* [PATCH v5 01/25] sysfs: export sysfs_remove_file_self()
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 16:26 ` Jinpu Wang
2019-12-20 15:50 ` [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections Jack Wang
` (25 subsequent siblings)
26 siblings, 1 reply; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev, Jack Wang, Roman Pen, linux-kernel
From: Jack Wang <jinpu.wang@cloud.iono.com>
Function is going to be used in transport over RDMA module
in subsequent patches, so export it to GPL modules.
Signed-off-by: Roman Pen <roman.penyaev@profitbricks.com>
Acked-by: Tejun Heo <tj@kernel.org>
Cc: linux-kernel@vger.kernel.org
[jwang: extend the commit message]
Signed-off-by: Jack Wang <jinpu.wang@cloud.iono.com>
---
fs/sysfs/file.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c
index 130fc6fbcc03..1ff4672d7746 100644
--- a/fs/sysfs/file.c
+++ b/fs/sysfs/file.c
@@ -492,6 +492,7 @@ bool sysfs_remove_file_self(struct kobject *kobj, const struct attribute *attr)
kernfs_put(kn);
return ret;
}
+EXPORT_SYMBOL_GPL(sysfs_remove_file_self);
void sysfs_remove_files(struct kobject *kobj, const struct attribute * const *ptr)
{
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* Re: [PATCH v5 01/25] sysfs: export sysfs_remove_file_self()
2019-12-20 15:50 ` [PATCH v5 01/25] sysfs: export sysfs_remove_file_self() Jack Wang
@ 2019-12-20 16:26 ` Jinpu Wang
0 siblings, 0 replies; 45+ messages in thread
From: Jinpu Wang @ 2019-12-20 16:26 UTC (permalink / raw)
To: Jack Wang
Cc: linux-block, linux-rdma, Jens Axboe, Christoph Hellwig,
Sagi Grimberg, Bart Van Assche, Leon Romanovsky, Doug Ledford,
Danil Kipnis, rpenyaev, Jack Wang, Roman Pen, linux-kernel
On Fri, Dec 20, 2019 at 4:51 PM Jack Wang <jinpuwang@gmail.com> wrote:
>
> From: Jack Wang <jinpu.wang@cloud.iono.com>
>
> Function is going to be used in transport over RDMA module
> in subsequent patches, so export it to GPL modules.
>
> Signed-off-by: Roman Pen <roman.penyaev@profitbricks.com>
> Acked-by: Tejun Heo <tj@kernel.org>
> Cc: linux-kernel@vger.kernel.org
> [jwang: extend the commit message]
> Signed-off-by: Jack Wang <jinpu.wang@cloud.iono.com>
Sorry, I made a mistake, the email address was wrong, should be cloud.ionos.com,
I've fixed it locally, it will be fixed next round if necessary.
Regards,
Jack Wang
^ permalink raw reply [flat|nested] 45+ messages in thread
* [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
2019-12-20 15:50 ` [PATCH v5 01/25] sysfs: export sysfs_remove_file_self() Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-21 10:15 ` Leon Romanovsky
2019-12-20 15:50 ` [PATCH v5 03/25] rtrs: private headers with rtrs protocol structs and helpers Jack Wang
` (24 subsequent siblings)
26 siblings, 1 reply; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
Introduce public header which provides set of API functions to
establish RDMA connections from client to server machine using
RTRS protocol, which manages RDMA connections for each session,
does multipathing and load balancing.
Main functions for client (active) side:
rtrs_clt_open() - Creates set of RDMA connections incapsulated
in IBTRS session and returns pointer on RTRS
session object.
rtrs_clt_close() - Closes RDMA connections associated with RTRS
session.
rtrs_clt_request() - Requests zero-copy RDMA transfer to/from
server.
Main functions for server (passive) side:
rtrs_srv_open() - Starts listening for RTRS clients on specified
port and invokes RTRS callbacks for incoming
RDMA requests or link events.
rtrs_srv_close() - Closes RTRS server context.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs.h | 334 +++++++++++++++++++++++++++++
1 file changed, 334 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs.h
diff --git a/drivers/infiniband/ulp/rtrs/rtrs.h b/drivers/infiniband/ulp/rtrs/rtrs.h
new file mode 100644
index 000000000000..5b55ad163505
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs.h
@@ -0,0 +1,334 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#ifndef RTRS_H
+#define RTRS_H
+
+#include <linux/socket.h>
+#include <linux/scatterlist.h>
+
+struct rtrs_permit;
+struct rtrs_clt;
+struct rtrs_srv_ctx;
+struct rtrs_srv;
+struct rtrs_srv_op;
+
+/*
+ * Here goes RTRS client API
+ */
+
+/**
+ * enum rtrs_clt_link_ev - Events about connectivity state of a client
+ * @RTRS_CLT_LINK_EV_RECONNECTED Client was reconnected.
+ * @RTRS_CLT_LINK_EV_DISCONNECTED Client was disconnected.
+ */
+enum rtrs_clt_link_ev {
+ RTRS_CLT_LINK_EV_RECONNECTED,
+ RTRS_CLT_LINK_EV_DISCONNECTED,
+};
+
+/**
+ * Source and destination address of a path to be established
+ */
+struct rtrs_addr {
+ struct sockaddr_storage *src;
+ struct sockaddr_storage *dst;
+};
+
+typedef void (link_clt_ev_fn)(void *priv, enum rtrs_clt_link_ev ev);
+/**
+ * rtrs_clt_open() - Open a session to an RTRS server
+ * @priv: User supplied private data.
+ * @link_ev: Event notification for connection state changes
+ * @priv: User supplied data that was passed to rtrs_clt_open()
+ * @ev: Occurred event
+ * @sessname: name of the session
+ * @paths: Paths to be established defined by their src and dst addresses
+ * @path_cnt: Number of elemnts in the @paths array
+ * @port: port to be used by the RTRS session
+ * @pdu_sz: Size of extra payload which can be accessed after permit allocation.
+ * @max_inflight_msg: Max. number of parallel inflight messages for the session
+ * @max_segments: Max. number of segments per IO request
+ * @reconnect_delay_sec: time between reconnect tries
+ * @max_reconnect_attempts: Number of times to reconnect on error before giving
+ * up, 0 for * disabled, -1 for forever
+ *
+ * Starts session establishment with the rtrs_server. The function can block
+ * up to ~2000ms until it returns.
+ *
+ * Return a valid pointer on success otherwise PTR_ERR.
+ */
+struct rtrs_clt *rtrs_clt_open(void *priv, link_clt_ev_fn *link_ev,
+ const char *sessname,
+ const struct rtrs_addr *paths,
+ size_t path_cnt, short port,
+ size_t pdu_sz, u8 reconnect_delay_sec,
+ u16 max_segments,
+ s16 max_reconnect_attempts);
+
+/**
+ * rtrs_clt_close() - Close a session
+ * @sess: Session handle. Session is freed on return.
+ */
+void rtrs_clt_close(struct rtrs_clt *sess);
+
+/**
+ * rtrs_permit_from_pdu() - converts opaque pdu pointer to rtrs_permit
+ * @pdu: opaque pointer
+ */
+struct rtrs_permit *rtrs_permit_from_pdu(void *pdu);
+
+/**
+ * rtrs_permit_to_pdu() - converts rtrs_permit to opaque pdu pointer
+ * @permit: RTRS permit pointer
+ */
+void *rtrs_permit_to_pdu(struct rtrs_permit *permit);
+
+enum {
+ RTRS_PERMIT_NOWAIT = 0,
+ RTRS_PERMIT_WAIT = 1,
+};
+
+/**
+ * enum rtrs_clt_con_type() type of ib connection to use with a given permit
+ * @USR_CON - use connection reserved vor "service" messages
+ * @IO_CON - use a connection reserved for IO
+ */
+enum rtrs_clt_con_type {
+ RTRS_USR_CON,
+ RTRS_IO_CON
+};
+
+/**
+ * rtrs_clt_get_permit() - allocates permit for future RDMA operation
+ * @sess: Current session
+ * @con_type: Type of connection to use with the permit
+ * @wait: Wait type
+ *
+ * Description:
+ * Allocates permit for the following RDMA operation. Permit is used
+ * to preallocate all resources and to propagate memory pressure
+ * up earlier.
+ *
+ * Context:
+ * Can sleep if @wait == RTRS_TAG_WAIT
+ */
+struct rtrs_permit *rtrs_clt_get_permit(struct rtrs_clt *sess,
+ enum rtrs_clt_con_type con_type,
+ int wait);
+
+/**
+ * rtrs_clt_put_permit() - puts allocated permit
+ * @sess: Current session
+ * @permit: Permit to be freed
+ *
+ * Context:
+ * Does not matter
+ */
+void rtrs_clt_put_permit(struct rtrs_clt *sess, struct rtrs_permit *permit);
+
+typedef void (rtrs_conf_fn)(void *priv, int errno);
+/**
+ * rtrs_clt_request() - Request data transfer to/from server via RDMA.
+ *
+ * @dir: READ/WRITE
+ * @conf: callback function to be called as confirmation
+ * @sess: Session
+ * @permit: Preallocated permit
+ * @priv: User provided data, passed back with corresponding
+ * @(conf) confirmation.
+ * @vec: Message that is send to server together with the request.
+ * Sum of len of all @vec elements limited to <= IO_MSG_SIZE.
+ * Since the msg is copied internally it can be allocated on stack.
+ * @nr: Number of elements in @vec.
+ * @len: length of data send to/from server
+ * @sg: Pages to be sent/received to/from server.
+ * @sg_cnt: Number of elements in the @sg
+ *
+ * Return:
+ * 0: Success
+ * <0: Error
+ *
+ * On dir=READ rtrs client will request a data transfer from Server to client.
+ * The data that the server will respond with will be stored in @sg when
+ * the user receives an %RTRS_CLT_RDMA_EV_RDMA_REQUEST_WRITE_COMPL event.
+ * On dir=WRITE rtrs client will rdma write data in sg to server side.
+ */
+int rtrs_clt_request(int dir, rtrs_conf_fn *conf, struct rtrs_clt *sess,
+ struct rtrs_permit *permit, void *priv,
+ const struct kvec *vec, size_t nr, size_t len,
+ struct scatterlist *sg, unsigned int sg_cnt);
+
+/**
+ * rtrs_attrs - RTRS session attributes
+ */
+struct rtrs_attrs {
+ u32 queue_depth;
+ u32 max_io_size;
+ u8 sessname[NAME_MAX];
+ struct kobject *sess_kobj;
+};
+
+/**
+ * rtrs_clt_query() - queries RTRS session attributes
+ *
+ * Returns:
+ * 0 on success
+ * -ECOMM no connection to the server
+ */
+int rtrs_clt_query(struct rtrs_clt *sess, struct rtrs_attrs *attr);
+
+/*
+ * Here goes RTRS server API
+ */
+
+/**
+ * enum rtrs_srv_link_ev - Server link events
+ * @RTRS_SRV_LINK_EV_CONNECTED: Connection from client established
+ * @RTRS_SRV_LINK_EV_DISCONNECTED: Connection was disconnected, all
+ * connection RTRS resources were freed.
+ */
+enum rtrs_srv_link_ev {
+ RTRS_SRV_LINK_EV_CONNECTED,
+ RTRS_SRV_LINK_EV_DISCONNECTED,
+};
+
+/**
+ * rdma_ev_fn(): Event notification for RDMA operations
+ * If the callback returns a value != 0, an error message
+ * for the data transfer will be sent to the client.
+
+ * @sess: Session
+ * @priv: Private data set by rtrs_srv_set_sess_priv()
+ * @id: internal RTRS operation id
+ * @dir: READ/WRITE
+ * @data: Pointer to (bidirectional) rdma memory area:
+ * - in case of %RTRS_SRV_RDMA_EV_RECV contains
+ * data sent by the client
+ * - in case of %RTRS_SRV_RDMA_EV_WRITE_REQ points to the
+ * memory area where the response is to be written to
+ * @datalen: Size of the memory area in @data
+ * @usr: The extra user message sent by the client (%vec)
+ * @usrlen: Size of the user message
+ */
+typedef int (rdma_ev_fn)(struct rtrs_srv *sess, void *priv,
+ struct rtrs_srv_op *id, int dir,
+ void *data, size_t datalen, const void *usr,
+ size_t usrlen);
+
+/**
+ * link_ev_fn(): Events about connective state changes
+ * If the callback returns != 0 and the event
+ * %RTRS_SRV_LINK_EV_CONNECTED the corresponding session
+ * will be destroyed.
+ * @sess: Session
+ * @ev: event
+ * @priv: Private data from user if previously set with
+ * rtrs_srv_set_sess_priv()
+ */
+typedef int (link_ev_fn)(struct rtrs_srv *sess, enum rtrs_srv_link_ev ev,
+ void *priv);
+
+/**
+ * rtrs_srv_open() - open RTRS server context
+ * @ops: callback functions
+ *
+ * Creates server context with specified callbacks.
+ *
+ * Return a valid pointer on success otherwise PTR_ERR.
+ */
+struct rtrs_srv_ctx *rtrs_srv_open(rdma_ev_fn *rdma_ev, link_ev_fn *link_ev,
+ unsigned int port);
+
+/**
+ * rtrs_srv_close() - close RTRS server context
+ * @ctx: pointer to server context
+ *
+ * Closes RTRS server context with all client sessions.
+ */
+void rtrs_srv_close(struct rtrs_srv_ctx *ctx);
+
+/**
+ * rtrs_srv_resp_rdma() - Finish an RDMA request
+ *
+ * @id: Internal RTRS operation identifier
+ * @errno: Response Code send to the other side for this operation.
+ * 0 = success, <=0 error
+ *
+ * Finish a RDMA operation. A message is sent to the client and the
+ * corresponding memory areas will be released.
+ */
+void rtrs_srv_resp_rdma(struct rtrs_srv_op *id, int errno);
+
+/**
+ * rtrs_srv_set_sess_priv() - Set private pointer in rtrs_srv.
+ * @sess: Session
+ * @priv: The private pointer that is associated with the session.
+ */
+void rtrs_srv_set_sess_priv(struct rtrs_srv *sess, void *priv);
+
+/**
+ * rtrs_srv_get_sess_qdepth() - Get rtrs_srv qdepth.
+ * @sess: Session
+ */
+int rtrs_srv_get_queue_depth(struct rtrs_srv *sess);
+
+/**
+ * rtrs_srv_get_sess_name() - Get rtrs_srv peer hostname.
+ * @sess: Session
+ * @sessname: Sessname buffer
+ * @len: Length of sessname buffer
+ */
+int rtrs_srv_get_sess_name(struct rtrs_srv *sess, char *sessname, size_t len);
+
+/**
+ * rtrs_addr_to_sockaddr() - convert path string "src,dst" to sockaddreses
+ * @str string containing source and destination addr of a path
+ * separated by comma. I.e. "ip:1.1.1.1,ip:1.1.1.2". If str
+ * contains only one address it's considered to be destination.
+ * @len string length
+ * @addr->dst will be set to the destination sockadddr.
+ * @addr->src will be set to the source address or to NULL
+ * if str doesn't contain any sorce address.
+ *
+ * Returns zero if conversion successful. Non-zero otherwise.
+ */
+int rtrs_addr_to_sockaddr(const char *str, size_t len, short port,
+ struct rtrs_addr *addr);
+
+/**
+ * sockaddr_to_str() - convert sockaddr to a string.
+ * @addr the sockadddr structure to be converted.
+ * @buf string containing socket addr.
+ * @len string length.
+ *
+ * The return value is the number of characters written into buf not
+ * including the trailing '\0'. If len is == 0 the function returns 0..
+ */
+int sockaddr_to_str(const struct sockaddr *addr, char *buf, size_t len);
+#endif
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* Re: [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections
2019-12-20 15:50 ` [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections Jack Wang
@ 2019-12-21 10:15 ` Leon Romanovsky
2019-12-21 14:27 ` Danil Kipnis
0 siblings, 1 reply; 45+ messages in thread
From: Leon Romanovsky @ 2019-12-21 10:15 UTC (permalink / raw)
To: Jack Wang
Cc: linux-block, linux-rdma, axboe, hch, sagi, bvanassche, dledford,
danil.kipnis, jinpu.wang, rpenyaev
On Fri, Dec 20, 2019 at 04:50:46PM +0100, Jack Wang wrote:
> From: Jack Wang <jinpu.wang@cloud.ionos.com>
>
> Introduce public header which provides set of API functions to
> establish RDMA connections from client to server machine using
> RTRS protocol, which manages RDMA connections for each session,
> does multipathing and load balancing.
>
> Main functions for client (active) side:
>
> rtrs_clt_open() - Creates set of RDMA connections incapsulated
> in IBTRS session and returns pointer on RTRS
> session object.
> rtrs_clt_close() - Closes RDMA connections associated with RTRS
> session.
> rtrs_clt_request() - Requests zero-copy RDMA transfer to/from
> server.
>
> Main functions for server (passive) side:
>
> rtrs_srv_open() - Starts listening for RTRS clients on specified
> port and invokes RTRS callbacks for incoming
> RDMA requests or link events.
> rtrs_srv_close() - Closes RTRS server context.
>
> Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
> ---
> drivers/infiniband/ulp/rtrs/rtrs.h | 334 +++++++++++++++++++++++++++++
> 1 file changed, 334 insertions(+)
> create mode 100644 drivers/infiniband/ulp/rtrs/rtrs.h
>
> diff --git a/drivers/infiniband/ulp/rtrs/rtrs.h b/drivers/infiniband/ulp/rtrs/rtrs.h
> new file mode 100644
> index 000000000000..5b55ad163505
> --- /dev/null
> +++ b/drivers/infiniband/ulp/rtrs/rtrs.h
> @@ -0,0 +1,334 @@
> +/* SPDX-License-Identifier: GPL-2.0-or-later */
> +/*
> + * InfiniBand Transport Layer
> + *
> + * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
> + * Authors: Fabian Holler <mail@fholler.de>
> + * Jack Wang <jinpu.wang@profitbricks.com>
> + * Kleber Souza <kleber.souza@profitbricks.com>
> + * Danil Kipnis <danil.kipnis@profitbricks.com>
> + * Roman Penyaev <roman.penyaev@profitbricks.com>
> + * Milind Dumbare <Milind.dumbare@gmail.com>
> + *
> + * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
> + * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
> + * Roman Penyaev <roman.penyaev@profitbricks.com>
> + *
> + * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
> + * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
> + * Jinpu Wang <jinpu.wang@cloud.ionos.com>
> + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> + */
> +
> +/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
> + * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
> + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> + * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
> + * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
> + */
Perhaps it is normal practice to write half a company as authors,
and I'm wrong in the following, but code authorship is determined by
multiple tags in the commit messages.
Thanks
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections
2019-12-21 10:15 ` Leon Romanovsky
@ 2019-12-21 14:27 ` Danil Kipnis
2019-12-22 7:36 ` Leon Romanovsky
0 siblings, 1 reply; 45+ messages in thread
From: Danil Kipnis @ 2019-12-21 14:27 UTC (permalink / raw)
To: Leon Romanovsky
Cc: Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Bart Van Assche, Doug Ledford,
Jack Wang, rpenyaev
Hi Leon,
On Sat, Dec 21, 2019 at 11:15 AM Leon Romanovsky <leon@kernel.org> wrote:
>
> Perhaps it is normal practice to write half a company as authors,
> and I'm wrong in the following, but code authorship is determined by
> multiple tags in the commit messages.
Different developers contributed to the driver over the last several
years. Currently they are not working any more on this code. What tags
in the commit message do you think would be appropriate to give those
people credit for their work?
Best,
Danil
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections
2019-12-21 14:27 ` Danil Kipnis
@ 2019-12-22 7:36 ` Leon Romanovsky
2019-12-23 7:38 ` Jinpu Wang
0 siblings, 1 reply; 45+ messages in thread
From: Leon Romanovsky @ 2019-12-22 7:36 UTC (permalink / raw)
To: Danil Kipnis
Cc: Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Bart Van Assche, Doug Ledford,
Jack Wang, rpenyaev
On Sat, Dec 21, 2019 at 03:27:55PM +0100, Danil Kipnis wrote:
> Hi Leon,
>
> On Sat, Dec 21, 2019 at 11:15 AM Leon Romanovsky <leon@kernel.org> wrote:
> >
> > Perhaps it is normal practice to write half a company as authors,
> > and I'm wrong in the following, but code authorship is determined by
> > multiple tags in the commit messages.
>
> Different developers contributed to the driver over the last several
> years. Currently they are not working any more on this code. What tags
> in the commit message do you think would be appropriate to give those
> people credit for their work?
Signed-of-by/Co-developed-../e.t.c
But honestly without looking in your company contract, I'm pretty sure
that those people are not eligible for special authorship rights and
credits beyond already payed by the employer.
Thanks
>
> Best,
> Danil
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections
2019-12-22 7:36 ` Leon Romanovsky
@ 2019-12-23 7:38 ` Jinpu Wang
2019-12-23 8:04 ` Leon Romanovsky
0 siblings, 1 reply; 45+ messages in thread
From: Jinpu Wang @ 2019-12-23 7:38 UTC (permalink / raw)
To: Leon Romanovsky
Cc: Danil Kipnis, Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Bart Van Assche, Doug Ledford,
rpenyaev
On Sun, Dec 22, 2019 at 8:36 AM Leon Romanovsky <leon@kernel.org> wrote:
>
> On Sat, Dec 21, 2019 at 03:27:55PM +0100, Danil Kipnis wrote:
> > Hi Leon,
> >
> > On Sat, Dec 21, 2019 at 11:15 AM Leon Romanovsky <leon@kernel.org> wrote:
> > >
> > > Perhaps it is normal practice to write half a company as authors,
> > > and I'm wrong in the following, but code authorship is determined by
> > > multiple tags in the commit messages.
> >
> > Different developers contributed to the driver over the last several
> > years. Currently they are not working any more on this code. What tags
> > in the commit message do you think would be appropriate to give those
> > people credit for their work?
>
> Signed-of-by/Co-developed-../e.t.c
>
> But honestly without looking in your company contract, I'm pretty sure
> that those people are not eligible for special authorship rights and
> credits beyond already payed by the employer.
>
Hi, Leon,
Thanks for the suggestion, how about only remove the authors for the
new entry, only keep the company copyright?
> +/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
> + * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
> + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> + * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
> + * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
> + */
The older entries were there, I think it's not polite to remove them.
Regards,
Jack
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections
2019-12-23 7:38 ` Jinpu Wang
@ 2019-12-23 8:04 ` Leon Romanovsky
2019-12-23 10:31 ` Jinpu Wang
0 siblings, 1 reply; 45+ messages in thread
From: Leon Romanovsky @ 2019-12-23 8:04 UTC (permalink / raw)
To: Jinpu Wang
Cc: Danil Kipnis, Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Bart Van Assche, Doug Ledford,
rpenyaev
On Mon, Dec 23, 2019 at 08:38:54AM +0100, Jinpu Wang wrote:
> On Sun, Dec 22, 2019 at 8:36 AM Leon Romanovsky <leon@kernel.org> wrote:
> >
> > On Sat, Dec 21, 2019 at 03:27:55PM +0100, Danil Kipnis wrote:
> > > Hi Leon,
> > >
> > > On Sat, Dec 21, 2019 at 11:15 AM Leon Romanovsky <leon@kernel.org> wrote:
> > > >
> > > > Perhaps it is normal practice to write half a company as authors,
> > > > and I'm wrong in the following, but code authorship is determined by
> > > > multiple tags in the commit messages.
> > >
> > > Different developers contributed to the driver over the last several
> > > years. Currently they are not working any more on this code. What tags
> > > in the commit message do you think would be appropriate to give those
> > > people credit for their work?
> >
> > Signed-of-by/Co-developed-../e.t.c
> >
> > But honestly without looking in your company contract, I'm pretty sure
> > that those people are not eligible for special authorship rights and
> > credits beyond already payed by the employer.
> >
> Hi, Leon,
>
> Thanks for the suggestion, how about only remove the authors for the
> new entry, only keep the company copyright?
> > +/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
> > + * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
> > + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> > + * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
> > + * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
> > + */
>
> The older entries were there, I think it's not polite to remove them.
From our point of view, this is brand new code and it doesn't matter how
many internal iterations you had prior submission. If you want to be
polite, your company shall issue official press release and mention
all those names there as main contributors for RTRS success.
You can find a lot of examples of "Authors:" in the kernel code, but
they one of two: code from pre-git era or copy/paste multiplied by
cargo cult.
Thanks
>
> Regards,
> Jack
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections
2019-12-23 8:04 ` Leon Romanovsky
@ 2019-12-23 10:31 ` Jinpu Wang
0 siblings, 0 replies; 45+ messages in thread
From: Jinpu Wang @ 2019-12-23 10:31 UTC (permalink / raw)
To: Leon Romanovsky
Cc: Danil Kipnis, Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Bart Van Assche, Doug Ledford,
rpenyaev
On Mon, Dec 23, 2019 at 9:04 AM Leon Romanovsky <leon@kernel.org> wrote:
>
> On Mon, Dec 23, 2019 at 08:38:54AM +0100, Jinpu Wang wrote:
> > On Sun, Dec 22, 2019 at 8:36 AM Leon Romanovsky <leon@kernel.org> wrote:
> > >
> > > On Sat, Dec 21, 2019 at 03:27:55PM +0100, Danil Kipnis wrote:
> > > > Hi Leon,
> > > >
> > > > On Sat, Dec 21, 2019 at 11:15 AM Leon Romanovsky <leon@kernel.org> wrote:
> > > > >
> > > > > Perhaps it is normal practice to write half a company as authors,
> > > > > and I'm wrong in the following, but code authorship is determined by
> > > > > multiple tags in the commit messages.
> > > >
> > > > Different developers contributed to the driver over the last several
> > > > years. Currently they are not working any more on this code. What tags
> > > > in the commit message do you think would be appropriate to give those
> > > > people credit for their work?
> > >
> > > Signed-of-by/Co-developed-../e.t.c
> > >
> > > But honestly without looking in your company contract, I'm pretty sure
> > > that those people are not eligible for special authorship rights and
> > > credits beyond already payed by the employer.
> > >
> > Hi, Leon,
> >
> > Thanks for the suggestion, how about only remove the authors for the
> > new entry, only keep the company copyright?
> > > +/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
> > > + * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
> > > + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> > > + * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
> > > + * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
> > > + */
> >
> > The older entries were there, I think it's not polite to remove them.
>
> From our point of view, this is brand new code and it doesn't matter how
> many internal iterations you had prior submission. If you want to be
> polite, your company shall issue official press release and mention
> all those names there as main contributors for RTRS success.
>
> You can find a lot of examples of "Authors:" in the kernel code, but
> they one of two: code from pre-git era or copy/paste multiplied by
> cargo cult.
>
> Thanks
>
Ok, will only keep the copyright lines, and remove Authors.
