On Thu, Oct 08, 2015 at 03:16:02AM +0000, Zha, Qipeng wrote: > > >>> + ipcdev.base[BIOS_MAILBOX] = addr; > >>> + addr += MAILBOX_REGISTER_SPACE; > >>> + ipcdev.base[GTDRIVER_MAILBOX] = addr; > >>> + addr += MAILBOX_REGISTER_SPACE; > >>> + ipcdev.base[ISPDRIVER_MAILBOX] = addr; > > >>Looks akward, does the platform have the several resources for different purpose? Why do you unify them (who does guarantee the non-breakable segment for all resources?) first and then split up? > >>Please, refactor. > > >From spec, these three parts are consecutive, so only define one acpi resource entry is reasonable way, > >But BIOS maintainer finally make the resource like this due to request from Window os driver, > >So can't treat these three as three separate parts. > > Andriy, Darren: this is my before feedback, and My apologies Qipeng, I missed this response from you. > a) Punit function is configured as ACPI device in BIOS, not PCI device(seems can't configure as PCI). > b) To make it compatible(same acpi entry) for WOS, BIOS define Punit mem resource as res0, so this driver is coding as this. > So this driver is depend on acpi entry in BIOS to get device resource like all acpi device drivers. > In future, I don't think BIOS will change current defined resource for Punit function. An acpidump would be good to see, along with the couple minor changes requested. I want this to get into next sooner rather than later, so I'm queueing to testing today. If we could get the next rev and the acpidump out soon, we can get to linux-next this weekend which would be ideal for 4.4. -- Darren Hart Intel Open Source Technology Center -- To unsubscribe from this list: send the line "unsubscribe platform-driver-x86" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
