Commit 6b3712c0 authored by Jason Gunthorpe's avatar Jason Gunthorpe

RDMA/core: Ensure that rdma_user_mmap_entry_remove() is a fence

The set of entry->driver_removed is missing locking, protect it with
xa_lock() which is held by the only reader.

Otherwise readers may continue to see driver_removed = false after
rdma_user_mmap_entry_remove() returns and may continue to try and
establish new mmaps.

Fixes: 3411f9f0 ("RDMA/core: Create mmap database and cookie helper functions")
Link: https://lore.kernel.org/r/20200115202041.GA17199@ziepe.caReviewed-by: default avatarGal Pressman <galpress@amazon.com>
Acked-by: default avatarMichal Kalderon <michal.kalderon@marvell.com>
Signed-off-by: default avatarJason Gunthorpe <jgg@mellanox.com>
parent e8b3a426
...@@ -232,7 +232,9 @@ void rdma_user_mmap_entry_remove(struct rdma_user_mmap_entry *entry) ...@@ -232,7 +232,9 @@ void rdma_user_mmap_entry_remove(struct rdma_user_mmap_entry *entry)
if (!entry) if (!entry)
return; return;
xa_lock(&entry->ucontext->mmap_xa);
entry->driver_removed = true; entry->driver_removed = true;
xa_unlock(&entry->ucontext->mmap_xa);
kref_put(&entry->ref, rdma_user_mmap_entry_free); kref_put(&entry->ref, rdma_user_mmap_entry_free);
} }
EXPORT_SYMBOL(rdma_user_mmap_entry_remove); EXPORT_SYMBOL(rdma_user_mmap_entry_remove);
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment