RV32 hypervisor
need not be limited in its ability to virtualize real 32-bit RISC-V machines,
even those with 33-bit or 34-bit physical addresses.
This may include the possibility of a machine virtualizing itself, if it
happens to use 33-bit or 34-bit physical addresses.
Multiplying the size and alignment of the root page table by a factor of four
is the cheapest way to extend Sv32 to cover a 34-bit address.
The possible wastage of 12~KiB for an unnecessarily large root page table is
expected to be of negligible consequence for most (maybe all) real uses.

A consistent ability to virtualize machines having as much as four times the
physical address space as virtual address space is believed to be of some
utility also for RV64.
For a machine supporting 39-bit virtual addresses (Sv39), for example, this
allows the hypervisor extension to support up to a 41-bit guest physical
address space without either necessitating hardware support for 48-bit virtual
addresses (Sv48) or falling back to emulating the larger address space with
shadow page tables.
\end{commentary}

The conversion of an Sv32x4, Sv39x4, or Sv48x4 guest physical address is
accomplished with the same algorithm used for Sv32, Sv39, or Sv48, as presented
in Section~\ref{sv32algorithm}, except that:
\begin{compactitem}
\item
in step~1, $a = \mbox{{\tt hgatp}.PPN}\times\mbox{PAGESIZE}$; and
\item
the current privilege mode is always taken to be U-mode.
\end{compactitem}

For guest physical address translation, all memory accesses (including those
made to access data structures for VS-level address translation) are considered
to be user-level accesses, as though executed in U-mode.
Access type permissions---readable, writable, or executable---are checked
during guest physical address translation the same as for VS-level address
translation.
For a memory access made to support VS-level address translation (such as to
read/write a VS-level page table), permissions are checked as though for a load
or store, not for the original access type.
However, any exception is always reported for the original access type
(instruction, load, or store/AMO).

Access faults and page faults raised by guest physical address translation are
treated as HS-level exceptions for the purpose of exception delegation, so are
not delegated to VS-mode, regardless of the setting of the {\tt hedeleg}
register.

\subsection{Memory-Management Fences}

The behavior of the SFENCE.VMA instruction is affected by the current
virtualization mode V.  When V=0, the virtual-address argument is an HS-level
virtual address, and the ASID argument is an HS-level ASID.
The instruction orders stores only to HS-level address-translation structures
with subsequent HS-level address translations.

When V=1, the virtual-address argument to SFENCE.VMA is a guest virtual
address within the current virtual machine, and the ASID argument is a VS-level
ASID within the current virtual machine.
The current virtual machine is identified by the VMID field of CSR {\tt hgatp},
and the effective ASID can be considered to be the combination of this VMID
with the VS-level ASID.
The SFENCE.VMA instruction orders stores only to the VS-level
address-translation structures with subsequent VS-level address translations
for the same virtual machine, i.e., only when {\tt hgatp}.VMID is the same as
when the SFENCE.VMA executed.

Hypervisor instructions HFENCE.GVMA and HFENCE.VVMA provide additional
memory-management fences to complement SFENCE.VMA.
These instructions are described in Section~\ref{sec:hfence.vma}.

Section~\ref{pmp-vmem} discusses the intersection between physical memory
protection (PMP) and page-based address translation.
It is noted there that, when PMP settings are modified in a manner that affects
either the physical memory that holds page tables or the physical memory to
which page tables point, M-mode software must synchronize the PMP settings with
the virtual memory system.
For HS-level address translation, this is accomplished by executing in M-mode
an SFENCE.VMA instruction with {\em rs1}={\tt x0} and {\em rs2}={\tt x0}, after
the PMP CSRs are written.
If guest physical address translation is in use, synchronization with its data
structures is also needed.
When PMP settings are modified in a manner that affects either the physical
memory that holds guest-physical page tables or the physical memory to which
guest-physical page tables point, an HFENCE.GVMA instruction with
{\em rs1}={\tt x0} and {\em rs2}={\tt x0} must be executed in M-mode after the
PMP CSRs are written.
An HFENCE.VVMA instruction is not required.

