Merge http://lia64.bkbits.net/to-linus-2.5

into ppc970.osdl.org:/home/torvalds/v2.6/linux

Documentation/cachetlb.txt

@@ -132,6 +132,17 @@ changes occur:
 	translations for software managed TLB configurations.
 	The sparc64 port currently does this.
 
+7) void tlb_migrate_finish(struct mm_struct *mm)
+
+	This interface is called at the end of an explicit
+	process migration. This interface provides a hook
+	to allow a platform to update TLB or context-specific
+	information for the address space.
+
+	The ia64 sn2 platform is one example of a platform
+	that uses this interface.
+
+
 Next, we have the cache flushing interfaces.  In general, when Linux
 is changing an existing virtual-->physical mapping to a new value,
 the sequence will be in one of the following forms:

arch/ia64/ia32/ia32_entry.S

@@ -32,7 +32,7 @@ ENTRY(ia32_execve)
 END(ia32_execve)
 
 ENTRY(ia32_clone)
-	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(2)
+	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(5)
 	alloc r16=ar.pfs,5,2,6,0
 	DO_SAVE_SWITCH_STACK
 	mov loc0=rp
@@ -110,7 +110,9 @@ GLOBAL_ENTRY(ia32_ret_from_clone)
 	ld4 r2=[r2]
 	;;
 	mov r8=0
-	tbit.nz p6,p0=r2,TIF_SYSCALL_TRACE
+	and r2=_TIF_SYSCALL_TRACEAUDIT,r2
+	;;
+	cmp.ne p6,p0=r2,r0
 (p6)	br.cond.spnt .ia32_strace_check_retval
 	;;					// prevent RAW on r8
 END(ia32_ret_from_clone)
@@ -142,7 +144,7 @@ GLOBAL_ENTRY(ia32_trace_syscall)
 	adds r2=IA64_PT_REGS_R8_OFFSET+16,sp
 	;;
 	st8 [r2]=r3				// initialize return code to -ENOSYS
-	br.call.sptk.few rp=syscall_trace	// give parent a chance to catch syscall args
+	br.call.sptk.few rp=syscall_trace_enter	// give parent a chance to catch syscall args
 .ret2:	// Need to reload arguments (they may be changed by the tracing process)
 	adds r2=IA64_PT_REGS_R1_OFFSET+16,sp	// r2 = &pt_regs.r1
 	adds r3=IA64_PT_REGS_R13_OFFSET+16,sp	// r3 = &pt_regs.r13
@@ -170,7 +172,7 @@ GLOBAL_ENTRY(ia32_trace_syscall)
 	adds r2=IA64_PT_REGS_R8_OFFSET+16,sp	// r2 = &pt_regs.r8
 	;;
 	st8.spill [r2]=r8			// store return value in slot for r8
-	br.call.sptk.few rp=syscall_trace	// give parent a chance to catch return value
+	br.call.sptk.few rp=syscall_trace_leave	// give parent a chance to catch return value
 .ret4:	alloc r2=ar.pfs,0,0,0,0			// drop the syscall argument frame
 	br.cond.sptk.many ia64_leave_kernel
 END(ia32_trace_syscall)

arch/ia64/ia32/sys_ia32.c

@@ -1996,18 +1996,19 @@ sys32_sigaltstack (ia32_stack_t *uss32, ia32_stack_t *uoss32,
 	int ret;
 	mm_segment_t old_fs = get_fs();
 
-	if (uss32)
+	if (uss32) {
 		if (copy_from_user(&buf32, uss32, sizeof(ia32_stack_t)))
 			return -EFAULT;
-	uss.ss_sp = (void *) (long) buf32.ss_sp;
-	uss.ss_flags = buf32.ss_flags;
-	/* MINSIGSTKSZ is different for ia32 vs ia64. We lie here to pass the 
-           check and set it to the user requested value later */
-	if ((buf32.ss_flags != SS_DISABLE) && (buf32.ss_size < MINSIGSTKSZ_IA32)) {
-		ret = -ENOMEM;
-		goto out;
+		uss.ss_sp = (void *) (long) buf32.ss_sp;
+		uss.ss_flags = buf32.ss_flags;
+		/* MINSIGSTKSZ is different for ia32 vs ia64. We lie here to pass the 
+	           check and set it to the user requested value later */
+		if ((buf32.ss_flags != SS_DISABLE) && (buf32.ss_size < MINSIGSTKSZ_IA32)) {
+			ret = -ENOMEM;
+			goto out;
+		}
+		uss.ss_size = MINSIGSTKSZ;
 	}
-	uss.ss_size = MINSIGSTKSZ;
 	set_fs(KERNEL_DS);
 	ret = do_sigaltstack(uss32 ? &uss : NULL, &uoss, pt->r12);
  	current->sas_ss_size = buf32.ss_size;	

arch/ia64/kernel/efi.c

@@ -43,18 +43,20 @@ static unsigned long mem_limit = ~0UL, max_addr = ~0UL;
 
 #define efi_call_virt(f, args...)	(*(f))(args)
 
