binutils-gdb/gas/doc/c-riscv.texi

@c Copyright (C) 2016-2022 Free Software Foundation, Inc.
@c This is part of the GAS anual.
@c For copying conditions, see the file as.texinfo
@c man end

@ifset GENERIC
@page
@node RISC-V-Dependent
@chapter RISC-V Dependent Features
@end ifset
@ifclear GENERIC
@node Machine Dependencies
@chapter RISC-V Dependent Features
@end ifclear

@cindex RISC-V support
@menu
* RISC-V-Options::        RISC-V Options
* RISC-V-Directives::     RISC-V Directives
* RISC-V-Modifiers::      RISC-V Assembler Modifiers
* RISC-V-Formats::        RISC-V Instruction Formats
* RISC-V-ATTRIBUTE::      RISC-V Object Attribute
@end menu

@node RISC-V-Options
@section RISC-V Options

The following table lists all available RISC-V specific options.

@c man begin OPTIONS
@table @gcctabopt

@cindex @samp{-fpic} option, RISC-V
@item -fpic
@itemx -fPIC
Generate position-independent code

@cindex @samp{-fno-pic} option, RISC-V
@item -fno-pic
Don't generate position-independent code (default)

@cindex @samp{-march=ISA} option, RISC-V
@item -march=ISA
Select the base isa, as specified by ISA.  For example -march=rv32ima.
If this option and the architecture attributes aren't set, then assembler
will check the default configure setting --with-arch=ISA.

@cindex @samp{-misa-spec=ISAspec} option, RISC-V
@item -misa-spec=ISAspec
Select the default isa spec version.  If the version of ISA isn't set
by -march, then assembler helps to set the version according to
the default chosen spec.  If this option isn't set, then assembler will
check the default configure setting --with-isa-spec=ISAspec.

@cindex @samp{-mpriv-spec=PRIVspec} option, RISC-V
@item -mpriv-spec=PRIVspec
Select the privileged spec version.  We can decide whether the CSR is valid or
not according to the chosen spec.  If this option and the privilege attributes
aren't set, then assembler will check the default configure setting
--with-priv-spec=PRIVspec.

@cindex @samp{-mabi=ABI} option, RISC-V
@item -mabi=ABI
Selects the ABI, which is either "ilp32" or "lp64", optionally followed
by "f", "d", or "q" to indicate single-precision, double-precision, or
quad-precision floating-point calling convention, or none to indicate
the soft-float calling convention.  Also, "ilp32" can optionally be followed
by "e" to indicate the RVE ABI, which is always soft-float.

@cindex @samp{-mrelax} option, RISC-V
@item -mrelax
Take advantage of linker relaxations to reduce the number of instructions
required to materialize symbol addresses. (default)

@cindex @samp{-mno-relax} option, RISC-V
@item -mno-relax
Don't do linker relaxations.

@cindex @samp{-march-attr} option, RISC-V
@item -march-attr
Generate the default contents for the riscv elf attribute section if the
.attribute directives are not set.  This section is used to record the
information that a linker or runtime loader needs to check compatibility.
This information includes ISA string, stack alignment requirement, unaligned
memory accesses, and the major, minor and revision version of privileged
specification.

@cindex @samp{-mno-arch-attr} option, RISC-V
@item -mno-arch-attr
Don't generate the default riscv elf attribute section if the .attribute
directives are not set.

@cindex @samp{-mcsr-check} option, RISC-V
@item -mcsr-check
Enable the CSR checking for the ISA-dependent CRS and the read-only CSR.
The ISA-dependent CSR are only valid when the specific ISA is set.  The
read-only CSR can not be written by the CSR instructions.

@cindex @samp{-mno-csr-check} option, RISC-V
@item -mno-csr-check
Don't do CSR checking.

@cindex @samp{-mlittle-endian} option, RISC-V
@item -mlittle-endian
Generate code for a little endian machine.

@cindex @samp{-mbig-endian} option, RISC-V
@item -mbig-endian
Generate code for a big endian machine.
@end table
@c man end

@node RISC-V-Directives
@section RISC-V Directives
@cindex machine directives, RISC-V
@cindex RISC-V machine directives

The following table lists all available RISC-V specific directives.

