On Wed, 2018-12-19 at 11:17 +0000, Daniel P. Berrangé wrote: > On Wed, Dec 19, 2018 at 10:42:14AM +0100, Gerd Hoffmann wrote: > > Hi, > > > > > > This is quite horrible though. I'm more inclined > > > > to revert this patch and find another way to fix the original problem > > > > which won't require the UI frontends to track modifier state. > > > > > > The UIs track modifier state anyway. > > > > > > I fact I have some WIP patches to add a generic keyboard state tracker, > > > so the UIs can use common code instead of having their own state > > > tracking code each. Also to make UI hotkeys configurable, consistent > > > across all UIs we have. Guess I should undust them, at least the state > > > tracking part of it. > > > > > > With that in place we can easily pass the full keyboard state to the > > > keymap code. > > > > Tried that: > > https://git.kraxel.org/cgit/qemu/log/?h=sirius/kbd-state > > I've had a look & think that code makes sense & ought to be able to > solve this problem.
I'll try and find a minute to do a scratch build with this patch (and without my keymap patch) and throw it on openQA staging, since openQA seems to do a great job of testing qemu input code. :P -- Adam Williamson Fedora QA Community Monkey IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net http://www.happyassassin.net
