11 files changed, 157 insertions, 130 deletions

diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index 29a7940f284f..1d035082e78e 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -430,6 +430,13 @@ config CMM_IUCV
           Select this option to enable the special message interface to
           the cooperative memory management.
 
+config PAGE_STATES
+        bool "Unused page notification"
+        help
+          This enables the notification of unused pages to the
+          hypervisor. The ESSA instruction is used to do the states
+          changes between a page that has content and the unused state.
+
 config VIRT_TIMER
         bool "Virtual CPU timer support"
         help
diff --git a/arch/s390/kernel/compat_wrapper.S b/arch/s390/kernel/compat_wrapper.S
index 743d54f0b8db..d003a6e16afb 100644
--- a/arch/s390/kernel/compat_wrapper.S
+++ b/arch/s390/kernel/compat_wrapper.S
@@ -121,7 +121,7 @@ sys32_ptrace_wrapper:
         lgfr    %r3,%r3                 # long
         llgtr   %r4,%r4                 # long
         llgfr   %r5,%r5                 # long
-        jg      sys_ptrace              # branch to system call
+        jg      compat_sys_ptrace       # branch to system call
 
         .globl  sys32_alarm_wrapper
 sys32_alarm_wrapper:
diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index bdbb3bcd78a5..708cf9cf9a35 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -279,8 +279,6 @@ sysc_do_restart:
         st      %r2,SP_R2(%r15)   # store return value (change R2 on stack)
 
 sysc_return:
-        tm      SP_PSW+1(%r15),0x01     # returning to user ?
-        bno     BASED(sysc_restore)
         tm      __TI_flags+3(%r9),_TIF_WORK_SVC
         bnz     BASED(sysc_work)  # there is work to do (signals etc.)
 sysc_restore:
@@ -312,6 +310,8 @@ sysc_work_loop:
 # One of the work bits is on. Find out which one.
 #
 sysc_work:
+        tm      SP_PSW+1(%r15),0x01     # returning to user ?
+        bno     BASED(sysc_restore)
         tm      __TI_flags+3(%r9),_TIF_MCCK_PENDING
         bo      BASED(sysc_mcck_pending)
         tm      __TI_flags+3(%r9),_TIF_NEED_RESCHED
@@ -602,12 +602,6 @@ io_no_vtime:
         la      %r2,SP_PTREGS(%r15)     # address of register-save area
         basr    %r14,%r1                # branch to standard irq handler
 io_return:
-        tm      SP_PSW+1(%r15),0x01     # returning to user ?
-#ifdef CONFIG_PREEMPT
-        bno     BASED(io_preempt)       # no -> check for preemptive scheduling
-#else
-        bno     BASED(io_restore)       # no-> skip resched & signal
-#endif
         tm      __TI_flags+3(%r9),_TIF_WORK_INT
         bnz     BASED(io_work)          # there is work to do (signals etc.)
 io_restore:
@@ -629,10 +623,18 @@ io_restore_trace_psw:
         .long   0, io_restore_trace + 0x80000000
 #endif
 
-#ifdef CONFIG_PREEMPT
-io_preempt:
+#
+# switch to kernel stack, then check the TIF bits
+#
+io_work:
+        tm      SP_PSW+1(%r15),0x01     # returning to user ?
+#ifndef CONFIG_PREEMPT
+        bno     BASED(io_restore)       # no-> skip resched & signal
+#else
+        bnz     BASED(io_work_user)     # no -> check for preemptive scheduling
+        # check for preemptive scheduling
         icm     %r0,15,__TI_precount(%r9)
-        bnz     BASED(io_restore)
+        bnz     BASED(io_restore)       # preemption disabled
         l       %r1,SP_R15(%r15)
         s       %r1,BASED(.Lc_spsize)
         mvc     SP_PTREGS(__PT_SIZE,%r1),SP_PTREGS(%r15)
@@ -646,10 +648,7 @@ io_resume_loop:
         br      %r1                     # call schedule
 #endif
 