@table @code

@cindex @code{align} directive
@item .align @var{size-log-2}
Align to the given boundary, with the size given as log2 the number of bytes to
align to.

@cindex Data directives
@item .half @var{value}
@itemx .word @var{value}
@itemx .dword @var{value}
Emits a half-word, word, or double-word value at the current position.

@cindex DTP-relative data directives
@item .dtprelword @var{value}
@itemx .dtpreldword @var{value}
Emits a DTP-relative word (or double-word) at the current position.  This is
meant to be used by the compiler in shared libraries for DWARF debug info for
thread local variables.

@cindex BSS directive
@item .bss
Sets the current section to the BSS section.

@cindex LEB128 directives
@item .uleb128 @var{value}
@itemx .sleb128 @var{value}
Emits a signed or unsigned LEB128 value at the current position.  This only
accepts constant expressions, because symbol addresses can change with
relaxation, and we don't support relocations to modify LEB128 values at link
time.

@cindex Option directive
@cindex @code{option} directive
@item .option @var{argument}
Modifies RISC-V specific assembler options inline with the assembly code.
This is used when particular instruction sequences must be assembled with a
specific set of options.  For example, since we relax addressing sequences to
shorter GP-relative sequences when possible the initial load of GP must not be
relaxed and should be emitted as something like

@smallexample
	.option push
	.option norelax
	la gp, __global_pointer$
	.option pop
@end smallexample

in order to produce after linker relaxation the expected

@smallexample
	auipc gp, %pcrel_hi(__global_pointer$)
	addi gp, gp, %pcrel_lo(__global_pointer$)
@end smallexample

instead of just

@smallexample
	addi gp, gp, 0
@end smallexample

It's not expected that options are changed in this manner during regular use,
but there are a handful of esoteric cases like the one above where users need
to disable particular features of the assembler for particular code sequences.
The complete list of option arguments is shown below:

@table @code
@item push
@itemx pop
Pushes or pops the current option stack.  These should be used whenever
changing an option in line with assembly code in order to ensure the user's
command-line options are respected for the bulk of the file being assembled.

@item rvc
@itemx norvc
Enables or disables the generation of compressed instructions.  Instructions
are opportunistically compressed by the RISC-V assembler when possible, but
sometimes this behavior is not desirable, especially when handling alignments.

@item pic
@itemx nopic
Enables or disables position-independent code generation.  Unless you really
know what you're doing, this should only be at the top of a file.

@item relax
@itemx norelax
Enables or disables relaxation.  The RISC-V assembler and linker
opportunistically relax some code sequences, but sometimes this behavior is not
desirable.

@item csr-check
@itemx no-csr-check
Enables or disables the CSR checking.

@item arch, @var{+extension[version]} [,...,@var{+extension_n[version_n]}]
@itemx arch, @var{-extension} [,...,@var{-extension_n}]
@itemx arch, @var{=ISA}
Enables or disables the extensions for specific code region.  For example,
@samp{.option arch, +m2p0} means add m extension with version 2.0, and
@samp{.option arch, -f, -d} means remove extensions, f and d, from the
architecture string.  Note that, @samp{.option arch, +c, -c} have the same
behavior as @samp{.option rvc, norvc}.  However, they are also undesirable
sometimes.  Besides, @samp{.option arch, -i} is illegal, since we cannot
remove the base i extension anytime.  If you want to reset the whole ISA
string, you can also use @samp{.option arch, =rv32imac} to overwrite the
previous settings.
@end table

@cindex INSN directives
@item .insn @var{type}, @var{operand} [,...,@var{operand_n}]
@itemx .insn @var{insn_length}, @var{value}
@itemx .insn @var{value}
This directive permits the numeric representation of an instructions
and makes the assembler insert the operands according to one of the
instruction formats for @samp{.insn} (@ref{RISC-V-Formats}).
For example, the instruction @samp{add a0, a1, a2} could be written as
@samp{.insn r 0x33, 0, 0, a0, a1, a2}.  But in fact, the instruction
formats are difficult to use for some users, so most of them are using
@samp{.word} to encode the instruction directly, rather than using
@samp{.insn}.  It is fine for now, but will be wrong when the mapping
symbols are supported, since @samp{.word} will not be shown as an
instruction, it should be shown as data.  Therefore, we also support
two more formats of the @samp{.insn}, the instruction @samp{add a0, a1, a2}
could also be written as @samp{.insn 0x4, 0xc58533} or @samp{.insn 0xc58533}.
When the @var{insn_length} is set, then assembler will check if the
@var{value} is a valid @var{insn_length} bytes instruction.

