AGESA F15: AGESA family15 model 00-0fh northbridge wrapper
authorKerry Sheh <shekairui@gmail.com>
Tue, 7 Feb 2012 12:31:40 +0000 (20:31 +0800)
committerMarc Jones <marcj303@gmail.com>
Thu, 16 Feb 2012 20:26:14 +0000 (21:26 +0100)
Change-Id: I87c4d47f19161c604b0285102bb3809c8337375a
Signed-off-by: Kerry Sheh <kerry.she@amd.com>
Signed-off-by: Kerry Sheh <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/556
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
src/northbridge/amd/agesa/Kconfig
src/northbridge/amd/agesa/Makefile.inc
src/northbridge/amd/agesa/family15/Kconfig [new file with mode: 0644]
src/northbridge/amd/agesa/family15/Makefile.inc [new file with mode: 0644]
src/northbridge/amd/agesa/family15/bootblock.c [new file with mode: 0644]
src/northbridge/amd/agesa/family15/chip.h [new file with mode: 0644]
src/northbridge/amd/agesa/family15/northbridge.c [new file with mode: 0644]
src/northbridge/amd/agesa/family15/northbridge.h [new file with mode: 0644]
src/northbridge/amd/agesa/family15/root_complex/Kconfig [new file with mode: 0644]
src/northbridge/amd/agesa/family15/root_complex/chip.h [new file with mode: 0644]

index 3bcb0bb74f4fdc0c569e2d9fc42d21be0ae2d18f..2ed9fd53095eac7462df0139ef91fc44aa51a1f3 100644 (file)
@@ -1,7 +1,7 @@
 #
 # This file is part of the coreboot project.
 #
-# Copyright (C) 2011 Advanced Micro Devices, Inc.
+# Copyright (C) 2011 - 2012 Advanced Micro Devices, Inc.
 #
 # 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
@@ -24,4 +24,5 @@ config CONSOLE_VGA_MULTI
 source src/northbridge/amd/agesa/family10/Kconfig
 source src/northbridge/amd/agesa/family12/Kconfig
 source src/northbridge/amd/agesa/family14/Kconfig
+source src/northbridge/amd/agesa/family15/Kconfig
 
index 1da8f60f17cfd819c408d011843eb9c9a8bbd147..eef1cd3cd166ca658d75af3079dc2f1856ffa90f 100644 (file)
@@ -1,7 +1,7 @@
 #
 # This file is part of the coreboot project.
 #
-# Copyright (C) 2011 Advanced Micro Devices, Inc.
+# Copyright (C) 2011 - 2012 Advanced Micro Devices, Inc.
 #
 # 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
@@ -19,3 +19,4 @@
 subdirs-$(CONFIG_NORTHBRIDGE_AMD_AGESA_FAMILY10) += family10
 subdirs-$(CONFIG_NORTHBRIDGE_AMD_AGESA_FAMILY12) += family12
 subdirs-$(CONFIG_NORTHBRIDGE_AMD_AGESA_FAMILY14) += family14
+subdirs-$(CONFIG_NORTHBRIDGE_AMD_AGESA_FAMILY15) += family15
diff --git a/src/northbridge/amd/agesa/family15/Kconfig b/src/northbridge/amd/agesa/family15/Kconfig
new file mode 100644 (file)
index 0000000..52f7a1e
--- /dev/null
@@ -0,0 +1,49 @@
+##
+## This file is part of the coreboot project.
+##
+## Copyright (C) 2007-2009 coresystems GmbH
+## Copyright (C) 2012 Advanced Micro Devices, Inc.
+##
+## 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
+##
+config NORTHBRIDGE_AMD_AGESA_FAMILY15
+       bool
+       select HAVE_DEBUG_RAM_SETUP
+       select HAVE_DEBUG_SMBUS
+       select HYPERTRANSPORT_PLUGIN_SUPPORT
+       select MMCONF_SUPPORT
+       select NORTHBRIDGE_AMD_AGESA_FAMILY15_ROOT_COMPLEX
+
+if NORTHBRIDGE_AMD_AGESA_FAMILY15
+config HT3_SUPPORT
+       bool
+       default y
+config HW_MEM_HOLE_SIZEK
+       hex
+       default 0x100000
+config HW_MEM_HOLE_SIZE_AUTO_INC
+       bool
+       default n
+config MMCONF_BASE_ADDRESS
+       hex
+       default 0xF8000000
+config MMCONF_BUS_NUMBER
+       int
+       default 64
+config BOOTBLOCK_NORTHBRIDGE_INIT
+  string
+  default "northbridge/amd/agesa/family15/bootblock.c"
+endif #NORTHBRIDGE_AMD_AGESA_FAMILY15
+
+source "src/northbridge/amd/agesa/family15/root_complex/Kconfig"
diff --git a/src/northbridge/amd/agesa/family15/Makefile.inc b/src/northbridge/amd/agesa/family15/Makefile.inc
new file mode 100644 (file)
index 0000000..255fe10
--- /dev/null
@@ -0,0 +1,20 @@
+#
+# This file is part of the coreboot project.
+#
+# Copyright (C) 2012 Advanced Micro Devices, Inc.
+#
+# 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
+#
+
+driver-y += northbridge.c
diff --git a/src/northbridge/amd/agesa/family15/bootblock.c b/src/northbridge/amd/agesa/family15/bootblock.c
new file mode 100644 (file)
index 0000000..fc62c3e
--- /dev/null
@@ -0,0 +1,25 @@
+/*
+ * This file is part of the coreboot project.
+ *
+ * Copyright (C) 2012 Advanced Micro Devices, Inc.
+ *
+ * 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
+ */
+
+#include <arch/io.h>
+#include <arch/romcc_io.h>
+#include <device/pci_def.h>
+
+static void bootblock_northbridge_init(void) {
+}
diff --git a/src/northbridge/amd/agesa/family15/chip.h b/src/northbridge/amd/agesa/family15/chip.h
new file mode 100644 (file)
index 0000000..cec1fc4
--- /dev/null
@@ -0,0 +1,24 @@
+/*
+ * This file is part of the coreboot project.
+ *
+ * Copyright (C) 2012 Advanced Micro Devices, Inc.
+ *
+ * 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
+ */
+
+struct northbridge_amd_agesa_family15_config
+{
+};
+
+extern struct chip_operations northbridge_amd_agesa_family15_ops;
diff --git a/src/northbridge/amd/agesa/family15/northbridge.c b/src/northbridge/amd/agesa/family15/northbridge.c
new file mode 100644 (file)
index 0000000..b0fc91d
--- /dev/null
@@ -0,0 +1,1168 @@
+/*
+ * This file is part of the coreboot project.
+ *
+ * Copyright (C) 2012 Advanced Micro Devices, Inc.