Thanks
^ permalink raw reply [flat|nested] 45+ messages in thread
* [PATCH v5 03/25] rtrs: private headers with rtrs protocol structs and helpers
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
2019-12-20 15:50 ` [PATCH v5 01/25] sysfs: export sysfs_remove_file_self() Jack Wang
2019-12-20 15:50 ` [PATCH v5 02/25] rtrs: public interface header to establish RDMA connections Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 04/25] rtrs: core: lib functions shared between client and server modules Jack Wang
` (23 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
These are common private headers with rtrs protocol structures,
logging, sysfs and other helper functions, which are used on
both client and server sides.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-log.h | 50 +++
drivers/infiniband/ulp/rtrs/rtrs-pri.h | 426 +++++++++++++++++++++++++
2 files changed, 476 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-log.h
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-pri.h
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-log.h b/drivers/infiniband/ulp/rtrs/rtrs-log.h
new file mode 100644
index 000000000000..69ab002f8b7d
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-log.h
@@ -0,0 +1,50 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#ifndef RTRS_LOG_H
+#define RTRS_LOG_H
+
+#define rtrs_prefix(obj) (obj->sessname)
+
+#define rtrs_log(fn, obj, fmt, ...) \
+ fn("<%s>: " fmt, rtrs_prefix(obj), ##__VA_ARGS__)
+
+#define rtrs_err(obj, fmt, ...) \
+ rtrs_log(pr_err, obj, fmt, ##__VA_ARGS__)
+#define rtrs_err_rl(obj, fmt, ...) \
+ rtrs_log(pr_err_ratelimited, obj, fmt, ##__VA_ARGS__)
+#define rtrs_wrn(obj, fmt, ...) \
+ rtrs_log(pr_warn, obj, fmt, ##__VA_ARGS__)
+#define rtrs_wrn_rl(obj, fmt, ...) \
+ rtrs_log(pr_warn_ratelimited, obj, fmt, ##__VA_ARGS__)
+#define rtrs_info(obj, fmt, ...) \
+ rtrs_log(pr_info, obj, fmt, ##__VA_ARGS__)
+#define rtrs_info_rl(obj, fmt, ...) \
+ rtrs_log(pr_info_ratelimited, obj, fmt, ##__VA_ARGS__)
+
+#endif /* RTRS_LOG_H */
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-pri.h b/drivers/infiniband/ulp/rtrs/rtrs-pri.h
new file mode 100644
index 000000000000..5ea335bdc9ea
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-pri.h
@@ -0,0 +1,426 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Swapnil Ingle <swapnil.ingle@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+
+#ifndef RTRS_PRI_H
+#define RTRS_PRI_H
+
+#include <linux/uuid.h>
+#include <rdma/rdma_cm.h>
+#include <rdma/ib_verbs.h>
+#include <rdma/ib.h>
+
+#include "rtrs.h"
+
+#define RTRS_PROTO_VER_MAJOR 2
+#define RTRS_PROTO_VER_MINOR 0
+
+#define RTRS_PROTO_VER_STRING __stringify(RTRS_PROTO_VER_MAJOR) "." \
+ __stringify(RTRS_PROTO_VER_MINOR)
+
+enum rtrs_imm_const {
+ MAX_IMM_TYPE_BITS = 4,
+ MAX_IMM_TYPE_MASK = ((1 << MAX_IMM_TYPE_BITS) - 1),
+ MAX_IMM_PAYL_BITS = 28,
+ MAX_IMM_PAYL_MASK = ((1 << MAX_IMM_PAYL_BITS) - 1),
+};
+
+enum rtrs_imm_type {
+ RTRS_IO_REQ_IMM = 0, /* client to server */
+ RTRS_IO_RSP_IMM = 1, /* server to client */
+ RTRS_IO_RSP_W_INV_IMM = 2, /* server to client */
+
+ RTRS_HB_MSG_IMM = 8,
+ RTRS_HB_ACK_IMM = 9,
+
+ RTRS_LAST_IMM,
+};
+
+enum {
+ SERVICE_CON_QUEUE_DEPTH = 512,
+
+ MIN_RTR_CNT = 1,
+ MAX_RTR_CNT = 7,
+
+ MAX_PATHS_NUM = 128,
+
+ /*
+ * With the current size of the tag allocated on the client, 4K
+ * is the maximum number of tags we can allocate. This number is
+ * also used on the client to allocate the IU for the user connection
+ * to receive the RDMA addresses from the server.
+ */
+ MAX_SESS_QUEUE_DEPTH = 4096,
+
+ RTRS_HB_INTERVAL_MS = 5000,
+ RTRS_HB_MISSED_MAX = 5,
+
+ RTRS_MAGIC = 0x1BBD,
+ RTRS_PROTO_VER = (RTRS_PROTO_VER_MAJOR << 8) | RTRS_PROTO_VER_MINOR,
+};
+
+struct rtrs_ib_dev;
+
+struct rtrs_ib_dev_pool_ops {
+ struct rtrs_ib_dev *(*alloc)(void);
+ void (*free)(struct rtrs_ib_dev *dev);
+ int (*init)(struct rtrs_ib_dev *dev);
+ void (*deinit)(struct rtrs_ib_dev *dev);
+};
+
+struct rtrs_ib_dev_pool {
+ struct mutex mutex;
+ struct list_head list;
+ enum ib_pd_flags pd_flags;
+ const struct rtrs_ib_dev_pool_ops *ops;
+};
+
+struct rtrs_ib_dev {
+ struct ib_device *ib_dev;
+ struct ib_pd *ib_pd;
+ struct kref ref;
+ struct list_head entry;
+ struct rtrs_ib_dev_pool *pool;
+};
+
+struct rtrs_con {
+ struct rtrs_sess *sess;
+ struct ib_qp *qp;
+ struct ib_cq *cq;
+ struct rdma_cm_id *cm_id;
+ unsigned int cid;
+};
+
+typedef void (rtrs_hb_handler_t)(struct rtrs_con *con);
+
+struct rtrs_sess {
+ struct list_head entry;
+ struct sockaddr_storage dst_addr;
+ struct sockaddr_storage src_addr;
+ char sessname[NAME_MAX];
+ uuid_t uuid;
+ struct rtrs_con **con;
+ unsigned int con_num;
+ unsigned int recon_cnt;
+ struct rtrs_ib_dev *dev;
+ int dev_ref;
+ struct ib_cqe *hb_cqe;
+ rtrs_hb_handler_t *hb_err_handler;
+ struct workqueue_struct *hb_wq;
+ struct delayed_work hb_dwork;
+ unsigned int hb_interval_ms;
+ unsigned int hb_missed_cnt;
+ unsigned int hb_missed_max;
+};
+
+struct rtrs_iu {
+ struct list_head list;
+ struct ib_cqe cqe;
+ dma_addr_t dma_addr;
+ void *buf;
+ size_t size;
+ enum dma_data_direction direction;
+};
+
+/**
+ * enum rtrs_msg_types - RTRS message types.
+ * @RTRS_MSG_INFO_REQ: Client additional info request to the server
+ * @RTRS_MSG_INFO_RSP: Server additional info response to the client
+ * @RTRS_MSG_WRITE: Client writes data per RDMA to server
+ * @RTRS_MSG_READ: Client requests data transfer from server
+ * @RTRS_MSG_RKEY_RSP: Server refreshed rkey for rbuf
+ */
+enum rtrs_msg_types {
+ RTRS_MSG_INFO_REQ,
+ RTRS_MSG_INFO_RSP,
+ RTRS_MSG_WRITE,
+ RTRS_MSG_READ,
+ RTRS_MSG_RKEY_RSP,
+};
+
+/**
+ * enum rtrs_msg_flags - RTRS message flags.
+ * @RTRS_NEED_INVAL: Send invalidation in response.
+ * @RTRS_MSG_NEW_RKEY_F: Send refreshed rkey in response.
+ */
+enum rtrs_msg_flags {
+ RTRS_MSG_NEED_INVAL_F = 1 << 0,
+ RTRS_MSG_NEW_RKEY_F = 1 << 1,
+};
+
+/**
+ * struct rtrs_sg_desc - RDMA-Buffer entry description
+ * @addr: Address of RDMA destination buffer
+ * @key: Authorization rkey to write to the buffer
+ * @len: Size of the buffer
+ */
+struct rtrs_sg_desc {
+ __le64 addr;
+ __le32 key;
+ __le32 len;
+};
+
+/**
+ * struct rtrs_msg_conn_req - Client connection request to the server
+ * @magic: RTRS magic
+ * @version: RTRS protocol version
+ * @cid: Current connection id
+ * @cid_num: Number of connections per session
+ * @recon_cnt: Reconnections counter
+ * @sess_uuid: UUID of a session (path)
+ * @paths_uuid: UUID of a group of sessions (paths)
+ *
+ * NOTE: max size 56 bytes, see man rdma_connect().
+ */
+struct rtrs_msg_conn_req {
+ u8 __cma_version; /* Is set to 0 by cma.c in case of
+ * AF_IB, do not touch that.
+ */
+ u8 __ip_version; /* On sender side that should be
+ * set to 0, or cma_save_ip_info()
+ * extract garbage and will fail.
+ */
+ __le16 magic;
+ __le16 version;
+ __le16 cid;
+ __le16 cid_num;
+ __le16 recon_cnt;
+ uuid_t sess_uuid;
+ uuid_t paths_uuid;
+ u8 reserved[12];
+};
+
+/**
+ * struct rtrs_msg_conn_rsp - Server connection response to the client
+ * @magic: RTRS magic
+ * @version: RTRS protocol version
+ * @errno: If rdma_accept() then 0, if rdma_reject() indicates error
+ * @queue_depth: max inflight messages (queue-depth) in this session
+ * @max_io_size: max io size server supports
+ * @max_hdr_size: max msg header size server supports
+ *
+ * NOTE: size is 56 bytes, max possible is 136 bytes, see man rdma_accept().
+ */
+struct rtrs_msg_conn_rsp {
+ __le16 magic;
+ __le16 version;
+ __le16 errno;
+ __le16 queue_depth;
+ __le32 max_io_size;
+ __le32 max_hdr_size;
+ __le32 flags;
+ u8 reserved[36];
+};
+
+/**
+ * struct rtrs_msg_info_req
+ * @type: @RTRS_MSG_INFO_REQ
+ * @sessname: Session name chosen by client
+ */
+struct rtrs_msg_info_req {
+ __le16 type;
+ u8 sessname[NAME_MAX];
+ u8 reserved[15];
+};
+
+/**
+ * struct rtrs_msg_info_rsp
+ * @type: @RTRS_MSG_INFO_RSP
+ * @sg_cnt: Number of @desc entries
+ * @desc: RDMA buffers where the client can write to server
+ */
+struct rtrs_msg_info_rsp {
+ __le16 type;
+ __le16 sg_cnt;
+ u8 reserved[4];
+ struct rtrs_sg_desc desc[];
+};
+
+/**
+ * struct rtrs_msg_rkey_rsp
+ * @type: @RTRS_MSG_RKEY_RSP
+ * @buf_id: RDMA buf_id of the new rkey
+ * @rkey: new remote key for RDMA buffers id from server
+ */
+struct rtrs_msg_rkey_rsp {
+ __le16 type;
+ __le16 buf_id;
+ __le32 rkey;
+};
+
+/**
+ * struct rtrs_msg_rdma_read - RDMA data transfer request from client
+ * @type: always @RTRS_MSG_READ
+ * @usr_len: length of user payload
+ * @sg_cnt: number of @desc entries
+ * @desc: RDMA buffers where the server can write the result to
+ */
+struct rtrs_msg_rdma_read {
+ __le16 type;
+ __le16 usr_len;
+ __le16 flags;
+ __le16 sg_cnt;
+ struct rtrs_sg_desc desc[];
+};
+
+/**
+ * struct_msg_rdma_write - Message transferred to server with RDMA-Write
+ * @type: always @RTRS_MSG_WRITE
+ * @usr_len: length of user payload
+ */
+struct rtrs_msg_rdma_write {
+ __le16 type;
+ __le16 usr_len;
+};
+
+/**
+ * struct_msg_rdma_hdr - header for read or write request
+ * @type: @RTRS_MSG_WRITE | @RTRS_MSG_READ
+ */
+struct rtrs_msg_rdma_hdr {
+ __le16 type;
+};
+
+/* rtrs.c */
+
+struct rtrs_iu *rtrs_iu_alloc(u32 queue_size, size_t size, gfp_t t,
+ struct ib_device *dev, enum dma_data_direction,
+ void (*done)(struct ib_cq *cq, struct ib_wc *wc));
+void rtrs_iu_free(struct rtrs_iu *iu, enum dma_data_direction dir,
+ struct ib_device *dev, u32 queue_size);
+int rtrs_iu_post_recv(struct rtrs_con *con, struct rtrs_iu *iu);
+int rtrs_iu_post_send(struct rtrs_con *con, struct rtrs_iu *iu, size_t size,
+ struct ib_send_wr *head);
+int rtrs_iu_post_rdma_write_imm(struct rtrs_con *con, struct rtrs_iu *iu,
+ struct ib_sge *sge, unsigned int num_sge,
+ u32 rkey, u64 rdma_addr, u32 imm_data,
+ enum ib_send_flags flags,
+ struct ib_send_wr *head);
+
+int rtrs_post_recv_empty(struct rtrs_con *con, struct ib_cqe *cqe);
+int rtrs_post_recv_empty_x2(struct rtrs_con *con, struct ib_cqe *cqe);
+int rtrs_post_rdma_write_imm_empty(struct rtrs_con *con, struct ib_cqe *cqe,
+ u32 imm_data, enum ib_send_flags flags,
+ struct ib_send_wr *head);
+
+int rtrs_cq_qp_create(struct rtrs_sess *rtrs_sess, struct rtrs_con *con,
+ u32 max_send_sge, int cq_vector, u16 cq_size,
+ u16 wr_queue_size, enum ib_poll_context poll_ctx);
+void rtrs_cq_qp_destroy(struct rtrs_con *con);
+
+void rtrs_init_hb(struct rtrs_sess *sess, struct ib_cqe *cqe,
+ unsigned int interval_ms, unsigned int missed_max,
+ rtrs_hb_handler_t *err_handler,
+ struct workqueue_struct *wq);
+void rtrs_start_hb(struct rtrs_sess *sess);
+void rtrs_stop_hb(struct rtrs_sess *sess);
+void rtrs_send_hb_ack(struct rtrs_sess *sess);
+
+void rtrs_ib_dev_pool_init(enum ib_pd_flags pd_flags,
+ struct rtrs_ib_dev_pool *pool);
+void rtrs_ib_dev_pool_deinit(struct rtrs_ib_dev_pool *pool);
+
+struct rtrs_ib_dev *rtrs_ib_dev_find_or_add(struct ib_device *ib_dev,
+ struct rtrs_ib_dev_pool *pool);
+int rtrs_ib_dev_put(struct rtrs_ib_dev *dev);
+
+static inline u32 rtrs_to_imm(u32 type, u32 payload)
+{
+ BUILD_BUG_ON(MAX_IMM_PAYL_BITS + MAX_IMM_TYPE_BITS != 32);
+ BUILD_BUG_ON(RTRS_LAST_IMM > (1<<MAX_IMM_TYPE_BITS));
+ return ((type & MAX_IMM_TYPE_MASK) << MAX_IMM_PAYL_BITS) |
+ (payload & MAX_IMM_PAYL_MASK);
+}
+
+static inline void rtrs_from_imm(u32 imm, u32 *type, u32 *payload)
+{
+ *payload = (imm & MAX_IMM_PAYL_MASK);
+ *type = (imm >> MAX_IMM_PAYL_BITS);
+}
+
+static inline u32 rtrs_to_io_req_imm(u32 addr)
+{
+ return rtrs_to_imm(RTRS_IO_REQ_IMM, addr);
+}
+
+static inline u32 rtrs_to_io_rsp_imm(u32 msg_id, int errno, bool w_inval)
+{
+ enum rtrs_imm_type type;
+ u32 payload;
+
+ /* 9 bits for errno, 19 bits for msg_id */
+ payload = (abs(errno) & 0x1ff) << 19 | (msg_id & 0x7ffff);
+ type = (w_inval ? RTRS_IO_RSP_W_INV_IMM : RTRS_IO_RSP_IMM);
+
+ return rtrs_to_imm(type, payload);
+}
+
+static inline void rtrs_from_io_rsp_imm(u32 payload, u32 *msg_id, int *errno)
+{
+ /* 9 bits for errno, 19 bits for msg_id */
+ *msg_id = (payload & 0x7ffff);
+ *errno = -(int)((payload >> 19) & 0x1ff);
+}
+
+#define STAT_STORE_FUNC(type, set_value, reset) \
+static ssize_t set_value##_store(struct kobject *kobj, \
+ struct kobj_attribute *attr, \
+ const char *buf, size_t count) \
+{ \
+ int ret = -EINVAL; \
+ type *sess = container_of(kobj, type, kobj_stats); \
+ \
+ if (sysfs_streq(buf, "1")) \
+ ret = reset(&sess->stats, true); \
+ else if (sysfs_streq(buf, "0")) \
+ ret = reset(&sess->stats, false); \
+ if (ret) \
+ return ret; \
+ \
+ return count; \
+}
+
+#define STAT_SHOW_FUNC(type, get_value, print) \
+static ssize_t get_value##_show(struct kobject *kobj, \
+ struct kobj_attribute *attr, \
+ char *page) \
+{ \
+ type *sess = container_of(kobj, type, kobj_stats); \
+ \
+ return print(&sess->stats, page, PAGE_SIZE); \
+}
+
+#define STAT_ATTR(type, stat, print, reset) \
+STAT_STORE_FUNC(type, stat, reset) \
+STAT_SHOW_FUNC(type, stat, print) \
+static struct kobj_attribute stat##_attr = \
+ __ATTR(stat, 0644, \
+ stat##_show, \
+ stat##_store)
+
+#endif /* RTRS_PRI_H */
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 04/25] rtrs: core: lib functions shared between client and server modules
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (2 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 03/25] rtrs: private headers with rtrs protocol structs and helpers Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 05/25] rtrs: client: private header with client structs and functions Jack Wang
` (22 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is a set of library functions existing as a rtrs-core module,
used by client and server modules.
Mainly these functions wrap IB and RDMA calls and provide a bit higher
abstraction for implementing of RTRS protocol on client or server
sides.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs.c | 646 +++++++++++++++++++++++++++++
1 file changed, 646 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs.c
diff --git a/drivers/infiniband/ulp/rtrs/rtrs.c b/drivers/infiniband/ulp/rtrs/rtrs.c
new file mode 100644
index 000000000000..34f4f641a935
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs.c
@@ -0,0 +1,646 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include <linux/module.h>
+#include <linux/inet.h>
+
+#include "rtrs-pri.h"
+#include "rtrs-log.h"
+
+MODULE_DESCRIPTION("RTRS Core");
+MODULE_LICENSE("GPL");
+
+struct rtrs_iu *rtrs_iu_alloc(u32 queue_size, size_t size, gfp_t gfp_mask,
+ struct ib_device *dma_dev,
+ enum dma_data_direction dir,
+ void (*done)(struct ib_cq *cq,
+ struct ib_wc *wc))
+{
+ struct rtrs_iu *ius, *iu;
+ int i;
+
+ WARN_ON(!queue_size);
+ ius = kcalloc(queue_size, sizeof(*ius), gfp_mask);
+
+ if (unlikely(!ius))
+ return NULL;
+ for (i = 0; i < queue_size; i++) {
+ iu = &ius[i];
+ iu->buf = kzalloc(size, gfp_mask);
+ if (unlikely(!iu->buf))
+ goto err;
+
+ iu->dma_addr = ib_dma_map_single(dma_dev, iu->buf, size, dir);
+ if (unlikely(ib_dma_mapping_error(dma_dev, iu->dma_addr)))
+ goto err;
+
+ iu->cqe.done = done;
+ iu->size = size;
+ iu->direction = dir;
+ }
+
+ return ius;
+
+err:
+ rtrs_iu_free(ius, dir, dma_dev, i);
+
+ return NULL;
+}
+EXPORT_SYMBOL_GPL(rtrs_iu_alloc);
+
+void rtrs_iu_free(struct rtrs_iu *ius, enum dma_data_direction dir,
+ struct ib_device *ibdev, u32 queue_size)
+{
+ struct rtrs_iu *iu;
+ int i;
+
+ if (!ius)
+ return;
+
+ for (i = 0; i < queue_size; i++) {
+ iu = &ius[i];
+ ib_dma_unmap_single(ibdev, iu->dma_addr, iu->size, dir);
+ kfree(iu->buf);
+ }
+ kfree(ius);
+}
+EXPORT_SYMBOL_GPL(rtrs_iu_free);
+
+int rtrs_iu_post_recv(struct rtrs_con *con, struct rtrs_iu *iu)
+{
+ struct rtrs_sess *sess = con->sess;
+ struct ib_recv_wr wr;
+ const struct ib_recv_wr *bad_wr;
+ struct ib_sge list;
+
+ list.addr = iu->dma_addr;
+ list.length = iu->size;
+ list.lkey = sess->dev->ib_pd->local_dma_lkey;
+
+ if (WARN_ON(list.length == 0)) {
+ rtrs_wrn(con->sess,
+ "Posting receive work request failed, sg list is empty\n");
+ return -EINVAL;
+ }
+
+ wr.next = NULL;
+ wr.wr_cqe = &iu->cqe;
+ wr.sg_list = &list;
+ wr.num_sge = 1;
+
+ return ib_post_recv(con->qp, &wr, &bad_wr);
+}
+EXPORT_SYMBOL_GPL(rtrs_iu_post_recv);
+
+int rtrs_post_recv_empty(struct rtrs_con *con, struct ib_cqe *cqe)
+{
+ struct ib_recv_wr wr;
+ const struct ib_recv_wr *bad_wr;
+
+ wr.next = NULL;
+ wr.wr_cqe = cqe;
+ wr.sg_list = NULL;
+ wr.num_sge = 0;
+
+ return ib_post_recv(con->qp, &wr, &bad_wr);
+}
+EXPORT_SYMBOL_GPL(rtrs_post_recv_empty);
+
+int rtrs_post_recv_empty_x2(struct rtrs_con *con, struct ib_cqe *cqe)
+{
+ struct ib_recv_wr wr_arr[2], *wr;
+ const struct ib_recv_wr *bad_wr;
+ int i;
+
+ memset(wr_arr, 0, sizeof(wr_arr));
+ for (i = 0; i < ARRAY_SIZE(wr_arr); i++) {
+ wr = &wr_arr[i];
+ wr->wr_cqe = cqe;
+ if (i)
+ /* Chain backwards */
+ wr->next = &wr_arr[i - 1];
+ }
+
+ return ib_post_recv(con->qp, wr, &bad_wr);
+}
+EXPORT_SYMBOL_GPL(rtrs_post_recv_empty_x2);
+
+int rtrs_iu_post_send(struct rtrs_con *con, struct rtrs_iu *iu, size_t size,
+ struct ib_send_wr *head)
+{
+ struct rtrs_sess *sess = con->sess;
+ struct ib_send_wr wr;
+ const struct ib_send_wr *bad_wr;
+ struct ib_sge list;
+
+ if ((WARN_ON(size == 0)))
+ return -EINVAL;
+
+ list.addr = iu->dma_addr;
+ list.length = size;
+ list.lkey = sess->dev->ib_pd->local_dma_lkey;
+
+ memset(&wr, 0, sizeof(wr));
+ wr.next = NULL;
+ wr.wr_cqe = &iu->cqe;
+ wr.sg_list = &list;
+ wr.num_sge = 1;
+ wr.opcode = IB_WR_SEND;
+ wr.send_flags = IB_SEND_SIGNALED;
+
+ if (head) {
+ struct ib_send_wr *tail = head;
+
+ while (tail->next)
+ tail = tail->next;
+ tail->next = ≀
+ } else {
+ head = ≀
+ }
+
+ return ib_post_send(con->qp, head, &bad_wr);
+}
+EXPORT_SYMBOL_GPL(rtrs_iu_post_send);
+
+int rtrs_iu_post_rdma_write_imm(struct rtrs_con *con, struct rtrs_iu *iu,
+ struct ib_sge *sge, unsigned int num_sge,
+ u32 rkey, u64 rdma_addr, u32 imm_data,
+ enum ib_send_flags flags,
+ struct ib_send_wr *head)
+{
+ const struct ib_send_wr *bad_wr;
+ struct ib_rdma_wr wr;
+ int i;
+
+ wr.wr.next = NULL;
+ wr.wr.wr_cqe = &iu->cqe;
+ wr.wr.sg_list = sge;
+ wr.wr.num_sge = num_sge;
+ wr.rkey = rkey;
+ wr.remote_addr = rdma_addr;
+ wr.wr.opcode = IB_WR_RDMA_WRITE_WITH_IMM;
+ wr.wr.ex.imm_data = cpu_to_be32(imm_data);
+ wr.wr.send_flags = flags;
+
+ /*
+ * If one of the sges has 0 size, the operation will fail with an
+ * length error
+ */
+ for (i = 0; i < num_sge; i++)
+ if (WARN_ON(sge[i].length == 0))
+ return -EINVAL;
+
+ if (head) {
+ struct ib_send_wr *tail = head;
+
+ while (tail->next)
+ tail = tail->next;
+ tail->next = &wr.wr;
+ } else {
+ head = &wr.wr;
+ }
+
+ return ib_post_send(con->qp, head, &bad_wr);
+}
+EXPORT_SYMBOL_GPL(rtrs_iu_post_rdma_write_imm);
+
+int rtrs_post_rdma_write_imm_empty(struct rtrs_con *con, struct ib_cqe *cqe,
+ u32 imm_data, enum ib_send_flags flags,
+ struct ib_send_wr *head)
+{
+ struct ib_send_wr wr;
+ const struct ib_send_wr *bad_wr;
+
+ memset(&wr, 0, sizeof(wr));
+ wr.wr_cqe = cqe;
+ wr.send_flags = flags;
+ wr.opcode = IB_WR_RDMA_WRITE_WITH_IMM;
+ wr.ex.imm_data = cpu_to_be32(imm_data);
+
+ if (head) {
+ struct ib_send_wr *tail = head;
+
+ while (tail->next)
+ tail = tail->next;
+ tail->next = ≀
+ } else {
+ head = ≀
+ }
+
+ return ib_post_send(con->qp, head, &bad_wr);
+}
+EXPORT_SYMBOL_GPL(rtrs_post_rdma_write_imm_empty);
+
+static void qp_event_handler(struct ib_event *ev, void *ctx)
+{
+ struct rtrs_con *con = ctx;
+
+ switch (ev->event) {
+ case IB_EVENT_COMM_EST:
+ rtrs_info(con->sess, "QP event %s (%d) received\n",
+ ib_event_msg(ev->event), ev->event);
+ rdma_notify(con->cm_id, IB_EVENT_COMM_EST);
+ break;
+ default:
+ rtrs_info(con->sess, "Unhandled QP event %s (%d) received\n",
+ ib_event_msg(ev->event), ev->event);
+ break;
+ }
+}
+
+static int create_cq(struct rtrs_con *con, int cq_vector, u16 cq_size,
+ enum ib_poll_context poll_ctx)
+{
+ struct rdma_cm_id *cm_id = con->cm_id;
+ struct ib_cq *cq;
+
+ cq = ib_alloc_cq(cm_id->device, con, cq_size,
+ cq_vector, poll_ctx);
+ if (IS_ERR(cq)) {
+ rtrs_err(con->sess, "Creating completion queue failed, errno: %ld\n",
+ PTR_ERR(cq));
+ return PTR_ERR(cq);
+ }
+ con->cq = cq;
+
+ return 0;
+}
+
+static int create_qp(struct rtrs_con *con, struct ib_pd *pd,
+ u16 wr_queue_size, u32 max_sge)
+{
+ struct ib_qp_init_attr init_attr = {NULL};
+ struct rdma_cm_id *cm_id = con->cm_id;
+ int ret;
+
+ init_attr.cap.max_send_wr = wr_queue_size;
+ init_attr.cap.max_recv_wr = wr_queue_size;
+ init_attr.cap.max_recv_sge = 1;
+ init_attr.event_handler = qp_event_handler;
+ init_attr.qp_context = con;
+#undef max_send_sge
+ init_attr.cap.max_send_sge = max_sge;
+
+ init_attr.qp_type = IB_QPT_RC;
+ init_attr.send_cq = con->cq;
+ init_attr.recv_cq = con->cq;
+ init_attr.sq_sig_type = IB_SIGNAL_REQ_WR;
+
+ ret = rdma_create_qp(cm_id, pd, &init_attr);
+ if (unlikely(ret)) {
+ rtrs_err(con->sess, "Creating QP failed, err: %d\n", ret);
+ return ret;
+ }
+ con->qp = cm_id->qp;
+
+ return ret;
+}
+
+int rtrs_cq_qp_create(struct rtrs_sess *sess, struct rtrs_con *con,
+ u32 max_send_sge, int cq_vector, u16 cq_size,
+ u16 wr_queue_size, enum ib_poll_context poll_ctx)
+{
+ int err;
+
+ err = create_cq(con, cq_vector, cq_size, poll_ctx);
+ if (unlikely(err))
+ return err;
+
+ err = create_qp(con, sess->dev->ib_pd, wr_queue_size, max_send_sge);
+ if (unlikely(err)) {
+ ib_free_cq(con->cq);
+ con->cq = NULL;
+ return err;
+ }
+ con->sess = sess;
+
+ return 0;
+}
+EXPORT_SYMBOL_GPL(rtrs_cq_qp_create);
+
+void rtrs_cq_qp_destroy(struct rtrs_con *con)
+{
+ if (con->qp) {
+ rdma_destroy_qp(con->cm_id);
+ con->qp = NULL;
+ }
+ if (con->cq) {
+ ib_free_cq(con->cq);
+ con->cq = NULL;
+ }
+}
+EXPORT_SYMBOL_GPL(rtrs_cq_qp_destroy);
+
+static void schedule_hb(struct rtrs_sess *sess)
+{
+ queue_delayed_work(sess->hb_wq, &sess->hb_dwork,
+ msecs_to_jiffies(sess->hb_interval_ms));
+}
+
+void rtrs_send_hb_ack(struct rtrs_sess *sess)
+{
+ struct rtrs_con *usr_con = sess->con[0];
+ u32 imm;
+ int err;
+
+ imm = rtrs_to_imm(RTRS_HB_ACK_IMM, 0);
+ err = rtrs_post_rdma_write_imm_empty(usr_con, sess->hb_cqe, imm,
+ IB_SEND_SIGNALED, NULL);
+ if (unlikely(err)) {
+ sess->hb_err_handler(usr_con);
+ return;
+ }
+}
+EXPORT_SYMBOL_GPL(rtrs_send_hb_ack);
+
+static void hb_work(struct work_struct *work)
+{
+ struct rtrs_con *usr_con;
+ struct rtrs_sess *sess;
+ u32 imm;
+ int err;
+
+ sess = container_of(to_delayed_work(work), typeof(*sess), hb_dwork);
+ usr_con = sess->con[0];
+
+ if (sess->hb_missed_cnt > sess->hb_missed_max) {
+ sess->hb_err_handler(usr_con);
+ return;
+ }
+ if (sess->hb_missed_cnt++) {
+ /* Reschedule work without sending hb */
+ schedule_hb(sess);
+ return;
+ }
+ imm = rtrs_to_imm(RTRS_HB_MSG_IMM, 0);
+ err = rtrs_post_rdma_write_imm_empty(usr_con, sess->hb_cqe, imm,
+ IB_SEND_SIGNALED, NULL);
+ if (unlikely(err)) {
+ sess->hb_err_handler(usr_con);
+ return;
+ }
+
+ schedule_hb(sess);
+}
+
+void rtrs_init_hb(struct rtrs_sess *sess, struct ib_cqe *cqe,
+ unsigned int interval_ms, unsigned int missed_max,
+ rtrs_hb_handler_t *err_handler,
+ struct workqueue_struct *wq)
+{
+ sess->hb_cqe = cqe;
+ sess->hb_interval_ms = interval_ms;
+ sess->hb_err_handler = err_handler;
+ sess->hb_wq = wq;
+ sess->hb_missed_max = missed_max;
+ sess->hb_missed_cnt = 0;
+ INIT_DELAYED_WORK(&sess->hb_dwork, hb_work);
+}
+EXPORT_SYMBOL_GPL(rtrs_init_hb);
+
+void rtrs_start_hb(struct rtrs_sess *sess)
+{
+ schedule_hb(sess);
+}
+EXPORT_SYMBOL_GPL(rtrs_start_hb);
+
+void rtrs_stop_hb(struct rtrs_sess *sess)
+{
+ cancel_delayed_work_sync(&sess->hb_dwork);
+ sess->hb_missed_cnt = 0;
+ sess->hb_missed_max = 0;
+}
+EXPORT_SYMBOL_GPL(rtrs_stop_hb);
+
+static int rtrs_str_gid_to_sockaddr(const char *addr, size_t len,
+ short port, struct sockaddr_storage *dst)
+{
+ struct sockaddr_ib *dst_ib = (struct sockaddr_ib *)dst;
+ int ret;
+
+ /*
+ * We can use some of the I6 functions since GID is a valid
+ * IPv6 address format
+ */
+ ret = in6_pton(addr, len, dst_ib->sib_addr.sib_raw, '\0', NULL);
+ if (ret == 0)
+ return -EINVAL;
+
+ dst_ib->sib_family = AF_IB;
+ /*
+ * Use the same TCP server port number as the IB service ID
+ * on the IB port space range
+ */
+ dst_ib->sib_sid = cpu_to_be64(RDMA_IB_IP_PS_IB | port);
+ dst_ib->sib_sid_mask = cpu_to_be64(0xffffffffffffffffULL);
+ dst_ib->sib_pkey = cpu_to_be16(0xffff);
+
+ return 0;
+}
+
+/**
+ * rtrs_str_to_sockaddr() - Convert rtrs address string to sockaddr
+ * @addr: String representation of an addr (IPv4, IPv6 or IB GID):
+ * - "ip:192.168.1.1"
+ * - "ip:fe80::200:5aee:feaa:20a2"
+ * - "gid:fe80::200:5aee:feaa:20a2"
+ * @len: String address length
+ * @port: Destination port
+ * @dst: Destination sockaddr structure
+ *
+ * Returns 0 if conversion successful. Non-zero on error.
