mlock: make mlock error return Posixly Correct

Rework Posix error return for mlock(). Posix requires error code for mlock*() system calls for some conditions that differ from what kernel low level functions, such as get_user_pages(), return for those conditions. For more info, see: http://marc.info/?l=linux-kernel&m=121750892930775&w=2 This patch provides the same translation of get_user_pages() error codes to posix specified error codes in the context of the mlock rework for unevictable lru. [akpm@linux-foundation.org: fix build] Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mlock: make mlock error return Posixly Correct
Rework Posix error return for mlock(). Posix requires error code for mlock*() system calls for some conditions that differ from what kernel low level functions, such as get_user_pages(), return for those conditions. For more info, see: http://marc.info/?l=linux-kernel&m=121750892930775&w=2 This patch provides the same translation of get_user_pages() error codes to posix specified error codes in the context of the mlock rework for unevictable lru. [akpm@linux-foundation.org: fix build] Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Lee Schermerhorn · Linus Torvalds
1 parent c11d69d8c8
Showing 2 changed files with 28 additions and 7 deletions Side-by-side Diff
mm/memory.c
mm/mlock.c
@@ -2821,7 +2821,7 @@
 			len, write, 0, NULL, NULL);
 	if (ret < 0)
 		return ret;
-	return ret == len ? 0 : -1;
+	return ret == len ? 0 : -EFAULT;
 }
  
 #if !defined(__HAVE_ARCH_GATE_AREA)
@@ -248,13 +248,26 @@
 			addr += PAGE_SIZE;	/* for next get_user_pages() */
 			nr_pages--;
 		}
+		ret = 0;
 	}
  
 	lru_add_drain_all();	/* to update stats */
  
-	return 0;	/* count entire vma as locked_vm */
+	return ret;	/* count entire vma as locked_vm */
 }
  
+/*
+ * convert get_user_pages() return value to posix mlock() error
+ */
+static int __mlock_posix_error_return(long retval)
+{
+	if (retval == -EFAULT)
+		retval = -ENOMEM;
+	else if (retval == -ENOMEM)
+		retval = -EAGAIN;
+	return retval;
+}
+
 #else /* CONFIG_UNEVICTABLE_LRU */
  
 /*
  
@@ -265,9 +278,15 @@
 				   int mlock)
 {
 	if (mlock && (vma->vm_flags & VM_LOCKED))
-		make_pages_present(start, end);
+		return make_pages_present(start, end);
 	return 0;
 }
+
+static inline int __mlock_posix_error_return(long retval)
+{
+	return 0;
+}
+
 #endif /* CONFIG_UNEVICTABLE_LRU */
  
 /**
@@ -434,10 +453,7 @@
 		downgrade_write(&mm->mmap_sem);
  
 		ret = __mlock_vma_pages_range(vma, start, end, 1);
-		if (ret > 0) {
-			mm->locked_vm -= ret;
-			ret = 0;
-		}
+
 		/*
 		 * Need to reacquire mmap sem in write mode, as our callers
 		 * expect this.  We have no support for atomically upgrading
@@ -451,6 +467,11 @@
 		/* non-NULL *prev must contain @start, but need to check @end */
 		if (!(*prev) || end > (*prev)->vm_end)
 			ret = -ENOMEM;
+		else if (ret > 0) {
+			mm->locked_vm -= ret;
+			ret = 0;
+		} else
+			ret = __mlock_posix_error_return(ret); /* translate if needed */
 	} else {
 		/*
 		 * TODO:  for unlocking, pages will already be resident, so
...	...	@@ -2821,7 +2821,7 @@
2821	2821	len, write, 0, NULL, NULL);
2822	2822	if (ret < 0)
2823	2823	return ret;
2824		- return ret == len ? 0 : -1;
	2824	+ return ret == len ? 0 : -EFAULT;
2825	2825	}
2826	2826
2827	2827	#if !defined(__HAVE_ARCH_GATE_AREA)
...	...	@@ -248,13 +248,26 @@
248	248	addr += PAGE_SIZE; /* for next get_user_pages() */
249	249	nr_pages--;
250	250	}
	251	+ ret = 0;
251	252	}
252	253
253	254	lru_add_drain_all(); /* to update stats */
254	255
255		- return 0; /* count entire vma as locked_vm */
	256	+ return ret; /* count entire vma as locked_vm */
256	257	}
257	258
	259	+/*
	260	+ * convert get_user_pages() return value to posix mlock() error
	261	+ */
	262	+static int __mlock_posix_error_return(long retval)
	263	+{
	264	+ if (retval == -EFAULT)
	265	+ retval = -ENOMEM;
	266	+ else if (retval == -ENOMEM)
	267	+ retval = -EAGAIN;
	268	+ return retval;
	269	+}
	270	+
258	271	#else /* CONFIG_UNEVICTABLE_LRU */
259	272
260	273	/*
261	274
...	...	@@ -265,9 +278,15 @@
265	278	int mlock)
266	279	{
267	280	if (mlock && (vma->vm_flags & VM_LOCKED))
268		- make_pages_present(start, end);
	281	+ return make_pages_present(start, end);
269	282	return 0;
270	283	}
	284	+
	285	+static inline int __mlock_posix_error_return(long retval)
	286	+{
	287	+ return 0;
	288	+}
	289	+
271	290	#endif /* CONFIG_UNEVICTABLE_LRU */
272	291
273	292	/**
...	...	@@ -434,10 +453,7 @@
434	453	downgrade_write(&mm->mmap_sem);
435	454
436	455	ret = __mlock_vma_pages_range(vma, start, end, 1);
437		- if (ret > 0) {
438		- mm->locked_vm -= ret;
439		- ret = 0;
440		- }
	456	+
441	457	/*
442	458	* Need to reacquire mmap sem in write mode, as our callers
443	459	* expect this. We have no support for atomically upgrading
...	...	@@ -451,6 +467,11 @@
451	467	/* non-NULL prev must contain @start, but need to check @end /
452	468	if (!(prev) \|\| end > (prev)->vm_end)
453	469	ret = -ENOMEM;
	470	+ else if (ret > 0) {
	471	+ mm->locked_vm -= ret;
	472	+ ret = 0;
	473	+ } else
	474	+ ret = __mlock_posix_error_return(ret); /* translate if needed */
454	475	} else {
455	476	/*
456	477	* TODO: for unlocking, pages will already be resident, so