Merge bk://kernel.bkbits.net/davem/net-2.5

into home.transmeta.com:/home/torvalds/v2.5/linux

arch/i386/kernel/entry.S

@@ -72,6 +72,12 @@ DF_MASK		= 0x00000400
 NT_MASK		= 0x00004000
 VM_MASK		= 0x00020000
 
+/*
+ * ESP0 is at offset 4. 0x100 is the size of the TSS, and
+ * also thus the top-of-stack pointer offset of SYSENTER_ESP
+ */
+TSS_ESP0_OFFSET = (4 - 0x100)
+
 #ifdef CONFIG_PREEMPT
 #define preempt_stop		cli
 #else
@@ -229,6 +235,8 @@ need_resched:
 
 	# sysenter call handler stub
 ENTRY(sysenter_entry)
+	movl TSS_ESP0_OFFSET(%esp),%esp
+sysenter_past_esp:
 	sti
 	pushl $(__USER_DS)
 	pushl %ebp
@@ -458,12 +466,36 @@ device_not_available_emulate:
 	addl $4, %esp
 	jmp ret_from_exception
 
+/*
+ * Debug traps and NMI can happen at the one SYSENTER instruction
+ * that sets up the real kernel stack. Check here, since we can't
+ * allow the wrong stack to be used.
+ *
+ * "TSS_ESP0_OFFSET+12" is because the NMI/debug handler will have
+ * already pushed 3 words if it hits on the sysenter instruction:
+ * eflags, cs and eip.
+ *
+ * We just load the right stack, and push the three (known) values
+ * by hand onto the new stack - while updating the return eip past
+ * the instruction that would have done it for sysenter.
+ */
+#define CHECK_SYSENTER_EIP			\
+	cmpl $sysenter_entry,(%esp);		\
+	jne 1f;					\
+	movl TSS_ESP0_OFFSET+12(%esp),%esp;	\
+	pushfl;					\
+	pushl $__KERNEL_CS;			\
+	pushl $sysenter_past_esp;		\
+1:
+
 ENTRY(debug)
+	CHECK_SYSENTER_EIP
 	pushl $0
 	pushl $do_debug
 	jmp error_code
 
 ENTRY(nmi)
+	CHECK_SYSENTER_EIP
 	pushl %eax
 	SAVE_ALL
 	movl %esp, %edx

arch/i386/kernel/sysenter.c

@@ -41,8 +41,9 @@ void enable_sep_cpu(void *info)
 	struct tss_struct *tss = init_tss + cpu;
 
 	tss->ss1 = __KERNEL_CS;
+	tss->esp1 = sizeof(struct tss_struct) + (unsigned long) tss;
 	wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0);
-	wrmsr(MSR_IA32_SYSENTER_ESP, tss->esp0, 0);
+	wrmsr(MSR_IA32_SYSENTER_ESP, tss->esp1, 0);
 	wrmsr(MSR_IA32_SYSENTER_EIP, (unsigned long) sysenter_entry, 0);
 
 	printk("Enabling SEP on CPU %d\n", cpu);

fs/locks.c

@@ -107,22 +107,22 @@
  *  Use slab allocator instead of kmalloc/kfree.
  *  Use generic list implementation from <linux/list.h>.
  *  Sped up posix_locks_deadlock by only considering blocked locks.
- *  Matthew Wilcox <willy@thepuffingroup.com>, March, 2000.
+ *  Matthew Wilcox <willy@debian.org>, March, 2000.
  *
  *  Leases and LOCK_MAND
- *  Matthew Wilcox <willy@linuxcare.com>, June, 2000.
+ *  Matthew Wilcox <willy@debian.org>, June, 2000.
  *  Stephen Rothwell <sfr@canb.auug.org.au>, June, 2000.
  */
 
-#include <linux/slab.h>
-#include <linux/file.h>
-#include <linux/smp_lock.h>
-#include <linux/init.h>
 #include <linux/capability.h>
-#include <linux/timer.h>
-#include <linux/time.h>
+#include <linux/file.h>
 #include <linux/fs.h>
+#include <linux/init.h>
+#include <linux/module.h>
 #include <linux/security.h>
+#include <linux/slab.h>
+#include <linux/smp_lock.h>
+#include <linux/time.h>
 
 #include <asm/semaphore.h>
 #include <asm/uaccess.h>
@@ -519,19 +519,11 @@ static void locks_delete_lock(struct file_lock **thisfl_p)
  */
 static int locks_conflict(struct file_lock *caller_fl, struct file_lock *sys_fl)
 {
-	switch (caller_fl->fl_type) {
-	case F_RDLCK:
-		return (sys_fl->fl_type == F_WRLCK);
-
-	case F_WRLCK:
-		return (1);
-
-	default:
-		printk(KERN_ERR "locks_conflict(): impossible lock type - %d\n",
-		       caller_fl->fl_type);
-		break;
-	}
-	return (0);	/* This should never happen */
+	if (sys_fl->fl_type == F_WRLCK)
+		return 1;
+	if (caller_fl->fl_type == F_WRLCK)
+		return 1;
+	return 0;
 }
 