\section{WFI in Virtual Operating Modes}

Executing instruction WFI in VS-mode or VU-mode causes an illegal instruction
exception, unless it completes within an implementation-specific, bounded time
limit.

\begin{commentary}
The behavior required of WFI in VS-mode and VU-mode is the same as required of
it in U-mode when S-mode exists.
\end{commentary}

\section{Traps}

The hypervisor extension augments the environment-call exception cause
encoding.  Environment calls from HS-mode use cause 9, whereas environment
calls from VS-mode use cause 10.  Table~\ref{hcauses} lists the
possible M-mode and HS-mode exception codes when the hypervisor extension is
present.

\begin{table*}[p]
\begin{center}
\begin{tabular}{|r|r|l|l|}

  \hline
  Interrupt & Exception Code  & Description \\
  \hline
  1         & 0               & {\em Reserved} \\
  1         & 1               & Supervisor software interrupt \\
  1         & 2               & {\em Reserved} \\
  1         & 3               & Machine software interrupt \\ \hline
  1         & 4               & {\em Reserved} \\
  1         & 5               & Supervisor timer interrupt \\
  1         & 6               & {\em Reserved} \\
  1         & 7               & Machine timer interrupt \\ \hline
  1         & 8               & {\em Reserved} \\
  1         & 9               & Supervisor external interrupt \\
  1         & 10              & {\em Reserved} \\
  1         & 11              & Machine external interrupt \\ \hline
  1         & 12--15          & {\em Reserved} \\
  1         & $\ge$16         & {\em Available for platform use} \\ \hline
  0         & 0               & Instruction address misaligned \\
  0         & 1               & Instruction access fault \\
  0         & 2               & Illegal instruction \\
  0         & 3               & Breakpoint \\
  0         & 4               & Load address misaligned \\
  0         & 5               & Load access fault \\
  0         & 6               & Store/AMO address misaligned \\
  0         & 7               & Store/AMO access fault \\
  0         & 8               & Environment call from U-mode or VU-mode \\
  0         & 9               & Environment call from HS-mode \\
  0         & 10              & Environment call from VS-mode \\
  0         & 11              & Environment call from M-mode \\
  0         & 12              & Instruction page fault \\
  0         & 13              & Load page fault \\
  0         & 14              & {\em Reserved} \\
  0         & 15              & Store/AMO page fault \\
  0         & 16--23          & {\em Reserved} \\
  0         & 24--31          & {\em Available for custom use} \\
  0         & 32--47          & {\em Reserved} \\
  0         & 48--63          & {\em Available for custom use} \\
  0         & $\ge$64         & {\em Reserved} \\
  \hline
\end{tabular}
\end{center}
\caption{Supervisor and machine cause register ({\tt scause} and {\tt mcause}) values when the hypervisor extension is enabled.}
\label{hcauses}
\end{table*}

\begin{commentary}
HS-mode and VS-mode ECALLs use different cause values so they can be delegated
separately.
\end{commentary}

When a trap occurs in HS-mode or U-mode, it goes to M-mode, unless
delegated by {\tt medeleg} or {\tt mideleg}, in which case it goes to HS-mode.
When a trap occurs in VS-mode or VU-mode, it goes to M-mode, unless
delegated by {\tt medeleg} or {\tt mideleg}, in which case it goes to HS-mode,
unless further delegated by {\tt hedeleg} or {\tt hideleg}, in which case it
goes to VS-mode.