@cindex @code{.attribute} directive, RISC-V
@item .attribute @var{tag}, @var{value}
Set the object attribute @var{tag} to @var{value}.

The @var{tag} is either an attribute number, or one of the following:
@code{Tag_RISCV_arch}, @code{Tag_RISCV_stack_align},
@code{Tag_RISCV_unaligned_access}, @code{Tag_RISCV_priv_spec},
@code{Tag_RISCV_priv_spec_minor}, @code{Tag_RISCV_priv_spec_revision}.

@end table

@node RISC-V-Modifiers
@section RISC-V Assembler Modifiers

The RISC-V assembler supports following modifiers for relocatable addresses
used in RISC-V instruction operands.  However, we also support some pseudo
instructions that are easier to use than these modifiers.

@table @code
@item %lo(@var{symbol})
The low 12 bits of absolute address for @var{symbol}.

@item %hi(@var{symbol})
The high 20 bits of absolute address for @var{symbol}.  This is usually
used with the %lo modifier to represent a 32-bit absolute address.

@smallexample
	lui        a0, %hi(@var{symbol})     // R_RISCV_HI20
	addi       a0, a0, %lo(@var{symbol}) // R_RISCV_LO12_I

	lui        a0, %hi(@var{symbol})     // R_RISCV_HI20
	load/store a0, %lo(@var{symbol})(a0) // R_RISCV_LO12_I/S
@end smallexample

@item %pcrel_lo(@var{label})
The low 12 bits of relative address between pc and @var{symbol}.
The @var{symbol} is related to the high part instruction which is marked
by @var{label}.

@item %pcrel_hi(@var{symbol})
The high 20 bits of relative address between pc and @var{symbol}.
This is usually used with the %pcrel_lo modifier to represent a +/-2GB
pc-relative range.

@smallexample
@var{label}:
	auipc      a0, %pcrel_hi(@var{symbol})    // R_RISCV_PCREL_HI20
	addi       a0, a0, %pcrel_lo(@var{label}) // R_RISCV_PCREL_LO12_I

@var{label}:
	auipc      a0, %pcrel_hi(@var{symbol})    // R_RISCV_PCREL_HI20
	load/store a0, %pcrel_lo(@var{label})(a0) // R_RISCV_PCREL_LO12_I/S
@end smallexample

Or you can use the pseudo lla/lw/sw/... instruction to do this.

@smallexample
	lla  a0, @var{symbol}
@end smallexample

@item %got_pcrel_hi(@var{symbol})
The high 20 bits of relative address between pc and the GOT entry of
@var{symbol}.  This is usually used with the %pcrel_lo modifier to access
the GOT entry.

@smallexample
@var{label}:
	auipc      a0, %got_pcrel_hi(@var{symbol}) // R_RISCV_GOT_HI20
	addi       a0, a0, %pcrel_lo(@var{label})  // R_RISCV_PCREL_LO12_I

@var{label}:
	auipc      a0, %got_pcrel_hi(@var{symbol}) // R_RISCV_GOT_HI20
	load/store a0, %pcrel_lo(@var{label})(a0)  // R_RISCV_PCREL_LO12_I/S
@end smallexample

Also, the pseudo la instruction with PIC has similar behavior.

@item %tprel_add(@var{symbol})
This is used purely to associate the R_RISCV_TPREL_ADD relocation for
TLS relaxation.  This one is only valid as the fourth operand to the normally
3 operand add instruction.

@item %tprel_lo(@var{symbol})
The low 12 bits of relative address between tp and @var{symbol}.

@item %tprel_hi(@var{symbol})
The high 20 bits of relative address between tp and @var{symbol}.  This is
usually used with the %tprel_lo and %tprel_add modifiers to access the thread
local variable @var{symbol} in TLS Local Exec.