-#define STUB_GET_TIME(prefix, adjust_arg)							\
-static efi_status_t										\
-prefix##_get_time (efi_time_t *tm, efi_time_cap_t *tc)						\
-{												\
-	struct ia64_fpreg fr[6];								\
-	efi_status_t ret;									\
-												\
-	ia64_save_scratch_fpregs(fr);								\
-	ret = efi_call_##prefix((efi_get_time_t *) __va(runtime->get_time), adjust_arg(tm),	\
-				adjust_arg(tc));						\
-	ia64_load_scratch_fpregs(fr);								\
-	return ret;										\
+#define STUB_GET_TIME(prefix, adjust_arg)							  \
+static efi_status_t										  \
+prefix##_get_time (efi_time_t *tm, efi_time_cap_t *tc)						  \
+{												  \
+	struct ia64_fpreg fr[6];								  \
+	efi_time_cap_t *atc = 0;								  \
+	efi_status_t ret;									  \
+												  \
+	if (tc)											  \
+		atc = adjust_arg(tc);								  \
+	ia64_save_scratch_fpregs(fr);								  \
+	ret = efi_call_##prefix((efi_get_time_t *) __va(runtime->get_time), adjust_arg(tm), atc); \
+	ia64_load_scratch_fpregs(fr);								  \
+	return ret;										  \
 }
 
 #define STUB_SET_TIME(prefix, adjust_arg)							\
@@ -89,11 +91,14 @@ static efi_status_t										\
 prefix##_set_wakeup_time (efi_bool_t enabled, efi_time_t *tm)					\
 {												\
 	struct ia64_fpreg fr[6];								\
+	efi_time_t *atm = 0;									\
 	efi_status_t ret;									\
 												\
+	if (tm)											\
+		atm = adjust_arg(tm);								\
 	ia64_save_scratch_fpregs(fr);								\
 	ret = efi_call_##prefix((efi_set_wakeup_time_t *) __va(runtime->set_wakeup_time),	\
-				enabled, adjust_arg(tm));					\
+				enabled, atm);							\
 	ia64_load_scratch_fpregs(fr);								\
 	return ret;										\
 }
@@ -104,11 +109,14 @@ prefix##_get_variable (efi_char16_t *name, efi_guid_t *vendor, u32 *attr,		\
 		       unsigned long *data_size, void *data)				\
 {											\
 	struct ia64_fpreg fr[6];							\
+	u32 *aattr = 0;									\
 	efi_status_t ret;								\
 											\
+	if (attr)									\
+		aattr = adjust_arg(attr);						\
 	ia64_save_scratch_fpregs(fr);							\
 	ret = efi_call_##prefix((efi_get_variable_t *) __va(runtime->get_variable),	\
-				adjust_arg(name), adjust_arg(vendor), adjust_arg(attr),	\
+				adjust_arg(name), adjust_arg(vendor), aattr,		\
 				adjust_arg(data_size), adjust_arg(data));		\
 	ia64_load_scratch_fpregs(fr);							\
 	return ret;									\
@@ -164,33 +172,41 @@ prefix##_reset_system (int reset_type, efi_status_t status,			\
 		       unsigned long data_size, efi_char16_t *data)		\
 {										\
 	struct ia64_fpreg fr[6];						\
+	efi_char16_t *adata = 0;						\
+										\
+	if (data)								\
+		adata = adjust_arg(data);					\
 										\
 	ia64_save_scratch_fpregs(fr);						\
 	efi_call_##prefix((efi_reset_system_t *) __va(runtime->reset_system),	\
-			  reset_type, status, data_size, adjust_arg(data));	\
+			  reset_type, status, data_size, adata);		\
 	/* should not return, but just in case... */				\
 	ia64_load_scratch_fpregs(fr);						\
 }
 
-STUB_GET_TIME(phys, __pa)
-STUB_SET_TIME(phys, __pa)
-STUB_GET_WAKEUP_TIME(phys, __pa)
-STUB_SET_WAKEUP_TIME(phys, __pa)
-STUB_GET_VARIABLE(phys, __pa)
-STUB_GET_NEXT_VARIABLE(phys, __pa)
-STUB_SET_VARIABLE(phys, __pa)
-STUB_GET_NEXT_HIGH_MONO_COUNT(phys, __pa)
-STUB_RESET_SYSTEM(phys, __pa)
-
-STUB_GET_TIME(virt, )
-STUB_SET_TIME(virt, )
-STUB_GET_WAKEUP_TIME(virt, )
-STUB_SET_WAKEUP_TIME(virt, )
-STUB_GET_VARIABLE(virt, )
-STUB_GET_NEXT_VARIABLE(virt, )
-STUB_SET_VARIABLE(virt, )
-STUB_GET_NEXT_HIGH_MONO_COUNT(virt, )
-STUB_RESET_SYSTEM(virt, )
+#define phys_ptr(arg)	((__typeof__(arg)) ia64_tpa(arg))
+
+STUB_GET_TIME(phys, phys_ptr)
+STUB_SET_TIME(phys, phys_ptr)
+STUB_GET_WAKEUP_TIME(phys, phys_ptr)
+STUB_SET_WAKEUP_TIME(phys, phys_ptr)
+STUB_GET_VARIABLE(phys, phys_ptr)
+STUB_GET_NEXT_VARIABLE(phys, phys_ptr)
+STUB_SET_VARIABLE(phys, phys_ptr)
+STUB_GET_NEXT_HIGH_MONO_COUNT(phys, phys_ptr)
+STUB_RESET_SYSTEM(phys, phys_ptr)
+
+#define id(arg)	arg
+
+STUB_GET_TIME(virt, id)
+STUB_SET_TIME(virt, id)
+STUB_GET_WAKEUP_TIME(virt, id)
+STUB_SET_WAKEUP_TIME(virt, id)
+STUB_GET_VARIABLE(virt, id)
+STUB_GET_NEXT_VARIABLE(virt, id)
+STUB_SET_VARIABLE(virt, id)
+STUB_GET_NEXT_HIGH_MONO_COUNT(virt, id)
+STUB_RESET_SYSTEM(virt, id)
 
 void
 efi_gettimeofday (struct timespec *ts)

