Re: [PATCH v2 2/4] Introduce event-loop-base abstract class

[Nicolas Saenz Julienne]

On Thu, 2022-03-10 at 10:25 +0000, Stefan Hajnoczi wrote:
> On Thu, Mar 03, 2022 at 03:58:20PM +0100, Nicolas Saenz Julienne wrote:
> > @@ -2935,13 +2947,6 @@ qemu_syms = custom_target('qemu.syms', output: 
> > 'qemu.syms',
> >                               capture: true,
> >                               command: [undefsym, nm, '@INPUT@'])
> >  
> > -qom_ss = qom_ss.apply(config_host, strict: false)
> > -libqom = static_library('qom', qom_ss.sources() + genh,
> > -                        dependencies: [qom_ss.dependencies()],
> > -                        name_suffix: 'fa')
> > -
> > -qom = declare_dependency(link_whole: libqom)
> > -
> 
> Why was it necessary to move qom_ss and subdir('hw') up? Can
> event_loop_base be defined down here instead?

The way I setup it up, qemuutil now depdens on event_loop_base which in turn
depends on qom. IIUC I can't declare dependencies without declaring first the
libraries and source sets. All has to happen sequencially. With this in mind,
almost all libraries depend on libqemuutil so moving it down isn't possible.

subdir('hw') was also moved up, as 'hw/nvram/meson.build' is introducing files
into qom_ss. This operation has to be performed before declaring libqom.

> (The benefit of less code churn is it reduces the risk of patch conflicts.)

Agree, I know how painful it can be for backports.

Thanks,

-- 
Nicolás Sáenz