When a trap is taken into M-mode, virtualization mode V gets set to~0,
and {\tt mstatus}.MPV and {\tt mstatus}.MPP are set according to
Table~\ref{h-mpp}.

\begin{table*}[h!]
\begin{center}
\begin{tabular}{|l|c|c|}
  \hline
  Previous Mode & MPV & MPP \\ \hline
  U-mode        & 0   & 0   \\
  HS-mode       & 0   & 1   \\
  M-mode        & 0   & 3   \\ \hline
  VU-mode       & 1   & 0   \\
  VS-mode       & 1   & 1   \\ \hline
\end{tabular}
\end{center}
\caption{Value of {\tt mstatus} fields MPV and MPP after a trap into M-mode.
Upon trap return, MPV is ignored when MPP=3.}
\label{h-mpp}
\end{table*}

When a trap is taken into HS-mode, virtualization mode V is first set
to~0,
then {\tt hstatus}.SP2V is set to {\tt hstatus}.SPV, {\tt hstatus}.SP2P is set
to {\tt sstatus}.SPP, and lastly {\tt hstatus}.SPV and {\tt sstatus}.SPP are
set according to Table~\ref{h-spp}.

\begin{table*}[h!]
\begin{center}
\begin{tabular}{|l|c|c|}
  \hline
  Previous Mode & SPV & SPP \\ \hline
  U-mode        & 0   & 0   \\
  HS-mode       & 0   & 1   \\ \hline
  VU-mode       & 1   & 0   \\
  VS-mode       & 1   & 1   \\ \hline
\end{tabular}
\end{center}
\caption{Value of {\tt hstatus} field SPV and {\tt sstatus} field SPP after a trap into HS-mode.}
\label{h-spp}
\end{table*}

When a trap is taken into VS-mode, {\tt vsstatus}.SPP is set according to
Table~\ref{h-vspp}.
Register {\tt hstatus} and the HS-level {\tt sstatus} are not modified,
and the virtualization mode V remains~1.

\begin{table*}[h!]
\begin{center}
\begin{tabular}{|l|c|c|}
  \hline
  Previous Mode & SPP \\ \hline
  VU-mode       & 0   \\
  VS-mode       & 1   \\ \hline
\end{tabular}
\end{center}
\caption{Value of {\tt vsstatus} field SPP after a trap into VS-mode.}
\label{h-vspp}
\end{table*}


\FloatBarrier

\section{Trap Return}

The MRET instruction is used to return from a trap taken into M-mode.
MRET first determines what the new operating mode will be according to
the values of MPP and MPV in {\tt mstatus}, as encoded in
Table~\ref{h-mpp}.
MRET then in {\tt mstatus} sets MPV=0, MPP=0, MIE=MPIE, and MPIE=1.
If the new operating mode will be U, VS, or VU, MRET also sets
{\tt hstatus}.SPRV=0.
Lastly, MRET sets the virtualization and privilege modes as previously
determined, and sets {\tt pc}={\tt mepc}.

The SRET instruction is used to return from a trap taken into HS-mode or
VS-mode.  Its behavior depends on the current virtualization mode.

When executed in M-mode or HS-mode (i.e., V=0), SRET first determines
what the new operating mode will be according to the values in
{\tt hstatus}.SPV and {\tt sstatus}.SPP, as encoded in Table~\ref{h-spp}.
SRET then sets {\tt hstatus}.SPV={\tt hstatus}.SP2V,
{\tt sstatus}.SPP={\tt hstatus}.SP2P, {\tt hstatus}.SP2V=0,
{\tt hstatus}.SP2P=0, {\tt sstatus}.SIE={\tt sstatus}.SPIE, and
{\tt sstatus}.SPIE=1.
If the new operating mode will be U, VS, or VU, SRET also sets
{\tt hstatus}.SPRV=0.
Lastly, SRET sets the virtualization and privilege modes as previously
determined, and sets {\tt pc}={\tt sepc}.

When executed in VS-mode (i.e., V=1), SRET sets the privilege mode according to
Table~\ref{h-vspp}, then in {\tt vsstatus} sets SPP=0, SIE=SPIE, and SPIE=1, and
lastly sets {\tt pc}={\tt vsepc}.