Am 25.07.2013 19:18, schrieb Michael S. Tsirkin: > On Thu, Jul 25, 2013 at 05:50:55PM +0200, Andreas Färber wrote: >> Am 24.07.2013 18:01, schrieb Michael S. Tsirkin: >>> This code can also be found here: >>> git://git.kernel.org/pub/scm/virt/kvm/mst/qemu.git acpi >>> >>> Please review, and consider for 1.6. >> >> Quite frankly, this is still not looking the way I imagined it based on >> the KVM call discussion and Anthony's comments that I remember: > > By the way, for examples which are exactly > like this code here, look no further than: > > FWCfgState *fw_cfg_find(void); > > and the use in hw/misc/pvpanic.c of said API. > > Here we have device implementing an API to find it, and through this > pointer, it is accessed by more APIs access its state. > > This is exactly what my patches do for piix etc. > > So if you don't like this approach, please change > existing code using it.
According to git-blame that code was recently added by you: http://repo.or.cz/w/qemu.git/commit/600c60b76d0682f6c39d19bfff79da9321e8cf86?f=hw/nvram/fw_cfg.c So the existing-code argument is kind of moot. ;) Further, that returns the device's state struct and does not copy fields from that state struct into another state struct. BTW if you want to avoid repeatedly hardcoding things, fw_cfg should be using object_resolve_path_component(), I'll prepare a patch. Andreas -- SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