-#
-# switch to kernel stack, then check the TIF bits
-#
-io_work:
+io_work_user:
         l       %r1,__LC_KERNEL_STACK
         s       %r1,BASED(.Lc_spsize)
         mvc     SP_PTREGS(__PT_SIZE,%r1),SP_PTREGS(%r15)
diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S
index 5a4a7bcd2bba..fee10177dbfc 100644
--- a/arch/s390/kernel/entry64.S
+++ b/arch/s390/kernel/entry64.S
@@ -271,8 +271,6 @@ sysc_noemu:
         stg     %r2,SP_R2(%r15) # store return value (change R2 on stack)
 
 sysc_return:
-        tm      SP_PSW+1(%r15),0x01     # returning to user ?
-        jno     sysc_restore
         tm      __TI_flags+7(%r9),_TIF_WORK_SVC
         jnz     sysc_work       # there is work to do (signals etc.)
 sysc_restore:
@@ -304,6 +302,8 @@ sysc_work_loop:
 # One of the work bits is on. Find out which one.
 #
 sysc_work:
+        tm      SP_PSW+1(%r15),0x01     # returning to user ?
+        jno     sysc_restore
         tm      __TI_flags+7(%r9),_TIF_MCCK_PENDING
         jo      sysc_mcck_pending
         tm      __TI_flags+7(%r9),_TIF_NEED_RESCHED
@@ -585,12 +585,6 @@ io_no_vtime:
         la      %r2,SP_PTREGS(%r15)     # address of register-save area
         brasl   %r14,do_IRQ             # call standard irq handler
 io_return:
-        tm      SP_PSW+1(%r15),0x01     # returning to user ?
-#ifdef CONFIG_PREEMPT
-        jno     io_preempt              # no -> check for preemptive scheduling
-#else
-        jno     io_restore              # no-> skip resched & signal
-#endif
         tm      __TI_flags+7(%r9),_TIF_WORK_INT
         jnz     io_work                 # there is work to do (signals etc.)
 io_restore:
@@ -612,10 +606,41 @@ io_restore_trace_psw:
         .quad   0, io_restore_trace
 #endif
 
-#ifdef CONFIG_PREEMPT
-io_preempt:
+#
+# There is work todo, we need to check if we return to userspace, then
+# check, if we are in SIE, if yes leave it
+#
+io_work:
+        tm      SP_PSW+1(%r15),0x01     # returning to user ?
+#ifndef CONFIG_PREEMPT
+#if defined(CONFIG_KVM) || defined(CONFIG_KVM_MODULE)
+        jnz     io_work_user            # yes -> no need to check for SIE
+        la      %r1, BASED(sie_opcode)  # we return to kernel here
+        lg      %r2, SP_PSW+8(%r15)
+        clc     0(2,%r1), 0(%r2)        # is current instruction = SIE?
+        jne     io_restore              # no-> return to kernel
+        lg      %r1, SP_PSW+8(%r15)     # yes-> add 4 bytes to leave SIE
+        aghi    %r1, 4
+        stg     %r1, SP_PSW+8(%r15)
+        j       io_restore              # return to kernel
+#else
+        jno     io_restore              # no-> skip resched & signal
+#endif
+#else
+        jnz     io_work_user            # yes -> do resched & signal
+#if defined(CONFIG_KVM) || defined(CONFIG_KVM_MODULE)
+        la      %r1, BASED(sie_opcode)
+        lg      %r2, SP_PSW+8(%r15)
+        clc     0(2,%r1), 0(%r2)        # is current instruction = SIE?
+        jne     0f                      # no -> leave PSW alone
+        lg      %r1, SP_PSW+8(%r15)     # yes-> add 4 bytes to leave SIE
+        aghi    %r1, 4
+        stg     %r1, SP_PSW+8(%r15)
+0:
+#endif
+        # check for preemptive scheduling
         icm     %r0,15,__TI_precount(%r9)
-        jnz     io_restore
+        jnz     io_restore              # preemption is disabled
         # switch to kernel stack
         lg      %r1,SP_R15(%r15)
         aghi    %r1,-SP_SIZE
@@ -629,10 +654,7 @@ io_resume_loop:
         jg      preempt_schedule_irq
 #endif
 
