[PATCH] x86-64 merge

x86_64 core updates.

 - Make it compile again (switch_to macros etc., add dummy suspend.h)
 - reenable strength reduce optimization
 - Fix ramdisk (patch from Mikael Pettersson)
 - Some merges from i386
 - Reimplement lazy iobitmap allocation.  I reimplemented it based
   on bcrl's idea.
 - Fix IPC 32bit emulation to actually work and move into own file
 - New fixed mtrr.c from DaveJ ported from 2.4 and reenable it.
 - Move tlbstate into PDA.
 - Add some changes that got lost during the last merge.
 - new memset that seems to actually work.
 - Align signal handler stack frames to 16 bytes.
 - Some more minor bugfixes.

arch/x86_64/Makefile

@@ -43,15 +43,9 @@ CFLAGS += -mcmodel=kernel
 CFLAGS += -pipe
 # this makes reading assembly source easier
 CFLAGS += -fno-reorder-blocks	
-# needed for later gcc 3.1
 CFLAGS += -finline-limit=2000
-# needed for earlier gcc 3.1
-#CFLAGS += -fno-strength-reduce
 #CFLAGS += -g
 
-# prevent gcc from keeping the stack 16 byte aligned (FIXME)
-#CFLAGS += -mpreferred-stack-boundary=2
-
 HEAD := arch/x86_64/kernel/head.o arch/x86_64/kernel/head64.o arch/x86_64/kernel/init_task.o
 
 SUBDIRS := arch/x86_64/tools $(SUBDIRS) arch/x86_64/kernel arch/x86_64/mm arch/x86_64/lib

arch/x86_64/boot/Makefile

@@ -21,10 +21,6 @@ ROOT_DEV := CURRENT
 
 SVGA_MODE := -DSVGA_MODE=NORMAL_VGA
 
-# If you want the RAM disk device, define this to be the size in blocks.
-
-RAMDISK := -DRAMDISK=512
-
 # ---------------------------------------------------------------------------
 
 BOOT_INCL =	$(TOPDIR)/include/linux/config.h \

arch/x86_64/config.in

@@ -47,8 +47,7 @@ define_bool CONFIG_EISA n
 define_bool CONFIG_X86_IO_APIC y
 define_bool CONFIG_X86_LOCAL_APIC y
 
-#currently broken:
-#bool 'MTRR (Memory Type Range Register) support' CONFIG_MTRR
+bool 'MTRR (Memory Type Range Register) support' CONFIG_MTRR
 bool 'Symmetric multi-processing support' CONFIG_SMP
 if [ "$CONFIG_SMP" = "n" ]; then
    bool 'Preemptible Kernel' CONFIG_PREEMPT
@@ -226,6 +225,7 @@ if [ "$CONFIG_DEBUG_KERNEL" != "n" ]; then
    bool '  Spinlock debugging' CONFIG_DEBUG_SPINLOCK
    bool '  Additional run-time checks' CONFIG_CHECKING
    bool '  Debug __init statements' CONFIG_INIT_DEBUG
+   bool '  Spinlock debugging' CONFIG_DEBUG_SPINLOCK
 fi
 endmenu
 

arch/x86_64/ia32/Makefile

@@ -9,8 +9,9 @@ export-objs := ia32_ioctl.o sys_ia32.o
 all: ia32.o
 
 O_TARGET := ia32.o
-obj-$(CONFIG_IA32_EMULATION) := ia32entry.o sys_ia32.o ia32_ioctl.o ia32_signal.o \
-	ia32_binfmt.o fpu32.o socket32.o ptrace32.o
+obj-$(CONFIG_IA32_EMULATION) := ia32entry.o sys_ia32.o ia32_ioctl.o \
+	ia32_signal.o \
+	ia32_binfmt.o fpu32.o socket32.o ptrace32.o ipc32.o
 
 clean::
 

arch/x86_64/kernel/ioport.c

@@ -14,6 +14,7 @@
 #include <linux/smp.h>
 #include <linux/smp_lock.h>
 #include <linux/stddef.h>
+#include <linux/slab.h>
 
 /* Set EXTENT bits starting at BASE in BITMAP to value TURN_ON. */
 static void set_bitmap(unsigned long *bitmap, short base, short extent, int new_value)
@@ -61,27 +62,19 @@ asmlinkage int sys_ioperm(unsigned long from, unsigned long num, int turn_on)
 		return -EINVAL;
 	if (turn_on && !capable(CAP_SYS_RAWIO))
 		return -EPERM;
-	/*
-	 * If it's the first ioperm() call in this thread's lifetime, set the
-	 * IO bitmap up. ioperm() is much less timing critical than clone(),
-	 * this is why we delay this operation until now:
-	 */
- 	if (!t->ioperm) {
-		/*
-		 * just in case ...
-		 */
-		memset(t->io_bitmap,0xff,(IO_BITMAP_SIZE+1)*4);
-		t->ioperm = 1;
-		/*
-		 * this activates it in the TSS
-		 */
+
+	if (!t->io_bitmap_ptr) { 
+		t->io_bitmap_ptr = kmalloc((IO_BITMAP_SIZE+1)*4, GFP_KERNEL);
+		if (!t->io_bitmap_ptr) 
+			return -ENOMEM; 
+		memset(t->io_bitmap_ptr,0xff,(IO_BITMAP_SIZE+1)*4);
 		tss->io_map_base = IO_BITMAP_OFFSET;
 	}
 