+ *
+ * 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
+ */
+
+#include <console/console.h>
+#include <arch/io.h>
+#include <stdint.h>
+#include <device/device.h>
+#include <device/pci.h>
+#include <device/pci_ids.h>
+#include <device/hypertransport.h>
+#include <stdlib.h>
+#include <string.h>
+#include <bitops.h>
+#include <cpu/cpu.h>
+
+#include <cpu/x86/lapic.h>
+
+#include <Porting.h>
+#include <AGESA.h>
+#include <Options.h>
+#include <Topology.h>
+#include <cpu/amd/amdfam15.h>
+#include <cpuRegisters.h>
+#include "agesawrapper.h"
+#include "root_complex/chip.h"
+#include "northbridge.h"
+#include "chip.h"
+
+#define MAX_NODE_NUMS (MAX_NODES * MAX_DIES)
+
+#if (defined CONFIG_EXT_CONF_SUPPORT) && CONFIG_EXT_CONF_SUPPORT == 1
+#error CONFIG_EXT_CONF_SUPPORT == 1 not support anymore!
+#endif
+
+typedef struct dram_base_mask {
+       u32 base; //[47:27] at [28:8]
+       u32 mask; //[47:27] at [28:8] and enable at bit 0
+} dram_base_mask_t;
+
+static unsigned node_nums;
+static unsigned sblink;
+static device_t __f0_dev[MAX_NODE_NUMS];
+static device_t __f1_dev[MAX_NODE_NUMS];
+static device_t __f2_dev[MAX_NODE_NUMS];
+static device_t __f4_dev[MAX_NODE_NUMS];
+static unsigned fx_devs = 0;
+
+
+static dram_base_mask_t get_dram_base_mask(u32 nodeid)
+{
+       device_t dev;
+       dram_base_mask_t d;
+       dev = __f1_dev[0];
+       u32 temp;
+       temp = pci_read_config32(dev, 0x44 + (nodeid << 3)); //[39:24] at [31:16]
+       d.mask = ((temp & 0xfff80000)>>(8+3)); // mask out  DramMask [26:24] too
+       temp = pci_read_config32(dev, 0x144 + (nodeid <<3)) & 0xff; //[47:40] at [7:0]
+       d.mask |= temp<<21;
+       temp = pci_read_config32(dev, 0x40 + (nodeid << 3)); //[39:24] at [31:16]
+       d.mask |= (temp & 1); // enable bit
+       d.base = ((temp & 0xfff80000)>>(8+3)); // mask out DramBase [26:24) too
+       temp = pci_read_config32(dev, 0x140 + (nodeid <<3)) & 0xff; //[47:40] at [7:0]
+       d.base |= temp<<21;
+       return d;
+}
+
+static void set_io_addr_reg(device_t dev, u32 nodeid, u32 linkn, u32 reg,
+                       u32 io_min, u32 io_max)
+{
+       u32 i;
+       u32 tempreg;
+       /* io range allocation */
+       tempreg = (nodeid&0xf) | ((nodeid & 0x30)<<(8-4)) | (linkn<<4) | ((io_max&0xf0)<<(12-4)); //limit
+       for (i=0; i<node_nums; i++)
+               pci_write_config32(__f1_dev[i], reg+4, tempreg);
+       tempreg = 3 /*| ( 3<<4)*/ | ((io_min&0xf0)<<(12-4));          //base :ISA and VGA ?
+#if 0
+       // FIXME: can we use VGA reg instead?
+       if (dev->link[link].bridge_ctrl & PCI_BRIDGE_CTL_VGA) {
+               printk(BIOS_SPEW, "%s, enabling legacy VGA IO forwarding for %s link %s\n",
+                               __func__, dev_path(dev), link);
+               tempreg |= PCI_IO_BASE_VGA_EN;
+       }
+       if (dev->link[link].bridge_ctrl & PCI_BRIDGE_CTL_NO_ISA) {
+               tempreg |= PCI_IO_BASE_NO_ISA;
+       }
+#endif
+       for (i=0; i<node_nums; i++)
+               pci_write_config32(__f1_dev[i], reg, tempreg);
+}
+
+static void set_mmio_addr_reg(u32 nodeid, u32 linkn, u32 reg, u32 index, u32 mmio_min, u32 mmio_max, u32 nodes)
+{
+       u32 i;
+       u32 tempreg;
+       /* io range allocation */
+       tempreg = (nodeid&0xf) | (linkn<<4) |    (mmio_max&0xffffff00); //limit
+       for (i=0; i<nodes; i++)
+               pci_write_config32(__f1_dev[i], reg+4, tempreg);
+       tempreg = 3 | (nodeid & 0x30) | (mmio_min&0xffffff00);
+       for (i=0; i<node_nums; i++)
+               pci_write_config32(__f1_dev[i], reg, tempreg);
+}
+
+static device_t get_node_pci(u32 nodeid, u32 fn)
+{
+#if MAX_NODE_NUMS == 64
+       if (nodeid < 32) {
+               return dev_find_slot(CONFIG_CBB, PCI_DEVFN(CONFIG_CDB + nodeid, fn));
+       } else {
+               return dev_find_slot(CONFIG_CBB-1, PCI_DEVFN(CONFIG_CDB + nodeid - 32, fn));
+       }
+#else
+       return dev_find_slot(CONFIG_CBB, PCI_DEVFN(CONFIG_CDB + nodeid, fn));
+#endif
+}
+
+static void get_fx_devs(void)
+{
+       int i;
+       for (i = 0; i < MAX_NODE_NUMS; i++) {
+               __f0_dev[i] = get_node_pci(i, 0);
+               __f1_dev[i] = get_node_pci(i, 1);
+               __f2_dev[i] = get_node_pci(i, 2);
+               __f4_dev[i] = get_node_pci(i, 4);
+               if (__f0_dev[i] != NULL && __f1_dev[i] != NULL)
+                       fx_devs = i+1;
+       }
+       if (__f1_dev[0] == NULL || __f0_dev[0] == NULL || fx_devs == 0) {
+               die("Cannot find 0:0x18.[0|1]\n");
+       }
+       printk(BIOS_DEBUG, "fx_devs=0x%x\n", fx_devs);
+}
+
+static u32 f1_read_config32(unsigned reg)
+{
+       if (fx_devs == 0)
+               get_fx_devs();
+       return pci_read_config32(__f1_dev[0], reg);
+}
+
+static void f1_write_config32(unsigned reg, u32 value)
+{
+       int i;
+       if (fx_devs == 0)
+               get_fx_devs();
+       for(i = 0; i < fx_devs; i++) {
+               device_t dev;
+               dev = __f1_dev[i];
+               if (dev && dev->enabled) {
+                       pci_write_config32(dev, reg, value);
+               }
+       }
+}
+
+static u32 amdfam15_nodeid(device_t dev)
+{
+#if MAX_NODE_NUMS == 64
+       unsigned busn;
+       busn = dev->bus->secondary;
+       if (busn != CONFIG_CBB) {
+               return (dev->path.pci.devfn >> 3) - CONFIG_CDB + 32;
+       } else {
+               return (dev->path.pci.devfn >> 3) - CONFIG_CDB;
+       }
+
+#else
+       return (dev->path.pci.devfn >> 3) - CONFIG_CDB;
+#endif
+}
+
+static void set_vga_enable_reg(u32 nodeid, u32 linkn)
+{
+       u32 val;
+
+       val =  1 | (nodeid<<4) | (linkn<<12);
+       /* it will routing
+        * (1)mmio 0xa0000:0xbffff
+        * (2)io   0x3b0:0x3bb, 0x3c0:0x3df
+        */
+       f1_write_config32(0xf4, val);
+
+}
+
+/**
+ * @return
+ *  @retval 2  resoure not exist, usable
+ *  @retval 0  resource exist, not usable
+ *  @retval 1  resource exist, resource has been allocated before
+ */
+static int reg_useable(unsigned reg, device_t goal_dev, unsigned goal_nodeid,
+                       unsigned goal_link)
+{
+       struct resource *res;