+ */
+static int rtrs_str_to_sockaddr(const char *addr, size_t len,
+ short port, struct sockaddr_storage *dst)
+{
+ if (strncmp(addr, "gid:", 4) == 0) {
+ return rtrs_str_gid_to_sockaddr(addr + 4, len - 4, port, dst);
+ } else if (strncmp(addr, "ip:", 3) == 0) {
+ char port_str[8];
+ char *cpy;
+ int err;
+
+ snprintf(port_str, sizeof(port_str), "%u", port);
+ cpy = kstrndup(addr + 3, len - 3, GFP_KERNEL);
+ err = cpy ? inet_pton_with_scope(&init_net, AF_UNSPEC,
+ cpy, port_str, dst) : -ENOMEM;
+ kfree(cpy);
+
+ return err;
+ }
+ return -EPROTONOSUPPORT;
+}
+
+int sockaddr_to_str(const struct sockaddr *addr, char *buf, size_t len)
+{
+ int cnt;
+
+ switch (addr->sa_family) {
+ case AF_IB:
+ cnt = scnprintf(buf, len, "gid:%pI6",
+ &((struct sockaddr_ib *)addr)->sib_addr.sib_raw);
+ return cnt;
+ case AF_INET:
+ cnt = scnprintf(buf, len, "ip:%pI4",
+ &((struct sockaddr_in *)addr)->sin_addr);
+ return cnt;
+ case AF_INET6:
+ cnt = scnprintf(buf, len, "ip:%pI6c",
+ &((struct sockaddr_in6 *)addr)->sin6_addr);
+ return cnt;
+ }
+ cnt = scnprintf(buf, len, "<invalid address family>");
+ pr_err("Invalid address family\n");
+ return cnt;
+}
+EXPORT_SYMBOL(sockaddr_to_str);
+
+int rtrs_addr_to_sockaddr(const char *str, size_t len, short port,
+ struct rtrs_addr *addr)
+{
+ const char *d;
+ int ret;
+
+ d = strchr(str, ',');
+ if (!d)
+ d = strchr(str, '@');
+ if (d) {
+ if (rtrs_str_to_sockaddr(str, d - str, 0, addr->src))
+ return -EINVAL;
+ d += 1;
+ len -= d - str;
+ str = d;
+
+ } else {
+ addr->src = NULL;
+ }
+ ret = rtrs_str_to_sockaddr(str, len, port, addr->dst);
+
+ return ret;
+}
+EXPORT_SYMBOL(rtrs_addr_to_sockaddr);
+
+void rtrs_ib_dev_pool_init(enum ib_pd_flags pd_flags,
+ struct rtrs_ib_dev_pool *pool)
+{
+ WARN_ON(pool->ops && (!pool->ops->alloc ^ !pool->ops->free));
+ INIT_LIST_HEAD(&pool->list);
+ mutex_init(&pool->mutex);
+ pool->pd_flags = pd_flags;
+}
+EXPORT_SYMBOL(rtrs_ib_dev_pool_init);
+
+void rtrs_ib_dev_pool_deinit(struct rtrs_ib_dev_pool *pool)
+{
+ WARN_ON(!list_empty(&pool->list));
+}
+EXPORT_SYMBOL(rtrs_ib_dev_pool_deinit);
+
+static void dev_free(struct kref *ref)
+{
+ struct rtrs_ib_dev_pool *pool;
+ struct rtrs_ib_dev *dev;
+
+ dev = container_of(ref, typeof(*dev), ref);
+ pool = dev->pool;
+
+ mutex_lock(&pool->mutex);
+ list_del(&dev->entry);
+ mutex_unlock(&pool->mutex);
+
+ if (pool->ops && pool->ops->deinit)
+ pool->ops->deinit(dev);
+
+ ib_dealloc_pd(dev->ib_pd);
+
+ if (pool->ops && pool->ops->free)
+ pool->ops->free(dev);
+ else
+ kfree(dev);
+}
+
+int rtrs_ib_dev_put(struct rtrs_ib_dev *dev)
+{
+ return kref_put(&dev->ref, dev_free);
+}
+EXPORT_SYMBOL(rtrs_ib_dev_put);
+
+static int rtrs_ib_dev_get(struct rtrs_ib_dev *dev)
+{
+ return kref_get_unless_zero(&dev->ref);
+}
+
+struct rtrs_ib_dev *
+rtrs_ib_dev_find_or_add(struct ib_device *ib_dev,
+ struct rtrs_ib_dev_pool *pool)
+{
+ struct rtrs_ib_dev *dev;
+
+ mutex_lock(&pool->mutex);
+ list_for_each_entry(dev, &pool->list, entry) {
+ if (dev->ib_dev->node_guid == ib_dev->node_guid &&
+ rtrs_ib_dev_get(dev))
+ goto out_unlock;
+ }
+ if (pool->ops && pool->ops->alloc)
+ dev = pool->ops->alloc();
+ else
+ dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+ if (IS_ERR_OR_NULL(dev))
+ goto out_err;
+
+ kref_init(&dev->ref);
+ dev->pool = pool;
+ dev->ib_dev = ib_dev;
+ dev->ib_pd = ib_alloc_pd(ib_dev, pool->pd_flags);
+ if (IS_ERR(dev->ib_pd))
+ goto out_free_dev;
+
+ if (pool->ops && pool->ops->init && pool->ops->init(dev))
+ goto out_free_pd;
+
+ list_add(&dev->entry, &pool->list);
+out_unlock:
+ mutex_unlock(&pool->mutex);
+ return dev;
+
+out_free_pd:
+ ib_dealloc_pd(dev->ib_pd);
+out_free_dev:
+ if (pool->ops && pool->ops->free)
+ pool->ops->free(dev);
+ else
+ kfree(dev);
+out_err:
+ mutex_unlock(&pool->mutex);
+ return NULL;
+}
+EXPORT_SYMBOL(rtrs_ib_dev_find_or_add);
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 05/25] rtrs: client: private header with client structs and functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (3 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 04/25] rtrs: core: lib functions shared between client and server modules Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 06/25] rtrs: client: main functionality Jack Wang
` (21 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This header describes main structs and functions used by rtrs-client
module, mainly for managing rtrs sessions, creating/destroying sysfs
entries, accounting statistics on client side.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-clt.h | 314 +++++++++++++++++++++++++
1 file changed, 314 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-clt.h
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-clt.h b/drivers/infiniband/ulp/rtrs/rtrs-clt.h
new file mode 100644
index 000000000000..df774ab3f9b5
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-clt.h
@@ -0,0 +1,314 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Swapnil Ingle <swapnil.ingle@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+
+#ifndef RTRS_CLT_H
+#define RTRS_CLT_H
+
+#include <linux/device.h>
+#include "rtrs-pri.h"
+
+/**
+ * enum rtrs_clt_state - Client states.
+ */
+enum rtrs_clt_state {
+ RTRS_CLT_CONNECTING,
+ RTRS_CLT_CONNECTING_ERR,
+ RTRS_CLT_RECONNECTING,
+ RTRS_CLT_CONNECTED,
+ RTRS_CLT_CLOSING,
+ RTRS_CLT_CLOSED,
+ RTRS_CLT_DEAD,
+};
+
+static inline const char *rtrs_clt_state_str(enum rtrs_clt_state state)
+{
+ switch (state) {
+ case RTRS_CLT_CONNECTING:
+ return "RTRS_CLT_CONNECTING";
+ case RTRS_CLT_CONNECTING_ERR:
+ return "RTRS_CLT_CONNECTING_ERR";
+ case RTRS_CLT_RECONNECTING:
+ return "RTRS_CLT_RECONNECTING";
+ case RTRS_CLT_CONNECTED:
+ return "RTRS_CLT_CONNECTED";
+ case RTRS_CLT_CLOSING:
+ return "RTRS_CLT_CLOSING";
+ case RTRS_CLT_CLOSED:
+ return "RTRS_CLT_CLOSED";
+ case RTRS_CLT_DEAD:
+ return "RTRS_CLT_DEAD";
+ default:
+ return "UNKNOWN";
+ }
+}
+
+enum rtrs_mp_policy {
+ MP_POLICY_RR,
+ MP_POLICY_MIN_INFLIGHT,
+};
+
+struct rtrs_clt_stats_reconnects {
+ int successful_cnt;
+ int fail_cnt;
+};
+
+struct rtrs_clt_stats_wc_comp {
+ u32 cnt;
+ u64 total_cnt;
+};
+
+struct rtrs_clt_stats_cpu_migr {
+ atomic_t from;
+ int to;
+};
+
+struct rtrs_clt_stats_rdma {
+ struct {
+ u64 cnt;
+ u64 size_total;
+ } dir[2];
+
+ u64 failover_cnt;
+};
+
+struct rtrs_clt_stats_rdma_lat {
+ u64 read;
+ u64 write;
+};
+
+#define MIN_LOG_SG 2
+#define MAX_LOG_SG 5
+#define MAX_LIN_SG BIT(MIN_LOG_SG)
+#define SG_DISTR_SZ (MAX_LOG_SG - MIN_LOG_SG + MAX_LIN_SG + 2)
+
+#define MAX_LOG_LAT 16
+#define MIN_LOG_LAT 0
+#define LOG_LAT_SZ (MAX_LOG_LAT - MIN_LOG_LAT + 2)
+
+struct rtrs_clt_stats_pcpu {
+ struct rtrs_clt_stats_cpu_migr cpu_migr;
+ struct rtrs_clt_stats_rdma rdma;
+ u64 sg_list_total;
+ u64 sg_list_distr[SG_DISTR_SZ];
+ struct rtrs_clt_stats_rdma_lat rdma_lat_distr[LOG_LAT_SZ];
+ struct rtrs_clt_stats_rdma_lat rdma_lat_max;
+ struct rtrs_clt_stats_wc_comp wc_comp;
+};
+
+struct rtrs_clt_stats {
+ bool enable_rdma_lat;
+ struct rtrs_clt_stats_pcpu __percpu *pcpu_stats;
+ struct rtrs_clt_stats_reconnects reconnects;
+ atomic_t inflight;
+};
+
+struct rtrs_clt_con {
+ struct rtrs_con c;
+ struct rtrs_iu *rsp_ius;
+ u32 queue_size;
+ unsigned int cpu;
+ atomic_t io_cnt;
+ int cm_err;
+};
+
+/**
+ * rtrs_permit - permits the memory allocation for future RDMA operation
+ */
+struct rtrs_permit {
+ enum rtrs_clt_con_type con_type;
+ unsigned int cpu_id;
+ unsigned int mem_id;
+ unsigned int mem_off;
+};
+
+/**
+ * rtrs_clt_io_req - describes one inflight IO request
+ */
+struct rtrs_clt_io_req {
+ struct list_head list;
+ struct rtrs_iu *iu;
+ struct scatterlist *sglist; /* list holding user data */
+ unsigned int sg_cnt;
+ unsigned int sg_size;
+ unsigned int data_len;
+ unsigned int usr_len;
+ void *priv;
+ bool in_use;
+ struct rtrs_clt_con *con;
+ struct rtrs_sg_desc *desc;
+ struct ib_sge *sge;
+ struct rtrs_permit *permit;
+ enum dma_data_direction dir;
+ rtrs_conf_fn *conf;
+ unsigned long start_jiffies;
+
+ struct ib_mr *mr;
+ struct ib_cqe inv_cqe;
+ struct completion inv_comp;
+ int inv_errno;
+ bool need_inv_comp;
+ bool need_inv;
+};
+
+struct rtrs_rbuf {
+ u64 addr;
+ u32 rkey;
+};
+
+struct rtrs_clt_sess {
+ struct rtrs_sess s;
+ struct rtrs_clt *clt;
+ wait_queue_head_t state_wq;
+ enum rtrs_clt_state state;
+ atomic_t connected_cnt;
+ struct mutex init_mutex;
+ struct rtrs_clt_io_req *reqs;
+ struct delayed_work reconnect_dwork;
+ struct work_struct close_work;
+ unsigned int reconnect_attempts;
+ bool established;
+ struct rtrs_rbuf *rbufs;
+ size_t max_io_size;
+ u32 max_hdr_size;
+ u32 chunk_size;
+ size_t queue_depth;
+ u32 max_pages_per_mr;
+ int max_send_sge;
+ u32 flags;
+ struct kobject kobj;
+ struct kobject kobj_stats;
+ struct rtrs_clt_stats stats;
+ /* cache hca_port and hca_name to display in sysfs */
+ u8 hca_port;
+ char hca_name[IB_DEVICE_NAME_MAX];
+ struct list_head __percpu
+ *mp_skip_entry;
+};
+
+struct rtrs_clt {
+ struct list_head /* __rcu */ paths_list;
+ size_t paths_num;
+ struct rtrs_clt_sess
+ __rcu * __percpu *pcpu_path;
+
+ bool opened;
+ uuid_t paths_uuid;
+ int paths_up;
+ struct mutex paths_mutex;
+ struct mutex paths_ev_mutex;
+ char sessname[NAME_MAX];
+ short port;
+ unsigned int max_reconnect_attempts;
+ unsigned int reconnect_delay_sec;
+ unsigned int max_segments;
+ void *permits;
+ unsigned long *permits_map;
+ size_t queue_depth;
+ size_t max_io_size;
+ wait_queue_head_t permits_wait;
+ size_t pdu_sz;
+ void *priv;
+ link_clt_ev_fn *link_ev;
+ struct device dev;
+ struct kobject kobj_paths;
+ enum rtrs_mp_policy mp_policy;
+};
+
+static inline struct rtrs_clt_con *to_clt_con(struct rtrs_con *c)
+{
+ return container_of(c, struct rtrs_clt_con, c);
+}
+
+static inline struct rtrs_clt_sess *to_clt_sess(struct rtrs_sess *s)
+{
+ return container_of(s, struct rtrs_clt_sess, s);
+}
+
+#define PERMIT_SIZE(clt) (sizeof(struct rtrs_permit) + (clt)->pdu_sz)
+#define GET_PERMIT(clt, idx) ((clt)->permits + PERMIT_SIZE(clt) * idx)
+
+int rtrs_clt_reconnect_from_sysfs(struct rtrs_clt_sess *sess);
+int rtrs_clt_disconnect_from_sysfs(struct rtrs_clt_sess *sess);
+int rtrs_clt_create_path_from_sysfs(struct rtrs_clt *clt,
+ struct rtrs_addr *addr);
+int rtrs_clt_remove_path_from_sysfs(struct rtrs_clt_sess *sess,
+ const struct attribute *sysfs_self);
+
+void rtrs_clt_set_max_reconnect_attempts(struct rtrs_clt *clt, int value);
+int rtrs_clt_get_max_reconnect_attempts(const struct rtrs_clt *clt);
+
+/* rtrs-clt-stats.c */
+
+int rtrs_clt_init_stats(struct rtrs_clt_stats *stats);
+void rtrs_clt_free_stats(struct rtrs_clt_stats *stats);
+
+void rtrs_clt_decrease_inflight(struct rtrs_clt_stats *s);
+void rtrs_clt_inc_failover_cnt(struct rtrs_clt_stats *s);
+
+void rtrs_clt_update_rdma_lat(struct rtrs_clt_stats *s, bool read,
+ unsigned long ms);
+void rtrs_clt_update_wc_stats(struct rtrs_clt_con *con);
+void rtrs_clt_update_all_stats(struct rtrs_clt_io_req *req, int dir);
+
+int rtrs_clt_reset_sg_list_distr_stats(struct rtrs_clt_stats *stats,
+ bool enable);
+int rtrs_clt_stats_sg_list_distr_to_str(struct rtrs_clt_stats *stats,
+ char *buf, size_t len);
+int rtrs_clt_reset_rdma_lat_distr_stats(struct rtrs_clt_stats *stats,
+ bool enable);
+ssize_t rtrs_clt_stats_rdma_lat_distr_to_str(struct rtrs_clt_stats *stats,
+ char *page, size_t len);
+int rtrs_clt_reset_cpu_migr_stats(struct rtrs_clt_stats *stats, bool enable);
+int rtrs_clt_stats_migration_cnt_to_str(struct rtrs_clt_stats *stats, char *buf,
+ size_t len);
+int rtrs_clt_reset_reconnects_stat(struct rtrs_clt_stats *stats, bool enable);
+int rtrs_clt_stats_reconnects_to_str(struct rtrs_clt_stats *stats, char *buf,
+ size_t len);
+int rtrs_clt_reset_wc_comp_stats(struct rtrs_clt_stats *stats, bool enable);
+int rtrs_clt_stats_wc_completion_to_str(struct rtrs_clt_stats *stats, char *buf,
+ size_t len);
+int rtrs_clt_reset_rdma_stats(struct rtrs_clt_stats *stats, bool enable);
+ssize_t rtrs_clt_stats_rdma_to_str(struct rtrs_clt_stats *stats,
+ char *page, size_t len);
+int rtrs_clt_reset_all_stats(struct rtrs_clt_stats *stats, bool enable);
+ssize_t rtrs_clt_reset_all_help(struct rtrs_clt_stats *stats,
+ char *page, size_t len);
+
+/* rtrs-clt-sysfs.c */
+
+int rtrs_clt_create_sysfs_root_folders(struct rtrs_clt *clt);
+int rtrs_clt_create_sysfs_root_files(struct rtrs_clt *clt);
+void rtrs_clt_destroy_sysfs_root_folders(struct rtrs_clt *clt);
+void rtrs_clt_destroy_sysfs_root_files(struct rtrs_clt *clt);
+
+int rtrs_clt_create_sess_files(struct rtrs_clt_sess *sess);
+void rtrs_clt_destroy_sess_files(struct rtrs_clt_sess *sess,
+ const struct attribute *sysfs_self);
+
+#endif /* RTRS_CLT_H */
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 06/25] rtrs: client: main functionality
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (4 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 05/25] rtrs: client: private header with client structs and functions Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 07/25] rtrs: client: statistics functions Jack Wang
` (20 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is main functionality of rtrs-client module, which manages
set of RDMA connections for each rtrs session, does multipathing,
load balancing and failover of RDMA requests.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-clt.c | 2952 ++++++++++++++++++++++++
1 file changed, 2952 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-clt.c
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-clt.c b/drivers/infiniband/ulp/rtrs/rtrs-clt.c
new file mode 100644
index 000000000000..aa435655496c
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-clt.c
@@ -0,0 +1,2952 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Swapnil Ingle <swapnil.ingle@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include <linux/module.h>
+#include <linux/rculist.h>
+#include <linux/blkdev.h> /* for BLK_MAX_SEGMENT_SIZE */
+
+#include "rtrs-clt.h"
+#include "rtrs-log.h"
+
+#define RTRS_CONNECT_TIMEOUT_MS 30000
+
+MODULE_DESCRIPTION("RTRS Client");
+MODULE_LICENSE("GPL");
+
+static ushort nr_cons_per_session;
+module_param(nr_cons_per_session, ushort, 0444);
+MODULE_PARM_DESC(nr_cons_per_session,
+ "Number of connections per session. (default: nr_cpu_ids)");
+
+static int retry_cnt = 7;
+module_param_named(retry_cnt, retry_cnt, int, 0644);
+MODULE_PARM_DESC(retry_cnt,
+ "Number of times to send the message if the remote side didn't respond with Ack or Nack (default: 7, min: "
+ __stringify(MIN_RTR_CNT) ", max: "
+ __stringify(MAX_RTR_CNT) ")");
+
+static int __read_mostly noreg_cnt;
+module_param_named(noreg_cnt, noreg_cnt, int, 0444);
+MODULE_PARM_DESC(noreg_cnt,
+ "Max number of SG entries when MR registration does not happen (default: 0)");
+
+static const struct rtrs_ib_dev_pool_ops dev_pool_ops;
+static struct rtrs_ib_dev_pool dev_pool = {
+ .ops = &dev_pool_ops
+};
+
+static struct workqueue_struct *rtrs_wq;
+static struct class *rtrs_dev_class;
+
+static inline bool rtrs_clt_is_connected(const struct rtrs_clt *clt)
+{
+ struct rtrs_clt_sess *sess;
+ bool connected = false;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(sess, &clt->paths_list, s.entry)
+ connected |= (READ_ONCE(sess->state) == RTRS_CLT_CONNECTED);
+ rcu_read_unlock();
+
+ return connected;
+}
+
+static inline struct rtrs_permit *
+__rtrs_get_permit(struct rtrs_clt *clt, enum rtrs_clt_con_type con_type)
+{
+ size_t max_depth = clt->queue_depth;
+ struct rtrs_permit *permit;
+ int cpu, bit;
+
+ cpu = get_cpu();
+ do {
+ bit = find_first_zero_bit(clt->permits_map, max_depth);
+ if (unlikely(bit >= max_depth)) {
+ put_cpu();
+ return NULL;
+ }
+
+ } while (unlikely(test_and_set_bit_lock(bit, clt->permits_map)));
+ put_cpu();
+
+ permit = GET_PERMIT(clt, bit);
+ WARN_ON(permit->mem_id != bit);
+ permit->cpu_id = cpu;
+ permit->con_type = con_type;
+
+ return permit;
+}
+
+static inline void __rtrs_put_permit(struct rtrs_clt *clt,
+ struct rtrs_permit *permit)
+{
+ clear_bit_unlock(permit->mem_id, clt->permits_map);
+}
+
+struct rtrs_permit *rtrs_clt_get_permit(struct rtrs_clt *clt,
+ enum rtrs_clt_con_type con_type,
+ int can_wait)
+{
+ struct rtrs_permit *permit;
+ DEFINE_WAIT(wait);
+
+ permit = __rtrs_get_permit(clt, con_type);
+ if (likely(permit) || !can_wait)
+ return permit;
+
+ do {
+ prepare_to_wait(&clt->permits_wait, &wait,
+ TASK_UNINTERRUPTIBLE);
+ permit = __rtrs_get_permit(clt, con_type);
+ if (likely(permit))
+ break;
+
+ io_schedule();
+ } while (1);
+
+ finish_wait(&clt->permits_wait, &wait);
+
+ return permit;
+}
+EXPORT_SYMBOL(rtrs_clt_get_permit);
+
+void rtrs_clt_put_permit(struct rtrs_clt *clt, struct rtrs_permit *permit)
+{
+ if (WARN_ON(!test_bit(permit->mem_id, clt->permits_map)))
+ return;
+
+ __rtrs_put_permit(clt, permit);
+
+ /*
+ * Putting a permit is a barrier, so we will observe
+ * new entry in the wait list, no worries.
+ */
+ if (waitqueue_active(&clt->permits_wait))
+ wake_up(&clt->permits_wait);
+}
+EXPORT_SYMBOL(rtrs_clt_put_permit);
+
+struct rtrs_permit *rtrs_permit_from_pdu(void *pdu)
+{
+ return pdu - sizeof(struct rtrs_permit);
+}
+EXPORT_SYMBOL(rtrs_permit_from_pdu);
+
+void *rtrs_permit_to_pdu(struct rtrs_permit *permit)
+{
+ return permit + 1;
+}
+EXPORT_SYMBOL(rtrs_permit_to_pdu);
+
+/**
+ * rtrs_permit_to_clt_con() - returns RDMA connection id by the permit
+ *
+ * Note:
+ * IO connection starts from 1.
+ * 0 connection is for user messages.