-#
-# switch to kernel stack, then check TIF bits
-#
-io_work:
+io_work_user:
         lg      %r1,__LC_KERNEL_STACK
         aghi    %r1,-SP_SIZE
         mvc     SP_PTREGS(__PT_SIZE,%r1),SP_PTREGS(%r15)
@@ -653,6 +675,11 @@ io_work_loop:
         j       io_restore
 io_work_done:
 
+#if defined(CONFIG_KVM) || defined(CONFIG_KVM_MODULE)
+sie_opcode:
+        .long 0xb2140000
+#endif
+
 #
 # _TIF_MCCK_PENDING is set, call handler
 #
diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c
index 7f4270163744..35827b9bd4d1 100644
--- a/arch/s390/kernel/ptrace.c
+++ b/arch/s390/kernel/ptrace.c
@@ -292,8 +292,7 @@ poke_user(struct task_struct *child, addr_t addr, addr_t data)
         return 0;
 }
 
-static int
-do_ptrace_normal(struct task_struct *child, long request, long addr, long data)
+long arch_ptrace(struct task_struct *child, long request, long addr, long data)
 {
         ptrace_area parea; 
         int copied, ret;
@@ -529,35 +528,19 @@ poke_user_emu31(struct task_struct *child, addr_t addr, addr_t data)
         return 0;
 }
 
-static int
-do_ptrace_emu31(struct task_struct *child, long request, long addr, long data)
+long compat_arch_ptrace(struct task_struct *child, compat_long_t request,
+                        compat_ulong_t caddr, compat_ulong_t cdata)
 {
-        unsigned int tmp;  /* 4 bytes !! */
+        unsigned long addr = caddr;
+        unsigned long data = cdata;
         ptrace_area_emu31 parea; 
         int copied, ret;
 
         switch (request) {
-        case PTRACE_PEEKTEXT:
-        case PTRACE_PEEKDATA:
-                /* read word at location addr. */
-                copied = access_process_vm(child, addr, &tmp, sizeof(tmp), 0);
-                if (copied != sizeof(tmp))
-                        return -EIO;
-                return put_user(tmp, (unsigned int __force __user *) data);
-
         case PTRACE_PEEKUSR:
                 /* read the word at location addr in the USER area. */
                 return peek_user_emu31(child, addr, data);
 
-        case PTRACE_POKETEXT:
-        case PTRACE_POKEDATA:
-                /* write the word at location addr. */
-                tmp = data;
-                copied = access_process_vm(child, addr, &tmp, sizeof(tmp), 1);
-                if (copied != sizeof(tmp))
-                        return -EIO;
-                return 0;
-
         case PTRACE_POKEUSR:
                 /* write the word at location addr in the USER area */
                 return poke_user_emu31(child, addr, data);
@@ -587,82 +570,11 @@ do_ptrace_emu31(struct task_struct *child, long request, long addr, long data)
                         copied += sizeof(unsigned int);
                 }
                 return 0;
-        case PTRACE_GETEVENTMSG:
-                return put_user((__u32) child->ptrace_message,
-                                (unsigned int __force __user *) data);
-        case PTRACE_GETSIGINFO:
-                if (child->last_siginfo == NULL)
-                        return -EINVAL;
-                return copy_siginfo_to_user32((compat_siginfo_t
-                                               __force __user *) data,
-                                              child->last_siginfo);
-        case PTRACE_SETSIGINFO:
-                if (child->last_siginfo == NULL)
-                        return -EINVAL;
-                return copy_siginfo_from_user32(child->last_siginfo,
-                                                (compat_siginfo_t
-                                                 __force __user *) data);
         }
-        return ptrace_request(child, request, addr, data);
+        return compat_ptrace_request(child, request, addr, data);
 }
 #endif
 
