summaryrefslogtreecommitdiff
path: root/arch/s390/mm
diff options
context:
space:
mode:
Diffstat (limited to 'arch/s390/mm')
-rw-r--r--arch/s390/mm/Makefile9
-rw-r--r--arch/s390/mm/cmm.c497
-rw-r--r--arch/s390/mm/extmem.c778
-rw-r--r--arch/s390/mm/fault.c664
-rw-r--r--arch/s390/mm/gup.c236
-rw-r--r--arch/s390/mm/hugetlbpage.c130
-rw-r--r--arch/s390/mm/init.c262
-rw-r--r--arch/s390/mm/maccess.c168
-rw-r--r--arch/s390/mm/mmap.c178
-rw-r--r--arch/s390/mm/page-states.c114
-rw-r--r--arch/s390/mm/pageattr.c62
-rw-r--r--arch/s390/mm/pgtable.c871
-rw-r--r--arch/s390/mm/vmem.c388
13 files changed, 4357 insertions, 0 deletions
diff --git a/arch/s390/mm/Makefile b/arch/s390/mm/Makefile
new file mode 100644
index 00000000..d98fe900
--- /dev/null
+++ b/arch/s390/mm/Makefile
@@ -0,0 +1,9 @@
+#
+# Makefile for the linux s390-specific parts of the memory manager.
+#
+
+obj-y := init.o fault.o extmem.o mmap.o vmem.o pgtable.o maccess.o \
+ page-states.o gup.o
+obj-$(CONFIG_CMM) += cmm.o
+obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o
+obj-$(CONFIG_DEBUG_SET_MODULE_RONX) += pageattr.o
diff --git a/arch/s390/mm/cmm.c b/arch/s390/mm/cmm.c
new file mode 100644
index 00000000..1f1dba9d
--- /dev/null
+++ b/arch/s390/mm/cmm.c
@@ -0,0 +1,497 @@
+/*
+ * Collaborative memory management interface.
+ *
+ * Copyright IBM Corp 2003,2010
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>,
+ *
+ */
+
+#include <linux/errno.h>
+#include <linux/fs.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/gfp.h>
+#include <linux/sched.h>
+#include <linux/sysctl.h>
+#include <linux/ctype.h>
+#include <linux/swap.h>
+#include <linux/kthread.h>
+#include <linux/oom.h>
+#include <linux/suspend.h>
+#include <linux/uaccess.h>
+
+#include <asm/pgalloc.h>
+#include <asm/diag.h>
+
+#ifdef CONFIG_CMM_IUCV
+static char *cmm_default_sender = "VMRMSVM";
+#endif
+static char *sender;
+module_param(sender, charp, 0400);
+MODULE_PARM_DESC(sender,
+ "Guest name that may send SMSG messages (default VMRMSVM)");
+
+#include "../../../drivers/s390/net/smsgiucv.h"
+
+#define CMM_NR_PAGES ((PAGE_SIZE / sizeof(unsigned long)) - 2)
+
+struct cmm_page_array {
+ struct cmm_page_array *next;
+ unsigned long index;
+ unsigned long pages[CMM_NR_PAGES];
+};
+
+static long cmm_pages;
+static long cmm_timed_pages;
+static volatile long cmm_pages_target;
+static volatile long cmm_timed_pages_target;
+static long cmm_timeout_pages;
+static long cmm_timeout_seconds;
+static int cmm_suspended;
+
+static struct cmm_page_array *cmm_page_list;
+static struct cmm_page_array *cmm_timed_page_list;
+static DEFINE_SPINLOCK(cmm_lock);
+
+static struct task_struct *cmm_thread_ptr;
+static DECLARE_WAIT_QUEUE_HEAD(cmm_thread_wait);
+static DEFINE_TIMER(cmm_timer, NULL, 0, 0);
+
+static void cmm_timer_fn(unsigned long);
+static void cmm_set_timer(void);
+
+static long cmm_alloc_pages(long nr, long *counter,
+ struct cmm_page_array **list)
+{
+ struct cmm_page_array *pa, *npa;
+ unsigned long addr;
+
+ while (nr) {
+ addr = __get_free_page(GFP_NOIO);
+ if (!addr)
+ break;
+ spin_lock(&cmm_lock);
+ pa = *list;
+ if (!pa || pa->index >= CMM_NR_PAGES) {
+ /* Need a new page for the page list. */
+ spin_unlock(&cmm_lock);
+ npa = (struct cmm_page_array *)
+ __get_free_page(GFP_NOIO);
+ if (!npa) {
+ free_page(addr);
+ break;
+ }
+ spin_lock(&cmm_lock);
+ pa = *list;
+ if (!pa || pa->index >= CMM_NR_PAGES) {
+ npa->next = pa;
+ npa->index = 0;
+ pa = npa;
+ *list = pa;
+ } else
+ free_page((unsigned long) npa);
+ }
+ diag10_range(addr >> PAGE_SHIFT, 1);
+ pa->pages[pa->index++] = addr;
+ (*counter)++;
+ spin_unlock(&cmm_lock);
+ nr--;
+ }
+ return nr;
+}
+
+static long cmm_free_pages(long nr, long *counter, struct cmm_page_array **list)
+{
+ struct cmm_page_array *pa;
+ unsigned long addr;
+
+ spin_lock(&cmm_lock);
+ pa = *list;
+ while (nr) {
+ if (!pa || pa->index <= 0)
+ break;
+ addr = pa->pages[--pa->index];
+ if (pa->index == 0) {
+ pa = pa->next;
+ free_page((unsigned long) *list);
+ *list = pa;
+ }
+ free_page(addr);
+ (*counter)--;
+ nr--;
+ }
+ spin_unlock(&cmm_lock);
+ return nr;
+}
+
+static int cmm_oom_notify(struct notifier_block *self,
+ unsigned long dummy, void *parm)
+{
+ unsigned long *freed = parm;
+ long nr = 256;
+
+ nr = cmm_free_pages(nr, &cmm_timed_pages, &cmm_timed_page_list);
+ if (nr > 0)
+ nr = cmm_free_pages(nr, &cmm_pages, &cmm_page_list);
+ cmm_pages_target = cmm_pages;
+ cmm_timed_pages_target = cmm_timed_pages;
+ *freed += 256 - nr;
+ return NOTIFY_OK;
+}
+
+static struct notifier_block cmm_oom_nb = {
+ .notifier_call = cmm_oom_notify,
+};
+
+static int cmm_thread(void *dummy)
+{
+ int rc;
+
+ while (1) {
+ rc = wait_event_interruptible(cmm_thread_wait,
+ (!cmm_suspended && (cmm_pages != cmm_pages_target ||
+ cmm_timed_pages != cmm_timed_pages_target)) ||
+ kthread_should_stop());
+ if (kthread_should_stop() || rc == -ERESTARTSYS) {
+ cmm_pages_target = cmm_pages;
+ cmm_timed_pages_target = cmm_timed_pages;
+ break;
+ }
+ if (cmm_pages_target > cmm_pages) {
+ if (cmm_alloc_pages(1, &cmm_pages, &cmm_page_list))
+ cmm_pages_target = cmm_pages;
+ } else if (cmm_pages_target < cmm_pages) {
+ cmm_free_pages(1, &cmm_pages, &cmm_page_list);
+ }
+ if (cmm_timed_pages_target > cmm_timed_pages) {
+ if (cmm_alloc_pages(1, &cmm_timed_pages,
+ &cmm_timed_page_list))
+ cmm_timed_pages_target = cmm_timed_pages;
+ } else if (cmm_timed_pages_target < cmm_timed_pages) {
+ cmm_free_pages(1, &cmm_timed_pages,
+ &cmm_timed_page_list);
+ }
+ if (cmm_timed_pages > 0 && !timer_pending(&cmm_timer))
+ cmm_set_timer();
+ }
+ return 0;
+}
+
+static void cmm_kick_thread(void)
+{
+ wake_up(&cmm_thread_wait);
+}
+
+static void cmm_set_timer(void)
+{
+ if (cmm_timed_pages_target <= 0 || cmm_timeout_seconds <= 0) {
+ if (timer_pending(&cmm_timer))
+ del_timer(&cmm_timer);
+ return;
+ }
+ if (timer_pending(&cmm_timer)) {
+ if (mod_timer(&cmm_timer, jiffies + cmm_timeout_seconds*HZ))
+ return;
+ }
+ cmm_timer.function = cmm_timer_fn;
+ cmm_timer.data = 0;
+ cmm_timer.expires = jiffies + cmm_timeout_seconds*HZ;
+ add_timer(&cmm_timer);
+}
+
+static void cmm_timer_fn(unsigned long ignored)
+{
+ long nr;
+
+ nr = cmm_timed_pages_target - cmm_timeout_pages;
+ if (nr < 0)
+ cmm_timed_pages_target = 0;
+ else
+ cmm_timed_pages_target = nr;
+ cmm_kick_thread();
+ cmm_set_timer();
+}
+
+static void cmm_set_pages(long nr)
+{
+ cmm_pages_target = nr;
+ cmm_kick_thread();
+}
+
+static long cmm_get_pages(void)
+{
+ return cmm_pages;
+}
+
+static void cmm_add_timed_pages(long nr)
+{
+ cmm_timed_pages_target += nr;
+ cmm_kick_thread();
+}
+
+static long cmm_get_timed_pages(void)
+{
+ return cmm_timed_pages;
+}
+
+static void cmm_set_timeout(long nr, long seconds)
+{
+ cmm_timeout_pages = nr;
+ cmm_timeout_seconds = seconds;
+ cmm_set_timer();
+}
+
+static int cmm_skip_blanks(char *cp, char **endp)
+{
+ char *str;
+
+ for (str = cp; *str == ' ' || *str == '\t'; str++)
+ ;
+ *endp = str;
+ return str != cp;
+}
+
+static struct ctl_table cmm_table[];
+
+static int cmm_pages_handler(ctl_table *ctl, int write, void __user *buffer,
+ size_t *lenp, loff_t *ppos)
+{
+ char buf[16], *p;
+ long nr;
+ int len;
+
+ if (!*lenp || (*ppos && !write)) {
+ *lenp = 0;
+ return 0;
+ }
+
+ if (write) {
+ len = *lenp;
+ if (copy_from_user(buf, buffer,
+ len > sizeof(buf) ? sizeof(buf) : len))
+ return -EFAULT;
+ buf[sizeof(buf) - 1] = '\0';
+ cmm_skip_blanks(buf, &p);
+ nr = simple_strtoul(p, &p, 0);
+ if (ctl == &cmm_table[0])
+ cmm_set_pages(nr);
+ else
+ cmm_add_timed_pages(nr);
+ } else {
+ if (ctl == &cmm_table[0])
+ nr = cmm_get_pages();
+ else
+ nr = cmm_get_timed_pages();
+ len = sprintf(buf, "%ld\n", nr);
+ if (len > *lenp)
+ len = *lenp;
+ if (copy_to_user(buffer, buf, len))
+ return -EFAULT;
+ }
+ *lenp = len;
+ *ppos += len;
+ return 0;
+}
+
+static int cmm_timeout_handler(ctl_table *ctl, int write, void __user *buffer,
+ size_t *lenp, loff_t *ppos)
+{
+ char buf[64], *p;
+ long nr, seconds;
+ int len;
+
+ if (!*lenp || (*ppos && !write)) {
+ *lenp = 0;
+ return 0;
+ }
+
+ if (write) {
+ len = *lenp;
+ if (copy_from_user(buf, buffer,
+ len > sizeof(buf) ? sizeof(buf) : len))
+ return -EFAULT;
+ buf[sizeof(buf) - 1] = '\0';
+ cmm_skip_blanks(buf, &p);
+ nr = simple_strtoul(p, &p, 0);
+ cmm_skip_blanks(p, &p);
+ seconds = simple_strtoul(p, &p, 0);
+ cmm_set_timeout(nr, seconds);
+ } else {
+ len = sprintf(buf, "%ld %ld\n",
+ cmm_timeout_pages, cmm_timeout_seconds);
+ if (len > *lenp)
+ len = *lenp;
+ if (copy_to_user(buffer, buf, len))
+ return -EFAULT;
+ }
+ *lenp = len;
+ *ppos += len;
+ return 0;
+}
+
+static struct ctl_table cmm_table[] = {
+ {
+ .procname = "cmm_pages",
+ .mode = 0644,
+ .proc_handler = cmm_pages_handler,
+ },
+ {
+ .procname = "cmm_timed_pages",
+ .mode = 0644,
+ .proc_handler = cmm_pages_handler,
+ },
+ {
+ .procname = "cmm_timeout",
+ .mode = 0644,
+ .proc_handler = cmm_timeout_handler,
+ },
+ { }
+};
+
+static struct ctl_table cmm_dir_table[] = {
+ {
+ .procname = "vm",
+ .maxlen = 0,
+ .mode = 0555,
+ .child = cmm_table,
+ },
+ { }
+};
+
+#ifdef CONFIG_CMM_IUCV
+#define SMSG_PREFIX "CMM"
+static void cmm_smsg_target(const char *from, char *msg)
+{
+ long nr, seconds;
+
+ if (strlen(sender) > 0 && strcmp(from, sender) != 0)
+ return;
+ if (!cmm_skip_blanks(msg + strlen(SMSG_PREFIX), &msg))
+ return;
+ if (strncmp(msg, "SHRINK", 6) == 0) {
+ if (!cmm_skip_blanks(msg + 6, &msg))
+ return;
+ nr = simple_strtoul(msg, &msg, 0);
+ cmm_skip_blanks(msg, &msg);
+ if (*msg == '\0')
+ cmm_set_pages(nr);
+ } else if (strncmp(msg, "RELEASE", 7) == 0) {
+ if (!cmm_skip_blanks(msg + 7, &msg))
+ return;
+ nr = simple_strtoul(msg, &msg, 0);
+ cmm_skip_blanks(msg, &msg);
+ if (*msg == '\0')
+ cmm_add_timed_pages(nr);
+ } else if (strncmp(msg, "REUSE", 5) == 0) {
+ if (!cmm_skip_blanks(msg + 5, &msg))
+ return;
+ nr = simple_strtoul(msg, &msg, 0);
+ if (!cmm_skip_blanks(msg, &msg))
+ return;
+ seconds = simple_strtoul(msg, &msg, 0);
+ cmm_skip_blanks(msg, &msg);
+ if (*msg == '\0')
+ cmm_set_timeout(nr, seconds);
+ }
+}
+#endif
+
+static struct ctl_table_header *cmm_sysctl_header;
+
+static int cmm_suspend(void)
+{
+ cmm_suspended = 1;
+ cmm_free_pages(cmm_pages, &cmm_pages, &cmm_page_list);
+ cmm_free_pages(cmm_timed_pages, &cmm_timed_pages, &cmm_timed_page_list);
+ return 0;
+}
+
+static int cmm_resume(void)
+{
+ cmm_suspended = 0;
+ cmm_kick_thread();
+ return 0;
+}
+
+static int cmm_power_event(struct notifier_block *this,
+ unsigned long event, void *ptr)
+{
+ switch (event) {
+ case PM_POST_HIBERNATION:
+ return cmm_resume();
+ case PM_HIBERNATION_PREPARE:
+ return cmm_suspend();
+ default:
+ return NOTIFY_DONE;
+ }
+}
+
+static struct notifier_block cmm_power_notifier = {
+ .notifier_call = cmm_power_event,
+};
+
+static int __init cmm_init(void)
+{
+ int rc = -ENOMEM;
+
+ cmm_sysctl_header = register_sysctl_table(cmm_dir_table);
+ if (!cmm_sysctl_header)
+ goto out_sysctl;
+#ifdef CONFIG_CMM_IUCV
+ /* convert sender to uppercase characters */
+ if (sender) {
+ int len = strlen(sender);
+ while (len--)
+ sender[len] = toupper(sender[len]);
+ } else {
+ sender = cmm_default_sender;
+ }
+
+ rc = smsg_register_callback(SMSG_PREFIX, cmm_smsg_target);
+ if (rc < 0)
+ goto out_smsg;
+#endif
+ rc = register_oom_notifier(&cmm_oom_nb);
+ if (rc < 0)
+ goto out_oom_notify;
+ rc = register_pm_notifier(&cmm_power_notifier);
+ if (rc)
+ goto out_pm;
+ cmm_thread_ptr = kthread_run(cmm_thread, NULL, "cmmthread");
+ rc = IS_ERR(cmm_thread_ptr) ? PTR_ERR(cmm_thread_ptr) : 0;
+ if (rc)
+ goto out_kthread;
+ return 0;
+
+out_kthread:
+ unregister_pm_notifier(&cmm_power_notifier);
+out_pm:
+ unregister_oom_notifier(&cmm_oom_nb);
+out_oom_notify:
+#ifdef CONFIG_CMM_IUCV
+ smsg_unregister_callback(SMSG_PREFIX, cmm_smsg_target);
+out_smsg:
+#endif
+ unregister_sysctl_table(cmm_sysctl_header);
+out_sysctl:
+ del_timer_sync(&cmm_timer);
+ return rc;
+}
+module_init(cmm_init);
+
+static void __exit cmm_exit(void)
+{
+ unregister_sysctl_table(cmm_sysctl_header);
+#ifdef CONFIG_CMM_IUCV
+ smsg_unregister_callback(SMSG_PREFIX, cmm_smsg_target);
+#endif
+ unregister_pm_notifier(&cmm_power_notifier);
+ unregister_oom_notifier(&cmm_oom_nb);
+ kthread_stop(cmm_thread_ptr);
+ del_timer_sync(&cmm_timer);
+ cmm_free_pages(cmm_pages, &cmm_pages, &cmm_page_list);
+ cmm_free_pages(cmm_timed_pages, &cmm_timed_pages, &cmm_timed_page_list);
+}
+module_exit(cmm_exit);
+
+MODULE_LICENSE("GPL");
diff --git a/arch/s390/mm/extmem.c b/arch/s390/mm/extmem.c
new file mode 100644
index 00000000..075ddada
--- /dev/null
+++ b/arch/s390/mm/extmem.c
@@ -0,0 +1,778 @@
+/*
+ * File...........: arch/s390/mm/extmem.c
+ * Author(s)......: Carsten Otte <cotte@de.ibm.com>
+ * Rob M van der Heij <rvdheij@nl.ibm.com>
+ * Steven Shultz <shultzss@us.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * (C) IBM Corporation 2002-2004
+ */
+
+#define KMSG_COMPONENT "extmem"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/spinlock.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <linux/bootmem.h>
+#include <linux/ctype.h>
+#include <linux/ioport.h>
+#include <asm/page.h>
+#include <asm/pgtable.h>
+#include <asm/ebcdic.h>
+#include <asm/errno.h>
+#include <asm/extmem.h>
+#include <asm/cpcmd.h>
+#include <asm/setup.h>
+
+#define DCSS_LOADSHR 0x00
+#define DCSS_LOADNSR 0x04
+#define DCSS_PURGESEG 0x08
+#define DCSS_FINDSEG 0x0c
+#define DCSS_LOADNOLY 0x10
+#define DCSS_SEGEXT 0x18
+#define DCSS_LOADSHRX 0x20
+#define DCSS_LOADNSRX 0x24
+#define DCSS_FINDSEGX 0x2c
+#define DCSS_SEGEXTX 0x38
+#define DCSS_FINDSEGA 0x0c
+
+struct qrange {
+ unsigned long start; /* last byte type */
+ unsigned long end; /* last byte reserved */
+};
+
+struct qout64 {
+ unsigned long segstart;
+ unsigned long segend;
+ int segcnt;
+ int segrcnt;
+ struct qrange range[6];
+};
+
+#ifdef CONFIG_64BIT
+struct qrange_old {
+ unsigned int start; /* last byte type */
+ unsigned int end; /* last byte reserved */
+};
+
+/* output area format for the Diag x'64' old subcode x'18' */
+struct qout64_old {
+ int segstart;
+ int segend;
+ int segcnt;
+ int segrcnt;
+ struct qrange_old range[6];
+};
+#endif
+
+struct qin64 {
+ char qopcode;
+ char rsrv1[3];
+ char qrcode;
+ char rsrv2[3];
+ char qname[8];
+ unsigned int qoutptr;
+ short int qoutlen;
+};
+
+struct dcss_segment {
+ struct list_head list;
+ char dcss_name[8];
+ char res_name[15];
+ unsigned long start_addr;
+ unsigned long end;
+ atomic_t ref_count;
+ int do_nonshared;
+ unsigned int vm_segtype;
+ struct qrange range[6];
+ int segcnt;
+ struct resource *res;
+};
+
+static DEFINE_MUTEX(dcss_lock);
+static LIST_HEAD(dcss_list);
+static char *segtype_string[] = { "SW", "EW", "SR", "ER", "SN", "EN", "SC",
+ "EW/EN-MIXED" };
+static int loadshr_scode, loadnsr_scode, findseg_scode;
+static int segext_scode, purgeseg_scode;
+static int scode_set;
+
+/* set correct Diag x'64' subcodes. */
+static int
+dcss_set_subcodes(void)
+{
+#ifdef CONFIG_64BIT
+ char *name = kmalloc(8 * sizeof(char), GFP_KERNEL | GFP_DMA);
+ unsigned long rx, ry;
+ int rc;
+
+ if (name == NULL)
+ return -ENOMEM;
+
+ rx = (unsigned long) name;
+ ry = DCSS_FINDSEGX;
+
+ strcpy(name, "dummy");
+ asm volatile(
+ " diag %0,%1,0x64\n"
+ "0: ipm %2\n"
+ " srl %2,28\n"
+ " j 2f\n"
+ "1: la %2,3\n"
+ "2:\n"
+ EX_TABLE(0b, 1b)
+ : "+d" (rx), "+d" (ry), "=d" (rc) : : "cc");
+
+ kfree(name);
+ /* Diag x'64' new subcodes are supported, set to new subcodes */
+ if (rc != 3) {
+ loadshr_scode = DCSS_LOADSHRX;
+ loadnsr_scode = DCSS_LOADNSRX;
+ purgeseg_scode = DCSS_PURGESEG;
+ findseg_scode = DCSS_FINDSEGX;
+ segext_scode = DCSS_SEGEXTX;
+ return 0;
+ }
+#endif
+ /* Diag x'64' new subcodes are not supported, set to old subcodes */
+ loadshr_scode = DCSS_LOADNOLY;
+ loadnsr_scode = DCSS_LOADNSR;
+ purgeseg_scode = DCSS_PURGESEG;
+ findseg_scode = DCSS_FINDSEG;
+ segext_scode = DCSS_SEGEXT;
+ return 0;
+}
+
+/*
+ * Create the 8 bytes, ebcdic VM segment name from
+ * an ascii name.