+       unsigned nodeid, link = 0;
+       int result;
+       res = 0;
+       for (nodeid = 0; !res && (nodeid < fx_devs); nodeid++) {
+               device_t dev;
+               dev = __f0_dev[nodeid];
+               if (!dev)
+                       continue;
+               for (link = 0; !res && (link < 8); link++) {
+                       res = probe_resource(dev, IOINDEX(0x1000 + reg, link));
+               }
+       }
+       result = 2;
+       if (res) {
+               result = 0;
+               if ((goal_link == (link - 1)) &&
+                       (goal_nodeid == (nodeid - 1)) &&
+                       (res->flags <= 1)) {
+                       result = 1;
+               }
+       }
+       return result;
+}
+
+static struct resource *amdfam15_find_iopair(device_t dev, unsigned nodeid, unsigned link)
+{
+       struct resource *resource;
+       u32 free_reg, reg;
+       resource = 0;
+       free_reg = 0;
+       for (reg = 0xc0; reg <= 0xd8; reg += 0x8) {
+               int result;
+               result = reg_useable(reg, dev, nodeid, link);
+               if (result == 1) {
+                       /* I have been allocated this one */
+                       break;
+               }
+               else if (result > 1) {
+                       /* I have a free register pair */
+                       free_reg = reg;
+               }
+       }
+       if (reg > 0xd8) {
+               reg = free_reg; // if no free, the free_reg still be 0
+       }
+
+       resource = new_resource(dev, IOINDEX(0x1000 + reg, link));
+
+       return resource;
+}
+
+static struct resource *amdfam15_find_mempair(device_t dev, u32 nodeid, u32 link)
+{
+       struct resource *resource;
+       u32 free_reg, reg;
+       resource = 0;
+       free_reg = 0;
+       for (reg = 0x80; reg <= 0xb8; reg += 0x8) {
+               int result;
+               result = reg_useable(reg, dev, nodeid, link);
+               if (result == 1) {
+                       /* I have been allocated this one */
+                       break;
+               }
+               else if (result > 1) {
+                       /* I have a free register pair */
+                       free_reg = reg;
+               }
+       }
+       if (reg > 0xb8) {
+               reg = free_reg;
+       }
+
+       resource = new_resource(dev, IOINDEX(0x1000 + reg, link));
+       return resource;
+}
+
+
+static void amdfam15_link_read_bases(device_t dev, u32 nodeid, u32 link)
+{
+       struct resource *resource;
+
+       /* Initialize the io space constraints on the current bus */
+       resource = amdfam15_find_iopair(dev, nodeid, link);
+       if (resource) {
+               u32 align;
+               align = log2(HT_IO_HOST_ALIGN);
+               resource->base  = 0;
+               resource->size  = 0;
+               resource->align = align;
+               resource->gran  = align;
+               resource->limit = 0xffffUL;
+               resource->flags = IORESOURCE_IO | IORESOURCE_BRIDGE;
+       }
+
+       /* Initialize the prefetchable memory constraints on the current bus */
+       resource = amdfam15_find_mempair(dev, nodeid, link);
+       if (resource) {
+               resource->base = 0;
+               resource->size = 0;
+               resource->align = log2(HT_MEM_HOST_ALIGN);
+               resource->gran = log2(HT_MEM_HOST_ALIGN);
+               resource->limit = 0xffffffffffULL;
+               resource->flags = IORESOURCE_MEM | IORESOURCE_PREFETCH;
+               resource->flags |= IORESOURCE_BRIDGE;
+       }
+
+
+       /* Initialize the memory constraints on the current bus */
+       resource = amdfam15_find_mempair(dev, nodeid, link);
+       if (resource) {
+               resource->base = 0;
+               resource->size = 0;
+               resource->align = log2(HT_MEM_HOST_ALIGN);
+               resource->gran = log2(HT_MEM_HOST_ALIGN);
+               resource->limit = 0xffffffffffULL;
+               resource->flags = IORESOURCE_MEM | IORESOURCE_BRIDGE;
+       }
+
+}
+
+
+static void read_resources(device_t dev)
+{
+       u32 nodeid;
+       struct bus *link;
+
+       nodeid = amdfam15_nodeid(dev);
+       for (link = dev->link_list; link; link = link->next) {
+               if (link->children) {
+                       amdfam15_link_read_bases(dev, nodeid, link->link_num);
+               }
+       }
+}
+
+
+static void set_resource(device_t dev, struct resource *resource, u32 nodeid)
+{
+       resource_t rbase, rend;
+       unsigned reg, link_num;
+       char buf[50];
+
+
+       /* Make certain the resource has actually been set */
+       if (!(resource->flags & IORESOURCE_ASSIGNED)) {
+               return;
+       }
+
+       /* If I have already stored this resource don't worry about it */
+       if (resource->flags & IORESOURCE_STORED) {
+               return;
+       }
+
+       /* Only handle PCI memory and IO resources */
+       if (!(resource->flags & (IORESOURCE_MEM | IORESOURCE_IO)))
+               return;
+
+       /* Ensure I am actually looking at a resource of function 1 */
+       if ((resource->index & 0xffff) < 0x1000) {
+               return;
+       }
+       /* Get the base address */
+       rbase = resource->base;
+
+       /* Get the limit (rounded up) */
+       rend  = resource_end(resource);
+
+       /* Get the register and link */
+       reg  = resource->index & 0xfff; // 4k
+       link_num = IOINDEX_LINK(resource->index);
+
+       if (resource->flags & IORESOURCE_IO) {
+               set_io_addr_reg(dev, nodeid, link_num, reg, rbase>>8, rend>>8);
+       }
+       else if (resource->flags & IORESOURCE_MEM) {
+               set_mmio_addr_reg(nodeid, link_num, reg, (resource->index >>24), rbase>>8, rend>>8, node_nums) ;// [39:8]
+       }
+       resource->flags |= IORESOURCE_STORED;
+       sprintf(buf, " <node %x link %x>",
+                       nodeid, link_num);
+       report_resource_stored(dev, resource, buf);
+}
+
+/**
+ * I tried to reuse the resource allocation code in set_resource()
+ * but it is too difficult to deal with the resource allocation magic.
