[PATCH] x86: make 64bit identify_cpu use cpu_dev

[PATCH] x86: make 64bit identify_cpu use cpu_dev

[Yinghai Lu]

may need to move some functions to common.c later

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

Index: linux-2.6/arch/x86/kernel/cpu/amd_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/amd_64.c
+++ linux-2.6/arch/x86/kernel/cpu/amd_64.c
@@ -7,8 +7,7 @@
 
 #include <mach_apic.h>
 
-extern int __cpuinit get_model_name(struct cpuinfo_x86 *c);
-extern void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c);
+#include "cpu.h"
 
 int force_mwait __cpuinitdata;
 
@@ -109,7 +108,7 @@ static void __cpuinit early_init_amd_mc(
 #endif
 }
 
-void __cpuinit early_init_amd(struct cpuinfo_x86 *c)
+static void __cpuinit early_init_amd(struct cpuinfo_x86 *c)
 {
 	early_init_amd_mc(c);
 
@@ -118,7 +117,7 @@ void __cpuinit early_init_amd(struct cpu
 		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
 }
 
-void __cpuinit init_amd(struct cpuinfo_x86 *c)
+static void __cpuinit init_amd(struct cpuinfo_x86 *c)
 {
 	unsigned level;
 
@@ -200,3 +199,13 @@ void __cpuinit init_amd(struct cpuinfo_x
 			set_memory_4k((unsigned long)__va(tseg), 1);
 	}
 }
+
+static struct cpu_dev amd_cpu_dev __cpuinitdata = {
+	.c_vendor	= "AMD",
+	.c_ident	= { "AuthenticAMD" },
+	.c_early_init   = early_init_amd,
+	.c_init		= init_amd,
+};
+
+cpu_vendor_dev_register(X86_VENDOR_AMD, &amd_cpu_dev);
+
Index: linux-2.6/arch/x86/kernel/cpu/centaur_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/centaur_64.c
+++ linux-2.6/arch/x86/kernel/cpu/centaur_64.c
@@ -4,13 +4,15 @@
 #include <asm/cpufeature.h>
 #include <asm/processor.h>
 
-void __cpuinit early_init_centaur(struct cpuinfo_x86 *c)
+#include "cpu.h"
+
+static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c)
 {
 	if (c->x86 == 0x6 && c->x86_model >= 0xf)
 		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
 }
 
-void __cpuinit init_centaur(struct cpuinfo_x86 *c)
+static void __cpuinit init_centaur(struct cpuinfo_x86 *c)
 {
 	/* Cache sizes */
 	unsigned n;
@@ -29,3 +31,13 @@ void __cpuinit init_centaur(struct cpuin
 	}
 	set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC);
 }
+
+static struct cpu_dev centaur_cpu_dev __cpuinitdata = {
+	.c_vendor	= "Centaur",
+	.c_ident	= { "CentaurHauls" },
+	.c_early_init	= early_init_centaur,
+	.c_init		= init_centaur,
+};
+
+cpu_vendor_dev_register(X86_VENDOR_CENTAUR, &centaur_cpu_dev);
+
Index: linux-2.6/arch/x86/kernel/cpu/cpu.h
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/cpu.h
+++ linux-2.6/arch/x86/kernel/cpu/cpu.h
@@ -1,4 +1,6 @@
-#ifdef CONFIG_X86_32
+#ifndef ARCH_X86_CPU_H
+
+#define ARCH_X86_CPU_H
 
 struct cpu_model_info {
 	int vendor;
@@ -38,4 +40,4 @@ extern struct cpu_vendor_dev __x86cpuven
 extern int get_model_name(struct cpuinfo_x86 *c);
 extern void display_cacheinfo(struct cpuinfo_x86 *c);
 
-#endif /* CONFIG_X86_32 */
+#endif
Index: linux-2.6/arch/x86/kernel/cpu/intel_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/intel_64.c
+++ linux-2.6/arch/x86/kernel/cpu/intel_64.c
@@ -5,7 +5,9 @@
 #include <asm/topology.h>
 #include <asm/numa_64.h>
 
-void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
+#include "cpu.h"
+
+static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
 {
 	if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
 	    (c->x86 == 0x6 && c->x86_model >= 0x0e))
@@ -48,7 +50,7 @@ static void __cpuinit srat_detect_node(v
 #endif
 }
 
-void __cpuinit init_intel(struct cpuinfo_x86 *c)
+static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 {
 	/* Cache sizes */
 	unsigned n;
@@ -90,3 +92,12 @@ void __cpuinit init_intel(struct cpuinfo
 
 	srat_detect_node();
 }
+
+static struct cpu_dev intel_cpu_dev __cpuinitdata = {
+	.c_vendor	= "Intel",
+	.c_ident	= { "GenuineIntel" },
+	.c_early_init   = early_init_intel,
+	.c_init		= init_intel,
+};
+cpu_vendor_dev_register(X86_VENDOR_INTEL, &intel_cpu_dev);
+
Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -81,6 +81,8 @@
 #define ARCH_SETUP
 #endif
 
+#include "cpu/cpu.h"
+
 /* We need valid kernel segments for data and code in long mode too
  * IRET will check the segment types  kkeil 2000/10/28
  * Also sysret mandates a special GDT layout
@@ -180,6 +182,7 @@ static struct resource bss_resource = {
 	.flags = IORESOURCE_RAM,
 };
 
+static void __init early_cpu_init(void);
 static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c);
 
 #ifdef CONFIG_PROC_VMCORE
@@ -341,6 +344,7 @@ void __init setup_arch(char **cmdline_p)
 	bss_resource.start = virt_to_phys(&__bss_start);
 	bss_resource.end = virt_to_phys(&__bss_stop)-1;
 
+	early_cpu_init();
 	early_identify_cpu(&boot_cpu_data);
 
 	strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
@@ -527,6 +531,19 @@ void __init setup_arch(char **cmdline_p)
 	check_enable_amd_mmconf_dmi();
 }
 
+struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {};
+
+static void __cpuinit default_init(struct cpuinfo_x86 *c)
+{
+	display_cacheinfo(c);
+}
+
+static struct cpu_dev __cpuinitdata default_cpu = {
+	.c_init	= default_init,
+	.c_vendor = "Unknown",
+};
+static struct cpu_dev *this_cpu __cpuinitdata = &default_cpu;
+
 int __cpuinit get_model_name(struct cpuinfo_x86 *c)
 {
 	unsigned int *v;
@@ -628,24 +645,37 @@ out:
 static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
 {
 	char *v = c->x86_vendor_id;
+	int i;
+	static int printed;
 
-	if (!strcmp(v, "AuthenticAMD"))
-		c->x86_vendor = X86_VENDOR_AMD;
-	else if (!strcmp(v, "GenuineIntel"))
-		c->x86_vendor = X86_VENDOR_INTEL;
-	else if (!strcmp(v, "CentaurHauls"))
-		c->x86_vendor = X86_VENDOR_CENTAUR;
-	else
-		c->x86_vendor = X86_VENDOR_UNKNOWN;
+	for (i = 0; i < X86_VENDOR_NUM; i++) {
+		if (cpu_devs[i]) {
+			if (!strcmp(v, cpu_devs[i]->c_ident[0]) ||
+			    (cpu_devs[i]->c_ident[1] &&
+			    !strcmp(v, cpu_devs[i]->c_ident[1]))) {
+				c->x86_vendor = i;
+				this_cpu = cpu_devs[i];
+				return;
+			}
+		}
+	}
+	if (!printed) {
+		printed++;
+		printk(KERN_ERR "CPU: Vendor unknown, using generic init.\n");
+		printk(KERN_ERR "CPU: Your system may be unstable.\n");
+	}
+	c->x86_vendor = X86_VENDOR_UNKNOWN;
 }
 
-// FIXME: Needs to use cpu_vendor_dev_register
-extern void __cpuinit early_init_amd(struct cpuinfo_x86 *c);
-extern void __cpuinit init_amd(struct cpuinfo_x86 *c);
-extern void __cpuinit early_init_intel(struct cpuinfo_x86 *c);
-extern void __cpuinit init_intel(struct cpuinfo_x86 *c);
-extern void __cpuinit early_init_centaur(struct cpuinfo_x86 *c);
-extern void __cpuinit init_centaur(struct cpuinfo_x86 *c);
+static void __init early_cpu_init(void)
+{
+        struct cpu_vendor_dev *cvdev;
+
+        for (cvdev = __x86cpuvendor_start ;
+             cvdev < __x86cpuvendor_end   ;
+             cvdev++)
+                cpu_devs[cvdev->vendor] = cvdev->cpu_dev;
+}
 
 /* Do some early cpuid on the boot CPU to get some parameter that are
    needed before check_bugs. Everything advanced is in identify_cpu
@@ -725,17 +755,9 @@ static void __cpuinit early_identify_cpu
 	if (c->extended_cpuid_level >= 0x80000007)
 		c->x86_power = cpuid_edx(0x80000007);
 
-	switch (c->x86_vendor) {
-	case X86_VENDOR_AMD:
-		early_init_amd(c);
-		break;
-	case X86_VENDOR_INTEL:
-		early_init_intel(c);
-		break;
-	case X86_VENDOR_CENTAUR:
-		early_init_centaur(c);
-		break;
-	}
+	if (c->x86_vendor != X86_VENDOR_UNKNOWN &&
+	    cpu_devs[c->x86_vendor]->c_early_init)
+		cpu_devs[c->x86_vendor]->c_early_init(c);
 
 	validate_pat_support(c);
 }
@@ -763,24 +785,8 @@ void __cpuinit identify_cpu(struct cpuin
 	 * At the end of this section, c->x86_capability better
 	 * indicate the features this CPU genuinely supports!
 	 */
-	switch (c->x86_vendor) {
-	case X86_VENDOR_AMD:
-		init_amd(c);
-		break;
-
-	case X86_VENDOR_INTEL:
-		init_intel(c);
-		break;
-
-	case X86_VENDOR_CENTAUR:
-		init_centaur(c);
-		break;
-
-	case X86_VENDOR_UNKNOWN:
-	default:
-		display_cacheinfo(c);
-		break;
-	}
+	if (this_cpu->c_init)
+		this_cpu->c_init(c);
 
 	detect_ht(c);
 

Re: [PATCH] x86: make 64bit identify_cpu use cpu_dev

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> may need to move some functions to common.c later
> 
> Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

applied to tip/x86/cpu, thanks Yinghai.

	Ingo

Re: [PATCH] x86: make 64bit identify_cpu use cpu_dev

[Ingo Molnar]

* Ingo Molnar <mingo@elte.hu> wrote:

> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
> 
> > may need to move some functions to common.c later
> > 
> > Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
> 
> applied to tip/x86/cpu, thanks Yinghai.

-tip testing found a boot failure on 64-bit, the bootup just hangs early 
during bootup. I bisected it down to this patch and reverting the patch 
solves the lockup.

Here is the config that triggers it:

 http://redhat.com/~mingo/misc/config-Thu_Jun_19_10_11_15_CEST_2008.bad

	Ingo

Re: [PATCH] x86: make 64bit identify_cpu use cpu_dev

[Yinghai Lu]

On Thu, Jun 19, 2008 at 2:03 AM, Ingo Molnar <mingo@elte.hu> wrote:
>
> * Ingo Molnar <mingo@elte.hu> wrote:
>
>> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>>
>> > may need to move some functions to common.c later
>> >
>> > Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
>>
>> applied to tip/x86/cpu, thanks Yinghai.
>
> -tip testing found a boot failure on 64-bit, the bootup just hangs early
> during bootup. I bisected it down to this patch and reverting the patch
> solves the lockup.
>
> Here is the config that triggers it:
>
>  http://redhat.com/~mingo/misc/config-Thu_Jun_19_10_11_15_CEST_2008.bad

can not duplicate that here.

can you send me the boot log on your system?

YH

[PATCH] x86: make 64bit identify_cpu use cpu_dev v2

[Yinghai Lu]

v2: fix early_panic on some conf
    reason : struct cpu_vendor_dev size is 16, need to make table to be 16
             byte alignment
    also print out the cpu supported...

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/cpu/amd_64.c     |   17 ++++-
 arch/x86/kernel/cpu/centaur_64.c |   16 ++++-
 arch/x86/kernel/cpu/cpu.h        |    6 +-
 arch/x86/kernel/cpu/intel_64.c   |   15 ++++-
 arch/x86/kernel/setup_64.c       |  114 +++++++++++++++++++++++----------------
 arch/x86/kernel/vmlinux_64.lds.S |    1 
 6 files changed, 115 insertions(+), 54 deletions(-)

Index: linux-2.6/arch/x86/kernel/cpu/amd_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/amd_64.c
+++ linux-2.6/arch/x86/kernel/cpu/amd_64.c
@@ -7,8 +7,7 @@
 
 #include <mach_apic.h>
 
-extern int __cpuinit get_model_name(struct cpuinfo_x86 *c);
-extern void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c);
+#include "cpu.h"
 
 int force_mwait __cpuinitdata;
 
@@ -109,7 +108,7 @@ static void __cpuinit early_init_amd_mc(
 #endif
 }
 
-void __cpuinit early_init_amd(struct cpuinfo_x86 *c)
+static void __cpuinit early_init_amd(struct cpuinfo_x86 *c)
 {
 	early_init_amd_mc(c);
 
@@ -118,7 +117,7 @@ void __cpuinit early_init_amd(struct cpu
 		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
 }
 
-void __cpuinit init_amd(struct cpuinfo_x86 *c)
+static void __cpuinit init_amd(struct cpuinfo_x86 *c)
 {
 	unsigned level;
 
@@ -200,3 +199,13 @@ void __cpuinit init_amd(struct cpuinfo_x
 			set_memory_4k((unsigned long)__va(tseg), 1);
 	}
 }
+
+static struct cpu_dev amd_cpu_dev __cpuinitdata = {
+	.c_vendor	= "AMD",
+	.c_ident	= { "AuthenticAMD" },
+	.c_early_init   = early_init_amd,
+	.c_init		= init_amd,
+};
+
+cpu_vendor_dev_register(X86_VENDOR_AMD, &amd_cpu_dev);
+
Index: linux-2.6/arch/x86/kernel/cpu/centaur_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/centaur_64.c
+++ linux-2.6/arch/x86/kernel/cpu/centaur_64.c
@@ -4,13 +4,15 @@
 #include <asm/cpufeature.h>
 #include <asm/processor.h>
 
-void __cpuinit early_init_centaur(struct cpuinfo_x86 *c)
+#include "cpu.h"
+
+static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c)
 {
 	if (c->x86 == 0x6 && c->x86_model >= 0xf)
 		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
 }
 
-void __cpuinit init_centaur(struct cpuinfo_x86 *c)
+static void __cpuinit init_centaur(struct cpuinfo_x86 *c)
 {
 	/* Cache sizes */
 	unsigned n;
@@ -29,3 +31,13 @@ void __cpuinit init_centaur(struct cpuin
 	}
 	set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC);
 }
+
+static struct cpu_dev centaur_cpu_dev __cpuinitdata = {
+	.c_vendor	= "Centaur",
+	.c_ident	= { "CentaurHauls" },
+	.c_early_init	= early_init_centaur,
+	.c_init		= init_centaur,
+};
+
+cpu_vendor_dev_register(X86_VENDOR_CENTAUR, &centaur_cpu_dev);
+
Index: linux-2.6/arch/x86/kernel/cpu/cpu.h
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/cpu.h
+++ linux-2.6/arch/x86/kernel/cpu/cpu.h
@@ -1,4 +1,6 @@
-#ifdef CONFIG_X86_32
+#ifndef ARCH_X86_CPU_H
+
+#define ARCH_X86_CPU_H
 
 struct cpu_model_info {
 	int vendor;
@@ -38,4 +40,4 @@ extern struct cpu_vendor_dev __x86cpuven
 extern int get_model_name(struct cpuinfo_x86 *c);
 extern void display_cacheinfo(struct cpuinfo_x86 *c);
 
-#endif /* CONFIG_X86_32 */
+#endif
Index: linux-2.6/arch/x86/kernel/cpu/intel_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/intel_64.c
+++ linux-2.6/arch/x86/kernel/cpu/intel_64.c
@@ -5,7 +5,9 @@
 #include <asm/topology.h>
 #include <asm/numa_64.h>
 
-void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
+#include "cpu.h"
+
+static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
 {
 	if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
 	    (c->x86 == 0x6 && c->x86_model >= 0x0e))
@@ -48,7 +50,7 @@ static void __cpuinit srat_detect_node(v
 #endif
 }
 
-void __cpuinit init_intel(struct cpuinfo_x86 *c)
+static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 {
 	/* Cache sizes */
 	unsigned n;
@@ -90,3 +92,12 @@ void __cpuinit init_intel(struct cpuinfo
 
 	srat_detect_node();
 }
+
+static struct cpu_dev intel_cpu_dev __cpuinitdata = {
+	.c_vendor	= "Intel",
+	.c_ident	= { "GenuineIntel" },
+	.c_early_init   = early_init_intel,
+	.c_init		= init_intel,
+};
+cpu_vendor_dev_register(X86_VENDOR_INTEL, &intel_cpu_dev);
+
Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -81,6 +81,8 @@
 #define ARCH_SETUP
 #endif
 
+#include "cpu/cpu.h"
+
 /* We need valid kernel segments for data and code in long mode too
  * IRET will check the segment types  kkeil 2000/10/28
  * Also sysret mandates a special GDT layout
@@ -180,6 +182,7 @@ static struct resource bss_resource = {
 	.flags = IORESOURCE_RAM,
 };
 
+static void __init early_cpu_init(void);
 static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c);
 
 #ifdef CONFIG_PROC_VMCORE
@@ -341,6 +344,7 @@ void __init setup_arch(char **cmdline_p)
 	bss_resource.start = virt_to_phys(&__bss_start);
 	bss_resource.end = virt_to_phys(&__bss_stop)-1;
 
+	early_cpu_init();
 	early_identify_cpu(&boot_cpu_data);
 
 	strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
@@ -527,6 +531,19 @@ void __init setup_arch(char **cmdline_p)
 	check_enable_amd_mmconf_dmi();
 }
 
+struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {};
+
+static void __cpuinit default_init(struct cpuinfo_x86 *c)
+{
+	display_cacheinfo(c);
+}
+
+static struct cpu_dev __cpuinitdata default_cpu = {
+	.c_init	= default_init,
+	.c_vendor = "Unknown",
+};
+static struct cpu_dev *this_cpu __cpuinitdata = &default_cpu;
+
 int __cpuinit get_model_name(struct cpuinfo_x86 *c)
 {
 	unsigned int *v;
@@ -628,24 +645,57 @@ out:
 static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
 {
 	char *v = c->x86_vendor_id;
+	int i;
+	static int printed;
 
-	if (!strcmp(v, "AuthenticAMD"))
-		c->x86_vendor = X86_VENDOR_AMD;
-	else if (!strcmp(v, "GenuineIntel"))
-		c->x86_vendor = X86_VENDOR_INTEL;
-	else if (!strcmp(v, "CentaurHauls"))
-		c->x86_vendor = X86_VENDOR_CENTAUR;
-	else
-		c->x86_vendor = X86_VENDOR_UNKNOWN;
+	for (i = 0; i < X86_VENDOR_NUM; i++) {
+		if (cpu_devs[i]) {
+			if (!strcmp(v, cpu_devs[i]->c_ident[0]) ||
+			    (cpu_devs[i]->c_ident[1] &&
+			    !strcmp(v, cpu_devs[i]->c_ident[1]))) {
+				c->x86_vendor = i;
+				this_cpu = cpu_devs[i];
+				return;
+			}
+		}
+	}
+	if (!printed) {
+		printed++;
+		printk(KERN_ERR "CPU: Vendor unknown, using generic init.\n");
+		printk(KERN_ERR "CPU: Your system may be unstable.\n");
+	}
+	c->x86_vendor = X86_VENDOR_UNKNOWN;
 }
 
-// FIXME: Needs to use cpu_vendor_dev_register
-extern void __cpuinit early_init_amd(struct cpuinfo_x86 *c);
-extern void __cpuinit init_amd(struct cpuinfo_x86 *c);
-extern void __cpuinit early_init_intel(struct cpuinfo_x86 *c);
-extern void __cpuinit init_intel(struct cpuinfo_x86 *c);
-extern void __cpuinit early_init_centaur(struct cpuinfo_x86 *c);
-extern void __cpuinit init_centaur(struct cpuinfo_x86 *c);
+static void __init early_cpu_support_print(void)
+{
+	int i,j;
+	struct cpu_dev *cpu_devx;
+
+	printk("KERNEL supported cpus:\n");
+	for (i = 0; i < X86_VENDOR_NUM; i++) {
+		cpu_devx = cpu_devs[i];
+		if (!cpu_devx)
+			continue;
+		for (j = 0; j < 2; j++) {
+			if (!cpu_devx->c_ident[j])
+				continue;
+			printk("  %s %s\n", cpu_devx->c_vendor,
+				cpu_devx->c_ident[j]);
+		}
+	}
+}
+
+static void __init early_cpu_init(void)
+{
+        struct cpu_vendor_dev *cvdev;
+
+        for (cvdev = __x86cpuvendor_start ;
+             cvdev < __x86cpuvendor_end   ;
+             cvdev++)
+                cpu_devs[cvdev->vendor] = cvdev->cpu_dev;
+	early_cpu_support_print();
+}
 
 /* Do some early cpuid on the boot CPU to get some parameter that are
    needed before check_bugs. Everything advanced is in identify_cpu
@@ -725,17 +775,9 @@ static void __cpuinit early_identify_cpu
 	if (c->extended_cpuid_level >= 0x80000007)
 		c->x86_power = cpuid_edx(0x80000007);
 
-	switch (c->x86_vendor) {
-	case X86_VENDOR_AMD:
-		early_init_amd(c);
-		break;
-	case X86_VENDOR_INTEL:
-		early_init_intel(c);
-		break;
-	case X86_VENDOR_CENTAUR:
-		early_init_centaur(c);
-		break;
-	}
+	if (c->x86_vendor != X86_VENDOR_UNKNOWN &&
+	    cpu_devs[c->x86_vendor]->c_early_init)
+		cpu_devs[c->x86_vendor]->c_early_init(c);
 
 	validate_pat_support(c);
 }
@@ -763,24 +805,8 @@ void __cpuinit identify_cpu(struct cpuin
 	 * At the end of this section, c->x86_capability better
 	 * indicate the features this CPU genuinely supports!
 	 */
-	switch (c->x86_vendor) {
-	case X86_VENDOR_AMD:
-		init_amd(c);
-		break;
-
-	case X86_VENDOR_INTEL:
-		init_intel(c);
-		break;
-
-	case X86_VENDOR_CENTAUR:
-		init_centaur(c);
-		break;
-
-	case X86_VENDOR_UNKNOWN:
-	default:
-		display_cacheinfo(c);
-		break;
-	}
+	if (this_cpu->c_init)
+		this_cpu->c_init(c);
 
 	detect_ht(c);
 
Index: linux-2.6/arch/x86/kernel/vmlinux_64.lds.S
===================================================================
--- linux-2.6.orig/arch/x86/kernel/vmlinux_64.lds.S
+++ linux-2.6/arch/x86/kernel/vmlinux_64.lds.S
@@ -168,6 +168,7 @@ SECTIONS
 	*(.con_initcall.init)
   }
   __con_initcall_end = .;
+  . = ALIGN(16);
   __x86cpuvendor_start = .;
   .x86cpuvendor.init : AT(ADDR(.x86cpuvendor.init) - LOAD_OFFSET) {
 	*(.x86cpuvendor.init)

Re: [PATCH] x86: make 64bit identify_cpu use cpu_dev v2

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> v2: fix early_panic on some conf
>     reason : struct cpu_vendor_dev size is 16, need to make table to be 16
>              byte alignment
>     also print out the cpu supported...

applied to tip/x86/cpu, thanks Yinghai.

> --- linux-2.6.orig/arch/x86/kernel/vmlinux_64.lds.S
> +++ linux-2.6/arch/x86/kernel/vmlinux_64.lds.S
> @@ -168,6 +168,7 @@ SECTIONS
>  	*(.con_initcall.init)
>    }
>    __con_initcall_end = .;
> +  . = ALIGN(16);
>    __x86cpuvendor_start = .;
>    .x86cpuvendor.init : AT(ADDR(.x86cpuvendor.init) - LOAD_OFFSET) {
>  	*(.x86cpuvendor.init)

ah, nice one!

we broke that via v2.6.25-37-g03ae576 "x86: use ELF section to list CPU 
vendor specific code" - should we backport your fix to mainline right 
now? I'm wondering why it only triggered now - did alignment change in 
this section due to your changes and thus this dormant bug became 
triggerable?

	Ingo

Re: [PATCH] x86: make 64bit identify_cpu use cpu_dev v2

[Yinghai Lu]

On Thu, Jun 19, 2008 at 11:29 PM, Ingo Molnar <mingo@elte.hu> wrote:
>
> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>
>> v2: fix early_panic on some conf
>>     reason : struct cpu_vendor_dev size is 16, need to make table to be 16
>>              byte alignment
>>     also print out the cpu supported...
>
> applied to tip/x86/cpu, thanks Yinghai.
>
>> --- linux-2.6.orig/arch/x86/kernel/vmlinux_64.lds.S
>> +++ linux-2.6/arch/x86/kernel/vmlinux_64.lds.S
>> @@ -168,6 +168,7 @@ SECTIONS
>>       *(.con_initcall.init)
>>    }
>>    __con_initcall_end = .;
>> +  . = ALIGN(16);
>>    __x86cpuvendor_start = .;
>>    .x86cpuvendor.init : AT(ADDR(.x86cpuvendor.init) - LOAD_OFFSET) {
>>       *(.x86cpuvendor.init)
>
> ah, nice one!
>
> we broke that via v2.6.25-37-g03ae576 "x86: use ELF section to list CPU
> vendor specific code" - should we backport your fix to mainline right
> now? I'm wondering why it only triggered now - did alignment change in
> this section due to your changes and thus this dormant bug became
> triggerable?

don't need,  that section is not used before this patch.
we my add some building check to verify those kind info to avoid same kind bug.

YH

[PATCH] x86: change identify_cpu to static

[Yinghai Lu]

[PATCH] x86: change identify_cpu to static

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/cpu/common.c    |    2 +-
 arch/x86/kernel/cpu/common_64.c |    2 +-
 include/asm-x86/processor.h     |    1 -
 3 files changed, 2 insertions(+), 3 deletions(-)

Index: linux-2.6/arch/x86/kernel/cpu/common_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/common_64.c
+++ linux-2.6/arch/x86/kernel/cpu/common_64.c
@@ -312,7 +312,7 @@ static void __cpuinit early_identify_cpu
 /*
  * This does the hard work of actually picking apart the CPU stuff...
  */
-void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
+static void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
 {
 	int i;
 
Index: linux-2.6/arch/x86/kernel/cpu/common.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/common.c
+++ linux-2.6/arch/x86/kernel/cpu/common.c
@@ -427,7 +427,7 @@ __setup("serialnumber", x86_serial_nr_se
 /*
  * This does the hard work of actually picking apart the CPU stuff...
  */
-void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
+static void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
 {
 	int i;
 
Index: linux-2.6/include/asm-x86/processor.h
===================================================================
--- linux-2.6.orig/include/asm-x86/processor.h
+++ linux-2.6/include/asm-x86/processor.h
@@ -155,7 +155,6 @@ static inline int hlt_works(int cpu)
 extern void cpu_detect(struct cpuinfo_x86 *c);
 
 extern void early_cpu_init(void);
-extern void identify_cpu(struct cpuinfo_x86 *);
 extern void identify_boot_cpu(void);
 extern void identify_secondary_cpu(struct cpuinfo_x86 *);
 extern void print_cpu_info(struct cpuinfo_x86 *);

[PATCH] x86: seperate funcs from setup_64 to cpu common_64.c

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@mail.com>

---
 arch/x86/kernel/cpu/Makefile    |    2 
 arch/x86/kernel/cpu/common_64.c |  406 ++++++++++++++++++++++++++++++++++++++++
 arch/x86/kernel/setup_64.c      |  393 --------------------------------------
 include/asm-x86/processor.h     |    1 
 4 files changed, 409 insertions(+), 393 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -41,7 +41,6 @@
 #include <linux/dmi.h>
 #include <linux/dma-mapping.h>
 #include <linux/ctype.h>
-#include <linux/sort.h>
 #include <linux/uaccess.h>
 #include <linux/init_ohci1394_dma.h>
 #include <linux/kvm_para.h>
@@ -52,7 +51,6 @@
 #include <asm/vsyscall.h>
 #include <asm/io.h>
 #include <asm/smp.h>
-#include <asm/msr.h>
 #include <asm/desc.h>
 #include <video/edid.h>
 #include <asm/e820.h>
@@ -67,11 +65,6 @@
 #include <asm/sections.h>
 #include <asm/dmi.h>
 #include <asm/cacheflush.h>
-#include <asm/mce.h>
-#include <asm/ds.h>
-#include <asm/topology.h>
-#include <asm/trampoline.h>
-#include <asm/pat.h>
 
 #include <mach_apic.h>
 #ifdef CONFIG_PARAVIRT
@@ -80,24 +73,6 @@
 #define ARCH_SETUP
 #endif
 
-#include "cpu/cpu.h"
-
-/* We need valid kernel segments for data and code in long mode too
- * IRET will check the segment types  kkeil 2000/10/28
- * Also sysret mandates a special GDT layout
- */
-/* The TLS descriptors are currently at a different place compared to i386.
-   Hopefully nobody expects them at a fixed place (Wine?) */
-DEFINE_PER_CPU(struct gdt_page, gdt_page) = { .gdt = {
-	[GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } },
-	[GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } },
-	[GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } },
-	[GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } },
-	[GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } },
-	[GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } },
-} };
-EXPORT_PER_CPU_SYMBOL_GPL(gdt_page);
-
 /*
  * Machine setup..
  */
@@ -105,8 +80,6 @@ EXPORT_PER_CPU_SYMBOL_GPL(gdt_page);
 struct cpuinfo_x86 boot_cpu_data __read_mostly;
 EXPORT_SYMBOL(boot_cpu_data);
 
-__u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata;
-
 unsigned long mmu_cr4_features;
 
 /* Boot loader ID as an integer, for the benefit of proc_dointvec */
@@ -181,9 +154,6 @@ static struct resource bss_resource = {
 	.flags = IORESOURCE_RAM,
 };
 
-static void __init early_cpu_init(void);
-static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c);
-
 #ifdef CONFIG_PROC_VMCORE
 /* elfcorehdr= specifies the location of elf core header
  * stored by the crashed kernel. This option will be passed
@@ -243,17 +213,6 @@ static inline void copy_edd(void)
 }
 #endif
 
-/* Current gdt points %fs at the "master" per-cpu area: after this,
- *  * it's on the real one. */
-void switch_to_new_gdt(void)
-{
-	struct desc_ptr gdt_descr;
-
-	gdt_descr.address = (long)get_cpu_gdt_table(smp_processor_id());
-	gdt_descr.size = GDT_SIZE - 1;
-	load_gdt(&gdt_descr);
-}
-
 /*
  * setup_arch - architecture-specific boot-time initializations
  *
@@ -304,7 +263,6 @@ void __init setup_arch(char **cmdline_p)
 	bss_resource.end = virt_to_phys(&__bss_stop)-1;
 
 	early_cpu_init();
-	early_identify_cpu(&boot_cpu_data);
 
 	strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
 	*cmdline_p = command_line;
@@ -488,353 +446,3 @@ void __init setup_arch(char **cmdline_p)
 #endif
 }
 
-struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {};
-
-static void __cpuinit default_init(struct cpuinfo_x86 *c)
-{
-	display_cacheinfo(c);
-}
-
-static struct cpu_dev __cpuinitdata default_cpu = {
-	.c_init	= default_init,
-	.c_vendor = "Unknown",
-};
-static struct cpu_dev *this_cpu __cpuinitdata = &default_cpu;
-
-int __cpuinit get_model_name(struct cpuinfo_x86 *c)
-{
-	unsigned int *v;
-
-	if (c->extended_cpuid_level < 0x80000004)
-		return 0;
-
-	v = (unsigned int *) c->x86_model_id;
-	cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]);
-	cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]);
-	cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]);
-	c->x86_model_id[48] = 0;
-	return 1;
-}
-
-
-void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
-{
-	unsigned int n, dummy, eax, ebx, ecx, edx;
-
-	n = c->extended_cpuid_level;
-
-	if (n >= 0x80000005) {
-		cpuid(0x80000005, &dummy, &ebx, &ecx, &edx);
-		printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), "
-		       "D cache %dK (%d bytes/line)\n",
-		       edx>>24, edx&0xFF, ecx>>24, ecx&0xFF);
-		c->x86_cache_size = (ecx>>24) + (edx>>24);
-		/* On K8 L1 TLB is inclusive, so don't count it */
-		c->x86_tlbsize = 0;
-	}
-
-	if (n >= 0x80000006) {
-		cpuid(0x80000006, &dummy, &ebx, &ecx, &edx);
-		ecx = cpuid_ecx(0x80000006);
-		c->x86_cache_size = ecx >> 16;
-		c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff);
-
-		printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
-		c->x86_cache_size, ecx & 0xFF);
-	}
-	if (n >= 0x80000008) {
-		cpuid(0x80000008, &eax, &dummy, &dummy, &dummy);
-		c->x86_virt_bits = (eax >> 8) & 0xff;
-		c->x86_phys_bits = eax & 0xff;
-	}
-}
-
-void __cpuinit detect_ht(struct cpuinfo_x86 *c)
-{
-#ifdef CONFIG_SMP
-	u32 eax, ebx, ecx, edx;
-	int index_msb, core_bits;
-
-	cpuid(1, &eax, &ebx, &ecx, &edx);
-
-
-	if (!cpu_has(c, X86_FEATURE_HT))
-		return;
-	if (cpu_has(c, X86_FEATURE_CMP_LEGACY))
-		goto out;
-
-	smp_num_siblings = (ebx & 0xff0000) >> 16;
-
-	if (smp_num_siblings == 1) {
-		printk(KERN_INFO  "CPU: Hyper-Threading is disabled\n");
-	} else if (smp_num_siblings > 1) {
-
-		if (smp_num_siblings > NR_CPUS) {
-			printk(KERN_WARNING "CPU: Unsupported number of "
-			       "siblings %d", smp_num_siblings);
-			smp_num_siblings = 1;
-			return;
-		}
-
-		index_msb = get_count_order(smp_num_siblings);
-		c->phys_proc_id = phys_pkg_id(index_msb);
-
-		smp_num_siblings = smp_num_siblings / c->x86_max_cores;
-
-		index_msb = get_count_order(smp_num_siblings);
-
-		core_bits = get_count_order(c->x86_max_cores);
-
-		c->cpu_core_id = phys_pkg_id(index_msb) &
-					       ((1 << core_bits) - 1);
-	}
-out:
-	if ((c->x86_max_cores * smp_num_siblings) > 1) {
-		printk(KERN_INFO  "CPU: Physical Processor ID: %d\n",
-		       c->phys_proc_id);
-		printk(KERN_INFO  "CPU: Processor Core ID: %d\n",
-		       c->cpu_core_id);
-	}
-
-#endif
-}
-
-static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
-{
-	char *v = c->x86_vendor_id;
-	int i;
-	static int printed;
-
-	for (i = 0; i < X86_VENDOR_NUM; i++) {
-		if (cpu_devs[i]) {
-			if (!strcmp(v, cpu_devs[i]->c_ident[0]) ||
-			    (cpu_devs[i]->c_ident[1] &&
-			    !strcmp(v, cpu_devs[i]->c_ident[1]))) {
-				c->x86_vendor = i;
-				this_cpu = cpu_devs[i];
-				return;
-			}
-		}
-	}
-	if (!printed) {
-		printed++;
-		printk(KERN_ERR "CPU: Vendor unknown, using generic init.\n");
-		printk(KERN_ERR "CPU: Your system may be unstable.\n");
-	}
-	c->x86_vendor = X86_VENDOR_UNKNOWN;
-}
-
-static void __init early_cpu_support_print(void)
-{
-	int i,j;
-	struct cpu_dev *cpu_devx;
-
-	printk("KERNEL supported cpus:\n");
-	for (i = 0; i < X86_VENDOR_NUM; i++) {
-		cpu_devx = cpu_devs[i];
-		if (!cpu_devx)
-			continue;
-		for (j = 0; j < 2; j++) {
-			if (!cpu_devx->c_ident[j])
-				continue;
-			printk("  %s %s\n", cpu_devx->c_vendor,
-				cpu_devx->c_ident[j]);
-		}
-	}
-}
-
-static void __init early_cpu_init(void)
-{
-        struct cpu_vendor_dev *cvdev;
-
-        for (cvdev = __x86cpuvendor_start ;
-             cvdev < __x86cpuvendor_end   ;
-             cvdev++)
-                cpu_devs[cvdev->vendor] = cvdev->cpu_dev;
-	early_cpu_support_print();
-}
-
-/* Do some early cpuid on the boot CPU to get some parameter that are
-   needed before check_bugs. Everything advanced is in identify_cpu
-   below. */
-static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c)
-{
-	u32 tfms, xlvl;
-
-	c->loops_per_jiffy = loops_per_jiffy;
-	c->x86_cache_size = -1;
-	c->x86_vendor = X86_VENDOR_UNKNOWN;
-	c->x86_model = c->x86_mask = 0;	/* So far unknown... */
-	c->x86_vendor_id[0] = '\0'; /* Unset */
-	c->x86_model_id[0] = '\0';  /* Unset */
-	c->x86_clflush_size = 64;
-	c->x86_cache_alignment = c->x86_clflush_size;
-	c->x86_max_cores = 1;
-	c->x86_coreid_bits = 0;
-	c->extended_cpuid_level = 0;
-	memset(&c->x86_capability, 0, sizeof c->x86_capability);
-
-	/* Get vendor name */
-	cpuid(0x00000000, (unsigned int *)&c->cpuid_level,
-	      (unsigned int *)&c->x86_vendor_id[0],
-	      (unsigned int *)&c->x86_vendor_id[8],
-	      (unsigned int *)&c->x86_vendor_id[4]);
-
-	get_cpu_vendor(c);
-
-	/* Initialize the standard set of capabilities */
-	/* Note that the vendor-specific code below might override */
-
-	/* Intel-defined flags: level 0x00000001 */
-	if (c->cpuid_level >= 0x00000001) {
-		__u32 misc;
-		cpuid(0x00000001, &tfms, &misc, &c->x86_capability[4],
-		      &c->x86_capability[0]);
-		c->x86 = (tfms >> 8) & 0xf;
-		c->x86_model = (tfms >> 4) & 0xf;
-		c->x86_mask = tfms & 0xf;
-		if (c->x86 == 0xf)
-			c->x86 += (tfms >> 20) & 0xff;
-		if (c->x86 >= 0x6)
-			c->x86_model += ((tfms >> 16) & 0xF) << 4;
-		if (test_cpu_cap(c, X86_FEATURE_CLFLSH))
-			c->x86_clflush_size = ((misc >> 8) & 0xff) * 8;
-	} else {
-		/* Have CPUID level 0 only - unheard of */
-		c->x86 = 4;
-	}
-
-	c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xff;
-#ifdef CONFIG_SMP
-	c->phys_proc_id = c->initial_apicid;
-#endif
-	/* AMD-defined flags: level 0x80000001 */
-	xlvl = cpuid_eax(0x80000000);
-	c->extended_cpuid_level = xlvl;
-	if ((xlvl & 0xffff0000) == 0x80000000) {
-		if (xlvl >= 0x80000001) {
-			c->x86_capability[1] = cpuid_edx(0x80000001);
-			c->x86_capability[6] = cpuid_ecx(0x80000001);
-		}
-		if (xlvl >= 0x80000004)
-			get_model_name(c); /* Default name */
-	}
-
-	/* Transmeta-defined flags: level 0x80860001 */
-	xlvl = cpuid_eax(0x80860000);
-	if ((xlvl & 0xffff0000) == 0x80860000) {
-		/* Don't set x86_cpuid_level here for now to not confuse. */
-		if (xlvl >= 0x80860001)
-			c->x86_capability[2] = cpuid_edx(0x80860001);
-	}
-
-	c->extended_cpuid_level = cpuid_eax(0x80000000);
-	if (c->extended_cpuid_level >= 0x80000007)
-		c->x86_power = cpuid_edx(0x80000007);
-
-	if (c->x86_vendor != X86_VENDOR_UNKNOWN &&
-	    cpu_devs[c->x86_vendor]->c_early_init)
-		cpu_devs[c->x86_vendor]->c_early_init(c);
-
-	validate_pat_support(c);
-
-	/* early_param could clear that, but recall get it set again */
-	if (disable_apic)
-		clear_cpu_cap(c, X86_FEATURE_APIC);
-}
-
-/*
- * This does the hard work of actually picking apart the CPU stuff...
- */
-void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
-{
-	int i;
-
-	early_identify_cpu(c);
-
-	init_scattered_cpuid_features(c);
-
-	c->apicid = phys_pkg_id(0);
-
-	/*
-	 * Vendor-specific initialization.  In this section we
-	 * canonicalize the feature flags, meaning if there are
-	 * features a certain CPU supports which CPUID doesn't
-	 * tell us, CPUID claiming incorrect flags, or other bugs,
-	 * we handle them here.
-	 *
-	 * At the end of this section, c->x86_capability better
-	 * indicate the features this CPU genuinely supports!
-	 */
-	if (this_cpu->c_init)
-		this_cpu->c_init(c);
-
-	detect_ht(c);
-
-	/*
-	 * On SMP, boot_cpu_data holds the common feature set between
-	 * all CPUs; so make sure that we indicate which features are
-	 * common between the CPUs.  The first time this routine gets
-	 * executed, c == &boot_cpu_data.
-	 */
-	if (c != &boot_cpu_data) {
-		/* AND the already accumulated flags with these */
-		for (i = 0; i < NCAPINTS; i++)
-			boot_cpu_data.x86_capability[i] &= c->x86_capability[i];
-	}
-
-	/* Clear all flags overriden by options */
-	for (i = 0; i < NCAPINTS; i++)
-		c->x86_capability[i] &= ~cleared_cpu_caps[i];
-
-#ifdef CONFIG_X86_MCE
-	mcheck_init(c);
-#endif
-	select_idle_routine(c);
-
-#ifdef CONFIG_NUMA
-	numa_add_cpu(smp_processor_id());
-#endif
-
-}
-
-void __cpuinit identify_boot_cpu(void)
-{
-	identify_cpu(&boot_cpu_data);
-}
-
-void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c)
-{
-	BUG_ON(c == &boot_cpu_data);
-	identify_cpu(c);
-	mtrr_ap_init();
-}
-
-static __init int setup_noclflush(char *arg)
-{
-	setup_clear_cpu_cap(X86_FEATURE_CLFLSH);
-	return 1;
-}
-__setup("noclflush", setup_noclflush);
-
-void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
-{
-	if (c->x86_model_id[0])
-		printk(KERN_CONT "%s", c->x86_model_id);
-
-	if (c->x86_mask || c->cpuid_level >= 0)
-		printk(KERN_CONT " stepping %02x\n", c->x86_mask);
-	else
-		printk(KERN_CONT "\n");
-}
-
-static __init int setup_disablecpuid(char *arg)
-{
-	int bit;
-	if (get_option(&arg, &bit) && bit < NCAPINTS*32)
-		setup_clear_cpu_cap(bit);
-	else
-		return 0;
-	return 1;
-}
-__setup("clearcpuid=", setup_disablecpuid);
Index: linux-2.6/arch/x86/kernel/cpu/common_64.c
===================================================================
--- /dev/null
+++ linux-2.6/arch/x86/kernel/cpu/common_64.c
@@ -0,0 +1,406 @@
+#include <linux/init.h>
+#include <linux/string.h>
+#include <linux/delay.h>
+#include <linux/smp.h>
+#include <linux/module.h>
+#include <linux/percpu.h>
+#include <linux/bootmem.h>
+#include <asm/processor.h>
+#include <asm/i387.h>
+#include <asm/msr.h>
+#include <asm/io.h>
+#include <asm/mmu_context.h>
+#include <asm/mtrr.h>
+#include <asm/mce.h>
+#include <asm/pat.h>
+#include <asm/numa.h>
+#ifdef CONFIG_X86_LOCAL_APIC
+#include <asm/mpspec.h>
+#include <asm/apic.h>
+#include <mach_apic.h>
+#endif
+
+#include "cpu.h"
+
+/* We need valid kernel segments for data and code in long mode too
+ * IRET will check the segment types  kkeil 2000/10/28
+ * Also sysret mandates a special GDT layout
+ */
+/* The TLS descriptors are currently at a different place compared to i386.
+   Hopefully nobody expects them at a fixed place (Wine?) */
+DEFINE_PER_CPU(struct gdt_page, gdt_page) = { .gdt = {
+	[GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } },
+	[GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } },
+	[GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } },
+	[GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } },
+	[GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } },
+	[GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } },
+} };
+EXPORT_PER_CPU_SYMBOL_GPL(gdt_page);
+
+__u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata;
+
+/* Current gdt points %fs at the "master" per-cpu area: after this,
+ *  * it's on the real one. */
+void switch_to_new_gdt(void)
+{
+	struct desc_ptr gdt_descr;
+
+	gdt_descr.address = (long)get_cpu_gdt_table(smp_processor_id());
+	gdt_descr.size = GDT_SIZE - 1;
+	load_gdt(&gdt_descr);
+}
+
+struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {};
+
+static void __cpuinit default_init(struct cpuinfo_x86 *c)
+{
+	display_cacheinfo(c);
+}
+
+static struct cpu_dev __cpuinitdata default_cpu = {
+	.c_init	= default_init,
+	.c_vendor = "Unknown",
+};
+static struct cpu_dev *this_cpu __cpuinitdata = &default_cpu;
+
+int __cpuinit get_model_name(struct cpuinfo_x86 *c)
+{
+	unsigned int *v;
+
+	if (c->extended_cpuid_level < 0x80000004)
+		return 0;
+
+	v = (unsigned int *) c->x86_model_id;
+	cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]);
+	cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]);
+	cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]);
+	c->x86_model_id[48] = 0;
+	return 1;
+}
+
+
+void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
+{
+	unsigned int n, dummy, eax, ebx, ecx, edx;
+
+	n = c->extended_cpuid_level;
+
+	if (n >= 0x80000005) {
+		cpuid(0x80000005, &dummy, &ebx, &ecx, &edx);
+		printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), "
+		       "D cache %dK (%d bytes/line)\n",
+		       edx>>24, edx&0xFF, ecx>>24, ecx&0xFF);
+		c->x86_cache_size = (ecx>>24) + (edx>>24);
+		/* On K8 L1 TLB is inclusive, so don't count it */
+		c->x86_tlbsize = 0;
+	}
+
+	if (n >= 0x80000006) {
+		cpuid(0x80000006, &dummy, &ebx, &ecx, &edx);
+		ecx = cpuid_ecx(0x80000006);
+		c->x86_cache_size = ecx >> 16;
+		c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff);
+
+		printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
+		c->x86_cache_size, ecx & 0xFF);
+	}
+	if (n >= 0x80000008) {
+		cpuid(0x80000008, &eax, &dummy, &dummy, &dummy);
+		c->x86_virt_bits = (eax >> 8) & 0xff;
+		c->x86_phys_bits = eax & 0xff;
+	}
+}
+
+void __cpuinit detect_ht(struct cpuinfo_x86 *c)
+{
+#ifdef CONFIG_SMP
+	u32 eax, ebx, ecx, edx;
+	int index_msb, core_bits;
+
+	cpuid(1, &eax, &ebx, &ecx, &edx);
+
+
+	if (!cpu_has(c, X86_FEATURE_HT))
+		return;
+	if (cpu_has(c, X86_FEATURE_CMP_LEGACY))
+		goto out;
+
+	smp_num_siblings = (ebx & 0xff0000) >> 16;
+
+	if (smp_num_siblings == 1) {
+		printk(KERN_INFO  "CPU: Hyper-Threading is disabled\n");
+	} else if (smp_num_siblings > 1) {
+
+		if (smp_num_siblings > NR_CPUS) {
+			printk(KERN_WARNING "CPU: Unsupported number of "
+			       "siblings %d", smp_num_siblings);
+			smp_num_siblings = 1;
+			return;
+		}
+
+		index_msb = get_count_order(smp_num_siblings);
+		c->phys_proc_id = phys_pkg_id(index_msb);
+
+		smp_num_siblings = smp_num_siblings / c->x86_max_cores;
+
+		index_msb = get_count_order(smp_num_siblings);
+
+		core_bits = get_count_order(c->x86_max_cores);
+
+		c->cpu_core_id = phys_pkg_id(index_msb) &
+					       ((1 << core_bits) - 1);
+	}
+out:
+	if ((c->x86_max_cores * smp_num_siblings) > 1) {
+		printk(KERN_INFO  "CPU: Physical Processor ID: %d\n",
+		       c->phys_proc_id);
+		printk(KERN_INFO  "CPU: Processor Core ID: %d\n",
+		       c->cpu_core_id);
+	}
+
+#endif
+}
+
+static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
+{
+	char *v = c->x86_vendor_id;
+	int i;
+	static int printed;
+
+	for (i = 0; i < X86_VENDOR_NUM; i++) {
+		if (cpu_devs[i]) {
+			if (!strcmp(v, cpu_devs[i]->c_ident[0]) ||
+			    (cpu_devs[i]->c_ident[1] &&
+			    !strcmp(v, cpu_devs[i]->c_ident[1]))) {
+				c->x86_vendor = i;
+				this_cpu = cpu_devs[i];
+				return;
+			}
+		}
+	}
+	if (!printed) {
+		printed++;
+		printk(KERN_ERR "CPU: Vendor unknown, using generic init.\n");
+		printk(KERN_ERR "CPU: Your system may be unstable.\n");
+	}
+	c->x86_vendor = X86_VENDOR_UNKNOWN;
+}
+
+static void __init early_cpu_support_print(void)
+{
+	int i,j;
+	struct cpu_dev *cpu_devx;
+
+	printk("KERNEL supported cpus:\n");
+	for (i = 0; i < X86_VENDOR_NUM; i++) {
+		cpu_devx = cpu_devs[i];
+		if (!cpu_devx)
+			continue;
+		for (j = 0; j < 2; j++) {
+			if (!cpu_devx->c_ident[j])
+				continue;
+			printk("  %s %s\n", cpu_devx->c_vendor,
+				cpu_devx->c_ident[j]);
+		}
+	}
+}
+
+static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c);
+
+void __init early_cpu_init(void)
+{
+        struct cpu_vendor_dev *cvdev;
+
+        for (cvdev = __x86cpuvendor_start ;
+             cvdev < __x86cpuvendor_end   ;
+             cvdev++)
+                cpu_devs[cvdev->vendor] = cvdev->cpu_dev;
+	early_cpu_support_print();
+	early_identify_cpu(&boot_cpu_data);
+}
+
+/* Do some early cpuid on the boot CPU to get some parameter that are
+   needed before check_bugs. Everything advanced is in identify_cpu
+   below. */
+static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c)
+{
+	u32 tfms, xlvl;
+
+	c->loops_per_jiffy = loops_per_jiffy;
+	c->x86_cache_size = -1;
+	c->x86_vendor = X86_VENDOR_UNKNOWN;
+	c->x86_model = c->x86_mask = 0;	/* So far unknown... */
+	c->x86_vendor_id[0] = '\0'; /* Unset */
+	c->x86_model_id[0] = '\0';  /* Unset */
+	c->x86_clflush_size = 64;
+	c->x86_cache_alignment = c->x86_clflush_size;
+	c->x86_max_cores = 1;
+	c->x86_coreid_bits = 0;
+	c->extended_cpuid_level = 0;
+	memset(&c->x86_capability, 0, sizeof c->x86_capability);
+
+	/* Get vendor name */
+	cpuid(0x00000000, (unsigned int *)&c->cpuid_level,
+	      (unsigned int *)&c->x86_vendor_id[0],
+	      (unsigned int *)&c->x86_vendor_id[8],
+	      (unsigned int *)&c->x86_vendor_id[4]);
+
+	get_cpu_vendor(c);
+
+	/* Initialize the standard set of capabilities */
+	/* Note that the vendor-specific code below might override */
+
+	/* Intel-defined flags: level 0x00000001 */
+	if (c->cpuid_level >= 0x00000001) {
+		__u32 misc;
+		cpuid(0x00000001, &tfms, &misc, &c->x86_capability[4],
+		      &c->x86_capability[0]);
+		c->x86 = (tfms >> 8) & 0xf;
+		c->x86_model = (tfms >> 4) & 0xf;
+		c->x86_mask = tfms & 0xf;
+		if (c->x86 == 0xf)
+			c->x86 += (tfms >> 20) & 0xff;
+		if (c->x86 >= 0x6)
+			c->x86_model += ((tfms >> 16) & 0xF) << 4;
+		if (test_cpu_cap(c, X86_FEATURE_CLFLSH))
+			c->x86_clflush_size = ((misc >> 8) & 0xff) * 8;
+	} else {
+		/* Have CPUID level 0 only - unheard of */
+		c->x86 = 4;
+	}
+
+	c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xff;
+#ifdef CONFIG_SMP
+	c->phys_proc_id = c->initial_apicid;
+#endif
+	/* AMD-defined flags: level 0x80000001 */
+	xlvl = cpuid_eax(0x80000000);
+	c->extended_cpuid_level = xlvl;
+	if ((xlvl & 0xffff0000) == 0x80000000) {
+		if (xlvl >= 0x80000001) {
+			c->x86_capability[1] = cpuid_edx(0x80000001);
+			c->x86_capability[6] = cpuid_ecx(0x80000001);
+		}
+		if (xlvl >= 0x80000004)
+			get_model_name(c); /* Default name */
+	}
+
+	/* Transmeta-defined flags: level 0x80860001 */
+	xlvl = cpuid_eax(0x80860000);
+	if ((xlvl & 0xffff0000) == 0x80860000) {
+		/* Don't set x86_cpuid_level here for now to not confuse. */
+		if (xlvl >= 0x80860001)
+			c->x86_capability[2] = cpuid_edx(0x80860001);
+	}
+
+	c->extended_cpuid_level = cpuid_eax(0x80000000);
+	if (c->extended_cpuid_level >= 0x80000007)
+		c->x86_power = cpuid_edx(0x80000007);
+
+	if (c->x86_vendor != X86_VENDOR_UNKNOWN &&
+	    cpu_devs[c->x86_vendor]->c_early_init)
+		cpu_devs[c->x86_vendor]->c_early_init(c);
+
+	validate_pat_support(c);
+
+	/* early_param could clear that, but recall get it set again */
+	if (disable_apic)
+		clear_cpu_cap(c, X86_FEATURE_APIC);
+}
+
+/*
+ * This does the hard work of actually picking apart the CPU stuff...
+ */
+void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
+{
+	int i;
+
+	early_identify_cpu(c);
+
+	init_scattered_cpuid_features(c);
+
+	c->apicid = phys_pkg_id(0);
+
+	/*
+	 * Vendor-specific initialization.  In this section we
+	 * canonicalize the feature flags, meaning if there are
+	 * features a certain CPU supports which CPUID doesn't
+	 * tell us, CPUID claiming incorrect flags, or other bugs,
+	 * we handle them here.
+	 *
+	 * At the end of this section, c->x86_capability better
+	 * indicate the features this CPU genuinely supports!
+	 */
+	if (this_cpu->c_init)
+		this_cpu->c_init(c);
+
+	detect_ht(c);
+
+	/*
+	 * On SMP, boot_cpu_data holds the common feature set between
+	 * all CPUs; so make sure that we indicate which features are
+	 * common between the CPUs.  The first time this routine gets
+	 * executed, c == &boot_cpu_data.
+	 */
+	if (c != &boot_cpu_data) {
+		/* AND the already accumulated flags with these */
+		for (i = 0; i < NCAPINTS; i++)
+			boot_cpu_data.x86_capability[i] &= c->x86_capability[i];
+	}
+
+	/* Clear all flags overriden by options */
+	for (i = 0; i < NCAPINTS; i++)
+		c->x86_capability[i] &= ~cleared_cpu_caps[i];
+
+#ifdef CONFIG_X86_MCE
+	mcheck_init(c);
+#endif
+	select_idle_routine(c);
+
+#ifdef CONFIG_NUMA
+	numa_add_cpu(smp_processor_id());
+#endif
+
+}
+
+void __cpuinit identify_boot_cpu(void)
+{
+	identify_cpu(&boot_cpu_data);
+}
+
+void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c)
+{
+	BUG_ON(c == &boot_cpu_data);
+	identify_cpu(c);
+	mtrr_ap_init();
+}
+
+static __init int setup_noclflush(char *arg)
+{
+	setup_clear_cpu_cap(X86_FEATURE_CLFLSH);
+	return 1;
+}
+__setup("noclflush", setup_noclflush);
+
+void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
+{
+	if (c->x86_model_id[0])
+		printk(KERN_CONT "%s", c->x86_model_id);
+
+	if (c->x86_mask || c->cpuid_level >= 0)
+		printk(KERN_CONT " stepping %02x\n", c->x86_mask);
+	else
+		printk(KERN_CONT "\n");
+}
+
+static __init int setup_disablecpuid(char *arg)
+{
+	int bit;
+	if (get_option(&arg, &bit) && bit < NCAPINTS*32)
+		setup_clear_cpu_cap(bit);
+	else
+		return 0;
+	return 1;
+}
+__setup("clearcpuid=", setup_disablecpuid);
Index: linux-2.6/arch/x86/kernel/cpu/Makefile
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/Makefile
+++ linux-2.6/arch/x86/kernel/cpu/Makefile
@@ -6,7 +6,7 @@ obj-y			:= intel_cacheinfo.o addon_cpuid
 obj-y			+= proc.o feature_names.o
 