 	/*
 	 * do it in the per-thread copy and in the TSS ...
 	 */
-	set_bitmap((unsigned long *) t->io_bitmap, from, num, !turn_on);
+	set_bitmap((unsigned long *) t->io_bitmap_ptr, from, num, !turn_on);
 	set_bitmap((unsigned long *) tss->io_bitmap, from, num, !turn_on);
 
 	return 0;

arch/x86_64/kernel/process.c

@@ -39,6 +39,7 @@
 #include <linux/reboot.h>
 #include <linux/init.h>
 #include <linux/ctype.h>
+#include <linux/slab.h>
 
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
@@ -320,9 +321,6 @@ void show_regs(struct pt_regs * regs)
 	printk("CR2: %016lx CR3: %016lx CR4: %016lx\n", cr2, cr3, cr4);
 }
 
-#define __STR(x) #x
-#define __STR2(x) __STR(x)
-
 extern void load_gs_index(unsigned);
 
 /*
@@ -330,7 +328,13 @@ extern void load_gs_index(unsigned);
  */
 void exit_thread(void)
 {
-	/* nothing to do ... */
+	struct task_struct *me = current;
+	if (me->thread.io_bitmap_ptr) { 
+		kfree(me->thread.io_bitmap_ptr); 
+		me->thread.io_bitmap_ptr = NULL;
+		(init_tss + smp_processor_id())->io_map_base = 
+			INVALID_IO_BITMAP_OFFSET;
+	}
 }
 
 void flush_thread(void)
@@ -392,6 +396,14 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long rsp,
 	unlazy_fpu(current);	
 	p->thread.i387 = current->thread.i387;
 
+	if (unlikely(me->thread.io_bitmap_ptr != NULL)) { 
+		p->thread.io_bitmap_ptr = kmalloc((IO_BITMAP_SIZE+1)*4, GFP_KERNEL);
+		if (!p->thread.io_bitmap_ptr) 
+			return -ENOMEM;
+		memcpy(p->thread.io_bitmap_ptr, me->thread.io_bitmap_ptr, 
+		       (IO_BITMAP_SIZE+1)*4);
+	} 
+
 	return 0;
 }
 
@@ -491,21 +503,14 @@ void __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
 	/* 
 	 * Handle the IO bitmap 
 	 */ 
-	if (unlikely(prev->ioperm || next->ioperm)) {
-		if (next->ioperm) {
+	if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) {
+		if (next->io_bitmap_ptr) {
 			/*
 			 * 4 cachelines copy ... not good, but not that
 			 * bad either. Anyone got something better?
 			 * This only affects processes which use ioperm().
-			 * [Putting the TSSs into 4k-tlb mapped regions
-			 * and playing VM tricks to switch the IO bitmap
-			 * is not really acceptable.]
-			 * On x86-64 we could put multiple bitmaps into 
-			 * the GDT and just switch offsets
-			 * This would require ugly special cases on overflow
-			 * though -AK 
 			 */
-			memcpy(tss->io_bitmap, next->io_bitmap,
+			memcpy(tss->io_bitmap, next->io_bitmap_ptr,
 				 IO_BITMAP_SIZE*sizeof(u32));
 			tss->io_map_base = IO_BITMAP_OFFSET;
 		} else {

arch/x86_64/kernel/setup64.c

@@ -91,6 +91,9 @@ void pda_init(int cpu)
 	pda->me = pda;
 	pda->cpudata_offset = 0;
 
+	pda->active_mm = &init_mm;
+	pda->mmu_state = 0;
+	
 	asm volatile("movl %0,%%fs ; movl %0,%%gs" :: "r" (0)); 
 	wrmsrl(MSR_GS_BASE, cpu_pda + cpu);
 } 

arch/x86_64/kernel/signal.c

@@ -84,7 +84,6 @@ struct rt_sigframe
 	char *pretcode;
 	struct ucontext uc;
 	struct siginfo info;
-	struct _fpstate fpstate;
 };
 
 static int
@@ -186,8 +185,7 @@ asmlinkage long sys_rt_sigreturn(struct pt_regs regs)
  */
 
 static int
-setup_sigcontext(struct sigcontext *sc, struct _fpstate *fpstate,
-		 struct pt_regs *regs, unsigned long mask)
+setup_sigcontext(struct sigcontext *sc, struct pt_regs *regs, unsigned long mask)
 {
 	int tmp, err = 0;
 	struct task_struct *me = current;
@@ -221,20 +219,17 @@ setup_sigcontext(struct sigcontext *sc, struct _fpstate *fpstate,
 	err |= __put_user(mask, &sc->oldmask);
 	err |= __put_user(me->thread.cr2, &sc->cr2);
 