-long arch_ptrace(struct task_struct *child, long request, long addr, long data)
-{
-        switch (request) {
-        case PTRACE_SYSCALL:
-                /* continue and stop at next (return from) syscall */
-        case PTRACE_CONT:
-                /* restart after signal. */
-                if (!valid_signal(data))
-                        return -EIO;
-                if (request == PTRACE_SYSCALL)
-                        set_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
-                else
-                        clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
-                child->exit_code = data;
-                /* make sure the single step bit is not set. */
-                user_disable_single_step(child);
-                wake_up_process(child);
-                return 0;
-
-        case PTRACE_KILL:
-                /*
-                 * make the child exit.  Best I can do is send it a sigkill. 
-                 * perhaps it should be put in the status that it wants to 
-                 * exit.
-                 */
-                if (child->exit_state == EXIT_ZOMBIE) /* already dead */
-                        return 0;
-                child->exit_code = SIGKILL;
-                /* make sure the single step bit is not set. */
-                user_disable_single_step(child);
-                wake_up_process(child);
-                return 0;
-
-        case PTRACE_SINGLESTEP:
-                /* set the trap flag. */
-                if (!valid_signal(data))
-                        return -EIO;
-                clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
-                child->exit_code = data;
-                user_enable_single_step(child);
-                /* give it a chance to run. */
-                wake_up_process(child);
-                return 0;
-
-        /* Do requests that differ for 31/64 bit */
-        default:
-#ifdef CONFIG_COMPAT
-                if (test_thread_flag(TIF_31BIT))
-                        return do_ptrace_emu31(child, request, addr, data);
-#endif
-                return do_ptrace_normal(child, request, addr, data);
-        }
-        /* Not reached.  */
-        return -EIO;
-}
-
 asmlinkage void
 syscall_trace(struct pt_regs *regs, int entryexit)
 {
diff --git a/arch/s390/kvm/Kconfig b/arch/s390/kvm/Kconfig
index 1761b74d639b..e051cad1f1e0 100644
--- a/arch/s390/kvm/Kconfig
+++ b/arch/s390/kvm/Kconfig
@@ -22,7 +22,6 @@ config KVM
         select PREEMPT_NOTIFIERS
         select ANON_INODES
         select S390_SWITCH_AMODE
-        select PREEMPT
         ---help---
           Support hosting paravirtualized guest machines using the SIE
           virtualization capability on the mainframe. This should work
diff --git a/arch/s390/kvm/intercept.c b/arch/s390/kvm/intercept.c
index 349581a26103..47a0b642174c 100644
--- a/arch/s390/kvm/intercept.c
+++ b/arch/s390/kvm/intercept.c
@@ -105,6 +105,9 @@ static intercept_handler_t instruction_handlers[256] = {
 static int handle_noop(struct kvm_vcpu *vcpu)
 {
         switch (vcpu->arch.sie_block->icptcode) {
+        case 0x0:
+                vcpu->stat.exit_null++;
+                break;
         case 0x10:
                 vcpu->stat.exit_external_request++;
                 break;
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 98d1e73e01f1..0ac36a649eba 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -31,6 +31,7 @@
 
 struct kvm_stats_debugfs_item debugfs_entries[] = {
         { "userspace_handled", VCPU_STAT(exit_userspace) },
+        { "exit_null", VCPU_STAT(exit_null) },
         { "exit_validity", VCPU_STAT(exit_validity) },
         { "exit_stop_request", VCPU_STAT(exit_stop_request) },
         { "exit_external_request", VCPU_STAT(exit_external_request) },
@@ -221,10 +222,6 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
         vcpu->arch.guest_fpregs.fpc &= FPC_VALID_MASK;
         restore_fp_regs(&vcpu->arch.guest_fpregs);
         restore_access_regs(vcpu->arch.guest_acrs);
-
-        if (signal_pending(current))
-                atomic_set_mask(CPUSTAT_STOP_INT,
-                        &vcpu->arch.sie_block->cpuflags);
 }
 
