From: Jason Gunthorpe jgg@nvidia.com
[ Upstream commit 98837c6c3d7285f6eca86480b6f7fac6880e27a8 ]
This value is locked under the file->mut, ensure it is held whenever touching it.
The case in ucma_migrate_id() is a race, while in ucma_free_uctx() it is already not possible for the write side to run, the movement is just for clarity.
Fixes: 88314e4dda1e ("RDMA/cma: add support for rdma_migrate_id()") Link: https://lore.kernel.org/r/20200818120526.702120-10-leon@kernel.org Signed-off-by: Leon Romanovsky leonro@mellanox.com Signed-off-by: Jason Gunthorpe jgg@nvidia.com Signed-off-by: Sasha Levin sashal@kernel.org --- drivers/infiniband/core/ucma.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c index d03dacaef7880..cf283b70bf877 100644 --- a/drivers/infiniband/core/ucma.c +++ b/drivers/infiniband/core/ucma.c @@ -586,6 +586,7 @@ static int ucma_free_ctx(struct ucma_context *ctx) list_move_tail(&uevent->list, &list); } list_del(&ctx->list); + events_reported = ctx->events_reported; mutex_unlock(&ctx->file->mut);
list_for_each_entry_safe(uevent, tmp, &list, list) { @@ -595,7 +596,6 @@ static int ucma_free_ctx(struct ucma_context *ctx) kfree(uevent); }
- events_reported = ctx->events_reported; mutex_destroy(&ctx->mutex); kfree(ctx); return events_reported; @@ -1678,7 +1678,9 @@ static ssize_t ucma_migrate_id(struct ucma_file *new_file,
cur_file = ctx->file; if (cur_file == new_file) { + mutex_lock(&cur_file->mut); resp.events_reported = ctx->events_reported; + mutex_unlock(&cur_file->mut); goto response; }