 obj-$(CONFIG_X86_32)	+= common.o bugs.o
-obj-$(CONFIG_X86_64)	+= bugs_64.o
+obj-$(CONFIG_X86_64)	+= common_64.o bugs_64.o
 obj-$(CONFIG_X86_32)	+= amd.o
 obj-$(CONFIG_X86_64)	+= amd_64.o
 obj-$(CONFIG_X86_32)	+= cyrix.o
Index: linux-2.6/include/asm-x86/processor.h
===================================================================
--- linux-2.6.orig/include/asm-x86/processor.h
+++ linux-2.6/include/asm-x86/processor.h
@@ -154,6 +154,7 @@ static inline int hlt_works(int cpu)
 
 extern void cpu_detect(struct cpuinfo_x86 *c);
 
+extern void early_cpu_init(void);
 extern void identify_cpu(struct cpuinfo_x86 *);
 extern void identify_boot_cpu(void);
 extern void identify_secondary_cpu(struct cpuinfo_x86 *);

[PATCH] x86: merge setup64.c into common_64.c

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/Makefile        |    2 
 arch/x86/kernel/cpu/common_64.c |  277 ++++++++++++++++++++++++++++++++++++++
 arch/x86/kernel/setup64.c       |  287 ----------------------------------------
 3 files changed, 277 insertions(+), 289 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup64.c
+++ /dev/null
@@ -1,287 +0,0 @@
-/* 
- * X86-64 specific CPU setup.
- * Copyright (C) 1995  Linus Torvalds
- * Copyright 2001, 2002, 2003 SuSE Labs / Andi Kleen.
- * See setup.c for older changelog.
- */ 
-#include <linux/init.h>
-#include <linux/kernel.h>
-#include <linux/sched.h>
-#include <linux/string.h>
-#include <linux/bootmem.h>
-#include <linux/bitops.h>
-#include <linux/module.h>
-#include <linux/kgdb.h>
-#include <linux/topology.h>
-#include <asm/pda.h>
-#include <asm/pgtable.h>
-#include <asm/processor.h>
-#include <asm/desc.h>
-#include <asm/atomic.h>
-#include <asm/mmu_context.h>
-#include <asm/smp.h>
-#include <asm/i387.h>
-#include <asm/percpu.h>
-#include <asm/proto.h>
-#include <asm/sections.h>
-#include <asm/setup.h>
-#include <asm/genapic.h>
-
-#ifndef CONFIG_DEBUG_BOOT_PARAMS
-struct boot_params __initdata boot_params;
-#else
-struct boot_params boot_params;
-#endif
-
-cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
-
-struct x8664_pda **_cpu_pda __read_mostly;
-EXPORT_SYMBOL(_cpu_pda);
-
-struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
-
-char boot_cpu_stack[IRQSTACKSIZE] __page_aligned_bss;
-
-unsigned long __supported_pte_mask __read_mostly = ~0UL;
-EXPORT_SYMBOL_GPL(__supported_pte_mask);
-
-static int do_not_nx __cpuinitdata = 0;
-
-/* noexec=on|off
-Control non executable mappings for 64bit processes.
-
-on	Enable(default)
-off	Disable
-*/ 
-static int __init nonx_setup(char *str)
-{
-	if (!str)
-		return -EINVAL;
-	if (!strncmp(str, "on", 2)) {
-                __supported_pte_mask |= _PAGE_NX; 
- 		do_not_nx = 0; 
-	} else if (!strncmp(str, "off", 3)) {
-		do_not_nx = 1;
-		__supported_pte_mask &= ~_PAGE_NX;
-        }
-	return 0;
-} 
-early_param("noexec", nonx_setup);
-
-int force_personality32 = 0; 
-
-/* noexec32=on|off
-Control non executable heap for 32bit processes.
-To control the stack too use noexec=off
-
-on	PROT_READ does not imply PROT_EXEC for 32bit processes (default)
-off	PROT_READ implies PROT_EXEC
-*/
-static int __init nonx32_setup(char *str)
-{
-	if (!strcmp(str, "on"))
-		force_personality32 &= ~READ_IMPLIES_EXEC;
-	else if (!strcmp(str, "off"))
-		force_personality32 |= READ_IMPLIES_EXEC;
-	return 1;
-}
-__setup("noexec32=", nonx32_setup);
-
-void pda_init(int cpu)
-{ 
-	struct x8664_pda *pda = cpu_pda(cpu);
-
-	/* Setup up data that may be needed in __get_free_pages early */
-	asm volatile("movl %0,%%fs ; movl %0,%%gs" :: "r" (0)); 
-	/* Memory clobbers used to order PDA accessed */
-	mb();
-	wrmsrl(MSR_GS_BASE, pda);
-	mb();
-
-	pda->cpunumber = cpu; 
-	pda->irqcount = -1;
-	pda->kernelstack = 
-		(unsigned long)stack_thread_info() - PDA_STACKOFFSET + THREAD_SIZE; 
-	pda->active_mm = &init_mm;
-	pda->mmu_state = 0;
-
-	if (cpu == 0) {
-		/* others are initialized in smpboot.c */
-		pda->pcurrent = &init_task;
-		pda->irqstackptr = boot_cpu_stack; 
-	} else {
-		pda->irqstackptr = (char *)
-			__get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER);
-		if (!pda->irqstackptr)
-			panic("cannot allocate irqstack for cpu %d", cpu); 
-
-		if (pda->nodenumber == 0 && cpu_to_node(cpu) != NUMA_NO_NODE)
-			pda->nodenumber = cpu_to_node(cpu);
-	}
-
-	pda->irqstackptr += IRQSTACKSIZE-64;
-} 
-
-char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]
-__attribute__((section(".bss.page_aligned")));
-
-extern asmlinkage void ignore_sysret(void);
-
-/* May not be marked __init: used by software suspend */
-void syscall_init(void)
-{
-	/* 
-	 * LSTAR and STAR live in a bit strange symbiosis.
-	 * They both write to the same internal register. STAR allows to set CS/DS
-	 * but only a 32bit target. LSTAR sets the 64bit rip. 	 
-	 */ 
-	wrmsrl(MSR_STAR,  ((u64)__USER32_CS)<<48  | ((u64)__KERNEL_CS)<<32); 
-	wrmsrl(MSR_LSTAR, system_call); 
-	wrmsrl(MSR_CSTAR, ignore_sysret);
-
-#ifdef CONFIG_IA32_EMULATION   		
-	syscall32_cpu_init ();
-#endif
-
-	/* Flags to clear on syscall */
-	wrmsrl(MSR_SYSCALL_MASK,
-	       X86_EFLAGS_TF|X86_EFLAGS_DF|X86_EFLAGS_IF|X86_EFLAGS_IOPL);
-}
-
-void __cpuinit check_efer(void)
-{
-	unsigned long efer;
-
-	rdmsrl(MSR_EFER, efer); 
-        if (!(efer & EFER_NX) || do_not_nx) { 
-                __supported_pte_mask &= ~_PAGE_NX; 
-        }       
-}
-
-unsigned long kernel_eflags;
-
-/*
- * Copies of the original ist values from the tss are only accessed during
- * debugging, no special alignment required.
- */
-DEFINE_PER_CPU(struct orig_ist, orig_ist);
-
-/*
- * cpu_init() initializes state that is per-CPU. Some data is already
- * initialized (naturally) in the bootstrap process, such as the GDT
- * and IDT. We reload them nevertheless, this function acts as a
- * 'CPU state barrier', nothing should get across.
- * A lot of state is already set up in PDA init.
- */
-void __cpuinit cpu_init (void)
-{
-	int cpu = stack_smp_processor_id();
-	struct tss_struct *t = &per_cpu(init_tss, cpu);
-	struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu);
-	unsigned long v; 
-	char *estacks = NULL; 
-	struct task_struct *me;
-	int i;
-
-	/* CPU 0 is initialised in head64.c */
-	if (cpu != 0) {
-		pda_init(cpu);
-	} else 
-		estacks = boot_exception_stacks; 
-
-	me = current;
-
-	if (cpu_test_and_set(cpu, cpu_initialized))
-		panic("CPU#%d already initialized!\n", cpu);
-
-	printk("Initializing CPU#%d\n", cpu);
-
-	clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
-
-	/*
-	 * Initialize the per-CPU GDT with the boot GDT,
-	 * and set up the GDT descriptor:
-	 */
-
-	switch_to_new_gdt();
-	load_idt((const struct desc_ptr *)&idt_descr);
-
-	memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
-	syscall_init();
-
-	wrmsrl(MSR_FS_BASE, 0);
-	wrmsrl(MSR_KERNEL_GS_BASE, 0);
-	barrier(); 
-
-	check_efer();
-
-	/*
-	 * set up and load the per-CPU TSS
-	 */
-	for (v = 0; v < N_EXCEPTION_STACKS; v++) {
-		static const unsigned int order[N_EXCEPTION_STACKS] = {
-			[0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER,
-			[DEBUG_STACK - 1] = DEBUG_STACK_ORDER
-		};
-		if (cpu) {
-			estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]);
-			if (!estacks)
-				panic("Cannot allocate exception stack %ld %d\n",
-				      v, cpu); 
-		}
-		estacks += PAGE_SIZE << order[v];
-		orig_ist->ist[v] = t->x86_tss.ist[v] = (unsigned long)estacks;
-	}
-
-	t->x86_tss.io_bitmap_base = offsetof(struct tss_struct, io_bitmap);
-	/*
-	 * <= is required because the CPU will access up to
-	 * 8 bits beyond the end of the IO permission bitmap.
-	 */
-	for (i = 0; i <= IO_BITMAP_LONGS; i++)
-		t->io_bitmap[i] = ~0UL;
-
-	atomic_inc(&init_mm.mm_count);
-	me->active_mm = &init_mm;
-	if (me->mm)
-		BUG();
-	enter_lazy_tlb(&init_mm, me);
-
-	load_sp0(t, &current->thread);
-	set_tss_desc(cpu, t);
-	load_TR_desc();
-	load_LDT(&init_mm.context);
-
-#ifdef CONFIG_KGDB
-	/*
-	 * If the kgdb is connected no debug regs should be altered.  This
-	 * is only applicable when KGDB and a KGDB I/O module are built
-	 * into the kernel and you are using early debugging with
-	 * kgdbwait. KGDB will control the kernel HW breakpoint registers.
-	 */
-	if (kgdb_connected && arch_kgdb_ops.correct_hw_break)
-		arch_kgdb_ops.correct_hw_break();
-	else {
-#endif
-	/*
-	 * Clear all 6 debug registers:
-	 */
-
-	set_debugreg(0UL, 0);
-	set_debugreg(0UL, 1);
-	set_debugreg(0UL, 2);
-	set_debugreg(0UL, 3);
-	set_debugreg(0UL, 6);
-	set_debugreg(0UL, 7);
-#ifdef CONFIG_KGDB
-	/* If the kgdb is connected no debug regs should be altered. */
-	}
-#endif
-
-	fpu_init(); 
-
-	raw_local_save_flags(kernel_eflags);
-
-	if (is_uv_system())
-		uv_cpu_init();
-}
Index: linux-2.6/arch/x86/kernel/Makefile
===================================================================
--- linux-2.6.orig/arch/x86/kernel/Makefile
+++ linux-2.6/arch/x86/kernel/Makefile
@@ -30,7 +30,7 @@ obj-y			+= setup_$(BITS).o i8259.o irqin
 obj-$(CONFIG_X86_32)	+= probe_roms_32.o
 obj-$(CONFIG_X86_32)	+= sys_i386_32.o i386_ksyms_32.o
 obj-$(CONFIG_X86_64)	+= sys_x86_64.o x8664_ksyms_64.o
-obj-$(CONFIG_X86_64)	+= syscall_64.o vsyscall_64.o setup64.o
+obj-$(CONFIG_X86_64)	+= syscall_64.o vsyscall_64.o
 obj-y			+= bootflag.o e820.o
 obj-y			+= pci-dma.o quirks.o i8237.o topology.o kdebugfs.o
 obj-y			+= alternative.o i8253.o pci-nommu.o
Index: linux-2.6/arch/x86/kernel/cpu/common_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/common_64.c
+++ linux-2.6/arch/x86/kernel/cpu/common_64.c
@@ -1,10 +1,17 @@
 #include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/string.h>
+#include <linux/bootmem.h>
+#include <linux/bitops.h>
+#include <linux/module.h>
+#include <linux/kgdb.h>
+#include <linux/topology.h>
 #include <linux/string.h>
 #include <linux/delay.h>
 #include <linux/smp.h>
 #include <linux/module.h>
 #include <linux/percpu.h>
-#include <linux/bootmem.h>
 #include <asm/processor.h>
 #include <asm/i387.h>
 #include <asm/msr.h>
@@ -19,6 +26,15 @@
 #include <asm/apic.h>
 #include <mach_apic.h>
 #endif
+#include <asm/pda.h>
+#include <asm/pgtable.h>
+#include <asm/processor.h>
+#include <asm/desc.h>
+#include <asm/atomic.h>
+#include <asm/proto.h>
+#include <asm/sections.h>
+#include <asm/setup.h>
+#include <asm/genapic.h>
 
