diff --git a/src/southbridge/intel/ibexpeak/me.c b/src/southbridge/intel/ibexpeak/me.c
new file mode 100644
index 0000000..bc56012
--- /dev/null
+++ b/src/southbridge/intel/ibexpeak/me.c
@@ -0,0 +1,781 @@
+/*
+ * This file is part of the coreboot project.
+ *
+ * Copyright (C) 2011 The Chromium OS Authors. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of
+ * the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston,
+ * MA 02110-1301 USA
+ */
+
+/*
+ * This is a ramstage driver for the Intel Management Engine found in the
+ * 6-series chipset.  It handles the required boot-time messages over the
+ * MMIO-based Management Engine Interface to tell the ME that the BIOS is
+ * finished with POST.  Additional messages are defined for debug but are
+ * not used unless the console loglevel is high enough.
+ */
+
+#include <arch/acpi.h>
+#include <arch/hlt.h>
+#include <arch/io.h>
+#include <console/console.h>
+#include <device/pci_ids.h>
+#include <device/pci_def.h>
+#include <string.h>
+#include <delay.h>
+#include <elog.h>
+
+#ifdef __SMM__
+#include <arch/pci_mmio_cfg.h>
+#else
+# include <device/device.h>
+# include <device/pci.h>
+#endif
+
+#include "me.h"
+#include "pch.h"
+
+#if CONFIG_CHROMEOS
+#include <vendorcode/google/chromeos/gnvs.h>
+#endif
+
+#ifndef __SMM__
+/* Path that the BIOS should take based on ME state */
+static const char *me_bios_path_values[] = {
+	[ME_NORMAL_BIOS_PATH]		= "Normal",
+	[ME_S3WAKE_BIOS_PATH]		= "S3 Wake",
+	[ME_ERROR_BIOS_PATH]		= "Error",
+	[ME_RECOVERY_BIOS_PATH]		= "Recovery",
+	[ME_DISABLE_BIOS_PATH]		= "Disable",
+	[ME_FIRMWARE_UPDATE_BIOS_PATH]	= "Firmware Update",
+};
+#endif
+
+/* MMIO base address for MEI interface */
+static u32 mei_base_address;
+
+#if CONFIG_DEBUG_INTEL_ME
+static void mei_dump(void *ptr, int dword, int offset, const char *type)
+{
+	struct mei_csr *csr;
+
+	printk(BIOS_SPEW, "%-9s[%02x] : ", type, offset);
+
+	switch (offset) {
+	case MEI_H_CSR:
+	case MEI_ME_CSR_HA:
+		csr = ptr;
+		if (!csr) {
+			printk(BIOS_SPEW, "ERROR: 0x%08x\n", dword);
+			break;
+		}
+		printk(BIOS_SPEW, "cbd=%u cbrp=%02u cbwp=%02u ready=%u "
+		       "reset=%u ig=%u is=%u ie=%u\n", csr->buffer_depth,
+		       csr->buffer_read_ptr, csr->buffer_write_ptr,
+		       csr->ready, csr->reset, csr->interrupt_generate,
+		       csr->interrupt_status, csr->interrupt_enable);
+		break;
+	case MEI_ME_CB_RW:
+	case MEI_H_CB_WW:
+		printk(BIOS_SPEW, "CB: 0x%08x\n", dword);
+		break;
+	default:
+		printk(BIOS_SPEW, "0x%08x\n", offset);
+		break;
+	}
+}
+#else
+# define mei_dump(ptr,dword,offset,type) do {} while (0)
+#endif
+
+/*
+ * ME/MEI access helpers using memcpy to avoid aliasing.