 /* Determine if lock sys_fl blocks lock caller_fl. POSIX specific
@@ -1966,3 +1958,13 @@ static int __init filelock_init(void)
 }
 
 module_init(filelock_init)
+
+EXPORT_SYMBOL(file_lock_list);
+EXPORT_SYMBOL(locks_init_lock);
+EXPORT_SYMBOL(locks_copy_lock);
+EXPORT_SYMBOL(posix_lock_file);
+EXPORT_SYMBOL(posix_test_lock);
+EXPORT_SYMBOL(posix_block_lock);
+EXPORT_SYMBOL(posix_unblock_lock);
+EXPORT_SYMBOL(posix_locks_deadlock);
+EXPORT_SYMBOL(locks_mandatory_area);

include/asm-i386/processor.h

@@ -404,6 +404,8 @@ struct thread_struct {
 #define INIT_TSS  {							\
 	.esp0		= sizeof(init_stack) + (long)&init_stack,	\
 	.ss0		= __KERNEL_DS,					\
+	.esp1		= sizeof(init_tss[0]) + (long)&init_tss[0],	\
+	.ss1		= __KERNEL_CS,					\
 	.ldt		= GDT_ENTRY_LDT,				\
 	.bitmap		= INVALID_IO_BITMAP_OFFSET,			\
 	.io_bitmap	= { [ 0 ... IO_BITMAP_SIZE ] = ~0 },		\
@@ -412,12 +414,10 @@ struct thread_struct {
 static inline void load_esp0(struct tss_struct *tss, unsigned long esp0)
 {
 	tss->esp0 = esp0;
-	if (cpu_has_sep) {
-		if (tss->ss1 != __KERNEL_CS) {
-			tss->ss1 = __KERNEL_CS;
-			wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0);
-		}
-		wrmsr(MSR_IA32_SYSENTER_ESP, esp0, 0);
+	/* This can only happen when SEP is enabled, no need to test "SEP"arately */
+	if (tss->ss1 != __KERNEL_CS) {
+		tss->ss1 = __KERNEL_CS;
+		wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0);
 	}
 }
 

include/asm-i386/xor.h

@@ -18,22 +18,6 @@
  * Copyright (C) 1998 Ingo Molnar.
  */
 
-#define FPU_SAVE							\
-  do {									\
-	preempt_disable();						\
-	if (!test_thread_flag(TIF_USEDFPU))				\
-		__asm__ __volatile__ (" clts;\n");			\
-	__asm__ __volatile__ ("fsave %0; fwait": "=m"(fpu_save[0]));	\
-  } while (0)
-
-#define FPU_RESTORE							\
-  do {									\
-	__asm__ __volatile__ ("frstor %0": : "m"(fpu_save[0]));		\
-	if (!test_thread_flag(TIF_USEDFPU))				\
-		stts();							\
-	preempt_enable();						\
-  } while (0)
-
 #define LD(x,y)		"       movq   8*("#x")(%1), %%mm"#y"   ;\n"
 #define ST(x,y)		"       movq %%mm"#y",   8*("#x")(%1)   ;\n"
 #define XO1(x,y)	"       pxor   8*("#x")(%2), %%mm"#y"   ;\n"
@@ -46,9 +30,8 @@ static void
 xor_pII_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 {
 	unsigned long lines = bytes >> 7;
-	char fpu_save[108];
 
-	FPU_SAVE;
+	kernel_fpu_begin();
 
 	__asm__ __volatile__ (
 #undef BLOCK
@@ -83,7 +66,7 @@ xor_pII_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 	:
 	: "memory");
 
-	FPU_RESTORE;
+	kernel_fpu_end();
 }
 
 static void
@@ -91,9 +74,8 @@ xor_pII_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	      unsigned long *p3)
 {
 	unsigned long lines = bytes >> 7;
-	char fpu_save[108];
 
-	FPU_SAVE;
+	kernel_fpu_begin();
 
 	__asm__ __volatile__ (
 #undef BLOCK
@@ -133,7 +115,7 @@ xor_pII_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	:
 	: "memory");
 
-	FPU_RESTORE;
+	kernel_fpu_end();
 }
 
 static void
@@ -141,9 +123,8 @@ xor_pII_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	      unsigned long *p3, unsigned long *p4)
 {
 	unsigned long lines = bytes >> 7;
-	char fpu_save[108];
 