+ */
+static void
+dcss_mkname(char *name, char *dcss_name)
+{
+ int i;
+
+ for (i = 0; i < 8; i++) {
+ if (name[i] == '\0')
+ break;
+ dcss_name[i] = toupper(name[i]);
+ };
+ for (; i < 8; i++)
+ dcss_name[i] = ' ';
+ ASCEBC(dcss_name, 8);
+}
+
+
+/*
+ * search all segments in dcss_list, and return the one
+ * namend *name. If not found, return NULL.
+ */
+static struct dcss_segment *
+segment_by_name (char *name)
+{
+ char dcss_name[9];
+ struct list_head *l;
+ struct dcss_segment *tmp, *retval = NULL;
+
+ BUG_ON(!mutex_is_locked(&dcss_lock));
+ dcss_mkname (name, dcss_name);
+ list_for_each (l, &dcss_list) {
+ tmp = list_entry (l, struct dcss_segment, list);
+ if (memcmp(tmp->dcss_name, dcss_name, 8) == 0) {
+ retval = tmp;
+ break;
+ }
+ }
+ return retval;
+}
+
+
+/*
+ * Perform a function on a dcss segment.
+ */
+static inline int
+dcss_diag(int *func, void *parameter,
+ unsigned long *ret1, unsigned long *ret2)
+{
+ unsigned long rx, ry;
+ int rc;
+
+ if (scode_set == 0) {
+ rc = dcss_set_subcodes();
+ if (rc < 0)
+ return rc;
+ scode_set = 1;
+ }
+ rx = (unsigned long) parameter;
+ ry = (unsigned long) *func;
+
+#ifdef CONFIG_64BIT
+ /* 64-bit Diag x'64' new subcode, keep in 64-bit addressing mode */
+ if (*func > DCSS_SEGEXT)
+ asm volatile(
+ " diag %0,%1,0x64\n"
+ " ipm %2\n"
+ " srl %2,28\n"
+ : "+d" (rx), "+d" (ry), "=d" (rc) : : "cc");
+ /* 31-bit Diag x'64' old subcode, switch to 31-bit addressing mode */
+ else
+ asm volatile(
+ " sam31\n"
+ " diag %0,%1,0x64\n"
+ " sam64\n"
+ " ipm %2\n"
+ " srl %2,28\n"
+ : "+d" (rx), "+d" (ry), "=d" (rc) : : "cc");
+#else
+ asm volatile(
+ " diag %0,%1,0x64\n"
+ " ipm %2\n"
+ " srl %2,28\n"
+ : "+d" (rx), "+d" (ry), "=d" (rc) : : "cc");
+#endif
+ *ret1 = rx;
+ *ret2 = ry;
+ return rc;
+}
+
+static inline int
+dcss_diag_translate_rc (int vm_rc) {
+ if (vm_rc == 44)
+ return -ENOENT;
+ return -EIO;
+}
+
+
+/* do a diag to get info about a segment.
+ * fills start_address, end and vm_segtype fields
+ */
+static int
+query_segment_type (struct dcss_segment *seg)
+{
+ unsigned long dummy, vmrc;
+ int diag_cc, rc, i;
+ struct qout64 *qout;
+ struct qin64 *qin;
+
+ qin = kmalloc(sizeof(*qin), GFP_KERNEL | GFP_DMA);
+ qout = kmalloc(sizeof(*qout), GFP_KERNEL | GFP_DMA);
+ if ((qin == NULL) || (qout == NULL)) {
+ rc = -ENOMEM;
+ goto out_free;
+ }
+
+ /* initialize diag input parameters */
+ qin->qopcode = DCSS_FINDSEGA;
+ qin->qoutptr = (unsigned long) qout;
+ qin->qoutlen = sizeof(struct qout64);
+ memcpy (qin->qname, seg->dcss_name, 8);
+
+ diag_cc = dcss_diag(&segext_scode, qin, &dummy, &vmrc);
+
+ if (diag_cc < 0) {
+ rc = diag_cc;
+ goto out_free;
+ }
+ if (diag_cc > 1) {
+ pr_warning("Querying a DCSS type failed with rc=%ld\n", vmrc);
+ rc = dcss_diag_translate_rc (vmrc);
+ goto out_free;
+ }
+
+#ifdef CONFIG_64BIT
+ /* Only old format of output area of Diagnose x'64' is supported,
+ copy data for the new format. */
+ if (segext_scode == DCSS_SEGEXT) {
+ struct qout64_old *qout_old;
+ qout_old = kzalloc(sizeof(*qout_old), GFP_KERNEL | GFP_DMA);
+ if (qout_old == NULL) {
+ rc = -ENOMEM;
+ goto out_free;
+ }
+ memcpy(qout_old, qout, sizeof(struct qout64_old));
+ qout->segstart = (unsigned long) qout_old->segstart;
+ qout->segend = (unsigned long) qout_old->segend;
+ qout->segcnt = qout_old->segcnt;
+ qout->segrcnt = qout_old->segrcnt;
+
+ if (qout->segcnt > 6)
+ qout->segrcnt = 6;
+ for (i = 0; i < qout->segrcnt; i++) {
+ qout->range[i].start =
+ (unsigned long) qout_old->range[i].start;
+ qout->range[i].end =
+ (unsigned long) qout_old->range[i].end;
+ }
+ kfree(qout_old);
+ }
+#endif
+ if (qout->segcnt > 6) {
+ rc = -EOPNOTSUPP;
+ goto out_free;
+ }
+
+ if (qout->segcnt == 1) {
+ seg->vm_segtype = qout->range[0].start & 0xff;
+ } else {
+ /* multi-part segment. only one type supported here:
+ - all parts are contiguous
+ - all parts are either EW or EN type
+ - maximum 6 parts allowed */
+ unsigned long start = qout->segstart >> PAGE_SHIFT;
+ for (i=0; i<qout->segcnt; i++) {
+ if (((qout->range[i].start & 0xff) != SEG_TYPE_EW) &&
+ ((qout->range[i].start & 0xff) != SEG_TYPE_EN)) {
+ rc = -EOPNOTSUPP;
+ goto out_free;
+ }
+ if (start != qout->range[i].start >> PAGE_SHIFT) {
+ rc = -EOPNOTSUPP;
+ goto out_free;
+ }
+ start = (qout->range[i].end >> PAGE_SHIFT) + 1;
+ }
+ seg->vm_segtype = SEG_TYPE_EWEN;
+ }
+
+ /* analyze diag output and update seg */
+ seg->start_addr = qout->segstart;
+ seg->end = qout->segend;
+
+ memcpy (seg->range, qout->range, 6*sizeof(struct qrange));
+ seg->segcnt = qout->segcnt;
+
+ rc = 0;
+
+ out_free:
+ kfree(qin);
+ kfree(qout);
+ return rc;
+}
+
+/*
+ * get info about a segment
+ * possible return values:
+ * -ENOSYS : we are not running on VM
+ * -EIO : could not perform query diagnose
+ * -ENOENT : no such segment
+ * -EOPNOTSUPP: multi-part segment cannot be used with linux
+ * -ENOMEM : out of memory
+ * 0 .. 6 : type of segment as defined in include/asm-s390/extmem.h
+ */
+int
+segment_type (char* name)
+{
+ int rc;
+ struct dcss_segment seg;
+
+ if (!MACHINE_IS_VM)
+ return -ENOSYS;
+
+ dcss_mkname(name, seg.dcss_name);
+ rc = query_segment_type (&seg);
+ if (rc < 0)
+ return rc;
+ return seg.vm_segtype;
+}
+
+/*
+ * check if segment collides with other segments that are currently loaded
+ * returns 1 if this is the case, 0 if no collision was found
+ */
+static int
+segment_overlaps_others (struct dcss_segment *seg)
+{
+ struct list_head *l;
+ struct dcss_segment *tmp;
+
+ BUG_ON(!mutex_is_locked(&dcss_lock));
+ list_for_each(l, &dcss_list) {
+ tmp = list_entry(l, struct dcss_segment, list);
+ if ((tmp->start_addr >> 20) > (seg->end >> 20))
+ continue;
+ if ((tmp->end >> 20) < (seg->start_addr >> 20))
+ continue;
+ if (seg == tmp)
+ continue;
+ return 1;
+ }
+ return 0;
+}
+
+/*
+ * real segment loading function, called from segment_load
+ */
+static int
+__segment_load (char *name, int do_nonshared, unsigned long *addr, unsigned long *end)
+{
+ unsigned long start_addr, end_addr, dummy;
+ struct dcss_segment *seg;
+ int rc, diag_cc;
+
+ start_addr = end_addr = 0;
+ seg = kmalloc(sizeof(*seg), GFP_KERNEL | GFP_DMA);
+ if (seg == NULL) {
+ rc = -ENOMEM;
+ goto out;
+ }
+ dcss_mkname (name, seg->dcss_name);
+ rc = query_segment_type (seg);
+ if (rc < 0)
+ goto out_free;
+
+ if (loadshr_scode == DCSS_LOADSHRX) {
+ if (segment_overlaps_others(seg)) {
+ rc = -EBUSY;
+ goto out_free;
+ }
+ }
+
+ rc = vmem_add_mapping(seg->start_addr, seg->end - seg->start_addr + 1);
+
+ if (rc)
+ goto out_free;
+
+ seg->res = kzalloc(sizeof(struct resource), GFP_KERNEL);
+ if (seg->res == NULL) {
+ rc = -ENOMEM;
+ goto out_shared;
+ }
+ seg->res->flags = IORESOURCE_BUSY | IORESOURCE_MEM;
+ seg->res->start = seg->start_addr;
+ seg->res->end = seg->end;
+ memcpy(&seg->res_name, seg->dcss_name, 8);
+ EBCASC(seg->res_name, 8);
+ seg->res_name[8] = '\0';
+ strncat(seg->res_name, " (DCSS)", 7);
+ seg->res->name = seg->res_name;
+ rc = seg->vm_segtype;
+ if (rc == SEG_TYPE_SC ||
+ ((rc == SEG_TYPE_SR || rc == SEG_TYPE_ER) && !do_nonshared))
+ seg->res->flags |= IORESOURCE_READONLY;
+ if (request_resource(&iomem_resource, seg->res)) {
+ rc = -EBUSY;
+ kfree(seg->res);
+ goto out_shared;
+ }
+
+ if (do_nonshared)
+ diag_cc = dcss_diag(&loadnsr_scode, seg->dcss_name,
+ &start_addr, &end_addr);
+ else
+ diag_cc = dcss_diag(&loadshr_scode, seg->dcss_name,
+ &start_addr, &end_addr);
+ if (diag_cc < 0) {
+ dcss_diag(&purgeseg_scode, seg->dcss_name,
+ &dummy, &dummy);
+ rc = diag_cc;
+ goto out_resource;
+ }
+ if (diag_cc > 1) {
+ pr_warning("Loading DCSS %s failed with rc=%ld\n", name,
+ end_addr);
+ rc = dcss_diag_translate_rc(end_addr);
+ dcss_diag(&purgeseg_scode, seg->dcss_name,
+ &dummy, &dummy);
+ goto out_resource;
+ }
+ seg->start_addr = start_addr;
+ seg->end = end_addr;
+ seg->do_nonshared = do_nonshared;
+ atomic_set(&seg->ref_count, 1);
+ list_add(&seg->list, &dcss_list);
+ *addr = seg->start_addr;
+ *end = seg->end;
+ if (do_nonshared)
+ pr_info("DCSS %s of range %p to %p and type %s loaded as "
+ "exclusive-writable\n", name, (void*) seg->start_addr,
+ (void*) seg->end, segtype_string[seg->vm_segtype]);
+ else {
+ pr_info("DCSS %s of range %p to %p and type %s loaded in "
+ "shared access mode\n", name, (void*) seg->start_addr,
+ (void*) seg->end, segtype_string[seg->vm_segtype]);
+ }
+ goto out;
+ out_resource:
+ release_resource(seg->res);
+ kfree(seg->res);
+ out_shared:
+ vmem_remove_mapping(seg->start_addr, seg->end - seg->start_addr + 1);
+ out_free:
+ kfree(seg);
+ out:
+ return rc;
+}
+
+/*
+ * this function loads a DCSS segment
+ * name : name of the DCSS
+ * do_nonshared : 0 indicates that the dcss should be shared with other linux images
+ * 1 indicates that the dcss should be exclusive for this linux image
+ * addr : will be filled with start address of the segment
+ * end : will be filled with end address of the segment
+ * return values:
+ * -ENOSYS : we are not running on VM
+ * -EIO : could not perform query or load diagnose
+ * -ENOENT : no such segment
+ * -EOPNOTSUPP: multi-part segment cannot be used with linux
+ * -ENOSPC : segment cannot be used (overlaps with storage)
+ * -EBUSY : segment can temporarily not be used (overlaps with dcss)
+ * -ERANGE : segment cannot be used (exceeds kernel mapping range)
+ * -EPERM : segment is currently loaded with incompatible permissions
+ * -ENOMEM : out of memory
+ * 0 .. 6 : type of segment as defined in include/asm-s390/extmem.h
+ */
+int
+segment_load (char *name, int do_nonshared, unsigned long *addr,
+ unsigned long *end)
+{
+ struct dcss_segment *seg;
+ int rc;
+
+ if (!MACHINE_IS_VM)
+ return -ENOSYS;
+
+ mutex_lock(&dcss_lock);
+ seg = segment_by_name (name);
+ if (seg == NULL)
+ rc = __segment_load (name, do_nonshared, addr, end);
+ else {
+ if (do_nonshared == seg->do_nonshared) {
+ atomic_inc(&seg->ref_count);
+ *addr = seg->start_addr;
+ *end = seg->end;
+ rc = seg->vm_segtype;
+ } else {
+ *addr = *end = 0;
+ rc = -EPERM;
+ }
+ }
+ mutex_unlock(&dcss_lock);
+ return rc;
+}
+
+/*
+ * this function modifies the shared state of a DCSS segment. note that
+ * name : name of the DCSS
+ * do_nonshared : 0 indicates that the dcss should be shared with other linux images
+ * 1 indicates that the dcss should be exclusive for this linux image
+ * return values:
+ * -EIO : could not perform load diagnose (segment gone!)
