Free the memory allocated by memdup_user() in fs/sysfs/bin.c
[linux-2.6.git] / fs / sysfs / bin.c
index 96cc2bf..2524714 100644 (file)
@@ -157,14 +157,9 @@ static ssize_t write(struct file *file, const char __user *userbuf,
                        count = size - offs;
        }
 
-       temp = kmalloc(count, GFP_KERNEL);
-       if (!temp)
-               return -ENOMEM;
-
-       if (copy_from_user(temp, userbuf, count)) {
-               count = -EFAULT;
-               goto out_free;
-       }
+       temp = memdup_user(userbuf, count);
+       if (IS_ERR(temp))
+               return PTR_ERR(temp);
 
        mutex_lock(&bb->mutex);
 
@@ -176,7 +171,6 @@ static ssize_t write(struct file *file, const char __user *userbuf,
        if (count > 0)
                *off = offs + count;
 
-out_free:
        kfree(temp);
        return count;
 }
@@ -234,20 +228,23 @@ static int bin_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
        return ret;
 }
 
-static int bin_page_mkwrite(struct vm_area_struct *vma, struct page *page)
+static int bin_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 {
        struct file *file = vma->vm_file;
        struct bin_buffer *bb = file->private_data;
        struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata;
        int ret;
 
-       if (!bb->vm_ops || !bb->vm_ops->page_mkwrite)
-               return -EINVAL;
+       if (!bb->vm_ops)
+               return VM_FAULT_SIGBUS;
+
+       if (!bb->vm_ops->page_mkwrite)
+               return 0;
 
        if (!sysfs_get_active_two(attr_sd))
-               return -EINVAL;
+               return VM_FAULT_SIGBUS;
 
-       ret = bb->vm_ops->page_mkwrite(vma, page);
+       ret = bb->vm_ops->page_mkwrite(vma, vmf);
 
        sysfs_put_active_two(attr_sd);
        return ret;
@@ -273,12 +270,78 @@ static int bin_access(struct vm_area_struct *vma, unsigned long addr,
        return ret;
 }
 
+#ifdef CONFIG_NUMA
+static int bin_set_policy(struct vm_area_struct *vma, struct mempolicy *new)
+{
+       struct file *file = vma->vm_file;
+       struct bin_buffer *bb = file->private_data;
+       struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata;
+       int ret;
+
+       if (!bb->vm_ops || !bb->vm_ops->set_policy)
+               return 0;
+
+       if (!sysfs_get_active_two(attr_sd))
+               return -EINVAL;
+
+       ret = bb->vm_ops->set_policy(vma, new);
+
+       sysfs_put_active_two(attr_sd);
+       return ret;
+}
+
+static struct mempolicy *bin_get_policy(struct vm_area_struct *vma,
+                                       unsigned long addr)
+{
+       struct file *file = vma->vm_file;
+       struct bin_buffer *bb = file->private_data;
+       struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata;
+       struct mempolicy *pol;
+
+       if (!bb->vm_ops || !bb->vm_ops->get_policy)
+               return vma->vm_policy;
+
+       if (!sysfs_get_active_two(attr_sd))
+               return vma->vm_policy;
+
+       pol = bb->vm_ops->get_policy(vma, addr);
+
+       sysfs_put_active_two(attr_sd);
+       return pol;
+}
+
+static int bin_migrate(struct vm_area_struct *vma, const nodemask_t *from,
+                       const nodemask_t *to, unsigned long flags)
+{
+       struct file *file = vma->vm_file;
+       struct bin_buffer *bb = file->private_data;
+       struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata;
+       int ret;
+
+       if (!bb->vm_ops || !bb->vm_ops->migrate)
+               return 0;
+
+       if (!sysfs_get_active_two(attr_sd))
+               return 0;
+
+       ret = bb->vm_ops->migrate(vma, from, to, flags);
+
+       sysfs_put_active_two(attr_sd);
+       return ret;
+}
+#endif
+
 static struct vm_operations_struct bin_vm_ops = {
        .open           = bin_vma_open,
        .close          = bin_vma_close,
        .fault          = bin_fault,
        .page_mkwrite   = bin_page_mkwrite,
        .access         = bin_access,
+#ifdef CONFIG_NUMA
+       .set_policy     = bin_set_policy,
+       .get_policy     = bin_get_policy,
+       .migrate        = bin_migrate,
+#endif
 };
 
 static int mmap(struct file *file, struct vm_area_struct *vma)
@@ -287,7 +350,6 @@ static int mmap(struct file *file, struct vm_area_struct *vma)
        struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata;
        struct bin_attribute *attr = attr_sd->s_bin_attr.bin_attr;
        struct kobject *kobj = attr_sd->s_parent->s_dir.kobj;
-       struct vm_operations_struct *vm_ops;
        int rc;
 
        mutex_lock(&bb->mutex);
@@ -302,24 +364,25 @@ static int mmap(struct file *file, struct vm_area_struct *vma)
                goto out_put;
 
        rc = attr->mmap(kobj, attr, vma);
-       vm_ops = vma->vm_ops;
-       vma->vm_ops = &bin_vm_ops;
        if (rc)
                goto out_put;
 
-       rc = -EINVAL;
-       if (bb->mmapped && bb->vm_ops != vma->vm_ops)
+       /*
+        * PowerPC's pci_mmap of legacy_mem uses shmem_zero_setup()
+        * to satisfy versions of X which crash if the mmap fails: that
+        * substitutes a new vm_file, and we don't then want bin_vm_ops.
+        */
+       if (vma->vm_file != file)
                goto out_put;
 
-#ifdef CONFIG_NUMA
        rc = -EINVAL;
-       if (vm_ops && ((vm_ops->set_policy || vm_ops->get_policy || vm_ops->migrate)))
+       if (bb->mmapped && bb->vm_ops != vma->vm_ops)
                goto out_put;
-#endif
 
        rc = 0;
        bb->mmapped = 1;
-       bb->vm_ops = vm_ops;
+       bb->vm_ops = vma->vm_ops;
+       vma->vm_ops = &bin_vm_ops;
 out_put:
        sysfs_put_active_two(attr_sd);
 out_unlock: