Hi Bin,
On Thu, 28 Nov 2019 at 22:49, Bin Meng bmeng.cn@gmail.com wrote:
Hi Simon,
On Mon, Nov 25, 2019 at 12:12 PM Simon Glass sjg@chromium.org wrote:
This subsystem is present on various Intel SoCs.
Add very basic support for taking an lpss device out of reset.
Signed-off-by: Simon Glass sjg@chromium.org
Changes in v5: None Changes in v4:
- Add support for updating power state
- Move this to intel_common
Changes in v3: None Changes in v2: None
arch/x86/cpu/intel_common/Makefile | 1 + arch/x86/cpu/intel_common/lpss.c | 44 ++++++++++++++++++++++++++++++ arch/x86/include/asm/lpss.h | 36 ++++++++++++++++++++++++ 3 files changed, 81 insertions(+) create mode 100644 arch/x86/cpu/intel_common/lpss.c create mode 100644 arch/x86/include/asm/lpss.h
diff --git a/arch/x86/cpu/intel_common/Makefile b/arch/x86/cpu/intel_common/Makefile index 09212cee04..cc4e1c962b 100644 --- a/arch/x86/cpu/intel_common/Makefile +++ b/arch/x86/cpu/intel_common/Makefile @@ -19,6 +19,7 @@ endif obj-y += cpu.o obj-y += fast_spi.o obj-y += lpc.o +obj-y += lpss.o ifndef CONFIG_TARGET_EFI_APP obj-$(CONFIG_$(SPL_TPL_)X86_32BIT_INIT) += microcode.o ifndef CONFIG_$(SPL_)X86_64 diff --git a/arch/x86/cpu/intel_common/lpss.c b/arch/x86/cpu/intel_common/lpss.c new file mode 100644 index 0000000000..26a2d2d1e3 --- /dev/null +++ b/arch/x86/cpu/intel_common/lpss.c @@ -0,0 +1,44 @@ +// SPDX-License-Identifier: GPL-2.0 +/*
- Special driver to handle of-platdata
- Copyright 2019 Google LLC
- Some code from coreboot lpss.c
- */
+#include <common.h> +#include <dm.h> +#include <pci.h> +#include <asm/io.h> +#include <asm/lpss.h>
+enum {
LPSS_RESET_CTL_REG = 0x204,/** Bit 1:0 controls LPSS controller reset.** 00 ->LPSS Host Controller is in reset (Reset Asserted)* 01/10 ->Reserved* 11 ->LPSS Host Controller is NOT at reset (Reset Released)*/LPSS_CNT_RST_RELEASE = 3,/* Power management control and status register */PME_CTRL_STATUS = 0x84,/* Bit 1:0 Powerstate, controls D0 and D3 state */POWER_STATE_MASK = 3,Could we use #defines? These macros are register offsets and bit defines and should not be put in one enum.
Well we can, but I think enums are better. The enum is anonymous so it can't really be referenced by anything. This is just a 'C' way of doing constants, I think.
Regards, Simon