+ */
+
+static void create_vga_resource(device_t dev, unsigned nodeid)
+{
+       struct bus *link;
+
+
+       /* find out which link the VGA card is connected,
+        * we only deal with the 'first' vga card */
+       for (link = dev->link_list; link; link = link->next) {
+               if (link->bridge_ctrl & PCI_BRIDGE_CTL_VGA) {
+#if CONFIG_MULTIPLE_VGA_ADAPTERS == 1
+                       extern device_t vga_pri; // the primary vga device, defined in device.c
+                       printk(BIOS_DEBUG, "VGA: vga_pri bus num = %d bus range [%d,%d]\n", vga_pri->bus->secondary,
+                                       link->secondary,link->subordinate);
+                       /* We need to make sure the vga_pri is under the link */
+                       if((vga_pri->bus->secondary >= link->secondary ) &&
+                                       (vga_pri->bus->secondary <= link->subordinate )
+                         )
+#endif
+                               break;
+               }
+       }
+
+       /* no VGA card installed */
+       if (link == NULL)
+               return;
+
+       printk(BIOS_DEBUG, "VGA: %s (aka node %d) link %d has VGA device\n", dev_path(dev), nodeid, sblink);
+       set_vga_enable_reg(nodeid, sblink);
+}
+
+
+static void set_resources(device_t dev)
+{
+       unsigned nodeid;
+       struct bus *bus;
+       struct resource *res;
+
+       /* Find the nodeid */
+       nodeid = amdfam15_nodeid(dev);
+
+       create_vga_resource(dev, nodeid); //TODO: do we need this?
+
+       /* Set each resource we have found */
+       for (res = dev->resource_list; res; res = res->next) {
+               set_resource(dev, res, nodeid);
+       }
+
+       for (bus = dev->link_list; bus; bus = bus->next) {
+               if (bus->children) {
+                       assign_resources(bus);
+               }
+       }
+}
+
+static void northbridge_init(struct device *dev)
+{
+}
+
+static unsigned scan_chains(device_t dev, unsigned max)
+{
+       unsigned nodeid;
+       struct bus *link;
+       device_t io_hub = NULL;
+       u32 next_unitid = 0x18;
+       nodeid = amdfam15_nodeid(dev);
+       if (nodeid == 0) {
+               for (link = dev->link_list; link; link = link->next) {
+                       //if (link->link_num == sblink) { /* devicetree put IO Hub on link_lsit[sblink] */
+                       if (link->link_num == 0) { /* devicetree put IO Hub on link_lsit[0] */
+                               io_hub = link->children;
+                               if (!io_hub || !io_hub->enabled) {
+                                       die("I can't find the IO Hub, or IO Hub not enabled, please check the device tree.\n");
+                               }
+                               /* Now that nothing is overlapping it is safe to scan the children. */
+                               max = pci_scan_bus(link, 0x00, ((next_unitid - 1) << 3) | 7, 0);
+                       }
+               }
+       }
+       return max;
+}
+
+static struct device_operations northbridge_operations = {
+       .read_resources   = read_resources,
+       .set_resources    = set_resources,
+       .enable_resources = pci_dev_enable_resources,
+       .init             = northbridge_init,
+       .scan_bus         = scan_chains,
+       .enable           = 0,
+       .ops_pci          = 0,
+};
+
+static const struct pci_driver family15_northbridge __pci_driver = {
+       .ops    = &northbridge_operations,
+       .vendor = PCI_VENDOR_ID_AMD,
+       .device = PCI_DEVICE_ID_AMD_15H_MODEL_000F_NB_HT,
+};
+
+static const struct pci_driver family10_northbridge __pci_driver = {
+       .ops    = &northbridge_operations,
+       .vendor = PCI_VENDOR_ID_AMD,
+       .device = PCI_DEVICE_ID_AMD_10H_NB_HT,
+};
+
+struct chip_operations northbridge_amd_agesa_family15_ops = {
+       CHIP_NAME("AMD FAM15 Northbridge")
+       .enable_dev = 0,
+};
+
+static void domain_read_resources(device_t dev)
+{
+       unsigned reg;
+
+
+       /* Find the already assigned resource pairs */
+       get_fx_devs();
+       for (reg = 0x80; reg <= 0xd8; reg+= 0x08) {
+               u32 base, limit;
+               base  = f1_read_config32(reg);
+               limit = f1_read_config32(reg + 0x04);
+               /* Is this register allocated? */
+               if ((base & 3) != 0) {
+                       unsigned nodeid, reg_link;
+                       device_t reg_dev;
+                       if (reg<0xc0) { // mmio
+                               nodeid = (limit & 0xf) + (base&0x30);
+                       } else { // io
+                               nodeid =  (limit & 0xf) + ((base>>4)&0x30);
+                       }
+                       reg_link = (limit >> 4) & 7;
+                       reg_dev = __f0_dev[nodeid];
+                       if (reg_dev) {
+                               /* Reserve the resource  */
+                               struct resource *res;
+                               res = new_resource(reg_dev, IOINDEX(0x1000 + reg, reg_link));
+                               if (res) {
+                                       res->flags = 1;
+                               }
+                       }
+               }
+       }
+       /* FIXME: do we need to check extend conf space?