arch/ia64/kernel/entry.S

@@ -508,7 +508,7 @@ GLOBAL_ENTRY(ia64_trace_syscall)
 	;;
  	stf.spill [r16]=f10
  	stf.spill [r17]=f11
-	br.call.sptk.many rp=syscall_trace // give parent a chance to catch syscall args
+	br.call.sptk.many rp=syscall_trace_enter // give parent a chance to catch syscall args
 	adds r16=PT(F6)+16,sp
 	adds r17=PT(F7)+16,sp
 	;;
@@ -548,7 +548,7 @@ GLOBAL_ENTRY(ia64_trace_syscall)
 .strace_save_retval:
 .mem.offset 0,0; st8.spill [r2]=r8		// store return value in slot for r8
 .mem.offset 8,0; st8.spill [r3]=r10		// clear error indication in slot for r10
-	br.call.sptk.many rp=syscall_trace // give parent a chance to catch return value
+	br.call.sptk.many rp=syscall_trace_leave // give parent a chance to catch return value
 .ret3:	br.cond.sptk ia64_leave_syscall
 
 strace_error:
@@ -575,7 +575,7 @@ GLOBAL_ENTRY(ia64_strace_leave_kernel)
 	 */
 	nop.m 0
 	nop.i 0
-	br.call.sptk.many rp=syscall_trace // give parent a chance to catch return value
+	br.call.sptk.many rp=syscall_trace_leave // give parent a chance to catch return value
 }
 .ret4:	br.cond.sptk ia64_leave_kernel
 END(ia64_strace_leave_kernel)
@@ -601,7 +601,9 @@ GLOBAL_ENTRY(ia64_ret_from_clone)
 	ld4 r2=[r2]
 	;;
 	mov r8=0
-	tbit.nz p6,p0=r2,TIF_SYSCALL_TRACE
+	and r2=_TIF_SYSCALL_TRACEAUDIT,r2
+	;;
+	cmp.ne p6,p0=r2,r0
 (p6)	br.cond.spnt .strace_check_retval
 	;;					// added stop bits to prevent r8 dependency
 END(ia64_ret_from_clone)
@@ -663,25 +665,31 @@ GLOBAL_ENTRY(ia64_leave_syscall)
 	PT_REGS_UNWIND_INFO(0)
 	/*
 	 * work.need_resched etc. mustn't get changed by this CPU before it returns to
-	 * user- or fsys-mode, hence we disable interrupts early on:
+	 * user- or fsys-mode, hence we disable interrupts early on.
+	 *
+	 * p6 controls whether current_thread_info()->flags needs to be check for
+	 * extra work.  We always check for extra work when returning to user-level.
+	 * With CONFIG_PREEMPT, we also check for extra work when the preempt_count
+	 * is 0.  After extra work processing has been completed, execution
+	 * resumes at .work_processed_syscall with p6 set to 1 if the extra-work-check
+	 * needs to be redone.
 	 */
 #ifdef CONFIG_PREEMPT
 	rsm psr.i				// disable interrupts
-#else
-(pUStk)	rsm psr.i
-#endif
 	cmp.eq pLvSys,p0=r0,r0			// pLvSys=1: leave from syscall
-(pUStk)	cmp.eq.unc p6,p0=r0,r0			// p6 <- pUStk
-.work_processed_syscall:
-#ifdef CONFIG_PREEMPT
 (pKStk) adds r20=TI_PRE_COUNT+IA64_TASK_SIZE,r13
 	;;
 	.pred.rel.mutex pUStk,pKStk
 (pKStk) ld4 r21=[r20]			// r21 <- preempt_count
 (pUStk)	mov r21=0			// r21 <- 0
 	;;
-(p6)	cmp.eq.unc p6,p0=r21,r0		// p6 <- p6 && (r21 == 0)
-#endif /* CONFIG_PREEMPT */
+	cmp.eq p6,p0=r21,r0		// p6 <- pUStk || (preempt_count == 0)
+#else /* !CONFIG_PREEMPT */
+(pUStk)	rsm psr.i
+	cmp.eq pLvSys,p0=r0,r0		// pLvSys=1: leave from syscall
+(pUStk)	cmp.eq.unc p6,p0=r0,r0		// p6 <- pUStk
+#endif
+.work_processed_syscall:
 	adds r16=PT(LOADRS)+16,r12
 	adds r17=PT(AR_BSPSTORE)+16,r12
 	adds r18=TI_FLAGS+IA64_TASK_SIZE,r13
@@ -776,26 +784,31 @@ GLOBAL_ENTRY(ia64_leave_kernel)
 	PT_REGS_UNWIND_INFO(0)
 	/*
 	 * work.need_resched etc. mustn't get changed by this CPU before it returns to
-	 * user- or fsys-mode, hence we disable interrupts early on:
+	 * user- or fsys-mode, hence we disable interrupts early on.
+	 *
+	 * p6 controls whether current_thread_info()->flags needs to be check for
+	 * extra work.  We always check for extra work when returning to user-level.
+	 * With CONFIG_PREEMPT, we also check for extra work when the preempt_count
+	 * is 0.  After extra work processing has been completed, execution
+	 * resumes at .work_processed_syscall with p6 set to 1 if the extra-work-check
+	 * needs to be redone.
 	 */
 #ifdef CONFIG_PREEMPT
 	rsm psr.i				// disable interrupts
-#else
-(pUStk)	rsm psr.i
-#endif
 	cmp.eq p0,pLvSys=r0,r0			// pLvSys=0: leave from kernel
-(pUStk)	cmp.eq.unc p6,p0=r0,r0			// p6 <- pUStk
-	;;
-.work_processed_kernel:
-#ifdef CONFIG_PREEMPT
-	adds r20=TI_PRE_COUNT+IA64_TASK_SIZE,r13
+(pKStk)	adds r20=TI_PRE_COUNT+IA64_TASK_SIZE,r13
 	;;
 	.pred.rel.mutex pUStk,pKStk
 (pKStk)	ld4 r21=[r20]			// r21 <- preempt_count
 (pUStk)	mov r21=0			// r21 <- 0
 	;;
-(p6)	cmp.eq.unc p6,p0=r21,r0		// p6 <- p6 && (r21 == 0)
-#endif /* CONFIG_PREEMPT */
+	cmp.eq p6,p0=r21,r0		// p6 <- pUStk || (preempt_count == 0)
+#else
+(pUStk)	rsm psr.i
+	cmp.eq p0,pLvSys=r0,r0		// pLvSys=0: leave from kernel
+(pUStk)	cmp.eq.unc p6,p0=r0,r0		// p6 <- pUStk
+#endif
+.work_processed_kernel:
 	adds r17=TI_FLAGS+IA64_TASK_SIZE,r13
 	;;
 (p6)	ld4 r31=[r17]				// load current_thread_info()->flags
@@ -1065,7 +1078,7 @@ skip_rbs_switch:
 	br.cond.sptk.many .work_processed_kernel	// re-check
 