-	tmp = save_i387(fpstate);
-	if (tmp < 0)
-	  err = 1;
-	else
-	  err |= __put_user(tmp ? fpstate : NULL, &sc->fpstate);
-
 	return err;
 }
 
 /*
  * Determine which stack to use..
  */
-static inline struct rt_sigframe *
-get_sigframe(struct k_sigaction *ka, struct pt_regs * regs)
+
+#define round_down(p, r) ((void *)  ((unsigned long)((p) - (r) + 1) & ~((r)-1)))
+
+static void *
+get_stack(struct k_sigaction *ka, struct pt_regs *regs, unsigned long size)
 {
 	unsigned long rsp;
 
@@ -247,22 +242,34 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs * regs)
 			rsp = current->sas_ss_sp + current->sas_ss_size;
 	}
 
-	rsp = (rsp - sizeof(struct _fpstate)) & ~(15UL); 
-	rsp -= offsetof(struct rt_sigframe, fpstate);
-
-	return (struct rt_sigframe *) rsp; 
+	return round_down(rsp - size, 16); 
 }
 
 static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 			   sigset_t *set, struct pt_regs * regs)
 {
-	struct rt_sigframe *frame;
+	struct rt_sigframe *frame = NULL;
+	struct _fpstate *fp = NULL; 
 	int err = 0;
 
-	frame = get_sigframe(ka, regs);
+	if (current->used_math) {
+		fp = get_stack(ka, regs, sizeof(struct _fpstate)); 
+		frame = round_down((char *)fp - sizeof(struct rt_sigframe), 16) - 8;
 
-	if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
+		if (!access_ok(VERIFY_WRITE, fp, sizeof(struct _fpstate))) { 
 		goto give_sigsegv;
+		}
+
+		if (save_i387(fp) < 0) 
+			err |= -1; 
+	}
+
+	if (!frame)
+		frame = get_stack(ka, regs, sizeof(struct rt_sigframe)) - 8;
+
+	if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) {
+		goto give_sigsegv;
+	}
 
 	if (ka->sa.sa_flags & SA_SIGINFO) { 
 		err |= copy_siginfo_to_user(&frame->info, info);
@@ -278,14 +285,10 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 	err |= __put_user(sas_ss_flags(regs->rsp),
 			  &frame->uc.uc_stack.ss_flags);
 	err |= __put_user(current->sas_ss_size, &frame->uc.uc_stack.ss_size);
-	err |= setup_sigcontext(&frame->uc.uc_mcontext, &frame->fpstate,
-			        regs, set->sig[0]);
+	err |= setup_sigcontext(&frame->uc.uc_mcontext, regs, set->sig[0]);
+	err |= __put_user(fp, &frame->uc.uc_mcontext.fpstate);
 	err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
 
-	if (err) { 
-		goto give_sigsegv;
-	} 
-
 	/* Set up to return from userspace.  If provided, use a stub
 	   already in userspace.  */
 	/* x86-64 should always use SA_RESTORER. */
@@ -297,7 +300,6 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 	}
 
 	if (err) { 
-		printk("fault 3\n"); 
 		goto give_sigsegv;
 	} 
 
@@ -305,7 +307,6 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 	printk("%d old rip %lx old rsp %lx old rax %lx\n", current->pid,regs->rip,regs->rsp,regs->rax);
 #endif
 
-
 	/* Set up registers for signal handler */
 	{ 
 		struct exec_domain *ed = current_thread_info()->exec_domain;
@@ -320,9 +321,10 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 	   next argument after the signal number on the stack. */
 	regs->rsi = (unsigned long)&frame->info; 
 	regs->rdx = (unsigned long)&frame->uc; 
-	regs->rsp = (unsigned long) frame;
 	regs->rip = (unsigned long) ka->sa.sa_handler;
 
+	regs->rsp = (unsigned long)frame;
+
 	set_fs(USER_DS);
 	regs->eflags &= ~TF_MASK;
 

arch/x86_64/kernel/smp.c

@@ -25,8 +25,6 @@
 /* The 'big kernel lock' */
 spinlock_t kernel_flag __cacheline_aligned_in_smp = SPIN_LOCK_UNLOCKED;
 