+          I don't believe that much preset value */
+
+#if CONFIG_PCI_64BIT_PREF_MEM == 0
+       pci_domain_read_resources(dev);
+
+
+#else
+       struct bus *link;
+       struct resource *resource;
+       for (link=dev->link_list; link; link = link->next) {
+               /* Initialize the system wide io space constraints */
+               resource = new_resource(dev, 0|(link->link_num<<2));
+               resource->base  = 0x400;
+               resource->limit = 0xffffUL;
+               resource->flags = IORESOURCE_IO;
+
+               /* Initialize the system wide prefetchable memory resources constraints */
+               resource = new_resource(dev, 1|(link->link_num<<2));
+               resource->limit = 0xfcffffffffULL;
+               resource->flags = IORESOURCE_MEM | IORESOURCE_PREFETCH;
+
+               /* Initialize the system wide memory resources constraints */
+               resource = new_resource(dev, 2|(link->link_num<<2));
+               resource->limit = 0xfcffffffffULL;
+               resource->flags = IORESOURCE_MEM;
+       }
+#endif
+}
+
+static void domain_enable_resources(device_t dev)
+{
+       u32 val;
+       /* Must be called after PCI enumeration and resource allocation */
+       printk(BIOS_DEBUG, "\nFam15 - domain_enable_resources: AmdInitMid.\n");
+       val = agesawrapper_amdinitmid();
+       if (val) {
+               printk(BIOS_DEBUG, "agesawrapper_amdinitmid failed: %x \n", val);
+       }
+       printk(BIOS_DEBUG, "  ader - leaving domain_enable_resources.\n");
+}
+
+
+#if CONFIG_HW_MEM_HOLE_SIZEK != 0
+struct hw_mem_hole_info {
+       unsigned hole_startk;
+       int node_id;
+};
+static struct hw_mem_hole_info get_hw_mem_hole_info(void)
+{
+       struct hw_mem_hole_info mem_hole;
+       int i;
+       mem_hole.hole_startk = CONFIG_HW_MEM_HOLE_SIZEK;
+       mem_hole.node_id = -1;
+       for (i = 0; i < node_nums; i++) {
+               dram_base_mask_t d;
+               u32 hole;
+               d = get_dram_base_mask(i);
+               if (!(d.mask & 1)) continue; // no memory on this node
+               hole = pci_read_config32(__f1_dev[i], 0xf0);
+               if (hole & 1) { // we find the hole
+                       mem_hole.hole_startk = (hole & (0xff<<24)) >> 10;
+                       mem_hole.node_id = i; // record the node No with hole
+                       break; // only one hole
+               }
+       }
+       //We need to double check if there is speical set on base reg and limit reg are not continous instead of hole, it will find out it's hole_startk
+       if (mem_hole.node_id == -1) {
+               resource_t limitk_pri = 0;
+               for (i=0; i<node_nums; i++) {
+                       dram_base_mask_t d;
+                       resource_t base_k, limit_k;
+                       d = get_dram_base_mask(i);
+                       if (!(d.base & 1)) continue;
+                       base_k = ((resource_t)(d.base & 0x1fffff00)) <<9;
+                       if (base_k > 4 *1024 * 1024) break; // don't need to go to check
+                       if (limitk_pri != base_k) { // we find the hole
+                               mem_hole.hole_startk = (unsigned)limitk_pri; // must beblow 4G
+                               mem_hole.node_id = i;
+                               break; //only one hole
+                       }
+                       limit_k = ((resource_t)((d.mask + 0x00000100) & 0x1fffff00)) << 9;
+                       limitk_pri = limit_k;
+               }
+       }
+       return mem_hole;
+}
+#endif
+#if CONFIG_WRITE_HIGH_TABLES==1
+#define HIGH_TABLES_SIZE 64    // maximum size of high tables in KB
+extern uint64_t high_tables_base, high_tables_size;
+#endif
+#if CONFIG_GFXUMA == 1
+extern uint64_t uma_memory_base, uma_memory_size;
+static void add_uma_resource(struct device *dev, int index)
+{
+       struct resource *resource;
+
+       printk(BIOS_DEBUG, "Adding UMA memory area\n");
+       resource = new_resource(dev, index);
+       resource->base = (resource_t) uma_memory_base;
+       resource->size = (resource_t) uma_memory_size;
+       resource->flags = IORESOURCE_MEM | IORESOURCE_RESERVE |
+                       IORESOURCE_FIXED | IORESOURCE_STORED |
+                       IORESOURCE_ASSIGNED;
+}
+#endif
+
+static void domain_set_resources(device_t dev)
+{
+#if CONFIG_PCI_64BIT_PREF_MEM == 1
+       struct resource *io, *mem1, *mem2;
+       struct resource *res;
+#endif
+       unsigned long mmio_basek;
+       u32 pci_tolm;
+       int i, idx;
+       struct bus *link;
+#if CONFIG_HW_MEM_HOLE_SIZEK != 0
+       struct hw_mem_hole_info mem_hole;
+       u32 reset_memhole = 1;
+#endif
+
+#if CONFIG_PCI_64BIT_PREF_MEM == 1
+
+       for (link = dev->link_list; link; link = link->next) {
+               /* Now reallocate the pci resources memory with the
+                * highest addresses I can manage.
+                */
+               mem1 = find_resource(dev, 1|(link->link_num<<2));
+               mem2 = find_resource(dev, 2|(link->link_num<<2));
+
+               printk(BIOS_DEBUG, "base1: 0x%08Lx limit1: 0x%08Lx size: 0x%08Lx align: %d\n",
+                               mem1->base, mem1->limit, mem1->size, mem1->align);
+               printk(BIOS_DEBUG, "base2: 0x%08Lx limit2: 0x%08Lx size: 0x%08Lx align: %d\n",
+                               mem2->base, mem2->limit, mem2->size, mem2->align);
+
+               /* See if both resources have roughly the same limits */
+               if (((mem1->limit <= 0xffffffff) && (mem2->limit <= 0xffffffff)) ||
+                               ((mem1->limit > 0xffffffff) && (mem2->limit > 0xffffffff)))
+               {
+                       /* If so place the one with the most stringent alignment first */
+                       if (mem2->align > mem1->align) {
+                               struct resource *tmp;
+                               tmp = mem1;
+                               mem1 = mem2;
+                               mem2 = tmp;
+                       }
+                       /* Now place the memory as high up as it will go */
+                       mem2->base = resource_max(mem2);
+                       mem1->limit = mem2->base - 1;
+                       mem1->base = resource_max(mem1);
+               }
+               else {
+                       /* Place the resources as high up as they will go */
+                       mem2->base = resource_max(mem2);
+                       mem1->base = resource_max(mem1);
+               }
+
+               printk(BIOS_DEBUG, "base1: 0x%08Lx limit1: 0x%08Lx size: 0x%08Lx align: %d\n",
+                               mem1->base, mem1->limit, mem1->size, mem1->align);
+               printk(BIOS_DEBUG, "base2: 0x%08Lx limit2: 0x%08Lx size: 0x%08Lx align: %d\n",
+                               mem2->base, mem2->limit, mem2->size, mem2->align);
+       }
+
+       for (res = &dev->resource_list; res; res = res->next)
+       {
+               res->flags |= IORESOURCE_ASSIGNED;
+               res->flags |= IORESOURCE_STORED;
+               report_resource_stored(dev, res, "");
+       }
+#endif
+
+       pci_tolm = 0xffffffffUL;
+       for (link = dev->link_list; link; link = link->next) {
+               pci_tolm = find_pci_tolm(link);
+       }
+
+       // FIXME handle interleaved nodes. If you fix this here, please fix
+       // amdk8, too.
+       mmio_basek = pci_tolm >> 10;
+       /* Round mmio_basek to something the processor can support */
+       mmio_basek &= ~((1 << 6) -1);
+
+       // FIXME improve mtrr.c so we don't use up all of the mtrrs with a 64M
+       // MMIO hole. If you fix this here, please fix amdk8, too.