+ * -ENOENT : no such segment (segment gone!)
+ * -EAGAIN : segment is in use by other exploiters, try later
+ * -EINVAL : no segment with the given name is currently loaded - name invalid
+ * -EBUSY : segment can temporarily not be used (overlaps with dcss)
+ * 0 : operation succeeded
+ */
+int
+segment_modify_shared (char *name, int do_nonshared)
+{
+ struct dcss_segment *seg;
+ unsigned long start_addr, end_addr, dummy;
+ int rc, diag_cc;
+
+ start_addr = end_addr = 0;
+ mutex_lock(&dcss_lock);
+ seg = segment_by_name (name);
+ if (seg == NULL) {
+ rc = -EINVAL;
+ goto out_unlock;
+ }
+ if (do_nonshared == seg->do_nonshared) {
+ pr_info("DCSS %s is already in the requested access "
+ "mode\n", name);
+ rc = 0;
+ goto out_unlock;
+ }
+ if (atomic_read (&seg->ref_count) != 1) {
+ pr_warning("DCSS %s is in use and cannot be reloaded\n",
+ name);
+ rc = -EAGAIN;
+ goto out_unlock;
+ }
+ release_resource(seg->res);
+ if (do_nonshared)
+ seg->res->flags &= ~IORESOURCE_READONLY;
+ else
+ if (seg->vm_segtype == SEG_TYPE_SR ||
+ seg->vm_segtype == SEG_TYPE_ER)
+ seg->res->flags |= IORESOURCE_READONLY;
+
+ if (request_resource(&iomem_resource, seg->res)) {
+ pr_warning("DCSS %s overlaps with used memory resources "
+ "and cannot be reloaded\n", name);
+ rc = -EBUSY;
+ kfree(seg->res);
+ goto out_del_mem;
+ }
+
+ dcss_diag(&purgeseg_scode, seg->dcss_name, &dummy, &dummy);
+ if (do_nonshared)
+ diag_cc = dcss_diag(&loadnsr_scode, seg->dcss_name,
+ &start_addr, &end_addr);
+ else
+ diag_cc = dcss_diag(&loadshr_scode, seg->dcss_name,
+ &start_addr, &end_addr);
+ if (diag_cc < 0) {
+ rc = diag_cc;
+ goto out_del_res;
+ }
+ if (diag_cc > 1) {
+ pr_warning("Reloading DCSS %s failed with rc=%ld\n", name,
+ end_addr);
+ rc = dcss_diag_translate_rc(end_addr);
+ goto out_del_res;
+ }
+ seg->start_addr = start_addr;
+ seg->end = end_addr;
+ seg->do_nonshared = do_nonshared;
+ rc = 0;
+ goto out_unlock;
+ out_del_res:
+ release_resource(seg->res);
+ kfree(seg->res);
+ out_del_mem:
+ vmem_remove_mapping(seg->start_addr, seg->end - seg->start_addr + 1);
+ list_del(&seg->list);
+ dcss_diag(&purgeseg_scode, seg->dcss_name, &dummy, &dummy);
+ kfree(seg);
+ out_unlock:
+ mutex_unlock(&dcss_lock);
+ return rc;
+}
+
+/*
+ * Decrease the use count of a DCSS segment and remove
+ * it from the address space if nobody is using it
+ * any longer.
+ */
+void
+segment_unload(char *name)
+{
+ unsigned long dummy;
+ struct dcss_segment *seg;
+
+ if (!MACHINE_IS_VM)
+ return;
+
+ mutex_lock(&dcss_lock);
+ seg = segment_by_name (name);
+ if (seg == NULL) {
+ pr_err("Unloading unknown DCSS %s failed\n", name);
+ goto out_unlock;
+ }
+ if (atomic_dec_return(&seg->ref_count) != 0)
+ goto out_unlock;
+ release_resource(seg->res);
+ kfree(seg->res);
+ vmem_remove_mapping(seg->start_addr, seg->end - seg->start_addr + 1);
+ list_del(&seg->list);
+ dcss_diag(&purgeseg_scode, seg->dcss_name, &dummy, &dummy);
+ kfree(seg);
+out_unlock:
+ mutex_unlock(&dcss_lock);
+}
+
+/*
+ * save segment content permanently
+ */
+void
+segment_save(char *name)
+{
+ struct dcss_segment *seg;
+ char cmd1[160];
+ char cmd2[80];
+ int i, response;
+
+ if (!MACHINE_IS_VM)
+ return;
+
+ mutex_lock(&dcss_lock);
+ seg = segment_by_name (name);
+
+ if (seg == NULL) {
+ pr_err("Saving unknown DCSS %s failed\n", name);
+ goto out;
+ }
+
+ sprintf(cmd1, "DEFSEG %s", name);
+ for (i=0; i<seg->segcnt; i++) {
+ sprintf(cmd1+strlen(cmd1), " %lX-%lX %s",
+ seg->range[i].start >> PAGE_SHIFT,
+ seg->range[i].end >> PAGE_SHIFT,
+ segtype_string[seg->range[i].start & 0xff]);
+ }
+ sprintf(cmd2, "SAVESEG %s", name);
+ response = 0;
+ cpcmd(cmd1, NULL, 0, &response);
+ if (response) {
+ pr_err("Saving a DCSS failed with DEFSEG response code "
+ "%i\n", response);
+ goto out;
+ }
+ cpcmd(cmd2, NULL, 0, &response);
+ if (response) {
+ pr_err("Saving a DCSS failed with SAVESEG response code "
+ "%i\n", response);
+ goto out;
+ }
+out:
+ mutex_unlock(&dcss_lock);
+}
+
+/*
+ * print appropriate error message for segment_load()/segment_type()
+ * return code
+ */
+void segment_warning(int rc, char *seg_name)
+{
+ switch (rc) {
+ case -ENOENT:
+ pr_err("DCSS %s cannot be loaded or queried\n", seg_name);
+ break;
+ case -ENOSYS:
+ pr_err("DCSS %s cannot be loaded or queried without "
+ "z/VM\n", seg_name);
+ break;
+ case -EIO:
+ pr_err("Loading or querying DCSS %s resulted in a "
+ "hardware error\n", seg_name);
+ break;
+ case -EOPNOTSUPP:
+ pr_err("DCSS %s has multiple page ranges and cannot be "
+ "loaded or queried\n", seg_name);
+ break;
+ case -ENOSPC:
+ pr_err("DCSS %s overlaps with used storage and cannot "
+ "be loaded\n", seg_name);
+ break;
+ case -EBUSY:
+ pr_err("%s needs used memory resources and cannot be "
+ "loaded or queried\n", seg_name);
+ break;
+ case -EPERM:
+ pr_err("DCSS %s is already loaded in a different access "
+ "mode\n", seg_name);
+ break;
+ case -ENOMEM:
+ pr_err("There is not enough memory to load or query "
+ "DCSS %s\n", seg_name);
+ break;
+ case -ERANGE:
+ pr_err("DCSS %s exceeds the kernel mapping range (%lu) "
+ "and cannot be loaded\n", seg_name, VMEM_MAX_PHYS);
+ break;
+ default:
+ break;
+ }
+}
+
+EXPORT_SYMBOL(segment_load);
+EXPORT_SYMBOL(segment_unload);
+EXPORT_SYMBOL(segment_save);
+EXPORT_SYMBOL(segment_type);
+EXPORT_SYMBOL(segment_modify_shared);
+EXPORT_SYMBOL(segment_warning);
diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
new file mode 100644
index 00000000..4e668600
--- /dev/null
+++ b/arch/s390/mm/fault.c
@@ -0,0 +1,664 @@
+/*
+ * arch/s390/mm/fault.c
+ *
+ * S390 version
+ * Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ * Author(s): Hartmut Penner (hp@de.ibm.com)
+ * Ulrich Weigand (uweigand@de.ibm.com)
+ *
+ * Derived from "arch/i386/mm/fault.c"
+ * Copyright (C) 1995 Linus Torvalds
+ */
+
+#include <linux/kernel_stat.h>
+#include <linux/perf_event.h>
+#include <linux/signal.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/ptrace.h>
+#include <linux/mman.h>
+#include <linux/mm.h>
+#include <linux/compat.h>
+#include <linux/smp.h>
+#include <linux/kdebug.h>
+#include <linux/init.h>
+#include <linux/console.h>
+#include <linux/module.h>
+#include <linux/hardirq.h>
+#include <linux/kprobes.h>
+#include <linux/uaccess.h>
+#include <linux/hugetlb.h>
+#include <asm/asm-offsets.h>
+#include <asm/pgtable.h>
+#include <asm/irq.h>
+#include <asm/mmu_context.h>
+#include <asm/facility.h>
+#include "../kernel/entry.h"
+
+#ifndef CONFIG_64BIT
+#define __FAIL_ADDR_MASK 0x7ffff000
+#define __SUBCODE_MASK 0x0200
+#define __PF_RES_FIELD 0ULL
+#else /* CONFIG_64BIT */
+#define __FAIL_ADDR_MASK -4096L
+#define __SUBCODE_MASK 0x0600
+#define __PF_RES_FIELD 0x8000000000000000ULL
+#endif /* CONFIG_64BIT */
+
+#define VM_FAULT_BADCONTEXT 0x010000
+#define VM_FAULT_BADMAP 0x020000
+#define VM_FAULT_BADACCESS 0x040000
+
+static unsigned long store_indication;
+
+void fault_init(void)
+{
+ if (test_facility(2) && test_facility(75))
+ store_indication = 0xc00;
+}
+
+static inline int notify_page_fault(struct pt_regs *regs)
+{
+ int ret = 0;
+
+ /* kprobe_running() needs smp_processor_id() */
+ if (kprobes_built_in() && !user_mode(regs)) {
+ preempt_disable();
+ if (kprobe_running() && kprobe_fault_handler(regs, 14))
+ ret = 1;
+ preempt_enable();
+ }
+ return ret;
+}
+
+
+/*
+ * Unlock any spinlocks which will prevent us from getting the
+ * message out.
+ */
+void bust_spinlocks(int yes)
+{
+ if (yes) {
+ oops_in_progress = 1;
+ } else {
+ int loglevel_save = console_loglevel;
+ console_unblank();
+ oops_in_progress = 0;
+ /*
+ * OK, the message is on the console. Now we call printk()
+ * without oops_in_progress set so that printk will give klogd
+ * a poke. Hold onto your hats...
+ */
+ console_loglevel = 15;
+ printk(" ");
+ console_loglevel = loglevel_save;
+ }
+}
+
+/*
+ * Returns the address space associated with the fault.
+ * Returns 0 for kernel space and 1 for user space.
+ */
+static inline int user_space_fault(unsigned long trans_exc_code)
+{
+ /*
+ * The lowest two bits of the translation exception
+ * identification indicate which paging table was used.
+ */
+ trans_exc_code &= 3;
+ if (trans_exc_code == 2)
+ /* Access via secondary space, set_fs setting decides */
+ return current->thread.mm_segment.ar4;
+ if (user_mode == HOME_SPACE_MODE)
+ /* User space if the access has been done via home space. */
+ return trans_exc_code == 3;
+ /*
+ * If the user space is not the home space the kernel runs in home
+ * space. Access via secondary space has already been covered,
+ * access via primary space or access register is from user space
+ * and access via home space is from the kernel.
+ */
+ return trans_exc_code != 3;
+}
+
+static inline void report_user_fault(struct pt_regs *regs, long signr)
+{
+ if ((task_pid_nr(current) > 1) && !show_unhandled_signals)
+ return;
+ if (!unhandled_signal(current, signr))
+ return;
+ if (!printk_ratelimit())
+ return;
+ printk(KERN_ALERT "User process fault: interruption code 0x%X ",
+ regs->int_code);
+ print_vma_addr(KERN_CONT "in ", regs->psw.addr & PSW_ADDR_INSN);
+ printk(KERN_CONT "\n");
+ printk(KERN_ALERT "failing address: %lX\n",
+ regs->int_parm_long & __FAIL_ADDR_MASK);
+ show_regs(regs);
+}
+
+/*
+ * Send SIGSEGV to task. This is an external routine
+ * to keep the stack usage of do_page_fault small.
+ */
+static noinline void do_sigsegv(struct pt_regs *regs, int si_code)
+{
+ struct siginfo si;
+
+ report_user_fault(regs, SIGSEGV);
+ si.si_signo = SIGSEGV;
+ si.si_code = si_code;
+ si.si_addr = (void __user *)(regs->int_parm_long & __FAIL_ADDR_MASK);
+ force_sig_info(SIGSEGV, &si, current);
+}
+
+static noinline void do_no_context(struct pt_regs *regs)
+{
+ const struct exception_table_entry *fixup;
+ unsigned long address;
+
+ /* Are we prepared to handle this kernel fault? */
+ fixup = search_exception_tables(regs->psw.addr & PSW_ADDR_INSN);
+ if (fixup) {
+ regs->psw.addr = fixup->fixup | PSW_ADDR_AMODE;
+ return;
+ }
+
+ /*
+ * Oops. The kernel tried to access some bad page. We'll have to
+ * terminate things with extreme prejudice.
+ */
+ address = regs->int_parm_long & __FAIL_ADDR_MASK;
+ if (!user_space_fault(regs->int_parm_long))
+ printk(KERN_ALERT "Unable to handle kernel pointer dereference"
+ " at virtual kernel address %p\n", (void *)address);
+ else
+ printk(KERN_ALERT "Unable to handle kernel paging request"
+ " at virtual user address %p\n", (void *)address);
+
+ die(regs, "Oops");
+ do_exit(SIGKILL);
+}
+
+static noinline void do_low_address(struct pt_regs *regs)
+{
+ /* Low-address protection hit in kernel mode means
+ NULL pointer write access in kernel mode. */
+ if (regs->psw.mask & PSW_MASK_PSTATE) {
+ /* Low-address protection hit in user mode 'cannot happen'. */
+ die (regs, "Low-address protection");
+ do_exit(SIGKILL);
+ }
+
+ do_no_context(regs);
+}
+
+static noinline void do_sigbus(struct pt_regs *regs)
+{
+ struct task_struct *tsk = current;
+ struct siginfo si;
+
+ /*
+ * Send a sigbus, regardless of whether we were in kernel
+ * or user mode.
+ */
+ si.si_signo = SIGBUS;
+ si.si_errno = 0;
+ si.si_code = BUS_ADRERR;
+ si.si_addr = (void __user *)(regs->int_parm_long & __FAIL_ADDR_MASK);
+ force_sig_info(SIGBUS, &si, tsk);
+}
+
+static noinline void do_fault_error(struct pt_regs *regs, int fault)
+{
+ int si_code;
+
+ switch (fault) {
+ case VM_FAULT_BADACCESS:
+ case VM_FAULT_BADMAP:
+ /* Bad memory access. Check if it is kernel or user space. */
+ if (regs->psw.mask & PSW_MASK_PSTATE) {
+ /* User mode accesses just cause a SIGSEGV */
+ si_code = (fault == VM_FAULT_BADMAP) ?
+ SEGV_MAPERR : SEGV_ACCERR;
+ do_sigsegv(regs, si_code);
+ return;
+ }
+ case VM_FAULT_BADCONTEXT:
+ do_no_context(regs);
+ break;
+ default: /* fault & VM_FAULT_ERROR */
+ if (fault & VM_FAULT_OOM) {
+ if (!(regs->psw.mask & PSW_MASK_PSTATE))
+ do_no_context(regs);
+ else
+ pagefault_out_of_memory();
+ } else if (fault & VM_FAULT_SIGBUS) {
+ /* Kernel mode? Handle exceptions or die */
+ if (!(regs->psw.mask & PSW_MASK_PSTATE))
+ do_no_context(regs);
+ else
+ do_sigbus(regs);
+ } else
+ BUG();
+ break;
+ }
+}
+
+/*
+ * This routine handles page faults. It determines the address,
+ * and the problem, and then passes it off to one of the appropriate
+ * routines.
+ *
+ * interruption code (int_code):
+ * 04 Protection -> Write-Protection (suprression)
+ * 10 Segment translation -> Not present (nullification)
+ * 11 Page translation -> Not present (nullification)
+ * 3b Region third trans. -> Not present (nullification)
+ */
+static inline int do_exception(struct pt_regs *regs, int access)
+{
+ struct task_struct *tsk;
+ struct mm_struct *mm;
+ struct vm_area_struct *vma;
+ unsigned long trans_exc_code;
+ unsigned long address;
+ unsigned int flags;
+ int fault;
+
+ if (notify_page_fault(regs))
+ return 0;
+
+ tsk = current;
+ mm = tsk->mm;
+ trans_exc_code = regs->int_parm_long;
+
+ /*
+ * Verify that the fault happened in user space, that
+ * we are not in an interrupt and that there is a
+ * user context.
+ */
+ fault = VM_FAULT_BADCONTEXT;
+ if (unlikely(!user_space_fault(trans_exc_code) || in_atomic() || !mm))
+ goto out;
+
+ address = trans_exc_code & __FAIL_ADDR_MASK;
+ perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
+ flags = FAULT_FLAG_ALLOW_RETRY;
+ if (access == VM_WRITE || (trans_exc_code & store_indication) == 0x400)
+ flags |= FAULT_FLAG_WRITE;
+ down_read(&mm->mmap_sem);
+
+#ifdef CONFIG_PGSTE
+ if (test_tsk_thread_flag(current, TIF_SIE) && S390_lowcore.gmap) {
+ address = __gmap_fault(address,
+ (struct gmap *) S390_lowcore.gmap);
+ if (address == -EFAULT) {
+ fault = VM_FAULT_BADMAP;
+ goto out_up;
+ }
+ if (address == -ENOMEM) {
+ fault = VM_FAULT_OOM;
+ goto out_up;
+ }
+ }
+#endif
+
+retry:
+ fault = VM_FAULT_BADMAP;
+ vma = find_vma(mm, address);
+ if (!vma)
+ goto out_up;
+
+ if (unlikely(vma->vm_start > address)) {
+ if (!(vma->vm_flags & VM_GROWSDOWN))
+ goto out_up;
+ if (expand_stack(vma, address))
+ goto out_up;
+ }
+
+ /*
+ * Ok, we have a good vm_area for this memory access, so
+ * we can handle it..
+ */
+ fault = VM_FAULT_BADACCESS;
+ if (unlikely(!(vma->vm_flags & access)))
+ goto out_up;
+
+ if (is_vm_hugetlb_page(vma))
+ address &= HPAGE_MASK;
+ /*
+ * If for any reason at all we couldn't handle the fault,
+ * make sure we exit gracefully rather than endlessly redo
+ * the fault.
