* Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
*
- * Copyright (c) 2012, 2016, Intel Corporation.
+ * Copyright (c) 2012, 2017, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
#include <obd.h>
#include <obd_class.h>
#include "ptlrpc_internal.h"
+#include <lnet/lib-lnet.h> /* for CFS_FAIL_PTLRPC_OST_BULK_CB2 */
/**
* Helper function. Sends \a len bytes from \a base at offset \a offset
break;
}
- /* LU-6441: last md is not sent and desc->bd_md_count == 1 */
- if (OBD_FAIL_CHECK_ORSET(OBD_FAIL_PTLRPC_CLIENT_BULK_CB3,
- CFS_FAIL_ONCE) &&
- total_md > 1 && posted_md == total_md - 1) {
- posted_md++;
- continue;
- }
+ /* sanity.sh 224c: lets skip last md */
+ if (posted_md == desc->bd_md_max_brw - 1)
+ OBD_FAIL_CHECK_RESET(OBD_FAIL_PTLRPC_CLIENT_BULK_CB3,
+ CFS_FAIL_PTLRPC_OST_BULK_CB2);
/* Network is about to get at the memory */
if (ptlrpc_is_bulk_put_source(desc->bd_type))
desc->bd_portal, mbits, 0, 0);
else
rc = LNetGet(self_nid, desc->bd_mds[posted_md],
- peer_id, desc->bd_portal, mbits, 0);
+ peer_id, desc->bd_portal, mbits, 0, false);
posted_md++;
if (rc != 0) {
/* cleanup the state of the bulk for it will be reused */
if (req->rq_resend || req->rq_send_state == LUSTRE_IMP_REPLAY)
desc->bd_nob_transferred = 0;
- else
- LASSERT(desc->bd_nob_transferred == 0);
+ else if (desc->bd_nob_transferred != 0)
+ /* If the network failed after an RPC was sent, this condition
+ * could happen. Rather than assert (was here before), return
+ * an EIO error. */
+ RETURN(-EIO);
desc->bd_failure = 0;
LNET_MD_OP_GET : LNET_MD_OP_PUT);
ptlrpc_fill_bulk_md(&md, desc, posted_md);
- rc = LNetMEAttach(desc->bd_portal, peer, mbits, 0,
+ if (posted_md > 0 && posted_md + 1 == total_md &&
+ OBD_FAIL_CHECK(OBD_FAIL_PTLRPC_BULK_ATTACH)) {
+ rc = -ENOMEM;
+ } else {
+ rc = LNetMEAttach(desc->bd_portal, peer, mbits, 0,
LNET_UNLINK, LNET_INS_AFTER, &me_h);
+ }
if (rc != 0) {
CERROR("%s: LNetMEAttach failed x%llu/%d: rc = %d\n",
desc->bd_import->imp_obd->obd_name, mbits,
LASSERT(desc->bd_md_count >= 0);
mdunlink_iterate_helper(desc->bd_mds, desc->bd_md_max_brw);
req->rq_status = -ENOMEM;
+ desc->bd_registered = 0;
RETURN(-ENOMEM);
}
wait_queue_head_t *wq = (req->rq_set != NULL) ?
&req->rq_set->set_waitq :
&req->rq_reply_waitq;
- /* Network access will complete in finite time but the HUGE
- * timeout lets us CWARN for visibility of sluggish NALs */
- lwi = LWI_TIMEOUT_INTERVAL(cfs_time_seconds(LONG_UNLINK),
- cfs_time_seconds(1), NULL, NULL);
- rc = l_wait_event(*wq, !ptlrpc_client_bulk_active(req), &lwi);
- if (rc == 0) {
- ptlrpc_rqphase_move(req, req->rq_next_phase);
- RETURN(1);
- }
+ /*
+ * Network access will complete in finite time but the HUGE
+ * timeout lets us CWARN for visibility of sluggish NALs.
+ */
+ lwi = LWI_TIMEOUT_INTERVAL(cfs_time_seconds(LONG_UNLINK),
+ cfs_time_seconds(1), NULL, NULL);
+ rc = l_wait_event(*wq, !ptlrpc_client_bulk_active(req), &lwi);
+ if (rc == 0) {
+ ptlrpc_rqphase_move(req, req->rq_next_phase);
+ RETURN(1);
+ }
- LASSERT(rc == -ETIMEDOUT);
- DEBUG_REQ(D_WARNING, req, "Unexpectedly long timeout: desc %p",
- desc);
- }
- RETURN(0);
+ LASSERT(rc == -ETIMEDOUT);
+ DEBUG_REQ(D_WARNING, req, "Unexpectedly long timeout: desc %p",
+ desc);
+ }
+ RETURN(0);
}
static void ptlrpc_at_set_reply(struct ptlrpc_request *req, int flags)
req->rq_export->exp_obd->obd_minor);
}
- /* In order to keep interoprability with the client (< 2.3) which
- * doesn't have pb_jobid in ptlrpc_body, We have to shrink the
- * ptlrpc_body in reply buffer to ptlrpc_body_v2, otherwise, the
- * reply buffer on client will be overflow.