@smallexample
	lui        a5, %tprel_hi(@var{symbol})          // R_RISCV_TPREL_HI20
	add        a5, a5, tp, %tprel_add(@var{symbol}) // R_RISCV_TPREL_ADD
	load/store t0, %tprel_lo(@var{symbol})(a5)      // R_RISCV_TPREL_LO12_I/S
@end smallexample

@item %tls_ie_pcrel_hi(@var{symbol})
The high 20 bits of relative address between pc and GOT entry.  It is
usually used with the %pcrel_lo modifier to access the thread local
variable @var{symbol} in TLS Initial Exec.

@smallexample
	la.tls.ie  a5, @var{symbol}
	add        a5, a5, tp
	load/store t0, 0(a5)
@end smallexample

The pseudo la.tls.ie instruction can be expended to

@smallexample
@var{label}:
	auipc a5, %tls_ie_pcrel_hi(@var{symbol}) // R_RISCV_TLS_GOT_HI20
	load  a5, %pcrel_lo(@var{label})(a5)     // R_RISCV_PCREL_LO12_I
@end smallexample

@item %tls_gd_pcrel_hi(@var{symbol})
The high 20 bits of relative address between pc and GOT entry.  It is
usually used with the %pcrel_lo modifier to access the thread local variable
@var{symbol} in TLS Global Dynamic.

@smallexample
	la.tls.gd  a0, @var{symbol}
	call       __tls_get_addr@@plt
	mv         a5, a0
	load/store t0, 0(a5)
@end smallexample

The pseudo la.tls.gd instruction can be expended to

@smallexample
@var{label}:
	auipc a0, %tls_gd_pcrel_hi(@var{symbol}) // R_RISCV_TLS_GD_HI20
	addi  a0, a0, %pcrel_lo(@var{label})     // R_RISCV_PCREL_LO12_I
@end smallexample

@end table

@node RISC-V-Formats
@section RISC-V Instruction Formats
@cindex instruction formats, risc-v
@cindex RISC-V instruction formats

The RISC-V Instruction Set Manual Volume I: User-Level ISA lists 15
instruction formats where some of the formats have multiple variants.
For the @samp{.insn} pseudo directive the assembler recognizes some
of the formats.
Typically, the most general variant of the instruction format is used
by the @samp{.insn} directive.

The following table lists the abbreviations used in the table of
instruction formats:

@display
@multitable @columnfractions .15 .40
@item opcode @tab Unsigned immediate or opcode name for 7-bits opcode.
@item opcode2 @tab Unsigned immediate or opcode name for 2-bits opcode.
@item func7 @tab Unsigned immediate for 7-bits function code.
@item func6 @tab Unsigned immediate for 6-bits function code.
@item func4 @tab Unsigned immediate for 4-bits function code.
@item func3 @tab Unsigned immediate for 3-bits function code.
@item func2 @tab Unsigned immediate for 2-bits function code.
@item rd @tab Destination register number for operand x, can be GPR or FPR.
@item rd' @tab Destination register number for operand x,
only accept s0-s1, a0-a5, fs0-fs1 and fa0-fa5.
@item rs1 @tab First source register number for operand x, can be GPR or FPR.
@item rs1' @tab First source register number for operand x,
only accept s0-s1, a0-a5, fs0-fs1 and fa0-fa5.
@item rs2 @tab Second source register number for operand x, can be GPR or FPR.
@item rs2' @tab Second source register number for operand x,
only accept s0-s1, a0-a5, fs0-fs1 and fa0-fa5.
@item simm12 @tab Sign-extended 12-bit immediate for operand x.
@item simm20 @tab Sign-extended 20-bit immediate for operand x.
@item simm6 @tab Sign-extended 6-bit immediate for operand x.
@item uimm5 @tab Unsigned 5-bit immediate for operand x.
@item uimm6 @tab Unsigned 6-bit immediate for operand x.
@item uimm8 @tab Unsigned 8-bit immediate for operand x.
@item symbol @tab Symbol or lable reference for operand x.
@end multitable
@end display

The following table lists all available opcode name:

@table @code
@item C0
@item C1
@item C2
Opcode space for compressed instructions.

@item LOAD
Opcode space for load instructions.

@item LOAD_FP
Opcode space for floating-point load instructions.

@item STORE
Opcode space for store instructions.

@item STORE_FP
Opcode space for floating-point store instructions.

@item AUIPC
Opcode space for auipc instruction.

@item LUI
Opcode space for lui instruction.

@item BRANCH
Opcode space for branch instructions.