+ */
+ fault = handle_mm_fault(mm, vma, address, flags);
+ if (unlikely(fault & VM_FAULT_ERROR))
+ goto out_up;
+
+ /*
+ * Major/minor page fault accounting is only done on the
+ * initial attempt. If we go through a retry, it is extremely
+ * likely that the page will be found in page cache at that point.
+ */
+ if (flags & FAULT_FLAG_ALLOW_RETRY) {
+ if (fault & VM_FAULT_MAJOR) {
+ tsk->maj_flt++;
+ perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1,
+ regs, address);
+ } else {
+ tsk->min_flt++;
+ perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1,
+ regs, address);
+ }
+ if (fault & VM_FAULT_RETRY) {
+ /* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk
+ * of starvation. */
+ flags &= ~FAULT_FLAG_ALLOW_RETRY;
+ down_read(&mm->mmap_sem);
+ goto retry;
+ }
+ }
+ /*
+ * The instruction that caused the program check will
+ * be repeated. Don't signal single step via SIGTRAP.
+ */
+ clear_tsk_thread_flag(tsk, TIF_PER_TRAP);
+ fault = 0;
+out_up:
+ up_read(&mm->mmap_sem);
+out:
+ return fault;
+}
+
+void __kprobes do_protection_exception(struct pt_regs *regs)
+{
+ unsigned long trans_exc_code;
+ int fault;
+
+ trans_exc_code = regs->int_parm_long;
+ /* Protection exception is suppressing, decrement psw address. */
+ regs->psw.addr = __rewind_psw(regs->psw, regs->int_code >> 16);
+ /*
+ * Check for low-address protection. This needs to be treated
+ * as a special case because the translation exception code
+ * field is not guaranteed to contain valid data in this case.
+ */
+ if (unlikely(!(trans_exc_code & 4))) {
+ do_low_address(regs);
+ return;
+ }
+ fault = do_exception(regs, VM_WRITE);
+ if (unlikely(fault))
+ do_fault_error(regs, fault);
+}
+
+void __kprobes do_dat_exception(struct pt_regs *regs)
+{
+ int access, fault;
+
+ access = VM_READ | VM_EXEC | VM_WRITE;
+ fault = do_exception(regs, access);
+ if (unlikely(fault))
+ do_fault_error(regs, fault);
+}
+
+#ifdef CONFIG_64BIT
+void __kprobes do_asce_exception(struct pt_regs *regs)
+{
+ struct mm_struct *mm = current->mm;
+ struct vm_area_struct *vma;
+ unsigned long trans_exc_code;
+
+ trans_exc_code = regs->int_parm_long;
+ if (unlikely(!user_space_fault(trans_exc_code) || in_atomic() || !mm))
+ goto no_context;
+
+ down_read(&mm->mmap_sem);
+ vma = find_vma(mm, trans_exc_code & __FAIL_ADDR_MASK);
+ up_read(&mm->mmap_sem);
+
+ if (vma) {
+ update_mm(mm, current);
+ return;
+ }
+
+ /* User mode accesses just cause a SIGSEGV */
+ if (regs->psw.mask & PSW_MASK_PSTATE) {
+ do_sigsegv(regs, SEGV_MAPERR);
+ return;
+ }
+
+no_context:
+ do_no_context(regs);
+}
+#endif
+
+int __handle_fault(unsigned long uaddr, unsigned long pgm_int_code, int write)
+{
+ struct pt_regs regs;
+ int access, fault;
+
+ regs.psw.mask = psw_kernel_bits | PSW_MASK_DAT | PSW_MASK_MCHECK;
+ if (!irqs_disabled())
+ regs.psw.mask |= PSW_MASK_IO | PSW_MASK_EXT;
+ regs.psw.addr = (unsigned long) __builtin_return_address(0);
+ regs.psw.addr |= PSW_ADDR_AMODE;
+ regs.int_code = pgm_int_code;
+ regs.int_parm_long = (uaddr & PAGE_MASK) | 2;
+ access = write ? VM_WRITE : VM_READ;
+ fault = do_exception(&regs, access);
+ if (unlikely(fault)) {
+ if (fault & VM_FAULT_OOM)
+ return -EFAULT;
+ else if (fault & VM_FAULT_SIGBUS)
+ do_sigbus(&regs);
+ }
+ return fault ? -EFAULT : 0;
+}
+
+#ifdef CONFIG_PFAULT
+/*
+ * 'pfault' pseudo page faults routines.
+ */
+static int pfault_disable;
+
+static int __init nopfault(char *str)
+{
+ pfault_disable = 1;
+ return 1;
+}
+
+__setup("nopfault", nopfault);
+
+struct pfault_refbk {
+ u16 refdiagc;
+ u16 reffcode;
+ u16 refdwlen;
+ u16 refversn;
+ u64 refgaddr;
+ u64 refselmk;
+ u64 refcmpmk;
+ u64 reserved;
+} __attribute__ ((packed, aligned(8)));
+
+int pfault_init(void)
+{
+ struct pfault_refbk refbk = {
+ .refdiagc = 0x258,
+ .reffcode = 0,
+ .refdwlen = 5,
+ .refversn = 2,
+ .refgaddr = __LC_CURRENT_PID,
+ .refselmk = 1ULL << 48,
+ .refcmpmk = 1ULL << 48,
+ .reserved = __PF_RES_FIELD };
+ int rc;
+
+ if (pfault_disable)
+ return -1;
+ asm volatile(
+ " diag %1,%0,0x258\n"
+ "0: j 2f\n"
+ "1: la %0,8\n"
+ "2:\n"
+ EX_TABLE(0b,1b)
+ : "=d" (rc) : "a" (&refbk), "m" (refbk) : "cc");
+ return rc;
+}
+
+void pfault_fini(void)
+{
+ struct pfault_refbk refbk = {
+ .refdiagc = 0x258,
+ .reffcode = 1,
+ .refdwlen = 5,
+ .refversn = 2,
+ };
+
+ if (pfault_disable)
+ return;
+ asm volatile(
+ " diag %0,0,0x258\n"
+ "0:\n"
+ EX_TABLE(0b,0b)
+ : : "a" (&refbk), "m" (refbk) : "cc");
+}
+
+static DEFINE_SPINLOCK(pfault_lock);
+static LIST_HEAD(pfault_list);
+
+static void pfault_interrupt(struct ext_code ext_code,
+ unsigned int param32, unsigned long param64)
+{
+ struct task_struct *tsk;
+ __u16 subcode;
+ pid_t pid;
+
+ /*
+ * Get the external interruption subcode & pfault
+ * initial/completion signal bit. VM stores this
+ * in the 'cpu address' field associated with the
+ * external interrupt.
+ */
+ subcode = ext_code.subcode;
+ if ((subcode & 0xff00) != __SUBCODE_MASK)
+ return;
+ kstat_cpu(smp_processor_id()).irqs[EXTINT_PFL]++;
+ if (subcode & 0x0080) {
+ /* Get the token (= pid of the affected task). */
+ pid = sizeof(void *) == 4 ? param32 : param64;
+ rcu_read_lock();
+ tsk = find_task_by_pid_ns(pid, &init_pid_ns);
+ if (tsk)
+ get_task_struct(tsk);
+ rcu_read_unlock();
+ if (!tsk)
+ return;
+ } else {
+ tsk = current;
+ }
+ spin_lock(&pfault_lock);
+ if (subcode & 0x0080) {
+ /* signal bit is set -> a page has been swapped in by VM */
+ if (tsk->thread.pfault_wait == 1) {
+ /* Initial interrupt was faster than the completion
+ * interrupt. pfault_wait is valid. Set pfault_wait
+ * back to zero and wake up the process. This can
+ * safely be done because the task is still sleeping
+ * and can't produce new pfaults. */
+ tsk->thread.pfault_wait = 0;
+ list_del(&tsk->thread.list);
+ wake_up_process(tsk);
+ put_task_struct(tsk);
+ } else {
+ /* Completion interrupt was faster than initial
+ * interrupt. Set pfault_wait to -1 so the initial
+ * interrupt doesn't put the task to sleep.
+ * If the task is not running, ignore the completion
+ * interrupt since it must be a leftover of a PFAULT
+ * CANCEL operation which didn't remove all pending
+ * completion interrupts. */
+ if (tsk->state == TASK_RUNNING)
+ tsk->thread.pfault_wait = -1;
+ }
+ put_task_struct(tsk);
+ } else {
+ /* signal bit not set -> a real page is missing. */
+ if (tsk->thread.pfault_wait == 1) {
+ /* Already on the list with a reference: put to sleep */
+ set_task_state(tsk, TASK_UNINTERRUPTIBLE);
+ set_tsk_need_resched(tsk);
+ } else if (tsk->thread.pfault_wait == -1) {
+ /* Completion interrupt was faster than the initial
+ * interrupt (pfault_wait == -1). Set pfault_wait
+ * back to zero and exit. */
+ tsk->thread.pfault_wait = 0;
+ } else {
+ /* Initial interrupt arrived before completion
+ * interrupt. Let the task sleep.
+ * An extra task reference is needed since a different
+ * cpu may set the task state to TASK_RUNNING again
+ * before the scheduler is reached. */
+ get_task_struct(tsk);
+ tsk->thread.pfault_wait = 1;
+ list_add(&tsk->thread.list, &pfault_list);
+ set_task_state(tsk, TASK_UNINTERRUPTIBLE);
+ set_tsk_need_resched(tsk);
+ }
+ }
+ spin_unlock(&pfault_lock);
+}
+
+static int __cpuinit pfault_cpu_notify(struct notifier_block *self,
+ unsigned long action, void *hcpu)
+{
+ struct thread_struct *thread, *next;
+ struct task_struct *tsk;
+
+ switch (action) {
+ case CPU_DEAD:
+ case CPU_DEAD_FROZEN:
+ spin_lock_irq(&pfault_lock);
+ list_for_each_entry_safe(thread, next, &pfault_list, list) {
+ thread->pfault_wait = 0;
+ list_del(&thread->list);
+ tsk = container_of(thread, struct task_struct, thread);
+ wake_up_process(tsk);
+ put_task_struct(tsk);
+ }
+ spin_unlock_irq(&pfault_lock);
+ break;
+ default:
+ break;
+ }
+ return NOTIFY_OK;
+}
+
+static int __init pfault_irq_init(void)
+{
+ int rc;
+
+ rc = register_external_interrupt(0x2603, pfault_interrupt);
+ if (rc)
+ goto out_extint;
+ rc = pfault_init() == 0 ? 0 : -EOPNOTSUPP;
+ if (rc)
+ goto out_pfault;
+ service_subclass_irq_register();
+ hotcpu_notifier(pfault_cpu_notify, 0);
+ return 0;
+
+out_pfault:
+ unregister_external_interrupt(0x2603, pfault_interrupt);
+out_extint:
+ pfault_disable = 1;
+ return rc;
+}
+early_initcall(pfault_irq_init);
+
+#endif /* CONFIG_PFAULT */
diff --git a/arch/s390/mm/gup.c b/arch/s390/mm/gup.c
new file mode 100644
index 00000000..65cb06e2
--- /dev/null
+++ b/arch/s390/mm/gup.c
@@ -0,0 +1,236 @@
+/*
+ * Lockless get_user_pages_fast for s390
+ *
+ * Copyright IBM Corp. 2010
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <linux/hugetlb.h>
+#include <linux/vmstat.h>
+#include <linux/pagemap.h>
+#include <linux/rwsem.h>
+#include <asm/pgtable.h>
+
+/*
+ * The performance critical leaf functions are made noinline otherwise gcc
+ * inlines everything into a single function which results in too much
+ * register pressure.
+ */
+static inline int gup_pte_range(pmd_t *pmdp, pmd_t pmd, unsigned long addr,
+ unsigned long end, int write, struct page **pages, int *nr)
+{
+ unsigned long mask;
+ pte_t *ptep, pte;
+ struct page *page;
+
+ mask = (write ? _PAGE_RO : 0) | _PAGE_INVALID | _PAGE_SPECIAL;
+
+ ptep = ((pte_t *) pmd_deref(pmd)) + pte_index(addr);
+ do {
+ pte = *ptep;
+ barrier();
+ if ((pte_val(pte) & mask) != 0)
+ return 0;
+ VM_BUG_ON(!pfn_valid(pte_pfn(pte)));
+ page = pte_page(pte);
+ if (!page_cache_get_speculative(page))
+ return 0;
+ if (unlikely(pte_val(pte) != pte_val(*ptep))) {
+ put_page(page);
+ return 0;
+ }
+ pages[*nr] = page;
+ (*nr)++;
+
+ } while (ptep++, addr += PAGE_SIZE, addr != end);
+
+ return 1;
+}
+
+static inline int gup_huge_pmd(pmd_t *pmdp, pmd_t pmd, unsigned long addr,
+ unsigned long end, int write, struct page **pages, int *nr)
+{
+ unsigned long mask, result;
+ struct page *head, *page, *tail;
+ int refs;
+
+ result = write ? 0 : _SEGMENT_ENTRY_RO;
+ mask = result | _SEGMENT_ENTRY_INV;
+ if ((pmd_val(pmd) & mask) != result)
+ return 0;
+ VM_BUG_ON(!pfn_valid(pmd_val(pmd) >> PAGE_SHIFT));
+
+ refs = 0;
+ head = pmd_page(pmd);
+ page = head + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
+ tail = page;
+ do {
+ VM_BUG_ON(compound_head(page) != head);
+ pages[*nr] = page;
+ (*nr)++;
+ page++;
+ refs++;
+ } while (addr += PAGE_SIZE, addr != end);
+
+ if (!page_cache_add_speculative(head, refs)) {
+ *nr -= refs;
+ return 0;
+ }
+
+ if (unlikely(pmd_val(pmd) != pmd_val(*pmdp))) {
+ *nr -= refs;
+ while (refs--)
+ put_page(head);
+ return 0;
+ }
+
+ /*
+ * Any tail page need their mapcount reference taken before we
+ * return.
+ */
+ while (refs--) {
+ if (PageTail(tail))
+ get_huge_page_tail(tail);
+ tail++;
+ }
+
+ return 1;
+}
+
+
+static inline int gup_pmd_range(pud_t *pudp, pud_t pud, unsigned long addr,
+ unsigned long end, int write, struct page **pages, int *nr)
+{
+ unsigned long next;
+ pmd_t *pmdp, pmd;
+
+ pmdp = (pmd_t *) pudp;
+#ifdef CONFIG_64BIT
+ if ((pud_val(pud) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R3)
+ pmdp = (pmd_t *) pud_deref(pud);
+ pmdp += pmd_index(addr);
+#endif
+ do {
+ pmd = *pmdp;
+ barrier();
+ next = pmd_addr_end(addr, end);
+ if (pmd_none(pmd))
+ return 0;
+ if (unlikely(pmd_huge(pmd))) {
+ if (!gup_huge_pmd(pmdp, pmd, addr, next,
+ write, pages, nr))
+ return 0;
+ } else if (!gup_pte_range(pmdp, pmd, addr, next,
+ write, pages, nr))
+ return 0;
+ } while (pmdp++, addr = next, addr != end);
+
+ return 1;
+}
+
+static inline int gup_pud_range(pgd_t *pgdp, pgd_t pgd, unsigned long addr,
+ unsigned long end, int write, struct page **pages, int *nr)
+{
+ unsigned long next;
+ pud_t *pudp, pud;
+
+ pudp = (pud_t *) pgdp;
+#ifdef CONFIG_64BIT
+ if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R2)
+ pudp = (pud_t *) pgd_deref(pgd);
+ pudp += pud_index(addr);
+#endif
+ do {
+ pud = *pudp;
+ barrier();
+ next = pud_addr_end(addr, end);
+ if (pud_none(pud))
+ return 0;
+ if (!gup_pmd_range(pudp, pud, addr, next, write, pages, nr))
+ return 0;
+ } while (pudp++, addr = next, addr != end);
+
+ return 1;
+}
+
+/**
+ * get_user_pages_fast() - pin user pages in memory
+ * @start: starting user address
+ * @nr_pages: number of pages from start to pin
+ * @write: whether pages will be written to
+ * @pages: array that receives pointers to the pages pinned.
+ * Should be at least nr_pages long.
+ *
+ * Attempt to pin user pages in memory without taking mm->mmap_sem.
+ * If not successful, it will fall back to taking the lock and
+ * calling get_user_pages().
+ *
+ * Returns number of pages pinned. This may be fewer than the number
+ * requested. If nr_pages is 0 or negative, returns 0. If no pages
+ * were pinned, returns -errno.
+ */
+int get_user_pages_fast(unsigned long start, int nr_pages, int write,
+ struct page **pages)
+{
+ struct mm_struct *mm = current->mm;
+ unsigned long addr, len, end;
+ unsigned long next;
+ pgd_t *pgdp, pgd;
+ int nr = 0;
+
+ start &= PAGE_MASK;
+ addr = start;
+ len = (unsigned long) nr_pages << PAGE_SHIFT;
+ end = start + len;
+ if (end < start)
+ goto slow_irqon;
+
+ /*
+ * local_irq_disable() doesn't prevent pagetable teardown, but does
+ * prevent the pagetables from being freed on s390.
+ *
+ * So long as we atomically load page table pointers versus teardown,
+ * we can follow the address down to the the page and take a ref on it.
+ */
+ local_irq_disable();
+ pgdp = pgd_offset(mm, addr);
+ do {
+ pgd = *pgdp;
+ barrier();
+ next = pgd_addr_end(addr, end);
+ if (pgd_none(pgd))
+ goto slow;
+ if (!gup_pud_range(pgdp, pgd, addr, next, write, pages, &nr))
+ goto slow;
+ } while (pgdp++, addr = next, addr != end);
+ local_irq_enable();
+
+ VM_BUG_ON(nr != (end - start) >> PAGE_SHIFT);
+ return nr;
+
+ {
+ int ret;
+slow:
+ local_irq_enable();
+slow_irqon:
+ /* Try to get the remaining pages with get_user_pages */
+ start += nr << PAGE_SHIFT;
+ pages += nr;
+
+ down_read(&mm->mmap_sem);
+ ret = get_user_pages(current, mm, start,
+ (end - start) >> PAGE_SHIFT, write, 0, pages, NULL);
+ up_read(&mm->mmap_sem);
+
+ /* Have to be a bit careful with return values */
+ if (nr > 0) {
+ if (ret < 0)
+ ret = nr;
+ else
+ ret += nr;
+ }
+
+ return ret;
+ }
+}
diff --git a/arch/s390/mm/hugetlbpage.c b/arch/s390/mm/hugetlbpage.c
new file mode 100644
index 00000000..597bb2d2
--- /dev/null
+++ b/arch/s390/mm/hugetlbpage.c
@@ -0,0 +1,130 @@
+/*
+ * IBM System z Huge TLB Page Support for Kernel.