 #include "cpu.h"
 
@@ -404,3 +420,262 @@ static __init int setup_disablecpuid(cha
 	return 1;
 }
 __setup("clearcpuid=", setup_disablecpuid);
+
+#ifndef CONFIG_DEBUG_BOOT_PARAMS
+struct boot_params __initdata boot_params;
+#else
+struct boot_params boot_params;
+#endif
+
+cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
+
+struct x8664_pda **_cpu_pda __read_mostly;
+EXPORT_SYMBOL(_cpu_pda);
+
+struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
+
+char boot_cpu_stack[IRQSTACKSIZE] __page_aligned_bss;
+
+unsigned long __supported_pte_mask __read_mostly = ~0UL;
+EXPORT_SYMBOL_GPL(__supported_pte_mask);
+
+static int do_not_nx __cpuinitdata;
+
+/* noexec=on|off
+Control non executable mappings for 64bit processes.
+
+on	Enable(default)
+off	Disable
+*/
+static int __init nonx_setup(char *str)
+{
+	if (!str)
+		return -EINVAL;
+	if (!strncmp(str, "on", 2)) {
+		__supported_pte_mask |= _PAGE_NX;
+		do_not_nx = 0;
+	} else if (!strncmp(str, "off", 3)) {
+		do_not_nx = 1;
+		__supported_pte_mask &= ~_PAGE_NX;
+	}
+	return 0;
+}
+early_param("noexec", nonx_setup);
+
+int force_personality32;
+
+/* noexec32=on|off
+Control non executable heap for 32bit processes.
+To control the stack too use noexec=off
+
+on	PROT_READ does not imply PROT_EXEC for 32bit processes (default)
+off	PROT_READ implies PROT_EXEC
+*/
+static int __init nonx32_setup(char *str)
+{
+	if (!strcmp(str, "on"))
+		force_personality32 &= ~READ_IMPLIES_EXEC;
+	else if (!strcmp(str, "off"))
+		force_personality32 |= READ_IMPLIES_EXEC;
+	return 1;
+}
+__setup("noexec32=", nonx32_setup);
+
+void pda_init(int cpu)
+{
+	struct x8664_pda *pda = cpu_pda(cpu);
+
+	/* Setup up data that may be needed in __get_free_pages early */
+	asm volatile("movl %0,%%fs ; movl %0,%%gs" :: "r" (0));
+	/* Memory clobbers used to order PDA accessed */
+	mb();
+	wrmsrl(MSR_GS_BASE, pda);
+	mb();
+
+	pda->cpunumber = cpu;
+	pda->irqcount = -1;
+	pda->kernelstack = (unsigned long)stack_thread_info() -
+				 PDA_STACKOFFSET + THREAD_SIZE;
+	pda->active_mm = &init_mm;
+	pda->mmu_state = 0;
+
+	if (cpu == 0) {
+		/* others are initialized in smpboot.c */
+		pda->pcurrent = &init_task;
+		pda->irqstackptr = boot_cpu_stack;
+	} else {
+		pda->irqstackptr = (char *)
+			__get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER);
+		if (!pda->irqstackptr)
+			panic("cannot allocate irqstack for cpu %d", cpu);
+
+		if (pda->nodenumber == 0 && cpu_to_node(cpu) != NUMA_NO_NODE)
+			pda->nodenumber = cpu_to_node(cpu);
+	}
+
+	pda->irqstackptr += IRQSTACKSIZE-64;
+}
+
+char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ +
+			   DEBUG_STKSZ]
+__attribute__((section(".bss.page_aligned")));
+
+extern asmlinkage void ignore_sysret(void);
+
+/* May not be marked __init: used by software suspend */
+void syscall_init(void)
+{
+	/*
+	 * LSTAR and STAR live in a bit strange symbiosis.
+	 * They both write to the same internal register. STAR allows to
+	 * set CS/DS but only a 32bit target. LSTAR sets the 64bit rip.
+	 */
+	wrmsrl(MSR_STAR,  ((u64)__USER32_CS)<<48  | ((u64)__KERNEL_CS)<<32);
+	wrmsrl(MSR_LSTAR, system_call);
+	wrmsrl(MSR_CSTAR, ignore_sysret);
+
+#ifdef CONFIG_IA32_EMULATION
+	syscall32_cpu_init();
+#endif
+
+	/* Flags to clear on syscall */
+	wrmsrl(MSR_SYSCALL_MASK,
+	       X86_EFLAGS_TF|X86_EFLAGS_DF|X86_EFLAGS_IF|X86_EFLAGS_IOPL);
+}
+
+void __cpuinit check_efer(void)
+{
+	unsigned long efer;
+
+	rdmsrl(MSR_EFER, efer);
+	if (!(efer & EFER_NX) || do_not_nx)
+		__supported_pte_mask &= ~_PAGE_NX;
+}
+
+unsigned long kernel_eflags;
+
+/*
+ * Copies of the original ist values from the tss are only accessed during
+ * debugging, no special alignment required.
+ */
+DEFINE_PER_CPU(struct orig_ist, orig_ist);
+
+/*
+ * cpu_init() initializes state that is per-CPU. Some data is already
+ * initialized (naturally) in the bootstrap process, such as the GDT
+ * and IDT. We reload them nevertheless, this function acts as a
+ * 'CPU state barrier', nothing should get across.
+ * A lot of state is already set up in PDA init.
+ */
+void __cpuinit cpu_init(void)
+{
+	int cpu = stack_smp_processor_id();
+	struct tss_struct *t = &per_cpu(init_tss, cpu);
+	struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu);
+	unsigned long v;
+	char *estacks = NULL;
+	struct task_struct *me;
+	int i;
+
+	/* CPU 0 is initialised in head64.c */
+	if (cpu != 0)
+		pda_init(cpu);
+	else
+		estacks = boot_exception_stacks;
+
+	me = current;
+
+	if (cpu_test_and_set(cpu, cpu_initialized))
+		panic("CPU#%d already initialized!\n", cpu);
+
+	printk(KERN_INFO "Initializing CPU#%d\n", cpu);
+
+	clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
+
+	/*
+	 * Initialize the per-CPU GDT with the boot GDT,
+	 * and set up the GDT descriptor:
+	 */
+
+	switch_to_new_gdt();
+	load_idt((const struct desc_ptr *)&idt_descr);
+
+	memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
+	syscall_init();
+
+	wrmsrl(MSR_FS_BASE, 0);
+	wrmsrl(MSR_KERNEL_GS_BASE, 0);
+	barrier();
+
+	check_efer();
+
+	/*
+	 * set up and load the per-CPU TSS
+	 */
+	for (v = 0; v < N_EXCEPTION_STACKS; v++) {
+		static const unsigned int order[N_EXCEPTION_STACKS] = {
+			[0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER,
+			[DEBUG_STACK - 1] = DEBUG_STACK_ORDER
+		};
+		if (cpu) {
+			estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]);
+			if (!estacks)
+				panic("Cannot allocate exception stack %ld %d\n",
+				      v, cpu);
+		}
+		estacks += PAGE_SIZE << order[v];
+		orig_ist->ist[v] = t->x86_tss.ist[v] = (unsigned long)estacks;
+	}
+
+	t->x86_tss.io_bitmap_base = offsetof(struct tss_struct, io_bitmap);
+	/*
+	 * <= is required because the CPU will access up to
+	 * 8 bits beyond the end of the IO permission bitmap.
+	 */
+	for (i = 0; i <= IO_BITMAP_LONGS; i++)
+		t->io_bitmap[i] = ~0UL;
+
+	atomic_inc(&init_mm.mm_count);
+	me->active_mm = &init_mm;
+	if (me->mm)
+		BUG();
+	enter_lazy_tlb(&init_mm, me);
+
+	load_sp0(t, &current->thread);
+	set_tss_desc(cpu, t);
+	load_TR_desc();
+	load_LDT(&init_mm.context);
+
+#ifdef CONFIG_KGDB
+	/*
+	 * If the kgdb is connected no debug regs should be altered.  This
+	 * is only applicable when KGDB and a KGDB I/O module are built
+	 * into the kernel and you are using early debugging with
+	 * kgdbwait. KGDB will control the kernel HW breakpoint registers.
+	 */
+	if (kgdb_connected && arch_kgdb_ops.correct_hw_break)
+		arch_kgdb_ops.correct_hw_break();
+	else {
+#endif
+	/*
+	 * Clear all 6 debug registers:
+	 */
+
+	set_debugreg(0UL, 0);
+	set_debugreg(0UL, 1);
+	set_debugreg(0UL, 2);
+	set_debugreg(0UL, 3);
+	set_debugreg(0UL, 6);
+	set_debugreg(0UL, 7);
+#ifdef CONFIG_KGDB
+	/* If the kgdb is connected no debug regs should be altered. */
+	}
+#endif
+
+	fpu_init();
+
+	raw_local_save_flags(kernel_eflags);
+
+	if (is_uv_system())
+		uv_cpu_init();
+}

[PATCH] x86: remove two duplicated func in setup_32.c

[Yinghai Lu]

early_cpu_init is declcared in processor.h
memory_setup is defined in e820.c

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

--- a/arch/x86/kernel/setup_32.c	2008-06-21 17:17:02.000000000 -0700
+++ b/arch/x86/kernel/setup_32.c	2008-06-21 17:17:51.000000000 -0700
@@ -207,7 +207,6 @@ struct ist_info ist_info;
 EXPORT_SYMBOL(ist_info);
 #endif
 
-extern void early_cpu_init(void);
 extern int root_mountflags;
 
 unsigned long saved_video_mode;
@@ -607,12 +606,6 @@ static void set_mca_bus(int x)
 static void set_mca_bus(int x) { }
 #endif
 
-/* Overridden in paravirt.c if CONFIG_PARAVIRT */
-char * __init __attribute__((weak)) memory_setup(void)
-{
-	return machine_specific_memory_setup();
-}
-
 /*
  * Determine if we were loaded by an EFI loader.  If so, then we have also been
  * passed the efi memmap, systab, etc., so we should use these data structures

[PATCH] x86: move reserve_standard_io_resource to setup.c

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/setup.c    |   33 ++++++++++++++++++++++++++
 arch/x86/kernel/setup_32.c |   57 ---------------------------------------------
 arch/x86/kernel/setup_64.c |   29 ----------------------
 include/asm-x86/setup.h    |    2 +
 4 files changed, 37 insertions(+), 84 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -110,29 +110,6 @@ extern int root_mountflags;
 
 static char __initdata command_line[COMMAND_LINE_SIZE];
 
-static struct resource standard_io_resources[] = {
-	{ .name = "dma1", .start = 0x00, .end = 0x1f,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "pic1", .start = 0x20, .end = 0x21,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "timer0", .start = 0x40, .end = 0x43,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "timer1", .start = 0x50, .end = 0x53,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "keyboard", .start = 0x60, .end = 0x60,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "keyboard", .start = 0x64, .end = 0x64,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "dma page reg", .start = 0x80, .end = 0x8f,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "pic2", .start = 0xa0, .end = 0xa1,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "dma2", .start = 0xc0, .end = 0xdf,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
-	{ .name = "fpu", .start = 0xf0, .end = 0xff,
-		.flags = IORESOURCE_BUSY | IORESOURCE_IO }
-};
-
 #define IORESOURCE_RAM (IORESOURCE_BUSY | IORESOURCE_MEM)
 
 static struct resource data_resource = {
@@ -220,8 +197,6 @@ static inline void copy_edd(void)
  */
 void __init setup_arch(char **cmdline_p)
 {
-	unsigned i;
-
 	printk(KERN_INFO "Command line: %s\n", boot_command_line);
 
 	ROOT_DEV = old_decode_dev(boot_params.hdr.root_dev);
@@ -430,9 +405,7 @@ void __init setup_arch(char **cmdline_p)
 	e820_reserve_resources();
 	e820_mark_nosave_regions(end_pfn);
 
-	/* request I/O space for devices used on all i[345]86 PCs */
-	for (i = 0; i < ARRAY_SIZE(standard_io_resources); i++)
-		request_resource(&ioport_resource, &standard_io_resources[i]);
+	reserve_standard_io_resources();
 
 	e820_setup_gap();
 
Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -109,58 +109,6 @@ static struct resource video_ram_resourc
 	.flags	= IORESOURCE_BUSY | IORESOURCE_MEM
 };
 
-static struct resource standard_io_resources[] = { {
-	.name	= "dma1",
-	.start	= 0x0000,
-	.end	= 0x001f,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name	= "pic1",
-	.start	= 0x0020,
-	.end	= 0x0021,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name   = "timer0",
-	.start	= 0x0040,
-	.end    = 0x0043,
-	.flags  = IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name   = "timer1",
-	.start  = 0x0050,
-	.end    = 0x0053,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name	= "keyboard",
-	.start	= 0x0060,
-	.end	= 0x0060,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name	= "keyboard",
-	.start	= 0x0064,
-	.end	= 0x0064,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name	= "dma page reg",
-	.start	= 0x0080,
-	.end	= 0x008f,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name	= "pic2",
-	.start	= 0x00a0,
-	.end	= 0x00a1,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name	= "dma2",
-	.start	= 0x00c0,
-	.end	= 0x00df,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-}, {
-	.name	= "fpu",
-	.start	= 0x00f0,
-	.end	= 0x00ff,
-	.flags	= IORESOURCE_BUSY | IORESOURCE_IO
-} };
-
 /* cpu data as detected by the assembly code in head.S */
 struct cpuinfo_x86 new_cpu_data __cpuinitdata = { 0, 0, 0, 0, -1, 1, 0, 0, -1 };
 /* common cpu data for all cpus */
@@ -615,7 +563,6 @@ static void set_mca_bus(int x) { }
  */
 void __init setup_arch(char **cmdline_p)
 {
-	int i;
 	unsigned long max_low_pfn;
 
 	memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data));
@@ -830,9 +777,7 @@ void __init setup_arch(char **cmdline_p)
 	e820_mark_nosave_regions(max_low_pfn);
 
 	request_resource(&iomem_resource, &video_ram_resource);
-	/* request I/O space for devices used on all i[345]86 PCs */
-	for (i = 0; i < ARRAY_SIZE(standard_io_resources); i++)
-		request_resource(&ioport_resource, &standard_io_resources[i]);
+	reserve_standard_io_resources();
 
 	e820_setup_gap();
 
Index: linux-2.6/arch/x86/kernel/setup.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup.c
+++ linux-2.6/arch/x86/kernel/setup.c
@@ -467,3 +467,36 @@ void __init reserve_crashkernel(void)
 void __init reserve_crashkernel(void)
 {}
 #endif
+
+static struct resource standard_io_resources[] = {
+	{ .name = "dma1", .start = 0x00, .end = 0x1f,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "pic1", .start = 0x20, .end = 0x21,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "timer0", .start = 0x40, .end = 0x43,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "timer1", .start = 0x50, .end = 0x53,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "keyboard", .start = 0x60, .end = 0x60,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "keyboard", .start = 0x64, .end = 0x64,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "dma page reg", .start = 0x80, .end = 0x8f,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "pic2", .start = 0xa0, .end = 0xa1,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "dma2", .start = 0xc0, .end = 0xdf,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
+	{ .name = "fpu", .start = 0xf0, .end = 0xff,
+		.flags = IORESOURCE_BUSY | IORESOURCE_IO }
+};
+
+void __init reserve_standard_io_resources(void)
+{
+	int i;
+
+	/* request I/O space for devices used on all i[345]86 PCs */
+	for (i = 0; i < ARRAY_SIZE(standard_io_resources); i++)
+		request_resource(&ioport_resource, &standard_io_resources[i]);
+
+}
Index: linux-2.6/include/asm-x86/setup.h
===================================================================
--- linux-2.6.orig/include/asm-x86/setup.h
+++ linux-2.6/include/asm-x86/setup.h
@@ -38,6 +38,8 @@ void reserve_crashkernel(void);
 #ifndef __ASSEMBLY__
 #include <asm/bootparam.h>
 
+void reserve_standard_io_resources(void);
+
 #ifndef _SETUP
 
 /*

[PATCH] x86: move elfcorehdr parsing to setup.c

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/setup.c    |   19 +++++++++++++++++++
 arch/x86/kernel/setup_32.c |   16 ----------------
 arch/x86/kernel/setup_64.c |   17 -----------------
 3 files changed, 19 insertions(+), 33 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -26,7 +26,6 @@
 #include <asm/processor.h>
 #include <linux/console.h>
 #include <linux/seq_file.h>
-#include <linux/crash_dump.h>
 #include <linux/root_dev.h>
 #include <linux/pci.h>
 #include <asm/pci-direct.h>
@@ -131,22 +130,6 @@ static struct resource bss_resource = {
 	.flags = IORESOURCE_RAM,
 };
 
-#ifdef CONFIG_PROC_VMCORE
-/* elfcorehdr= specifies the location of elf core header
- * stored by the crashed kernel. This option will be passed
- * by kexec loader to the capture kernel.
- */
-static int __init setup_elfcorehdr(char *arg)
-{
-	char *end;
-	if (!arg)
-		return -EINVAL;
-	elfcorehdr_addr = memparse(arg, &end);
-	return end > arg ? 0 : -EINVAL;
-}
-early_param("elfcorehdr", setup_elfcorehdr);
-#endif
-
 #ifndef CONFIG_NUMA
 static void __init
 contig_initmem_init(unsigned long start_pfn, unsigned long end_pfn)
Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -42,7 +42,6 @@
 #include <linux/iscsi_ibft.h>
 #include <linux/nodemask.h>
 #include <linux/kexec.h>
-#include <linux/crash_dump.h>
 #include <linux/dmi.h>
 #include <linux/pfn.h>
 #include <linux/pci.h>
@@ -195,21 +194,6 @@ static inline void copy_edd(void)
 }
 #endif
 
