Limit repeated accesses to broken devices.

Signed-off-by: Takahiro Yasui <takahiro.yasui@hds.com>
Reviewed-by: Petr Rockai <prockai@redhat.com>

4 files changed, 55 insertions, 2 deletions

diff --git a/lib/device/dev-cache.c b/lib/device/dev-cache.c
index d43fcc02..ef93f4d6 100644
--- a/lib/device/dev-cache.c
+++ b/lib/device/dev-cache.c
@@ -104,6 +104,8 @@ struct device *dev_create_file(const char *filename, struct device *dev,
 	dev->dev = 0;
 	dev->fd = -1;
 	dev->open_count = 0;
+	dev->error_count = 0;
+	dev->max_error_count = NO_DEV_ERROR_COUNT_LIMIT;
 	dev->block_size = -1;
 	dev->read_ahead = -1;
 	memset(dev->pvid, 0, sizeof(dev->pvid));
@@ -125,6 +127,7 @@ static struct device *_dev_create(dev_t d)
 	dev->dev = d;
 	dev->fd = -1;
 	dev->open_count = 0;
+	dev->max_error_count = dev_disable_after_error_count();
 	dev->block_size = -1;
 	dev->read_ahead = -1;
 	dev->end = UINT64_C(0);
@@ -845,6 +848,22 @@ struct device *dev_iter_get(struct dev_iter *iter)
 	return NULL;
 }
 
+void dev_reset_error_count(struct cmd_context *cmd)
+{
+	struct dev_iter *iter;
+	struct device *dev;
+
+	if (!(iter = dev_iter_create(cmd->filter, 0))) {
+		log_error("Resetting device error count failed");
+		return;
+	}
+
+	for (dev = dev_iter_get(iter); dev; dev = dev_iter_get(iter))
+		dev->error_count = 0;
+
+	dev_iter_destroy(iter);
+}
+
 int dev_fd(struct device *dev)
 {
 	return dev->fd;
diff --git a/lib/device/dev-cache.h b/lib/device/dev-cache.h
index 0ade053c..c1c86d6a 100644
--- a/lib/device/dev-cache.h
+++ b/lib/device/dev-cache.h
@@ -53,4 +53,6 @@ struct dev_iter *dev_iter_create(struct dev_filter *f, int dev_scan);
 void dev_iter_destroy(struct dev_iter *iter);
 struct device *dev_iter_get(struct dev_iter *iter);
 
+void dev_reset_error_count(struct cmd_context *cmd);
+
 #endif
diff --git a/lib/device/dev-io.c b/lib/device/dev-io.c
index 1995bdea..eb80a894 100644
--- a/lib/device/dev-io.c
+++ b/lib/device/dev-io.c
@@ -603,18 +603,40 @@ void dev_close_all(void)
 	}
 }
 
+static inline int _dev_is_valid(struct device *dev)
+{
+	return (dev->max_error_count == NO_DEV_ERROR_COUNT_LIMIT ||
+		dev->error_count < dev->max_error_count);
+}
+
+static void _dev_inc_error_count(struct device *dev)
+{
+	if (++dev->error_count == dev->max_error_count)
+		log_warn("WARNING: Error counts reached a limit of %d. "
+			 "Device %s was disabled",
+			 dev->max_error_count, dev_name(dev));
+}
+
 int dev_read(struct device *dev, uint64_t offset, size_t len, void *buffer)
 {
 	struct device_area where;
+	int ret;
 
 	if (!dev->open_count)
 		return_0;
 
+	if (!_dev_is_valid(dev))
+		return 0;
+
 	where.dev = dev;
 	where.start = offset;
 	where.size = len;
 
-	return _aligned_io(&where, buffer, 0);
+	ret = _aligned_io(&where, buffer, 0);
+	if (!ret)
+		_dev_inc_error_count(dev);
+
+	return ret;
 }
 
 /*
@@ -670,17 +692,25 @@ int dev_append(struct device *dev, size_t len, void *buffer)
 int dev_write(struct device *dev, uint64_t offset, size_t len, void *buffer)
 {
 	struct device_area where;
+	int ret;
 
 	if (!dev->open_count)
 		return_0;
 
+	if (!_dev_is_valid(dev))
+		return 0;
+
 	where.dev = dev;
 	where.start = offset;
 	where.size = len;
 
 	dev->flags |= DEV_ACCESSED_W;
 
-	return _aligned_io(&where, buffer, 1);
+	ret = _aligned_io(&where, buffer, 1);
+	if (!ret)
+		_dev_inc_error_count(dev);
+
+	return ret;
 }
 
 int dev_set(struct device *dev, uint64_t offset, size_t len, int value)
diff --git a/lib/device/device.h b/lib/device/device.h
index 5a599508..694f503f 100644
--- a/lib/device/device.h
+++ b/lib/device/device.h
@@ -39,6 +39,8 @@ struct device {
 	/* private */
 	int fd;
 	int open_count;
+	int error_count;
+	int max_error_count;
 	int block_size;
 	int read_ahead;
 	uint32_t flags;