Messages in this thread | | | Date | Fri, 18 Jan 2008 13:25:40 +0100 | From | Ingo Molnar <> | Subject | Re: [PATCH] x86: fix unconditional arch/x86/kernel/pcspeaker.c?compiling |
| |
* Michael Opdenacker <michael-lists@free-electrons.com> wrote:
> On 01/18/2008 12:02 PM, Ingo Molnar wrote: > > > > why didnt you make this: > > > > obj-$(CONFIG_INPUT_PCSPKR) += pcspeaker.o > > > > ? > > > Many thanks for your feedback. > > That's what I did first, but if CONFIG_INPUT_PCSPKR=m, > arch/x86/kernel/pcspeaker.c gets compiled as a module. While compiling > doesn't fail, is this still a valid module? It defines no init and > exit functions, and it defines an initcall, which only makes sense at > boot time. > > We could make pcspeaker.c depend on another switch, like > CONFIG_PCSPEAKER on mips. We could offer the possibility to disable it > when CONFIG_EMBEDDED is set.
i'm confused, the .ko definitely exists:
europe:~> uname -r 2.6.24-0.123.rc6.fc9
europe:~> lsmod | grep pcspkr pcspkr 6400 0
ah, this is a different .ko.
perhaps the right solution would be to only build it in if CONFIG_PCSPEAKER is "y" or "m". I.e. your original patch?
Ingo
| |