-struct tlb_state cpu_tlbstate[NR_CPUS] = {[0 ... NR_CPUS-1] = { &init_mm, 0 }};
-
 /*
  * the following functions deal with sending IPIs between CPUs.
  *
@@ -147,9 +145,9 @@ static spinlock_t tlbstate_lock = SPIN_LOCK_UNLOCKED;
  */
 static void inline leave_mm (unsigned long cpu)
 {
-	if (cpu_tlbstate[cpu].state == TLBSTATE_OK)
+	if (read_pda(mmu_state) == TLBSTATE_OK)
 		BUG();
-	clear_bit(cpu, &cpu_tlbstate[cpu].active_mm->cpu_vm_mask);
+	clear_bit(cpu, &read_pda(active_mm)->cpu_vm_mask);
 	__flush_tlb();
 }
 
@@ -164,18 +162,18 @@ static void inline leave_mm (unsigned long cpu)
  * 	the other cpus, but smp_invalidate_interrupt ignore flush ipis
  * 	for the wrong mm, and in the worst case we perform a superflous
  * 	tlb flush.
- * 1a2) set cpu_tlbstate to TLBSTATE_OK
+ * 1a2) set cpu mmu_state to TLBSTATE_OK
  * 	Now the smp_invalidate_interrupt won't call leave_mm if cpu0
  *	was in lazy tlb mode.
- * 1a3) update cpu_tlbstate[].active_mm
+ * 1a3) update cpu active_mm
  * 	Now cpu0 accepts tlb flushes for the new mm.
  * 1a4) set_bit(cpu, &new_mm->cpu_vm_mask);
  * 	Now the other cpus will send tlb flush ipis.
  * 1a4) change cr3.
  * 1b) thread switch without mm change
- *	cpu_tlbstate[].active_mm is correct, cpu0 already handles
+ *	cpu active_mm is correct, cpu0 already handles
  *	flush ipis.
- * 1b1) set cpu_tlbstate to TLBSTATE_OK
+ * 1b1) set cpu mmu_state to TLBSTATE_OK
  * 1b2) test_and_set the cpu bit in cpu_vm_mask.
  * 	Atomically set the bit [other cpus will start sending flush ipis],
  * 	and test the bit.
@@ -188,7 +186,7 @@ static void inline leave_mm (unsigned long cpu)
  *   runs in kernel space, the cpu could load tlb entries for user space
  *   pages.
  *
- * The good news is that cpu_tlbstate is local to each cpu, no
+ * The good news is that cpu mmu_state is local to each cpu, no
  * write/read ordering problems.
  */
 
@@ -216,8 +214,8 @@ asmlinkage void smp_invalidate_interrupt (void)
 		 * BUG();
 		 */
 		 
-	if (flush_mm == cpu_tlbstate[cpu].active_mm) {
-		if (cpu_tlbstate[cpu].state == TLBSTATE_OK) {
+	if (flush_mm == read_pda(active_mm)) {
+		if (read_pda(mmu_state) == TLBSTATE_OK) {
 			if (flush_va == FLUSH_ALL)
 				local_flush_tlb();
 			else
@@ -335,7 +333,7 @@ static inline void do_flush_tlb_all_local(void)
 	unsigned long cpu = smp_processor_id();
 
 	__flush_tlb_all();
-	if (cpu_tlbstate[cpu].state == TLBSTATE_LAZY)
+	if (read_pda(mmu_state) == TLBSTATE_LAZY)
 		leave_mm(cpu);
 }
 

arch/x86_64/kernel/vsyscall.c

@@ -47,7 +47,7 @@
 
 #define __vsyscall(nr) __attribute__ ((unused,__section__(".vsyscall_" #nr)))
 
-#define NO_VSYSCALL 1
+//#define NO_VSYSCALL 1
 
 #ifdef NO_VSYSCALL
 #include <asm/unistd.h>

arch/x86_64/kernel/x8664_ksyms.c

@@ -189,3 +189,5 @@ EXPORT_SYMBOL_NOVERS(do_softirq_thunk);
 
 void out_of_line_bug(void);
 EXPORT_SYMBOL(out_of_line_bug);
+
+EXPORT_SYMBOL(init_level4_pgt);

arch/x86_64/lib/Makefile

@@ -12,7 +12,7 @@ obj-y  = csum-partial.o csum-copy.o csum-wrappers.o delay.o \
 	thunk.o io.o clear_page.o copy_page.o
 obj-y += memcpy.o
 obj-y += memmove.o
-#obj-y += memset.o
+obj-y += memset.o
 obj-y += copy_user.o
 
 export-objs := io.o csum-wrappers.o csum-partial.o

arch/x86_64/lib/memset.S

-/* Copyright 2002 Andi Kleen, SuSE Labs */
-
-	// #define FIX_ALIGNMENT 1
+/* Copyright 2002 Andi Kleen */
 