 .notify:
-	br.call.spnt.many rp=notify_resume_user
+(pUStk)	br.call.spnt.many rp=notify_resume_user
 .ret10:	cmp.ne p6,p0=r0,r0				// p6 <- 0
 (pLvSys)br.cond.sptk.many .work_processed_syscall	// don't re-check
 	br.cond.sptk.many .work_processed_kernel	// don't re-check

arch/ia64/kernel/fsys.S

@@ -165,7 +165,6 @@ ENTRY(fsys_gettimeofday)
 	add r9=TI_FLAGS+IA64_TASK_SIZE,r16
 	addl r3=THIS_CPU(cpu_info),r0
 
-	mov.m r31=ar.itc		// put time stamp into r31 (ITC) == now		(35 cyc)
 #ifdef CONFIG_SMP
 	movl r10=__per_cpu_offset
 	movl r2=sal_platform_features
@@ -240,12 +239,13 @@ EX(.fail_efault, probe.w.fault r10, 3)		// this must come _after_ NaT-check
 	;;
 
 	ldf8 f8=[r21]			// f8 now contains itm_next
+	mov.m r31=ar.itc		// put time stamp into r31 (ITC) == now
 	sub r28=r29, r28, 1		// r28 now contains "-(lost + 1)"
-	tbit.nz p9, p10=r23, 0		// p9 <- is_odd(r23), p10 <- is_even(r23)
 	;;
 
 	ld8 r2=[r19]			// r2 = sec = xtime.tv_sec
 	ld8 r29=[r20]			// r29 = nsec = xtime.tv_nsec
+	tbit.nz p9, p10=r23, 0		// p9 <- is_odd(r23), p10 <- is_even(r23)
 
 	setf.sig f6=r28			// f6 <- -(lost + 1)				(6 cyc)
 	;;
@@ -260,7 +260,6 @@ EX(.fail_efault, probe.w.fault r10, 3)		// this must come _after_ NaT-check
 	nop 0
 	;;
 
-	mov r31=ar.itc			// re-read ITC in case we .retry		(35 cyc)
 	xma.l f8=f11, f8, f12	// f8 (elapsed_cycles) <- (-1*last_tick + now) = (now - last_tick)
 	nop 0
 	;;

arch/ia64/kernel/ivt.S

@@ -752,7 +752,9 @@ ENTRY(break_fault)
 	;;
 	ld4 r2=[r2]				// r2 = current_thread_info()->flags
 	;;
-	tbit.z p8,p0=r2,TIF_SYSCALL_TRACE
+	and r2=_TIF_SYSCALL_TRACEAUDIT,r2	// mask trace or audit
+	;;
+	cmp.eq p8,p0=r2,r0
 	mov b6=r20
 	;;
 (p8)	br.call.sptk.many b6=b6			// ignore this return addr
@@ -1573,10 +1575,11 @@ ENTRY(dispatch_to_ia32_handler)
 	ld4 r2=[r2]		// r2 = current_thread_info()->flags
 	;;
 	ld8 r16=[r16]
-	tbit.z p8,p0=r2,TIF_SYSCALL_TRACE
+	and r2=_TIF_SYSCALL_TRACEAUDIT,r2	// mask trace or audit
 	;;
 	mov b6=r16
 	movl r15=ia32_ret_from_syscall
+	cmp.eq p8,p0=r2,r0
 	;;
 	mov rp=r15
 (p8)	br.call.sptk.many b6=b6

arch/ia64/kernel/machvec.c

@@ -43,12 +43,6 @@ machvec_init (const char *name)
 
 #endif /* CONFIG_IA64_GENERIC */
 
-void
-machvec_noop (void)
-{
-}
-EXPORT_SYMBOL(machvec_noop);
-
 void
 machvec_setup (char **arg)
 {

arch/ia64/kernel/mca.c

@@ -247,7 +247,9 @@ ia64_mca_log_sal_error_record(int sal_info_type)
 	u8 *buffer;
 	u64 size;
 	int irq_safe = sal_info_type != SAL_INFO_TYPE_MCA && sal_info_type != SAL_INFO_TYPE_INIT;
+#ifdef IA64_MCA_DEBUG_INFO
 	static const char * const rec_name[] = { "MCA", "INIT", "CMC", "CPE" };
+#endif
 
 	size = ia64_log_get(sal_info_type, &buffer, irq_safe);
 	if (!size)
@@ -596,7 +598,7 @@ ia64_mca_cmc_vector_disable (void *dummy)
 	cmcv = (cmcv_reg_t)ia64_getreg(_IA64_REG_CR_CMCV);
 
 	cmcv.cmcv_mask = 1; /* Mask/disable interrupt */
-	ia64_setreg(_IA64_REG_CR_CMCV, cmcv.cmcv_regval)
+	ia64_setreg(_IA64_REG_CR_CMCV, cmcv.cmcv_regval);
 