+       /* Round the mmio hole to 64M */
+       mmio_basek &= ~((64*1024) - 1);
+
+#if CONFIG_HW_MEM_HOLE_SIZEK != 0
+       /* if the hw mem hole is already set in raminit stage, here we will compare
+        * mmio_basek and hole_basek. if mmio_basek is bigger that hole_basek and will
+        * use hole_basek as mmio_basek and we don't need to reset hole.
+        * otherwise We reset the hole to the mmio_basek
+        */
+
+       mem_hole = get_hw_mem_hole_info();
+
+       // Use hole_basek as mmio_basek, and we don't need to reset hole anymore
+       if ((mem_hole.node_id !=  -1) && (mmio_basek > mem_hole.hole_startk)) {
+               mmio_basek = mem_hole.hole_startk;
+               reset_memhole = 0;
+       }
+#endif
+
+       idx = 0x10;
+       for (i = 0; i < node_nums; i++) {
+               dram_base_mask_t d;
+               resource_t basek, limitk, sizek; // 4 1T
+
+               d = get_dram_base_mask(i);
+
+               if (!(d.mask & 1)) continue;
+               basek = ((resource_t)(d.base & 0x1fffff00)) << 9; // could overflow, we may lost 6 bit here
+               limitk = ((resource_t)((d.mask + 0x00000100) & 0x1fffff00)) << 9 ;
+
+               sizek = limitk - basek;
+
+
+               /* see if we need a hole from 0xa0000 to 0xbffff */
+               if ((basek < ((8*64)+(8*16))) && (sizek > ((8*64)+(16*16)))) {
+                       ram_resource(dev, (idx | i), basek, ((8*64)+(8*16)) - basek);
+                       idx += 0x10;
+                       basek = (8*64)+(16*16);
+                       sizek = limitk - ((8*64)+(16*16));
+
+               }
+
+               //printk(BIOS_DEBUG, "node %d : mmio_basek=%08lx, basek=%08llx, limitk=%08llx\n", i, mmio_basek, basek, limitk);
+
+               /* split the region to accomodate pci memory space */
+               if ((basek < 4*1024*1024 ) && (limitk > mmio_basek)) {
+                       if (basek <= mmio_basek) {
+                               unsigned pre_sizek;
+                               pre_sizek = mmio_basek - basek;
+                               if (pre_sizek>0) {
+                                       ram_resource(dev, (idx | i), basek, pre_sizek);
+                                       idx += 0x10;
+                                       sizek -= pre_sizek;
+#if CONFIG_WRITE_HIGH_TABLES==1
+                                       if (high_tables_base==0) {
+                                               /* Leave some space for ACPI, PIRQ and MP tables */
+#if CONFIG_GFXUMA == 1
+                                               high_tables_base = uma_memory_base - (HIGH_TABLES_SIZE * 1024);
+#else
+                                               high_tables_base = (mmio_basek - HIGH_TABLES_SIZE) * 1024;
+#endif
+                                               high_tables_size = HIGH_TABLES_SIZE * 1024;
+                                               printk(BIOS_DEBUG, " split: %dK table at =%08llx\n",
+                                                        HIGH_TABLES_SIZE, high_tables_base);
+                                       }
+#endif
+                               }
+                               basek = mmio_basek;
+                       }
+                       if ((basek + sizek) <= 4*1024*1024) {
+                               sizek = 0;
+                       }
+                       else {
+                               basek = 4*1024*1024;
+                               sizek -= (4*1024*1024 - mmio_basek);
+                       }
+               }
+
+#if CONFIG_GFXUMA == 1
+               /* Deduct uma memory before reporting because
+                * this is what the mtrr code expects */
+               sizek -= uma_memory_size / 1024;
+#endif
+               ram_resource(dev, (idx | i), basek, sizek);
+               idx += 0x10;
+#if CONFIG_WRITE_HIGH_TABLES==1
+               printk(BIOS_DEBUG, "node %d: mmio_basek=%08lx, basek=%08llx, limitk=%08llx\n",
+                               i, mmio_basek, basek, limitk);
+               if (high_tables_base==0) {
+                       /* Leave some space for ACPI, PIRQ and MP tables */
+#if CONFIG_GFXUMA == 1
+                       high_tables_base = uma_memory_base - (HIGH_TABLES_SIZE * 1024);
+#else
+                       high_tables_base = (limitk - HIGH_TABLES_SIZE) * 1024;
+#endif
+                       high_tables_size = HIGH_TABLES_SIZE * 1024;
+               }
+#endif
+       }
+
+#if CONFIG_GFXUMA == 1
+       add_uma_resource(dev, 7);
+#endif
+
+       for(link = dev->link_list; link; link = link->next) {
+               if (link->children) {
+                       assign_resources(link);
+               }
+       }
+}
+
+
+static struct device_operations pci_domain_ops = {
+       .read_resources   = domain_read_resources,
+       .set_resources    = domain_set_resources,
+       .enable_resources = domain_enable_resources,
+       .init             = NULL,
+       .scan_bus         = pci_domain_scan_bus,
+
+#if CONFIG_MMCONF_SUPPORT_DEFAULT
+       .ops_pci_bus      = &pci_ops_mmconf,
+#else
+       .ops_pci_bus      = &pci_cf8_conf1,
+#endif
+};
+
+
+static void sysconf_init(device_t dev) // first node
+{
+       sblink = (pci_read_config32(dev, 0x64)>>8) & 7; // don't forget sublink1
+       node_nums = ((pci_read_config32(dev, 0x60)>>4) & 7) + 1; //NodeCnt[2:0]
+}
+
+static void add_more_links(device_t dev, unsigned total_links)
+{
+       struct bus *link, *last = NULL;
+       int link_num;
+
+       for (link = dev->link_list; link; link = link->next)
+               last = link;
+
+       if (last) {
+               int links = total_links - last->link_num;
+               link_num = last->link_num;
+               if (links > 0) {
+                       link = malloc(links*sizeof(*link));
+                       if (!link)
+                               die("Couldn't allocate more links!\n");
+                       memset(link, 0, links*sizeof(*link));
+                       last->next = link;
+               }
+       }
+       else {
+               link_num = -1;
+               link = malloc(total_links*sizeof(*link));
+               memset(link, 0, total_links*sizeof(*link));
+               dev->link_list = link;
+       }
+
+       for (link_num = link_num + 1; link_num < total_links; link_num++) {
+               link->link_num = link_num;
+               link->dev = dev;
+               link->next = link + 1;
+               last = link;
+               link = link->next;
+       }
+       last->next = NULL;
+}
+
+/* dummy read_resources */
+static void lapic_read_resources(device_t dev)
+{
+}
+
+static struct device_operations lapic_ops = {
+       .read_resources   = lapic_read_resources,
+       .set_resources    = pci_dev_set_resources,
+       .enable_resources = pci_dev_enable_resources,
+       .init             = 0,
+       .scan_bus         = 0,
+       .enable           = 0,
+       .ops_pci          = 0,
+};
+