+ */
+
+static inline void mei_read_dword_ptr(void *ptr, int offset)
+{
+	u32 dword = read32(mei_base_address + offset);
+	memcpy(ptr, &dword, sizeof(dword));
+	mei_dump(ptr, dword, offset, "READ");
+}
+
+static inline void mei_write_dword_ptr(void *ptr, int offset)
+{
+	u32 dword = 0;
+	memcpy(&dword, ptr, sizeof(dword));
+	write32(mei_base_address + offset, dword);
+	mei_dump(ptr, dword, offset, "WRITE");
+}
+
+#ifndef __SMM__
+static inline void pci_read_dword_ptr(device_t dev, void *ptr, int offset)
+{
+	u32 dword = pci_read_config32(dev, offset);
+	memcpy(ptr, &dword, sizeof(dword));
+	mei_dump(ptr, dword, offset, "PCI READ");
+}
+#endif
+
+static inline void read_host_csr(struct mei_csr *csr)
+{
+	mei_read_dword_ptr(csr, MEI_H_CSR);
+}
+
+static inline void write_host_csr(struct mei_csr *csr)
+{
+	mei_write_dword_ptr(csr, MEI_H_CSR);
+}
+
+static inline void read_me_csr(struct mei_csr *csr)
+{
+	mei_read_dword_ptr(csr, MEI_ME_CSR_HA);
+}
+
+static inline void write_cb(u32 dword)
+{
+	write32(mei_base_address + MEI_H_CB_WW, dword);
+	mei_dump(NULL, dword, MEI_H_CB_WW, "WRITE");
+}
+
+static inline u32 read_cb(void)
+{
+	u32 dword = read32(mei_base_address + MEI_ME_CB_RW);
+	mei_dump(NULL, dword, MEI_ME_CB_RW, "READ");
+	return dword;
+}
+
+/* Wait for ME ready bit to be asserted */
+static int mei_wait_for_me_ready(void)
+{
+	struct mei_csr me;
+	unsigned try = ME_RETRY;
+
+	while (try--) {
+		read_me_csr(&me);
+		if (me.ready)
+			return 0;
+		udelay(ME_DELAY);
+	}
+
+	printk(BIOS_ERR, "ME: failed to become ready\n");
+	return -1;
+}
+
+static void mei_reset(void)
+{
+	struct mei_csr host;
+
+	if (mei_wait_for_me_ready() < 0)
+		return;
+
+	/* Reset host and ME circular buffers for next message */
+	read_host_csr(&host);
+	host.reset = 1;
+	host.interrupt_generate = 1;
+	write_host_csr(&host);
+
+	if (mei_wait_for_me_ready() < 0)
+		return;
+
+	/* Re-init and indicate host is ready */
+	read_host_csr(&host);
+	host.interrupt_generate = 1;
+	host.ready = 1;
+	host.reset = 0;
+	write_host_csr(&host);
+}
+
+static int mei_send_msg(struct mei_header *mei, struct mkhi_header *mkhi,
+			void *req_data)
+{
+	struct mei_csr host;
+	unsigned ndata, n;
+	u32 *data;
+
+	/* Number of dwords to write, ignoring MKHI */
+	ndata = mei->length >> 2;
+
+	/* Pad non-dword aligned request message length */
+	if (mei->length & 3)
+		ndata++;
+	if (!ndata) {
+		printk(BIOS_DEBUG, "ME: request does not include MKHI\n");
+		return -1;
+	}
+	ndata++; /* Add MEI header */
+
+	/*
+	 * Make sure there is still room left in the circular buffer.
+	 * Reset the buffer pointers if the requested message will not fit.
+	 */
+	read_host_csr(&host);
+	if ((host.buffer_depth - host.buffer_write_ptr) < ndata) {
+		printk(BIOS_ERR, "ME: circular buffer full, resetting...\n");
+		mei_reset();
+		read_host_csr(&host);
+	}
+
+	/*
+	 * This implementation does not handle splitting large messages
+	 * across multiple transactions.  Ensure the requested length
+	 * will fit in the available circular buffer depth.
