Make fence.tso a separate inst, fix fence bug

[dhower-qc]

fixes #428

[ThinkOpenly]

Do we want to use > here?

[dhower-qc]

I don't think so; the line spacing has syntactic meaning for asciidoc, so we don't want to fold anything.

[ThinkOpenly]

My understanding of the folding rules is that the text below this would be folded to (hopefully this renders well enough here):

Orders memory operations.

fence.tso orders all load operations in its predecessor set before all memory operations in its successor set, and all store operations in its predecessor set before all store operations in its successor set. This leaves non-AMO store operations in the 'fence.tso's predecessor set unordered with non-AMO loads in its successor set.

The rs1 and rd fields are unused and ignored.

In modes other than M-mode, fence.tso is further affected by menvcfg.FIOM, senvcfg.FIOM<% if ext?(:H) %>, and/or henvcfg.FIOM<% end %>.

Is that not preferred?

[dhower-qc]

I think it actually renders as:

Orders memory operations.
fence.tso orders all load operations in its predecessor set before all memory operations in its successor set, and all store operations in its predecessor set before all store operations in its successor set. This leaves non-AMO store operations in the 'fence.tso's predecessor set unordered with non-AMO loads in its successor set.
The rs1 and rd fields are unused and ignored.
In modes other than M-mode, fence.tso is further affected by menvcfg.FIOM, senvcfg.FIOM<% if ext?(:H) %>, and/or henvcfg.FIOM<% end %>.

Which in asciidoc will make all the text show up in a single paragraph.