aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/infiniband/core/uverbs_cmd.c
diff options
context:
space:
mode:
authorJason Gunthorpe <jgg@mellanox.com>2020-01-08 19:22:06 +0200
committerJason Gunthorpe <jgg@mellanox.com>2020-01-13 16:20:16 -0400
commit5c55cfd6a553d008fcd54e4a4e3fed1340ee5090 (patch)
tree0dd95d4166c15b2bee2fbb220f9f0d5e7d12d08f /drivers/infiniband/core/uverbs_cmd.c
parentRDMA/core: Make ib_uverbs_async_event_file into a uobject (diff)
downloadlinux-dev-5c55cfd6a553d008fcd54e4a4e3fed1340ee5090.tar.xz
linux-dev-5c55cfd6a553d008fcd54e4a4e3fed1340ee5090.zip
RDMA/core: Use READ_ONCE for ib_ufile.async_file
The writer for async_file holds the ucontext_lock, while the readers are left unlocked. Most readers rely on an implicit locking, either by having a uobject (which cannot be created before a context) or by holding the ib_ufile kref. However ib_uverbs_free_hw_resources() has no implicit lock and has a possible race. Make this all clear and sane by using READ_ONCE consistently. Link: https://lore.kernel.org/r/1578504126-9400-15-git-send-email-yishaih@mellanox.com Signed-off-by: Yishai Hadas <yishaih@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Diffstat (limited to 'drivers/infiniband/core/uverbs_cmd.c')
-rw-r--r--drivers/infiniband/core/uverbs_cmd.c2
1 files changed, 1 insertions, 1 deletions
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index ced1384d316b..29b1b5ad8836 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -1056,7 +1056,7 @@ err_free:
kfree(cq);
err_file:
if (ev_file)
- ib_uverbs_release_ucq(attrs->ufile, ev_file, obj);
+ ib_uverbs_release_ucq(ev_file, obj);
err:
uobj_alloc_abort(&obj->uevent.uobject, attrs);