[PATCH] i386/x86-64: Don't call change_page_attr with a spinlock held

It's illegal because it can sleep.

Use a two step lookup scheme instead.  First look up the vm_struct, then
change the direct mapping, then finally unmap it.  That's ok because nobody
can change the particular virtual address range as long as the vm_struct is
still in the global list.

Also added some LinuxDoc documentation to iounmap.

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

arch/i386/mm/ioremap.c

@@ -223,9 +223,15 @@ void __iomem *ioremap_nocache (unsigned long phys_addr, unsigned long size)
 }
 EXPORT_SYMBOL(ioremap_nocache);
 
+/**
+ * iounmap - Free a IO remapping
+ * @addr: virtual address from ioremap_*
+ *
+ * Caller must ensure there is only one unmapping for the same pointer.
+ */
 void iounmap(volatile void __iomem *addr)
 {
-	struct vm_struct *p;
+	struct vm_struct *p, *o;
 
 	if ((void __force *)addr <= high_memory)
 		return;
@@ -239,22 +245,37 @@ void iounmap(volatile void __iomem *addr)
 			addr < phys_to_virt(ISA_END_ADDRESS))
 		return;
 
-	write_lock(&vmlist_lock);
-	p = __remove_vm_area((void *)(PAGE_MASK & (unsigned long __force)addr));
-	if (!p) { 
-		printk(KERN_WARNING "iounmap: bad address %p\n", addr);
+	addr = (volatile void *)(PAGE_MASK & (unsigned long __force)addr);
+
+	/* Use the vm area unlocked, assuming the caller
+	   ensures there isn't another iounmap for the same address
+	   in parallel. Reuse of the virtual address is prevented by
+	   leaving it in the global lists until we're done with it.
+	   cpa takes care of the direct mappings. */
+	read_lock(&vmlist_lock);
+	for (p = vmlist; p; p = p->next) {
+		if (p->addr == addr)
+			break;
+	}
+	read_unlock(&vmlist_lock);
+
+	if (!p) {
+		printk("iounmap: bad address %p\n", addr);
 		dump_stack();
-		goto out_unlock;
+		return;
 	}
 
+	/* Reset the direct mapping. Can block */
 	if ((p->flags >> 20) && p->phys_addr < virt_to_phys(high_memory) - 1) {
 		change_page_attr(virt_to_page(__va(p->phys_addr)),
 				 p->size >> PAGE_SHIFT,
 				 PAGE_KERNEL);
 		global_flush_tlb();
 	} 
-out_unlock:
-	write_unlock(&vmlist_lock);
+
+	/* Finally remove it */
+	o = remove_vm_area((void *)addr);
+	BUG_ON(p != o || o == NULL);
 	kfree(p); 
 }
 EXPORT_SYMBOL(iounmap);

arch/x86_64/mm/ioremap.c

@@ -247,22 +247,47 @@ void __iomem *ioremap_nocache (unsigned long phys_addr, unsigned long size)
 	return __ioremap(phys_addr, size, _PAGE_PCD);
 }
 
+/**
+ * iounmap - Free a IO remapping
+ * @addr: virtual address from ioremap_*
+ *
+ * Caller must ensure there is only one unmapping for the same pointer.
+ */
 void iounmap(volatile void __iomem *addr)
 {
-	struct vm_struct *p;
+	struct vm_struct *p, *o;
 
 	if (addr <= high_memory) 
 		return; 
 	if (addr >= phys_to_virt(ISA_START_ADDRESS) &&
 		addr < phys_to_virt(ISA_END_ADDRESS))
 		return;
 
-	write_lock(&vmlist_lock);
-	p = __remove_vm_area((void *)((unsigned long)addr & PAGE_MASK));
-	if (!p)
+	addr = (volatile void *)(PAGE_MASK & (unsigned long __force)addr);
+	/* Use the vm area unlocked, assuming the caller
+	   ensures there isn't another iounmap for the same address
+	   in parallel. Reuse of the virtual address is prevented by
+	   leaving it in the global lists until we're done with it.
+	   cpa takes care of the direct mappings. */
+	read_lock(&vmlist_lock);
+	for (p = vmlist; p; p = p->next) {
+		if (p->addr == addr)
+			break;
+	}
+	read_unlock(&vmlist_lock);
+
+	if (!p) {
 		printk("iounmap: bad address %p\n", addr);
-	else if (p->flags >> 20)
+		dump_stack();
+		return;
+	}
+
+	/* Reset the direct mapping. Can block */
+	if (p->flags >> 20)
 		ioremap_change_attr(p->phys_addr, p->size, 0);
-	write_unlock(&vmlist_lock);
+
+	/* Finally remove it */
+	o = remove_vm_area((void *)addr);
+	BUG_ON(p != o || o == NULL);
 	kfree(p); 
 }