Also the debugger should write dmcontrol=0 to perform a reset, instead
of "may".

Fixes #1021.

Alternate to #1040, after some discussion in the meeting. #1040 didn't
address what happens while dmactive is changing. This version is not a
substantive change, so also a bit easier to merge.

Fix for mepc example typo

Update maskmax6 detection algorithm to always work

Simplify WARL description for dmode=0 and action=1.

Clarify timing support suggestion for a chain of mcontrol load triggers

Eliminate redundant statement about AMO's stored value availability

Replace "count" with a reference to 'icount.count'

Replace single quotes with backticks in `mstatus` reference

Restore reference to `time`

Fix references to `sbcs.sbaccess64/128`

Fix formatting for `mconfigptr`

[3.14.6. Abstract Control and Status] make `busy` a reference

Restore references to `pc` in [3.8. Program Buffer]

Fix a typo in [3.12. Security]

Restore reference to `dmstatus.impebreak`

Restore references to `command.cmdtype` and `tab:cmdtype` table

Without the additional formatting the
> but count must be able to count

part seems a bit strange.

Seems like these references were lost durining convertion from LaTeX.

See https://github.com/riscv/riscv-debug-spec/blob/f510a7dd33317d0eee0f26b4fa082cd43a5ac7ea/Sdext.tex#L47-L49

LaTeX refernces didn't use numbers and replaced them with text.

Seems like this caused an error during convertion and references to
`sbcs.sbaccess64/128` started to refer to `sbcs.sbaccess` field with an
addition of `Sixtyfour` and `OneTwentyeight` respectively.

Names of registers not defined in this specification are highlighted
using backticks throughout the specification.

`mconfigptr` used to be an exception.

This makes it clear that the statatement refers to a field of a
register, not a name of a value of `abstractcs.cmderr` field.

Moreover, references are used throughout the spec in similar cases.

Seems like these references were lost durining convertion from LaTeX.

See https://github.com/riscv/riscv-debug-spec/blob/f510a7dd33317d0eee0f26b4fa082cd43a5ac7ea/debug_module.tex#L468

Seems like these references were lost durining convertion from LaTeX.

See https://github.com/riscv/riscv-debug-spec/blob/f510a7dd33317d0eee0f26b4fa082cd43a5ac7ea/debug_module.tex#L442

Seems like these references were lost durining convertion from LaTeX.
See https://github.com/riscv/riscv-debug-spec/blob/f510a7dd33317d0eee0f26b4fa082cd43a5ac7ea/debug_module.tex#L405-L406

Don't trademark the trigger module

Signed-off-by: Alexandre Joannou <gameboo@users.noreply.github.com>

1.0.0-rc2

Add support for Smdbltrp.

Restore references to `mhartid` in `debug_module.adoc`

"Native Triggers": Add in mepc/mcause back in.

It is covered by
> Whether data store triggers match on AMOs is UNSPECIFIED.

Seems like these references were lost durining convertion from LaTeX.
See https://github.com/riscv/riscv-debug-spec/blob/f510a7dd33317d0eee0f26b4fa082cd43a5ac7ea/debug_module.tex#L125-L126

These got deleted by the asciidoc conversion.

Fixes #1015.

Previously, the first write of match=1 might fail because the value in
tdata2 is incompatible.

Addresses #1014

fix a reference to `mcontrol.store` field