-	FPU_SAVE;
+	kernel_fpu_begin();
 
 	__asm__ __volatile__ (
 #undef BLOCK
@@ -188,7 +169,7 @@ xor_pII_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	:
 	: "memory");
 
-	FPU_RESTORE;
+	kernel_fpu_end();
 }
 
 
@@ -197,9 +178,8 @@ xor_pII_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	      unsigned long *p3, unsigned long *p4, unsigned long *p5)
 {
 	unsigned long lines = bytes >> 7;
-	char fpu_save[108];
 
-	FPU_SAVE;
+	kernel_fpu_begin();
 
 	/* need to save/restore p4/p5 manually otherwise gcc's 10 argument
 	   limit gets exceeded (+ counts as two arguments) */
@@ -255,7 +235,7 @@ xor_pII_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	: "r" (p4), "r" (p5) 
 	: "memory");
 
-	FPU_RESTORE;
+	kernel_fpu_end();
 }
 
 #undef LD
@@ -270,9 +250,8 @@ static void
 xor_p5_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 {
 	unsigned long lines = bytes >> 6;
-	char fpu_save[108];
 
-	FPU_SAVE;
+	kernel_fpu_begin();
 
 	__asm__ __volatile__ (
 	" .align 32	             ;\n"
@@ -311,7 +290,7 @@ xor_p5_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 	:
 	: "memory");
 
-	FPU_RESTORE;
+	kernel_fpu_end();
 }
 
 static void
@@ -319,9 +298,8 @@ xor_p5_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	     unsigned long *p3)
 {
 	unsigned long lines = bytes >> 6;
-	char fpu_save[108];
 
-	FPU_SAVE;
+	kernel_fpu_begin();
 
 	__asm__ __volatile__ (
 	" .align 32,0x90             ;\n"
@@ -369,7 +347,7 @@ xor_p5_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	:
 	: "memory" );
 
-	FPU_RESTORE;
+	kernel_fpu_end();
 }
 
 static void
@@ -377,9 +355,8 @@ xor_p5_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	     unsigned long *p3, unsigned long *p4)
 {
 	unsigned long lines = bytes >> 6;
-	char fpu_save[108];
 
-	FPU_SAVE;
+	kernel_fpu_begin();
 
 	__asm__ __volatile__ (
 	" .align 32,0x90             ;\n"
@@ -436,7 +413,7 @@ xor_p5_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	:
 	: "memory");
 
-	FPU_RESTORE;
+	kernel_fpu_end();
 }
 
 static void
@@ -444,9 +421,8 @@ xor_p5_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	     unsigned long *p3, unsigned long *p4, unsigned long *p5)
 {
 	unsigned long lines = bytes >> 6;
-	char fpu_save[108];
 
-	FPU_SAVE;
+	kernel_fpu_begin();
 
 	/* need to save p4/p5 manually to not exceed gcc's 10 argument limit */
 	__asm__ __volatile__ (
@@ -517,7 +493,7 @@ xor_p5_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 	: "r" (p4), "r" (p5)
 	: "memory");
 
-	FPU_RESTORE;
+	kernel_fpu_end();
 }
 
 static struct xor_block_template xor_block_pII_mmx = {
@@ -536,9 +512,6 @@ static struct xor_block_template xor_block_p5_mmx = {
 	.do_5 = xor_p5_mmx_5,
 };
 
-#undef FPU_SAVE
-#undef FPU_RESTORE
-
 /*
  * Cache avoiding checksumming functions utilizing KNI instructions
  * Copyright (C) 1999 Zach Brown (with obvious credit due Ingo)

kernel/ksyms.c

@@ -235,15 +235,6 @@ EXPORT_SYMBOL(generic_file_write_nolock);
 EXPORT_SYMBOL(generic_file_mmap);
 EXPORT_SYMBOL(generic_file_readonly_mmap);
 EXPORT_SYMBOL(generic_ro_fops);
-EXPORT_SYMBOL(file_lock_list);
-EXPORT_SYMBOL(locks_init_lock);
-EXPORT_SYMBOL(locks_copy_lock);
-EXPORT_SYMBOL(posix_lock_file);
-EXPORT_SYMBOL(posix_test_lock);
-EXPORT_SYMBOL(posix_block_lock);
-EXPORT_SYMBOL(posix_unblock_lock);
-EXPORT_SYMBOL(posix_locks_deadlock);
-EXPORT_SYMBOL(locks_mandatory_area);
 EXPORT_SYMBOL(dput);
 EXPORT_SYMBOL(have_submounts);
 EXPORT_SYMBOL(d_find_alias);