 	IA64_MCA_DEBUG("%s: CPU %d corrected "
 		       "machine check vector %#x disabled.\n",
@@ -623,7 +625,7 @@ ia64_mca_cmc_vector_enable (void *dummy)
 	cmcv = (cmcv_reg_t)ia64_getreg(_IA64_REG_CR_CMCV);
 
 	cmcv.cmcv_mask = 0; /* Unmask/enable interrupt */
-	ia64_setreg(_IA64_REG_CR_CMCV, cmcv.cmcv_regval)
+	ia64_setreg(_IA64_REG_CR_CMCV, cmcv.cmcv_regval);
 
 	IA64_MCA_DEBUG("%s: CPU %d corrected "
 		       "machine check vector %#x enabled.\n",

arch/ia64/kernel/module.c

@@ -656,26 +656,18 @@ do_reloc (struct module *mod, uint8_t r_type, Elf64_Sym *sym, uint64_t addend,
 	      case RV_PCREL:
 		switch (r_type) {
 		      case R_IA64_PCREL21B:
-			if (in_init(mod, val)) {
-				/* Calls to init code from core are bad news */
-				if (in_core(mod, (uint64_t)location)) {
-					printk(KERN_ERR "%s: init symbol 0x%lx used in module code at %p\n",
-						mod->name, val, location);
-					return -ENOEXEC;
-				}
-			} else if (in_core(mod, val)) {
+			if ((in_init(mod, val) && in_core(mod, (uint64_t)location)) ||
+			    (in_core(mod, val) && in_init(mod, (uint64_t)location))) {
 				/*
 				 * Init section may have been allocated far away from core,
 				 * if the branch won't reach, then allocate a plt for it.
 				 */
-				if (in_init(mod, (uint64_t)location)) {
-					uint64_t delta = ((int64_t)val - (int64_t)location) / 16;
-					if (delta + (1 << 20) >= (1 << 21)) {
-						val = get_fdesc(mod, val, &ok);
-						val = get_plt(mod, location, val, &ok);
-					}
+				uint64_t delta = ((int64_t)val - (int64_t)location) / 16;
+				if (delta + (1 << 20) >= (1 << 21)) {
+					val = get_fdesc(mod, val, &ok);
+					val = get_plt(mod, location, val, &ok);
 				}
-			} else
+			} else if (!is_internal(mod, val))
 				val = get_plt(mod, location, val, &ok);
 			/* FALL THROUGH */
 		      default:

arch/ia64/kernel/pal.S

@@ -54,7 +54,7 @@ END(ia64_pal_default_handler)
  *
  */
 GLOBAL_ENTRY(ia64_pal_call_static)
-	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(6)
+	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(5)
 	alloc loc1 = ar.pfs,5,5,0,0
 	movl loc2 = pal_entry_point
 1:	{
@@ -100,7 +100,7 @@ END(ia64_pal_call_static)
  * 	in2 - in3   Remaning PAL arguments
  */
 GLOBAL_ENTRY(ia64_pal_call_stacked)
-	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(5)
+	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(4)
 	alloc loc1 = ar.pfs,4,4,4,0
 	movl loc2 = pal_entry_point
 
@@ -147,7 +147,7 @@ END(ia64_pal_call_stacked)
 
 
 GLOBAL_ENTRY(ia64_pal_call_phys_static)
-	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(6)
+	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(4)
 	alloc loc1 = ar.pfs,4,7,0,0
 	movl loc2 = pal_entry_point
 1:	{

arch/ia64/kernel/perfmon.c

@@ -4702,21 +4702,22 @@ static int
 pfm_check_task_state(pfm_context_t *ctx, int cmd, unsigned long flags)
 {
 	struct task_struct *task;
-	int state;
+	int state, old_state;
 
+recheck:
 	state = ctx->ctx_state;
+	task  = ctx->ctx_task;
 
-	task = PFM_CTX_TASK(ctx);
 	if (task == NULL) {
 		DPRINT(("context %d no task, state=%d\n", ctx->ctx_fd, state));
 		return 0;
 	}
 
 	DPRINT(("context %d state=%d [%d] task_state=%ld must_stop=%d\n",
-				ctx->ctx_fd,
-				state,
-				task->pid,
-				task->state, PFM_CMD_STOPPED(cmd)));
+		ctx->ctx_fd,
+		state,
+		task->pid,
+		task->state, PFM_CMD_STOPPED(cmd)));
 
 	/*
 	 * self-monitoring always ok.
@@ -4728,31 +4729,61 @@ pfm_check_task_state(pfm_context_t *ctx, int cmd, unsigned long flags)
 	if (task == current || ctx->ctx_fl_system) return 0;
 
 	/*
-	 * context is UNLOADED, MASKED we are safe to go
+	 * no command can operate on a zombie context
 	 */
-	if (state != PFM_CTX_LOADED) return 0;
+	if (state == PFM_CTX_ZOMBIE) {
+		DPRINT(("cmd %d state zombie cannot operate on context\n", cmd));
+		return -EINVAL;
+	}
 
-	if (state == PFM_CTX_ZOMBIE) return -EINVAL;
+	/*
+	 * if context is UNLOADED, MASKED we are safe to go
+	 */
+	if (state != PFM_CTX_LOADED) return 0;
 