+ */
+static
+struct rtrs_clt_con *rtrs_permit_to_clt_con(struct rtrs_clt_sess *sess,
+ struct rtrs_permit *permit)
+{
+ int id = 0;
+
+ if (likely(permit->con_type == RTRS_IO_CON))
+ id = (permit->cpu_id % (sess->s.con_num - 1)) + 1;
+
+ return to_clt_con(sess->s.con[id]);
+}
+
+static bool __rtrs_clt_change_state(struct rtrs_clt_sess *sess,
+ enum rtrs_clt_state new_state)
+{
+ enum rtrs_clt_state old_state;
+ bool changed = false;
+
+ lockdep_assert_held(&sess->state_wq.lock);
+
+ old_state = sess->state;
+ switch (new_state) {
+ case RTRS_CLT_CONNECTING:
+ switch (old_state) {
+ case RTRS_CLT_RECONNECTING:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ case RTRS_CLT_RECONNECTING:
+ switch (old_state) {
+ case RTRS_CLT_CONNECTED:
+ case RTRS_CLT_CONNECTING_ERR:
+ case RTRS_CLT_CLOSED:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ case RTRS_CLT_CONNECTED:
+ switch (old_state) {
+ case RTRS_CLT_CONNECTING:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ case RTRS_CLT_CONNECTING_ERR:
+ switch (old_state) {
+ case RTRS_CLT_CONNECTING:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ case RTRS_CLT_CLOSING:
+ switch (old_state) {
+ case RTRS_CLT_CONNECTING:
+ case RTRS_CLT_CONNECTING_ERR:
+ case RTRS_CLT_RECONNECTING:
+ case RTRS_CLT_CONNECTED:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ case RTRS_CLT_CLOSED:
+ switch (old_state) {
+ case RTRS_CLT_CLOSING:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ case RTRS_CLT_DEAD:
+ switch (old_state) {
+ case RTRS_CLT_CLOSED:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ default:
+ break;
+ }
+ if (changed) {
+ sess->state = new_state;
+ wake_up_locked(&sess->state_wq);
+ }
+
+ return changed;
+}
+
+static bool rtrs_clt_change_state_from_to(struct rtrs_clt_sess *sess,
+ enum rtrs_clt_state old_state,
+ enum rtrs_clt_state new_state)
+{
+ bool changed = false;
+
+ spin_lock_irq(&sess->state_wq.lock);
+ if (sess->state == old_state)
+ changed = __rtrs_clt_change_state(sess, new_state);
+ spin_unlock_irq(&sess->state_wq.lock);
+
+ return changed;
+}
+
+static void rtrs_rdma_error_recovery(struct rtrs_clt_con *con)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+
+ if (rtrs_clt_change_state_from_to(sess,
+ RTRS_CLT_CONNECTED,
+ RTRS_CLT_RECONNECTING)) {
+ /*
+ * Normal scenario, reconnect if we were successfully connected
+ */
+ queue_delayed_work(rtrs_wq, &sess->reconnect_dwork, 0);
+ } else {
+ /*
+ * Error can happen just on establishing new connection,
+ * so notify waiter with error state, waiter is responsible
+ * for cleaning the rest and reconnect if needed.
+ */
+ rtrs_clt_change_state_from_to(sess,
+ RTRS_CLT_CONNECTING,
+ RTRS_CLT_CONNECTING_ERR);
+ }
+}
+
+static void rtrs_clt_fast_reg_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_clt_con *con = cq->cq_context;
+
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ rtrs_err(con->c.sess, "Failed IB_WR_REG_MR: %s\n",
+ ib_wc_status_msg(wc->status));
+ rtrs_rdma_error_recovery(con);
+ }
+}
+
+static struct ib_cqe fast_reg_cqe = {
+ .done = rtrs_clt_fast_reg_done
+};
+
+static void complete_rdma_req(struct rtrs_clt_io_req *req, int errno,
+ bool notify, bool can_wait);
+
+static void rtrs_clt_inv_rkey_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_clt_io_req *req =
+ container_of(wc->wr_cqe, typeof(*req), inv_cqe);
+ struct rtrs_clt_con *con = cq->cq_context;
+
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ rtrs_err(con->c.sess, "Failed IB_WR_LOCAL_INV: %s\n",
+ ib_wc_status_msg(wc->status));
+ rtrs_rdma_error_recovery(con);
+ }
+ req->need_inv = false;
+ if (likely(req->need_inv_comp))
+ complete(&req->inv_comp);
+ else
+ /* Complete request from INV callback */
+ complete_rdma_req(req, req->inv_errno, true, false);
+}
+
+static int rtrs_inv_rkey(struct rtrs_clt_io_req *req)
+{
+ struct rtrs_clt_con *con = req->con;
+ const struct ib_send_wr *bad_wr;
+ struct ib_send_wr wr = {
+ .opcode = IB_WR_LOCAL_INV,
+ .wr_cqe = &req->inv_cqe,
+ .next = NULL,
+ .num_sge = 0,
+ .send_flags = IB_SEND_SIGNALED,
+ .ex.invalidate_rkey = req->mr->rkey,
+ };
+ req->inv_cqe.done = rtrs_clt_inv_rkey_done;
+
+ return ib_post_send(con->c.qp, &wr, &bad_wr);
+}
+
+static void complete_rdma_req(struct rtrs_clt_io_req *req, int errno,
+ bool notify, bool can_wait)
+{
+ struct rtrs_clt_con *con = req->con;
+ struct rtrs_clt_sess *sess;
+ int err;
+
+ if (WARN_ON(!req->in_use))
+ return;
+ if (WARN_ON(!req->con))
+ return;
+ sess = to_clt_sess(con->c.sess);
+
+ if (req->sg_cnt) {
+ if (unlikely(req->dir == DMA_FROM_DEVICE && req->need_inv)) {
+ /*
+ * We are here to invalidate RDMA read requests
+ * ourselves. In normal scenario server should
+ * send INV for all requested RDMA reads, but
+ * we are here, thus two things could happen:
+ *
+ * 1. this is failover, when errno != 0
+ * and can_wait == 1,
+ *
+ * 2. something totally bad happened and
+ * server forgot to send INV, so we
+ * should do that ourselves.
+ */
+
+ if (likely(can_wait)) {
+ req->need_inv_comp = true;
+ } else {
+ /* This should be IO path, so always notify */
+ WARN_ON(!notify);
+ /* Save errno for INV callback */
+ req->inv_errno = errno;
+ }
+
+ err = rtrs_inv_rkey(req);
+ if (unlikely(err)) {
+ rtrs_err(con->c.sess, "Send INV WR key=%#x: %d\n",
+ req->mr->rkey, err);
+ } else if (likely(can_wait)) {
+ wait_for_completion(&req->inv_comp);
+ } else {
+ /*
+ * Something went wrong, so request will be
+ * completed from INV callback.
+ */
+ WARN_ON_ONCE(1);
+
+ return;
+ }
+ }
+ ib_dma_unmap_sg(sess->s.dev->ib_dev, req->sglist,
+ req->sg_cnt, req->dir);
+ }
+ if (sess->stats.enable_rdma_lat)
+ rtrs_clt_update_rdma_lat(&sess->stats,
+ req->dir == DMA_FROM_DEVICE,
+ jiffies_to_msecs(jiffies -
+ req->start_jiffies));
+ rtrs_clt_decrease_inflight(&sess->stats);
+
+ req->in_use = false;
+ req->con = NULL;
+
+ if (notify)
+ req->conf(req->priv, errno);
+}
+
+static int rtrs_post_send_rdma(struct rtrs_clt_con *con,
+ struct rtrs_clt_io_req *req,
+ struct rtrs_rbuf *rbuf, u32 off,
+ u32 imm, struct ib_send_wr *wr)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ enum ib_send_flags flags;
+ struct ib_sge sge;
+
+ if (unlikely(!req->sg_size)) {
+ rtrs_wrn(con->c.sess,
+ "Doing RDMA Write failed, no data supplied\n");
+ return -EINVAL;
+ }
+
+ /* user data and user message in the first list element */
+ sge.addr = req->iu->dma_addr;
+ sge.length = req->sg_size;
+ sge.lkey = sess->s.dev->ib_pd->local_dma_lkey;
+
+ /*
+ * From time to time we have to post signalled sends,
+ * or send queue will fill up and only QP reset can help.
+ */
+ flags = atomic_inc_return(&con->io_cnt) % sess->queue_depth ?
+ 0 : IB_SEND_SIGNALED;
+
+ ib_dma_sync_single_for_device(sess->s.dev->ib_dev, req->iu->dma_addr,
+ req->sg_size, DMA_TO_DEVICE);
+
+ return rtrs_iu_post_rdma_write_imm(&con->c, req->iu, &sge, 1,
+ rbuf->rkey, rbuf->addr + off,
+ imm, flags, wr);
+}
+
+static void process_io_rsp(struct rtrs_clt_sess *sess, u32 msg_id,
+ s16 errno, bool w_inval)
+{
+ struct rtrs_clt_io_req *req;
+
+ if (WARN_ON(msg_id >= sess->queue_depth))
+ return;
+
+ req = &sess->reqs[msg_id];
+ /* Drop need_inv if server responsed with invalidation */
+ req->need_inv &= !w_inval;
+ complete_rdma_req(req, errno, true, false);
+}
+
+static void rtrs_clt_recv_done(struct rtrs_clt_con *con, struct ib_wc *wc)
+{
+ struct rtrs_iu *iu;
+ int err;
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+
+ WARN_ON(sess->flags != RTRS_MSG_NEW_RKEY_F);
+ iu = container_of(wc->wr_cqe, struct rtrs_iu,
+ cqe);
+ err = rtrs_iu_post_recv(&con->c, iu);
+ if (unlikely(err)) {
+ rtrs_err(con->c.sess, "post iu failed %d\n", err);
+ rtrs_rdma_error_recovery(con);
+ }
+}
+
+static void rtrs_clt_rkey_rsp_done(struct rtrs_clt_con *con, struct ib_wc *wc)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ struct rtrs_msg_rkey_rsp *msg;
+ u32 imm_type, imm_payload;
+ bool w_inval = false;
+ struct rtrs_iu *iu;
+ u32 buf_id;
+ int err;
+
+ WARN_ON(sess->flags != RTRS_MSG_NEW_RKEY_F);
+
+ iu = container_of(wc->wr_cqe, struct rtrs_iu, cqe);
+
+ if (unlikely(wc->byte_len < sizeof(*msg))) {
+ rtrs_err(con->c.sess, "rkey response is malformed: size %d\n",
+ wc->byte_len);
+ goto out;
+ }
+ ib_dma_sync_single_for_cpu(sess->s.dev->ib_dev, iu->dma_addr,
+ iu->size, DMA_FROM_DEVICE);
+ msg = iu->buf;
+ if (unlikely(le16_to_cpu(msg->type) != RTRS_MSG_RKEY_RSP)) {
+ rtrs_err(sess->clt, "rkey response is malformed: type %d\n",
+ le16_to_cpu(msg->type));
+ goto out;
+ }
+ buf_id = le16_to_cpu(msg->buf_id);
+ if (WARN_ON(buf_id >= sess->queue_depth))
+ goto out;
+
+ rtrs_from_imm(be32_to_cpu(wc->ex.imm_data), &imm_type, &imm_payload);
+ if (likely(imm_type == RTRS_IO_RSP_IMM ||
+ imm_type == RTRS_IO_RSP_W_INV_IMM)) {
+ u32 msg_id;
+
+ w_inval = (imm_type == RTRS_IO_RSP_W_INV_IMM);
+ rtrs_from_io_rsp_imm(imm_payload, &msg_id, &err);
+
+ if (WARN_ON(buf_id != msg_id))
+ goto out;
+ sess->rbufs[buf_id].rkey = le32_to_cpu(msg->rkey);
+ process_io_rsp(sess, msg_id, err, w_inval);
+ }
+ ib_dma_sync_single_for_device(sess->s.dev->ib_dev, iu->dma_addr,
+ iu->size, DMA_FROM_DEVICE);
+ return rtrs_clt_recv_done(con, wc);
+out:
+ rtrs_rdma_error_recovery(con);
+}
+
+static void rtrs_clt_rdma_done(struct ib_cq *cq, struct ib_wc *wc);
+
+static struct ib_cqe io_comp_cqe = {
+ .done = rtrs_clt_rdma_done
+};
+
+static void rtrs_clt_rdma_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_clt_con *con = cq->cq_context;
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ u32 imm_type, imm_payload;
+ bool w_inval = false;
+ int err;
+
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ if (wc->status != IB_WC_WR_FLUSH_ERR) {
+ rtrs_err(sess->clt, "RDMA failed: %s\n",
+ ib_wc_status_msg(wc->status));
+ rtrs_rdma_error_recovery(con);
+ }
+ return;
+ }
+ rtrs_clt_update_wc_stats(con);
+
+ switch (wc->opcode) {
+ case IB_WC_RECV_RDMA_WITH_IMM:
+ /*
+ * post_recv() RDMA write completions of IO reqs (read/write)
+ * and hb
+ */
+ if (WARN_ON(wc->wr_cqe->done != rtrs_clt_rdma_done))
+ return;
+ rtrs_from_imm(be32_to_cpu(wc->ex.imm_data),
+ &imm_type, &imm_payload);
+ if (likely(imm_type == RTRS_IO_RSP_IMM ||
+ imm_type == RTRS_IO_RSP_W_INV_IMM)) {
+ u32 msg_id;
+
+ w_inval = (imm_type == RTRS_IO_RSP_W_INV_IMM);
+ rtrs_from_io_rsp_imm(imm_payload, &msg_id, &err);
+
+ process_io_rsp(sess, msg_id, err, w_inval);
+ } else if (imm_type == RTRS_HB_MSG_IMM) {
+ WARN_ON(con->c.cid);
+ rtrs_send_hb_ack(&sess->s);
+ if (sess->flags == RTRS_MSG_NEW_RKEY_F)
+ return rtrs_clt_recv_done(con, wc);
+ } else if (imm_type == RTRS_HB_ACK_IMM) {
+ WARN_ON(con->c.cid);
+ sess->s.hb_missed_cnt = 0;
+ if (sess->flags == RTRS_MSG_NEW_RKEY_F)
+ return rtrs_clt_recv_done(con, wc);
+ } else {
+ rtrs_wrn(con->c.sess, "Unknown IMM type %u\n",
+ imm_type);
+ }
+ if (w_inval)
+ /*
+ * Post x2 empty WRs: first is for this RDMA with IMM,
+ * second is for RECV with INV, which happened earlier.
+ */
+ err = rtrs_post_recv_empty_x2(&con->c, &io_comp_cqe);
+ else
+ err = rtrs_post_recv_empty(&con->c, &io_comp_cqe);
+ if (unlikely(err)) {
+ rtrs_err(con->c.sess, "rtrs_post_recv_empty(): %d\n",
+ err);
+ rtrs_rdma_error_recovery(con);
+ break;
+ }
+ break;
+ case IB_WC_RECV:
+ /*
+ * Key invalidations from server side
+ */
+ WARN_ON(!(wc->wc_flags & IB_WC_WITH_INVALIDATE ||
+ wc->wc_flags & IB_WC_WITH_IMM));
+ WARN_ON(wc->wr_cqe->done != rtrs_clt_rdma_done);
+ if (sess->flags == RTRS_MSG_NEW_RKEY_F) {
+ if (wc->wc_flags & IB_WC_WITH_INVALIDATE)
+ return rtrs_clt_recv_done(con, wc);
+
+ return rtrs_clt_rkey_rsp_done(con, wc);
+ }
+ break;
+ case IB_WC_RDMA_WRITE:
+ /*
+ * post_send() RDMA write completions of IO reqs (read/write)
+ * and hb
+ */
+ break;
+
+ default:
+ rtrs_wrn(sess->clt, "Unexpected WC type: %d\n", wc->opcode);
+ return;
+ }
+}
+
+static int post_recv_io(struct rtrs_clt_con *con, size_t q_size)
+{
+ int err, i;
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+
+ for (i = 0; i < q_size; i++) {
+ if (sess->flags == RTRS_MSG_NEW_RKEY_F) {
+ struct rtrs_iu *iu = &con->rsp_ius[i];
+
+ err = rtrs_iu_post_recv(&con->c, iu);
+ } else {
+ err = rtrs_post_recv_empty(&con->c, &io_comp_cqe);
+ }
+ if (unlikely(err))
+ return err;
+ }
+
+ return 0;
+}
+
+static int post_recv_sess(struct rtrs_clt_sess *sess)
+{
+ size_t q_size = 0;
+ int err, cid;
+
+ for (cid = 0; cid < sess->s.con_num; cid++) {
+ if (cid == 0)
+ q_size = SERVICE_CON_QUEUE_DEPTH;
+ else
+ q_size = sess->queue_depth;
+
+ /*
+ * x2 for RDMA read responses + FR key invalidations,
+ * RDMA writes do not require any FR registrations.
+ */
+ q_size *= 2;
+
+ err = post_recv_io(to_clt_con(sess->s.con[cid]), q_size);
+ if (unlikely(err)) {
+ rtrs_err(sess->clt, "post_recv_io(), err: %d\n", err);
+ return err;
+ }
+ }
+
+ return 0;
+}
+
+struct path_it {
+ int i;
+ struct list_head skip_list;
+ struct rtrs_clt *clt;
+ struct rtrs_clt_sess *(*next_path)(struct path_it *it);
+};
+
+#define do_each_path(path, clt, it) { \
+ path_it_init(it, clt); \
+ rcu_read_lock(); \
+ for ((it)->i = 0; ((path) = ((it)->next_path)(it)) && \
+ (it)->i < (it)->clt->paths_num; \
+ (it)->i++)
+
+#define while_each_path(it) \
+ path_it_deinit(it); \
+ rcu_read_unlock(); \
+ }
+
+/**
+ * list_next_or_null_rr_rcu - get next list element in round-robin fashion.
+ * @head: the head for the list.
+ * @ptr: the list head to take the next element from.
+ * @type: the type of the struct this is embedded in.
+ * @memb: the name of the list_head within the struct.
+ *
+ * Next element returned in round-robin fashion, i.e. head will be skipped,
+ * but if list is observed as empty, NULL will be returned.
+ *
+ * This primitive may safely run concurrently with the _rcu list-mutation
+ * primitives such as list_add_rcu() as long as it's guarded by rcu_read_lock().
+ */
+#define list_next_or_null_rr_rcu(head, ptr, type, memb) \
+({ \
+ list_next_or_null_rcu(head, ptr, type, memb) ?: \
+ list_next_or_null_rcu(head, READ_ONCE((ptr)->next), \
+ type, memb); \
+})
+
+/**
+ * get_next_path_rr() - Returns path in round-robin fashion.
+ * @it the path pointer
+ *
+ * Related to @MP_POLICY_RR
+ *
+ * Locks:
+ * rcu_read_lock() must be hold.
+ */
+static struct rtrs_clt_sess *get_next_path_rr(struct path_it *it)
+{
+ struct rtrs_clt_sess __rcu **ppcpu_path;
+ struct rtrs_clt_sess *path;
+ struct rtrs_clt *clt;
+
+ clt = it->clt;
+
+ /*
+ * Here we use two RCU objects: @paths_list and @pcpu_path
+ * pointer. See rtrs_clt_remove_path_from_arr() for details
+ * how that is handled.
+ */
+
+ ppcpu_path = this_cpu_ptr(clt->pcpu_path);
+ path = rcu_dereference(*ppcpu_path);
+ if (unlikely(!path))
+ path = list_first_or_null_rcu(&clt->paths_list,
+ typeof(*path), s.entry);
+ else
+ path = list_next_or_null_rr_rcu(&clt->paths_list,
+ &path->s.entry,
+ typeof(*path),
+ s.entry);
+ rcu_assign_pointer(*ppcpu_path, path);
+
+ return path;
+}
+
+/**
+ * get_next_path_min_inflight() - Returns path with minimal inflight count.
+ * @it the path pointer
+ *
+ * Related to @MP_POLICY_MIN_INFLIGHT
+ *
+ * Locks:
+ * rcu_read_lock() must be hold.
+ */
+static struct rtrs_clt_sess *get_next_path_min_inflight(struct path_it *it)
+{
+ struct rtrs_clt_sess *min_path = NULL;
+ struct rtrs_clt *clt = it->clt;
+ struct rtrs_clt_sess *sess;
+ int min_inflight = INT_MAX;
+ int inflight;
+
+ list_for_each_entry_rcu(sess, &clt->paths_list, s.entry) {
+ if (unlikely(!list_empty(raw_cpu_ptr(sess->mp_skip_entry))))
+ continue;
+
+ inflight = atomic_read(&sess->stats.inflight);
+
+ if (inflight < min_inflight) {
+ min_inflight = inflight;
+ min_path = sess;
+ }
+ }
+
+ /*
+ * add the path to the skip list, so that next time we can get
+ * a different one
+ */
+ if (min_path)
+ list_add(raw_cpu_ptr(min_path->mp_skip_entry), &it->skip_list);
+
+ return min_path;
+}
+
+static inline void path_it_init(struct path_it *it, struct rtrs_clt *clt)
+{
+ INIT_LIST_HEAD(&it->skip_list);
+ it->clt = clt;
+ it->i = 0;
+
+ if (clt->mp_policy == MP_POLICY_RR)
+ it->next_path = get_next_path_rr;
+ else
+ it->next_path = get_next_path_min_inflight;
+}
+
+static inline void path_it_deinit(struct path_it *it)
+{
+ struct list_head *skip, *tmp;
+ /*
+ * The skip_list is used only for the MIN_INFLIGHT policy.
+ * We need to remove paths from it, so that next IO can insert
+ * paths (->mp_skip_entry) into a skip_list again.
+ */
+ list_for_each_safe(skip, tmp, &it->skip_list)
+ list_del_init(skip);
+}
+
+/**
+ * rtrs_clt_init_req() Initialize an rtrs_clt_io_req holding information
+ * about an inflight IO.
+ * The user buffer holding user control message (not data) is copied into
+ * the corresponding buffer of rtrs_iu (req->iu->buf), which later on will
+ * also hold the control message of rtrs.
+ */
+static inline void rtrs_clt_init_req(struct rtrs_clt_io_req *req,
+ struct rtrs_clt_sess *sess,
+ rtrs_conf_fn *conf,
+ struct rtrs_permit *permit, void *priv,
+ const struct kvec *vec, size_t usr_len,
+ struct scatterlist *sg, size_t sg_cnt,
+ size_t data_len, int dir)
+{
+ struct iov_iter iter;
+ size_t len;
+
+ req->permit = permit;
+ req->in_use = true;
+ req->usr_len = usr_len;
+ req->data_len = data_len;
+ req->sglist = sg;
+ req->sg_cnt = sg_cnt;
+ req->priv = priv;
+ req->dir = dir;
+ req->con = rtrs_permit_to_clt_con(sess, permit);
+ req->conf = conf;
+ req->need_inv = false;
+ req->need_inv_comp = false;
+ req->inv_errno = 0;
+
+ iov_iter_kvec(&iter, READ, vec, 1, usr_len);
+ len = _copy_from_iter(req->iu->buf, usr_len, &iter);
+ WARN_ON(len != usr_len);
+
+ reinit_completion(&req->inv_comp);
+ if (sess->stats.enable_rdma_lat)
+ req->start_jiffies = jiffies;
+}
+
+static inline struct rtrs_clt_io_req *
+rtrs_clt_get_req(struct rtrs_clt_sess *sess, rtrs_conf_fn *conf,
+ struct rtrs_permit *permit, void *priv,
+ const struct kvec *vec, size_t usr_len,
+ struct scatterlist *sg, size_t sg_cnt,
+ size_t data_len, int dir)
+{
+ struct rtrs_clt_io_req *req;
+
+ req = &sess->reqs[permit->mem_id];
+ rtrs_clt_init_req(req, sess, conf, permit, priv, vec, usr_len,
+ sg, sg_cnt, data_len, dir);
+ return req;
+}
+
+static inline struct rtrs_clt_io_req *
+rtrs_clt_get_copy_req(struct rtrs_clt_sess *alive_sess,
+ struct rtrs_clt_io_req *fail_req)
+{
+ struct rtrs_clt_io_req *req;
+ struct kvec vec = {
+ .iov_base = fail_req->iu->buf,
+ .iov_len = fail_req->usr_len
+ };
+
+ req = &alive_sess->reqs[fail_req->permit->mem_id];
+ rtrs_clt_init_req(req, alive_sess, fail_req->conf, fail_req->permit,
+ fail_req->priv, &vec, fail_req->usr_len,
+ fail_req->sglist, fail_req->sg_cnt,
+ fail_req->data_len, fail_req->dir);
+ return req;
+}
+
+static int rtrs_post_rdma_write_sg(struct rtrs_clt_con *con,
+ struct rtrs_clt_io_req *req,
+ struct rtrs_rbuf *rbuf,
+ u32 size, u32 imm)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ struct ib_sge *sge = req->sge;
+ enum ib_send_flags flags;
+ struct scatterlist *sg;
+ size_t num_sge;
+ int i;
+
+ for_each_sg(req->sglist, sg, req->sg_cnt, i) {
+ sge[i].addr = sg_dma_address(sg);
+ sge[i].length = sg_dma_len(sg);
+ sge[i].lkey = sess->s.dev->ib_pd->local_dma_lkey;
+ }
+ sge[i].addr = req->iu->dma_addr;
+ sge[i].length = size;
+ sge[i].lkey = sess->s.dev->ib_pd->local_dma_lkey;
+
+ num_sge = 1 + req->sg_cnt;
+
+ /*
+ * From time to time we have to post signalled sends,
+ * or send queue will fill up and only QP reset can help.
+ */
+ flags = atomic_inc_return(&con->io_cnt) % sess->queue_depth ?
+ 0 : IB_SEND_SIGNALED;
+
+ ib_dma_sync_single_for_device(sess->s.dev->ib_dev, req->iu->dma_addr,
+ size, DMA_TO_DEVICE);
+
+ return rtrs_iu_post_rdma_write_imm(&con->c, req->iu, sge, num_sge,
+ rbuf->rkey, rbuf->addr, imm,
+ flags, NULL);
+}
+
+static int rtrs_clt_write_req(struct rtrs_clt_io_req *req)
+{
+ struct rtrs_clt_con *con = req->con;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_clt_sess *sess = to_clt_sess(s);
+ struct rtrs_msg_rdma_write *msg;
+
+ struct rtrs_rbuf *rbuf;
+ int ret, count = 0;
+ u32 imm, buf_id;
+
+ const size_t tsize = sizeof(*msg) + req->data_len + req->usr_len;
+
+ if (unlikely(tsize > sess->chunk_size)) {
+ rtrs_wrn(s, "Write request failed, size too big %zu > %d\n",
+ tsize, sess->chunk_size);
+ return -EMSGSIZE;
+ }
+ if (req->sg_cnt) {
+ count = ib_dma_map_sg(sess->s.dev->ib_dev, req->sglist,
+ req->sg_cnt, req->dir);
+ if (unlikely(!count)) {
+ rtrs_wrn(s, "Write request failed, map failed\n");
+ return -EINVAL;
+ }
+ }
+ /* put rtrs msg after sg and user message */
+ msg = req->iu->buf + req->usr_len;
+ msg->type = cpu_to_le16(RTRS_MSG_WRITE);
+ msg->usr_len = cpu_to_le16(req->usr_len);
+
+ /* rtrs message on server side will be after user data and message */
+ imm = req->permit->mem_off + req->data_len + req->usr_len;
+ imm = rtrs_to_io_req_imm(imm);
+ buf_id = req->permit->mem_id;
+ req->sg_size = tsize;
+ rbuf = &sess->rbufs[buf_id];
+
+ /*
+ * Update stats now, after request is successfully sent it is not
+ * safe anymore to touch it.