@item JAL
Opcode space for jal instruction.

@item JALR
Opcode space for jalr instruction.

@item OP
Opcode space for ALU instructions.

@item OP_32
Opcode space for 32-bits ALU instructions.

@item OP_IMM
Opcode space for ALU with immediate instructions.

@item OP_IMM_32
Opcode space for 32-bits ALU with immediate instructions.

@item OP_FP
Opcode space for floating-point operation instructions.

@item MADD
Opcode space for madd instruction.

@item MSUB
Opcode space for msub instruction.

@item NMADD
Opcode space for nmadd instruction.

@item NMSUB
Opcode space for msub instruction.

@item AMO
Opcode space for atomic memory operation instructions.

@item MISC_MEM
Opcode space for misc instructions.

@item SYSTEM
Opcode space for system instructions.

@item CUSTOM_0
@item CUSTOM_1
@item CUSTOM_2
@item CUSTOM_3
Opcode space for customize instructions.

@end table

An instruction is two or four bytes in length and must be aligned
on a 2 byte boundary. The first two bits of the instruction specify the
length of the instruction, 00, 01 and 10 indicates a two byte instruction,
11 indicates a four byte instruction.

The following table lists the RISC-V instruction formats that are available
with the @samp{.insn} pseudo directive:

@table @code
@item R type: .insn r opcode6, func3, func7, rd, rs1, rs2
@verbatim
+-------+-----+-----+-------+----+---------+
| func7 | rs2 | rs1 | func3 | rd | opcode6 |
+-------+-----+-----+-------+----+---------+
31      25    20    15      12   7        0
@end verbatim

@item R type with 4 register operands: .insn r opcode6, func3, func2, rd, rs1, rs2, rs3
@itemx R4 type: .insn r4 opcode6, func3, func2, rd, rs1, rs2, rs3
@verbatim
+-----+-------+-----+-----+-------+----+---------+
| rs3 | func2 | rs2 | rs1 | func3 | rd | opcode6 |
+-----+-------+-----+-----+-------+----+---------+
31    27      25    20    15      12   7         0
@end verbatim

@item I type: .insn i opcode6, func3, rd, rs1, simm12
@itemx I type: .insn i opcode6, func3, rd, simm12(rs1)
@verbatim
+--------------+-----+-------+----+---------+
| simm12[11:0] | rs1 | func3 | rd | opcode6 |
+--------------+-----+-------+----+---------+
31             20    15      12   7         0
@end verbatim

@item S type: .insn s opcode6, func3, rs2, simm12(rs1)
@verbatim
+--------------+-----+-----+-------+-------------+---------+
| simm12[11:5] | rs2 | rs1 | func3 | simm12[4:0] | opcode6 |
+--------------+-----+-----+-------+-------------+---------+
31             25    20    15      12            7         0
@end verbatim

@item B type: .insn s opcode6, func3, rs1, rs2, symbol
@itemx SB type: .insn sb opcode6, func3, rs1, rs2, symbol
@verbatim
+-----------------+-----+-----+-------+----------------+---------+
| simm12[12|10:5] | rs2 | rs1 | func3 | simm12[4:1|11] | opcode6 |
+-----------------+-----+-----+-------+----------------+---------+
31                25    20    15      12               7         0
@end verbatim

@item U type: .insn u opcode6, rd, simm20
@verbatim
+--------------------------+----+---------+
| simm20[20|10:1|11|19:12] | rd | opcode6 |
+--------------------------+----+---------+
31                         12   7         0
@end verbatim

@item J type: .insn j opcode6, rd, symbol
@itemx UJ type: .insn uj opcode6, rd, symbol
@verbatim
+------------+--------------+------------+---------------+----+---------+
| simm20[20] | simm20[10:1] | simm20[11] | simm20[19:12] | rd | opcode6 |
+------------+--------------+------------+---------------+----+---------+
31           30             21           20              12   7         0
@end verbatim

@item CR type: .insn cr opcode2, func4, rd, rs2
@verbatim
+-------+--------+-----+---------+
| func4 | rd/rs1 | rs2 | opcode2 |
+-------+--------+-----+---------+
15      12       7     2        0
@end verbatim

