s390/mm: enable fixup_user_fault retrying
authorDominik Dingel <dingel@linux.vnet.ibm.com>
Sat, 16 Jan 2016 00:57:07 +0000 (16:57 -0800)
committerLinus Torvalds <torvalds@linux-foundation.org>
Sat, 16 Jan 2016 01:56:32 +0000 (17:56 -0800)
By passing a non-null flag we allow fixup_user_fault to retry, which
enables userfaultfd.  As during these retries we might drop the mmap_sem
we need to check if that happened and redo the complete chain of
actions.

Signed-off-by: Dominik Dingel <dingel@linux.vnet.ibm.com>
Reviewed-by: Andrea Arcangeli <aarcange@redhat.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: "Jason J. Herne" <jjherne@linux.vnet.ibm.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Eric B Munson <emunson@akamai.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Dominik Dingel <dingel@linux.vnet.ibm.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
arch/s390/mm/pgtable.c

index 84bddda8d412410e5260c3e5e61cfff04d7e6387..a809fa8e6f8bd01d2c0ad90891060d6aabd36a80 100644 (file)
@@ -578,17 +578,29 @@ int gmap_fault(struct gmap *gmap, unsigned long gaddr,
 {
        unsigned long vmaddr;
        int rc;
+       bool unlocked;
 
        down_read(&gmap->mm->mmap_sem);
+
+retry:
+       unlocked = false;
        vmaddr = __gmap_translate(gmap, gaddr);
        if (IS_ERR_VALUE(vmaddr)) {
                rc = vmaddr;
                goto out_up;
        }
-       if (fixup_user_fault(current, gmap->mm, vmaddr, fault_flags, NULL)) {
+       if (fixup_user_fault(current, gmap->mm, vmaddr, fault_flags,
+                            &unlocked)) {
                rc = -EFAULT;
                goto out_up;
        }
+       /*
+        * In the case that fixup_user_fault unlocked the mmap_sem during
+        * faultin redo __gmap_translate to not race with a map/unmap_segment.
+        */
+       if (unlocked)
+               goto retry;
+
        rc = __gmap_link(gmap, gaddr, vmaddr);
 out_up:
        up_read(&gmap->mm->mmap_sem);
@@ -714,12 +726,14 @@ int gmap_ipte_notify(struct gmap *gmap, unsigned long gaddr, unsigned long len)
        spinlock_t *ptl;
        pte_t *ptep, entry;
        pgste_t pgste;
+       bool unlocked;
        int rc = 0;
 
        if ((gaddr & ~PAGE_MASK) || (len & ~PAGE_MASK))
                return -EINVAL;
        down_read(&gmap->mm->mmap_sem);
        while (len) {
+               unlocked = false;
                /* Convert gmap address and connect the page tables */
                addr = __gmap_translate(gmap, gaddr);
                if (IS_ERR_VALUE(addr)) {
@@ -728,10 +742,13 @@ int gmap_ipte_notify(struct gmap *gmap, unsigned long gaddr, unsigned long len)
                }
                /* Get the page mapped */
                if (fixup_user_fault(current, gmap->mm, addr, FAULT_FLAG_WRITE,
-                                    NULL)) {
+                                    &unlocked)) {
                        rc = -EFAULT;
                        break;
                }
+               /* While trying to map mmap_sem got unlocked. Let us retry */
+               if (unlocked)
+                       continue;
                rc = __gmap_link(gmap, gaddr, addr);
                if (rc)
                        break;
@@ -792,9 +809,11 @@ int set_guest_storage_key(struct mm_struct *mm, unsigned long addr,
        spinlock_t *ptl;
        pgste_t old, new;
        pte_t *ptep;
+       bool unlocked;
 
        down_read(&mm->mmap_sem);
 retry:
+       unlocked = false;
        ptep = get_locked_pte(mm, addr, &ptl);
        if (unlikely(!ptep)) {
                up_read(&mm->mmap_sem);
@@ -803,8 +822,12 @@ retry:
        if (!(pte_val(*ptep) & _PAGE_INVALID) &&
             (pte_val(*ptep) & _PAGE_PROTECT)) {
                pte_unmap_unlock(ptep, ptl);
+               /*
+                * We do not really care about unlocked. We will retry either
+                * way. But this allows fixup_user_fault to enable userfaultfd.
+                */
                if (fixup_user_fault(current, mm, addr, FAULT_FLAG_WRITE,
-                                    NULL)) {
+                                    &unlocked)) {
                        up_read(&mm->mmap_sem);
                        return -EFAULT;
                }