 	/*
-	 * context is loaded, we must make sure the task is stopped
+	 * context is LOADED, we must make sure the task is stopped
 	 * We could lift this restriction for UP but it would mean that
 	 * the user has no guarantee the task would not run between
 	 * two successive calls to perfmonctl(). That's probably OK.
 	 * If this user wants to ensure the task does not run, then
 	 * the task must be stopped.
 	 */
-	if (PFM_CMD_STOPPED(cmd) && task->state != TASK_STOPPED) {
-		DPRINT(("[%d] task not in stopped state\n", task->pid));
-		return -EBUSY;
-	}
+	if (PFM_CMD_STOPPED(cmd)) {
+		if (task->state != TASK_STOPPED) {
+			DPRINT(("[%d] task not in stopped state\n", task->pid));
+			return -EBUSY;
+		}
+		/*
+		 * task is now stopped, wait for ctxsw out
+		 *
+		 * This is an interesting point in the code.
+		 * We need to unprotect the context because
+		 * the pfm_save_regs() routines needs to grab
+		 * the same lock. There are danger in doing
+		 * this because it leaves a window open for
+		 * another task to get access to the context
+		 * and possibly change its state. The one thing
+		 * that is not possible is for the context to disappear
+		 * because we are protected by the VFS layer, i.e.,
+		 * get_fd()/put_fd().
+		 */
+		old_state = state;
 
-	UNPROTECT_CTX(ctx, flags);
+		UNPROTECT_CTX(ctx, flags);
 
-	wait_task_inactive(task);
+		wait_task_inactive(task);
 
-	PROTECT_CTX(ctx, flags);
+		PROTECT_CTX(ctx, flags);
 
+		/*
+		 * we must recheck to verify if state has changed
+		 */
+		if (ctx->ctx_state != old_state) {
+			DPRINT(("old_state=%d new_state=%d\n", old_state, ctx->ctx_state));
+			goto recheck;
+		}
+	}
 	return 0;
 }
 

arch/ia64/kernel/ptrace.c

@@ -1447,9 +1447,8 @@ sys_ptrace (long request, pid_t pid, unsigned long addr, unsigned long data,
 	return ret;
 }
 
-/* "asmlinkage" so the input arguments are preserved... */
 
-asmlinkage void
+void
 syscall_trace (void)
 {
 	if (!test_thread_flag(TIF_SYSCALL_TRACE))
@@ -1472,3 +1471,38 @@ syscall_trace (void)
 		current->exit_code = 0;
 	}
 }
+
+/* "asmlinkage" so the input arguments are preserved... */
+
+asmlinkage void
+syscall_trace_enter (long arg0, long arg1, long arg2, long arg3,
+		     long arg4, long arg5, long arg6, long arg7, long stack)
+{
+	struct pt_regs *regs = (struct pt_regs *) &stack;
+	long syscall;
+
+	if (unlikely(current->audit_context)) {
+		if (IS_IA32_PROCESS(regs))
+			syscall = regs->r1;
+		else
+			syscall = regs->r15;
+
+		audit_syscall_entry(current, syscall, arg0, arg1, arg2, arg3);
+	}
+
+	if (test_thread_flag(TIF_SYSCALL_TRACE) && (current->ptrace & PT_PTRACED))
+		syscall_trace();
+}
+
+/* "asmlinkage" so the input arguments are preserved... */
+
+asmlinkage void
+syscall_trace_leave (long arg0, long arg1, long arg2, long arg3,
+		     long arg4, long arg5, long arg6, long arg7, long stack)
+{
+	if (unlikely(current->audit_context))
+		audit_syscall_exit(current, ((struct pt_regs *) &stack)->r8);
+
+	if (test_thread_flag(TIF_SYSCALL_TRACE) && (current->ptrace & PT_PTRACED))
+		syscall_trace();
+}

arch/ia64/kernel/sal.c

@@ -188,6 +188,27 @@ sal_desc_ap_wakeup (void *p)
 		break;
 	}
 }
+
+static void __init
+chk_nointroute_opt(void)
+{
+	char *cp;
+	extern char saved_command_line[];
+
+	for (cp = saved_command_line; *cp; ) {
+		if (memcmp(cp, "nointroute", 10) == 0) {
+			no_int_routing = 1;
+			printk ("no_int_routing on\n");
+			break;
+		} else {
+			while (*cp != ' ' && *cp)
+				++cp;
+			while (*cp == ' ')
+				++cp;
+		}
+	}
+}
+
 #else
 static void __init sal_desc_ap_wakeup(void *p) { }
 #endif
@@ -207,6 +228,9 @@ ia64_sal_init (struct ia64_sal_systab *systab)
 		printk(KERN_ERR "bad signature in system table!");
 
 	check_versions(systab);
+#ifdef CONFIG_SMP
+	chk_nointroute_opt();
+#endif
 
 	/* revisions are coded in BCD, so %x does the job for us */
 	printk(KERN_INFO "SAL %x.%x: %.32s %.32s%sversion %x.%x\n",

arch/ia64/sn/kernel/irq.c

@@ -72,7 +72,7 @@ sn_enable_irq(unsigned int irq)
 {
 }
 
-static inline void move_irq(int irq)
+static inline void sn_move_irq(int irq)
 {
 	/* note - we hold desc->lock */
 	cpumask_t tmp;
@@ -110,7 +110,7 @@ sn_ack_irq(unsigned int irq)
 	}
 	HUB_S((unsigned long *)GLOBAL_MMR_ADDR(nasid, SH_EVENT_OCCURRED_ALIAS), mask );
 	__set_bit(irq, (volatile void *)pda->sn_in_service_ivecs);
-	move_irq(irq);
+	sn_move_irq(irq);
 }
 
 static void

arch/ia64/sn/kernel/sn2/sn2_smp.c

@@ -27,6 +27,7 @@
 #include <asm/delay.h>
 #include <asm/io.h>
 #include <asm/smp.h>
+#include <asm/tlb.h>
 #include <asm/numa.h>
 #include <asm/bitops.h>
 #include <asm/hw_irq.h>
@@ -60,6 +61,13 @@ wait_piowc(void)
 }
 
