Re: [RFC PATCH 00/24] plugins: Allow to read registers

[Alex Bennée]

Akihiko Odaki <akihiko.odaki@daynix.com> writes:

> I and other people in the University of Tokyo, where I research processor
> design, found TCG plugins are very useful for processor design
> exploration.

Thanks for the submission - I've finished my initial review pass.

I think introducing register introspection into the plugins subsystem is
a very worthwhile addition. I'm also happy (for now) to use the
underlying gdb support for it in lieu of a greater refactoring of QEMU's
multiple register introspection features.

> The feature we find missing is the capability to read registers from
> plugins. In this series, I propose to add such a capability by reusing
> gdbstub code.
>
> The reuse of gdbstub code ensures the long-term stability of the TCG plugin
> interface for register access without incurring a burden to maintain yet
> another interface for register access.

However I don't want to expose the gdb detail to plugins to leave us a
free hand in further internal clean-ups later on.

> This process to add TCG plugin involves four major changes. The first one
> is to add GDBFeature structure that represents a GDB feature, which usually
> includes registers. GDBFeature can be generated from static XML files or
> dynamically generated by architecture-specific code. In fact, this is a
> refactoring independent of the feature this series adds, and potentially
> it's benefitial even without the plugin feature. The plugin feature will
> utilize this new structure to describe registers exposed to plugins.

I think we can get cleanups to this handling in ahead of the wider
plugin feature. Ideally it would be nice to push the XML generation into
gdbstub itself but that might be more of a refactor than you are willing
to pursue for the time being.

> The second one is to make gdb_read_register/gdb_write_register usable
> outside of gdbstub context.
>
> The third one is to actually make registers readable for plugins.

Modulo isolating the plugin API from gdb specifics I'm happy with this
approach.

> The last one is to allow to implement a QEMU plugin in C++. A plugin that
> I'll describe later is written in C++.

I would want a more compelling reason that a hello world plugin for
this. Only because QEMU has removed a bunch of C++ dependency over the
last few years so I don't think we are in a rush to re-introduce it. 

Are you OK to do a re-spin addressing the comments so far?

-- 
Alex Bennée
Virtualisation Tech Lead @ Linaro