-#ifdef CONFIG_PROC_VMCORE
-/* elfcorehdr= specifies the location of elf core header
- * stored by the crashed kernel.
- */
-static int __init parse_elfcorehdr(char *arg)
-{
-	if (!arg)
-		return -EINVAL;
-
-	elfcorehdr_addr = memparse(arg, &arg);
-	return 0;
-}
-early_param("elfcorehdr", parse_elfcorehdr);
-#endif /* CONFIG_PROC_VMCORE */
-
 /*
  * highmem=size forces highmem to be exactly 'size' bytes.
  * This works even on boxes that have no highmem otherwise.
Index: linux-2.6/arch/x86/kernel/setup.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup.c
+++ linux-2.6/arch/x86/kernel/setup.c
@@ -4,6 +4,7 @@
 #include <linux/bootmem.h>
 #include <linux/percpu.h>
 #include <linux/kexec.h>
+#include <linux/crash_dump.h>
 #include <asm/smp.h>
 #include <asm/percpu.h>
 #include <asm/sections.h>
@@ -500,3 +501,21 @@ void __init reserve_standard_io_resource
 		request_resource(&ioport_resource, &standard_io_resources[i]);
 
 }
+
+#ifdef CONFIG_PROC_VMCORE
+/* elfcorehdr= specifies the location of elf core header
+ * stored by the crashed kernel. This option will be passed
+ * by kexec loader to the capture kernel.
+ */
+static int __init setup_elfcorehdr(char *arg)
+{
+	char *end;
+	if (!arg)
+		return -EINVAL;
+	elfcorehdr_addr = memparse(arg, &end);
+	return end > arg ? 0 : -EINVAL;
+}
+early_param("elfcorehdr", setup_elfcorehdr);
+#endif
+
+

[PATCH] x86: introduce initmem_init for 64 bit

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/setup_64.c |   25 +------------------------
 arch/x86/mm/init_64.c      |   16 ++++++++++++++++
 arch/x86/mm/numa_64.c      |    2 +-
 include/asm-x86/numa_64.h  |    1 -
 include/asm-x86/page_64.h  |    1 +
 5 files changed, 19 insertions(+), 26 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -130,25 +130,6 @@ static struct resource bss_resource = {
 	.flags = IORESOURCE_RAM,
 };
 
-#ifndef CONFIG_NUMA
-static void __init
-contig_initmem_init(unsigned long start_pfn, unsigned long end_pfn)
-{
-	unsigned long bootmap_size, bootmap;
-
-	bootmap_size = bootmem_bootmap_pages(end_pfn)<<PAGE_SHIFT;
-	bootmap = find_e820_area(0, end_pfn<<PAGE_SHIFT, bootmap_size,
-				 PAGE_SIZE);
-	if (bootmap == -1L)
-		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
-	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, end_pfn);
-	e820_register_active_regions(0, start_pfn, end_pfn);
-	free_bootmem_with_active_regions(0, end_pfn);
-	early_res_to_bootmem(0, end_pfn<<PAGE_SHIFT);
-	reserve_bootmem(bootmap, bootmap_size, BOOTMEM_DEFAULT);
-}
-#endif
-
 #if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
 struct edd edd;
 #ifdef CONFIG_EDD_MODULE
@@ -309,11 +290,7 @@ void __init setup_arch(char **cmdline_p)
 	acpi_numa_init();
 #endif
 
-#ifdef CONFIG_NUMA
-	numa_initmem_init(0, end_pfn);
-#else
-	contig_initmem_init(0, end_pfn);
-#endif
+	initmem_init(0, end_pfn);
 
 	dma32_reserve_bootmem();
 
Index: linux-2.6/arch/x86/mm/init_64.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/init_64.c
+++ linux-2.6/arch/x86/mm/init_64.c
@@ -624,6 +624,22 @@ unsigned long __init_refok init_memory_m
 }
 
 #ifndef CONFIG_NUMA
+void __init initmem_init(unsigned long start_pfn, unsigned long end_pfn)
+{
+	unsigned long bootmap_size, bootmap;
+
+	bootmap_size = bootmem_bootmap_pages(end_pfn)<<PAGE_SHIFT;
+	bootmap = find_e820_area(0, end_pfn<<PAGE_SHIFT, bootmap_size,
+				 PAGE_SIZE);
+	if (bootmap == -1L)
+		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
+	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, end_pfn);
+	e820_register_active_regions(0, start_pfn, end_pfn);
+	free_bootmem_with_active_regions(0, end_pfn);
+	early_res_to_bootmem(0, end_pfn<<PAGE_SHIFT);
+	reserve_bootmem(bootmap, bootmap_size, BOOTMEM_DEFAULT);
+}
+
 void __init paging_init(void)
 {
 	unsigned long max_zone_pfns[MAX_NR_ZONES];
Index: linux-2.6/arch/x86/mm/numa_64.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/numa_64.c
+++ linux-2.6/arch/x86/mm/numa_64.c
@@ -514,7 +514,7 @@ out:
 }
 #endif /* CONFIG_NUMA_EMU */
 
-void __init numa_initmem_init(unsigned long start_pfn, unsigned long last_pfn)
+void __init initmem_init(unsigned long start_pfn, unsigned long last_pfn)
 {
 	int i;
 
Index: linux-2.6/include/asm-x86/numa_64.h
===================================================================
--- linux-2.6.orig/include/asm-x86/numa_64.h
+++ linux-2.6/include/asm-x86/numa_64.h
@@ -22,7 +22,6 @@ extern int hotadd_percent;
 
 extern s16 apicid_to_node[MAX_LOCAL_APIC];
 
-extern void numa_initmem_init(unsigned long start_pfn, unsigned long end_pfn);
 extern unsigned long numa_free_all_bootmem(void);
 extern void setup_node_bootmem(int nodeid, unsigned long start,
 			       unsigned long end);
Index: linux-2.6/include/asm-x86/page_64.h
===================================================================
--- linux-2.6.orig/include/asm-x86/page_64.h
+++ linux-2.6/include/asm-x86/page_64.h
@@ -83,6 +83,7 @@ typedef struct { pteval_t pte; } pte_t;
 extern unsigned long init_memory_mapping(unsigned long start,
 					 unsigned long end);
 
+extern void initmem_init(unsigned long start_pfn, unsigned long end_pfn);
 #endif	/* !__ASSEMBLY__ */
 
 #ifdef CONFIG_FLATMEM

[PATCH] x86: introduce initmem_init for 32 bit

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
---
 arch/x86/kernel/setup_32.c |   94 ---------------------------------------------
 arch/x86/mm/discontig_32.c |    4 -
 arch/x86/mm/init_32.c      |   90 +++++++++++++++++++++++++++++++++++++++++++
 include/asm-x86/page_32.h  |    5 ++
 4 files changed, 99 insertions(+), 94 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -301,71 +301,11 @@ unsigned long __init find_max_low_pfn(vo
 	return max_low_pfn;
 }
 
-#ifndef CONFIG_NEED_MULTIPLE_NODES
-static void __init setup_bootmem_allocator(void);
-static unsigned long __init setup_memory(void)
-{
-	/*
-	 * partially used pages are not usable - thus
-	 * we are rounding upwards:
-	 */
-	min_low_pfn = PFN_UP(init_pg_tables_end);
-
-	max_low_pfn = find_max_low_pfn();
-
-#ifdef CONFIG_HIGHMEM
-	highstart_pfn = highend_pfn = max_pfn;
-	if (max_pfn > max_low_pfn) {
-		highstart_pfn = max_low_pfn;
-	}
-	memory_present(0, 0, highend_pfn);
-	printk(KERN_NOTICE "%ldMB HIGHMEM available.\n",
-		pages_to_mb(highend_pfn - highstart_pfn));
-	num_physpages = highend_pfn;
-	high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1;
-#else
-	memory_present(0, 0, max_low_pfn);
-	num_physpages = max_low_pfn;
-	high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1;
-#endif
-#ifdef CONFIG_FLATMEM
-	max_mapnr = num_physpages;
-#endif
-	printk(KERN_NOTICE "%ldMB LOWMEM available.\n",
-			pages_to_mb(max_low_pfn));
-
-	setup_bootmem_allocator();
-
-	return max_low_pfn;
-}
-
-static void __init zone_sizes_init(void)
-{
-	unsigned long max_zone_pfns[MAX_NR_ZONES];
-	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
-	max_zone_pfns[ZONE_DMA] =
-		virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT;
-	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
-	remove_all_active_ranges();
-#ifdef CONFIG_HIGHMEM
-	max_zone_pfns[ZONE_HIGHMEM] = highend_pfn;
-	e820_register_active_regions(0, 0, highend_pfn);
-#else
-	e820_register_active_regions(0, 0, max_low_pfn);
-#endif
-
-	free_area_init_nodes(max_zone_pfns);
-}
-#else
-extern unsigned long __init setup_memory(void);
-extern void zone_sizes_init(void);
-#endif /* !CONFIG_NEED_MULTIPLE_NODES */
-
 #ifdef CONFIG_BLK_DEV_INITRD
 
 static bool do_relocate_initrd = false;
 
-static void __init reserve_initrd(void)
+void __init reserve_initrd(void)
 {
 	u64 ramdisk_image = boot_params.hdr.ramdisk_image;
 	u64 ramdisk_size  = boot_params.hdr.ramdisk_size;
@@ -481,36 +421,6 @@ static void __init relocate_initrd(void)
 
 #endif /* CONFIG_BLK_DEV_INITRD */
 
-void __init setup_bootmem_allocator(void)
-{
-	int i;
-	unsigned long bootmap_size, bootmap;
-	/*
-	 * Initialize the boot-time allocator (with low memory only):
-	 */
-	bootmap_size = bootmem_bootmap_pages(max_low_pfn)<<PAGE_SHIFT;
-	bootmap = find_e820_area(min_low_pfn<<PAGE_SHIFT,
-				 max_pfn_mapped<<PAGE_SHIFT, bootmap_size,
-				 PAGE_SIZE);
-	if (bootmap == -1L)
-		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
-	reserve_early(bootmap, bootmap + bootmap_size, "BOOTMAP");
-#ifdef CONFIG_BLK_DEV_INITRD
-	reserve_initrd();
-#endif
-	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, max_low_pfn);
-	printk(KERN_INFO "  mapped low ram: 0 - %08lx\n",
-		 max_pfn_mapped<<PAGE_SHIFT);
-	printk(KERN_INFO "  low ram: %08lx - %08lx\n",
-		 min_low_pfn<<PAGE_SHIFT, max_low_pfn<<PAGE_SHIFT);
-	printk(KERN_INFO "  bootmap %08lx - %08lx\n",
-		 bootmap, bootmap + bootmap_size);
-	for_each_online_node(i)
-		free_bootmem_with_active_regions(i, max_low_pfn);
-	early_res_to_bootmem(0, max_low_pfn<<PAGE_SHIFT);
-
-}
-
 /*
  * The node 0 pgdat is initialized before all of these because
  * it's needed for bootmem.  node>0 pgdats have their virtual
@@ -672,7 +582,7 @@ void __init setup_arch(char **cmdline_p)
         acpi_numa_init();
 #endif
 
-	max_low_pfn = setup_memory();
+	max_low_pfn = initmem_init(0, max_pfn);
 
 #ifdef CONFIG_ACPI_SLEEP
 	/*
Index: linux-2.6/arch/x86/mm/discontig_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/discontig_32.c
+++ linux-2.6/arch/x86/mm/discontig_32.c
@@ -309,8 +309,8 @@ static void init_remap_allocator(int nid
 		(ulong) node_remap_end_vaddr[nid]);
 }
 
-extern void setup_bootmem_allocator(void);
-unsigned long __init setup_memory(void)
+unsigned long __init initmem_init(unsigned long start_pfn,
+				  unsigned long end_pfn)
 {
 	int nid;
 	unsigned long system_start_pfn, system_max_low_pfn;
Index: linux-2.6/arch/x86/mm/init_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/init_32.c
+++ linux-2.6/arch/x86/mm/init_32.c
@@ -540,6 +540,96 @@ static void __init set_nx(void)
 }
 #endif
 
+#ifndef CONFIG_NEED_MULTIPLE_NODES
+extern unsigned long find_max_low_pfn(void);
+unsigned long __init initmem_init(unsigned long start_pfn,
+				  unsigned long end_pfn)
+{
+	/*
+	 * partially used pages are not usable - thus
+	 * we are rounding upwards:
+	 */
+	min_low_pfn = PFN_UP(init_pg_tables_end);
+
+	max_low_pfn = find_max_low_pfn();
+
+#ifdef CONFIG_HIGHMEM
+	highstart_pfn = highend_pfn = max_pfn;
+	if (max_pfn > max_low_pfn)
+		highstart_pfn = max_low_pfn;
+	memory_present(0, 0, highend_pfn);
+	printk(KERN_NOTICE "%ldMB HIGHMEM available.\n",
+		pages_to_mb(highend_pfn - highstart_pfn));
+	num_physpages = highend_pfn;
+	high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1;
+#else
+	memory_present(0, 0, max_low_pfn);
+	num_physpages = max_low_pfn;
+	high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1;
+#endif
+#ifdef CONFIG_FLATMEM
+	max_mapnr = num_physpages;
+#endif
+	printk(KERN_NOTICE "%ldMB LOWMEM available.\n",
+			pages_to_mb(max_low_pfn));
+
+	setup_bootmem_allocator();
+
+	return max_low_pfn;
+}
+
+void __init zone_sizes_init(void)
+{
+	unsigned long max_zone_pfns[MAX_NR_ZONES];
+	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
+	max_zone_pfns[ZONE_DMA] =
+		virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT;
+	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
+	remove_all_active_ranges();
+#ifdef CONFIG_HIGHMEM
+	max_zone_pfns[ZONE_HIGHMEM] = highend_pfn;
+	e820_register_active_regions(0, 0, highend_pfn);
+#else
+	e820_register_active_regions(0, 0, max_low_pfn);
+#endif
+
+	free_area_init_nodes(max_zone_pfns);
+}
+#endif /* !CONFIG_NEED_MULTIPLE_NODES */
+
+extern void reserve_initrd(void);
+
+void __init setup_bootmem_allocator(void)
+{
+	int i;
+	unsigned long bootmap_size, bootmap;
+	/*
+	 * Initialize the boot-time allocator (with low memory only):
+	 */
+	bootmap_size = bootmem_bootmap_pages(max_low_pfn)<<PAGE_SHIFT;
+	bootmap = find_e820_area(min_low_pfn<<PAGE_SHIFT,
+				 max_pfn_mapped<<PAGE_SHIFT, bootmap_size,
+				 PAGE_SIZE);
+	if (bootmap == -1L)
+		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
+	reserve_early(bootmap, bootmap + bootmap_size, "BOOTMAP");
+#ifdef CONFIG_BLK_DEV_INITRD
+	reserve_initrd();
+#endif
+	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, max_low_pfn);
+	printk(KERN_INFO "  mapped low ram: 0 - %08lx\n",
+		 max_pfn_mapped<<PAGE_SHIFT);
+	printk(KERN_INFO "  low ram: %08lx - %08lx\n",
+		 min_low_pfn<<PAGE_SHIFT, max_low_pfn<<PAGE_SHIFT);
+	printk(KERN_INFO "  bootmap %08lx - %08lx\n",
+		 bootmap, bootmap + bootmap_size);
+	for_each_online_node(i)
+		free_bootmem_with_active_regions(i, max_low_pfn);
+	early_res_to_bootmem(0, max_low_pfn<<PAGE_SHIFT);
+
+}
+
+
 /*
  * paging_init() sets up the page tables - note that the first 8MB are
  * already mapped by head.S.
Index: linux-2.6/include/asm-x86/page_32.h
===================================================================
--- linux-2.6.orig/include/asm-x86/page_32.h
+++ linux-2.6/include/asm-x86/page_32.h
@@ -97,6 +97,11 @@ extern int sysctl_legacy_va_layout;
 #define VMALLOC_RESERVE		((unsigned long)__VMALLOC_RESERVE)
 #define MAXMEM			(-__PAGE_OFFSET - __VMALLOC_RESERVE)
 
+extern unsigned long initmem_init(unsigned long, unsigned long);
+extern void zone_sizes_init(void);
+extern void setup_bootmem_allocator(void);
+
+
 #ifdef CONFIG_X86_USE_3DNOW
 #include <asm/mmx.h>
 

[PATCH] x86: introduce reserve_initrd

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/setup_32.c |   16 ++++++++-----
 arch/x86/kernel/setup_64.c |   53 ++++++++++++++++++++++++---------------------
 arch/x86/mm/init_32.c      |    6 +----
 include/asm-x86/setup.h    |    2 +
 4 files changed, 43 insertions(+), 34 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -337,7 +337,7 @@ void __init reserve_initrd(void)
 		 * in i386_start_kernel
 		 */
 		initrd_start = ramdisk_image + PAGE_OFFSET;
-		initrd_end = initrd_start+ramdisk_size;
+		initrd_end = initrd_start + ramdisk_size;
 		return;
 	}
 
@@ -364,7 +364,7 @@ void __init reserve_initrd(void)
 
 #define MAX_MAP_CHUNK	(NR_FIX_BTMAPS << PAGE_SHIFT)
 
-static void __init relocate_initrd(void)
+static void __init post_reserve_initrd(void)
 {
 	u64 ramdisk_image = boot_params.hdr.ramdisk_image;
 	u64 ramdisk_size  = boot_params.hdr.ramdisk_size;
@@ -418,7 +418,13 @@ static void __init relocate_initrd(void)
 	/* need to free that, otherwise init highmem will reserve it again */
 	free_early(ramdisk_image, ramdisk_image+ramdisk_size);
 }
-
+#else
+void __init reserve_initrd(void)
+{
+}
+static void __init post_reserve_initrd(void)
+{
+}
 #endif /* CONFIG_BLK_DEV_INITRD */
 