@item CI type: .insn ci opcode2, func3, rd, simm6
@verbatim
+-------+----------+--------+------------+---------+
| func3 | simm6[5] | rd/rs1 | simm6[4:0] | opcode2 |
+-------+----------+--------+------------+---------+
15      13         12       7            2         0
@end verbatim

@item CIW type: .insn ciw opcode2, func3, rd', uimm8
@verbatim
+-------+------------+-----+---------+
| func3 | uimm8[7:0] | rd' | opcode2 |
+-------+-------- ---+-----+---------+
15      13           5     2         0
@end verbatim

@item CSS type: .insn css opcode2, func3, rd, uimm6
@verbatim
+-------+------------+----+---------+
| func3 | uimm6[5:0] | rd | opcode2 |
+-------+------------+----+---------+
15      13           7    2         0
@end verbatim

@item CL type: .insn cl opcode2, func3, rd', uimm5(rs1')
@verbatim
+-------+------------+------+------------+------+---------+
| func3 | uimm5[4:2] | rs1' | uimm5[1:0] |  rd' | opcode2 |
+-------+------------+------+------------+------+---------+
15      13           10     7            5      2         0
@end verbatim

@item CS type: .insn cs opcode2, func3, rs2', uimm5(rs1')
@verbatim
+-------+------------+------+------------+------+---------+
| func3 | uimm5[4:2] | rs1' | uimm5[1:0] | rs2' | opcode2 |
+-------+------------+------+------------+------+---------+
15      13           10     7            5      2         0
@end verbatim

@item CA type: .insn ca opcode2, func6, func2, rd', rs2'
@verbatim
+-- ----+----------+-------+------+---------+
| func6 | rd'/rs1' | func2 | rs2' | opcode2 |
+-------+----------+-------+------+---------+
15      10         7       5      2         0
@end verbatim

@item CB type: .insn cb opcode2, func3, rs1', symbol
@verbatim
+-------+--------------+------+------------------+---------+
| func3 | simm8[8|4:3] | rs1' | simm8[7:6|2:1|5] | opcode2 |
+-------+--------------+------+------------------+---------+
15      13             10     7                  2         0
@end verbatim

@item CJ type: .insn cj opcode2, symbol
@verbatim
+-------+-------------------------------+---------+
| func3 | simm11[11|4|9:8|10|6|7|3:1|5] | opcode2 |
+-------+-------------------------------+---------+
15      13                              2         0
@end verbatim


@end table

For the complete list of all instruction format variants see
The RISC-V Instruction Set Manual Volume I: User-Level ISA.

@node RISC-V-ATTRIBUTE
@section RISC-V Object Attribute
@cindex Object Attribute, RISC-V

RISC-V attributes have a string value if the tag number is odd and an integer
value if the tag number is even.

@table @r
@item Tag_RISCV_stack_align (4)
Tag_RISCV_strict_align records the N-byte stack alignment for this object.  The
default value is 16 for RV32I or RV64I, and 4 for RV32E.

The smallest value will be used if object files with different
Tag_RISCV_stack_align values are merged.

@item Tag_RISCV_arch (5)
Tag_RISCV_arch contains a string for the target architecture taken from the
option @option{-march}.  Different architectures will be integrated into a
superset when object files are merged.

Note that the version information of the target architecture must be presented
explicitly in the attribute and abbreviations must be expanded.  The version
information, if not given by @option{-march}, must be in accordance with the
default specified by the tool.  For example, the architecture @code{RV32I} has
to be recorded in the attribute as @code{RV32I2P0} in which @code{2P0} stands
for the default version of its base ISA.  On the other hand, the architecture
@code{RV32G} has to be presented as @code{RV32I2P0_M2P0_A2P0_F2P0_D2P0} in
which the abbreviation @code{G} is expanded to the @code{IMAFD} combination
with default versions of the standard extensions.

@item Tag_RISCV_unaligned_access (6)
Tag_RISCV_unaligned_access is 0 for files that do not allow any unaligned
memory accesses, and 1 for files that do allow unaligned memory accesses.

@item Tag_RISCV_priv_spec (8)
@item Tag_RISCV_priv_spec_minor (10)
@item Tag_RISCV_priv_spec_revision (12)
Tag_RISCV_priv_spec contains the major/minor/revision version information of
the privileged specification.  It will report errors if object files of
different privileged specification versions are merged.

@end table