 /*
  * ISO C memset - set a memory block to a byte value.
@@ -11,51 +9,51 @@
  * 
  * rax   original destination
  */	
-	.globl ____memset
+ 	.globl __memset
+	.globl memset
 	.p2align
-____memset:
-	movq %rdi,%r10		/* save destination for return address */
-	movq %rdx,%r11		/* save count */ 
+memset:	
+__memset:
+	movq %rdi,%r10
+	movq %rdx,%r11
 
 	/* expand byte value  */
-	movzbl %sil,%ecx	/* zero extend char value */
-	movabs $0x0101010101010101,%rax		/* expansion pattern */
-	mul    %rcx		/* expand with rax, clobbers rdx */
+	movzbl %sil,%ecx
+	movabs $0x0101010101010101,%rax
+	mul    %rcx		/* with rax, clobbers rdx */
 
-#ifdef FIX_ALIGNMENT
 	/* align dst */
 	movl  %edi,%r9d		
-	andl  $7,%r9d		/* test unaligned bits */
+	andl  $7,%r9d	
 	jnz  bad_alignment
 after_bad_alignment:
-#endif
 	
-	movq %r11,%rcx		/* restore count */
-	shrq $6,%rcx		/* divide by 64 */
-	jz	 handle_tail	/* block smaller than 64 bytes? */
-	movl $64,%r8d		/* CSE loop block size */
+	movq %r11,%rcx
+	movl $64,%r8d
+	shrq $6,%rcx
+	jz	 handle_tail
 
 loop_64:	
-	movnti  %rax,0*8(%rdi) 
-	movnti  %rax,1*8(%rdi) 
-	movnti  %rax,2*8(%rdi) 
-	movnti  %rax,3*8(%rdi) 
-	movnti  %rax,4*8(%rdi) 
-	movnti  %rax,5*8(%rdi) 
-	movnti  %rax,6*8(%rdi) 
-	movnti  %rax,7*8(%rdi) 	/* clear 64 byte blocks */
-	addq    %r8,%rdi	/* increase pointer by 64 bytes */
-	loop	loop_64		/* decrement rcx and if not zero loop */
+	movnti  %rax,(%rdi) 
+	movnti  %rax,8(%rdi) 
+	movnti  %rax,16(%rdi) 
+	movnti  %rax,24(%rdi) 
+	movnti  %rax,32(%rdi) 
+	movnti  %rax,40(%rdi) 
+	movnti  %rax,48(%rdi) 
+	movnti  %rax,56(%rdi) 
+	addq    %r8,%rdi
+	loop	loop_64
 
 	/* Handle tail in loops. The loops should be faster than hard
 	   to predict jump tables. */ 
 handle_tail:
 	movl	%r11d,%ecx
-	andl    $63,%ecx
-	shrl	$3,%ecx
+	andl    $63&(~7),%ecx
 	jz 	handle_7
+	shrl	$3,%ecx
 loop_8:
-	movnti  %rax,(%rdi) 	/* long words */
+	movnti  %rax,(%rdi) 
 	addq    $8,%rdi
 	loop 	loop_8   
 
@@ -64,22 +62,20 @@ handle_7:
 	andl	$7,%ecx
 	jz      ende
 loop_1:
-	movb 	%al,(%rdi)	/* bytes */
-	incq	%rdi
+	movb 	%al,(%rdi)
+	addq	$1,%rdi
 	loop	loop_1
 	
 ende:	
 	movq	%r10,%rax
 	ret
 
-#ifdef FIX_ALIGNMENT
 bad_alignment:
-	andq $-8,%r11		/* shorter than 8 bytes */
-	jz      handle_7	/* if yes handle it in the tail code */
-	movnti %rax,(%rdi)	/* unaligned store of 8 bytes */
+	cmpq $7,%r11
+	jbe	handle_7
+	movnti %rax,(%rdi)	/* unaligned store */
 	movq $8,%r8			
-	subq %r9,%r8 		/* compute alignment (8-misalignment) */
-	addq %r8,%rdi		/* fix destination */
-	subq %r8,%r11		/* fix count */
+	subq %r9,%r8 
+	addq %r8,%rdi
+	subq %r8,%r11
 	jmp after_bad_alignment
-#endif

include/asm-x86_64/i387.h

@@ -16,11 +16,22 @@
 #include <asm/processor.h>
 #include <asm/sigcontext.h>
 #include <asm/user.h>
+#include <asm/thread_info.h>
 
 extern void fpu_init(void);
 extern void init_fpu(void);
 int save_i387(struct _fpstate *buf);
 