+ */
+ rtrs_clt_update_all_stats(req, WRITE);
+
+ ret = rtrs_post_rdma_write_sg(req->con, req, rbuf,
+ req->usr_len + sizeof(*msg),
+ imm);
+ if (unlikely(ret)) {
+ rtrs_err(s, "Write request failed: %d\n", ret);
+ rtrs_clt_decrease_inflight(&sess->stats);
+ if (req->sg_cnt)
+ ib_dma_unmap_sg(sess->s.dev->ib_dev, req->sglist,
+ req->sg_cnt, req->dir);
+ }
+
+ return ret;
+}
+
+static int rtrs_map_sg_fr(struct rtrs_clt_io_req *req, size_t count)
+{
+ int nr;
+
+ /* Align the MR to a 4K page size to match the block virt boundary */
+ nr = ib_map_mr_sg(req->mr, req->sglist, count, NULL, SZ_4K);
+ if (unlikely(nr < req->sg_cnt)) {
+ if (nr < 0)
+ return nr;
+ return -EINVAL;
+ }
+ ib_update_fast_reg_key(req->mr, ib_inc_rkey(req->mr->rkey));
+
+ return nr;
+}
+
+static int rtrs_clt_read_req(struct rtrs_clt_io_req *req)
+{
+ struct rtrs_clt_con *con = req->con;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_clt_sess *sess = to_clt_sess(s);
+ struct rtrs_msg_rdma_read *msg;
+ struct rtrs_ib_dev *dev;
+ struct scatterlist *sg;
+
+ struct ib_reg_wr rwr;
+ struct ib_send_wr *wr = NULL;
+
+ int i, ret, count = 0;
+ u32 imm, buf_id;
+
+ const size_t tsize = sizeof(*msg) + req->data_len + req->usr_len;
+
+ s = &sess->s;
+ dev = sess->s.dev;
+
+ if (unlikely(tsize > sess->chunk_size)) {
+ rtrs_wrn(s,
+ "Read request failed, message size is %zu, bigger than CHUNK_SIZE %d\n",
+ tsize, sess->chunk_size);
+ return -EMSGSIZE;
+ }
+
+ if (req->sg_cnt) {
+ count = ib_dma_map_sg(dev->ib_dev, req->sglist, req->sg_cnt,
+ req->dir);
+ if (unlikely(!count)) {
+ rtrs_wrn(s,
+ "Read request failed, dma map failed\n");
+ return -EINVAL;
+ }
+ }
+ /* put our message into req->buf after user message*/
+ msg = req->iu->buf + req->usr_len;
+ msg->type = cpu_to_le16(RTRS_MSG_READ);
+ msg->usr_len = cpu_to_le16(req->usr_len);
+
+ if (count > noreg_cnt) {
+ ret = rtrs_map_sg_fr(req, count);
+ if (ret < 0) {
+ rtrs_err_rl(s,
+ "Read request failed, failed to map fast reg. data, err: %d\n",
+ ret);
+ ib_dma_unmap_sg(dev->ib_dev, req->sglist, req->sg_cnt,
+ req->dir);
+ return ret;
+ }
+ memset(&rwr, 0, sizeof(rwr));
+ rwr.wr.next = NULL;
+ rwr.wr.opcode = IB_WR_REG_MR;
+ rwr.wr.wr_cqe = &fast_reg_cqe;
+ rwr.wr.num_sge = 0;
+ rwr.mr = req->mr;
+ rwr.key = req->mr->rkey;
+ rwr.access = (IB_ACCESS_LOCAL_WRITE |
+ IB_ACCESS_REMOTE_WRITE);
+ wr = &rwr.wr;
+
+ msg->sg_cnt = cpu_to_le16(1);
+ msg->flags = cpu_to_le16(RTRS_MSG_NEED_INVAL_F);
+
+ msg->desc[0].addr = cpu_to_le64(req->mr->iova);
+ msg->desc[0].key = cpu_to_le32(req->mr->rkey);
+ msg->desc[0].len = cpu_to_le32(req->mr->length);
+
+ /* Further invalidation is required */
+ req->need_inv = !!RTRS_MSG_NEED_INVAL_F;
+
+ } else {
+ msg->sg_cnt = cpu_to_le16(count);
+ msg->flags = 0;
+
+ for_each_sg(req->sglist, sg, req->sg_cnt, i) {
+ msg->desc[i].addr = cpu_to_le64(sg_dma_address(sg));
+ msg->desc[i].key =
+ cpu_to_le32(dev->ib_pd->unsafe_global_rkey);
+ msg->desc[i].len = cpu_to_le32(sg_dma_len(sg));
+ }
+ }
+ /*
+ * rtrs message will be after the space reserved for disk data and
+ * user message
+ */
+ imm = req->permit->mem_off + req->data_len + req->usr_len;
+ imm = rtrs_to_io_req_imm(imm);
+ buf_id = req->permit->mem_id;
+
+ req->sg_size = sizeof(*msg);
+ req->sg_size += le16_to_cpu(msg->sg_cnt) * sizeof(struct rtrs_sg_desc);
+ req->sg_size += req->usr_len;
+
+ /*
+ * Update stats now, after request is successfully sent it is not
+ * safe anymore to touch it.
+ */
+ rtrs_clt_update_all_stats(req, READ);
+
+ ret = rtrs_post_send_rdma(req->con, req, &sess->rbufs[buf_id],
+ req->data_len, imm, wr);
+ if (unlikely(ret)) {
+ rtrs_err(s, "Read request failed: %d\n", ret);
+ rtrs_clt_decrease_inflight(&sess->stats);
+ req->need_inv = false;
+ if (req->sg_cnt)
+ ib_dma_unmap_sg(dev->ib_dev, req->sglist,
+ req->sg_cnt, req->dir);
+ }
+
+ return ret;
+}
+
+/**
+ * rtrs_clt_failover_req() Try to find an active path for a failed request
+ */
+static int rtrs_clt_failover_req(struct rtrs_clt *clt,
+ struct rtrs_clt_io_req *fail_req)
+{
+ struct rtrs_clt_sess *alive_sess;
+ struct rtrs_clt_io_req *req;
+ int err = -ECONNABORTED;
+ struct path_it it;
+
+ do_each_path(alive_sess, clt, &it) {
+ if (unlikely(READ_ONCE(alive_sess->state) !=
+ RTRS_CLT_CONNECTED))
+ continue;
+ req = rtrs_clt_get_copy_req(alive_sess, fail_req);
+ if (req->dir == DMA_TO_DEVICE)
+ err = rtrs_clt_write_req(req);
+ else
+ err = rtrs_clt_read_req(req);
+ if (unlikely(err)) {
+ req->in_use = false;
+ continue;
+ }
+ /* Success path */
+ rtrs_clt_inc_failover_cnt(&alive_sess->stats);
+ break;
+ } while_each_path(&it);
+
+ return err;
+}
+
+static void fail_all_outstanding_reqs(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt *clt = sess->clt;
+ struct rtrs_clt_io_req *req;
+ int i, err;
+
+ if (!sess->reqs)
+ return;
+ for (i = 0; i < sess->queue_depth; ++i) {
+ req = &sess->reqs[i];
+ if (!req->in_use)
+ continue;
+
+ /*
+ * Safely (without notification) complete failed request.
+ * After completion this request is still useble and can
+ * be failovered to another path.
+ */
+ complete_rdma_req(req, -ECONNABORTED, false, true);
+
+ err = rtrs_clt_failover_req(clt, req);
+ if (unlikely(err))
+ /* Failover failed, notify anyway */
+ req->conf(req->priv, err);
+ }
+}
+
+static void free_sess_reqs(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt_io_req *req;
+ int i;
+
+ if (!sess->reqs)
+ return;
+ for (i = 0; i < sess->queue_depth; ++i) {
+ req = &sess->reqs[i];
+ if (req->mr)
+ ib_dereg_mr(req->mr);
+ kfree(req->sge);
+ rtrs_iu_free(req->iu, DMA_TO_DEVICE,
+ sess->s.dev->ib_dev, 1);
+ }
+ kfree(sess->reqs);
+ sess->reqs = NULL;
+}
+
+static int alloc_sess_reqs(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt_io_req *req;
+ struct rtrs_clt *clt = sess->clt;
+ int i, err = -ENOMEM;
+
+ sess->reqs = kcalloc(sess->queue_depth, sizeof(*sess->reqs),
+ GFP_KERNEL);
+ if (unlikely(!sess->reqs))
+ return -ENOMEM;
+
+ for (i = 0; i < sess->queue_depth; ++i) {
+ req = &sess->reqs[i];
+ req->iu = rtrs_iu_alloc(1, sess->max_hdr_size, GFP_KERNEL,
+ sess->s.dev->ib_dev,
+ DMA_TO_DEVICE,
+ rtrs_clt_rdma_done);
+ if (unlikely(!req->iu))
+ goto out;
+
+ req->sge = kmalloc_array(clt->max_segments + 1,
+ sizeof(*req->sge), GFP_KERNEL);
+ if (unlikely(!req->sge))
+ goto out;
+
+ req->mr = ib_alloc_mr(sess->s.dev->ib_pd, IB_MR_TYPE_MEM_REG,
+ sess->max_pages_per_mr);
+ if (IS_ERR(req->mr)) {
+ err = PTR_ERR(req->mr);
+ req->mr = NULL;
+ pr_err("Failed to alloc sess->max_pages_per_mr %d\n",
+ sess->max_pages_per_mr);
+ goto out;
+ }
+
+ init_completion(&req->inv_comp);
+ }
+
+ return 0;
+
+out:
+ free_sess_reqs(sess);
+
+ return err;
+}
+
+static int alloc_permits(struct rtrs_clt *clt)
+{
+ unsigned int chunk_bits;
+ int err, i;
+
+ clt->permits_map = kcalloc(BITS_TO_LONGS(clt->queue_depth),
+ sizeof(long), GFP_KERNEL);
+ if (unlikely(!clt->permits_map)) {
+ err = -ENOMEM;
+ goto out_err;
+ }
+ clt->permits = kcalloc(clt->queue_depth, PERMIT_SIZE(clt), GFP_KERNEL);
+ if (unlikely(!clt->permits)) {
+ err = -ENOMEM;
+ goto err_map;
+ }
+ chunk_bits = ilog2(clt->queue_depth - 1) + 1;
+ for (i = 0; i < clt->queue_depth; i++) {
+ struct rtrs_permit *permit;
+
+ permit = GET_PERMIT(clt, i);
+ permit->mem_id = i;
+ permit->mem_off = i << (MAX_IMM_PAYL_BITS - chunk_bits);
+ }
+
+ return 0;
+
+err_map:
+ kfree(clt->permits_map);
+ clt->permits_map = NULL;
+out_err:
+ return err;
+}
+
+static void free_permits(struct rtrs_clt *clt)
+{
+ kfree(clt->permits_map);
+ clt->permits_map = NULL;
+ kfree(clt->permits);
+ clt->permits = NULL;
+}
+
+static void query_fast_reg_mode(struct rtrs_clt_sess *sess)
+{
+ struct ib_device *ib_dev;
+ u64 max_pages_per_mr;
+ int mr_page_shift;
+
+ ib_dev = sess->s.dev->ib_dev;
+
+ /*
+ * Use the smallest page size supported by the HCA, down to a
+ * minimum of 4096 bytes. We're unlikely to build large sglists
+ * out of smaller entries.
+ */
+ mr_page_shift = max(12, ffs(ib_dev->attrs.page_size_cap) - 1);
+ max_pages_per_mr = ib_dev->attrs.max_mr_size;
+ do_div(max_pages_per_mr, (1ull << mr_page_shift));
+ sess->max_pages_per_mr =
+ min3(sess->max_pages_per_mr, (u32)max_pages_per_mr,
+ ib_dev->attrs.max_fast_reg_page_list_len);
+ sess->max_send_sge = ib_dev->attrs.max_send_sge;
+}
+
+static bool rtrs_clt_change_state_get_old(struct rtrs_clt_sess *sess,
+ enum rtrs_clt_state new_state,
+ enum rtrs_clt_state *old_state)
+{
+ bool changed;
+
+ spin_lock_irq(&sess->state_wq.lock);
+ *old_state = sess->state;
+ changed = __rtrs_clt_change_state(sess, new_state);
+ spin_unlock_irq(&sess->state_wq.lock);
+
+ return changed;
+}
+
+static bool rtrs_clt_change_state(struct rtrs_clt_sess *sess,
+ enum rtrs_clt_state new_state)
+{
+ enum rtrs_clt_state old_state;
+
+ return rtrs_clt_change_state_get_old(sess, new_state, &old_state);
+}
+
+static void rtrs_clt_hb_err_handler(struct rtrs_con *c)
+{
+ struct rtrs_clt_con *con = container_of(c, typeof(*con), c);
+
+ rtrs_rdma_error_recovery(con);
+}
+
+static void rtrs_clt_init_hb(struct rtrs_clt_sess *sess)
+{
+ rtrs_init_hb(&sess->s, &io_comp_cqe,
+ RTRS_HB_INTERVAL_MS,
+ RTRS_HB_MISSED_MAX,
+ rtrs_clt_hb_err_handler,
+ rtrs_wq);
+}
+
+static void rtrs_clt_start_hb(struct rtrs_clt_sess *sess)
+{
+ rtrs_start_hb(&sess->s);
+}
+
+static void rtrs_clt_stop_hb(struct rtrs_clt_sess *sess)
+{
+ rtrs_stop_hb(&sess->s);
+}
+
+static void rtrs_clt_reconnect_work(struct work_struct *work);
+static void rtrs_clt_close_work(struct work_struct *work);
+
+static struct rtrs_clt_sess *alloc_sess(struct rtrs_clt *clt,
+ const struct rtrs_addr *path,
+ size_t con_num, u16 max_segments)
+{
+ struct rtrs_clt_sess *sess;
+ int err = -ENOMEM;
+ int cpu;
+
+ sess = kzalloc(sizeof(*sess), GFP_KERNEL);
+ if (unlikely(!sess))
+ goto err;
+
+ /* Extra connection for user messages */
+ con_num += 1;
+
+ sess->s.con = kcalloc(con_num, sizeof(*sess->s.con), GFP_KERNEL);
+ if (unlikely(!sess->s.con))
+ goto err_free_sess;
+
+ mutex_init(&sess->init_mutex);
+ uuid_gen(&sess->s.uuid);
+ memcpy(&sess->s.dst_addr, path->dst,
+ rdma_addr_size((struct sockaddr *)path->dst));
+
+ /*
+ * rdma_resolve_addr() passes src_addr to cma_bind_addr, which
+ * checks the sa_family to be non-zero. If user passed src_addr=NULL
+ * the sess->src_addr will contain only zeros, which is then fine.
+ */
+ if (path->src)
+ memcpy(&sess->s.src_addr, path->src,
+ rdma_addr_size((struct sockaddr *)path->src));
+ strlcpy(sess->s.sessname, clt->sessname, sizeof(sess->s.sessname));
+ sess->s.con_num = con_num;
+ sess->clt = clt;
+ sess->max_pages_per_mr = max_segments * BLK_MAX_SEGMENT_SIZE >> 12;
+ init_waitqueue_head(&sess->state_wq);
+ sess->state = RTRS_CLT_CONNECTING;
+ atomic_set(&sess->connected_cnt, 0);
+ INIT_WORK(&sess->close_work, rtrs_clt_close_work);
+ INIT_DELAYED_WORK(&sess->reconnect_dwork, rtrs_clt_reconnect_work);
+ rtrs_clt_init_hb(sess);
+
+ sess->mp_skip_entry = alloc_percpu(typeof(*sess->mp_skip_entry));
+ if (unlikely(!sess->mp_skip_entry))
+ goto err_free_con;
+
+ for_each_possible_cpu(cpu)
+ INIT_LIST_HEAD(per_cpu_ptr(sess->mp_skip_entry, cpu));
+
+ err = rtrs_clt_init_stats(&sess->stats);
+ if (unlikely(err))
+ goto err_free_percpu;
+
+ return sess;
+
+err_free_percpu:
+ free_percpu(sess->mp_skip_entry);
+err_free_con:
+ kfree(sess->s.con);
+err_free_sess:
+ kfree(sess);
+err:
+ return ERR_PTR(err);
+}
+
+static void free_sess(struct rtrs_clt_sess *sess)
+{
+ rtrs_clt_free_stats(&sess->stats);
+ free_percpu(sess->mp_skip_entry);
+ kfree(sess->s.con);
+ kfree(sess->rbufs);
+ kfree(sess);
+}
+
+static int create_con(struct rtrs_clt_sess *sess, unsigned int cid)
+{
+ struct rtrs_clt_con *con;
+
+ con = kzalloc(sizeof(*con), GFP_KERNEL);
+ if (unlikely(!con))
+ return -ENOMEM;
+
+ /* Map first two connections to the first CPU */
+ con->cpu = (cid ? cid - 1 : 0) % nr_cpu_ids;
+ con->c.cid = cid;
+ con->c.sess = &sess->s;
+ atomic_set(&con->io_cnt, 0);
+
+ sess->s.con[cid] = &con->c;
+
+ return 0;
+}
+
+static void destroy_con(struct rtrs_clt_con *con)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+
+ sess->s.con[con->c.cid] = NULL;
+ kfree(con);
+}
+
+static int create_con_cq_qp(struct rtrs_clt_con *con)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ u16 wr_queue_size;
+ int err, cq_vector;
+ struct rtrs_msg_rkey_rsp *rsp;
+
+ /*
+ * This function can fail, but still destroy_con_cq_qp() should
+ * be called, this is because create_con_cq_qp() is called on cm
+ * event path, thus caller/waiter never knows: have we failed before
+ * create_con_cq_qp() or after. To solve this dilemma without
+ * creating any additional flags just allow destroy_con_cq_qp() be
+ * called many times.
+ */
+
+ if (con->c.cid == 0) {
+ /*
+ * One completion for each receive and two for each send
+ * (send request + registration)
+ * + 2 for drain and heartbeat
+ * in case qp gets into error state
+ */
+ wr_queue_size = SERVICE_CON_QUEUE_DEPTH * 3 + 2;
+ /* We must be the first here */
+ if (WARN_ON(sess->s.dev))
+ return -EINVAL;
+
+ /*
+ * The whole session uses device from user connection.
+ * Be careful not to close user connection before ib dev
+ * is gracefully put.
+ */
+ sess->s.dev = rtrs_ib_dev_find_or_add(con->c.cm_id->device,
+ &dev_pool);
+ if (unlikely(!sess->s.dev)) {
+ rtrs_wrn(sess->clt,
+ "rtrs_ib_dev_find_get_or_add(): no memory\n");
+ return -ENOMEM;
+ }
+ sess->s.dev_ref = 1;
+ query_fast_reg_mode(sess);
+ } else {
+ /*
+ * Here we assume that session members are correctly set.
+ * This is always true if user connection (cid == 0) is
+ * established first.
+ */
+ if (WARN_ON(!sess->s.dev))
+ return -EINVAL;
+ if (WARN_ON(!sess->queue_depth))
+ return -EINVAL;
+
+ /* Shared between connections */
+ sess->s.dev_ref++;
+ wr_queue_size =
+ min_t(int, sess->s.dev->ib_dev->attrs.max_qp_wr,
+ /* QD * (REQ + RSP + FR REGS or INVS) + drain */
+ sess->queue_depth * 3 + 1);
+ }
+ /* alloc iu to recv new rkey reply when server reports flags set */
+ if (sess->flags == RTRS_MSG_NEW_RKEY_F || con->c.cid == 0) {
+ con->rsp_ius = rtrs_iu_alloc(wr_queue_size, sizeof(*rsp),
+ GFP_KERNEL, sess->s.dev->ib_dev,
+ DMA_FROM_DEVICE,
+ rtrs_clt_rdma_done);
+ if (unlikely(!con->rsp_ius))
+ return -ENOMEM;
+ con->queue_size = wr_queue_size;
+ }
+ cq_vector = con->cpu % sess->s.dev->ib_dev->num_comp_vectors;
+ err = rtrs_cq_qp_create(&sess->s, &con->c, sess->max_send_sge,
+ cq_vector, wr_queue_size, wr_queue_size,
+ IB_POLL_SOFTIRQ);
+ /*
+ * In case of error we do not bother to clean previous allocations,
+ * since destroy_con_cq_qp() must be called.
+ */
+
+ if (unlikely(err))
+ return err;
+ return err;
+}
+
+static void destroy_con_cq_qp(struct rtrs_clt_con *con)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+
+ /*
+ * Be careful here: destroy_con_cq_qp() can be called even
+ * create_con_cq_qp() failed, see comments there.
+ */
+
+ rtrs_cq_qp_destroy(&con->c);
+ if (con->rsp_ius) {
+ rtrs_iu_free(con->rsp_ius, DMA_FROM_DEVICE,
+ sess->s.dev->ib_dev, con->queue_size);
+ con->rsp_ius = NULL;
+ con->queue_size = 0;
+ }
+ if (sess->s.dev_ref && !--sess->s.dev_ref) {
+ rtrs_ib_dev_put(sess->s.dev);
+ sess->s.dev = NULL;
+ }
+}
+
+static void stop_cm(struct rtrs_clt_con *con)
+{
+ rdma_disconnect(con->c.cm_id);
+ if (con->c.qp)
+ ib_drain_qp(con->c.qp);
+}
+
+static void destroy_cm(struct rtrs_clt_con *con)
+{
+ rdma_destroy_id(con->c.cm_id);
+ con->c.cm_id = NULL;
+}
+
+static int rtrs_rdma_addr_resolved(struct rtrs_clt_con *con)
+{
+ struct rtrs_sess *s = con->c.sess;
+ int err;
+
+ err = create_con_cq_qp(con);
+ if (unlikely(err)) {
+ rtrs_err(s, "create_con_cq_qp(), err: %d\n", err);
+ return err;
+ }
+ err = rdma_resolve_route(con->c.cm_id, RTRS_CONNECT_TIMEOUT_MS);
+ if (unlikely(err)) {
+ rtrs_err(s, "Resolving route failed, err: %d\n", err);
+ destroy_con_cq_qp(con);
+ }
+
+ return err;
+}
+
+static int rtrs_rdma_route_resolved(struct rtrs_clt_con *con)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ struct rtrs_clt *clt = sess->clt;
+ struct rtrs_msg_conn_req msg;
+ struct rdma_conn_param param;
+
+ int err;
+
+ memset(¶m, 0, sizeof(param));
+ param.retry_count = clamp(retry_cnt, MIN_RTR_CNT, MAX_RTR_CNT);
+ param.rnr_retry_count = 7;
+ param.private_data = &msg;
+ param.private_data_len = sizeof(msg);
+
+ /*
+ * Those two are the part of struct cma_hdr which is shared
+ * with private_data in case of AF_IB, so put zeroes to avoid
+ * wrong validation inside cma.c on receiver side.
+ */
+ msg.__cma_version = 0;
+ msg.__ip_version = 0;
+ msg.magic = cpu_to_le16(RTRS_MAGIC);
+ msg.version = cpu_to_le16(RTRS_PROTO_VER);
+ msg.cid = cpu_to_le16(con->c.cid);
+ msg.cid_num = cpu_to_le16(sess->s.con_num);
+ msg.recon_cnt = cpu_to_le16(sess->s.recon_cnt);
+ uuid_copy(&msg.sess_uuid, &sess->s.uuid);
+ uuid_copy(&msg.paths_uuid, &clt->paths_uuid);
+
+ err = rdma_connect(con->c.cm_id, ¶m);
+ if (err)
+ rtrs_err(clt, "rdma_connect(): %d\n", err);
+
+ return err;
+}
+
+static int rtrs_rdma_conn_established(struct rtrs_clt_con *con,
+ struct rdma_cm_event *ev)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ struct rtrs_clt *clt = sess->clt;
+ const struct rtrs_msg_conn_rsp *msg;
+ u16 version, queue_depth;
+ int errno;
+ u8 len;
+
+ msg = ev->param.conn.private_data;
+ len = ev->param.conn.private_data_len;
+ if (unlikely(len < sizeof(*msg))) {
+ rtrs_err(clt, "Invalid RTRS connection response\n");
+ return -ECONNRESET;
+ }
+ if (unlikely(le16_to_cpu(msg->magic) != RTRS_MAGIC)) {
+ rtrs_err(clt, "Invalid RTRS magic\n");
+ return -ECONNRESET;
+ }
+ version = le16_to_cpu(msg->version);
+ if (unlikely(version >> 8 != RTRS_PROTO_VER_MAJOR)) {
+ rtrs_err(clt, "Unsupported major RTRS version: %d, expected %d\n",
+ version >> 8, RTRS_PROTO_VER_MAJOR);
+ return -ECONNRESET;
+ }
+ errno = le16_to_cpu(msg->errno);
+ if (unlikely(errno)) {
+ rtrs_err(clt, "Invalid RTRS message: errno %d\n",
+ errno);
+ return -ECONNRESET;
+ }
+ if (con->c.cid == 0) {
+ queue_depth = le16_to_cpu(msg->queue_depth);
+
+ if (queue_depth > MAX_SESS_QUEUE_DEPTH) {
+ rtrs_err(clt, "Invalid RTRS message: queue=%d\n",
+ queue_depth);
+ return -ECONNRESET;
+ }
+ if (!sess->rbufs || sess->queue_depth < queue_depth) {
+ kfree(sess->rbufs);
+ sess->rbufs = kcalloc(queue_depth, sizeof(*sess->rbufs),
+ GFP_KERNEL);
+ if (unlikely(!sess->rbufs)) {
+ rtrs_err(clt,
+ "Failed to allocate queue_depth=%d\n",
+ queue_depth);
+ return -ENOMEM;
+ }
+ }
+ sess->queue_depth = queue_depth;
+ sess->max_hdr_size = le32_to_cpu(msg->max_hdr_size);
+ sess->max_io_size = le32_to_cpu(msg->max_io_size);
+ sess->flags = le32_to_cpu(msg->flags);
+ sess->chunk_size = sess->max_io_size + sess->max_hdr_size;
+
+ /*
+ * Global queue depth and IO size is always a minimum.
+ * If while a reconnection server sends us a value a bit
+ * higher - client does not care and uses cached minimum.
+ *
+ * Since we can have several sessions (paths) restablishing
+ * connections in parallel, use lock.
+ */
+ mutex_lock(&clt->paths_mutex);
+ clt->queue_depth = min_not_zero(sess->queue_depth,
+ clt->queue_depth);
+ clt->max_io_size = min_not_zero(sess->max_io_size,
+ clt->max_io_size);
+ mutex_unlock(&clt->paths_mutex);
+
+ /*
+ * Cache the hca_port and hca_name for sysfs
+ */
+ sess->hca_port = con->c.cm_id->port_num;
+ scnprintf(sess->hca_name, sizeof(sess->hca_name),
+ sess->s.dev->ib_dev->name);
+ sess->s.src_addr = con->c.cm_id->route.addr.src_addr;
+ }
+
+ return 0;
+}
+
+static inline void flag_success_on_conn(struct rtrs_clt_con *con)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+
+ atomic_inc(&sess->connected_cnt);
+ con->cm_err = 1;
+}
+
+static int rtrs_rdma_conn_rejected(struct rtrs_clt_con *con,
+ struct rdma_cm_event *ev)
+{
+ struct rtrs_sess *s = con->c.sess;
+ const struct rtrs_msg_conn_rsp *msg;
+ const char *rej_msg;
+ int status, errno;
+ u8 data_len;
+
+ status = ev->status;
+ rej_msg = rdma_reject_msg(con->c.cm_id, status);
+ msg = rdma_consumer_reject_data(con->c.cm_id, ev, &data_len);
+
+ if (msg && data_len >= sizeof(*msg)) {
+ errno = (int16_t)le16_to_cpu(msg->errno);
+ if (errno == -EBUSY)
+ rtrs_err(s,
+ "Previous session is still exists on the server, please reconnect later\n");
+ else
+ rtrs_err(s,
+ "Connect rejected: status %d (%s), rtrs errno %d\n",
+ status, rej_msg, errno);
+ } else {
+ rtrs_err(s,
+ "Connect rejected but with malformed message: status %d (%s)\n",
+ status, rej_msg);
+ }
+
+ return -ECONNRESET;
+}
+
+static void rtrs_clt_close_conns(struct rtrs_clt_sess *sess, bool wait)
+{
+ if (rtrs_clt_change_state(sess, RTRS_CLT_CLOSING))
+ queue_work(rtrs_wq, &sess->close_work);
+ if (wait)
+ flush_work(&sess->close_work);
+}
+
+static inline void flag_error_on_conn(struct rtrs_clt_con *con, int cm_err)
+{
+ if (con->cm_err == 1) {
+ struct rtrs_clt_sess *sess;
+
+ sess = to_clt_sess(con->c.sess);
+ if (atomic_dec_and_test(&sess->connected_cnt))
+ wake_up(&sess->state_wq);
+ }
+ con->cm_err = cm_err;
+}
+
+static int rtrs_clt_rdma_cm_handler(struct rdma_cm_id *cm_id,
+ struct rdma_cm_event *ev)
+{
+ struct rtrs_clt_con *con = cm_id->context;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_clt_sess *sess = to_clt_sess(s);
+ int cm_err = 0;
+
+ switch (ev->event) {
+ case RDMA_CM_EVENT_ADDR_RESOLVED:
+ cm_err = rtrs_rdma_addr_resolved(con);
+ break;
+ case RDMA_CM_EVENT_ROUTE_RESOLVED:
+ cm_err = rtrs_rdma_route_resolved(con);
+ break;
+ case RDMA_CM_EVENT_ESTABLISHED:
+ con->cm_err = rtrs_rdma_conn_established(con, ev);
+ if (likely(!con->cm_err)) {
+ /*
+ * Report success and wake up. Here we abuse state_wq,
+ * i.e. wake up without state change, but we set cm_err.
