git://git.onelab.eu
/
linux-2.6.git
/ blobdiff
commit
grep
author
committer
pickaxe
?
search:
re
summary
|
shortlog
|
log
|
commit
|
commitdiff
|
tree
raw
|
inline
| side by side
linux 2.6.16.38 w/ vs2.0.3-rc1
[linux-2.6.git]
/
include
/
asm-i386
/
processor.h
diff --git
a/include/asm-i386/processor.h
b/include/asm-i386/processor.h
index
e6712d6
..
bb5f709
100644
(file)
--- a/
include/asm-i386/processor.h
+++ b/
include/asm-i386/processor.h
@@
-20,7
+20,6
@@
#include <linux/config.h>
#include <linux/threads.h>
#include <asm/percpu.h>
#include <linux/config.h>
#include <linux/threads.h>
#include <asm/percpu.h>
-#include <linux/cpumask.h>
/* flag for disabling the tsc */
extern int tsc_disable;
/* flag for disabling the tsc */
extern int tsc_disable;
@@
-68,9
+67,6
@@
struct cpuinfo_x86 {
char pad0;
int x86_power;
unsigned long loops_per_jiffy;
char pad0;
int x86_power;
unsigned long loops_per_jiffy;
-#ifdef CONFIG_SMP
- cpumask_t llc_shared_map; /* cpus sharing the last level cache */
-#endif
unsigned char x86_max_cores; /* cpuid returned max cores value */
unsigned char booted_cores; /* number of cores as seen by OS */
unsigned char apicid;
unsigned char x86_max_cores; /* cpuid returned max cores value */
unsigned char booted_cores; /* number of cores as seen by OS */
unsigned char apicid;
@@
-107,7
+103,6
@@
extern struct cpuinfo_x86 cpu_data[];
extern int phys_proc_id[NR_CPUS];
extern int cpu_core_id[NR_CPUS];
extern int phys_proc_id[NR_CPUS];
extern int cpu_core_id[NR_CPUS];
-extern int cpu_llc_id[NR_CPUS];
extern char ignore_fpu_irq;
extern void identify_cpu(struct cpuinfo_x86 *);
extern char ignore_fpu_irq;
extern void identify_cpu(struct cpuinfo_x86 *);
@@
-321,17
+316,15
@@
extern unsigned int mca_pentium_flag;
extern int bootloader_type;
/*
extern int bootloader_type;
/*
- * User space process size:
3GB (
default).
+ * User space process size:
(3GB
default).
*/
*/
-#define TASK_SIZE (PAGE_OFFSET)
+#define __TASK_SIZE (__PAGE_OFFSET)
+#define TASK_SIZE ((unsigned long)__TASK_SIZE)
/* This decides where the kernel will search for a free chunk of vm
* space during mmap's.
*/
/* This decides where the kernel will search for a free chunk of vm
* space during mmap's.
*/
-#define TASK_UNMAPPED_BASE PAGE_ALIGN(TASK_SIZE/3)
-
-#define __HAVE_ARCH_ALIGN_STACK
-extern unsigned long arch_align_stack(unsigned long sp);
+#define TASK_UNMAPPED_BASE (PAGE_ALIGN(TASK_SIZE / 3))
#define HAVE_ARCH_PICK_MMAP_LAYOUT
#define HAVE_ARCH_PICK_MMAP_LAYOUT
@@
-513,9
+506,6
@@
static inline void load_esp0(struct tss_struct *tss, struct thread_struct *threa
regs->xcs = __USER_CS; \
regs->eip = new_eip; \
regs->esp = new_esp; \
regs->xcs = __USER_CS; \
regs->eip = new_eip; \
regs->esp = new_esp; \
- preempt_disable(); \
- load_user_cs_desc(smp_processor_id(), current->mm); \
- preempt_enable(); \
} while (0)
/*
} while (0)
/*
@@
-627,6
+617,8
@@
struct extended_sigtable {
unsigned int reserved[3];
struct extended_signature sigs[0];
};
unsigned int reserved[3];
struct extended_signature sigs[0];
};
+/* '6' because it used to be for P6 only (but now covers Pentium 4 as well) */
+#define MICROCODE_IOCFREE _IO('6',0)
/* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */
static inline void rep_nop(void)
/* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */
static inline void rep_nop(void)