 /*
@@ -638,9 +644,7 @@ void __init setup_arch(char **cmdline_p)
 	 * NOTE: at this point the bootmem allocator is fully available.
 	 */
 
-#ifdef CONFIG_BLK_DEV_INITRD
-	relocate_initrd();
-#endif
+	post_reserve_initrd();
 
 	remapped_pgdat_init();
 	sparse_init();
Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -154,6 +154,33 @@ static inline void copy_edd(void)
 }
 #endif
 
+void __init reserve_initrd(void)
+{
+#ifdef CONFIG_BLK_DEV_INITRD
+	if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) {
+		unsigned long ramdisk_image = boot_params.hdr.ramdisk_image;
+		unsigned long ramdisk_size  = boot_params.hdr.ramdisk_size;
+		unsigned long ramdisk_end   = ramdisk_image + ramdisk_size;
+		unsigned long end_of_mem    = end_pfn << PAGE_SHIFT;
+
+		if (ramdisk_end <= end_of_mem) {
+			/*
+			 * don't need to reserve again, already reserved early
+			 * in x86_64_start_kernel, and early_res_to_bootmem
+			 * will convert that to reserved in bootmem
+			 */
+			initrd_start = ramdisk_image + PAGE_OFFSET;
+			initrd_end = initrd_start+ramdisk_size;
+		} else {
+			free_early(ramdisk_image, ramdisk_end);
+			printk(KERN_ERR "initrd extends beyond end of memory "
+			       "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
+			       ramdisk_end, end_of_mem);
+			initrd_start = 0;
+		}
+	}
+#endif
+}
 /*
  * setup_arch - architecture-specific boot-time initializations
  *
@@ -251,6 +278,8 @@ void __init setup_arch(char **cmdline_p)
 		end_pfn = e820_end_of_ram();
 	}
 
+	reserve_initrd();
+
 	num_physpages = end_pfn;
 
 	check_efer();
@@ -307,30 +336,6 @@ void __init setup_arch(char **cmdline_p)
 	*/
 	find_smp_config();
 #endif
-#ifdef CONFIG_BLK_DEV_INITRD
-	if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) {
-		unsigned long ramdisk_image = boot_params.hdr.ramdisk_image;
-		unsigned long ramdisk_size  = boot_params.hdr.ramdisk_size;
-		unsigned long ramdisk_end   = ramdisk_image + ramdisk_size;
-		unsigned long end_of_mem    = end_pfn << PAGE_SHIFT;
-
-		if (ramdisk_end <= end_of_mem) {
-			/*
-			 * don't need to reserve again, already reserved early
-			 * in x86_64_start_kernel, and early_res_to_bootmem
-			 * convert that to reserved in bootmem
-			 */
-			initrd_start = ramdisk_image + PAGE_OFFSET;
-			initrd_end = initrd_start+ramdisk_size;
-		} else {
-			free_bootmem(ramdisk_image, ramdisk_size);
-			printk(KERN_ERR "initrd extends beyond end of memory "
-			       "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
-			       ramdisk_end, end_of_mem);
-			initrd_start = 0;
-		}
-	}
-#endif
 	reserve_crashkernel();
 
 	reserve_ibft_region();
Index: linux-2.6/arch/x86/mm/init_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/init_32.c
+++ linux-2.6/arch/x86/mm/init_32.c
@@ -597,8 +597,6 @@ void __init zone_sizes_init(void)
 }
 #endif /* !CONFIG_NEED_MULTIPLE_NODES */
 
-extern void reserve_initrd(void);
-
 void __init setup_bootmem_allocator(void)
 {
 	int i;
@@ -613,9 +611,9 @@ void __init setup_bootmem_allocator(void
 	if (bootmap == -1L)
 		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
 	reserve_early(bootmap, bootmap + bootmap_size, "BOOTMAP");
-#ifdef CONFIG_BLK_DEV_INITRD
+
 	reserve_initrd();
-#endif
+
 	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, max_low_pfn);
 	printk(KERN_INFO "  mapped low ram: 0 - %08lx\n",
 		 max_pfn_mapped<<PAGE_SHIFT);
Index: linux-2.6/include/asm-x86/setup.h
===================================================================
--- linux-2.6.orig/include/asm-x86/setup.h
+++ linux-2.6/include/asm-x86/setup.h
@@ -39,6 +39,8 @@ void reserve_crashkernel(void);
 #include <asm/bootparam.h>
 
 void reserve_standard_io_resources(void);
+void reserve_initrd(void);
+
 
 #ifndef _SETUP
 

[PATCH] x86: move boot_params declaring to setup.c

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/cpu/common_64.c |    6 ------
 arch/x86/kernel/setup.c         |    6 ++++++
 arch/x86/kernel/setup_32.c      |    6 ------
 3 files changed, 6 insertions(+), 12 deletions(-)

Index: linux-2.6/arch/x86/kernel/cpu/common_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/common_64.c
+++ linux-2.6/arch/x86/kernel/cpu/common_64.c
@@ -421,12 +421,6 @@ static __init int setup_disablecpuid(cha
 }
 __setup("clearcpuid=", setup_disablecpuid);
 
-#ifndef CONFIG_DEBUG_BOOT_PARAMS
-struct boot_params __initdata boot_params;
-#else
-struct boot_params boot_params;
-#endif
-
 cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
 
 struct x8664_pda **_cpu_pda __read_mostly;
Index: linux-2.6/arch/x86/kernel/setup.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup.c
+++ linux-2.6/arch/x86/kernel/setup.c
@@ -15,6 +15,12 @@
 #include <asm/apicdef.h>
 #include <asm/highmem.h>
 
+#ifndef CONFIG_DEBUG_BOOT_PARAMS
+struct boot_params __initdata boot_params;
+#else
+struct boot_params boot_params;
+#endif
+
 #ifdef CONFIG_X86_LOCAL_APIC
 unsigned int num_processors;
 unsigned disabled_cpus __cpuinitdata;
Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -161,12 +161,6 @@ unsigned long saved_video_mode;
 
 static char __initdata command_line[COMMAND_LINE_SIZE];
 
-#ifndef CONFIG_DEBUG_BOOT_PARAMS
-struct boot_params __initdata boot_params;
-#else
-struct boot_params boot_params;
-#endif
-
 #if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
 struct edd edd;
 #ifdef CONFIG_EDD_MODULE

[PATCH] x86: move reservetop and vmalloc parsing to pgtable_32.c

[Yinghai Lu]

also change reserve_top_address to __init attibute

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/setup_32.c |   33 ---------------------------------
 arch/x86/mm/pgtable_32.c   |   35 ++++++++++++++++++++++++++++++++++-
 2 files changed, 34 insertions(+), 34 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -185,39 +185,6 @@ static inline void copy_edd(void)
 }
 #endif
 
-/*
- * vmalloc=size forces the vmalloc area to be exactly 'size'
- * bytes. This can be used to increase (or decrease) the
- * vmalloc area - the default is 128m.
- */
-static int __init parse_vmalloc(char *arg)
-{
-	if (!arg)
-		return -EINVAL;
-
-	__VMALLOC_RESERVE = memparse(arg, &arg);
-	return 0;
-}
-early_param("vmalloc", parse_vmalloc);
-
-/*
- * reservetop=size reserves a hole at the top of the kernel address space which
- * a hypervisor can load into later.  Needed for dynamically loaded hypervisors,
- * so relocating the fixmap can be done before paging initialization.
- */
-static int __init parse_reservetop(char *arg)
-{
-	unsigned long address;
-
-	if (!arg)
-		return -EINVAL;
-
-	address = memparse(arg, &arg);
-	reserve_top_address(address);
-	return 0;
-}
-early_param("reservetop", parse_reservetop);
-
 #ifdef CONFIG_BLK_DEV_INITRD
 
 static bool do_relocate_initrd = false;
Index: linux-2.6/arch/x86/mm/pgtable_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/pgtable_32.c
+++ linux-2.6/arch/x86/mm/pgtable_32.c
@@ -152,7 +152,7 @@ EXPORT_SYMBOL(__FIXADDR_TOP);
  * Can be used to relocate the fixmap area and poke a hole in the top
  * of kernel address space to make room for a hypervisor.
  */
-void reserve_top_address(unsigned long reserve)
+void __init reserve_top_address(unsigned long reserve)
 {
 	BUG_ON(fixmaps_set > 0);
 	printk(KERN_INFO "Reserving virtual address space above 0x%08x\n",
@@ -160,3 +160,36 @@ void reserve_top_address(unsigned long r
 	__FIXADDR_TOP = -reserve - PAGE_SIZE;
 	__VMALLOC_RESERVE += reserve;
 }
+
+/*
+ * vmalloc=size forces the vmalloc area to be exactly 'size'
+ * bytes. This can be used to increase (or decrease) the
+ * vmalloc area - the default is 128m.
+ */
+static int __init parse_vmalloc(char *arg)
+{
+	if (!arg)
+		return -EINVAL;
+
+	__VMALLOC_RESERVE = memparse(arg, &arg);
+	return 0;
+}
+early_param("vmalloc", parse_vmalloc);
+
+/*
+ * reservetop=size reserves a hole at the top of the kernel address space which
+ * a hypervisor can load into later.  Needed for dynamically loaded hypervisors,
+ * so relocating the fixmap can be done before paging initialization.
+ */
+static int __init parse_reservetop(char *arg)
+{
+	unsigned long address;
+
+	if (!arg)
+		return -EINVAL;
+
+	address = memparse(arg, &arg);
+	reserve_top_address(address);
+	return 0;
+}
+early_param("reservetop", parse_reservetop);

[PATCH] x86: introduce reserve_initrd

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/setup_32.c |   16 ++++++++-----
 arch/x86/kernel/setup_64.c |   53 ++++++++++++++++++++++++---------------------
 arch/x86/mm/init_32.c      |    6 +----
 include/asm-x86/setup.h    |    2 +
 4 files changed, 43 insertions(+), 34 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -337,7 +337,7 @@ void __init reserve_initrd(void)
 		 * in i386_start_kernel
 		 */
 		initrd_start = ramdisk_image + PAGE_OFFSET;
-		initrd_end = initrd_start+ramdisk_size;
+		initrd_end = initrd_start + ramdisk_size;
 		return;
 	}
 
@@ -364,7 +364,7 @@ void __init reserve_initrd(void)
 
 #define MAX_MAP_CHUNK	(NR_FIX_BTMAPS << PAGE_SHIFT)
 
-static void __init relocate_initrd(void)
+static void __init post_reserve_initrd(void)
 {
 	u64 ramdisk_image = boot_params.hdr.ramdisk_image;
 	u64 ramdisk_size  = boot_params.hdr.ramdisk_size;
@@ -418,7 +418,13 @@ static void __init relocate_initrd(void)
 	/* need to free that, otherwise init highmem will reserve it again */
 	free_early(ramdisk_image, ramdisk_image+ramdisk_size);
 }
-
+#else
+void __init reserve_initrd(void)
+{
+}
+static void __init post_reserve_initrd(void)
+{
+}
 #endif /* CONFIG_BLK_DEV_INITRD */
 
 /*
@@ -638,9 +644,7 @@ void __init setup_arch(char **cmdline_p)
 	 * NOTE: at this point the bootmem allocator is fully available.
 	 */
 
-#ifdef CONFIG_BLK_DEV_INITRD
-	relocate_initrd();
-#endif
+	post_reserve_initrd();
 
 	remapped_pgdat_init();
 	sparse_init();
Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -154,6 +154,33 @@ static inline void copy_edd(void)
 }
 #endif
 
+void __init reserve_initrd(void)
+{
+#ifdef CONFIG_BLK_DEV_INITRD
+	if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) {
+		unsigned long ramdisk_image = boot_params.hdr.ramdisk_image;
+		unsigned long ramdisk_size  = boot_params.hdr.ramdisk_size;
+		unsigned long ramdisk_end   = ramdisk_image + ramdisk_size;
+		unsigned long end_of_mem    = end_pfn << PAGE_SHIFT;
+
+		if (ramdisk_end <= end_of_mem) {
+			/*
+			 * don't need to reserve again, already reserved early
+			 * in x86_64_start_kernel, and early_res_to_bootmem
+			 * will convert that to reserved in bootmem
+			 */
+			initrd_start = ramdisk_image + PAGE_OFFSET;
+			initrd_end = initrd_start+ramdisk_size;
+		} else {
+			free_early(ramdisk_image, ramdisk_end);
+			printk(KERN_ERR "initrd extends beyond end of memory "
+			       "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
+			       ramdisk_end, end_of_mem);
+			initrd_start = 0;
+		}
+	}
+#endif
+}
 /*
  * setup_arch - architecture-specific boot-time initializations
  *
@@ -251,6 +278,8 @@ void __init setup_arch(char **cmdline_p)
 		end_pfn = e820_end_of_ram();
 	}
 
+	reserve_initrd();
+
 	num_physpages = end_pfn;
 
 	check_efer();
@@ -307,30 +336,6 @@ void __init setup_arch(char **cmdline_p)
 	*/
 	find_smp_config();
 #endif
-#ifdef CONFIG_BLK_DEV_INITRD
-	if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) {
-		unsigned long ramdisk_image = boot_params.hdr.ramdisk_image;
-		unsigned long ramdisk_size  = boot_params.hdr.ramdisk_size;
-		unsigned long ramdisk_end   = ramdisk_image + ramdisk_size;
-		unsigned long end_of_mem    = end_pfn << PAGE_SHIFT;
-
-		if (ramdisk_end <= end_of_mem) {
-			/*
-			 * don't need to reserve again, already reserved early
-			 * in x86_64_start_kernel, and early_res_to_bootmem
-			 * convert that to reserved in bootmem
-			 */
-			initrd_start = ramdisk_image + PAGE_OFFSET;
-			initrd_end = initrd_start+ramdisk_size;
-		} else {
-			free_bootmem(ramdisk_image, ramdisk_size);
-			printk(KERN_ERR "initrd extends beyond end of memory "
-			       "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
-			       ramdisk_end, end_of_mem);
-			initrd_start = 0;
-		}
-	}
-#endif
 	reserve_crashkernel();
 
 	reserve_ibft_region();
Index: linux-2.6/arch/x86/mm/init_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/init_32.c
+++ linux-2.6/arch/x86/mm/init_32.c
@@ -597,8 +597,6 @@ void __init zone_sizes_init(void)
 }
 #endif /* !CONFIG_NEED_MULTIPLE_NODES */
 
-extern void reserve_initrd(void);
-
 void __init setup_bootmem_allocator(void)
 {
 	int i;
@@ -613,9 +611,9 @@ void __init setup_bootmem_allocator(void
 	if (bootmap == -1L)
 		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
 	reserve_early(bootmap, bootmap + bootmap_size, "BOOTMAP");
-#ifdef CONFIG_BLK_DEV_INITRD
+
 	reserve_initrd();
-#endif
+
 	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, max_low_pfn);
 	printk(KERN_INFO "  mapped low ram: 0 - %08lx\n",
 		 max_pfn_mapped<<PAGE_SHIFT);
Index: linux-2.6/include/asm-x86/setup.h
===================================================================
--- linux-2.6.orig/include/asm-x86/setup.h
+++ linux-2.6/include/asm-x86/setup.h
@@ -39,6 +39,8 @@ void reserve_crashkernel(void);
 #include <asm/bootparam.h>
 
 void reserve_standard_io_resources(void);
+void reserve_initrd(void);
+
 
 #ifndef _SETUP
 

[PATCH] x86: move reservetop and vmalloc parsing to pgtable_32.c

[Yinghai Lu]

also change reserve_top_address to __init attibute

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/setup_32.c |   33 ---------------------------------
 arch/x86/mm/pgtable_32.c   |   35 ++++++++++++++++++++++++++++++++++-
 2 files changed, 34 insertions(+), 34 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -185,39 +185,6 @@ static inline void copy_edd(void)
 }
 #endif
 
-/*
- * vmalloc=size forces the vmalloc area to be exactly 'size'
- * bytes. This can be used to increase (or decrease) the
- * vmalloc area - the default is 128m.
- */
-static int __init parse_vmalloc(char *arg)
-{
-	if (!arg)
-		return -EINVAL;
-
-	__VMALLOC_RESERVE = memparse(arg, &arg);
-	return 0;
-}
-early_param("vmalloc", parse_vmalloc);
-
-/*
- * reservetop=size reserves a hole at the top of the kernel address space which
- * a hypervisor can load into later.  Needed for dynamically loaded hypervisors,
- * so relocating the fixmap can be done before paging initialization.
- */
-static int __init parse_reservetop(char *arg)
-{
-	unsigned long address;
-
-	if (!arg)
-		return -EINVAL;
-
-	address = memparse(arg, &arg);
-	reserve_top_address(address);
-	return 0;
-}
-early_param("reservetop", parse_reservetop);
-
 #ifdef CONFIG_BLK_DEV_INITRD
 
 static bool do_relocate_initrd = false;
Index: linux-2.6/arch/x86/mm/pgtable_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/pgtable_32.c
+++ linux-2.6/arch/x86/mm/pgtable_32.c
@@ -152,7 +152,7 @@ EXPORT_SYMBOL(__FIXADDR_TOP);
  * Can be used to relocate the fixmap area and poke a hole in the top
  * of kernel address space to make room for a hypervisor.
  */
-void reserve_top_address(unsigned long reserve)
+void __init reserve_top_address(unsigned long reserve)
 {
 	BUG_ON(fixmaps_set > 0);
 	printk(KERN_INFO "Reserving virtual address space above 0x%08x\n",
@@ -160,3 +160,36 @@ void reserve_top_address(unsigned long r
 	__FIXADDR_TOP = -reserve - PAGE_SIZE;
 	__VMALLOC_RESERVE += reserve;
 }
+
+/*
+ * vmalloc=size forces the vmalloc area to be exactly 'size'
+ * bytes. This can be used to increase (or decrease) the
+ * vmalloc area - the default is 128m.
+ */
+static int __init parse_vmalloc(char *arg)
+{
+	if (!arg)
+		return -EINVAL;
+
+	__VMALLOC_RESERVE = memparse(arg, &arg);
+	return 0;
+}
+early_param("vmalloc", parse_vmalloc);
+
+/*
+ * reservetop=size reserves a hole at the top of the kernel address space which
+ * a hypervisor can load into later.  Needed for dynamically loaded hypervisors,
+ * so relocating the fixmap can be done before paging initialization.
+ */
+static int __init parse_reservetop(char *arg)
+{
+	unsigned long address;
+
+	if (!arg)
+		return -EINVAL;
+
+	address = memparse(arg, &arg);
+	reserve_top_address(address);
+	return 0;
+}
+early_param("reservetop", parse_reservetop);

[PATCH] x86: cleanup using max_low_pfn for 32 bit

[Yinghai Lu]

for max_low_pfn is not changed after it is set.
so we can move that early and out of initmem_init.

could call find_low_pfn_range just after max_pfn is set.

also could move reserve_initrd out of setup_bootmem_allocator

so 32bit is more like 64bit.