+ */
+ flag_success_on_conn(con);
+ wake_up(&sess->state_wq);
+ return 0;
+ }
+ break;
+ case RDMA_CM_EVENT_REJECTED:
+ cm_err = rtrs_rdma_conn_rejected(con, ev);
+ break;
+ case RDMA_CM_EVENT_CONNECT_ERROR:
+ case RDMA_CM_EVENT_UNREACHABLE:
+ rtrs_wrn(s, "CM error event %d\n", ev->event);
+ cm_err = -ECONNRESET;
+ break;
+ case RDMA_CM_EVENT_ADDR_ERROR:
+ case RDMA_CM_EVENT_ROUTE_ERROR:
+ cm_err = -EHOSTUNREACH;
+ break;
+ case RDMA_CM_EVENT_DISCONNECTED:
+ case RDMA_CM_EVENT_ADDR_CHANGE:
+ case RDMA_CM_EVENT_TIMEWAIT_EXIT:
+ cm_err = -ECONNRESET;
+ break;
+ case RDMA_CM_EVENT_DEVICE_REMOVAL:
+ /*
+ * Device removal is a special case. Queue close and return 0.
+ */
+ rtrs_clt_close_conns(sess, false);
+ return 0;
+ default:
+ rtrs_err(s, "Unexpected RDMA CM event (%d)\n", ev->event);
+ cm_err = -ECONNRESET;
+ break;
+ }
+
+ if (cm_err) {
+ /*
+ * cm error makes sense only on connection establishing,
+ * in other cases we rely on normal procedure of reconnecting.
+ */
+ flag_error_on_conn(con, cm_err);
+ rtrs_rdma_error_recovery(con);
+ }
+
+ return 0;
+}
+
+static int create_cm(struct rtrs_clt_con *con)
+{
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_clt_sess *sess = to_clt_sess(s);
+ struct rdma_cm_id *cm_id;
+ int err;
+
+ cm_id = rdma_create_id(&init_net, rtrs_clt_rdma_cm_handler, con,
+ sess->s.dst_addr.ss_family == AF_IB ?
+ RDMA_PS_IB : RDMA_PS_TCP, IB_QPT_RC);
+ if (IS_ERR(cm_id)) {
+ err = PTR_ERR(cm_id);
+ rtrs_err(s, "Failed to create CM ID, err: %d\n", err);
+
+ return err;
+ }
+ con->c.cm_id = cm_id;
+ con->cm_err = 0;
+ /* allow the port to be reused */
+ err = rdma_set_reuseaddr(cm_id, 1);
+ if (err != 0) {
+ rtrs_err(s, "Set address reuse failed, err: %d\n", err);
+ goto destroy_cm;
+ }
+ err = rdma_resolve_addr(cm_id, (struct sockaddr *)&sess->s.src_addr,
+ (struct sockaddr *)&sess->s.dst_addr,
+ RTRS_CONNECT_TIMEOUT_MS);
+ if (unlikely(err)) {
+ rtrs_err(s, "Failed to resolve address, err: %d\n", err);
+ goto destroy_cm;
+ }
+ /*
+ * Combine connection status and session events. This is needed
+ * for waiting two possible cases: cm_err has something meaningful
+ * or session state was really changed to error by device removal.
+ */
+ err = wait_event_interruptible_timeout(
+ sess->state_wq,
+ con->cm_err || sess->state != RTRS_CLT_CONNECTING,
+ msecs_to_jiffies(RTRS_CONNECT_TIMEOUT_MS));
+ if (unlikely(err == 0 || err == -ERESTARTSYS)) {
+ if (err == 0)
+ err = -ETIMEDOUT;
+ /* Timedout or interrupted */
+ goto errr;
+ }
+ if (unlikely(con->cm_err < 0)) {
+ err = con->cm_err;
+ goto errr;
+ }
+ if (unlikely(READ_ONCE(sess->state) != RTRS_CLT_CONNECTING)) {
+ /* Device removal */
+ err = -ECONNABORTED;
+ goto errr;
+ }
+
+ return 0;
+
+errr:
+ stop_cm(con);
+ /* Is safe to call destroy if cq_qp is not inited */
+ destroy_con_cq_qp(con);
+destroy_cm:
+ destroy_cm(con);
+
+ return err;
+}
+
+static void rtrs_clt_sess_up(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt *clt = sess->clt;
+ int up;
+
+ /*
+ * We can fire RECONNECTED event only when all paths were
+ * connected on rtrs_clt_open(), then each was disconnected
+ * and the first one connected again. That's why this nasty
+ * game with counter value.
+ */
+
+ mutex_lock(&clt->paths_ev_mutex);
+ up = ++clt->paths_up;
+ /*
+ * Here it is safe to access paths num directly since up counter
+ * is greater than MAX_PATHS_NUM only while rtrs_clt_open() is
+ * in progress, thus paths removals are impossible.
+ */
+ if (up > MAX_PATHS_NUM && up == MAX_PATHS_NUM + clt->paths_num)
+ clt->paths_up = clt->paths_num;
+ else if (up == 1)
+ clt->link_ev(clt->priv, RTRS_CLT_LINK_EV_RECONNECTED);
+ mutex_unlock(&clt->paths_ev_mutex);
+
+ /* Mark session as established */
+ sess->established = true;
+ sess->reconnect_attempts = 0;
+ sess->stats.reconnects.successful_cnt++;
+}
+
+static void rtrs_clt_sess_down(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt *clt = sess->clt;
+
+ if (!sess->established)
+ return;
+
+ sess->established = false;
+ mutex_lock(&clt->paths_ev_mutex);
+ WARN_ON(!clt->paths_up);
+ if (--clt->paths_up == 0)
+ clt->link_ev(clt->priv, RTRS_CLT_LINK_EV_DISCONNECTED);
+ mutex_unlock(&clt->paths_ev_mutex);
+}
+
+static void rtrs_clt_stop_and_destroy_conns(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt_con *con;
+ unsigned int cid;
+
+ WARN_ON(READ_ONCE(sess->state) == RTRS_CLT_CONNECTED);
+
+ /*
+ * Possible race with rtrs_clt_open(), when DEVICE_REMOVAL comes
+ * exactly in between. Start destroying after it finishes.
+ */
+ mutex_lock(&sess->init_mutex);
+ mutex_unlock(&sess->init_mutex);
+
+ /*
+ * All IO paths must observe !CONNECTED state before we
+ * free everything.
+ */
+ synchronize_rcu();
+
+ rtrs_clt_stop_hb(sess);
+
+ /*
+ * The order it utterly crucial: firstly disconnect and complete all
+ * rdma requests with error (thus set in_use=false for requests),
+ * then fail outstanding requests checking in_use for each, and
+ * eventually notify upper layer about session disconnection.
+ */
+
+ for (cid = 0; cid < sess->s.con_num; cid++) {
+ if (!sess->s.con[cid])
+ break;
+ con = to_clt_con(sess->s.con[cid]);
+ stop_cm(con);
+ }
+ fail_all_outstanding_reqs(sess);
+ free_sess_reqs(sess);
+ rtrs_clt_sess_down(sess);
+
+ /*
+ * Wait for graceful shutdown, namely when peer side invokes
+ * rdma_disconnect(). 'connected_cnt' is decremented only on
+ * CM events, thus if other side had crashed and hb has detected
+ * something is wrong, here we will stuck for exactly timeout ms,
+ * since CM does not fire anything. That is fine, we are not in
+ * hurry.
+ */
+ wait_event_timeout(sess->state_wq, !atomic_read(&sess->connected_cnt),
+ msecs_to_jiffies(RTRS_CONNECT_TIMEOUT_MS));
+
+ for (cid = 0; cid < sess->s.con_num; cid++) {
+ if (!sess->s.con[cid])
+ break;
+ con = to_clt_con(sess->s.con[cid]);
+ destroy_con_cq_qp(con);
+ destroy_cm(con);
+ destroy_con(con);
+ }
+}
+
+static inline bool xchg_sessions(struct rtrs_clt_sess __rcu **rcu_ppcpu_path,
+ struct rtrs_clt_sess *sess,
+ struct rtrs_clt_sess *next)
+{
+ struct rtrs_clt_sess **ppcpu_path;
+
+ /* Call cmpxchg() without sparse warnings */
+ ppcpu_path = (typeof(ppcpu_path))rcu_ppcpu_path;
+ return (sess == cmpxchg(ppcpu_path, sess, next));
+}
+
+static void rtrs_clt_remove_path_from_arr(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt *clt = sess->clt;
+ struct rtrs_clt_sess *next;
+ bool wait_for_grace = false;
+ int cpu;
+
+ mutex_lock(&clt->paths_mutex);
+ list_del_rcu(&sess->s.entry);
+
+ /* Make sure everybody observes path removal. */
+ synchronize_rcu();
+
+ /*
+ * At this point nobody sees @sess in the list, but still we have
+ * dangling pointer @pcpu_path which _can_ point to @sess. Since
+ * nobody can observe @sess in the list, we guarantee that IO path
+ * will not assign @sess to @pcpu_path, i.e. @pcpu_path can be equal
+ * to @sess, but can never again become @sess.
+ */
+
+ /*
+ * Decrement paths number only after grace period, because
+ * caller of do_each_path() must firstly observe list without
+ * path and only then decremented paths number.
+ *
+ * Otherwise there can be the following situation:
+ * o Two paths exist and IO is coming.
+ * o One path is removed:
+ * CPU#0 CPU#1
+ * do_each_path(): rtrs_clt_remove_path_from_arr():
+ * path = get_next_path()
+ * ^^^ list_del_rcu(path)
+ * [!CONNECTED path] clt->paths_num--
+ * ^^^^^^^^^
+ * load clt->paths_num from 2 to 1
+ * ^^^^^^^^^
+ * sees 1
+ *
+ * path is observed as !CONNECTED, but do_each_path() loop
+ * ends, because expression i < clt->paths_num is false.
+ */
+ clt->paths_num--;
+
+ /*
+ * Get @next connection from current @sess which is going to be
+ * removed. If @sess is the last element, then @next is NULL.
+ */
+ next = list_next_or_null_rr_rcu(&clt->paths_list, &sess->s.entry,
+ typeof(*next), s.entry);
+
+ /*
+ * @pcpu paths can still point to the path which is going to be
+ * removed, so change the pointer manually.
+ */
+ for_each_possible_cpu(cpu) {
+ struct rtrs_clt_sess __rcu **ppcpu_path;
+
+ ppcpu_path = per_cpu_ptr(clt->pcpu_path, cpu);
+ if (rcu_dereference(*ppcpu_path) != sess)
+ /*
+ * synchronize_rcu() was called just after deleting
+ * entry from the list, thus IO code path cannot
+ * change pointer back to the pointer which is going
+ * to be removed, we are safe here.
+ */
+ continue;
+
+ /*
+ * We race with IO code path, which also changes pointer,
+ * thus we have to be careful not to overwrite it.
+ */
+ if (xchg_sessions(ppcpu_path, sess, next))
+ /*
+ * @ppcpu_path was successfully replaced with @next,
+ * that means that someone could also pick up the
+ * @sess and dereferencing it right now, so wait for
+ * a grace period is required.
+ */
+ wait_for_grace = true;
+ }
+ if (wait_for_grace)
+ synchronize_rcu();
+
+ mutex_unlock(&clt->paths_mutex);
+}
+
+static void rtrs_clt_add_path_to_arr(struct rtrs_clt_sess *sess,
+ struct rtrs_addr *addr)
+{
+ struct rtrs_clt *clt = sess->clt;
+
+ mutex_lock(&clt->paths_mutex);
+ clt->paths_num++;
+
+ /*
+ * Firstly increase paths_num, wait for GP and then
+ * add path to the list. Why? Since we add path with
+ * !CONNECTED state explanation is similar to what has
+ * been written in rtrs_clt_remove_path_from_arr().
+ */
+ synchronize_rcu();
+
+ list_add_tail_rcu(&sess->s.entry, &clt->paths_list);
+ mutex_unlock(&clt->paths_mutex);
+}
+
+static void rtrs_clt_close_work(struct work_struct *work)
+{
+ struct rtrs_clt_sess *sess;
+
+ sess = container_of(work, struct rtrs_clt_sess, close_work);
+
+ cancel_delayed_work_sync(&sess->reconnect_dwork);
+ rtrs_clt_stop_and_destroy_conns(sess);
+ /*
+ * Sounds stupid, huh? No, it is not. Consider this sequence:
+ *
+ * #CPU0 #CPU1
+ * 1. CONNECTED->RECONNECTING
+ * 2. RECONNECTING->CLOSING
+ * 3. queue_work(&reconnect_dwork)
+ * 4. queue_work(&close_work);
+ * 5. reconnect_work(); close_work();
+ *
+ * To avoid that case do cancel twice: before and after.
+ */
+ cancel_delayed_work_sync(&sess->reconnect_dwork);
+ rtrs_clt_change_state(sess, RTRS_CLT_CLOSED);
+}
+
+static int init_conns(struct rtrs_clt_sess *sess)
+{
+ unsigned int cid;
+ int err;
+
+ /*
+ * On every new session connections increase reconnect counter
+ * to avoid clashes with previous sessions not yet closed
+ * sessions on a server side.
+ */
+ sess->s.recon_cnt++;
+
+ /* Establish all RDMA connections */
+ for (cid = 0; cid < sess->s.con_num; cid++) {
+ err = create_con(sess, cid);
+ if (unlikely(err))
+ goto destroy;
+
+ err = create_cm(to_clt_con(sess->s.con[cid]));
+ if (unlikely(err)) {
+ destroy_con(to_clt_con(sess->s.con[cid]));
+ goto destroy;
+ }
+ }
+ err = alloc_sess_reqs(sess);
+ if (unlikely(err))
+ goto destroy;
+
+ rtrs_clt_start_hb(sess);
+
+ return 0;
+
+destroy:
+ while (cid--) {
+ struct rtrs_clt_con *con = to_clt_con(sess->s.con[cid]);
+
+ stop_cm(con);
+ destroy_con_cq_qp(con);
+ destroy_cm(con);
+ destroy_con(con);
+ }
+ /*
+ * If we've never taken async path and got an error, say,
+ * doing rdma_resolve_addr(), switch to CONNECTION_ERR state
+ * manually to keep reconnecting.
+ */
+ rtrs_clt_change_state(sess, RTRS_CLT_CONNECTING_ERR);
+
+ return err;
+}
+
+static void rtrs_clt_info_req_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_clt_con *con = cq->cq_context;
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ struct rtrs_iu *iu;
+
+ iu = container_of(wc->wr_cqe, struct rtrs_iu, cqe);
+ rtrs_iu_free(iu, DMA_TO_DEVICE, sess->s.dev->ib_dev, 1);
+
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ rtrs_err(sess->clt, "Sess info request send failed: %s\n",
+ ib_wc_status_msg(wc->status));
+ rtrs_clt_change_state(sess, RTRS_CLT_CONNECTING_ERR);
+ return;
+ }
+
+ rtrs_clt_update_wc_stats(con);
+}
+
+static int process_info_rsp(struct rtrs_clt_sess *sess,
+ const struct rtrs_msg_info_rsp *msg)
+{
+ unsigned int sg_cnt, total_len;
+ int i, sgi;
+
+ sg_cnt = le16_to_cpu(msg->sg_cnt);
+ if (unlikely(!sg_cnt))
+ return -EINVAL;
+ /*
+ * Check if IB immediate data size is enough to hold the mem_id and
+ * the offset inside the memory chunk.
+ */
+ if (unlikely((ilog2(sg_cnt - 1) + 1) +
+ (ilog2(sess->chunk_size - 1) + 1) >
+ MAX_IMM_PAYL_BITS)) {
+ rtrs_err(sess->clt,
+ "RDMA immediate size (%db) not enough to encode %d buffers of size %dB\n",
+ MAX_IMM_PAYL_BITS, sg_cnt, sess->chunk_size);
+ return -EINVAL;
+ }
+ if (unlikely(!sg_cnt || (sess->queue_depth % sg_cnt))) {
+ rtrs_err(sess->clt, "Incorrect sg_cnt %d, is not multiple\n",
+ sg_cnt);
+ return -EINVAL;
+ }
+ total_len = 0;
+ for (sgi = 0, i = 0; sgi < sg_cnt && i < sess->queue_depth; sgi++) {
+ const struct rtrs_sg_desc *desc = &msg->desc[sgi];
+ u32 len, rkey;
+ u64 addr;
+
+ addr = le64_to_cpu(desc->addr);
+ rkey = le32_to_cpu(desc->key);
+ len = le32_to_cpu(desc->len);
+
+ total_len += len;
+
+ if (unlikely(!len || (len % sess->chunk_size))) {
+ rtrs_err(sess->clt, "Incorrect [%d].len %d\n", sgi,
+ len);
+ return -EINVAL;
+ }
+ for ( ; len && i < sess->queue_depth; i++) {
+ sess->rbufs[i].addr = addr;
+ sess->rbufs[i].rkey = rkey;
+
+ len -= sess->chunk_size;
+ addr += sess->chunk_size;
+ }
+ }
+ /* Sanity check */
+ if (unlikely(sgi != sg_cnt || i != sess->queue_depth)) {
+ rtrs_err(sess->clt, "Incorrect sg vector, not fully mapped\n");
+ return -EINVAL;
+ }
+ if (unlikely(total_len != sess->chunk_size * sess->queue_depth)) {
+ rtrs_err(sess->clt, "Incorrect total_len %d\n", total_len);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static void rtrs_clt_info_rsp_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_clt_con *con = cq->cq_context;
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ struct rtrs_msg_info_rsp *msg;
+ enum rtrs_clt_state state;
+ struct rtrs_iu *iu;
+ size_t rx_sz;
+ int err;
+
+ state = RTRS_CLT_CONNECTING_ERR;
+
+ WARN_ON(con->c.cid);
+ iu = container_of(wc->wr_cqe, struct rtrs_iu, cqe);
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ rtrs_err(sess->clt, "Sess info response recv failed: %s\n",
+ ib_wc_status_msg(wc->status));
+ goto out;
+ }
+ WARN_ON(wc->opcode != IB_WC_RECV);
+
+ if (unlikely(wc->byte_len < sizeof(*msg))) {
+ rtrs_err(sess->clt, "Sess info response is malformed: size %d\n",
+ wc->byte_len);
+ goto out;
+ }
+ ib_dma_sync_single_for_cpu(sess->s.dev->ib_dev, iu->dma_addr,
+ iu->size, DMA_FROM_DEVICE);
+ msg = iu->buf;
+ if (unlikely(le16_to_cpu(msg->type) != RTRS_MSG_INFO_RSP)) {
+ rtrs_err(sess->clt, "Sess info response is malformed: type %d\n",
+ le16_to_cpu(msg->type));
+ goto out;
+ }
+ rx_sz = sizeof(*msg);
+ rx_sz += sizeof(msg->desc[0]) * le16_to_cpu(msg->sg_cnt);
+ if (unlikely(wc->byte_len < rx_sz)) {
+ rtrs_err(sess->clt, "Sess info response is malformed: size %d\n",
+ wc->byte_len);
+ goto out;
+ }
+ err = process_info_rsp(sess, msg);
+ if (unlikely(err))
+ goto out;
+
+ err = post_recv_sess(sess);
+ if (unlikely(err))
+ goto out;
+
+ state = RTRS_CLT_CONNECTED;
+
+out:
+ rtrs_clt_update_wc_stats(con);
+ rtrs_iu_free(iu, DMA_FROM_DEVICE, sess->s.dev->ib_dev, 1);
+ rtrs_clt_change_state(sess, state);
+}
+
+static int rtrs_send_sess_info(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt_con *usr_con = to_clt_con(sess->s.con[0]);
+ struct rtrs_msg_info_req *msg;
+ struct rtrs_iu *tx_iu, *rx_iu;
+ size_t rx_sz;
+ int err;
+
+ rx_sz = sizeof(struct rtrs_msg_info_rsp);
+ rx_sz += sizeof(u64) * MAX_SESS_QUEUE_DEPTH;
+
+ tx_iu = rtrs_iu_alloc(1, sizeof(struct rtrs_msg_info_req), GFP_KERNEL,
+ sess->s.dev->ib_dev, DMA_TO_DEVICE,
+ rtrs_clt_info_req_done);
+ rx_iu = rtrs_iu_alloc(1, rx_sz, GFP_KERNEL, sess->s.dev->ib_dev,
+ DMA_FROM_DEVICE, rtrs_clt_info_rsp_done);
+ if (unlikely(!tx_iu || !rx_iu)) {
+ rtrs_err(sess->clt, "rtrs_iu_alloc(): no memory\n");
+ err = -ENOMEM;
+ goto out;
+ }
+ /* Prepare for getting info response */
+ err = rtrs_iu_post_recv(&usr_con->c, rx_iu);
+ if (unlikely(err)) {
+ rtrs_err(sess->clt, "rtrs_iu_post_recv(), err: %d\n", err);
+ goto out;
+ }
+ rx_iu = NULL;
+
+ msg = tx_iu->buf;
+ msg->type = cpu_to_le16(RTRS_MSG_INFO_REQ);
+ memcpy(msg->sessname, sess->s.sessname, sizeof(msg->sessname));
+
+ ib_dma_sync_single_for_device(sess->s.dev->ib_dev, tx_iu->dma_addr,
+ tx_iu->size, DMA_TO_DEVICE);
+
+ /* Send info request */
+ err = rtrs_iu_post_send(&usr_con->c, tx_iu, sizeof(*msg), NULL);
+ if (unlikely(err)) {
+ rtrs_err(sess->clt, "rtrs_iu_post_send(), err: %d\n", err);
+ goto out;
+ }
+ tx_iu = NULL;
+
+ /* Wait for state change */
+ wait_event_interruptible_timeout(sess->state_wq,
+ sess->state != RTRS_CLT_CONNECTING,
+ msecs_to_jiffies(
+ RTRS_CONNECT_TIMEOUT_MS));
+ if (unlikely(READ_ONCE(sess->state) != RTRS_CLT_CONNECTED)) {
+ if (READ_ONCE(sess->state) == RTRS_CLT_CONNECTING_ERR)
+ err = -ECONNRESET;
+ else
+ err = -ETIMEDOUT;
+ goto out;
+ }
+
+out:
+ if (tx_iu)
+ rtrs_iu_free(tx_iu, DMA_TO_DEVICE, sess->s.dev->ib_dev, 1);
+ if (rx_iu)
+ rtrs_iu_free(rx_iu, DMA_FROM_DEVICE, sess->s.dev->ib_dev, 1);
+ if (unlikely(err))
+ /* If we've never taken async path because of malloc problems */
+ rtrs_clt_change_state(sess, RTRS_CLT_CONNECTING_ERR);
+
+ return err;
+}
+
+/**
+ * init_sess() - establishes all session connections and does handshake
+ *
+ * In case of error full close or reconnect procedure should be taken,
+ * because reconnect or close async works can be started.
+ */
+static int init_sess(struct rtrs_clt_sess *sess)
+{
+ int err;
+
+ mutex_lock(&sess->init_mutex);
+ err = init_conns(sess);
+ if (unlikely(err)) {
+ rtrs_err(sess->clt, "init_conns(), err: %d\n", err);
+ goto out;
+ }
+ err = rtrs_send_sess_info(sess);
+ if (unlikely(err)) {
+ rtrs_err(sess->clt, "rtrs_send_sess_info(), err: %d\n", err);
+ goto out;
+ }
+ rtrs_clt_sess_up(sess);
+out:
+ mutex_unlock(&sess->init_mutex);
+
+ return err;
+}
+
+static void rtrs_clt_reconnect_work(struct work_struct *work)
+{
+ struct rtrs_clt_sess *sess;
+ struct rtrs_clt *clt;
+ unsigned int delay_ms;
+ int err;
+
+ sess = container_of(to_delayed_work(work), struct rtrs_clt_sess,
+ reconnect_dwork);
+ clt = sess->clt;
+
+ if (READ_ONCE(sess->state) == RTRS_CLT_CLOSING)
+ /* User requested closing */
+ return;
+
+ if (sess->reconnect_attempts >= clt->max_reconnect_attempts) {
+ /* Close a session completely if max attempts is reached */
+ rtrs_clt_close_conns(sess, false);
+ return;
+ }
+ sess->reconnect_attempts++;
+
+ /* Stop everything */
+ rtrs_clt_stop_and_destroy_conns(sess);
+ rtrs_clt_change_state(sess, RTRS_CLT_CONNECTING);
+
+ err = init_sess(sess);
+ if (unlikely(err))
+ goto reconnect_again;
+
+ return;
+
+reconnect_again:
+ if (rtrs_clt_change_state(sess, RTRS_CLT_RECONNECTING)) {
+ sess->stats.reconnects.fail_cnt++;
+ delay_ms = clt->reconnect_delay_sec * 1000;
+ queue_delayed_work(rtrs_wq, &sess->reconnect_dwork,
+ msecs_to_jiffies(delay_ms));
+ }
+}
+
+static void rtrs_clt_dev_release(struct device *dev)
+{
+ struct rtrs_clt *clt = container_of(dev, struct rtrs_clt, dev);
+
+ kfree(clt);
+}
+
+static struct rtrs_clt *alloc_clt(const char *sessname, size_t paths_num,
+ short port, size_t pdu_sz,
+ void *priv, link_clt_ev_fn *link_ev,
+ unsigned int max_segments,
+ unsigned int reconnect_delay_sec,
+ unsigned int max_reconnect_attempts)
+{
+ struct rtrs_clt *clt;
+ int err;
+
+ if (unlikely(!paths_num || paths_num > MAX_PATHS_NUM))
+ return ERR_PTR(-EINVAL);
+
+ if (unlikely(strlen(sessname) >= sizeof(clt->sessname)))
+ return ERR_PTR(-EINVAL);
+
+ clt = kzalloc(sizeof(*clt), GFP_KERNEL);
+ if (unlikely(!clt))
+ return ERR_PTR(-ENOMEM);
+
+ clt->pcpu_path = alloc_percpu(typeof(*clt->pcpu_path));
+ if (unlikely(!clt->pcpu_path)) {
+ kfree(clt);
+ return ERR_PTR(-ENOMEM);
+ }
+
+ uuid_gen(&clt->paths_uuid);
+ INIT_LIST_HEAD_RCU(&clt->paths_list);
+ clt->paths_num = paths_num;
+ clt->paths_up = MAX_PATHS_NUM;
+ clt->port = port;
+ clt->pdu_sz = pdu_sz;
+ clt->max_segments = max_segments;
+ clt->reconnect_delay_sec = reconnect_delay_sec;
+ clt->max_reconnect_attempts = max_reconnect_attempts;
+ clt->priv = priv;
+ clt->link_ev = link_ev;
+ clt->mp_policy = MP_POLICY_MIN_INFLIGHT;
+ strlcpy(clt->sessname, sessname, sizeof(clt->sessname));
+ init_waitqueue_head(&clt->permits_wait);
+ mutex_init(&clt->paths_ev_mutex);
+ mutex_init(&clt->paths_mutex);
+
+ clt->dev.class = rtrs_dev_class;
+ clt->dev.release = rtrs_clt_dev_release;
+ dev_set_name(&clt->dev, "%s", sessname);
+
+ err = device_register(&clt->dev);
+ if (unlikely(err))
+ goto percpu_free;
+
+ err = rtrs_clt_create_sysfs_root_folders(clt);
+ if (unlikely(err))
+ goto dev_unregister;
+
+ return clt;
+
+dev_unregister:
+ device_unregister(&clt->dev);
+percpu_free:
+ free_percpu(clt->pcpu_path);
+ kfree(clt);
+ return ERR_PTR(err);
+}
+
+static void wait_for_inflight_permits(struct rtrs_clt *clt)
+{
+ if (clt->permits_map) {
+ size_t sz = clt->queue_depth;
+
+ wait_event(clt->permits_wait,
+ find_first_bit(clt->permits_map, sz) >= sz);
+ }
+}
+
+static void free_clt(struct rtrs_clt *clt)
+{
+ rtrs_clt_destroy_sysfs_root_folders(clt);
+ wait_for_inflight_permits(clt);
+ free_permits(clt);
+ free_percpu(clt->pcpu_path);
+ /* release callback will free clt in last put */
+ device_unregister(&clt->dev);
+}
+
+struct rtrs_clt *rtrs_clt_open(void *priv, link_clt_ev_fn *link_ev,
+ const char *sessname,
+ const struct rtrs_addr *paths,
+ size_t paths_num,
+ short port,
+ size_t pdu_sz, u8 reconnect_delay_sec,
+ u16 max_segments,
+ s16 max_reconnect_attempts)
+{
+ struct rtrs_clt_sess *sess, *tmp;
+ struct rtrs_clt *clt;
+ int err, i;
+
+ clt = alloc_clt(sessname, paths_num, port, pdu_sz, priv, link_ev,
+ max_segments, reconnect_delay_sec,
+ max_reconnect_attempts);
+ if (IS_ERR(clt)) {
+ err = PTR_ERR(clt);
+ goto out;
+ }
+ for (i = 0; i < paths_num; i++) {
+ struct rtrs_clt_sess *sess;
+
+ sess = alloc_sess(clt, &paths[i], nr_cons_per_session,
+ max_segments);
+ if (IS_ERR(sess)) {
+ err = PTR_ERR(sess);
+ rtrs_err(clt, "alloc_sess(), err: %d\n", err);
+ goto close_all_sess;
+ }
+ list_add_tail_rcu(&sess->s.entry, &clt->paths_list);
+
+ err = init_sess(sess);
+ if (unlikely(err))
+ goto close_all_sess;
+
+ err = rtrs_clt_create_sess_files(sess);
+ if (unlikely(err))
+ goto close_all_sess;
+ }
+ err = alloc_permits(clt);
+ if (unlikely(err)) {
+ rtrs_err(clt, "alloc_permits(), err: %d\n", err);
+ goto close_all_sess;
+ }
+ err = rtrs_clt_create_sysfs_root_files(clt);
+ if (unlikely(err))
+ goto close_all_sess;
+
+ /*
+ * There is a race if someone decides to completely remove just
+ * newly created path using sysfs entry. To avoid the race we
+ * use simple 'opened' flag, see rtrs_clt_remove_path_from_sysfs().