+	 */
+	if ((host.buffer_depth - host.buffer_write_ptr) < ndata) {
+		printk(BIOS_ERR, "ME: message (%u) too large for buffer (%u)\n",
+		       ndata + 2, host.buffer_depth);
+		return -1;
+	}
+
+	/* Write MEI header */
+	mei_write_dword_ptr(mei, MEI_H_CB_WW);
+	ndata--;
+
+	/* Write MKHI header */
+	mei_write_dword_ptr(mkhi, MEI_H_CB_WW);
+	ndata--;
+
+	/* Write message data */
+	data = req_data;
+	for (n = 0; n < ndata; ++n)
+		write_cb(*data++);
+
+	/* Generate interrupt to the ME */
+	read_host_csr(&host);
+	host.interrupt_generate = 1;
+	write_host_csr(&host);
+
+	/* Make sure ME is ready after sending request data */
+	return mei_wait_for_me_ready();
+}
+
+static int mei_recv_msg(struct mei_header *mei, struct mkhi_header *mkhi,
+			void *rsp_data, int rsp_bytes)
+{
+	struct mei_header mei_rsp;
+	struct mkhi_header mkhi_rsp;
+	struct mei_csr me, host;
+	unsigned ndata, n;
+	unsigned expected;
+	u32 *data;
+
+	/* Total number of dwords to read from circular buffer */
+	expected = (rsp_bytes + sizeof(mei_rsp) + sizeof(mkhi_rsp)) >> 2;
+	if (rsp_bytes & 3)
+		expected++;
+
+	/*
+	 * The interrupt status bit does not appear to indicate that the
+	 * message has actually been received.  Instead we wait until the
+	 * expected number of dwords are present in the circular buffer.
+	 */
+	for (n = ME_RETRY; n; --n) {
+		read_me_csr(&me);
+		if ((me.buffer_write_ptr - me.buffer_read_ptr) >= expected)
+			break;
+		udelay(ME_DELAY);
+	}
+	if (!n) {
+		printk(BIOS_ERR, "ME: timeout waiting for data: expected "
+		       "%u, available %u\n", expected,
+		       me.buffer_write_ptr - me.buffer_read_ptr);
+		return -1;
+	}
+
+	/* Read and verify MEI response header from the ME */
+	mei_read_dword_ptr(&mei_rsp, MEI_ME_CB_RW);
+	if (!mei_rsp.is_complete) {
+		printk(BIOS_ERR, "ME: response is not complete\n");
+		return -1;
+	}
+
+	/* Handle non-dword responses and expect at least MKHI header */
+	ndata = mei_rsp.length >> 2;
+	if (mei_rsp.length & 3)
+		ndata++;
+	if (ndata != (expected - 1)) {
+		printk(BIOS_ERR, "ME: response is missing data\n");
+		return -1;
+	}
+
+	/* Read and verify MKHI response header from the ME */
+	mei_read_dword_ptr(&mkhi_rsp, MEI_ME_CB_RW);
+	if (!mkhi_rsp.is_response ||
+	    mkhi->group_id != mkhi_rsp.group_id ||
+	    mkhi->command != mkhi_rsp.command) {
+		printk(BIOS_ERR, "ME: invalid response, group %u ?= %u, "
+		       "command %u ?= %u, is_response %u\n", mkhi->group_id,
+		       mkhi_rsp.group_id, mkhi->command, mkhi_rsp.command,
+		       mkhi_rsp.is_response);
+		return -1;
+	}
+	ndata--; /* MKHI header has been read */
+
+	/* Make sure caller passed a buffer with enough space */
+	if (ndata != (rsp_bytes >> 2)) {
+		printk(BIOS_ERR, "ME: not enough room in response buffer: "
+		       "%u != %u\n", ndata, rsp_bytes >> 2);
+		return -1;
+	}
+
+	/* Read response data from the circular buffer */
+	data = rsp_data;
+	for (n = 0; n < ndata; ++n)
+		*data++ = read_cb();
+
+	/* Tell the ME that we have consumed the response */
