On Tue, 08/15 11:53, Daniel P. Berrange wrote: > On Tue, Aug 15, 2017 at 06:47:22PM +0800, Fam Zheng wrote: > > On Tue, 08/15 11:04, Daniel P. Berrange wrote: > > > On Tue, Aug 15, 2017 at 10:36:04AM +0100, Daniel P. Berrange wrote: > > > > The https://gitlab.com/keycodemap/keycodemapdb/ repo contains a > > > > data file mapping between all the different scancode/keycode/keysym > > > > sets that are known, and a tool to auto-generate lookup tables for > > > > different combinations. > > > > > > > > It is used by GTK-VNC, SPICE-GTK and libvirt for mapping keys. > > > > Using it in QEMU will let us replace many hand written lookup > > > > tables with auto-generated tables from a master data source, > > > > reducing bugs. Adding new QKeyCodes will now only require the > > > > master table to be updated, all ~20 other tables will be > > > > automatically updated to follow. > > > > > > > > Signed-off-by: Daniel P. Berrange <berra...@redhat.com> > > > > --- > > > > .gitignore | 2 ++ > > > > .gitmodules | 3 +++ > > > > configure | 2 ++ > > > > tests/docker/Makefile.include | 11 +++++++---- > > > > tests/docker/run | 4 +++- > > > > ui/Makefile.objs | 18 ++++++++++++++++++ > > > > ui/keycodemapdb | 1 + > > > > 7 files changed, 36 insertions(+), 5 deletions(-) > > > > create mode 160000 ui/keycodemapdb > > > > > > > > > > > diff --git a/configure b/configure > > > > index dd73cce62f..bd373ec2b4 100755 > > > > --- a/configure > > > > +++ b/configure > > > > @@ -5258,6 +5258,8 @@ echo_version() { > > > > fi > > > > } > > > > > > > > +(cd $source_path && git submodule update --init ui/keycodemapdb) > > > > + > > > > > > Urgh, no, this won't work because of course you don't have to > > > have a git checkout when running configure. > > > > > > Any suggestions on the "best" way to ensure that the ui/keycodemapdb > > > git submodule is always checked out, without requiring developers to > > > do something manually ? > > > > > > I could try > > > > > > (cd $source_path && test -d .git && git submodule update --init > > > ui/keycodemapdb) > > > > > > but wonder if there's a better way ? > > > > I think the way dtc handles this is okay: probe headers/libs, if failed, > > hint > > the "git submodule update" command in the error message. This is also a > > familiar/friendly user interface to developers. > > I don't think that's acceptable in this case. Few people will ever need > to setup the dtc submodule as distros commonly have that available.
It's not available in the case of RHEL/CentOS 7, so it is similar in a degree which is why I mentioned it. But certainly I'll be happy to see a way to have this automated. FWIW I've no objection to the "test -d .git" idea. For docker tests to work, I'm inclined to change the Makefile rules so that every initialized submodule under $SRC_PATH is git-archived and copied in. Fam
