summaryrefslogtreecommitdiffstats
path: root/include/rdma
diff options
context:
space:
mode:
authorNiels Dossche <dossche.niels@gmail.com>2022-02-28 20:51:44 +0100
committerJason Gunthorpe <jgg@nvidia.com>2022-04-04 14:45:51 -0300
commit22cbc6c2681a0a4fe76150270426e763d52353a4 (patch)
treebd907b1bdb5f54e7d414fb252d979a09a0ea95b9 /include/rdma
parentf543a3e82bb275349961f8507ee195f34132ffb4 (diff)
downloadlinux-stable-22cbc6c2681a0a4fe76150270426e763d52353a4.tar.gz
linux-stable-22cbc6c2681a0a4fe76150270426e763d52353a4.tar.bz2
linux-stable-22cbc6c2681a0a4fe76150270426e763d52353a4.zip
IB/rdmavt: add missing locks in rvt_ruc_loopback
The documentation of the function rvt_error_qp says both r_lock and s_lock need to be held when calling that function. It also asserts using lockdep that both of those locks are held. rvt_error_qp is called form rvt_send_cq, which is called from rvt_qp_complete_swqe, which is called from rvt_send_complete, which is called from rvt_ruc_loopback in two places. Both of these places do not hold r_lock. Fix this by acquiring a spin_lock of r_lock in both of these places. The r_lock acquiring cannot be added in rvt_qp_complete_swqe because some of its other callers already have r_lock acquired. Link: https://lore.kernel.org/r/20220228195144.71946-1-dossche.niels@gmail.com Signed-off-by: Niels Dossche <dossche.niels@gmail.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Diffstat (limited to 'include/rdma')
0 files changed, 0 insertions, 0 deletions