+	read_host_csr(&host);
+	host.interrupt_status = 1;
+	host.interrupt_generate = 1;
+	write_host_csr(&host);
+
+	return mei_wait_for_me_ready();
+}
+
+static inline int mei_sendrecv(struct mei_header *mei, struct mkhi_header *mkhi,
+			       void *req_data, void *rsp_data, int rsp_bytes)
+{
+	if (mei_send_msg(mei, mkhi, req_data) < 0)
+		return -1;
+	if (mei_recv_msg(mei, mkhi, rsp_data, rsp_bytes) < 0)
+		return -1;
+	return 0;
+}
+
+#ifdef __SMM__
+/* Send END OF POST message to the ME */
+static int mkhi_end_of_post(void)
+{
+	struct mkhi_header mkhi = {
+		.group_id	= MKHI_GROUP_ID_GEN,
+		.command	= MKHI_END_OF_POST,
+	};
+	struct mei_header mei = {
+		.is_complete	= 1,
+		.host_address	= MEI_HOST_ADDRESS,
+		.client_address	= MEI_ADDRESS_MKHI,
+		.length		= sizeof(mkhi),
+	};
+
+	/* Send request and wait for response */
+	if (mei_sendrecv(&mei, &mkhi, NULL, NULL, 0) < 0) {
+		printk(BIOS_ERR, "ME: END OF POST message failed\n");
+		return -1;
+	}
+
+	printk(BIOS_INFO, "ME: END OF POST message successful\n");
+	return 0;
+}
+#endif
+
+#if (CONFIG_DEFAULT_CONSOLE_LOGLEVEL >= BIOS_DEBUG) && !defined(__SMM__) && (CONFIG_NORTHBRIDGE_INTEL_SANDYBRIDGE || CONFIG_NORTHBRIDGE_INTEL_IVYBRIDGE)
+/* Get ME firmware version */
+static int mkhi_get_fw_version(void)
+{
+	struct me_fw_version version;
+	struct mkhi_header mkhi = {
+		.group_id	= MKHI_GROUP_ID_GEN,
+		.command	= MKHI_GET_FW_VERSION,
+	};
+	struct mei_header mei = {
+		.is_complete	= 1,
+		.host_address	= MEI_HOST_ADDRESS,
+		.client_address	= MEI_ADDRESS_MKHI,
+		.length		= sizeof(mkhi),
+	};
+
+	/* Send request and wait for response */
+	if (mei_sendrecv(&mei, &mkhi, NULL, &version, sizeof(version)) < 0) {
+		printk(BIOS_ERR, "ME: GET FW VERSION message failed\n");
+		return -1;
+	}
+
+	printk(BIOS_INFO, "ME: Firmware Version %u.%u.%u.%u (code) "
+	       "%u.%u.%u.%u (recovery)\n",
+	       version.code_major, version.code_minor,
+	       version.code_build_number, version.code_hot_fix,
+	       version.recovery_major, version.recovery_minor,
+	       version.recovery_build_number, version.recovery_hot_fix);
+
+	return 0;
+}
+
+static inline void print_cap(const char *name, int state)
+{
+	printk(BIOS_DEBUG, "ME Capability: %-30s : %sabled\n",
+	       name, state ? "en" : "dis");
+}
+
+/* Get ME Firmware Capabilities */
+static int mkhi_get_fwcaps(void)
+{
+	u32 rule_id = 0;
+	struct me_fwcaps cap;
+	struct mkhi_header mkhi = {
+		.group_id	= MKHI_GROUP_ID_FWCAPS,
+		.command	= MKHI_FWCAPS_GET_RULE,
+	};
+	struct mei_header mei = {
+		.is_complete	= 1,
+		.host_address	= MEI_HOST_ADDRESS,
+		.client_address	= MEI_ADDRESS_MKHI,
+		.length		= sizeof(mkhi) + sizeof(rule_id),
+	};
+
+	/* Send request and wait for response */
+	if (mei_sendrecv(&mei, &mkhi, &rule_id, &cap, sizeof(cap)) < 0) {
+		printk(BIOS_ERR, "ME: GET FWCAPS message failed\n");
+		return -1;
+	}
+
+	print_cap("Full Network manageability", cap.caps_sku.full_net);
+	print_cap("Regular Network manageability", cap.caps_sku.std_net);
+	print_cap("Manageability", cap.caps_sku.manageability);
+	print_cap("Small business technology", cap.caps_sku.small_business);
+	print_cap("Level III manageability", cap.caps_sku.l3manageability);
+	print_cap("IntelR Anti-Theft (AT)", cap.caps_sku.intel_at);
+	print_cap("IntelR Capability Licensing Service (CLS)",
+		  cap.caps_sku.intel_cls);
+	print_cap("IntelR Power Sharing Technology (MPC)",
+		  cap.caps_sku.intel_mpc);
+	print_cap("ICC Over Clocking", cap.caps_sku.icc_over_clocking);
+        print_cap("Protected Audio Video Path (PAVP)", cap.caps_sku.pavp);
+	print_cap("IPV6", cap.caps_sku.ipv6);
+	print_cap("KVM Remote Control (KVM)", cap.caps_sku.kvm);
+	print_cap("Outbreak Containment Heuristic (OCH)", cap.caps_sku.och);
+	print_cap("Virtual LAN (VLAN)", cap.caps_sku.vlan);
+	print_cap("TLS", cap.caps_sku.tls);
+	print_cap("Wireless LAN (WLAN)", cap.caps_sku.wlan);
+
+	return 0;
+}
+#endif
+
+#if CONFIG_CHROMEOS && 0 /* DISABLED */
+/* Tell ME to issue a global reset */
+int mkhi_global_reset(void)
+{
+	struct me_global_reset reset = {
+		.request_origin	= GLOBAL_RESET_BIOS_POST,
+		.reset_type	= CBM_RR_GLOBAL_RESET,
+	};
+	struct mkhi_header mkhi = {
+		.group_id	= MKHI_GROUP_ID_CBM,
+		.command	= MKHI_GLOBAL_RESET,
+	};
+	struct mei_header mei = {
+		.is_complete	= 1,
+		.length		= sizeof(mkhi) + sizeof(reset),
+		.host_address	= MEI_HOST_ADDRESS,
+		.client_address	= MEI_ADDRESS_MKHI,
+	};
+
+	printk(BIOS_NOTICE, "ME: Requesting global reset\n");
+
+	/* Send request and wait for response */
+	if (mei_sendrecv(&mei, &mkhi, &reset, NULL, 0) < 0) {
+		/* No response means reset will happen shortly... */
+		hlt();
+	}
+
+	/* If the ME responded it rejected the reset request */
+	printk(BIOS_ERR, "ME: Global Reset failed\n");
+	return -1;
+}
+#endif
+
+#ifdef __SMM__
+static void intel_me7_finalize_smm(void)
+{
+	struct me_hfs hfs;
+	u32 reg32;
+
+	mei_base_address =
+		pci_read_config32(PCH_ME_DEV, PCI_BASE_ADDRESS_0) & ~0xf;
+
+	/* S3 path will have hidden this device already */
+	if (!mei_base_address || mei_base_address == 0xfffffff0)
+		return;
+
+	/* Make sure ME is in a mode that expects EOP */
+	reg32 = pci_read_config32(PCH_ME_DEV, PCI_ME_HFS);
+	memcpy(&hfs, &reg32, sizeof(u32));
+
+	/* Abort and leave device alone if not normal mode */
+	if (hfs.fpt_bad ||
+	    hfs.working_state != ME_HFS_CWS_NORMAL ||
+	    hfs.operation_mode != ME_HFS_MODE_NORMAL)
+		return;
+
+	/* Try to send EOP command so ME stops accepting other commands */
+	mkhi_end_of_post();
+
+	/* Make sure IO is disabled */
+	reg32 = pci_read_config32(PCH_ME_DEV, PCI_COMMAND);
+	reg32 &= ~(PCI_COMMAND_MASTER |
+		   PCI_COMMAND_MEMORY | PCI_COMMAND_IO);
+	pci_write_config32(PCH_ME_DEV, PCI_COMMAND, reg32);
+
+	/* Hide the PCI device */
+	RCBA32_OR(FD2, PCH_DISABLE_MEI1);
+}
+
+void intel_me_finalize_smm(void)
+{
+	u32 did = pci_read_config32(PCH_ME_DEV, PCI_VENDOR_ID);
+	switch (did) {
+	case 0x1c3a8086:
+		intel_me7_finalize_smm();
+		break;
+	case 0x1e3a8086:
+		intel_me8_finalize_smm();
+		break;
+	default:
+		printk(BIOS_ERR, "No finalize handler for ME %08x.\n", did);
+	}
+}
+#else /* !__SMM__ */
+
+/* Determine the path that we should take based on ME status */
+static me_bios_path intel_me_path(device_t dev)
+{
+	me_bios_path path = ME_DISABLE_BIOS_PATH;
+	struct me_hfs hfs;
+	struct me_gmes gmes;
+
+#if CONFIG_HAVE_ACPI_RESUME
+	/* S3 wake skips all MKHI messages */
+	if (acpi_slp_type == 3) {
+		return ME_S3WAKE_BIOS_PATH;
+	}
+#endif
+
+	pci_read_dword_ptr(dev, &hfs, PCI_ME_HFS);
+	pci_read_dword_ptr(dev, &gmes, PCI_ME_GMES);
+
+	/* Check and dump status */
+	intel_me_status(&hfs, &gmes);
+
+	/* Check Current Working State */
+	switch (hfs.working_state) {
+	case ME_HFS_CWS_NORMAL:
+		path = ME_NORMAL_BIOS_PATH;
+		break;
+	case ME_HFS_CWS_REC:
+		path = ME_RECOVERY_BIOS_PATH;
+		break;
+	default:
+		path = ME_DISABLE_BIOS_PATH;
+		break;
+	}
+
+	/* Check Current Operation Mode */
+	switch (hfs.operation_mode) {
+	case ME_HFS_MODE_NORMAL:
+		break;
+	case ME_HFS_MODE_DEBUG:
+	case ME_HFS_MODE_DIS:
+	case ME_HFS_MODE_OVER_JMPR:
+	case ME_HFS_MODE_OVER_MEI:
+	default:
+		path = ME_DISABLE_BIOS_PATH;
+		break;
+	}
+
+	/* Check for any error code and valid firmware */
+	if (hfs.error_code || hfs.fpt_bad)
+		path = ME_ERROR_BIOS_PATH;
+
+#if CONFIG_ELOG
+	if (path != ME_NORMAL_BIOS_PATH) {
+		struct elog_event_data_me_extended data = {
+			.current_working_state = hfs.working_state,
+			.operation_state       = hfs.operation_state,
+			.operation_mode        = hfs.operation_mode,
+			.error_code            = hfs.error_code,
+			.progress_code         = gmes.progress_code,
+			.current_pmevent       = gmes.current_pmevent,
+			.current_state         = gmes.current_state,
+		};
+		elog_add_event_byte(ELOG_TYPE_MANAGEMENT_ENGINE, path);
+		elog_add_event_raw(ELOG_TYPE_MANAGEMENT_ENGINE_EXT,
+				   &data, sizeof(data));
+	}
+#endif
+
+	return path;
+}
+
+/* Prepare ME for MEI messages */
+static int intel_mei_setup(device_t dev)
+{
+	struct resource *res;
+	struct mei_csr host;
+	u32 reg32;
+
+	/* Find the MMIO base for the ME interface */
+	res = find_resource(dev, PCI_BASE_ADDRESS_0);
+	if (!res || res->base == 0 || res->size == 0) {
+		printk(BIOS_DEBUG, "ME: MEI resource not present!\n");
+		return -1;
+	}
+	mei_base_address = res->base;
+
+	/* Ensure Memory and Bus Master bits are set */
+	reg32 = pci_read_config32(dev, PCI_COMMAND);
+	reg32 |= PCI_COMMAND_MASTER | PCI_COMMAND_MEMORY;
+	pci_write_config32(dev, PCI_COMMAND, reg32);
+
+	/* Clean up status for next message */
+	read_host_csr(&host);
+	host.interrupt_generate = 1;
+	host.ready = 1;
+	host.reset = 0;
+	write_host_csr(&host);
+
+	return 0;
+}
+
+/* Read the Extend register hash of ME firmware */
+static int intel_me_extend_valid(device_t dev)
+{
+	struct me_heres status;
+	u32 extend[8] = {0};
+	int i, count = 0;
+
+	pci_read_dword_ptr(dev, &status, PCI_ME_HERES);
+	if (!status.extend_feature_present) {
