Message ID | 20190210130617.8392-1-tomli@tomli.me (mailing list archive) |
---|---|
Headers | show |
Series | mips: loongson64: move EC header to include/asm/mach-loongson64 | expand |
Hi, On Sun, Feb 10, 2019 at 09:06:16PM +0800, Yifeng Li wrote: > In order to operate the Embedded Controller from multiple platform > drivers, it should be possible to include lemote-2f/ec_kb3310b.h > from everywhere. This commits move it from lemote-2f/ec_kb3310b.h > to include/asm/mach-loongson64/. > > This simple patch immediately enables the implementation of two > platform drivers. if there's no objection from the maintainers, > please consider to prioritize it for mips-next, thanks. > > Yifeng Li (1): > mips: loongson64: move EC header to include/asm/mach-loongson64 This probably should be MFD driver under drivers/mfd. It's a longer road, though... A.
> Hi, > > Yifeng Li (1): > > mips: loongson64: move EC header to include/asm/mach-loongson64 > > This probably should be MFD driver under drivers/mfd. It's a longer > road, though... > > A. The problem of converting it to a MFD driver, is that there's still something doesn't fit together. The core CPU idle code in loongson64/common calls the platform code in arch/mips/loongson64/lemote-2f to set up an interrupt handler in i8259, and inspect if the source of interrupt is an lid open event by querying the ENE KB3310B controller. But normally an MFD driver is only used to share the underlying hardware between multiple drivers in each subsystem, not to provide shared utilily functions. But after reading drivers/mfd for a while, apparently we do have drivers that "EXPORT_SYMBOL()" to other parts of the kernel. So I think it's a matter of fact that can be discussed with MFD maintainers. So yes, I'm converting all the EC code to a MFD driver. I have withdrawn this patch, do not merge it. The new MFD patch is coming soon. Thanks, Tom Li