+static inline int need_signal_i387(struct task_struct *me) 
+{ 
+	if (!me->used_math)
+		return 0;
+	me->used_math = 0; 
+	if (!test_thread_flag(TIF_USEDFPU))
+		return 0;
+	return 1;
+} 
+
 /*
  * FPU lazy state save handling...
  */

include/asm-x86_64/ia32.h

@@ -18,7 +18,9 @@ typedef int		       __kernel_clock_t32;
 typedef int		       __kernel_pid_t32;
 typedef unsigned short	       __kernel_ipc_pid_t32;
 typedef unsigned short	       __kernel_uid_t32;
+typedef unsigned 				__kernel_uid32_t32;
 typedef unsigned short	       __kernel_gid_t32;
+typedef unsigned 				__kernel_gid32_t32;
 typedef unsigned short	       __kernel_dev_t32;
 typedef unsigned int	       __kernel_ino_t32;
 typedef unsigned short	       __kernel_mode_t32;

include/asm-x86_64/ipc.h

 #ifndef __i386_IPC_H__
 #define __i386_IPC_H__
 
-/* 
- * These are used to wrap system calls on x86.
- *
- * See arch/i386/kernel/sys_i386.c for ugly details..
- *
- * (on x86-64 only used for 32bit emulation) 
- */
-
-struct ipc_kludge {
-	struct msgbuf *msgp;
-	long msgtyp;
-};
-
-#define SEMOP		 1
-#define SEMGET		 2
-#define SEMCTL		 3
-#define MSGSND		11
-#define MSGRCV		12
-#define MSGGET		13
-#define MSGCTL		14
-#define SHMAT		21
-#define SHMDT		22
-#define SHMGET		23
-#define SHMCTL		24
-
-/* Used by the DIPC package, try and avoid reusing it */
-#define DIPC            25
-
-#define IPCCALL(version,op)	((version)<<16 | (op))
+/* dummy */
 
 #endif

include/asm-x86_64/mmu_context.h

@@ -19,8 +19,8 @@ int init_new_context(struct task_struct *tsk, struct mm_struct *mm);
 
 static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk, unsigned cpu)
 {
-	if(cpu_tlbstate[cpu].state == TLBSTATE_OK)
-		cpu_tlbstate[cpu].state = TLBSTATE_LAZY;	
+	if (read_pda(mmu_state) == TLBSTATE_OK) 
+		write_pda(mmu_state, TLBSTATE_LAZY);
 }
 #else
 static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk, unsigned cpu)
@@ -35,8 +35,8 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
 		/* stop flush ipis for the previous mm */
 		clear_bit(cpu, &prev->cpu_vm_mask);
 #ifdef CONFIG_SMP
-		cpu_tlbstate[cpu].state = TLBSTATE_OK;
-		cpu_tlbstate[cpu].active_mm = next;
+		write_pda(mmu_state, TLBSTATE_OK);
+		write_pda(active_mm, next);
 #endif
 		set_bit(cpu, &next->cpu_vm_mask);
 		/* Re-load page tables */
@@ -48,8 +48,8 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
 	}
 #ifdef CONFIG_SMP
 	else {
-		cpu_tlbstate[cpu].state = TLBSTATE_OK;
-		if(cpu_tlbstate[cpu].active_mm != next)
+		write_pda(mmu_state, TLBSTATE_OK);
+		if (read_pda(active_mm) != next)
 			out_of_line_bug();
 		if(!test_and_set_bit(cpu, &next->cpu_vm_mask)) {
 			/* We were in lazy tlb mode and leave_mm disabled 

include/asm-x86_64/msr.h

@@ -95,6 +95,7 @@
 #define MSR_IA32_PERFCTR0      0xc1
 #define MSR_IA32_PERFCTR1      0xc2
 
+#define MSR_MTRRcap		0x0fe
 #define MSR_IA32_BBL_CR_CTL        0x119
 
 #define MSR_IA32_MCG_CAP       0x179
@@ -110,6 +111,19 @@
 #define MSR_IA32_LASTINTFROMIP     0x1dd
 #define MSR_IA32_LASTINTTOIP       0x1de
 
+#define MSR_MTRRfix64K_00000	0x250
+#define MSR_MTRRfix16K_80000	0x258
+#define MSR_MTRRfix16K_A0000	0x259
+#define MSR_MTRRfix4K_C0000	0x268
+#define MSR_MTRRfix4K_C8000	0x269
+#define MSR_MTRRfix4K_D0000	0x26a
+#define MSR_MTRRfix4K_D8000	0x26b
+#define MSR_MTRRfix4K_E0000	0x26c
+#define MSR_MTRRfix4K_E8000	0x26d
+#define MSR_MTRRfix4K_F0000	0x26e
+#define MSR_MTRRfix4K_F8000	0x26f
+#define MSR_MTRRdefType		0x2ff
+
 #define MSR_IA32_MC0_CTL       0x400
 #define MSR_IA32_MC0_STATUS        0x401
 #define MSR_IA32_MC0_ADDR      0x402
@@ -171,11 +185,4 @@
 #define MSR_IA32_APICBASE_ENABLE        (1<<11)
 #define MSR_IA32_APICBASE_BASE          (0xfffff<<12)
 