+static u32 cpu_bus_scan(device_t dev, u32 max)
+{
+       struct bus *cpu_bus;
+       device_t dev_mc;
+#if CONFIG_CBB
+       device_t pci_domain;
+#endif
+       int i,j;
+       int coreid_bits;
+       int core_max = 0;
+       unsigned ApicIdCoreIdSize;
+       unsigned core_nums;
+       int siblings = 0;
+       unsigned int family;
+
+#if CONFIG_CBB
+       dev_mc = dev_find_slot(0, PCI_DEVFN(CONFIG_CDB, 0)); //0x00
+       if (dev_mc && dev_mc->bus) {
+               printk(BIOS_DEBUG, "%s found", dev_path(dev_mc));
+               pci_domain = dev_mc->bus->dev;
+               if (pci_domain && (pci_domain->path.type == DEVICE_PATH_PCI_DOMAIN)) {
+                       printk(BIOS_DEBUG, "\n%s move to ",dev_path(dev_mc));
+                       dev_mc->bus->secondary = CONFIG_CBB; // move to 0xff
+                       printk(BIOS_DEBUG, "%s",dev_path(dev_mc));
+               } else {
+                       printk(BIOS_DEBUG, " but it is not under pci_domain directly ");
+               }
+               printk(BIOS_DEBUG, "\n");
+       }
+       dev_mc = dev_find_slot(CONFIG_CBB, PCI_DEVFN(CONFIG_CDB, 0));
+       if (!dev_mc) {
+               dev_mc = dev_find_slot(0, PCI_DEVFN(0x18, 0));
+               if (dev_mc && dev_mc->bus) {
+                       printk(BIOS_DEBUG, "%s found\n", dev_path(dev_mc));
+                       pci_domain = dev_mc->bus->dev;
+                       if (pci_domain && (pci_domain->path.type == DEVICE_PATH_PCI_DOMAIN)) {
+                               if ((pci_domain->link_list) && (pci_domain->link_list->children == dev_mc)) {
+                                       printk(BIOS_DEBUG, "%s move to ",dev_path(dev_mc));
+                                       dev_mc->bus->secondary = CONFIG_CBB; // move to 0xff
+                                       printk(BIOS_DEBUG, "%s\n",dev_path(dev_mc));
+                                       while (dev_mc) {
+                                               printk(BIOS_DEBUG, "%s move to ",dev_path(dev_mc));
+                                               dev_mc->path.pci.devfn -= PCI_DEVFN(0x18,0);
+                                               printk(BIOS_DEBUG, "%s\n",dev_path(dev_mc));
+                                               dev_mc = dev_mc->sibling;
+                                       }
+                               }
+                       }
+               }
+       }
+#endif
+       dev_mc = dev_find_slot(CONFIG_CBB, PCI_DEVFN(CONFIG_CDB, 0));
+       if (!dev_mc) {
+               printk(BIOS_ERR, "%02x:%02x.0 not found", CONFIG_CBB, CONFIG_CDB);
+               die("");
+       }
+       sysconf_init(dev_mc);
+#if CONFIG_CBB && (MAX_NODE_NUMS > 32)
+       if (node_nums>32) { // need to put node 32 to node 63 to bus 0xfe
+               if (pci_domain->link_list && !pci_domain->link_list->next) {
+                       struct bus *new_link = new_link(pci_domain);
+                       pci_domain->link_list->next = new_link;
+                       new_link->link_num = 1;
+                       new_link->dev = pci_domain;
+                       new_link->children = 0;
+                       printk(BIOS_DEBUG, "%s links now 2\n", dev_path(pci_domain));
+               }
+               pci_domain->link_list->next->secondary = CONFIG_CBB - 1;
+       }
+#endif
+
+       /* Get Max Number of cores(MNC) */
+       coreid_bits = (cpuid_ecx(AMD_CPUID_ASIZE_PCCOUNT) & 0x0000F000) >> 12;
+       core_max = 1 << (coreid_bits & 0x000F); //mnc
+
+       ApicIdCoreIdSize = ((cpuid_ecx(0x80000008)>>12) & 0xF);
+       if (ApicIdCoreIdSize) {
+               core_nums = (1 << ApicIdCoreIdSize) - 1;
+       } else {
+               core_nums = 3; //quad core
+       }
+
+       /* Find which cpus are present */
+       cpu_bus = dev->link_list;
+       for (i = 0; i < node_nums; i++) {
+               device_t cdb_dev, cpu;
+               struct device_path cpu_path;
+               unsigned busn, devn;
+               struct bus *pbus;
+
+               busn = CONFIG_CBB;
+               devn = CONFIG_CDB + i;
+               pbus = dev_mc->bus;
+#if CONFIG_CBB && (MAX_NODE_NUMS > 32)
+               if (i >= 32) {
+                       busn--;
+                       devn -= 32;
+                       pbus = pci_domain->link_list->next;
+               }
+#endif
+
+               /* Find the cpu's pci device */
+               cdb_dev = dev_find_slot(busn, PCI_DEVFN(devn, 0));
+               if (!cdb_dev) {
+                       /* If I am probing things in a weird order
+                        * ensure all of the cpu's pci devices are found.
+                        */
+                       int fn;
+                       for(fn = 0; fn <= 5; fn++) { //FBDIMM?
+                               cdb_dev = pci_probe_dev(NULL, pbus,
+                                                       PCI_DEVFN(devn, fn));
+                       }
+                       cdb_dev = dev_find_slot(busn, PCI_DEVFN(devn, 0));
+               } else {
+                       /* Ok, We need to set the links for that device.
+                        * otherwise the device under it will not be scanned
+                        */
+                       int linknum;
+#if CONFIG_HT3_SUPPORT==1
+                       linknum = 8;
+#else
+                       linknum = 4;
+#endif
+                       add_more_links(cdb_dev, linknum);
+               }
+
+               family = cpuid_eax(1);
+               family = (family >> 20) & 0xFF;
+               if (family == 1) { //f10
+                       u32 dword;
+                       cdb_dev = dev_find_slot(busn, PCI_DEVFN(devn, 3));
+                       dword = pci_read_config32(cdb_dev, 0xe8);
+                       siblings = ((dword & BIT15) >> 13) | ((dword & (BIT13 | BIT12)) >> 12);
+               } else if (family == 6) {//f15
+                       cdb_dev = dev_find_slot(busn, PCI_DEVFN(devn, 5));
+                       if (cdb_dev && cdb_dev->enabled) {
+                               siblings = pci_read_config32(cdb_dev, 0x84);
+                               siblings &= 0xFF;
+                       }
+               } else {
+                       siblings = 0; //default one core
+               }
+               printk(BIOS_SPEW, "%s family%xh, core_max=0x%x, core_nums=0x%x, siblings=0x%x\n",
+                               dev_path(cdb_dev), 0x0f + family, core_max, core_nums, siblings);
+
+               for (j = 0; j <= siblings; j++ ) {
+                       extern CONST OPTIONS_CONFIG_TOPOLOGY ROMDATA TopologyConfiguration;
+                       u32 modules = TopologyConfiguration.PlatformNumberOfModules;
+                       u32 lapicid_start = 0;
+
+                       /* Build the cpu device path */
+                       cpu_path.type = DEVICE_PATH_APIC;
+                       /*
+                        * APIC ID calucation is tightly coupled with AGESA v5 code.