+ *
+ * Copyright 2007 IBM Corp.
+ * Author(s): Gerald Schaefer <gerald.schaefer@de.ibm.com>
+ */
+
+#include <linux/mm.h>
+#include <linux/hugetlb.h>
+
+
+void set_huge_pte_at(struct mm_struct *mm, unsigned long addr,
+ pte_t *pteptr, pte_t pteval)
+{
+ pmd_t *pmdp = (pmd_t *) pteptr;
+ unsigned long mask;
+
+ if (!MACHINE_HAS_HPAGE) {
+ pteptr = (pte_t *) pte_page(pteval)[1].index;
+ mask = pte_val(pteval) &
+ (_SEGMENT_ENTRY_INV | _SEGMENT_ENTRY_RO);
+ pte_val(pteval) = (_SEGMENT_ENTRY + __pa(pteptr)) | mask;
+ }
+
+ pmd_val(*pmdp) = pte_val(pteval);
+}
+
+int arch_prepare_hugepage(struct page *page)
+{
+ unsigned long addr = page_to_phys(page);
+ pte_t pte;
+ pte_t *ptep;
+ int i;
+
+ if (MACHINE_HAS_HPAGE)
+ return 0;
+
+ ptep = (pte_t *) pte_alloc_one(&init_mm, addr);
+ if (!ptep)
+ return -ENOMEM;
+
+ pte = mk_pte(page, PAGE_RW);
+ for (i = 0; i < PTRS_PER_PTE; i++) {
+ set_pte_at(&init_mm, addr + i * PAGE_SIZE, ptep + i, pte);
+ pte_val(pte) += PAGE_SIZE;
+ }
+ page[1].index = (unsigned long) ptep;
+ return 0;
+}
+
+void arch_release_hugepage(struct page *page)
+{
+ pte_t *ptep;
+
+ if (MACHINE_HAS_HPAGE)
+ return;
+
+ ptep = (pte_t *) page[1].index;
+ if (!ptep)
+ return;
+ page_table_free(&init_mm, (unsigned long *) ptep);
+ page[1].index = 0;
+}
+
+pte_t *huge_pte_alloc(struct mm_struct *mm,
+ unsigned long addr, unsigned long sz)
+{
+ pgd_t *pgdp;
+ pud_t *pudp;
+ pmd_t *pmdp = NULL;
+
+ pgdp = pgd_offset(mm, addr);
+ pudp = pud_alloc(mm, pgdp, addr);
+ if (pudp)
+ pmdp = pmd_alloc(mm, pudp, addr);
+ return (pte_t *) pmdp;
+}
+
+pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr)
+{
+ pgd_t *pgdp;
+ pud_t *pudp;
+ pmd_t *pmdp = NULL;
+
+ pgdp = pgd_offset(mm, addr);
+ if (pgd_present(*pgdp)) {
+ pudp = pud_offset(pgdp, addr);
+ if (pud_present(*pudp))
+ pmdp = pmd_offset(pudp, addr);
+ }
+ return (pte_t *) pmdp;
+}
+
+int huge_pmd_unshare(struct mm_struct *mm, unsigned long *addr, pte_t *ptep)
+{
+ return 0;
+}
+
+struct page *follow_huge_addr(struct mm_struct *mm, unsigned long address,
+ int write)
+{
+ return ERR_PTR(-EINVAL);
+}
+
+int pmd_huge(pmd_t pmd)
+{
+ if (!MACHINE_HAS_HPAGE)
+ return 0;
+
+ return !!(pmd_val(pmd) & _SEGMENT_ENTRY_LARGE);
+}
+
+int pud_huge(pud_t pud)
+{
+ return 0;
+}
+
+struct page *follow_huge_pmd(struct mm_struct *mm, unsigned long address,
+ pmd_t *pmdp, int write)
+{
+ struct page *page;
+
+ if (!MACHINE_HAS_HPAGE)
+ return NULL;
+
+ page = pmd_page(*pmdp);
+ if (page)
+ page += ((address & ~HPAGE_MASK) >> PAGE_SHIFT);
+ return page;
+}
diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
new file mode 100644
index 00000000..2bea0605
--- /dev/null
+++ b/arch/s390/mm/init.c
@@ -0,0 +1,262 @@
+/*
+ * arch/s390/mm/init.c
+ *
+ * S390 version
+ * Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ * Author(s): Hartmut Penner (hp@de.ibm.com)
+ *
+ * Derived from "arch/i386/mm/init.c"
+ * Copyright (C) 1995 Linus Torvalds
+ */
+
+#include <linux/signal.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/ptrace.h>
+#include <linux/mman.h>
+#include <linux/mm.h>
+#include <linux/swap.h>
+#include <linux/smp.h>
+#include <linux/init.h>
+#include <linux/pagemap.h>
+#include <linux/bootmem.h>
+#include <linux/pfn.h>
+#include <linux/poison.h>
+#include <linux/initrd.h>
+#include <linux/export.h>
+#include <linux/gfp.h>
+#include <asm/processor.h>
+#include <asm/uaccess.h>
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <asm/dma.h>
+#include <asm/lowcore.h>
+#include <asm/tlb.h>
+#include <asm/tlbflush.h>
+#include <asm/sections.h>
+#include <asm/ctl_reg.h>
+
+pgd_t swapper_pg_dir[PTRS_PER_PGD] __attribute__((__aligned__(PAGE_SIZE)));
+
+unsigned long empty_zero_page, zero_page_mask;
+EXPORT_SYMBOL(empty_zero_page);
+
+static unsigned long setup_zero_pages(void)
+{
+ struct cpuid cpu_id;
+ unsigned int order;
+ unsigned long size;
+ struct page *page;
+ int i;
+
+ get_cpu_id(&cpu_id);
+ switch (cpu_id.machine) {
+ case 0x9672: /* g5 */
+ case 0x2064: /* z900 */
+ case 0x2066: /* z900 */
+ case 0x2084: /* z990 */
+ case 0x2086: /* z990 */
+ case 0x2094: /* z9-109 */
+ case 0x2096: /* z9-109 */
+ order = 0;
+ break;
+ case 0x2097: /* z10 */
+ case 0x2098: /* z10 */
+ default:
+ order = 2;
+ break;
+ }
+
+ empty_zero_page = __get_free_pages(GFP_KERNEL | __GFP_ZERO, order);
+ if (!empty_zero_page)
+ panic("Out of memory in setup_zero_pages");
+
+ page = virt_to_page((void *) empty_zero_page);
+ split_page(page, order);
+ for (i = 1 << order; i > 0; i--) {
+ SetPageReserved(page);
+ page++;
+ }
+
+ size = PAGE_SIZE << order;
+ zero_page_mask = (size - 1) & PAGE_MASK;
+
+ return 1UL << order;
+}
+
+/*
+ * paging_init() sets up the page tables
+ */
+void __init paging_init(void)
+{
+ unsigned long max_zone_pfns[MAX_NR_ZONES];
+ unsigned long pgd_type, asce_bits;
+
+ init_mm.pgd = swapper_pg_dir;
+#ifdef CONFIG_64BIT
+ if (VMALLOC_END > (1UL << 42)) {
+ asce_bits = _ASCE_TYPE_REGION2 | _ASCE_TABLE_LENGTH;
+ pgd_type = _REGION2_ENTRY_EMPTY;
+ } else {
+ asce_bits = _ASCE_TYPE_REGION3 | _ASCE_TABLE_LENGTH;
+ pgd_type = _REGION3_ENTRY_EMPTY;
+ }
+#else
+ asce_bits = _ASCE_TABLE_LENGTH;
+ pgd_type = _SEGMENT_ENTRY_EMPTY;
+#endif
+ S390_lowcore.kernel_asce = (__pa(init_mm.pgd) & PAGE_MASK) | asce_bits;
+ clear_table((unsigned long *) init_mm.pgd, pgd_type,
+ sizeof(unsigned long)*2048);
+ vmem_map_init();
+
+ /* enable virtual mapping in kernel mode */
+ __ctl_load(S390_lowcore.kernel_asce, 1, 1);
+ __ctl_load(S390_lowcore.kernel_asce, 7, 7);
+ __ctl_load(S390_lowcore.kernel_asce, 13, 13);
+ arch_local_irq_restore(4UL << (BITS_PER_LONG - 8));
+
+ atomic_set(&init_mm.context.attach_count, 1);
+
+ sparse_memory_present_with_active_regions(MAX_NUMNODES);
+ sparse_init();
+ memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
+ max_zone_pfns[ZONE_DMA] = PFN_DOWN(MAX_DMA_ADDRESS);
+ max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
+ free_area_init_nodes(max_zone_pfns);
+ fault_init();
+}
+
+void __init mem_init(void)
+{
+ unsigned long codesize, reservedpages, datasize, initsize;
+
+ max_mapnr = num_physpages = max_low_pfn;
+ high_memory = (void *) __va(max_low_pfn * PAGE_SIZE);
+
+ /* Setup guest page hinting */
+ cmma_init();
+
+ /* this will put all low memory onto the freelists */
+ totalram_pages += free_all_bootmem();
+ totalram_pages -= setup_zero_pages(); /* Setup zeroed pages. */
+
+ reservedpages = 0;
+
+ codesize = (unsigned long) &_etext - (unsigned long) &_text;
+ datasize = (unsigned long) &_edata - (unsigned long) &_etext;
+ initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
+ printk("Memory: %luk/%luk available (%ldk kernel code, %ldk reserved, %ldk data, %ldk init)\n",
+ nr_free_pages() << (PAGE_SHIFT-10),
+ max_mapnr << (PAGE_SHIFT-10),
+ codesize >> 10,
+ reservedpages << (PAGE_SHIFT-10),
+ datasize >>10,
+ initsize >> 10);
+ printk("Write protected kernel read-only data: %#lx - %#lx\n",
+ (unsigned long)&_stext,
+ PFN_ALIGN((unsigned long)&_eshared) - 1);
+}
+
+#ifdef CONFIG_DEBUG_PAGEALLOC
+void kernel_map_pages(struct page *page, int numpages, int enable)
+{
+ pgd_t *pgd;
+ pud_t *pud;
+ pmd_t *pmd;
+ pte_t *pte;
+ unsigned long address;
+ int i;
+
+ for (i = 0; i < numpages; i++) {
+ address = page_to_phys(page + i);
+ pgd = pgd_offset_k(address);
+ pud = pud_offset(pgd, address);
+ pmd = pmd_offset(pud, address);
+ pte = pte_offset_kernel(pmd, address);
+ if (!enable) {
+ __ptep_ipte(address, pte);
+ pte_val(*pte) = _PAGE_TYPE_EMPTY;
+ continue;
+ }
+ *pte = mk_pte_phys(address, __pgprot(_PAGE_TYPE_RW));
+ /* Flush cpu write queue. */
+ mb();
+ }
+}
+#endif
+
+void free_init_pages(char *what, unsigned long begin, unsigned long end)
+{
+ unsigned long addr = begin;
+
+ if (begin >= end)
+ return;
+ for (; addr < end; addr += PAGE_SIZE) {
+ ClearPageReserved(virt_to_page(addr));
+ init_page_count(virt_to_page(addr));
+ memset((void *)(addr & PAGE_MASK), POISON_FREE_INITMEM,
+ PAGE_SIZE);
+ free_page(addr);
+ totalram_pages++;
+ }
+ printk(KERN_INFO "Freeing %s: %luk freed\n", what, (end - begin) >> 10);
+}
+
+void free_initmem(void)
+{
+ free_init_pages("unused kernel memory",
+ (unsigned long)&__init_begin,
+ (unsigned long)&__init_end);
+}
+
+#ifdef CONFIG_BLK_DEV_INITRD
+void free_initrd_mem(unsigned long start, unsigned long end)
+{
+ free_init_pages("initrd memory", start, end);
+}
+#endif
+
+#ifdef CONFIG_MEMORY_HOTPLUG
+int arch_add_memory(int nid, u64 start, u64 size)
+{
+ unsigned long zone_start_pfn, zone_end_pfn, nr_pages;
+ unsigned long start_pfn = PFN_DOWN(start);
+ unsigned long size_pages = PFN_DOWN(size);
+ struct zone *zone;
+ int rc;
+
+ rc = vmem_add_mapping(start, size);
+ if (rc)
+ return rc;
+ for_each_zone(zone) {
+ if (zone_idx(zone) != ZONE_MOVABLE) {
+ /* Add range within existing zone limits */
+ zone_start_pfn = zone->zone_start_pfn;
+ zone_end_pfn = zone->zone_start_pfn +
+ zone->spanned_pages;
+ } else {
+ /* Add remaining range to ZONE_MOVABLE */
+ zone_start_pfn = start_pfn;
+ zone_end_pfn = start_pfn + size_pages;
+ }
+ if (start_pfn < zone_start_pfn || start_pfn >= zone_end_pfn)
+ continue;
+ nr_pages = (start_pfn + size_pages > zone_end_pfn) ?
+ zone_end_pfn - start_pfn : size_pages;
+ rc = __add_pages(nid, zone, start_pfn, nr_pages);
+ if (rc)
+ break;
+ start_pfn += nr_pages;
+ size_pages -= nr_pages;
+ if (!size_pages)
+ break;
+ }
+ if (rc)
+ vmem_remove_mapping(start, size);
+ return rc;
+}
+#endif /* CONFIG_MEMORY_HOTPLUG */
diff --git a/arch/s390/mm/maccess.c b/arch/s390/mm/maccess.c
new file mode 100644
index 00000000..e1335dc2
--- /dev/null
+++ b/arch/s390/mm/maccess.c
@@ -0,0 +1,168 @@
+/*
+ * Access kernel memory without faulting -- s390 specific implementation.
+ *
+ * Copyright IBM Corp. 2009
+ *
+ * Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>,
+ *
+ */
+
+#include <linux/uaccess.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/gfp.h>
+#include <asm/ctl_reg.h>
+
+/*
+ * This function writes to kernel memory bypassing DAT and possible
+ * write protection. It copies one to four bytes from src to dst
+ * using the stura instruction.
+ * Returns the number of bytes copied or -EFAULT.
+ */
+static long probe_kernel_write_odd(void *dst, const void *src, size_t size)
+{
+ unsigned long count, aligned;
+ int offset, mask;
+ int rc = -EFAULT;
+
+ aligned = (unsigned long) dst & ~3UL;
+ offset = (unsigned long) dst & 3;
+ count = min_t(unsigned long, 4 - offset, size);
+ mask = (0xf << (4 - count)) & 0xf;
+ mask >>= offset;
+ asm volatile(
+ " bras 1,0f\n"
+ " icm 0,0,0(%3)\n"
+ "0: l 0,0(%1)\n"
+ " lra %1,0(%1)\n"
+ "1: ex %2,0(1)\n"
+ "2: stura 0,%1\n"
+ " la %0,0\n"
+ "3:\n"
+ EX_TABLE(0b,3b) EX_TABLE(1b,3b) EX_TABLE(2b,3b)
+ : "+d" (rc), "+a" (aligned)
+ : "a" (mask), "a" (src) : "cc", "memory", "0", "1");
+ return rc ? rc : count;
+}
+
+long probe_kernel_write(void *dst, const void *src, size_t size)
+{
+ long copied = 0;
+
+ while (size) {
+ copied = probe_kernel_write_odd(dst, src, size);
+ if (copied < 0)
+ break;
+ dst += copied;
+ src += copied;
+ size -= copied;
+ }
+ return copied < 0 ? -EFAULT : 0;
+}
+
+static int __memcpy_real(void *dest, void *src, size_t count)
+{
+ register unsigned long _dest asm("2") = (unsigned long) dest;
+ register unsigned long _len1 asm("3") = (unsigned long) count;
+ register unsigned long _src asm("4") = (unsigned long) src;
+ register unsigned long _len2 asm("5") = (unsigned long) count;
+ int rc = -EFAULT;
+
+ asm volatile (
+ "0: mvcle %1,%2,0x0\n"
+ "1: jo 0b\n"
+ " lhi %0,0x0\n"
+ "2:\n"
+ EX_TABLE(1b,2b)
+ : "+d" (rc), "+d" (_dest), "+d" (_src), "+d" (_len1),
+ "+d" (_len2), "=m" (*((long *) dest))
+ : "m" (*((long *) src))
+ : "cc", "memory");
+ return rc;
+}
+
+/*
+ * Copy memory in real mode (kernel to kernel)
+ */
+int memcpy_real(void *dest, void *src, size_t count)
+{
+ unsigned long flags;
+ int rc;
+
+ if (!count)
+ return 0;
+ local_irq_save(flags);
+ __arch_local_irq_stnsm(0xfbUL);
+ rc = __memcpy_real(dest, src, count);
+ local_irq_restore(flags);
+ return rc;
+}
+
+/*
+ * Copy memory to absolute zero
+ */
+void copy_to_absolute_zero(void *dest, void *src, size_t count)
+{
+ unsigned long cr0;
+
+ BUG_ON((unsigned long) dest + count >= sizeof(struct _lowcore));
+ preempt_disable();
+ __ctl_store(cr0, 0, 0);
+ __ctl_clear_bit(0, 28); /* disable lowcore protection */
+ memcpy_real(dest + store_prefix(), src, count);
+ __ctl_load(cr0, 0, 0);
+ preempt_enable();
+}
+
+/*
+ * Copy memory from kernel (real) to user (virtual)
+ */
+int copy_to_user_real(void __user *dest, void *src, size_t count)
+{
+ int offs = 0, size, rc;
+ char *buf;
+
+ buf = (char *) __get_free_page(GFP_KERNEL);
+ if (!buf)
+ return -ENOMEM;
+ rc = -EFAULT;
+ while (offs < count) {
+ size = min(PAGE_SIZE, count - offs);
+ if (memcpy_real(buf, src + offs, size))
+ goto out;
+ if (copy_to_user(dest + offs, buf, size))
+ goto out;
+ offs += size;
+ }
+ rc = 0;
+out:
+ free_page((unsigned long) buf);
+ return rc;
+}
+
+/*
+ * Copy memory from user (virtual) to kernel (real)
+ */
+int copy_from_user_real(void *dest, void __user *src, size_t count)
+{
+ int offs = 0, size, rc;
+ char *buf;
+
+ buf = (char *) __get_free_page(GFP_KERNEL);
+ if (!buf)
+ return -ENOMEM;
+ rc = -EFAULT;
+ while (offs < count) {
+ size = min(PAGE_SIZE, count - offs);
+ if (copy_from_user(buf, src + offs, size))
+ goto out;
+ if (memcpy_real(dest + offs, buf, size))
+ goto out;
+ offs += size;
+ }
+ rc = 0;
+out:
+ free_page((unsigned long) buf);
+ return rc;
+}
diff --git a/arch/s390/mm/mmap.c b/arch/s390/mm/mmap.c
new file mode 100644
index 00000000..2857c484
--- /dev/null
+++ b/arch/s390/mm/mmap.c
@@ -0,0 +1,178 @@
+/*
+ * linux/arch/s390/mm/mmap.c
+ *
+ * flexible mmap layout support
+ *
+ * Copyright 2003-2004 Red Hat Inc., Durham, North Carolina.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ *
+ * Started by Ingo Molnar <mingo@elte.hu>
+ */
+
+#include <linux/personality.h>
+#include <linux/mm.h>
+#include <linux/mman.h>
+#include <linux/module.h>
+#include <linux/random.h>
+#include <linux/compat.h>
+#include <asm/pgalloc.h>
+
+static unsigned long stack_maxrandom_size(void)
+{
+ if (!(current->flags & PF_RANDOMIZE))
+ return 0;
+ if (current->personality & ADDR_NO_RANDOMIZE)
+ return 0;
+ return STACK_RND_MASK << PAGE_SHIFT;
+}
+
+/*
+ * Top of mmap area (just below the process stack).