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/setup_32.c |   16 ++++++++++------
 arch/x86/kernel/setup_64.c |    2 +-
 arch/x86/mm/discontig_32.c |   22 +++++++---------------
 arch/x86/mm/init_32.c      |   25 +++++++++----------------
 include/asm-x86/page_32.h  |    3 ++-
 include/asm-x86/setup.h    |    2 --
 6 files changed, 29 insertions(+), 41 deletions(-)

Index: linux-2.6/arch/x86/kernel/setup_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_32.c
+++ linux-2.6/arch/x86/kernel/setup_32.c
@@ -189,13 +189,14 @@ static inline void copy_edd(void)
 
 static bool do_relocate_initrd = false;
 
-void __init reserve_initrd(void)
+static void __init reserve_initrd(void)
 {
 	u64 ramdisk_image = boot_params.hdr.ramdisk_image;
 	u64 ramdisk_size  = boot_params.hdr.ramdisk_size;
 	u64 ramdisk_end   = ramdisk_image + ramdisk_size;
 	u64 end_of_lowmem = max_low_pfn << PAGE_SHIFT;
 	u64 ramdisk_here;
+	u64 ramdisk_target;
 
 	if (!boot_params.hdr.type_of_loader ||
 	    !ramdisk_image || !ramdisk_size)
@@ -203,7 +204,7 @@ void __init reserve_initrd(void)
 
 	initrd_start = 0;
 
-	if (ramdisk_size >= end_of_lowmem/2) {
+	if (ramdisk_size >= (end_of_lowmem>>1)) {
 		free_early(ramdisk_image, ramdisk_end);
 		printk(KERN_ERR "initrd too large to handle, "
 		       "disabling initrd\n");
@@ -226,7 +227,8 @@ void __init reserve_initrd(void)
 	}
 
 	/* We need to move the initrd down into lowmem */
-	ramdisk_here = find_e820_area(min_low_pfn<<PAGE_SHIFT,
+	ramdisk_target = max_pfn_mapped<<PAGE_SHIFT;
+	ramdisk_here = find_e820_area(min(ramdisk_target, end_of_lowmem>>1),
 				 end_of_lowmem, ramdisk_size,
 				 PAGE_SIZE);
 
@@ -347,8 +349,6 @@ static void set_mca_bus(int x) { }
  */
 void __init setup_arch(char **cmdline_p)
 {
-	unsigned long max_low_pfn;
-
 	memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data));
 	pre_setup_arch_hook();
 	early_cpu_init();
@@ -456,6 +456,10 @@ void __init setup_arch(char **cmdline_p)
 		max_pfn = e820_end_of_ram();
 	}
 
+	find_low_pfn_range();
+
+	reserve_initrd();
+
 	dmi_scan_machine();
 
 	io_delay_init();
@@ -472,7 +476,7 @@ void __init setup_arch(char **cmdline_p)
         acpi_numa_init();
 #endif
 
-	max_low_pfn = initmem_init(0, max_pfn);
+	initmem_init(0, max_pfn);
 
 #ifdef CONFIG_ACPI_SLEEP
 	/*
Index: linux-2.6/arch/x86/kernel/setup_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/setup_64.c
+++ linux-2.6/arch/x86/kernel/setup_64.c
@@ -154,7 +154,7 @@ static inline void copy_edd(void)
 }
 #endif
 
-void __init reserve_initrd(void)
+static void __init reserve_initrd(void)
 {
 #ifdef CONFIG_BLK_DEV_INITRD
 	if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) {
Index: linux-2.6/arch/x86/mm/discontig_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/discontig_32.c
+++ linux-2.6/arch/x86/mm/discontig_32.c
@@ -309,11 +309,10 @@ static void init_remap_allocator(int nid
 		(ulong) node_remap_end_vaddr[nid]);
 }
 
-unsigned long __init initmem_init(unsigned long start_pfn,
+void __init initmem_init(unsigned long start_pfn,
 				  unsigned long end_pfn)
 {
 	int nid;
-	unsigned long system_start_pfn, system_max_low_pfn;
 	long kva_target_pfn;
 
 	/*
@@ -324,17 +323,11 @@ unsigned long __init initmem_init(unsign
 	 * and ZONE_HIGHMEM.
 	 */
 
-	/* call find_max_low_pfn at first, it could update max_pfn */
-	system_max_low_pfn = max_low_pfn = find_max_low_pfn();
-
 	remove_all_active_ranges();
 	get_memcfg_numa();
 
 	kva_pages = round_up(calculate_numa_remap_pages(), PTRS_PER_PTE);
 
-	/* partially used pages are not usable - thus round upwards */
-	system_start_pfn = min_low_pfn = PFN_UP(init_pg_tables_end);
-
 	kva_target_pfn = round_down(max_low_pfn - kva_pages, PTRS_PER_PTE);
 	do {
 		kva_start_pfn = find_e820_area(kva_target_pfn<<PAGE_SHIFT,
@@ -357,19 +350,19 @@ unsigned long __init initmem_init(unsign
 		     "KVA PG");
 #ifdef CONFIG_HIGHMEM
 	highstart_pfn = highend_pfn = max_pfn;
-	if (max_pfn > system_max_low_pfn)
-		highstart_pfn = system_max_low_pfn;
+	if (max_pfn > max_low_pfn)
+		highstart_pfn = max_low_pfn;
 	printk(KERN_NOTICE "%ldMB HIGHMEM available.\n",
 	       pages_to_mb(highend_pfn - highstart_pfn));
 	num_physpages = highend_pfn;
 	high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1;
 #else
-	num_physpages = system_max_low_pfn;
-	high_memory = (void *) __va(system_max_low_pfn * PAGE_SIZE - 1) + 1;
+	num_physpages = max_low_pfn;
+	high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1;
 #endif
 	printk(KERN_NOTICE "%ldMB LOWMEM available.\n",
-			pages_to_mb(system_max_low_pfn));
-	printk("min_low_pfn = %ld, max_low_pfn = %ld, highstart_pfn = %ld\n", 
+			pages_to_mb(max_low_pfn));
+	printk("min_low_pfn = %ld, max_low_pfn = %ld, highstart_pfn = %ld\n",
 			min_low_pfn, max_low_pfn, highstart_pfn);
 
 	printk("Low memory ends at vaddr %08lx\n",
@@ -387,7 +380,6 @@ unsigned long __init initmem_init(unsign
 	memset(NODE_DATA(0), 0, sizeof(struct pglist_data));
 	NODE_DATA(0)->bdata = &node0_bdata;
 	setup_bootmem_allocator();
-	return max_low_pfn;
 }
 
 void __init zone_sizes_init(void)
Index: linux-2.6/arch/x86/mm/init_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/init_32.c
+++ linux-2.6/arch/x86/mm/init_32.c
@@ -561,9 +561,15 @@ early_param("highmem", parse_highmem);
 /*
  * Determine low and high memory ranges:
  */
-unsigned long __init find_max_low_pfn(void)
+void __init find_low_pfn_range(void)
 {
-	unsigned long max_low_pfn;
+	/* it could update max_pfn */
+
+	/*
+	 * partially used pages are not usable - thus
+	 * we are rounding upwards:
+	 */
+	min_low_pfn = PFN_UP(init_pg_tables_end);
 
 	max_low_pfn = max_pfn;
 	if (max_low_pfn > MAXMEM_PFN) {
@@ -625,21 +631,12 @@ unsigned long __init find_max_low_pfn(vo
 					" kernel!\n");
 #endif
 	}
-	return max_low_pfn;
 }
 
 #ifndef CONFIG_NEED_MULTIPLE_NODES
-unsigned long __init initmem_init(unsigned long start_pfn,
+void __init initmem_init(unsigned long start_pfn,
 				  unsigned long end_pfn)
 {
-	/*
-	 * partially used pages are not usable - thus
-	 * we are rounding upwards:
-	 */
-	min_low_pfn = PFN_UP(init_pg_tables_end);
-
-	max_low_pfn = find_max_low_pfn();
-
 #ifdef CONFIG_HIGHMEM
 	highstart_pfn = highend_pfn = max_pfn;
 	if (max_pfn > max_low_pfn)
@@ -661,8 +658,6 @@ unsigned long __init initmem_init(unsign
 			pages_to_mb(max_low_pfn));
 
 	setup_bootmem_allocator();
-
-	return max_low_pfn;
 }
 
 void __init zone_sizes_init(void)
@@ -699,8 +694,6 @@ void __init setup_bootmem_allocator(void
 		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
 	reserve_early(bootmap, bootmap + bootmap_size, "BOOTMAP");
 
-	reserve_initrd();
-
 	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, max_low_pfn);
 	printk(KERN_INFO "  mapped low ram: 0 - %08lx\n",
 		 max_pfn_mapped<<PAGE_SHIFT);
Index: linux-2.6/include/asm-x86/page_32.h
===================================================================
--- linux-2.6.orig/include/asm-x86/page_32.h
+++ linux-2.6/include/asm-x86/page_32.h
@@ -97,7 +97,8 @@ extern int sysctl_legacy_va_layout;
 #define VMALLOC_RESERVE		((unsigned long)__VMALLOC_RESERVE)
 #define MAXMEM			(-__PAGE_OFFSET - __VMALLOC_RESERVE)
 
-extern unsigned long initmem_init(unsigned long, unsigned long);
+extern void find_low_pfn_range(void);
+extern void initmem_init(unsigned long, unsigned long);
 extern void zone_sizes_init(void);
 extern void setup_bootmem_allocator(void);
 
Index: linux-2.6/include/asm-x86/setup.h
===================================================================
--- linux-2.6.orig/include/asm-x86/setup.h
+++ linux-2.6/include/asm-x86/setup.h
@@ -39,8 +39,6 @@ void reserve_crashkernel(void);
 #include <asm/bootparam.h>
 
 void reserve_standard_io_resources(void);
-void reserve_initrd(void);
-
 
 #ifndef _SETUP
 

[PATCH] x86: cleanup min_low_pfn

[Yinghai Lu]

for 32bit
we already had early_res support, so don't need to track min_low_pfn.
keep it to 0 always.

also use init_bootmem_node instead of init_bootmem, so don't touch
min_low_pfn.

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/mm/init_32.c |   10 ++++------
 arch/x86/mm/init_64.c |    4 +++-
 2 files changed, 7 insertions(+), 7 deletions(-)

Index: linux-2.6/arch/x86/mm/init_32.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/init_32.c
+++ linux-2.6/arch/x86/mm/init_32.c
@@ -565,11 +565,7 @@ void __init find_low_pfn_range(void)
 {
 	/* it could update max_pfn */
 
-	/*
-	 * partially used pages are not usable - thus
-	 * we are rounding upwards:
-	 */
-	min_low_pfn = PFN_UP(init_pg_tables_end);
+	/* max_low_pfn is 0, we already have early_res support */
 
 	max_low_pfn = max_pfn;
 	if (max_low_pfn > MAXMEM_PFN) {
@@ -694,7 +690,9 @@ void __init setup_bootmem_allocator(void
 		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
 	reserve_early(bootmap, bootmap + bootmap_size, "BOOTMAP");
 
-	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, max_low_pfn);
+	/* don't touch min_low_pfn */
+	bootmap_size = init_bootmem_node(NODE_DATA(0), bootmap >> PAGE_SHIFT,
+					 min_low_pfn, max_low_pfn);
 	printk(KERN_INFO "  mapped low ram: 0 - %08lx\n",
 		 max_pfn_mapped<<PAGE_SHIFT);
 	printk(KERN_INFO "  low ram: %08lx - %08lx\n",
Index: linux-2.6/arch/x86/mm/init_64.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/init_64.c
+++ linux-2.6/arch/x86/mm/init_64.c
@@ -633,7 +633,9 @@ void __init initmem_init(unsigned long s
 				 PAGE_SIZE);
 	if (bootmap == -1L)
 		panic("Cannot find bootmem map of size %ld\n", bootmap_size);
-	bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, end_pfn);
+	/* don't touch min_low_pfn */
+	bootmap_size = init_bootmem_node(NODE_DATA(0), bootmap >> PAGE_SHIFT,
+					 0, end_pfn);
 	e820_register_active_regions(0, start_pfn, end_pfn);
 	free_bootmem_with_active_regions(0, end_pfn);
 	early_res_to_bootmem(0, end_pfn<<PAGE_SHIFT);

Re: [PATCH] x86: cleanup using max_low_pfn for 32 bit

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> for max_low_pfn is not changed after it is set. so we can move that 
> early and out of initmem_init.
> 
> could call find_low_pfn_range just after max_pfn is set.
> 
> also could move reserve_initrd out of setup_bootmem_allocator
> 
> so 32bit is more like 64bit.

applied, thanks Yinghai.

i've picked up the following patches from you into a new 
tip/x86/setup-memory topic:

Yinghai Lu (19):
      x86: check command line when CONFIG_X86_MPPARSE is not set, v2
      x86: clean up init_amd()
      x86: remove some acpi ifdefs in setup_32/64
      x86: seperate funcs from setup_64 to cpu common_64.c
      x86: change identify_cpu to static
      x86: add e820_remove_range
      x86: seperate probe_roms into another file
      x86: merge setup64.c into common_64.c
      x86: remove two duplicated funcs in setup_32.c
      x86: move reserve_standard_io_resource to setup.c
      x86: move elfcorehdr parsing to setup.c
      x86: introduce initmem_init for 64 bit
      x86: introduce initmem_init for 32 bit
      x86: introduce reserve_initrd
      x86: move boot_params declaring to setup.c
      x86: move find_max_low_pfn to init_32.c
      x86: move reservetop and vmalloc parsing to pgtable_32.c
      x86: clean up using max_low_pfn on 32-bit
      x86: clean up min_low_pfn

very nice stuff - and it's working pretty well in my testing so far!

	Ingo

Re: [PATCH] x86: cleanup using max_low_pfn for 32 bit

[H. Peter Anvin]

Ingo Molnar wrote:
> 
> very nice stuff - and it's working pretty well in my testing so far!
> 

Very nice stuff indeed!

	-hpa

Re: [PATCH] x86: cleanup using max_low_pfn for 32 bit

[Ingo Molnar]

>       x86: clean up using max_low_pfn on 32-bit

this one needed the small fix below.

	Ingo

----------->
commit 749cffb6c81f6637bbd054a4db246196eaa69ccc
Author: Ingo Molnar <mingo@elte.hu>
Date:   Mon Jun 23 22:19:22 2008 +0200

    x86: build fix
    
    fix:
    
    arch/x86/kernel/setup_32.c:409: error: 'enable_local_apic' undeclared (first use in this function)
    arch/x86/kernel/setup_32.c:409: error: (Each undeclared identifier is reported only once
    arch/x86/kernel/setup_32.c:409: error: for each function it appears in.)
    
    Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c
index b42f570..1e67037 100644
--- a/arch/x86/kernel/setup_32.c
+++ b/arch/x86/kernel/setup_32.c
@@ -406,7 +406,9 @@ void __init setup_arch(char **cmdline_p)
 	parse_early_param();
 
 	if (acpi_mps_check()){
+#ifdef CONFIG_X86_LOCAL_APIC
 		enable_local_apic = -1;
+#endif
 		clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC);
 	}
 

Re: [PATCH] x86: cleanup using max_low_pfn for 32 bit

[Yinghai Lu]

On Mon, Jun 23, 2008 at 1:22 PM, Ingo Molnar <mingo@elte.hu> wrote:
>
>>       x86: clean up using max_low_pfn on 32-bit
>
> this one needed the small fix below.
>
>        Ingo
>
> ----------->
> commit 749cffb6c81f6637bbd054a4db246196eaa69ccc
> Author: Ingo Molnar <mingo@elte.hu>
> Date:   Mon Jun 23 22:19:22 2008 +0200
>
>    x86: build fix
>
>    fix:
>
>    arch/x86/kernel/setup_32.c:409: error: 'enable_local_apic' undeclared (first use in this function)
>    arch/x86/kernel/setup_32.c:409: error: (Each undeclared identifier is reported only once
>    arch/x86/kernel/setup_32.c:409: error: for each function it appears in.)
>
>    Signed-off-by: Ingo Molnar <mingo@elte.hu>
>
> diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c
> index b42f570..1e67037 100644
> --- a/arch/x86/kernel/setup_32.c
> +++ b/arch/x86/kernel/setup_32.c
> @@ -406,7 +406,9 @@ void __init setup_arch(char **cmdline_p)
>        parse_early_param();
>
>        if (acpi_mps_check()){
> +#ifdef CONFIG_X86_LOCAL_APIC
>                enable_local_apic = -1;
> +#endif
>                clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC);
>        }

need to merge enable_local_apic (32bit) and disable_apic (64 bit).

YH