| Message ID | 71ab90d6d2639d812587bf7afa280e2f27d21b1e.1535730440.git.swise@opengridcomputing.com (mailing list archive) |
|---|---|
| State | Accepted |
| Delegated to: | Jason Gunthorpe |
| Headers | show |
| Series | iw_cxgb4: only allow 1 flush on user qps | expand |
On Fri, Aug 31, 2018 at 07:15:56AM -0700, Steve Wise wrote: > Once the qp has been flushed, it cannot be flushed again. The user > qp flush logic wasn't enforcing it however. The bug can cause > touch-after-free crashes like: > > Unable to handle kernel paging request for data at address 0x000001ec > Faulting instruction address: 0xc008000016069100 > Oops: Kernel access of bad area, sig: 11 [#1] > ... > NIP [c008000016069100] flush_qp+0x80/0x480 [iw_cxgb4] > LR [c00800001606cd6c] c4iw_modify_qp+0x71c/0x11d0 [iw_cxgb4] > Call Trace: > [c00800001606cd6c] c4iw_modify_qp+0x71c/0x11d0 [iw_cxgb4] > [c00800001606e868] c4iw_ib_modify_qp+0x118/0x200 [iw_cxgb4] > [c0080000119eae80] ib_security_modify_qp+0xd0/0x3d0 [ib_core] > [c0080000119c4e24] ib_modify_qp+0xc4/0x2c0 [ib_core] > [c008000011df0284] iwcm_modify_qp_err+0x44/0x70 [iw_cm] > [c008000011df0fec] destroy_cm_id+0xcc/0x370 [iw_cm] > [c008000011ed4358] rdma_destroy_id+0x3c8/0x520 [rdma_cm] > [c0080000134b0540] ucma_close+0x90/0x1b0 [rdma_ucm] > [c000000000444da4] __fput+0xe4/0x2f0 > > So fix flush_qp() to only flush the wq once. > > Cc: stable@vger.kernel.org > Signed-off-by: Steve Wise <swise@opengridcomputing.com> > --- > drivers/infiniband/hw/cxgb4/qp.c | 6 ++++++ > 1 file changed, 6 insertions(+) Applied to for-rc, thanks Jason
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c index b3203afa3b1d..347fe18b1a41 100644 --- a/drivers/infiniband/hw/cxgb4/qp.c +++ b/drivers/infiniband/hw/cxgb4/qp.c @@ -1685,6 +1685,12 @@ static void flush_qp(struct c4iw_qp *qhp) schp = to_c4iw_cq(qhp->ibqp.send_cq); if (qhp->ibqp.uobject) { + + /* for user qps, qhp->wq.flushed is protected by qhp->mutex */ + if (qhp->wq.flushed) + return; + + qhp->wq.flushed = 1; t4_set_wq_in_error(&qhp->wq, 0); t4_set_cq_in_error(&rchp->cq); spin_lock_irqsave(&rchp->comp_handler_lock, flag);
Once the qp has been flushed, it cannot be flushed again. The user qp flush logic wasn't enforcing it however. The bug can cause touch-after-free crashes like: Unable to handle kernel paging request for data at address 0x000001ec Faulting instruction address: 0xc008000016069100 Oops: Kernel access of bad area, sig: 11 [#1] ... NIP [c008000016069100] flush_qp+0x80/0x480 [iw_cxgb4] LR [c00800001606cd6c] c4iw_modify_qp+0x71c/0x11d0 [iw_cxgb4] Call Trace: [c00800001606cd6c] c4iw_modify_qp+0x71c/0x11d0 [iw_cxgb4] [c00800001606e868] c4iw_ib_modify_qp+0x118/0x200 [iw_cxgb4] [c0080000119eae80] ib_security_modify_qp+0xd0/0x3d0 [ib_core] [c0080000119c4e24] ib_modify_qp+0xc4/0x2c0 [ib_core] [c008000011df0284] iwcm_modify_qp_err+0x44/0x70 [iw_cm] [c008000011df0fec] destroy_cm_id+0xcc/0x370 [iw_cm] [c008000011ed4358] rdma_destroy_id+0x3c8/0x520 [rdma_cm] [c0080000134b0540] ucma_close+0x90/0x1b0 [rdma_ucm] [c000000000444da4] __fput+0xe4/0x2f0 So fix flush_qp() to only flush the wq once. Cc: stable@vger.kernel.org Signed-off-by: Steve Wise <swise@opengridcomputing.com> --- drivers/infiniband/hw/cxgb4/qp.c | 6 ++++++ 1 file changed, 6 insertions(+)