 void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
diff --git a/arch/s390/mm/Makefile b/arch/s390/mm/Makefile
index fb988a48a754..2a7458134544 100644
--- a/arch/s390/mm/Makefile
+++ b/arch/s390/mm/Makefile
@@ -5,3 +5,4 @@
 obj-y    := init.o fault.o extmem.o mmap.o vmem.o pgtable.o
 obj-$(CONFIG_CMM) += cmm.o
 obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o
+obj-$(CONFIG_PAGE_STATES) += page-states.o
diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
index fa31de6ae97a..29f3a63806b9 100644
--- a/arch/s390/mm/init.c
+++ b/arch/s390/mm/init.c
@@ -126,6 +126,9 @@ void __init mem_init(void)
         /* clear the zero-page */
         memset(empty_zero_page, 0, PAGE_SIZE);
 
+        /* Setup guest page hinting */
+        cmma_init();
+
         /* this will put all low memory onto the freelists */
         totalram_pages += free_all_bootmem();
 
diff --git a/arch/s390/mm/page-states.c b/arch/s390/mm/page-states.c
new file mode 100644
index 000000000000..fc0ad73ffd90
--- /dev/null
+++ b/arch/s390/mm/page-states.c
@@ -0,0 +1,79 @@
+/*
+ * arch/s390/mm/page-states.c
+ *
+ * Copyright IBM Corp. 2008
+ *
+ * Guest page hinting for unused pages.
+ *
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/mm.h>
+#include <linux/init.h>
+
+#define ESSA_SET_STABLE         1
+#define ESSA_SET_UNUSED         2
+
+static int cmma_flag;
+
+static int __init cmma(char *str)
+{
+        char *parm;
+        parm = strstrip(str);
+        if (strcmp(parm, "yes") == 0 || strcmp(parm, "on") == 0) {
+                cmma_flag = 1;
+                return 1;
+        }
+        cmma_flag = 0;
+        if (strcmp(parm, "no") == 0 || strcmp(parm, "off") == 0)
+                return 1;
+        return 0;
+}
+
+__setup("cmma=", cmma);
+
+void __init cmma_init(void)
+{
+        register unsigned long tmp asm("0") = 0;
+        register int rc asm("1") = -EOPNOTSUPP;
+
+        if (!cmma_flag)
+                return;
+        asm volatile(
+                "       .insn rrf,0xb9ab0000,%1,%1,0,0\n"
+                "0:     la      %0,0\n"
+                "1:\n"
+                EX_TABLE(0b,1b)
+                : "+&d" (rc), "+&d" (tmp));
+        if (rc)
+                cmma_flag = 0;
+}
+
+void arch_free_page(struct page *page, int order)
+{
+        int i, rc;
+
+        if (!cmma_flag)
+                return;
+        for (i = 0; i < (1 << order); i++)
+                asm volatile(".insn rrf,0xb9ab0000,%0,%1,%2,0"
+                             : "=&d" (rc)
+                             : "a" ((page_to_pfn(page) + i) << PAGE_SHIFT),
+                               "i" (ESSA_SET_UNUSED));
+}
+
+void arch_alloc_page(struct page *page, int order)
+{
+        int i, rc;
+
+        if (!cmma_flag)
+                return;
+        for (i = 0; i < (1 << order); i++)
+                asm volatile(".insn rrf,0xb9ab0000,%0,%1,%2,0"
+                             : "=&d" (rc)
+                             : "a" ((page_to_pfn(page) + i) << PAGE_SHIFT),
+                               "i" (ESSA_SET_STABLE));
+}