+ *
+ * Leave at least a ~32 MB hole.
+ */
+#define MIN_GAP (32*1024*1024)
+#define MAX_GAP (STACK_TOP/6*5)
+
+static inline int mmap_is_legacy(void)
+{
+ if (current->personality & ADDR_COMPAT_LAYOUT)
+ return 1;
+ if (rlimit(RLIMIT_STACK) == RLIM_INFINITY)
+ return 1;
+ return sysctl_legacy_va_layout;
+}
+
+static unsigned long mmap_rnd(void)
+{
+ if (!(current->flags & PF_RANDOMIZE))
+ return 0;
+ /* 8MB randomization for mmap_base */
+ return (get_random_int() & 0x7ffUL) << PAGE_SHIFT;
+}
+
+static inline unsigned long mmap_base(void)
+{
+ unsigned long gap = rlimit(RLIMIT_STACK);
+
+ if (gap < MIN_GAP)
+ gap = MIN_GAP;
+ else if (gap > MAX_GAP)
+ gap = MAX_GAP;
+ gap &= PAGE_MASK;
+ return STACK_TOP - stack_maxrandom_size() - mmap_rnd() - gap;
+}
+
+#ifndef CONFIG_64BIT
+
+/*
+ * This function, called very early during the creation of a new
+ * process VM image, sets up which VM layout function to use:
+ */
+void arch_pick_mmap_layout(struct mm_struct *mm)
+{
+ /*
+ * Fall back to the standard layout if the personality
+ * bit is set, or if the expected stack growth is unlimited:
+ */
+ if (mmap_is_legacy()) {
+ mm->mmap_base = TASK_UNMAPPED_BASE;
+ mm->get_unmapped_area = arch_get_unmapped_area;
+ mm->unmap_area = arch_unmap_area;
+ } else {
+ mm->mmap_base = mmap_base();
+ mm->get_unmapped_area = arch_get_unmapped_area_topdown;
+ mm->unmap_area = arch_unmap_area_topdown;
+ }
+}
+
+#else
+
+int s390_mmap_check(unsigned long addr, unsigned long len)
+{
+ if (!is_compat_task() &&
+ len >= TASK_SIZE && TASK_SIZE < (1UL << 53))
+ return crst_table_upgrade(current->mm, 1UL << 53);
+ return 0;
+}
+
+static unsigned long
+s390_get_unmapped_area(struct file *filp, unsigned long addr,
+ unsigned long len, unsigned long pgoff, unsigned long flags)
+{
+ struct mm_struct *mm = current->mm;
+ unsigned long area;
+ int rc;
+
+ area = arch_get_unmapped_area(filp, addr, len, pgoff, flags);
+ if (!(area & ~PAGE_MASK))
+ return area;
+ if (area == -ENOMEM && !is_compat_task() && TASK_SIZE < (1UL << 53)) {
+ /* Upgrade the page table to 4 levels and retry. */
+ rc = crst_table_upgrade(mm, 1UL << 53);
+ if (rc)
+ return (unsigned long) rc;
+ area = arch_get_unmapped_area(filp, addr, len, pgoff, flags);
+ }
+ return area;
+}
+
+static unsigned long
+s390_get_unmapped_area_topdown(struct file *filp, const unsigned long addr,
+ const unsigned long len, const unsigned long pgoff,
+ const unsigned long flags)
+{
+ struct mm_struct *mm = current->mm;
+ unsigned long area;
+ int rc;
+
+ area = arch_get_unmapped_area_topdown(filp, addr, len, pgoff, flags);
+ if (!(area & ~PAGE_MASK))
+ return area;
+ if (area == -ENOMEM && !is_compat_task() && TASK_SIZE < (1UL << 53)) {
+ /* Upgrade the page table to 4 levels and retry. */
+ rc = crst_table_upgrade(mm, 1UL << 53);
+ if (rc)
+ return (unsigned long) rc;
+ area = arch_get_unmapped_area_topdown(filp, addr, len,
+ pgoff, flags);
+ }
+ return area;
+}
+/*
+ * This function, called very early during the creation of a new
+ * process VM image, sets up which VM layout function to use:
+ */
+void arch_pick_mmap_layout(struct mm_struct *mm)
+{
+ /*
+ * Fall back to the standard layout if the personality
+ * bit is set, or if the expected stack growth is unlimited:
+ */
+ if (mmap_is_legacy()) {
+ mm->mmap_base = TASK_UNMAPPED_BASE;
+ mm->get_unmapped_area = s390_get_unmapped_area;
+ mm->unmap_area = arch_unmap_area;
+ } else {
+ mm->mmap_base = mmap_base();
+ mm->get_unmapped_area = s390_get_unmapped_area_topdown;
+ mm->unmap_area = arch_unmap_area_topdown;
+ }
+}
+
+#endif
diff --git a/arch/s390/mm/page-states.c b/arch/s390/mm/page-states.c
new file mode 100644
index 00000000..a90d45e9
--- /dev/null
+++ b/arch/s390/mm/page-states.c
@@ -0,0 +1,114 @@
+/*
+ * 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/gfp.h>
+#include <linux/init.h>
+
+#define ESSA_SET_STABLE 1
+#define ESSA_SET_UNUSED 2
+
+static int cmma_flag = 1;
+
+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;
+}
+
+static inline void set_page_unstable(struct page *page, int order)
+{
+ int i, rc;
+
+ for (i = 0; i < (1 << order); i++)
+ asm volatile(".insn rrf,0xb9ab0000,%0,%1,%2,0"
+ : "=&d" (rc)
+ : "a" (page_to_phys(page + i)),
+ "i" (ESSA_SET_UNUSED));
+}
+
+void arch_free_page(struct page *page, int order)
+{
+ if (!cmma_flag)
+ return;
+ set_page_unstable(page, order);
+}
+
+static inline void set_page_stable(struct page *page, int order)
+{
+ int i, rc;
+
+ for (i = 0; i < (1 << order); i++)
+ asm volatile(".insn rrf,0xb9ab0000,%0,%1,%2,0"
+ : "=&d" (rc)
+ : "a" (page_to_phys(page + i)),
+ "i" (ESSA_SET_STABLE));
+}
+
+void arch_alloc_page(struct page *page, int order)
+{
+ if (!cmma_flag)
+ return;
+ set_page_stable(page, order);
+}
+
+void arch_set_page_states(int make_stable)
+{
+ unsigned long flags, order, t;
+ struct list_head *l;
+ struct page *page;
+ struct zone *zone;
+
+ if (!cmma_flag)
+ return;
+ if (make_stable)
+ drain_local_pages(NULL);
+ for_each_populated_zone(zone) {
+ spin_lock_irqsave(&zone->lock, flags);
+ for_each_migratetype_order(order, t) {
+ list_for_each(l, &zone->free_area[order].free_list[t]) {
+ page = list_entry(l, struct page, lru);
+ if (make_stable)
+ set_page_stable(page, order);
+ else
+ set_page_unstable(page, order);
+ }
+ }
+ spin_unlock_irqrestore(&zone->lock, flags);
+ }
+}
diff --git a/arch/s390/mm/pageattr.c b/arch/s390/mm/pageattr.c
new file mode 100644
index 00000000..b36537a5
--- /dev/null
+++ b/arch/s390/mm/pageattr.c
@@ -0,0 +1,62 @@
+/*
+ * Copyright IBM Corp. 2011
+ * Author(s): Jan Glauber <jang@linux.vnet.ibm.com>
+ */
+#include <linux/module.h>
+#include <linux/mm.h>
+#include <linux/hugetlb.h>
+#include <asm/cacheflush.h>
+#include <asm/pgtable.h>
+
+static void change_page_attr(unsigned long addr, int numpages,
+ pte_t (*set) (pte_t))
+{
+ pte_t *ptep, pte;
+ pmd_t *pmdp;
+ pud_t *pudp;
+ pgd_t *pgdp;
+ int i;
+
+ for (i = 0; i < numpages; i++) {
+ pgdp = pgd_offset(&init_mm, addr);
+ pudp = pud_offset(pgdp, addr);
+ pmdp = pmd_offset(pudp, addr);
+ if (pmd_huge(*pmdp)) {
+ WARN_ON_ONCE(1);
+ continue;
+ }
+ ptep = pte_offset_kernel(pmdp, addr);
+
+ pte = *ptep;
+ pte = set(pte);
+ __ptep_ipte(addr, ptep);
+ *ptep = pte;
+ addr += PAGE_SIZE;
+ }
+}
+
+int set_memory_ro(unsigned long addr, int numpages)
+{
+ change_page_attr(addr, numpages, pte_wrprotect);
+ return 0;
+}
+EXPORT_SYMBOL_GPL(set_memory_ro);
+
+int set_memory_rw(unsigned long addr, int numpages)
+{
+ change_page_attr(addr, numpages, pte_mkwrite);
+ return 0;
+}
+EXPORT_SYMBOL_GPL(set_memory_rw);
+
+/* not possible */
+int set_memory_nx(unsigned long addr, int numpages)
+{
+ return 0;
+}
+EXPORT_SYMBOL_GPL(set_memory_nx);
+
+int set_memory_x(unsigned long addr, int numpages)
+{
+ return 0;
+}
diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c
new file mode 100644
index 00000000..6e765bf0
--- /dev/null
+++ b/arch/s390/mm/pgtable.c
@@ -0,0 +1,871 @@
+/*
+ * Copyright IBM Corp. 2007,2011
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/gfp.h>
+#include <linux/mm.h>
+#include <linux/swap.h>
+#include <linux/smp.h>
+#include <linux/highmem.h>
+#include <linux/pagemap.h>
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/quicklist.h>
+#include <linux/rcupdate.h>
+#include <linux/slab.h>
+
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <asm/tlb.h>
+#include <asm/tlbflush.h>
+#include <asm/mmu_context.h>
+
+#ifndef CONFIG_64BIT
+#define ALLOC_ORDER 1
+#define FRAG_MASK 0x0f
+#else
+#define ALLOC_ORDER 2
+#define FRAG_MASK 0x03
+#endif
+
+
+unsigned long *crst_table_alloc(struct mm_struct *mm)
+{
+ struct page *page = alloc_pages(GFP_KERNEL, ALLOC_ORDER);
+
+ if (!page)
+ return NULL;
+ return (unsigned long *) page_to_phys(page);
+}
+
+void crst_table_free(struct mm_struct *mm, unsigned long *table)
+{
+ free_pages((unsigned long) table, ALLOC_ORDER);
+}
+
+#ifdef CONFIG_64BIT
+int crst_table_upgrade(struct mm_struct *mm, unsigned long limit)
+{
+ unsigned long *table, *pgd;
+ unsigned long entry;
+
+ BUG_ON(limit > (1UL << 53));
+repeat:
+ table = crst_table_alloc(mm);
+ if (!table)
+ return -ENOMEM;
+ spin_lock_bh(&mm->page_table_lock);
+ if (mm->context.asce_limit < limit) {
+ pgd = (unsigned long *) mm->pgd;
+ if (mm->context.asce_limit <= (1UL << 31)) {
+ entry = _REGION3_ENTRY_EMPTY;
+ mm->context.asce_limit = 1UL << 42;
+ mm->context.asce_bits = _ASCE_TABLE_LENGTH |
+ _ASCE_USER_BITS |
+ _ASCE_TYPE_REGION3;
+ } else {
+ entry = _REGION2_ENTRY_EMPTY;
+ mm->context.asce_limit = 1UL << 53;
+ mm->context.asce_bits = _ASCE_TABLE_LENGTH |
+ _ASCE_USER_BITS |
+ _ASCE_TYPE_REGION2;
+ }
+ crst_table_init(table, entry);
+ pgd_populate(mm, (pgd_t *) table, (pud_t *) pgd);
+ mm->pgd = (pgd_t *) table;
+ mm->task_size = mm->context.asce_limit;
+ table = NULL;
+ }
+ spin_unlock_bh(&mm->page_table_lock);
+ if (table)
+ crst_table_free(mm, table);
+ if (mm->context.asce_limit < limit)
+ goto repeat;
+ update_mm(mm, current);
+ return 0;
+}
+
+void crst_table_downgrade(struct mm_struct *mm, unsigned long limit)
+{
+ pgd_t *pgd;
+
+ if (mm->context.asce_limit <= limit)
+ return;
+ __tlb_flush_mm(mm);
+ while (mm->context.asce_limit > limit) {
+ pgd = mm->pgd;
+ switch (pgd_val(*pgd) & _REGION_ENTRY_TYPE_MASK) {
+ case _REGION_ENTRY_TYPE_R2:
+ mm->context.asce_limit = 1UL << 42;
+ mm->context.asce_bits = _ASCE_TABLE_LENGTH |
+ _ASCE_USER_BITS |
+ _ASCE_TYPE_REGION3;
+ break;
+ case _REGION_ENTRY_TYPE_R3:
+ mm->context.asce_limit = 1UL << 31;
+ mm->context.asce_bits = _ASCE_TABLE_LENGTH |
+ _ASCE_USER_BITS |
+ _ASCE_TYPE_SEGMENT;
+ break;
+ default:
+ BUG();
+ }
+ mm->pgd = (pgd_t *) (pgd_val(*pgd) & _REGION_ENTRY_ORIGIN);
+ mm->task_size = mm->context.asce_limit;
+ crst_table_free(mm, (unsigned long *) pgd);
+ }
+ update_mm(mm, current);
+}
+#endif
+
+#ifdef CONFIG_PGSTE
+
+/**
+ * gmap_alloc - allocate a guest address space
+ * @mm: pointer to the parent mm_struct
+ *
+ * Returns a guest address space structure.
+ */
+struct gmap *gmap_alloc(struct mm_struct *mm)
+{
+ struct gmap *gmap;
+ struct page *page;
+ unsigned long *table;
+
+ gmap = kzalloc(sizeof(struct gmap), GFP_KERNEL);
+ if (!gmap)
+ goto out;
+ INIT_LIST_HEAD(&gmap->crst_list);
+ gmap->mm = mm;
+ page = alloc_pages(GFP_KERNEL, ALLOC_ORDER);
+ if (!page)
+ goto out_free;
+ list_add(&page->lru, &gmap->crst_list);
+ table = (unsigned long *) page_to_phys(page);
+ crst_table_init(table, _REGION1_ENTRY_EMPTY);
+ gmap->table = table;
+ gmap->asce = _ASCE_TYPE_REGION1 | _ASCE_TABLE_LENGTH |
+ _ASCE_USER_BITS | __pa(table);
+ list_add(&gmap->list, &mm->context.gmap_list);
+ return gmap;
+
+out_free:
+ kfree(gmap);
+out:
+ return NULL;
+}
+EXPORT_SYMBOL_GPL(gmap_alloc);
+
+static int gmap_unlink_segment(struct gmap *gmap, unsigned long *table)
+{
+ struct gmap_pgtable *mp;
+ struct gmap_rmap *rmap;
+ struct page *page;
+
+ if (*table & _SEGMENT_ENTRY_INV)
+ return 0;
+ page = pfn_to_page(*table >> PAGE_SHIFT);
+ mp = (struct gmap_pgtable *) page->index;
+ list_for_each_entry(rmap, &mp->mapper, list) {
+ if (rmap->entry != table)
+ continue;
+ list_del(&rmap->list);
+ kfree(rmap);
+ break;
+ }
+ *table = _SEGMENT_ENTRY_INV | _SEGMENT_ENTRY_RO | mp->vmaddr;
+ return 1;
+}
+
+static void gmap_flush_tlb(struct gmap *gmap)
+{
+ if (MACHINE_HAS_IDTE)
+ __tlb_flush_idte((unsigned long) gmap->table |
+ _ASCE_TYPE_REGION1);
+ else
+ __tlb_flush_global();
+}
+
+/**
+ * gmap_free - free a guest address space
+ * @gmap: pointer to the guest address space structure
+ */
+void gmap_free(struct gmap *gmap)
+{
+ struct page *page, *next;
+ unsigned long *table;
+ int i;
+
+
+ /* Flush tlb. */
+ if (MACHINE_HAS_IDTE)
+ __tlb_flush_idte((unsigned long) gmap->table |
+ _ASCE_TYPE_REGION1);
+ else
+ __tlb_flush_global();
+
+ /* Free all segment & region tables. */
+ down_read(&gmap->mm->mmap_sem);
+ spin_lock(&gmap->mm->page_table_lock);
+ list_for_each_entry_safe(page, next, &gmap->crst_list, lru) {
+ table = (unsigned long *) page_to_phys(page);
+ if ((*table & _REGION_ENTRY_TYPE_MASK) == 0)
+ /* Remove gmap rmap structures for segment table. */
+ for (i = 0; i < PTRS_PER_PMD; i++, table++)
+ gmap_unlink_segment(gmap, table);
+ __free_pages(page, ALLOC_ORDER);
+ }
+ spin_unlock(&gmap->mm->page_table_lock);
+ up_read(&gmap->mm->mmap_sem);
+ list_del(&gmap->list);
+ kfree(gmap);
+}
+EXPORT_SYMBOL_GPL(gmap_free);
+
+/**
+ * gmap_enable - switch primary space to the guest address space
+ * @gmap: pointer to the guest address space structure
+ */
+void gmap_enable(struct gmap *gmap)
+{
+ S390_lowcore.gmap = (unsigned long) gmap;
+}
+EXPORT_SYMBOL_GPL(gmap_enable);
+
+/**
+ * gmap_disable - switch back to the standard primary address space
+ * @gmap: pointer to the guest address space structure
+ */
+void gmap_disable(struct gmap *gmap)
+{
+ S390_lowcore.gmap = 0UL;
+}
+EXPORT_SYMBOL_GPL(gmap_disable);
+
+/*
+ * gmap_alloc_table is assumed to be called with mmap_sem held
+ */
+static int gmap_alloc_table(struct gmap *gmap,
+ unsigned long *table, unsigned long init)
+{
+ struct page *page;
+ unsigned long *new;
+
+ /* since we dont free the gmap table until gmap_free we can unlock */
+ spin_unlock(&gmap->mm->page_table_lock);
+ page = alloc_pages(GFP_KERNEL, ALLOC_ORDER);
+ spin_lock(&gmap->mm->page_table_lock);
+ if (!page)
+ return -ENOMEM;
+ new = (unsigned long *) page_to_phys(page);
+ crst_table_init(new, init);
+ if (*table & _REGION_ENTRY_INV) {
+ list_add(&page->lru, &gmap->crst_list);
+ *table = (unsigned long) new | _REGION_ENTRY_LENGTH |
+ (*table & _REGION_ENTRY_TYPE_MASK);
+ } else
+ __free_pages(page, ALLOC_ORDER);
+ return 0;
+}
+
+/**
+ * gmap_unmap_segment - unmap segment from the guest address space
+ * @gmap: pointer to the guest address space structure
+ * @addr: address in the guest address space
+ * @len: length of the memory area to unmap
+ *
+ * Returns 0 if the unmap succeded, -EINVAL if not.