-
-#define MSR_IA32_THERM_CONTROL		0x19a
-#define MSR_IA32_THERM_INTERRUPT	0x19b
-#define MSR_IA32_THERM_STATUS		0x19c
-#define MSR_IA32_MISC_ENABLE		0x1a0
-
-
 #endif

include/asm-x86_64/mtrr.h

@@ -30,16 +30,16 @@
 
 struct mtrr_sentry
 {
-    unsigned long base;    /*  Base address     */
-    unsigned long size;    /*  Size of region   */
+    __u64 base;    /*  Base address     */
+    __u32 size;    /*  Size of region   */
     unsigned int type;     /*  Type of region   */
 };
 
 struct mtrr_gentry
 {
+    __u64 base;    /*  Base address     */
+    __u32 size;    /*  Size of region   */
     unsigned int regnum;   /*  Register number  */
-    unsigned long base;    /*  Base address     */
-    unsigned long size;    /*  Size of region   */
     unsigned int type;     /*  Type of region   */
 };
 
@@ -81,46 +81,38 @@ static char *mtrr_strings[MTRR_NUM_TYPES] =
 #ifdef __KERNEL__
 
 /*  The following functions are for use by other drivers  */
-# ifdef CONFIG_MTRR
-extern int mtrr_add (unsigned long base, unsigned long size,
-		     unsigned int type, char increment);
-extern int mtrr_add_page (unsigned long base, unsigned long size,
-		     unsigned int type, char increment);
-extern int mtrr_del (int reg, unsigned long base, unsigned long size);
-extern int mtrr_del_page (int reg, unsigned long base, unsigned long size);
-extern void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi);
-#  else
-static __inline__ int mtrr_add (unsigned long base, unsigned long size,
+#ifdef CONFIG_MTRR
+extern int mtrr_add (__u64 base, __u32 size, unsigned int type, char increment);
+extern int mtrr_add_page (__u64 base, __u32 size, unsigned int type, char increment);
+extern int mtrr_del (int reg, __u64 base, __u32 size);
+extern int mtrr_del_page (int reg, __u64 base, __u32 size);
+#else
+static __inline__ int mtrr_add (__u64 base, __u32 size,
 				unsigned int type, char increment)
 {
     return -ENODEV;
 }
-static __inline__ int mtrr_add_page (unsigned long base, unsigned long size,
+static __inline__ int mtrr_add_page (__u64 base, __u32 size,
 				unsigned int type, char increment)
 {
     return -ENODEV;
 }
-static __inline__ int mtrr_del (int reg, unsigned long base,
-				unsigned long size)
+static __inline__ int mtrr_del (int reg, __u64 base, __u32 size)
 {
     return -ENODEV;
 }
-static __inline__ int mtrr_del_page (int reg, unsigned long base,
-				unsigned long size)
+static __inline__ int mtrr_del_page (int reg, __u64 base, __u32 size)
 {
     return -ENODEV;
 }
-
-static __inline__ void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi) {;}
-
-#  endif
+#endif
 
 /*  The following functions are for initialisation: don't use them!  */
 extern int mtrr_init (void);
-#  if defined(CONFIG_SMP) && defined(CONFIG_MTRR)
+#if defined(CONFIG_SMP) && defined(CONFIG_MTRR)
 extern void mtrr_init_boot_cpu (void);
 extern void mtrr_init_secondary_cpu (void);
-#  endif
+#endif
 
 #endif
 

include/asm-x86_64/pda.h

@@ -22,6 +22,8 @@ struct x8664_pda {
 	unsigned int __local_bh_count;
 	unsigned int __nmi_count;	/* arch dependent */
 	struct task_struct * __ksoftirqd_task; /* waitqueue is too large */
+	struct mm_struct *active_mm;
+	int mmu_state;     
 } ____cacheline_aligned;
 
 #define PDA_STACKOFFSET (5*8)

include/asm-x86_64/processor.h

@@ -45,21 +45,12 @@ struct cpuinfo_x86 {
 	__u8	x86_vendor;	/* CPU vendor */
 	__u8	x86_model;
 	__u8	x86_mask;
-  /* We know that wp_works_ok = 1, hlt_works_ok = 1, hard_math = 1,
-     etc... */
-	char	wp_works_ok;	/* It doesn't on 386's */
-	char	hlt_works_ok;	/* Problems on some 486Dx4's and old 386's */
-	char	hard_math;
-	char	rfu;
 	int	cpuid_level;	/* Maximum supported CPUID level, -1=no CPUID */
 	__u32	x86_capability[NCAPINTS];
 	char	x86_vendor_id[16];
 	char	x86_model_id[64];
 	int 	x86_cache_size;  /* in KB - valid for CPUS which support this
 				    call  */
-	int	fdiv_bug;
-	int	f00f_bug;
-	int	coma_bug;
 	unsigned long loops_per_jiffy;
 } ____cacheline_aligned;
 
