Staging: binder: mmap fixes.

Only allow a binder file pointer to be mmapped once. The buffer management
code cannot deal with more then one area.
Also remove leftover mutex_unlock if mmap fails.

Signed-off-by: Arve Hjønnevåg <arve@android.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This commit is contained in:
Arve Hjønnevåg 2009-04-06 15:12:55 -07:00 committed by Greg Kroah-Hartman
parent 85e0b0cbbf
commit 4ddfc911e3
1 changed files with 9 additions and 1 deletions

View File

@ -2694,6 +2694,12 @@ static int binder_mmap(struct file *filp, struct vm_area_struct *vma)
}
vma->vm_flags = (vma->vm_flags | VM_DONTCOPY) & ~VM_MAYWRITE;
if (proc->buffer) {
ret = -EBUSY;
failure_string = "already mapped";
goto err_already_mapped;
}
area = get_vm_area(vma->vm_end - vma->vm_start, VM_IOREMAP);
if (area == NULL) {
ret = -ENOMEM;
@ -2741,10 +2747,12 @@ static int binder_mmap(struct file *filp, struct vm_area_struct *vma)
err_alloc_small_buf_failed:
kfree(proc->pages);
proc->pages = NULL;
err_alloc_pages_failed:
vfree(proc->buffer);
proc->buffer = NULL;
err_get_vm_area_failed:
mutex_unlock(&binder_lock);
err_already_mapped:
err_bad_arg:
printk(KERN_ERR "binder_mmap: %d %lx-%lx %s failed %d\n", proc->pid, vma->vm_start, vma->vm_end, failure_string, ret);
return ret;