Issue 201007.1: Wide Registers in Location Description Expressions
Section 2.5.1.3, pg 29
Some architectures support registers that are wider than the DWARF stack.
Compilers may choose to store a vector of values in those registers that
are relevant for describing the location of objects, such as addresses or
offsets from a common base address.
To support extracting those values from such wide registers and using them
in DWARF expressions, we propose a family of operators.
===
Section 2.5.1.3, pg. 29.
Rephrase
DW_OP_regval_type pushes the contents of the register together with
the given base type, while the other operations push the result of
adding the contents of a register to a given signed offset together
with the generic type.
into
DW_OP_regval_type pushes the contents of the register together with
the given base type. DW_OP_regval_bits pushes the partial contents of
the register together with the generic type. The other operations
push the result of adding the contents of a register to a given signed
offset together with the generic type.
Add
5. DW_OP_regval_bits
The DW_OP_regval_bits operation takes a single 1-byte unsigned
integer operand, which gives the number of bits to read, which must
be smaller or equal to the bit size of the generic type. It pops
the top two stack elements and interprets the top element as an
unsigned bit offset from the least significant bit end and the
other as a register number identifying the register from which to
extract the value. If the extracted value is smaller than the size
of the generic type, it is zero extended.
Section 7.7.1, p223ff
Add
DW_OP_regval_bits | 0xaa | 1 | 1-byte bit size
to Table 7.9.
--
2022-01-25: Revised. Previous version: http://dwarfstd.org/issues/201007.1-1.html
2022-03-17: Revised. Removed DW_OP_regval_bits_stack_offset and DW_OP_regval_bits_stack.
Previous version: http://dwarfstd.org/issues/201007.1-2.html
2022-03-21: Accepted with modification: Change ubyte operand to ULEB128.