+ */
+int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len)
+{
+ unsigned long *table;
+ unsigned long off;
+ int flush;
+
+ if ((to | len) & (PMD_SIZE - 1))
+ return -EINVAL;
+ if (len == 0 || to + len < to)
+ return -EINVAL;
+
+ flush = 0;
+ down_read(&gmap->mm->mmap_sem);
+ spin_lock(&gmap->mm->page_table_lock);
+ for (off = 0; off < len; off += PMD_SIZE) {
+ /* Walk the guest addr space page table */
+ table = gmap->table + (((to + off) >> 53) & 0x7ff);
+ if (*table & _REGION_ENTRY_INV)
+ goto out;
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + (((to + off) >> 42) & 0x7ff);
+ if (*table & _REGION_ENTRY_INV)
+ goto out;
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + (((to + off) >> 31) & 0x7ff);
+ if (*table & _REGION_ENTRY_INV)
+ goto out;
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + (((to + off) >> 20) & 0x7ff);
+
+ /* Clear segment table entry in guest address space. */
+ flush |= gmap_unlink_segment(gmap, table);
+ *table = _SEGMENT_ENTRY_INV;
+ }
+out:
+ spin_unlock(&gmap->mm->page_table_lock);
+ up_read(&gmap->mm->mmap_sem);
+ if (flush)
+ gmap_flush_tlb(gmap);
+ return 0;
+}
+EXPORT_SYMBOL_GPL(gmap_unmap_segment);
+
+/**
+ * gmap_mmap_segment - map a segment to the guest address space
+ * @gmap: pointer to the guest address space structure
+ * @from: source address in the parent address space
+ * @to: target address in the guest address space
+ *
+ * Returns 0 if the mmap succeded, -EINVAL or -ENOMEM if not.
+ */
+int gmap_map_segment(struct gmap *gmap, unsigned long from,
+ unsigned long to, unsigned long len)
+{
+ unsigned long *table;
+ unsigned long off;
+ int flush;
+
+ if ((from | to | len) & (PMD_SIZE - 1))
+ return -EINVAL;
+ if (len == 0 || from + len > PGDIR_SIZE ||
+ from + len < from || to + len < to)
+ return -EINVAL;
+
+ flush = 0;
+ down_read(&gmap->mm->mmap_sem);
+ spin_lock(&gmap->mm->page_table_lock);
+ for (off = 0; off < len; off += PMD_SIZE) {
+ /* Walk the gmap address space page table */
+ table = gmap->table + (((to + off) >> 53) & 0x7ff);
+ if ((*table & _REGION_ENTRY_INV) &&
+ gmap_alloc_table(gmap, table, _REGION2_ENTRY_EMPTY))
+ goto out_unmap;
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + (((to + off) >> 42) & 0x7ff);
+ if ((*table & _REGION_ENTRY_INV) &&
+ gmap_alloc_table(gmap, table, _REGION3_ENTRY_EMPTY))
+ goto out_unmap;
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + (((to + off) >> 31) & 0x7ff);
+ if ((*table & _REGION_ENTRY_INV) &&
+ gmap_alloc_table(gmap, table, _SEGMENT_ENTRY_EMPTY))
+ goto out_unmap;
+ table = (unsigned long *) (*table & _REGION_ENTRY_ORIGIN);
+ table = table + (((to + off) >> 20) & 0x7ff);
+
+ /* Store 'from' address in an invalid segment table entry. */
+ flush |= gmap_unlink_segment(gmap, table);
+ *table = _SEGMENT_ENTRY_INV | _SEGMENT_ENTRY_RO | (from + off);
+ }
+ spin_unlock(&gmap->mm->page_table_lock);
+ up_read(&gmap->mm->mmap_sem);
+ if (flush)
+ gmap_flush_tlb(gmap);
+ return 0;
+
+out_unmap:
+ spin_unlock(&gmap->mm->page_table_lock);
+ up_read(&gmap->mm->mmap_sem);
+ gmap_unmap_segment(gmap, to, len);
+ return -ENOMEM;
+}
+EXPORT_SYMBOL_GPL(gmap_map_segment);
+
+/*
+ * this function is assumed to be called with mmap_sem held
+ */
+unsigned long __gmap_fault(unsigned long address, struct gmap *gmap)
+{
+ unsigned long *table, vmaddr, segment;
+ struct mm_struct *mm;
+ struct gmap_pgtable *mp;
+ struct gmap_rmap *rmap;
+ struct vm_area_struct *vma;
+ struct page *page;
+ pgd_t *pgd;
+ pud_t *pud;
+ pmd_t *pmd;
+
+ current->thread.gmap_addr = address;
+ mm = gmap->mm;
+ /* Walk the gmap address space page table */
+ table = gmap->table + ((address >> 53) & 0x7ff);
+ if (unlikely(*table & _REGION_ENTRY_INV))
+ return -EFAULT;
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + ((address >> 42) & 0x7ff);
+ if (unlikely(*table & _REGION_ENTRY_INV))
+ return -EFAULT;
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + ((address >> 31) & 0x7ff);
+ if (unlikely(*table & _REGION_ENTRY_INV))
+ return -EFAULT;
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + ((address >> 20) & 0x7ff);
+
+ /* Convert the gmap address to an mm address. */
+ segment = *table;
+ if (likely(!(segment & _SEGMENT_ENTRY_INV))) {
+ page = pfn_to_page(segment >> PAGE_SHIFT);
+ mp = (struct gmap_pgtable *) page->index;
+ return mp->vmaddr | (address & ~PMD_MASK);
+ } else if (segment & _SEGMENT_ENTRY_RO) {
+ vmaddr = segment & _SEGMENT_ENTRY_ORIGIN;
+ vma = find_vma(mm, vmaddr);
+ if (!vma || vma->vm_start > vmaddr)
+ return -EFAULT;
+
+ /* Walk the parent mm page table */
+ pgd = pgd_offset(mm, vmaddr);
+ pud = pud_alloc(mm, pgd, vmaddr);
+ if (!pud)
+ return -ENOMEM;
+ pmd = pmd_alloc(mm, pud, vmaddr);
+ if (!pmd)
+ return -ENOMEM;
+ if (!pmd_present(*pmd) &&
+ __pte_alloc(mm, vma, pmd, vmaddr))
+ return -ENOMEM;
+ /* pmd now points to a valid segment table entry. */
+ rmap = kmalloc(sizeof(*rmap), GFP_KERNEL|__GFP_REPEAT);
+ if (!rmap)
+ return -ENOMEM;
+ /* Link gmap segment table entry location to page table. */
+ page = pmd_page(*pmd);
+ mp = (struct gmap_pgtable *) page->index;
+ rmap->entry = table;
+ spin_lock(&mm->page_table_lock);
+ list_add(&rmap->list, &mp->mapper);
+ spin_unlock(&mm->page_table_lock);
+ /* Set gmap segment table entry to page table. */
+ *table = pmd_val(*pmd) & PAGE_MASK;
+ return vmaddr | (address & ~PMD_MASK);
+ }
+ return -EFAULT;
+}
+
+unsigned long gmap_fault(unsigned long address, struct gmap *gmap)
+{
+ unsigned long rc;
+
+ down_read(&gmap->mm->mmap_sem);
+ rc = __gmap_fault(address, gmap);
+ up_read(&gmap->mm->mmap_sem);
+
+ return rc;
+}
+EXPORT_SYMBOL_GPL(gmap_fault);
+
+void gmap_discard(unsigned long from, unsigned long to, struct gmap *gmap)
+{
+
+ unsigned long *table, address, size;
+ struct vm_area_struct *vma;
+ struct gmap_pgtable *mp;
+ struct page *page;
+
+ down_read(&gmap->mm->mmap_sem);
+ address = from;
+ while (address < to) {
+ /* Walk the gmap address space page table */
+ table = gmap->table + ((address >> 53) & 0x7ff);
+ if (unlikely(*table & _REGION_ENTRY_INV)) {
+ address = (address + PMD_SIZE) & PMD_MASK;
+ continue;
+ }
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + ((address >> 42) & 0x7ff);
+ if (unlikely(*table & _REGION_ENTRY_INV)) {
+ address = (address + PMD_SIZE) & PMD_MASK;
+ continue;
+ }
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + ((address >> 31) & 0x7ff);
+ if (unlikely(*table & _REGION_ENTRY_INV)) {
+ address = (address + PMD_SIZE) & PMD_MASK;
+ continue;
+ }
+ table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+ table = table + ((address >> 20) & 0x7ff);
+ if (unlikely(*table & _SEGMENT_ENTRY_INV)) {
+ address = (address + PMD_SIZE) & PMD_MASK;
+ continue;
+ }
+ page = pfn_to_page(*table >> PAGE_SHIFT);
+ mp = (struct gmap_pgtable *) page->index;
+ vma = find_vma(gmap->mm, mp->vmaddr);
+ size = min(to - address, PMD_SIZE - (address & ~PMD_MASK));
+ zap_page_range(vma, mp->vmaddr | (address & ~PMD_MASK),
+ size, NULL);
+ address = (address + PMD_SIZE) & PMD_MASK;
+ }
+ up_read(&gmap->mm->mmap_sem);
+}
+EXPORT_SYMBOL_GPL(gmap_discard);
+
+void gmap_unmap_notifier(struct mm_struct *mm, unsigned long *table)
+{
+ struct gmap_rmap *rmap, *next;
+ struct gmap_pgtable *mp;
+ struct page *page;
+ int flush;
+
+ flush = 0;
+ spin_lock(&mm->page_table_lock);
+ page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+ mp = (struct gmap_pgtable *) page->index;
+ list_for_each_entry_safe(rmap, next, &mp->mapper, list) {
+ *rmap->entry =
+ _SEGMENT_ENTRY_INV | _SEGMENT_ENTRY_RO | mp->vmaddr;
+ list_del(&rmap->list);
+ kfree(rmap);
+ flush = 1;
+ }
+ spin_unlock(&mm->page_table_lock);
+ if (flush)
+ __tlb_flush_global();
+}
+
+static inline unsigned long *page_table_alloc_pgste(struct mm_struct *mm,
+ unsigned long vmaddr)
+{
+ struct page *page;
+ unsigned long *table;
+ struct gmap_pgtable *mp;
+
+ page = alloc_page(GFP_KERNEL|__GFP_REPEAT);
+ if (!page)
+ return NULL;
+ mp = kmalloc(sizeof(*mp), GFP_KERNEL|__GFP_REPEAT);
+ if (!mp) {
+ __free_page(page);
+ return NULL;
+ }
+ pgtable_page_ctor(page);
+ mp->vmaddr = vmaddr & PMD_MASK;
+ INIT_LIST_HEAD(&mp->mapper);
+ page->index = (unsigned long) mp;
+ atomic_set(&page->_mapcount, 3);
+ table = (unsigned long *) page_to_phys(page);
+ clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE/2);
+ clear_table(table + PTRS_PER_PTE, 0, PAGE_SIZE/2);
+ return table;
+}
+
+static inline void page_table_free_pgste(unsigned long *table)
+{
+ struct page *page;
+ struct gmap_pgtable *mp;
+
+ page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+ mp = (struct gmap_pgtable *) page->index;
+ BUG_ON(!list_empty(&mp->mapper));
+ pgtable_page_dtor(page);
+ atomic_set(&page->_mapcount, -1);
+ kfree(mp);
+ __free_page(page);
+}
+
+#else /* CONFIG_PGSTE */
+
+static inline unsigned long *page_table_alloc_pgste(struct mm_struct *mm,
+ unsigned long vmaddr)
+{
+ return NULL;
+}
+
+static inline void page_table_free_pgste(unsigned long *table)
+{
+}
+
+static inline void gmap_unmap_notifier(struct mm_struct *mm,
+ unsigned long *table)
+{
+}
+
+#endif /* CONFIG_PGSTE */
+
+static inline unsigned int atomic_xor_bits(atomic_t *v, unsigned int bits)
+{
+ unsigned int old, new;
+
+ do {
+ old = atomic_read(v);
+ new = old ^ bits;
+ } while (atomic_cmpxchg(v, old, new) != old);
+ return new;
+}
+
+/*
+ * page table entry allocation/free routines.
+ */
+unsigned long *page_table_alloc(struct mm_struct *mm, unsigned long vmaddr)
+{
+ struct page *page;
+ unsigned long *table;
+ unsigned int mask, bit;
+
+ if (mm_has_pgste(mm))
+ return page_table_alloc_pgste(mm, vmaddr);
+ /* Allocate fragments of a 4K page as 1K/2K page table */
+ spin_lock_bh(&mm->context.list_lock);
+ mask = FRAG_MASK;
+ if (!list_empty(&mm->context.pgtable_list)) {
+ page = list_first_entry(&mm->context.pgtable_list,
+ struct page, lru);
+ table = (unsigned long *) page_to_phys(page);
+ mask = atomic_read(&page->_mapcount);
+ mask = mask | (mask >> 4);
+ }
+ if ((mask & FRAG_MASK) == FRAG_MASK) {
+ spin_unlock_bh(&mm->context.list_lock);
+ page = alloc_page(GFP_KERNEL|__GFP_REPEAT);
+ if (!page)
+ return NULL;
+ pgtable_page_ctor(page);
+ atomic_set(&page->_mapcount, 1);
+ table = (unsigned long *) page_to_phys(page);
+ clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE);
+ spin_lock_bh(&mm->context.list_lock);
+ list_add(&page->lru, &mm->context.pgtable_list);
+ } else {
+ for (bit = 1; mask & bit; bit <<= 1)
+ table += PTRS_PER_PTE;
+ mask = atomic_xor_bits(&page->_mapcount, bit);
+ if ((mask & FRAG_MASK) == FRAG_MASK)
+ list_del(&page->lru);
+ }
+ spin_unlock_bh(&mm->context.list_lock);
+ return table;
+}
+
+void page_table_free(struct mm_struct *mm, unsigned long *table)
+{
+ struct page *page;
+ unsigned int bit, mask;
+
+ if (mm_has_pgste(mm)) {
+ gmap_unmap_notifier(mm, table);
+ return page_table_free_pgste(table);
+ }
+ /* Free 1K/2K page table fragment of a 4K page */
+ page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+ bit = 1 << ((__pa(table) & ~PAGE_MASK)/(PTRS_PER_PTE*sizeof(pte_t)));
+ spin_lock_bh(&mm->context.list_lock);
+ if ((atomic_read(&page->_mapcount) & FRAG_MASK) != FRAG_MASK)
+ list_del(&page->lru);
+ mask = atomic_xor_bits(&page->_mapcount, bit);
+ if (mask & FRAG_MASK)
+ list_add(&page->lru, &mm->context.pgtable_list);
+ spin_unlock_bh(&mm->context.list_lock);
+ if (mask == 0) {
+ pgtable_page_dtor(page);
+ atomic_set(&page->_mapcount, -1);
+ __free_page(page);
+ }
+}
+
+static void __page_table_free_rcu(void *table, unsigned bit)
+{
+ struct page *page;
+
+ if (bit == FRAG_MASK)
+ return page_table_free_pgste(table);
+ /* Free 1K/2K page table fragment of a 4K page */
+ page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+ if (atomic_xor_bits(&page->_mapcount, bit) == 0) {
+ pgtable_page_dtor(page);
+ atomic_set(&page->_mapcount, -1);
+ __free_page(page);
+ }
+}
+
+void page_table_free_rcu(struct mmu_gather *tlb, unsigned long *table)
+{
+ struct mm_struct *mm;
+ struct page *page;
+ unsigned int bit, mask;
+
+ mm = tlb->mm;
+ if (mm_has_pgste(mm)) {
+ gmap_unmap_notifier(mm, table);
+ table = (unsigned long *) (__pa(table) | FRAG_MASK);
+ tlb_remove_table(tlb, table);
+ return;
+ }
+ bit = 1 << ((__pa(table) & ~PAGE_MASK) / (PTRS_PER_PTE*sizeof(pte_t)));
+ page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+ spin_lock_bh(&mm->context.list_lock);
+ if ((atomic_read(&page->_mapcount) & FRAG_MASK) != FRAG_MASK)
+ list_del(&page->lru);
+ mask = atomic_xor_bits(&page->_mapcount, bit | (bit << 4));
+ if (mask & FRAG_MASK)
+ list_add_tail(&page->lru, &mm->context.pgtable_list);
+ spin_unlock_bh(&mm->context.list_lock);
+ table = (unsigned long *) (__pa(table) | (bit << 4));
+ tlb_remove_table(tlb, table);
+}
+
+void __tlb_remove_table(void *_table)
+{
+ const unsigned long mask = (FRAG_MASK << 4) | FRAG_MASK;
+ void *table = (void *)((unsigned long) _table & ~mask);
+ unsigned type = (unsigned long) _table & mask;
+
+ if (type)
+ __page_table_free_rcu(table, type);
+ else
+ free_pages((unsigned long) table, ALLOC_ORDER);
+}
+
+static void tlb_remove_table_smp_sync(void *arg)
+{
+ /* Simply deliver the interrupt */
+}
+
+static void tlb_remove_table_one(void *table)
+{
+ /*
+ * This isn't an RCU grace period and hence the page-tables cannot be
+ * assumed to be actually RCU-freed.