@@ -323,7 +314,7 @@ struct thread_struct {
 /* IO permissions. the bitmap could be moved into the GDT, that would make
    switch faster for a limited number of ioperm using tasks. -AK */
 	int		ioperm;
-	u32	io_bitmap[IO_BITMAP_SIZE+1];
+	u32	*io_bitmap_ptr;
 };
 
 #define INIT_THREAD  {				\

include/asm-x86_64/spinlock.h

@@ -15,7 +15,7 @@ extern int printk(const char * fmt, ...)
 
 typedef struct {
 	volatile unsigned int lock;
-#ifdef CONFIG_DEBUG_SPINLOCK
+#if SPINLOCK_DEBUG
 	unsigned magic;
 #endif
 } spinlock_t;
@@ -39,7 +39,7 @@ typedef struct {
  * We make no fairness assumptions. They have a cost.
  */
 
-#define spin_is_locked(x)	(*(volatile char *)(&(x)->lock) <= 0)
+#define spin_is_locked(x)	(*(volatile signed char *)(&(x)->lock) <= 0)
 #define spin_unlock_wait(x)	do { barrier(); } while(spin_is_locked(x))
 
 #define spin_lock_string \
@@ -62,7 +62,7 @@ typedef struct {
 
 static inline int _raw_spin_trylock(spinlock_t *lock)
 {
-	char oldval;
+	signed char oldval;
 	__asm__ __volatile__(
 		"xchgb %b0,%1"
 		:"=q" (oldval), "=m" (lock->lock)

include/asm-x86_64/string.h

@@ -40,18 +40,9 @@ extern void *__memcpy(void *to, const void *from, size_t len);
 		 __ret = __builtin_memcpy((dst),(src),__len);	\
 	   __ret; }) 
 
-#if 0
+
 #define __HAVE_ARCH_MEMSET
-extern void *__memset(void *mem, int val, size_t len); 
-#define memset(dst,val,len)					\
-	({ size_t __len = (len);				\
-	   void *__ret;						\
-	   if (__builtin_constant_p(len) && __len >= 64)	\
-		   __ret = __memset((dst),(val),__len);		\
-	   else							\
-		   __ret = __builtin_memset((dst),(val),__len);	\
-	   __ret; }) 
-#endif	   
+#define memset __builtin_memset
 
 #define __HAVE_ARCH_MEMMOVE
 void * memmove(void * dest,const void *src,size_t count);

include/asm-x86_64/suspend.h

+#ifndef SUSPEND_H
+#define SUSPEND_H 1
+
+/* dummy for now */
+
+#endif

include/asm-x86_64/system.h

@@ -13,7 +13,10 @@
 #define LOCK_PREFIX ""
 #endif
 
-#define prepare_to_switch() do {} while(0)
+#define prepare_arch_schedule(prev)            do { } while(0)
+#define finish_arch_schedule(prev)             do { } while(0)
+#define prepare_arch_switch(rq)                        do { } while(0)
+#define finish_arch_switch(rq)                 spin_unlock_irq(&(rq)->lock)
 
 #define __STR(x) #x
 #define STR(x) __STR(x)
@@ -41,7 +44,7 @@
 	__POP(rax) __POP(r15) __POP(r14) __POP(r13) __POP(r12) __POP(r11) __POP(r10) \
 	__POP(r9) __POP(r8)
 
-#define switch_to(prev,next) \
+#define switch_to(prev,next,last) \
 	asm volatile(SAVE_CONTEXT						    \
 		     "movq %%rsp,%[prevrsp]\n\t"				    \
 		     "movq %[nextrsp],%%rsp\n\t"				    \

include/asm-x86_64/timex.h

@@ -48,6 +48,4 @@ static inline cycles_t get_cycles (void)
 
 extern unsigned int cpu_khz;
 
-#define ARCH_HAS_JIFFIES_64
-
 #endif

include/asm-x86_64/tlbflush.h

@@ -106,15 +106,6 @@ static inline void flush_tlb_range(struct vm_area_struct * vma, unsigned long st
 #define TLBSTATE_OK	1
 #define TLBSTATE_LAZY	2
 
-struct tlb_state
-{
-	struct mm_struct *active_mm;
-	int state;
-	char __cacheline_padding[24];
-};
-extern struct tlb_state cpu_tlbstate[NR_CPUS];
-
-
 #endif
 
 #define flush_tlb_kernel_range(start, end) flush_tlb_all()