- *
- * XXX Remove this whenver we drop the interoprability with such client.
- */
- req->rq_replen = lustre_shrink_msg(req->rq_repmsg, 0,
- sizeof(struct ptlrpc_body_v2), 1);
-
if (req->rq_type != PTL_RPC_MSG_ERR)
req->rq_type = PTL_RPC_MSG_REPLY;
spin_unlock(&imp->imp_lock);
lustre_msg_set_last_xid(request->rq_reqmsg, min_xid);
- DEBUG_REQ(D_RPCTRACE, request, "Allocating new xid for "
- "resend on EINPROGRESS");
+ DEBUG_REQ(D_RPCTRACE, request,
+ "Allocating new XID for resend on EINPROGRESS");
}
if (request->rq_bulk != NULL) {
if (list_empty(&request->rq_unreplied_list) ||
request->rq_xid <= imp->imp_known_replied_xid) {
- DEBUG_REQ(D_ERROR, request, "xid: %llu, replied: %llu, "
- "list_empty:%d\n", request->rq_xid,
- imp->imp_known_replied_xid,
+ DEBUG_REQ(D_ERROR, request,
+ "xid=%llu, replied=%llu, list_empty=%d",
+ request->rq_xid, imp->imp_known_replied_xid,
list_empty(&request->rq_unreplied_list));
LBUG();
}
if (request->rq_resend_cb != NULL)
request->rq_resend_cb(request, &request->rq_async_args);
}
- if (request->rq_memalloc)
- mpflag = cfs_memory_pressure_get_and_set();
+ if (request->rq_memalloc)
+ mpflag = cfs_memory_pressure_get_and_set();
rc = sptlrpc_cli_wrap_request(request);
if (rc)
if (request->rq_bulk != NULL) {
rc = ptlrpc_register_bulk (request);
if (rc != 0)
- GOTO(out, rc);
+ GOTO(cleanup_bulk, rc);
/*
* All the mds in the request will have the same cpt
* encoded in the cookie. So we can just get the first
spin_lock(&request->rq_lock);
request->rq_err = 1;
spin_unlock(&request->rq_lock);
- request->rq_status = rc;
- GOTO(cleanup_bulk, rc);
- }
- } else {
- request->rq_repdata = NULL;
- request->rq_repmsg = NULL;
- }
+ request->rq_status = rc;
+ GOTO(cleanup_bulk, rc);
+ }
+ } else {
+ request->rq_repdata = NULL;
+ request->rq_repmsg = NULL;
+ }
rc = LNetMEAttach(request->rq_reply_portal,/*XXX FIXME bug 249*/
connection->c_peer, request->rq_xid, 0,
ptlrpc_pinger_sending_on_import(imp);
- DEBUG_REQ(D_INFO, request, "send flg=%x",
+ DEBUG_REQ(D_INFO, request, "send flags=%x",
lustre_msg_get_flags(request->rq_reqmsg));
rc = ptl_send_buf(&request->rq_req_md_h,
request->rq_reqbuf, request->rq_reqdata_len,
GOTO(out, rc);
cleanup_me:
- /* MEUnlink is safe; the PUT didn't even get off the ground, and
- * nobody apart from the PUT's target has the right nid+XID to
- * access the reply buffer. */
- rc2 = LNetMEUnlink(reply_me_h);
- LASSERT (rc2 == 0);
- /* UNLINKED callback called synchronously */
- LASSERT(!request->rq_receiving_reply);
+ /* MEUnlink is safe; the PUT didn't even get off the ground, and
+ * nobody apart from the PUT's target has the right nid+XID to
+ * access the reply buffer. */
+ rc2 = LNetMEUnlink(reply_me_h);
+ LASSERT (rc2 == 0);
+ /* UNLINKED callback called synchronously */
+ LASSERT(!request->rq_receiving_reply);
cleanup_bulk:
- /* We do sync unlink here as there was no real transfer here so
- * the chance to have long unlink to sluggish net is smaller here. */
+ /* We do sync unlink here as there was no real transfer here so
+ * the chance to have long unlink to sluggish net is smaller here. */
ptlrpc_unregister_bulk(request, 0);
+ if (request->rq_bulk != NULL)
+ request->rq_bulk->bd_registered = 0;
out:
if (rc == -ENOMEM) {
/* set rq_sent so that this request is treated
int ptlrpc_register_rqbd(struct ptlrpc_request_buffer_desc *rqbd)
{
struct ptlrpc_service *service = rqbd->rqbd_svcpt->scp_service;
- static struct lnet_process_id match_id = {LNET_NID_ANY, LNET_PID_ANY};
+ static struct lnet_process_id match_id = {
+ .nid = LNET_NID_ANY,
+ .pid = LNET_PID_ANY
+ };
int rc;
struct lnet_md md;
struct lnet_handle_me me_h;