All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH 1/5] libnvdimm, namespace: release labels properly on error
@ 2019-01-28  0:30 Wei Yang
  2019-01-28  0:30 ` [PATCH 2/5] libnvdimm, namespace: remove special handling when count == 0 Wei Yang
                   ` (4 more replies)
  0 siblings, 5 replies; 6+ messages in thread
From: Wei Yang @ 2019-01-28  0:30 UTC (permalink / raw)
  To: linux-nvdimm; +Cc: zwisler

In init_active_labels(), it iterates on ndr_mappings to create its
corresponding labels. When there is an error, it is supposed to release
those labels created. But current implementation doesn't handle this
well in two aspects:

  * when error happens during ndd check, labels are not released
  * just labels on current nd_mapping released, previous ones are lost

This patch extracts labels releasing code to error branch and release
labels on all nd_mapping besides only current one. By goto error branch
on error, it release all labels allocated.

Signed-off-by: Wei Yang <richardw.yang@linux.intel.com>
---
 drivers/nvdimm/namespace_devs.c | 19 ++++++++++++-------
 1 file changed, 12 insertions(+), 7 deletions(-)

diff --git a/drivers/nvdimm/namespace_devs.c b/drivers/nvdimm/namespace_devs.c
index 9471b9ca04f5..234c0c79726a 100644
--- a/drivers/nvdimm/namespace_devs.c
+++ b/drivers/nvdimm/namespace_devs.c
@@ -2451,7 +2451,7 @@ static struct device **create_namespaces(struct nd_region *nd_region)
 
 static int init_active_labels(struct nd_region *nd_region)
 {
-	int i;
+	int i, errno = -ENOMEM;
 
 	for (i = 0; i < nd_region->ndr_mappings; i++) {
 		struct nd_mapping *nd_mapping = &nd_region->mapping[i];
@@ -2476,7 +2476,8 @@ static int init_active_labels(struct nd_region *nd_region)
 					dev_name(&nd_mapping->nvdimm->dev),
 					test_bit(NDD_LOCKED, &nvdimm->flags)
 					? "locked" : "disabled");
-			return -ENXIO;
+			errno = -ENXIO;
+			goto error;
 		}
 		nd_mapping->ndd = ndd;
 		atomic_inc(&nvdimm->busy);
@@ -2500,16 +2501,20 @@ static int init_active_labels(struct nd_region *nd_region)
 			mutex_unlock(&nd_mapping->lock);
 		}
 
-		if (j >= count)
-			continue;
+		if (j < count)
+			goto error;
+	}
+
+	return 0;
 
+error:
+	for (; i >= 0; i--) {
+		struct nd_mapping *nd_mapping = &nd_region->mapping[i];
 		mutex_lock(&nd_mapping->lock);
 		nd_mapping_free_labels(nd_mapping);
 		mutex_unlock(&nd_mapping->lock);
-		return -ENOMEM;
 	}
-
-	return 0;
+	return errno;
 }
 
 int nd_region_register_namespaces(struct nd_region *nd_region, int *err)
-- 
2.19.1

_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm

^ permalink raw reply related	[flat|nested] 6+ messages in thread
end of thread, other threads:[~2019-02-01  6:42 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-01-28  0:30 [PATCH 1/5] libnvdimm, namespace: release labels properly on error Wei Yang
2019-01-28  0:30 ` [PATCH 2/5] libnvdimm, namespace: remove special handling when count == 0 Wei Yang
2019-01-28  0:30 ` [PATCH 3/5] libnvdimm, label: return nd_label directly instead of calculating it again Wei Yang
2019-01-28  0:30 ` [PATCH 4/5] libnvdimm, namespace: extract __init_active_labels to initialize labels for one nd_mapping Wei Yang
2019-01-28  0:30 ` [PATCH 5/5] libnvdimm, namespace: allocate devs for namespace just once Wei Yang
2019-02-01  6:42 ` [PATCH 1/5] libnvdimm, namespace: release labels properly on error Wei Yang

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.