+ *
+ * It is however sufficient for software page-table walkers that rely
+ * on IRQ disabling. See the comment near struct mmu_table_batch.
+ */
+ smp_call_function(tlb_remove_table_smp_sync, NULL, 1);
+ __tlb_remove_table(table);
+}
+
+static void tlb_remove_table_rcu(struct rcu_head *head)
+{
+ struct mmu_table_batch *batch;
+ int i;
+
+ batch = container_of(head, struct mmu_table_batch, rcu);
+
+ for (i = 0; i < batch->nr; i++)
+ __tlb_remove_table(batch->tables[i]);
+
+ free_page((unsigned long)batch);
+}
+
+void tlb_table_flush(struct mmu_gather *tlb)
+{
+ struct mmu_table_batch **batch = &tlb->batch;
+
+ if (*batch) {
+ __tlb_flush_mm(tlb->mm);
+ call_rcu_sched(&(*batch)->rcu, tlb_remove_table_rcu);
+ *batch = NULL;
+ }
+}
+
+void tlb_remove_table(struct mmu_gather *tlb, void *table)
+{
+ struct mmu_table_batch **batch = &tlb->batch;
+
+ if (*batch == NULL) {
+ *batch = (struct mmu_table_batch *)
+ __get_free_page(GFP_NOWAIT | __GFP_NOWARN);
+ if (*batch == NULL) {
+ __tlb_flush_mm(tlb->mm);
+ tlb_remove_table_one(table);
+ return;
+ }
+ (*batch)->nr = 0;
+ }
+ (*batch)->tables[(*batch)->nr++] = table;
+ if ((*batch)->nr == MAX_TABLE_BATCH)
+ tlb_table_flush(tlb);
+}
+
+/*
+ * switch on pgstes for its userspace process (for kvm)
+ */
+int s390_enable_sie(void)
+{
+ struct task_struct *tsk = current;
+ struct mm_struct *mm, *old_mm;
+
+ /* Do we have switched amode? If no, we cannot do sie */
+ if (user_mode == HOME_SPACE_MODE)
+ return -EINVAL;
+
+ /* Do we have pgstes? if yes, we are done */
+ if (mm_has_pgste(tsk->mm))
+ return 0;
+
+ /* lets check if we are allowed to replace the mm */
+ task_lock(tsk);
+ if (!tsk->mm || atomic_read(&tsk->mm->mm_users) > 1 ||
+#ifdef CONFIG_AIO
+ !hlist_empty(&tsk->mm->ioctx_list) ||
+#endif
+ tsk->mm != tsk->active_mm) {
+ task_unlock(tsk);
+ return -EINVAL;
+ }
+ task_unlock(tsk);
+
+ /* we copy the mm and let dup_mm create the page tables with_pgstes */
+ tsk->mm->context.alloc_pgste = 1;
+ mm = dup_mm(tsk);
+ tsk->mm->context.alloc_pgste = 0;
+ if (!mm)
+ return -ENOMEM;
+
+ /* Now lets check again if something happened */
+ task_lock(tsk);
+ if (!tsk->mm || atomic_read(&tsk->mm->mm_users) > 1 ||
+#ifdef CONFIG_AIO
+ !hlist_empty(&tsk->mm->ioctx_list) ||
+#endif
+ tsk->mm != tsk->active_mm) {
+ mmput(mm);
+ task_unlock(tsk);
+ return -EINVAL;
+ }
+
+ /* ok, we are alone. No ptrace, no threads, etc. */
+ old_mm = tsk->mm;
+ tsk->mm = tsk->active_mm = mm;
+ preempt_disable();
+ update_mm(mm, tsk);
+ atomic_inc(&mm->context.attach_count);
+ atomic_dec(&old_mm->context.attach_count);
+ cpumask_set_cpu(smp_processor_id(), mm_cpumask(mm));
+ preempt_enable();
+ task_unlock(tsk);
+ mmput(old_mm);
+ return 0;
+}
+EXPORT_SYMBOL_GPL(s390_enable_sie);
+
+#if defined(CONFIG_DEBUG_PAGEALLOC) && defined(CONFIG_HIBERNATION)
+bool kernel_page_present(struct page *page)
+{
+ unsigned long addr;
+ int cc;
+
+ addr = page_to_phys(page);
+ asm volatile(
+ " lra %1,0(%1)\n"
+ " ipm %0\n"
+ " srl %0,28"
+ : "=d" (cc), "+a" (addr) : : "cc");
+ return cc == 0;
+}
+#endif /* CONFIG_HIBERNATION && CONFIG_DEBUG_PAGEALLOC */
diff --git a/arch/s390/mm/vmem.c b/arch/s390/mm/vmem.c
new file mode 100644
index 00000000..4799383e
--- /dev/null
+++ b/arch/s390/mm/vmem.c
@@ -0,0 +1,388 @@
+/*
+ * arch/s390/mm/vmem.c
+ *
+ * Copyright IBM Corp. 2006
+ * Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>
+ */
+
+#include <linux/bootmem.h>
+#include <linux/pfn.h>
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/hugetlb.h>
+#include <linux/slab.h>
+#include <asm/pgalloc.h>
+#include <asm/pgtable.h>
+#include <asm/setup.h>
+#include <asm/tlbflush.h>
+#include <asm/sections.h>
+
+static DEFINE_MUTEX(vmem_mutex);
+
+struct memory_segment {
+ struct list_head list;
+ unsigned long start;
+ unsigned long size;
+};
+
+static LIST_HEAD(mem_segs);
+
+static void __ref *vmem_alloc_pages(unsigned int order)
+{
+ if (slab_is_available())
+ return (void *)__get_free_pages(GFP_KERNEL, order);
+ return alloc_bootmem_pages((1 << order) * PAGE_SIZE);
+}
+
+static inline pud_t *vmem_pud_alloc(void)
+{
+ pud_t *pud = NULL;
+
+#ifdef CONFIG_64BIT
+ pud = vmem_alloc_pages(2);
+ if (!pud)
+ return NULL;
+ clear_table((unsigned long *) pud, _REGION3_ENTRY_EMPTY, PAGE_SIZE * 4);
+#endif
+ return pud;
+}
+
+static inline pmd_t *vmem_pmd_alloc(void)
+{
+ pmd_t *pmd = NULL;
+
+#ifdef CONFIG_64BIT
+ pmd = vmem_alloc_pages(2);
+ if (!pmd)
+ return NULL;
+ clear_table((unsigned long *) pmd, _SEGMENT_ENTRY_EMPTY, PAGE_SIZE * 4);
+#endif
+ return pmd;
+}
+
+static pte_t __ref *vmem_pte_alloc(unsigned long address)
+{
+ pte_t *pte;
+
+ if (slab_is_available())
+ pte = (pte_t *) page_table_alloc(&init_mm, address);
+ else
+ pte = alloc_bootmem(PTRS_PER_PTE * sizeof(pte_t));
+ if (!pte)
+ return NULL;
+ clear_table((unsigned long *) pte, _PAGE_TYPE_EMPTY,
+ PTRS_PER_PTE * sizeof(pte_t));
+ return pte;
+}
+
+/*
+ * Add a physical memory range to the 1:1 mapping.
+ */
+static int vmem_add_mem(unsigned long start, unsigned long size, int ro)
+{
+ unsigned long address;
+ pgd_t *pg_dir;
+ pud_t *pu_dir;
+ pmd_t *pm_dir;
+ pte_t *pt_dir;
+ pte_t pte;
+ int ret = -ENOMEM;
+
+ for (address = start; address < start + size; address += PAGE_SIZE) {
+ pg_dir = pgd_offset_k(address);
+ if (pgd_none(*pg_dir)) {
+ pu_dir = vmem_pud_alloc();
+ if (!pu_dir)
+ goto out;
+ pgd_populate(&init_mm, pg_dir, pu_dir);
+ }
+
+ pu_dir = pud_offset(pg_dir, address);
+ if (pud_none(*pu_dir)) {
+ pm_dir = vmem_pmd_alloc();
+ if (!pm_dir)
+ goto out;
+ pud_populate(&init_mm, pu_dir, pm_dir);
+ }
+
+ pte = mk_pte_phys(address, __pgprot(ro ? _PAGE_RO : 0));
+ pm_dir = pmd_offset(pu_dir, address);
+
+#ifdef __s390x__
+ if (MACHINE_HAS_HPAGE && !(address & ~HPAGE_MASK) &&
+ (address + HPAGE_SIZE <= start + size) &&
+ (address >= HPAGE_SIZE)) {
+ pte_val(pte) |= _SEGMENT_ENTRY_LARGE;
+ pmd_val(*pm_dir) = pte_val(pte);
+ address += HPAGE_SIZE - PAGE_SIZE;
+ continue;
+ }
+#endif
+ if (pmd_none(*pm_dir)) {
+ pt_dir = vmem_pte_alloc(address);
+ if (!pt_dir)
+ goto out;
+ pmd_populate(&init_mm, pm_dir, pt_dir);
+ }
+
+ pt_dir = pte_offset_kernel(pm_dir, address);
+ *pt_dir = pte;
+ }
+ ret = 0;
+out:
+ flush_tlb_kernel_range(start, start + size);
+ return ret;
+}
+
+/*
+ * Remove a physical memory range from the 1:1 mapping.
+ * Currently only invalidates page table entries.
+ */
+static void vmem_remove_range(unsigned long start, unsigned long size)
+{
+ unsigned long address;
+ pgd_t *pg_dir;
+ pud_t *pu_dir;
+ pmd_t *pm_dir;
+ pte_t *pt_dir;
+ pte_t pte;
+
+ pte_val(pte) = _PAGE_TYPE_EMPTY;
+ for (address = start; address < start + size; address += PAGE_SIZE) {
+ pg_dir = pgd_offset_k(address);
+ pu_dir = pud_offset(pg_dir, address);
+ if (pud_none(*pu_dir))
+ continue;
+ pm_dir = pmd_offset(pu_dir, address);
+ if (pmd_none(*pm_dir))
+ continue;
+
+ if (pmd_huge(*pm_dir)) {
+ pmd_clear(pm_dir);
+ address += HPAGE_SIZE - PAGE_SIZE;
+ continue;
+ }
+
+ pt_dir = pte_offset_kernel(pm_dir, address);
+ *pt_dir = pte;
+ }
+ flush_tlb_kernel_range(start, start + size);
+}
+
+/*
+ * Add a backed mem_map array to the virtual mem_map array.
+ */
+int __meminit vmemmap_populate(struct page *start, unsigned long nr, int node)
+{
+ unsigned long address, start_addr, end_addr;
+ pgd_t *pg_dir;
+ pud_t *pu_dir;
+ pmd_t *pm_dir;
+ pte_t *pt_dir;
+ pte_t pte;
+ int ret = -ENOMEM;
+
+ start_addr = (unsigned long) start;
+ end_addr = (unsigned long) (start + nr);
+
+ for (address = start_addr; address < end_addr; address += PAGE_SIZE) {
+ pg_dir = pgd_offset_k(address);
+ if (pgd_none(*pg_dir)) {
+ pu_dir = vmem_pud_alloc();
+ if (!pu_dir)
+ goto out;
+ pgd_populate(&init_mm, pg_dir, pu_dir);
+ }
+
+ pu_dir = pud_offset(pg_dir, address);
+ if (pud_none(*pu_dir)) {
+ pm_dir = vmem_pmd_alloc();
+ if (!pm_dir)
+ goto out;
+ pud_populate(&init_mm, pu_dir, pm_dir);
+ }
+
+ pm_dir = pmd_offset(pu_dir, address);
+ if (pmd_none(*pm_dir)) {
+ pt_dir = vmem_pte_alloc(address);
+ if (!pt_dir)
+ goto out;
+ pmd_populate(&init_mm, pm_dir, pt_dir);
+ }
+
+ pt_dir = pte_offset_kernel(pm_dir, address);
+ if (pte_none(*pt_dir)) {
+ unsigned long new_page;
+
+ new_page =__pa(vmem_alloc_pages(0));
+ if (!new_page)
+ goto out;
+ pte = pfn_pte(new_page >> PAGE_SHIFT, PAGE_KERNEL);
+ *pt_dir = pte;
+ }
+ }
+ memset(start, 0, nr * sizeof(struct page));
+ ret = 0;
+out:
+ flush_tlb_kernel_range(start_addr, end_addr);
+ return ret;
+}
+
+/*
+ * Add memory segment to the segment list if it doesn't overlap with
+ * an already present segment.
+ */
+static int insert_memory_segment(struct memory_segment *seg)
+{
+ struct memory_segment *tmp;
+
+ if (seg->start + seg->size > VMEM_MAX_PHYS ||
+ seg->start + seg->size < seg->start)
+ return -ERANGE;
+
+ list_for_each_entry(tmp, &mem_segs, list) {
+ if (seg->start >= tmp->start + tmp->size)
+ continue;
+ if (seg->start + seg->size <= tmp->start)
+ continue;
+ return -ENOSPC;
+ }
+ list_add(&seg->list, &mem_segs);
+ return 0;
+}
+
+/*
+ * Remove memory segment from the segment list.
+ */
+static void remove_memory_segment(struct memory_segment *seg)
+{
+ list_del(&seg->list);
+}
+
+static void __remove_shared_memory(struct memory_segment *seg)
+{
+ remove_memory_segment(seg);
+ vmem_remove_range(seg->start, seg->size);
+}
+
+int vmem_remove_mapping(unsigned long start, unsigned long size)
+{
+ struct memory_segment *seg;
+ int ret;
+
+ mutex_lock(&vmem_mutex);
+
+ ret = -ENOENT;
+ list_for_each_entry(seg, &mem_segs, list) {
+ if (seg->start == start && seg->size == size)
+ break;
+ }
+
+ if (seg->start != start || seg->size != size)
+ goto out;
+
+ ret = 0;
+ __remove_shared_memory(seg);
+ kfree(seg);
+out:
+ mutex_unlock(&vmem_mutex);
+ return ret;
+}
+
+int vmem_add_mapping(unsigned long start, unsigned long size)
+{
+ struct memory_segment *seg;
+ int ret;
+
+ mutex_lock(&vmem_mutex);
+ ret = -ENOMEM;
+ seg = kzalloc(sizeof(*seg), GFP_KERNEL);
+ if (!seg)
+ goto out;
+ seg->start = start;
+ seg->size = size;
+
+ ret = insert_memory_segment(seg);
+ if (ret)
+ goto out_free;
+
+ ret = vmem_add_mem(start, size, 0);
+ if (ret)
+ goto out_remove;
+ goto out;
+
+out_remove:
+ __remove_shared_memory(seg);
+out_free:
+ kfree(seg);
+out:
+ mutex_unlock(&vmem_mutex);
+ return ret;
+}
+
+/*
+ * map whole physical memory to virtual memory (identity mapping)
+ * we reserve enough space in the vmalloc area for vmemmap to hotplug
+ * additional memory segments.
+ */
+void __init vmem_map_init(void)
+{
+ unsigned long ro_start, ro_end;
+ unsigned long start, end;
+ int i;
+
+ ro_start = ((unsigned long)&_stext) & PAGE_MASK;
+ ro_end = PFN_ALIGN((unsigned long)&_eshared);
+ for (i = 0; i < MEMORY_CHUNKS && memory_chunk[i].size > 0; i++) {
+ if (memory_chunk[i].type == CHUNK_CRASHK ||
+ memory_chunk[i].type == CHUNK_OLDMEM)
+ continue;
+ start = memory_chunk[i].addr;
+ end = memory_chunk[i].addr + memory_chunk[i].size;
+ if (start >= ro_end || end <= ro_start)
+ vmem_add_mem(start, end - start, 0);
+ else if (start >= ro_start && end <= ro_end)
+ vmem_add_mem(start, end - start, 1);
+ else if (start >= ro_start) {
+ vmem_add_mem(start, ro_end - start, 1);
+ vmem_add_mem(ro_end, end - ro_end, 0);
+ } else if (end < ro_end) {
+ vmem_add_mem(start, ro_start - start, 0);
+ vmem_add_mem(ro_start, end - ro_start, 1);
+ } else {
+ vmem_add_mem(start, ro_start - start, 0);
+ vmem_add_mem(ro_start, ro_end - ro_start, 1);
+ vmem_add_mem(ro_end, end - ro_end, 0);
+ }
+ }
+}
+
+/*
+ * Convert memory chunk array to a memory segment list so there is a single
+ * list that contains both r/w memory and shared memory segments.
+ */
+static int __init vmem_convert_memory_chunk(void)
+{
+ struct memory_segment *seg;
+ int i;
+
+ mutex_lock(&vmem_mutex);
+ for (i = 0; i < MEMORY_CHUNKS; i++) {
+ if (!memory_chunk[i].size)
+ continue;
+ if (memory_chunk[i].type == CHUNK_CRASHK ||
+ memory_chunk[i].type == CHUNK_OLDMEM)
+ continue;
+ seg = kzalloc(sizeof(*seg), GFP_KERNEL);
+ if (!seg)
+ panic("Out of memory...\n");
+ seg->start = memory_chunk[i].addr;
+ seg->size = memory_chunk[i].size;
+ insert_memory_segment(seg);
+ }
+ mutex_unlock(&vmem_mutex);
+ return 0;
+}
+
+core_initcall(vmem_convert_memory_chunk);
generated by cgit (git 2.34.1) at 2025-01-11 01:06:38 +0000