+                        * This calculation MUST match the assignment calculation done
+                        * in LocalApicInitializationAtEarly() function.
+                        * And reference GetLocalApicIdForCore()
+                        *
+                        * Apply apic enumeration rules
+                        * For systems with >= 16 APICs, put the IO-APICs at 0..n and
+                        * put the local-APICs at m..z
+                        *
+                        * This is needed because many IO-APIC devices only have 4 bits
+                        * for their APIC id and therefore must reside at 0..15
+                         */
+#ifndef CFG_PLAT_NUM_IO_APICS /* defined in mainboard buildOpts.c */
+#define CFG_PLAT_NUM_IO_APICS 3
+#endif
+                       if ((node_nums * core_max) + CFG_PLAT_NUM_IO_APICS >= 0x10) {
+                               lapicid_start = (CFG_PLAT_NUM_IO_APICS - 1) / core_max;
+                               lapicid_start = (lapicid_start + 1) * core_max;
+                               printk(BIOS_SPEW, "lpaicid_start=0x%x ", lapicid_start);
+                       }
+                       cpu_path.apic.apic_id = (lapicid_start * (i/modules + 1)) + ((i % modules) ? (j + (siblings + 1)) : j);
+                       printk(BIOS_SPEW, "node 0x%x core 0x%x apicid=0x%x\n",
+                                       i, j, cpu_path.apic.apic_id);
+
+                       /* See if I can find the cpu */
+                       cpu = find_dev_path(cpu_bus, &cpu_path);
+                       /* Enable the cpu if I have the processor */
+                       if (cdb_dev && cdb_dev->enabled) {
+                               if (!cpu) {
+                                       cpu = alloc_dev(cpu_bus, &cpu_path);
+                               }
+                               if (cpu) {
+                                       cpu->enabled = 1;
+                               }
+                       }
+                       /* Disable the cpu if I don't have the processor */
+                       if (cpu && (!cdb_dev || !cdb_dev->enabled)) {
+                               cpu->enabled = 0;
+                       }
+                       /* Report what I have done */
+                       if (cpu) {
+                               cpu->path.apic.node_id = i;
+                               cpu->path.apic.core_id = j;
+                               if (cpu->path.type == DEVICE_PATH_APIC) {
+                                       cpu->ops = &lapic_ops;
+                               }
+                               printk(BIOS_DEBUG, "CPU: %s %s\n",
+                                       dev_path(cpu), cpu->enabled?"enabled":"disabled");
+                       }
+               } //j
+       }
+       return max;
+}
+
+static void cpu_bus_init(device_t dev)
+{
+       initialize_cpus(dev->link_list);
+}
+
+static void cpu_bus_noop(device_t dev)
+{
+}
+
+static void cpu_bus_read_resources(device_t dev)
+{
+#if CONFIG_MMCONF_SUPPORT
+       struct resource *resource = new_resource(dev, 0xc0010058);
+       resource->base = CONFIG_MMCONF_BASE_ADDRESS;
+       resource->size = CONFIG_MMCONF_BUS_NUMBER * 4096*256;
+       resource->flags = IORESOURCE_MEM | IORESOURCE_RESERVE |
+               IORESOURCE_FIXED | IORESOURCE_STORED |  IORESOURCE_ASSIGNED;
+#endif
+}
+
+static void cpu_bus_set_resources(device_t dev)
+{
+       struct resource *resource = find_resource(dev, 0xc0010058);
+       if (resource) {
+               report_resource_stored(dev, resource, " <mmconfig>");
+       }
+       pci_dev_set_resources(dev);
+}
+
+static struct device_operations cpu_bus_ops = {
+       .read_resources   = cpu_bus_read_resources,
+       .set_resources    = cpu_bus_set_resources,
+       .enable_resources = cpu_bus_noop,
+       .init             = cpu_bus_init,
+       .scan_bus         = cpu_bus_scan,
+};
+
+
+static void root_complex_enable_dev(struct device *dev)
+{
+       /* Set the operations if it is a special bus type */
+       if (dev->path.type == DEVICE_PATH_PCI_DOMAIN) {
+               dev->ops = &pci_domain_ops;
+       } else if (dev->path.type == DEVICE_PATH_APIC_CLUSTER) {
+               dev->ops = &cpu_bus_ops;
+       }
+}
+
+
+struct chip_operations northbridge_amd_agesa_family15_root_complex_ops = {
+       CHIP_NAME("AMD FAM15 Root Complex")
+       .enable_dev = root_complex_enable_dev,
+};
diff --git a/src/northbridge/amd/agesa/family15/northbridge.h b/src/northbridge/amd/agesa/family15/northbridge.h
new file mode 100644 (file)
index 0000000..7606b32
--- /dev/null
@@ -0,0 +1,26 @@
+/*
+ * This file is part of the coreboot project.
+ *
+ * Copyright (C) 2012 Advanced Micro Devices, Inc.
+ *
+ * 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
+ */
+
+#ifndef NORTHBRIDGE_AMD_AGESA_FAM15_H
+#define NORTHBRIDGE_AMD_AGESA_FAM15_H
+
+static struct device_operations pci_domain_ops;
+static struct device_operations cpu_bus_ops;
+
+#endif /* NORTHBRIDGE_AMD_AGESA_FAM15_H */
diff --git a/src/northbridge/amd/agesa/family15/root_complex/Kconfig b/src/northbridge/amd/agesa/family15/root_complex/Kconfig
new file mode 100644 (file)
index 0000000..032b836
--- /dev/null
@@ -0,0 +1,2 @@
+config NORTHBRIDGE_AMD_AGESA_FAMILY15_ROOT_COMPLEX
+       bool
diff --git a/src/northbridge/amd/agesa/family15/root_complex/chip.h b/src/northbridge/amd/agesa/family15/root_complex/chip.h
new file mode 100644 (file)
index 0000000..06b3510
--- /dev/null
@@ -0,0 +1,24 @@
+/*
+ * This file is part of the coreboot project.
+ *
+ * Copyright (C) 2012 Advanced Micro Devices, Inc.
+ *
+ * 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
+ */
+
+struct northbridge_amd_agesa_family15_root_complex_config
+{
+};
+
+extern struct chip_operations northbridge_amd_agesa_family15_root_complex_ops;