 
+void
+sn_tlb_migrate_finish(struct mm_struct *mm)
+{
+	if (mm == current->mm)
+		flush_tlb_mm(mm);
+}
+
 
 /**
  * sn2_global_tlb_purge - globally purge translation cache of virtual address range
@@ -114,6 +122,13 @@ sn2_global_tlb_purge (unsigned long start, unsigned long end, unsigned long nbit
 		return;
 	}
 
+	if (atomic_read(&mm->mm_users) == 1) {
+		flush_tlb_mm(mm);
+		preempt_enable();
+		return;
+	}
+
+
 	nix = 0;
 	for (cnode=find_first_bit(&nodes_flushed, NR_NODES); cnode < NR_NODES; 
 			cnode=find_next_bit(&nodes_flushed, NR_NODES, ++cnode))

include/asm-generic/tlb.h

@@ -147,4 +147,6 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 		__pmd_free_tlb(tlb, pmdp);			\
 	} while (0)
 
+#define tlb_migrate_finish(mm) do {} while (0)
+
 #endif /* _ASM_GENERIC__TLB_H */

include/asm-ia64/atomic.h

@@ -56,7 +56,7 @@ ia64_atomic64_add (__s64 i, atomic64_t *v)
 		CMPXCHG_BUGCHECK(v);
 		old = atomic_read(v);
 		new = old + i;
-	} while (ia64_cmpxchg(acq, v, old, new, sizeof(atomic_t)) != old);
+	} while (ia64_cmpxchg(acq, v, old, new, sizeof(atomic64_t)) != old);
 	return new;
 }
 
@@ -84,7 +84,7 @@ ia64_atomic64_sub (__s64 i, atomic64_t *v)
 		CMPXCHG_BUGCHECK(v);
 		old = atomic_read(v);
 		new = old - i;
-	} while (ia64_cmpxchg(acq, v, old, new, sizeof(atomic_t)) != old);
+	} while (ia64_cmpxchg(acq, v, old, new, sizeof(atomic64_t)) != old);
 	return new;
 }
 

include/asm-ia64/dma-mapping.h

@@ -6,6 +6,8 @@
  *	David Mosberger-Tang <davidm@hpl.hp.com>
  */
 
+#include <asm/machvec.h>
+
 #define dma_alloc_coherent	platform_dma_alloc_coherent
 #define dma_alloc_noncoherent	platform_dma_alloc_coherent	/* coherent mem. is cheap */
 #define dma_free_coherent	platform_dma_free_coherent

include/asm-ia64/machvec.h

@@ -19,6 +19,7 @@ struct pt_regs;
 struct scatterlist;
 struct irq_desc;
 struct page;
+struct mm_struct;
 
 typedef void ia64_mv_setup_t (char **);
 typedef void ia64_mv_cpu_init_t (void);
@@ -26,6 +27,7 @@ typedef void ia64_mv_irq_init_t (void);
 typedef void ia64_mv_send_ipi_t (int, int, int, int);
 typedef void ia64_mv_timer_interrupt_t (int, void *, struct pt_regs *);
 typedef void ia64_mv_global_tlb_purge_t (unsigned long, unsigned long, unsigned long);
+typedef void ia64_mv_tlb_migrate_finish_t (struct mm_struct *);
 typedef struct irq_desc *ia64_mv_irq_desc (unsigned int);
 typedef u8 ia64_mv_irq_to_vector (u8);
 typedef unsigned int ia64_mv_local_vector_to_irq (u8 vector);
@@ -69,11 +71,21 @@ typedef unsigned short ia64_mv_readw_relaxed_t (void *);
 typedef unsigned int ia64_mv_readl_relaxed_t (void *);
 typedef unsigned long ia64_mv_readq_relaxed_t (void *);
 
-extern void machvec_noop (void);
+static inline void
+machvec_noop (void)
+{
+}
+
+static inline void
+machvec_noop_mm (struct mm_struct *mm)
+{
+}
+
 extern void machvec_setup (char **);
 extern void machvec_timer_interrupt (int, void *, struct pt_regs *);
 extern void machvec_dma_sync_single (struct device *, dma_addr_t, size_t, int);
 extern void machvec_dma_sync_sg (struct device *, struct scatterlist *, int, int);
+extern void machvec_tlb_migrate_finish (struct mm_struct *);
 