+		printk(BIOS_ERR, "ME: Extend Feature not present\n");
+		return -1;
+	}
+
+	if (!status.extend_reg_valid) {
+		printk(BIOS_ERR, "ME: Extend Register not valid\n");
+		return -1;
+	}
+
+	switch (status.extend_reg_algorithm) {
+	case PCI_ME_EXT_SHA1:
+		count = 5;
+		printk(BIOS_DEBUG, "ME: Extend SHA-1: ");
+		break;
+	case PCI_ME_EXT_SHA256:
+		count = 8;
+		printk(BIOS_DEBUG, "ME: Extend SHA-256: ");
+		break;
+	default:
+		printk(BIOS_ERR, "ME: Extend Algorithm %d unknown\n",
+		       status.extend_reg_algorithm);
+		return -1;
+	}
+
+	for (i = 0; i < count; ++i) {
+		extend[i] = pci_read_config32(dev, PCI_ME_HER(i));
+		printk(BIOS_DEBUG, "%08x", extend[i]);
+	}
+	printk(BIOS_DEBUG, "\n");
+
+#if CONFIG_CHROMEOS
+	/* Save hash in NVS for the OS to verify */
+	chromeos_set_me_hash(extend, count);
+#endif
+
+	return 0;
+}
+
+/* Hide the ME virtual PCI devices */
+static void intel_me_hide(device_t dev)
+{
+	dev->enabled = 0;
+	pch_enable(dev);
+}
+
+/* Check whether ME is present and do basic init */
+static void intel_me_init(device_t dev)
+{
+	me_bios_path path = intel_me_path(dev);
+
+	/* Do initial setup and determine the BIOS path */
+	printk(BIOS_NOTICE, "ME: BIOS path: %s\n", me_bios_path_values[path]);
+
+	switch (path) {
+	case ME_S3WAKE_BIOS_PATH:
+		intel_me_hide(dev);
+		break;
+
+	case ME_NORMAL_BIOS_PATH:
+		/* Validate the extend register */
+		if (intel_me_extend_valid(dev) < 0)
+			break; /* TODO: force recovery mode */
+
+		/* Prepare MEI MMIO interface */
+		if (intel_mei_setup(dev) < 0)
+			break;
+
+#if (CONFIG_DEFAULT_CONSOLE_LOGLEVEL >= BIOS_DEBUG) && (CONFIG_NORTHBRIDGE_INTEL_SANDYBRIDGE || CONFIG_NORTHBRIDGE_INTEL_IVYBRIDGE)
+		/* Print ME firmware version */
+		mkhi_get_fw_version();
+		/* Print ME firmware capabilities */
+		mkhi_get_fwcaps();
+#endif
+
+		/*
+		 * Leave the ME unlocked in this path.
+		 * It will be locked via SMI command later.
+		 */
+		break;
+
+	case ME_ERROR_BIOS_PATH:
+	case ME_RECOVERY_BIOS_PATH:
+	case ME_DISABLE_BIOS_PATH:
+	case ME_FIRMWARE_UPDATE_BIOS_PATH:
+		break;
+	}
+}
+
+static void set_subsystem(device_t dev, unsigned vendor, unsigned device)
+{
+	if (!vendor || !device) {
+		pci_write_config32(dev, PCI_SUBSYSTEM_VENDOR_ID,
+			   pci_read_config32(dev, PCI_VENDOR_ID));
+	} else {
+		pci_write_config32(dev, PCI_SUBSYSTEM_VENDOR_ID,
+			   ((device & 0xffff) << 16) | (vendor & 0xffff));
+	}
+}
+
+static struct pci_operations pci_ops = {
+	.set_subsystem = set_subsystem,
+};
+
+static struct device_operations device_ops = {
+	.read_resources		= pci_dev_read_resources,
+	.set_resources		= pci_dev_set_resources,
+	.enable_resources	= pci_dev_enable_resources,
+	.init			= intel_me_init,
+	.scan_bus		= scan_static_bus,
+	.ops_pci		= &pci_ops,
+};
+
+static const unsigned short pci_device_ids[] = { 0x1c3a, 0x3b64,
+						 0 };
+
+
+static const struct pci_driver intel_me __pci_driver = {
+	.ops	= &device_ops,
+	.vendor	= PCI_VENDOR_ID_INTEL,
+	.devices	= pci_device_ids
+};
+
+#endif /* !__SMM__ */