From: Oleg Drokin <green@linuxhacker.ru>
To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
linux-kernel@vger.kernel.org, devel@driverdev.osuosl.org,
Andreas Dilger <andreas.dilger@intel.com>
Cc: Hugues <morisset.hugues@gmail.com>,
Mikhail Pershin <mike.pershin@intel.com>,
Oleg Drokin <oleg.drokin@intel.com>
Subject: [PATCH 1/2] staging/lustre: use osc_reply_portal for OUT services
Date: Mon, 8 Sep 2014 21:41:27 -0400 [thread overview]
Message-ID: <1410226888-20377-2-git-send-email-green@linuxhacker.ru> (raw)
In-Reply-To: <1410226888-20377-1-git-send-email-green@linuxhacker.ru>
From: Mikhail Pershin <mike.pershin@intel.com>
OUT service is used to server both MDS-MDS updates and MDS-OST
therefore services on MDT and OST are set to use the same request
and reply portals to be fully unified and able to serve any type
of requests.
Signed-off-by: Mikhail Pershin <mike.pershin@intel.com>
Reviewed-on: http://review.whamcloud.com/8390
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3467
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
---
drivers/staging/lustre/lustre/fld/fld_request.c | 1 +
.../lustre/lustre/include/lustre/lustre_idl.h | 2 +-
drivers/staging/lustre/lustre/ldlm/ldlm_lib.c | 29 ++++++++++++++--------
3 files changed, 20 insertions(+), 12 deletions(-)
diff --git a/drivers/staging/lustre/lustre/fld/fld_request.c b/drivers/staging/lustre/lustre/fld/fld_request.c
index 02454db..0d361ff 100644
--- a/drivers/staging/lustre/lustre/fld/fld_request.c
+++ b/drivers/staging/lustre/lustre/fld/fld_request.c
@@ -426,6 +426,7 @@ int fld_client_rpc(struct obd_export *exp,
ptlrpc_request_set_replen(req);
req->rq_request_portal = FLD_REQUEST_PORTAL;
+ req->rq_reply_portal = MDC_REPLY_PORTAL;
ptlrpc_at_set_req_timeout(req);
if (fld_op == FLD_LOOKUP &&
diff --git a/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h b/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
index e529ad4..7b7457c 100644
--- a/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
+++ b/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
@@ -128,7 +128,7 @@
//#define PTLBD_BULK_PORTAL 21
#define MDS_SETATTR_PORTAL 22
#define MDS_READPAGE_PORTAL 23
-#define MDS_MDS_PORTAL 24
+#define OUT_PORTAL 24
#define MGC_REPLY_PORTAL 25
#define MGS_REQUEST_PORTAL 26
diff --git a/drivers/staging/lustre/lustre/ldlm/ldlm_lib.c b/drivers/staging/lustre/lustre/ldlm/ldlm_lib.c
index 9b3dac4..c21e30a 100644
--- a/drivers/staging/lustre/lustre/ldlm/ldlm_lib.c
+++ b/drivers/staging/lustre/lustre/ldlm/ldlm_lib.c
@@ -267,15 +267,10 @@ int client_obd_setup(struct obd_device *obddev, struct lustre_cfg *lcfg)
char *name = obddev->obd_type->typ_name;
ldlm_ns_type_t ns_type = LDLM_NS_TYPE_UNKNOWN;
int rc;
- char *cli_name = lustre_cfg_buf(lcfg, 0);
/* In a more perfect world, we would hang a ptlrpc_client off of
* obd_type and just use the values from there. */
- if (!strcmp(name, LUSTRE_OSC_NAME) ||
- (!(strcmp(name, LUSTRE_OSP_NAME)) &&
- (is_osp_on_mdt(cli_name) &&
- strstr(lustre_cfg_buf(lcfg, 1), "OST") != NULL))) {
- /* OSC or OSP_on_MDT for OSTs */
+ if (!strcmp(name, LUSTRE_OSC_NAME)) {
rq_portal = OST_REQUEST_PORTAL;
rp_portal = OSC_REPLY_PORTAL;
connect_op = OST_CONNECT;
@@ -283,17 +278,29 @@ int client_obd_setup(struct obd_device *obddev, struct lustre_cfg *lcfg)
cli->cl_sp_to = LUSTRE_SP_OST;
ns_type = LDLM_NS_TYPE_OSC;
} else if (!strcmp(name, LUSTRE_MDC_NAME) ||
- !strcmp(name, LUSTRE_LWP_NAME) ||
- (!strcmp(name, LUSTRE_OSP_NAME) &&
- (is_osp_on_mdt(cli_name) &&
- strstr(lustre_cfg_buf(lcfg, 1), "OST") == NULL))) {
- /* MDC or OSP_on_MDT for other MDTs */
+ !strcmp(name, LUSTRE_LWP_NAME)) {
rq_portal = MDS_REQUEST_PORTAL;
rp_portal = MDC_REPLY_PORTAL;
connect_op = MDS_CONNECT;
cli->cl_sp_me = LUSTRE_SP_CLI;
cli->cl_sp_to = LUSTRE_SP_MDT;
ns_type = LDLM_NS_TYPE_MDC;
+ } else if (!strcmp(name, LUSTRE_OSP_NAME)) {
+ if (strstr(lustre_cfg_buf(lcfg, 1), "OST") == NULL) {
+ /* OSP_on_MDT for other MDTs */
+ connect_op = MDS_CONNECT;
+ cli->cl_sp_to = LUSTRE_SP_MDT;
+ ns_type = LDLM_NS_TYPE_MDC;
+ rq_portal = OUT_PORTAL;
+ } else {
+ /* OSP on MDT for OST */
+ connect_op = OST_CONNECT;
+ cli->cl_sp_to = LUSTRE_SP_OST;
+ ns_type = LDLM_NS_TYPE_OSC;
+ rq_portal = OST_REQUEST_PORTAL;
+ }
+ rp_portal = OSC_REPLY_PORTAL;
+ cli->cl_sp_me = LUSTRE_SP_CLI;
} else if (!strcmp(name, LUSTRE_MGC_NAME)) {
rq_portal = MGS_REQUEST_PORTAL;
rp_portal = MGC_REPLY_PORTAL;
--
1.9.3
next prev parent reply other threads:[~2014-09-09 1:42 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-09-09 1:41 [PATCH 0/2] Lustre: get rid of is_osp_on_mdt Oleg Drokin
2014-09-09 1:41 ` Oleg Drokin [this message]
2014-09-09 1:41 ` [PATCH 2/2] staging/lustre: remove unused function is_osp_on_mdt Oleg Drokin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1410226888-20377-2-git-send-email-green@linuxhacker.ru \
--to=green@linuxhacker.ru \
--cc=andreas.dilger@intel.com \
--cc=devel@driverdev.osuosl.org \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mike.pershin@intel.com \
--cc=morisset.hugues@gmail.com \
--cc=oleg.drokin@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®