 # if defined (CONFIG_IA64_HP_SIM)
 #  include <asm/machvec_hpsim.h>
@@ -95,6 +107,7 @@ extern void machvec_dma_sync_sg (struct device *, struct scatterlist *, int, int
 #  define platform_send_ipi	ia64_mv.send_ipi
 #  define platform_timer_interrupt	ia64_mv.timer_interrupt
 #  define platform_global_tlb_purge	ia64_mv.global_tlb_purge
+#  define platform_tlb_migrate_finish	ia64_mv.tlb_migrate_finish
 #  define platform_dma_init		ia64_mv.dma_init
 #  define platform_dma_alloc_coherent	ia64_mv.dma_alloc_coherent
 #  define platform_dma_free_coherent	ia64_mv.dma_free_coherent
@@ -140,6 +153,7 @@ struct ia64_machine_vector {
 	ia64_mv_send_ipi_t *send_ipi;
 	ia64_mv_timer_interrupt_t *timer_interrupt;
 	ia64_mv_global_tlb_purge_t *global_tlb_purge;
+	ia64_mv_tlb_migrate_finish_t *tlb_migrate_finish;
 	ia64_mv_dma_init *dma_init;
 	ia64_mv_dma_alloc_coherent *dma_alloc_coherent;
 	ia64_mv_dma_free_coherent *dma_free_coherent;
@@ -181,6 +195,7 @@ struct ia64_machine_vector {
 	platform_send_ipi,			\
 	platform_timer_interrupt,		\
 	platform_global_tlb_purge,		\
+	platform_tlb_migrate_finish,		\
 	platform_dma_init,			\
 	platform_dma_alloc_coherent,		\
 	platform_dma_free_coherent,		\
@@ -260,6 +275,9 @@ extern ia64_mv_dma_supported		swiotlb_dma_supported;
 #ifndef platform_global_tlb_purge
 # define platform_global_tlb_purge	ia64_global_tlb_purge /* default to architected version */
 #endif
+#ifndef platform_tlb_migrate_finish
+# define platform_tlb_migrate_finish	machvec_noop_mm
+#endif
 #ifndef platform_dma_init
 # define platform_dma_init		swiotlb_init
 #endif

include/asm-ia64/machvec_sn2.h

@@ -39,6 +39,7 @@ extern ia64_mv_irq_init_t sn_irq_init;
 extern ia64_mv_send_ipi_t sn2_send_IPI;
 extern ia64_mv_timer_interrupt_t sn_timer_interrupt;
 extern ia64_mv_global_tlb_purge_t sn2_global_tlb_purge;
+extern ia64_mv_tlb_migrate_finish_t	sn_tlb_migrate_finish;
 extern ia64_mv_irq_desc sn_irq_desc;
 extern ia64_mv_irq_to_vector sn_irq_to_vector;
 extern ia64_mv_local_vector_to_irq sn_local_vector_to_irq;
@@ -83,6 +84,7 @@ extern ia64_mv_dma_supported		sn_dma_supported;
 #define platform_send_ipi		sn2_send_IPI
 #define platform_timer_interrupt	sn_timer_interrupt
 #define platform_global_tlb_purge       sn2_global_tlb_purge
+#define platform_tlb_migrate_finish	sn_tlb_migrate_finish
 #define platform_pci_fixup		sn_pci_fixup
 #define platform_inb			__sn_inb
 #define platform_inw			__sn_inw

include/asm-ia64/thread_info.h

@@ -73,12 +73,15 @@ struct thread_info {
 #define TIF_SIGPENDING		1	/* signal pending */
 #define TIF_NEED_RESCHED	2	/* rescheduling necessary */
 #define TIF_SYSCALL_TRACE	3	/* syscall trace active */
+#define TIF_SYSCALL_AUDIT	4	/* syscall auditing active */
 #define TIF_POLLING_NRFLAG	16	/* true if poll_idle() is polling TIF_NEED_RESCHED */
 
 #define TIF_WORK_MASK		0x7	/* like TIF_ALLWORK_BITS but sans TIF_SYSCALL_TRACE */
-#define TIF_ALLWORK_MASK	0xf	/* bits 0..3 are "work to do on user-return" bits */
+#define TIF_ALLWORK_MASK	0x1f	/* bits 0..4 are "work to do on user-return" bits */
 
 #define _TIF_SYSCALL_TRACE	(1 << TIF_SYSCALL_TRACE)
+#define _TIF_SYSCALL_AUDIT	(1 << TIF_SYSCALL_AUDIT)
+#define _TIF_SYSCALL_TRACEAUDIT	(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT)
 #define _TIF_NOTIFY_RESUME	(1 << TIF_NOTIFY_RESUME)
 #define _TIF_SIGPENDING		(1 << TIF_SIGPENDING)
 #define _TIF_NEED_RESCHED	(1 << TIF_NEED_RESCHED)

include/asm-ia64/tlb.h

@@ -44,6 +44,7 @@
 #include <asm/pgalloc.h>
 #include <asm/processor.h>
 #include <asm/tlbflush.h>
+#include <asm/machvec.h>
 
 #ifdef CONFIG_SMP
 # define FREE_PTE_NR		2048
@@ -211,6 +212,8 @@ __tlb_remove_tlb_entry (struct mmu_gather *tlb, pte_t *ptep, unsigned long addre
 	tlb->end_addr = address + PAGE_SIZE;
 }
 
+#define tlb_migrate_finish(mm)	platform_tlb_migrate_finish(mm)
+
 #define tlb_start_vma(tlb, vma)			do { } while (0)
 #define tlb_end_vma(tlb, vma)			do { } while (0)
 

init/Kconfig

@@ -162,7 +162,7 @@ config AUDIT
 
 config AUDITSYSCALL
 	bool "Enable system-call auditing support"
-	depends on AUDIT && (X86 || PPC64 || ARCH_S390)
+	depends on AUDIT && (X86 || PPC64 || ARCH_S390 || IA64)
 	default y if SECURITY_SELINUX
 	default n
 	help

kernel/sched.c

@@ -40,6 +40,7 @@
 #include <linux/cpu.h>
 #include <linux/percpu.h>
 #include <linux/kthread.h>
+#include <asm/tlb.h>
 
 #include <asm/unistd.h>
 
@@ -3349,6 +3350,7 @@ int set_cpus_allowed(task_t *p, cpumask_t new_mask)
 		task_rq_unlock(rq, &flags);
 		wake_up_process(rq->migration_thread);
 		wait_for_completion(&req.done);
+		tlb_migrate_finish(p->mm);
 		return 0;
 	}
 out: