That's the problem with coding in the middle of the night. Thanks. Bob -----Original Message----- From: Dan Carpenter <dan.carpenter@xxxxxxxxxx> Sent: Thursday, April 22, 2021 5:42 AM To: kbuild@xxxxxxxxxxxx; Bob Pearson <rpearsonhpe@xxxxxxxxx>; jgg@xxxxxxxxxx; zyjzyj2000@xxxxxxxxx; linux-rdma@xxxxxxxxxxxxxxx Cc: lkp@xxxxxxxxx; kbuild-all@xxxxxxxxxxxx; Pearson, Robert B <robert.pearson2@xxxxxxx> Subject: [kbuild] Re: [PATCH for-next v4 10/10] RDMA/rxe: Disallow MR dereg and invalidate when bound Hi Bob, url: https://github.com/0day-ci/linux/commits/Bob-Pearson/RDMA-rxe-Implement-memory-windows/20210422-132958 base: https://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma.git for-next config: x86_64-randconfig-m001-20210421 (attached as .config) compiler: gcc-9 (Debian 9.3.0-22) 9.3.0 If you fix the issue, kindly add following tag as appropriate Reported-by: kernel test robot <lkp@xxxxxxxxx> Reported-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx> smatch warnings: drivers/infiniband/sw/rxe/rxe_mr.c:553 rxe_invalidate_mr() warn: ignoring unreachable code. vim +553 drivers/infiniband/sw/rxe/rxe_mr.c 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 529 int rxe_invalidate_mr(struct rxe_qp *qp, u32 rkey) 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 530 { 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 531 struct rxe_dev *rxe = to_rdev(qp->ibqp.device); 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 532 struct rxe_mr *mr; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 533 int ret; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 534 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 535 mr = rxe_pool_get_index(&rxe->mr_pool, rkey >> 8); 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 536 if (!mr) { 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 537 pr_err("%s: No MR for rkey %#x\n", __func__, rkey); 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 538 ret = -EINVAL; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 539 goto err; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 540 } 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 541 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 542 if (rkey != mr->ibmr.rkey) { 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 543 pr_err("%s: rkey (%#x) doesn't match mr->ibmr.rkey (%#x)\n", 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 544 __func__, rkey, mr->ibmr.rkey); 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 545 ret = -EINVAL; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 546 goto err_drop_ref; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 547 } 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 548 740526f943e87ee Bob Pearson 2021-04-22 549 if (atomic_read(&mr->num_mw) > 0) { 740526f943e87ee Bob Pearson 2021-04-22 550 pr_warn("%s: Attempt to invalidate an MR while bound to MWs\n", 740526f943e87ee Bob Pearson 2021-04-22 551 __func__); 740526f943e87ee Bob Pearson 2021-04-22 552 return -EINVAL; 740526f943e87ee Bob Pearson 2021-04-22 @553 goto err_drop_ref; Either the goto or the return should be deleted. I would have expected that it would be the return to be deleted, otherwise it needs a big comment why we're holding on to the reference. 740526f943e87ee Bob Pearson 2021-04-22 554 } 740526f943e87ee Bob Pearson 2021-04-22 555 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 556 mr->state = RXE_MR_STATE_FREE; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 557 ret = 0; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 558 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 559 err_drop_ref: 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 560 rxe_drop_ref(mr); 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 561 err: 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 562 return ret; 5f1e7f44d3a9cdb Bob Pearson 2021-04-22 563 } --- 0-DAY CI Kernel Test Service, Intel Corporation https://lists.01.org/hyperkitty/list/kbuild-all@xxxxxxxxxxxx