+ */
+ clt->opened = true;
+
+ /* Do not let module be unloaded if client is alive */
+ __module_get(THIS_MODULE);
+
+ return clt;
+
+close_all_sess:
+ list_for_each_entry_safe(sess, tmp, &clt->paths_list, s.entry) {
+ rtrs_clt_destroy_sess_files(sess, NULL);
+ rtrs_clt_close_conns(sess, true);
+ free_sess(sess);
+ }
+ free_clt(clt);
+
+out:
+ return ERR_PTR(err);
+}
+EXPORT_SYMBOL(rtrs_clt_open);
+
+void rtrs_clt_close(struct rtrs_clt *clt)
+{
+ struct rtrs_clt_sess *sess, *tmp;
+
+ /* Firstly forbid sysfs access */
+ rtrs_clt_destroy_sysfs_root_files(clt);
+ rtrs_clt_destroy_sysfs_root_folders(clt);
+
+ /* Now it is safe to iterate over all paths without locks */
+ list_for_each_entry_safe(sess, tmp, &clt->paths_list, s.entry) {
+ rtrs_clt_destroy_sess_files(sess, NULL);
+ rtrs_clt_close_conns(sess, true);
+ free_sess(sess);
+ }
+ free_clt(clt);
+ module_put(THIS_MODULE);
+}
+EXPORT_SYMBOL(rtrs_clt_close);
+
+int rtrs_clt_reconnect_from_sysfs(struct rtrs_clt_sess *sess)
+{
+ enum rtrs_clt_state old_state;
+ int err = -EBUSY;
+ bool changed;
+
+ changed = rtrs_clt_change_state_get_old(sess, RTRS_CLT_RECONNECTING,
+ &old_state);
+ if (changed) {
+ sess->reconnect_attempts = 0;
+ queue_delayed_work(rtrs_wq, &sess->reconnect_dwork, 0);
+ }
+ if (changed || old_state == RTRS_CLT_RECONNECTING) {
+ /*
+ * flush_delayed_work() queues pending work for immediate
+ * execution, so do the flush if we have queued something
+ * right now or work is pending.
+ */
+ flush_delayed_work(&sess->reconnect_dwork);
+ err = (READ_ONCE(sess->state) ==
+ RTRS_CLT_CONNECTED ? 0 : -ENOTCONN);
+ }
+
+ return err;
+}
+
+int rtrs_clt_disconnect_from_sysfs(struct rtrs_clt_sess *sess)
+{
+ rtrs_clt_close_conns(sess, true);
+
+ return 0;
+}
+
+int rtrs_clt_remove_path_from_sysfs(struct rtrs_clt_sess *sess,
+ const struct attribute *sysfs_self)
+{
+ struct rtrs_clt *clt = sess->clt;
+ enum rtrs_clt_state old_state;
+ bool changed;
+
+ /*
+ * That can happen only when userspace tries to remove path
+ * very early, when rtrs_clt_open() is not yet finished.
+ */
+ if (unlikely(!clt->opened))
+ return -EBUSY;
+
+ /*
+ * Continue stopping path till state was changed to DEAD or
+ * state was observed as DEAD:
+ * 1. State was changed to DEAD - we were fast and nobody
+ * invoked rtrs_clt_reconnect(), which can again start
+ * reconnecting.
+ * 2. State was observed as DEAD - we have someone in parallel
+ * removing the path.
+ */
+ do {
+ rtrs_clt_close_conns(sess, true);
+ } while (!(changed = rtrs_clt_change_state_get_old(sess,
+ RTRS_CLT_DEAD,
+ &old_state)) &&
+ old_state != RTRS_CLT_DEAD);
+
+ /*
+ * If state was successfully changed to DEAD, commit suicide.
+ */
+ if (likely(changed)) {
+ rtrs_clt_destroy_sess_files(sess, sysfs_self);
+ rtrs_clt_remove_path_from_arr(sess);
+ free_sess(sess);
+ }
+
+ return 0;
+}
+
+void rtrs_clt_set_max_reconnect_attempts(struct rtrs_clt *clt, int value)
+{
+ clt->max_reconnect_attempts = (unsigned int)value;
+}
+
+int rtrs_clt_get_max_reconnect_attempts(const struct rtrs_clt *clt)
+{
+ return (int)clt->max_reconnect_attempts;
+}
+
+int rtrs_clt_request(int dir, rtrs_conf_fn *conf, struct rtrs_clt *clt,
+ struct rtrs_permit *permit, void *priv,
+ const struct kvec *vec, size_t nr, size_t data_len,
+ struct scatterlist *sg, unsigned int sg_cnt)
+{
+ struct rtrs_clt_io_req *req;
+ struct rtrs_clt_sess *sess;
+
+ enum dma_data_direction dma_dir;
+ int err = -ECONNABORTED, i;
+ size_t usr_len, hdr_len;
+ struct path_it it;
+
+ /* Get kvec length */
+ for (i = 0, usr_len = 0; i < nr; i++)
+ usr_len += vec[i].iov_len;
+
+ if (dir == READ) {
+ hdr_len = sizeof(struct rtrs_msg_rdma_read) +
+ sg_cnt * sizeof(struct rtrs_sg_desc);
+ dma_dir = DMA_FROM_DEVICE;
+ } else {
+ hdr_len = sizeof(struct rtrs_msg_rdma_write);
+ dma_dir = DMA_TO_DEVICE;
+ }
+
+ do_each_path(sess, clt, &it) {
+ if (unlikely(READ_ONCE(sess->state) != RTRS_CLT_CONNECTED))
+ continue;
+
+ if (unlikely(usr_len + hdr_len > sess->max_hdr_size)) {
+ rtrs_wrn_rl(sess->clt,
+ "%s request failed, user message size is %zu and header length %zu, but max size is %u\n",
+ dir == READ ? "Read" : "Write",
+ usr_len, hdr_len, sess->max_hdr_size);
+ err = -EMSGSIZE;
+ break;
+ }
+ req = rtrs_clt_get_req(sess, conf, permit, priv, vec, usr_len,
+ sg, sg_cnt, data_len, dma_dir);
+ if (dir == READ)
+ err = rtrs_clt_read_req(req);
+ else
+ err = rtrs_clt_write_req(req);
+ if (unlikely(err)) {
+ req->in_use = false;
+ continue;
+ }
+ /* Success path */
+ break;
+ } while_each_path(&it);
+
+ return err;
+}
+EXPORT_SYMBOL(rtrs_clt_request);
+
+int rtrs_clt_query(struct rtrs_clt *clt, struct rtrs_attrs *attr)
+{
+ if (unlikely(!rtrs_clt_is_connected(clt)))
+ return -ECOMM;
+
+ attr->queue_depth = clt->queue_depth;
+ attr->max_io_size = clt->max_io_size;
+ attr->sess_kobj = &clt->dev.kobj;
+ strlcpy(attr->sessname, clt->sessname, sizeof(attr->sessname));
+
+ return 0;
+}
+EXPORT_SYMBOL(rtrs_clt_query);
+
+int rtrs_clt_create_path_from_sysfs(struct rtrs_clt *clt,
+ struct rtrs_addr *addr)
+{
+ struct rtrs_clt_sess *sess;
+ int err;
+
+ sess = alloc_sess(clt, addr, nr_cons_per_session, clt->max_segments);
+ if (IS_ERR(sess))
+ return PTR_ERR(sess);
+
+ /*
+ * It is totally safe to add path in CONNECTING state: coming
+ * IO will never grab it. Also it is very important to add
+ * path before init, since init fires LINK_CONNECTED event.
+ */
+ rtrs_clt_add_path_to_arr(sess, addr);
+
+ err = init_sess(sess);
+ if (unlikely(err))
+ goto close_sess;
+
+ err = rtrs_clt_create_sess_files(sess);
+ if (unlikely(err))
+ goto close_sess;
+
+ return 0;
+
+close_sess:
+ rtrs_clt_remove_path_from_arr(sess);
+ rtrs_clt_close_conns(sess, true);
+ free_sess(sess);
+
+ return err;
+}
+
+static int check_module_params(void)
+{
+ if (nr_cons_per_session == 0)
+ nr_cons_per_session = min_t(unsigned int, nr_cpu_ids, U16_MAX);
+
+ return 0;
+}
+
+static int rtrs_clt_ib_dev_init(struct rtrs_ib_dev *dev)
+{
+ if (!(dev->ib_dev->attrs.device_cap_flags &
+ IB_DEVICE_MEM_MGT_EXTENSIONS)) {
+ pr_err("Memory registrations not supported.\n");
+ return -ENOTSUPP;
+ }
+
+ return 0;
+}
+
+static const struct rtrs_ib_dev_pool_ops dev_pool_ops = {
+ .init = rtrs_clt_ib_dev_init
+};
+
+static int __init rtrs_client_init(void)
+{
+ int err;
+
+ pr_info("Loading module %s, proto %s: (retry_cnt: %d, noreg_cnt: %d)\n",
+ KBUILD_MODNAME, RTRS_PROTO_VER_STRING,
+ retry_cnt, noreg_cnt);
+
+ rtrs_ib_dev_pool_init(noreg_cnt ? IB_PD_UNSAFE_GLOBAL_RKEY : 0,
+ &dev_pool);
+
+ err = check_module_params();
+ if (unlikely(err)) {
+ pr_err("Failed to load module, invalid module parameters, err: %d\n",
+ err);
+ return err;
+ }
+ rtrs_dev_class = class_create(THIS_MODULE, "rtrs-client");
+ if (IS_ERR(rtrs_dev_class)) {
+ pr_err("Failed to create rtrs-client dev class\n");
+ return PTR_ERR(rtrs_dev_class);
+ }
+ rtrs_wq = alloc_workqueue("rtrs_client_wq", WQ_MEM_RECLAIM, 0);
+ if (unlikely(!rtrs_wq)) {
+ pr_err("Failed to load module, alloc rtrs_client_wq failed\n");
+ class_destroy(rtrs_dev_class);
+ return -ENOMEM;
+ }
+
+ return 0;
+}
+
+static void __exit rtrs_client_exit(void)
+{
+ destroy_workqueue(rtrs_wq);
+ class_destroy(rtrs_dev_class);
+ rtrs_ib_dev_pool_deinit(&dev_pool);
+}
+
+module_init(rtrs_client_init);
+module_exit(rtrs_client_exit);
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 07/25] rtrs: client: statistics functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (5 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 06/25] rtrs: client: main functionality Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 08/25] rtrs: client: sysfs interface functions Jack Wang
` (19 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This introduces set of functions used on client side to account
statistics of RDMA data sent/received, amount of IOs inflight,
latency, cpu migrations, etc. Almost all statistics is collected
using percpu variables.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-clt-stats.c | 453 +++++++++++++++++++
1 file changed, 453 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-clt-stats.c
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-clt-stats.c b/drivers/infiniband/ulp/rtrs/rtrs-clt-stats.c
new file mode 100644
index 000000000000..421898f941a1
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-clt-stats.c
@@ -0,0 +1,453 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include "rtrs-clt.h"
+
+static inline int rtrs_clt_ms_to_id(unsigned long ms)
+{
+ int id = ms ? ilog2(ms) - MIN_LOG_LAT + 1 : 0;
+
+ return clamp(id, 0, LOG_LAT_SZ - 1);
+}
+
+void rtrs_clt_update_rdma_lat(struct rtrs_clt_stats *stats, bool read,
+ unsigned long ms)
+{
+ struct rtrs_clt_stats_pcpu *s;
+ int id;
+
+ id = rtrs_clt_ms_to_id(ms);
+ s = this_cpu_ptr(stats->pcpu_stats);
+ if (read) {
+ s->rdma_lat_distr[id].read++;
+ if (s->rdma_lat_max.read < ms)
+ s->rdma_lat_max.read = ms;
+ } else {
+ s->rdma_lat_distr[id].write++;
+ if (s->rdma_lat_max.write < ms)
+ s->rdma_lat_max.write = ms;
+ }
+}
+
+void rtrs_clt_decrease_inflight(struct rtrs_clt_stats *stats)
+{
+ atomic_dec(&stats->inflight);
+}
+
+void rtrs_clt_update_wc_stats(struct rtrs_clt_con *con)
+{
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ struct rtrs_clt_stats *stats = &sess->stats;
+ struct rtrs_clt_stats_pcpu *s;
+ int cpu;
+
+ cpu = raw_smp_processor_id();
+ s = this_cpu_ptr(stats->pcpu_stats);
+ s->wc_comp.cnt++;
+ s->wc_comp.total_cnt++;
+ if (unlikely(con->cpu != cpu)) {
+ s->cpu_migr.to++;
+
+ /* Careful here, override s pointer */
+ s = per_cpu_ptr(stats->pcpu_stats, con->cpu);
+ atomic_inc(&s->cpu_migr.from);
+ }
+}
+
+void rtrs_clt_inc_failover_cnt(struct rtrs_clt_stats *stats)
+{
+ struct rtrs_clt_stats_pcpu *s;
+
+ s = this_cpu_ptr(stats->pcpu_stats);
+ s->rdma.failover_cnt++;
+}
+
+static inline u32 rtrs_clt_stats_get_avg_wc_cnt(struct rtrs_clt_stats *stats)
+{
+ u32 cnt = 0;
+ u64 sum = 0;
+ int cpu;
+
+ for_each_possible_cpu(cpu) {
+ struct rtrs_clt_stats_pcpu *s;
+
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ sum += s->wc_comp.total_cnt;
+ cnt += s->wc_comp.cnt;
+ }
+
+ return cnt ? sum / cnt : 0;
+}
+
+int rtrs_clt_stats_wc_completion_to_str(struct rtrs_clt_stats *stats,
+ char *buf, size_t len)
+{
+ return scnprintf(buf, len, "%u\n",
+ rtrs_clt_stats_get_avg_wc_cnt(stats));
+}
+
+ssize_t rtrs_clt_stats_rdma_lat_distr_to_str(struct rtrs_clt_stats *stats,
+ char *page, size_t len)
+{
+ struct rtrs_clt_stats_rdma_lat res[LOG_LAT_SZ];
+ struct rtrs_clt_stats_rdma_lat max;
+ struct rtrs_clt_stats_pcpu *s;
+
+ ssize_t cnt = 0;
+ int i, cpu;
+
+ max.write = 0;
+ max.read = 0;
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+
+ if (max.write < s->rdma_lat_max.write)
+ max.write = s->rdma_lat_max.write;
+ if (max.read < s->rdma_lat_max.read)
+ max.read = s->rdma_lat_max.read;
+ }
+ for (i = 0; i < ARRAY_SIZE(res); i++) {
+ res[i].write = 0;
+ res[i].read = 0;
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+
+ res[i].write += s->rdma_lat_distr[i].write;
+ res[i].read += s->rdma_lat_distr[i].read;
+ }
+ }
+
+ for (i = 0; i < ARRAY_SIZE(res) - 1; i++)
+ cnt += scnprintf(page + cnt, len - cnt,
+ "< %6d ms: %llu %llu\n",
+ 1 << (i + MIN_LOG_LAT), res[i].read,
+ res[i].write);
+ cnt += scnprintf(page + cnt, len - cnt, ">= %5d ms: %llu %llu\n",
+ 1 << (i - 1 + MIN_LOG_LAT), res[i].read,
+ res[i].write);
+ cnt += scnprintf(page + cnt, len - cnt, " maximum ms: %llu %llu\n",
+ max.read, max.write);
+
+ return cnt;
+}
+
+int rtrs_clt_stats_migration_cnt_to_str(struct rtrs_clt_stats *stats,
+ char *buf, size_t len)
+{
+ struct rtrs_clt_stats_pcpu *s;
+
+ size_t used;
+ int cpu;
+
+ used = scnprintf(buf, len, " ");
+ for_each_possible_cpu(cpu)
+ used += scnprintf(buf + used, len - used, " CPU%u", cpu);
+
+ used += scnprintf(buf + used, len - used, "\nfrom:");
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ used += scnprintf(buf + used, len - used, " %d",
+ atomic_read(&s->cpu_migr.from));
+ }
+
+ used += scnprintf(buf + used, len - used, "\nto :");
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ used += scnprintf(buf + used, len - used, " %d",
+ s->cpu_migr.to);
+ }
+ used += scnprintf(buf + used, len - used, "\n");
+
+ return used;
+}
+
+int rtrs_clt_stats_reconnects_to_str(struct rtrs_clt_stats *stats, char *buf,
+ size_t len)
+{
+ return scnprintf(buf, len, "%d %d\n",
+ stats->reconnects.successful_cnt,
+ stats->reconnects.fail_cnt);
+}
+
+ssize_t rtrs_clt_stats_rdma_to_str(struct rtrs_clt_stats *stats,
+ char *page, size_t len)
+{
+ struct rtrs_clt_stats_rdma sum;
+ struct rtrs_clt_stats_rdma *r;
+ int cpu;
+
+ memset(&sum, 0, sizeof(sum));
+
+ for_each_possible_cpu(cpu) {
+ r = &per_cpu_ptr(stats->pcpu_stats, cpu)->rdma;
+
+ sum.dir[READ].cnt += r->dir[READ].cnt;
+ sum.dir[READ].size_total += r->dir[READ].size_total;
+ sum.dir[WRITE].cnt += r->dir[WRITE].cnt;
+ sum.dir[WRITE].size_total += r->dir[WRITE].size_total;
+ sum.failover_cnt += r->failover_cnt;
+ }
+
+ return scnprintf(page, len, "%llu %llu %llu %llu %u %llu\n",
+ sum.dir[READ].cnt, sum.dir[READ].size_total,
+ sum.dir[WRITE].cnt, sum.dir[WRITE].size_total,
+ atomic_read(&stats->inflight), sum.failover_cnt);
+}
+
+int rtrs_clt_stats_sg_list_distr_to_str(struct rtrs_clt_stats *stats,
+ char *buf, size_t len)
+{
+ struct rtrs_clt_stats_pcpu *s;
+
+ int i, cpu, cnt;
+
+ cnt = scnprintf(buf, len, "n\\cpu:");
+ for_each_possible_cpu(cpu)
+ cnt += scnprintf(buf + cnt, len - cnt, "%5d", cpu);
+
+ for (i = 0; i < SG_DISTR_SZ; i++) {
+ if (i <= MAX_LIN_SG)
+ cnt += scnprintf(buf + cnt, len - cnt, "\n= %3d:", i);
+ else if (i < SG_DISTR_SZ - 1)
+ cnt += scnprintf(buf + cnt, len - cnt, "\n< %3d:",
+ 1 << (i + MIN_LOG_SG - MAX_LIN_SG));
+ else
+ cnt += scnprintf(buf + cnt, len - cnt, "\n>=%3d:",
+ 1 << (i + MIN_LOG_SG -
+ MAX_LIN_SG - 1));
+
+ for_each_possible_cpu(cpu) {
+ unsigned int p, p_i, p_f;
+ u64 total, distr;
+
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ total = s->sg_list_total;
+ distr = s->sg_list_distr[i];
+
+ p = total ? distr * 1000 / total : 0;
+ p_i = p / 10;
+ p_f = p % 10;
+
+ if (distr)
+ cnt += scnprintf(buf + cnt, len - cnt,
+ " %2u.%01u", p_i, p_f);
+ else
+ cnt += scnprintf(buf + cnt, len - cnt, " 0");
+ }
+ }
+
+ cnt += scnprintf(buf + cnt, len - cnt, "\ntotal:");
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ cnt += scnprintf(buf + cnt, len - cnt, " %llu",
+ s->sg_list_total);
+ }
+ cnt += scnprintf(buf + cnt, len - cnt, "\n");
+
+ return cnt;
+}
+
+ssize_t rtrs_clt_reset_all_help(struct rtrs_clt_stats *s,
+ char *page, size_t len)
+{
+ return scnprintf(page, len, "echo 1 to reset all statistics\n");
+}
+
+int rtrs_clt_reset_rdma_stats(struct rtrs_clt_stats *stats, bool enable)
+{
+ struct rtrs_clt_stats_pcpu *s;
+ int cpu;
+
+ if (unlikely(!enable))
+ return -EINVAL;
+
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ memset(&s->rdma, 0, sizeof(s->rdma));
+ }
+
+ return 0;
+}
+
+int rtrs_clt_reset_rdma_lat_distr_stats(struct rtrs_clt_stats *stats,
+ bool enable)
+{
+ struct rtrs_clt_stats_pcpu *s;
+ int cpu;
+
+ if (enable) {
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ memset(&s->rdma_lat_max, 0, sizeof(s->rdma_lat_max));
+ memset(&s->rdma_lat_distr, 0,
+ sizeof(s->rdma_lat_distr));
+ }
+ }
+ stats->enable_rdma_lat = enable;
+
+ return 0;
+}
+
+int rtrs_clt_reset_sg_list_distr_stats(struct rtrs_clt_stats *stats,
+ bool enable)
+{
+ struct rtrs_clt_stats_pcpu *s;
+ int cpu;
+
+ if (unlikely(!enable))
+ return -EINVAL;
+
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ memset(&s->sg_list_total, 0, sizeof(s->sg_list_total));
+ memset(&s->sg_list_distr, 0, sizeof(s->sg_list_distr));
+ }
+
+ return 0;
+}
+
+int rtrs_clt_reset_cpu_migr_stats(struct rtrs_clt_stats *stats, bool enable)
+{
+ struct rtrs_clt_stats_pcpu *s;
+ int cpu;
+
+ if (unlikely(!enable))
+ return -EINVAL;
+
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ memset(&s->cpu_migr, 0, sizeof(s->cpu_migr));
+ }
+
+ return 0;
+}
+
+int rtrs_clt_reset_reconnects_stat(struct rtrs_clt_stats *stats, bool enable)
+{
+ if (unlikely(!enable))
+ return -EINVAL;
+
+ memset(&stats->reconnects, 0, sizeof(stats->reconnects));
+
+ return 0;
+}
+
+int rtrs_clt_reset_wc_comp_stats(struct rtrs_clt_stats *stats, bool enable)
+{
+ struct rtrs_clt_stats_pcpu *s;
+ int cpu;
+
+ if (unlikely(!enable))
+ return -EINVAL;
+
+ for_each_possible_cpu(cpu) {
+ s = per_cpu_ptr(stats->pcpu_stats, cpu);
+ memset(&s->wc_comp, 0, sizeof(s->wc_comp));
+ }
+
+ return 0;
+}
+
+int rtrs_clt_reset_all_stats(struct rtrs_clt_stats *s, bool enable)
+{
+ if (enable) {
+ rtrs_clt_reset_rdma_stats(s, enable);
+ rtrs_clt_reset_rdma_lat_distr_stats(s, enable);
+ rtrs_clt_reset_sg_list_distr_stats(s, enable);
+ rtrs_clt_reset_cpu_migr_stats(s, enable);
+ rtrs_clt_reset_reconnects_stat(s, enable);
+ rtrs_clt_reset_wc_comp_stats(s, enable);
+ atomic_set(&s->inflight, 0);
+ return 0;
+ }
+
+ return -EINVAL;
+}
+
+static inline void rtrs_clt_record_sg_distr(u64 stat[SG_DISTR_SZ], u64 *total,
+ unsigned int cnt)
+{
+ int i;
+
+ i = cnt > MAX_LIN_SG ? ilog2(cnt) + MAX_LIN_SG - MIN_LOG_SG + 1 : cnt;
+ i = i < SG_DISTR_SZ ? i : SG_DISTR_SZ - 1;
+
+ stat[i]++;
+ (*total)++;
+}
+
+static inline void rtrs_clt_update_rdma_stats(struct rtrs_clt_stats *stats,
+ size_t size, int d)
+{
+ struct rtrs_clt_stats_pcpu *s;
+
+ s = this_cpu_ptr(stats->pcpu_stats);
+ s->rdma.dir[d].cnt++;
+ s->rdma.dir[d].size_total += size;
+}
+
+void rtrs_clt_update_all_stats(struct rtrs_clt_io_req *req, int dir)
+{
+ struct rtrs_clt_con *con = req->con;
+ struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
+ struct rtrs_clt_stats *stats = &sess->stats;
+ unsigned int len;
+
+ struct rtrs_clt_stats_pcpu *s;
+
+ s = this_cpu_ptr(stats->pcpu_stats);
+ rtrs_clt_record_sg_distr(s->sg_list_distr, &s->sg_list_total,
+ req->sg_cnt);
+ len = req->usr_len + req->data_len;
+ rtrs_clt_update_rdma_stats(stats, len, dir);
+ atomic_inc(&stats->inflight);
+}
+
+int rtrs_clt_init_stats(struct rtrs_clt_stats *stats)
+{
+ stats->enable_rdma_lat = false;
+ stats->pcpu_stats = alloc_percpu(typeof(*stats->pcpu_stats));
+ if (unlikely(!stats->pcpu_stats))
+ return -ENOMEM;
+
+ /*
+ * successful_cnt will be set to 0 after session
+ * is established for the first time
+ */
+ stats->reconnects.successful_cnt = -1;
+
+ return 0;
+}
+
+void rtrs_clt_free_stats(struct rtrs_clt_stats *stats)
+{
+ free_percpu(stats->pcpu_stats);
+}
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 08/25] rtrs: client: sysfs interface functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (6 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 07/25] rtrs: client: statistics functions Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 09/25] rtrs: server: private header with server structs and functions Jack Wang
` (18 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is the sysfs interface to rtrs sessions on client side:
/sys/devices/virtual/rtrs-client/<SESS-NAME>/
*** rtrs session created by rtrs_clt_open() API call
|
|- max_reconnect_attempts
| *** number of reconnect attempts for session
|
|- add_path
| *** adds another connection path into rtrs session
|
|- paths/<SRC@DST>/
*** established paths to server in a session
|
|- disconnect
| *** disconnect path
|
|- reconnect
| *** reconnect path
|
|- remove_path
| *** remove current path
|
|- state
| *** retrieve current path state
|
|- hca_port
| *** HCA port number
|
|- hca_name
| *** HCA name
|
|- stats/
*** current path statistics
|
|- cpu_migration
|- rdma
|- rdma_lat
|- reconnects
|- reset_all
|- sg_entries
|- wc_completions
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-clt-sysfs.c | 519 +++++++++++++++++++
1 file changed, 519 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-clt-sysfs.c
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-clt-sysfs.c b/drivers/infiniband/ulp/rtrs/rtrs-clt-sysfs.c
new file mode 100644
index 000000000000..bc2cb7638e3f
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-clt-sysfs.c
@@ -0,0 +1,519 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include "rtrs-pri.h"
+#include "rtrs-clt.h"
+#include "rtrs-log.h"
+
+#define MIN_MAX_RECONN_ATT -1
+#define MAX_MAX_RECONN_ATT 9999
+
+static struct kobj_type ktype = {
+ .sysfs_ops = &kobj_sysfs_ops,
+};
+
+static ssize_t max_reconnect_attempts_show(struct device *dev,
+ struct device_attribute *attr,
+ char *page)
+{
+ struct rtrs_clt *clt;
+
+ clt = container_of(dev, struct rtrs_clt, dev);
+
+ return sprintf(page, "%d\n", rtrs_clt_get_max_reconnect_attempts(clt));
+}
+
+static ssize_t max_reconnect_attempts_store(struct device *dev,
+ struct device_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ struct rtrs_clt *clt;
+ int value;
+ int ret;
+
+ clt = container_of(dev, struct rtrs_clt, dev);
+
+ ret = kstrtoint(buf, 10, &value);
+ if (unlikely(ret)) {
+ rtrs_err(clt, "%s: failed to convert string '%s' to int\n",
+ attr->attr.name, buf);
+ return ret;
+ }
+ if (unlikely(value > MAX_MAX_RECONN_ATT ||
+ value < MIN_MAX_RECONN_ATT)) {
+ rtrs_err(clt,
+ "%s: invalid range (provided: '%s', accepted: min: %d, max: %d)\n",
+ attr->attr.name, buf, MIN_MAX_RECONN_ATT,
+ MAX_MAX_RECONN_ATT);
+ return -EINVAL;
+ }
+ rtrs_clt_set_max_reconnect_attempts(clt, value);
+
+ return count;
+}
+
+static DEVICE_ATTR_RW(max_reconnect_attempts);
+
+static ssize_t mpath_policy_show(struct device *dev,
+ struct device_attribute *attr,
+ char *page)
+{
+ struct rtrs_clt *clt;
+
+ clt = container_of(dev, struct rtrs_clt, dev);
+
+ switch (clt->mp_policy) {
+ case MP_POLICY_RR:
+ return sprintf(page, "round-robin (RR: %d)\n", clt->mp_policy);
+ case MP_POLICY_MIN_INFLIGHT:
+ return sprintf(page, "min-inflight (MI: %d)\n", clt->mp_policy);
+ default:
+ return sprintf(page, "Unknown (%d)\n", clt->mp_policy);
+ }
+}
+
+static ssize_t mpath_policy_store(struct device *dev,
+ struct device_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ struct rtrs_clt *clt;
+ int value;
+ int ret;
+
+ clt = container_of(dev, struct rtrs_clt, dev);
+
+ ret = kstrtoint(buf, 10, &value);
+ if (!ret && (value == MP_POLICY_RR ||
+ value == MP_POLICY_MIN_INFLIGHT)) {
+ clt->mp_policy = value;
+ return count;
+ }
+
+ if (!strncasecmp(buf, "round-robin", 11) ||
+ !strncasecmp(buf, "rr", 2))
+ clt->mp_policy = MP_POLICY_RR;
+ else if (!strncasecmp(buf, "min-inflight", 12) ||
+ !strncasecmp(buf, "mi", 2))
+ clt->mp_policy = MP_POLICY_MIN_INFLIGHT;
+ else
+ return -EINVAL;
+
+ return count;
+}
+
+static DEVICE_ATTR_RW(mpath_policy);
+
+static ssize_t add_path_show(struct device *dev,
+ struct device_attribute *attr, char *page)
+{
+ return scnprintf(page, PAGE_SIZE,
+ "Usage: echo [<source addr>@]<destination addr> > %s\n\n*addr ::= [ ip:<ipv4|ipv6> | gid:<gid> ]\n",
+ attr->attr.name);
+}
+
+static ssize_t add_path_store(struct device *dev,
+ struct device_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct sockaddr_storage srcaddr, dstaddr;
+ struct rtrs_addr addr = {
+ .src = &srcaddr,
+ .dst = &dstaddr
+ };
+ struct rtrs_clt *clt;
+ const char *nl;
+ size_t len;
+ int err;
+
+ clt = container_of(dev, struct rtrs_clt, dev);
+
+ nl = strchr(buf, '\n');
+ if (nl)
+ len = nl - buf;
+ else
+ len = count;
+ err = rtrs_addr_to_sockaddr(buf, len, clt->port, &addr);
+ if (unlikely(err))
+ return -EINVAL;
+
+ err = rtrs_clt_create_path_from_sysfs(clt, &addr);
+ if (unlikely(err))
+ return err;
+
+ return count;
+}
+
+static DEVICE_ATTR_RW(add_path);
+
+static ssize_t rtrs_clt_state_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *page)
+{
+ struct rtrs_clt_sess *sess;
+
+ sess = container_of(kobj, struct rtrs_clt_sess, kobj);
+ if (sess->state == RTRS_CLT_CONNECTED)
+ return sprintf(page, "connected\n");
+
+ return sprintf(page, "disconnected\n");
+}
+
+static struct kobj_attribute rtrs_clt_state_attr =
+ __ATTR(state, 0444, rtrs_clt_state_show, NULL);
+
+static ssize_t rtrs_clt_reconnect_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ return scnprintf(page, PAGE_SIZE, "Usage: echo 1 > %s\n",
+ attr->attr.name);
+}
+
+static ssize_t rtrs_clt_reconnect_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct rtrs_clt_sess *sess;
+ int ret;
+
+ sess = container_of(kobj, struct rtrs_clt_sess, kobj);
+ if (!sysfs_streq(buf, "1")) {
+ rtrs_err(sess->clt, "%s: unknown value: '%s'\n",
+ attr->attr.name, buf);
+ return -EINVAL;
+ }
+ ret = rtrs_clt_reconnect_from_sysfs(sess);
+ if (unlikely(ret))
+ return ret;
+
+ return count;
+}
+
+static struct kobj_attribute rtrs_clt_reconnect_attr =
+ __ATTR(reconnect, 0644, rtrs_clt_reconnect_show,
+ rtrs_clt_reconnect_store);
+
+static ssize_t rtrs_clt_disconnect_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ return scnprintf(page, PAGE_SIZE, "Usage: echo 1 > %s\n",
+ attr->attr.name);
+}
+
+static ssize_t rtrs_clt_disconnect_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct rtrs_clt_sess *sess;
+ int ret;
+
+ sess = container_of(kobj, struct rtrs_clt_sess, kobj);
+ if (!sysfs_streq(buf, "1")) {
+ rtrs_err(sess->clt, "%s: unknown value: '%s'\n",
+ attr->attr.name, buf);
+ return -EINVAL;
+ }
+ ret = rtrs_clt_disconnect_from_sysfs(sess);
+ if (unlikely(ret))
+ return ret;
+
+ return count;
+}
+
+static struct kobj_attribute rtrs_clt_disconnect_attr =
+ __ATTR(disconnect, 0644, rtrs_clt_disconnect_show,
+ rtrs_clt_disconnect_store);
+
+static ssize_t rtrs_clt_remove_path_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ return scnprintf(page, PAGE_SIZE, "Usage: echo 1 > %s\n",
+ attr->attr.name);
+}
+
+static ssize_t rtrs_clt_remove_path_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct rtrs_clt_sess *sess;
+ int ret;
+
+ sess = container_of(kobj, struct rtrs_clt_sess, kobj);
+ if (!sysfs_streq(buf, "1")) {
+ rtrs_err(sess->clt, "%s: unknown value: '%s'\n",
+ attr->attr.name, buf);
+ return -EINVAL;
+ }
+ ret = rtrs_clt_remove_path_from_sysfs(sess, &attr->attr);
+ if (unlikely(ret))
+ return ret;
+
+ return count;
+}
+
+static struct kobj_attribute rtrs_clt_remove_path_attr =
+ __ATTR(remove_path, 0644, rtrs_clt_remove_path_show,
+ rtrs_clt_remove_path_store);
+
+STAT_ATTR(struct rtrs_clt_sess, cpu_migration,
+ rtrs_clt_stats_migration_cnt_to_str,
+ rtrs_clt_reset_cpu_migr_stats);
+
+STAT_ATTR(struct rtrs_clt_sess, sg_entries,
+ rtrs_clt_stats_sg_list_distr_to_str,
+ rtrs_clt_reset_sg_list_distr_stats);
+
+STAT_ATTR(struct rtrs_clt_sess, reconnects,
+ rtrs_clt_stats_reconnects_to_str,
+ rtrs_clt_reset_reconnects_stat);
+
+STAT_ATTR(struct rtrs_clt_sess, rdma_lat,
+ rtrs_clt_stats_rdma_lat_distr_to_str,
+ rtrs_clt_reset_rdma_lat_distr_stats);
+
+STAT_ATTR(struct rtrs_clt_sess, wc_completion,
+ rtrs_clt_stats_wc_completion_to_str,
+ rtrs_clt_reset_wc_comp_stats);
+
+STAT_ATTR(struct rtrs_clt_sess, rdma,
+ rtrs_clt_stats_rdma_to_str,
+ rtrs_clt_reset_rdma_stats);
+
+STAT_ATTR(struct rtrs_clt_sess, reset_all,
+ rtrs_clt_reset_all_help,
+ rtrs_clt_reset_all_stats);
+
+static struct attribute *rtrs_clt_stats_attrs[] = {
+ &sg_entries_attr.attr,
+ &cpu_migration_attr.attr,
+ &reconnects_attr.attr,
+ &rdma_lat_attr.attr,
+ &wc_completion_attr.attr,
+ &rdma_attr.attr,
+ &reset_all_attr.attr,
+ NULL,
+};
+
+static struct attribute_group rtrs_clt_stats_attr_group = {
+ .attrs = rtrs_clt_stats_attrs,
+};
+
+static int rtrs_clt_create_stats_files(struct kobject *kobj,
+ struct kobject *kobj_stats)
+{
+ int ret;
+
+ ret = kobject_init_and_add(kobj_stats, &ktype, kobj, "stats");
+ if (ret) {
+ pr_err("Failed to init and add stats kobject, err: %d\n",
+ ret);
+ return ret;
+ }
+
+ ret = sysfs_create_group(kobj_stats, &rtrs_clt_stats_attr_group);
+ if (ret) {
+ pr_err("failed to create stats sysfs group, err: %d\n",
+ ret);
+ goto err;
+ }
+
+ return 0;
+
+err:
+ kobject_del(kobj_stats);
+ kobject_put(kobj_stats);
+
+ return ret;
+}
+
+static ssize_t rtrs_clt_hca_port_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rtrs_clt_sess *sess;
+
+ sess = container_of(kobj, typeof(*sess), kobj);
+
+ return scnprintf(page, PAGE_SIZE, "%u\n", sess->hca_port);
+}
+
+static struct kobj_attribute rtrs_clt_hca_port_attr =
+ __ATTR(hca_port, 0444, rtrs_clt_hca_port_show, NULL);
+
+static ssize_t rtrs_clt_hca_name_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rtrs_clt_sess *sess;
+
+ sess = container_of(kobj, struct rtrs_clt_sess, kobj);
+
+ return scnprintf(page, PAGE_SIZE, "%s\n", sess->hca_name);
+}
+
+static struct kobj_attribute rtrs_clt_hca_name_attr =
+ __ATTR(hca_name, 0444, rtrs_clt_hca_name_show, NULL);
+
+static ssize_t rtrs_clt_src_addr_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rtrs_clt_sess *sess;
+ int cnt;
+
+ sess = container_of(kobj, struct rtrs_clt_sess, kobj);
+ cnt = sockaddr_to_str((struct sockaddr *)&sess->s.src_addr,
+ page, PAGE_SIZE);
+ return cnt + scnprintf(page + cnt, PAGE_SIZE - cnt, "\n");
+}
+
+static struct kobj_attribute rtrs_clt_src_addr_attr =
+ __ATTR(src_addr, 0444, rtrs_clt_src_addr_show, NULL);
+
+static ssize_t rtrs_clt_dst_addr_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rtrs_clt_sess *sess;
+ int cnt;
+
+ sess = container_of(kobj, struct rtrs_clt_sess, kobj);
+ cnt = sockaddr_to_str((struct sockaddr *)&sess->s.dst_addr,
+ page, PAGE_SIZE);
+ return cnt + scnprintf(page + cnt, PAGE_SIZE - cnt, "\n");
+}
+
+static struct kobj_attribute rtrs_clt_dst_addr_attr =
+ __ATTR(dst_addr, 0444, rtrs_clt_dst_addr_show, NULL);
+
+static struct attribute *rtrs_clt_sess_attrs[] = {
+ &rtrs_clt_hca_name_attr.attr,
+ &rtrs_clt_hca_port_attr.attr,
+ &rtrs_clt_src_addr_attr.attr,
+ &rtrs_clt_dst_addr_attr.attr,
+ &rtrs_clt_state_attr.attr,
+ &rtrs_clt_reconnect_attr.attr,
+ &rtrs_clt_disconnect_attr.attr,
+ &rtrs_clt_remove_path_attr.attr,
+ NULL,
+};
+
+static struct attribute_group rtrs_clt_sess_attr_group = {
+ .attrs = rtrs_clt_sess_attrs,
+};
+
+int rtrs_clt_create_sess_files(struct rtrs_clt_sess *sess)
+{
+ struct rtrs_clt *clt = sess->clt;
+ char str[NAME_MAX];
+ int err, cnt;
+
+ cnt = sockaddr_to_str((struct sockaddr *)&sess->s.src_addr,
+ str, sizeof(str));
+ cnt += scnprintf(str + cnt, sizeof(str) - cnt, "@");
+ sockaddr_to_str((struct sockaddr *)&sess->s.dst_addr,
+ str + cnt, sizeof(str) - cnt);
+
+ err = kobject_init_and_add(&sess->kobj, &ktype, &clt->kobj_paths,
+ "%s", str);
+ if (unlikely(err)) {
+ pr_err("kobject_init_and_add: %d\n", err);
+ return err;
+ }
+ err = sysfs_create_group(&sess->kobj, &rtrs_clt_sess_attr_group);
+ if (unlikely(err)) {
+ pr_err("sysfs_create_group(): %d\n", err);
+ goto put_kobj;
+ }
+ err = rtrs_clt_create_stats_files(&sess->kobj, &sess->kobj_stats);
+ if (unlikely(err))
+ goto put_kobj;
+
+ return 0;
+
+put_kobj:
+ kobject_del(&sess->kobj);
+ kobject_put(&sess->kobj);
+
+ return err;
+}
+
+void rtrs_clt_destroy_sess_files(struct rtrs_clt_sess *sess,
+ const struct attribute *sysfs_self)
+{
+ if (sess->kobj.state_in_sysfs) {
+ kobject_del(&sess->kobj_stats);
+ kobject_put(&sess->kobj_stats);
+ if (sysfs_self)
+ /* To avoid deadlock firstly commit suicide */
+ sysfs_remove_file_self(&sess->kobj, sysfs_self);
+ kobject_del(&sess->kobj);
+ kobject_put(&sess->kobj);
+ }
+}
+
+static struct attribute *rtrs_clt_attrs[] = {
+ &dev_attr_max_reconnect_attempts.attr,
+ &dev_attr_mpath_policy.attr,
+ &dev_attr_add_path.attr,
+ NULL,
+};
+
+static struct attribute_group rtrs_clt_attr_group = {
+ .attrs = rtrs_clt_attrs,
+};
+
+int rtrs_clt_create_sysfs_root_folders(struct rtrs_clt *clt)
+{
+ return kobject_init_and_add(&clt->kobj_paths, &ktype,
+ &clt->dev.kobj, "paths");
+}
+
+int rtrs_clt_create_sysfs_root_files(struct rtrs_clt *clt)
+{
+ return sysfs_create_group(&clt->dev.kobj, &rtrs_clt_attr_group);
+}
+
+void rtrs_clt_destroy_sysfs_root_folders(struct rtrs_clt *clt)
+{
+ if (clt->kobj_paths.state_in_sysfs) {
+ kobject_del(&clt->kobj_paths);
+ kobject_put(&clt->kobj_paths);
+ }
+}
+
+void rtrs_clt_destroy_sysfs_root_files(struct rtrs_clt *clt)
+{
+ sysfs_remove_group(&clt->dev.kobj, &rtrs_clt_attr_group);
+}
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 09/25] rtrs: server: private header with server structs and functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (7 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 08/25] rtrs: client: sysfs interface functions Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 10/25] rtrs: server: main functionality Jack Wang
` (17 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This header describes main structs and functions used by rtrs-server
module, mainly for accepting rtrs sessions, creating/destroying
sysfs entries, accounting statistics on server side.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-srv.h | 159 +++++++++++++++++++++++++
1 file changed, 159 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-srv.h
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-srv.h b/drivers/infiniband/ulp/rtrs/rtrs-srv.h
new file mode 100644
index 000000000000..5124140652e4
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-srv.h
@@ -0,0 +1,159 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Swapnil Ingle <swapnil.ingle@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+
+#ifndef RTRS_SRV_H
+#define RTRS_SRV_H
+
+#include <linux/device.h>
+#include <linux/refcount.h>
+#include "rtrs-pri.h"
+
+/**
+ * enum rtrs_srv_state - Server states.
+ */
+enum rtrs_srv_state {
+ RTRS_SRV_CONNECTING,
+ RTRS_SRV_CONNECTED,
+ RTRS_SRV_CLOSING,
+ RTRS_SRV_CLOSED,
+};
+
+struct rtrs_stats_wc_comp {
+ atomic64_t calls;
+ atomic64_t total_wc_cnt;
+};
+
+struct rtrs_srv_stats_rdma_stats {
+ struct {
+ atomic64_t cnt;
+ atomic64_t size_total;
+ } dir[2];
+};
+
+struct rtrs_srv_stats {
+ struct rtrs_srv_stats_rdma_stats rdma_stats;
+ struct rtrs_stats_wc_comp wc_comp;
+};
+
+struct rtrs_srv_con {
+ struct rtrs_con c;
+ atomic_t wr_cnt;
+};
+
+struct rtrs_srv_op {
+ struct rtrs_srv_con *con;
+ u32 msg_id;
+ u8 dir;
+ struct rtrs_msg_rdma_read *rd_msg;
+ struct ib_rdma_wr *tx_wr;
+ struct ib_sge *tx_sg;
+};
+
+struct rtrs_srv_mr {
+ struct ib_mr *mr;
+ struct sg_table sgt;
+ struct ib_cqe inv_cqe; /* only for always_invalidate=true */
+ u32 msg_id; /* only for always_invalidate=true */
+ u32 msg_off; /* only for always_invalidate=true */
+ struct rtrs_iu *iu; /* send buffer for new rkey msg */
+};
+
+struct rtrs_srv_sess {
+ struct rtrs_sess s;
+ struct rtrs_srv *srv;
+ struct work_struct close_work;
+ enum rtrs_srv_state state;
+ spinlock_t state_lock;
+ int cur_cq_vector;
+ struct rtrs_srv_op **ops_ids;
+ atomic_t ids_inflight;
+ wait_queue_head_t ids_waitq;
+ struct rtrs_srv_mr *mrs;
+ unsigned int mrs_num;
+ dma_addr_t *dma_addr;
+ bool established;
+ unsigned int mem_bits;
+ struct kobject kobj;
+ struct kobject kobj_stats;
+ struct rtrs_srv_stats stats;
+};
+
+struct rtrs_srv {
+ struct list_head paths_list;
+ int paths_up;
+ struct mutex paths_ev_mutex;
+ size_t paths_num;
+ struct mutex paths_mutex;
+ uuid_t paths_uuid;
+ refcount_t refcount;
+ struct rtrs_srv_ctx *ctx;
+ struct list_head ctx_list;
+ void *priv;
+ size_t queue_depth;
+ struct page **chunks;
+ struct device dev;
+ unsigned int dev_ref;
+ struct kobject kobj_paths;
+};
+
+struct rtrs_srv_ctx {
+ rdma_ev_fn *rdma_ev;
+ link_ev_fn *link_ev;
+ struct rdma_cm_id *cm_id_ip;
+ struct rdma_cm_id *cm_id_ib;
+ struct mutex srv_mutex;
+ struct list_head srv_list;
+};
+
+extern struct class *rtrs_dev_class;
+
+void close_sess(struct rtrs_srv_sess *sess);
+
+/* rtrs-srv-stats.c */
+
+void rtrs_srv_update_rdma_stats(struct rtrs_srv_stats *s, size_t size, int d);
+void rtrs_srv_update_wc_stats(struct rtrs_srv_stats *s);
+
+int rtrs_srv_reset_rdma_stats(struct rtrs_srv_stats *stats, bool enable);
+ssize_t rtrs_srv_stats_rdma_to_str(struct rtrs_srv_stats *stats,
+ char *page, size_t len);
+int rtrs_srv_reset_wc_completion_stats(struct rtrs_srv_stats *stats,
+ bool enable);
+int rtrs_srv_stats_wc_completion_to_str(struct rtrs_srv_stats *stats, char *buf,
+ size_t len);
+int rtrs_srv_reset_all_stats(struct rtrs_srv_stats *stats, bool enable);
+ssize_t rtrs_srv_reset_all_help(struct rtrs_srv_stats *stats,
+ char *page, size_t len);
+
+/* rtrs-srv-sysfs.c */
+
+int rtrs_srv_create_sess_files(struct rtrs_srv_sess *sess);
+void rtrs_srv_destroy_sess_files(struct rtrs_srv_sess *sess);
+
+#endif /* RTRS_SRV_H */
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 10/25] rtrs: server: main functionality
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (8 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 09/25] rtrs: server: private header with server structs and functions Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 11/25] rtrs: server: statistics functions Jack Wang
` (16 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is main functionality of rtrs-server module, which accepts
set of RDMA connections (so called rtrs session), creates/destroys
sysfs entries associated with rtrs session and notifies upper layer
(user of RTRS API) about RDMA requests or link events.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-srv.c | 2187 ++++++++++++++++++++++++
1 file changed, 2187 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-srv.c
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-srv.c b/drivers/infiniband/ulp/rtrs/rtrs-srv.c
new file mode 100644
index 000000000000..24b69379022c
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-srv.c
@@ -0,0 +1,2187 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Swapnil Ingle <swapnil.ingle@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include <linux/module.h>
+#include <linux/mempool.h>
+
+#include "rtrs-srv.h"
+#include "rtrs-log.h"
+
+MODULE_DESCRIPTION("RTRS Server");
+MODULE_LICENSE("GPL");
+
+/* Must be power of 2, see mask from mr->page_size in ib_sg_to_pages() */
+#define DEFAULT_MAX_CHUNK_SIZE (128 << 10)
+#define DEFAULT_SESS_QUEUE_DEPTH 512
+#define MAX_HDR_SIZE PAGE_SIZE
+#define MAX_SG_COUNT ((MAX_HDR_SIZE - sizeof(struct rtrs_msg_rdma_read)) \
+ / sizeof(struct rtrs_sg_desc))
+
+/* We guarantee to serve 10 paths at least */
+#define CHUNK_POOL_SZ 10
+
+static struct rtrs_ib_dev_pool dev_pool;
+static mempool_t *chunk_pool;
+struct class *rtrs_dev_class;
+
+static int __read_mostly max_chunk_size = DEFAULT_MAX_CHUNK_SIZE;
+static int __read_mostly sess_queue_depth = DEFAULT_SESS_QUEUE_DEPTH;
+
+static bool always_invalidate = true;
+module_param(always_invalidate, bool, 0444);
+MODULE_PARM_DESC(always_invalidate,
+ "Invalidate memory registration for contiguous memory regions before accessing.");
+
+module_param_named(max_chunk_size, max_chunk_size, int, 0444);
+MODULE_PARM_DESC(max_chunk_size,
+ "Max size for each IO request, when change the unit is in byte (default: "
+ __stringify(DEFAULT_MAX_CHUNK_SIZE) "KB)");
+
+module_param_named(sess_queue_depth, sess_queue_depth, int, 0444);
+MODULE_PARM_DESC(sess_queue_depth,
+ "Number of buffers for pending I/O requests to allocate per session. Maximum: "
+ __stringify(MAX_SESS_QUEUE_DEPTH) " (default: "
+ __stringify(DEFAULT_SESS_QUEUE_DEPTH) ")");
+
+static char cq_affinity_list[256];
+static cpumask_t cq_affinity_mask = { CPU_BITS_ALL };
+
+static void init_cq_affinity(void)
+{
+ sprintf(cq_affinity_list, "0-%d", nr_cpu_ids - 1);
+}
+
+static int cq_affinity_list_set(const char *val, const struct kernel_param *kp)
+{
+ int ret = 0, len = strlen(val);
+ cpumask_var_t new_value;
+
+ init_cq_affinity();
+
+ if (len >= sizeof(cq_affinity_list))
+ return -EINVAL;
+ if (!alloc_cpumask_var(&new_value, GFP_KERNEL))
+ return -ENOMEM;
+
+ ret = cpulist_parse(val, new_value);
+ if (ret) {
+ pr_err("Can't set cq_affinity_list \"%s\": %d\n", val,
+ ret);
+ goto free_cpumask;
+ }
+
+ strlcpy(cq_affinity_list, val, sizeof(cq_affinity_list));
+ *strchrnul(cq_affinity_list, '\n') = '\0';
+ cpumask_copy(&cq_affinity_mask, new_value);
+
+ pr_info("cq_affinity_list changed to %*pbl\n",
+ cpumask_pr_args(&cq_affinity_mask));
+free_cpumask:
+ free_cpumask_var(new_value);
+ return ret;
+}
+
+static struct kparam_string cq_affinity_list_kparam_str = {
+ .maxlen = sizeof(cq_affinity_list),
+ .string = cq_affinity_list
+};
+
+static const struct kernel_param_ops cq_affinity_list_ops = {
+ .set = cq_affinity_list_set,
+ .get = param_get_string,
+};
+
+module_param_cb(cq_affinity_list, &cq_affinity_list_ops,
+ &cq_affinity_list_kparam_str, 0644);
+MODULE_PARM_DESC(cq_affinity_list,
+ "Sets the list of cpus to use as cq vectors. (default: use all possible CPUs)");
+
+static struct workqueue_struct *rtrs_wq;
+
+static inline struct rtrs_srv_con *to_srv_con(struct rtrs_con *c)
+{
+ return container_of(c, struct rtrs_srv_con, c);
+}
+
+static inline struct rtrs_srv_sess *to_srv_sess(struct rtrs_sess *s)
+{
+ return container_of(s, struct rtrs_srv_sess, s);
+}
+
+static bool __rtrs_srv_change_state(struct rtrs_srv_sess *sess,
+ enum rtrs_srv_state new_state)
+{
+ enum rtrs_srv_state old_state;
+ bool changed = false;
+
+ lockdep_assert_held(&sess->state_lock);
+ old_state = sess->state;
+ switch (new_state) {
+ case RTRS_SRV_CONNECTED:
+ switch (old_state) {
+ case RTRS_SRV_CONNECTING:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ case RTRS_SRV_CLOSING:
+ switch (old_state) {
+ case RTRS_SRV_CONNECTING:
+ case RTRS_SRV_CONNECTED:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ case RTRS_SRV_CLOSED:
+ switch (old_state) {
+ case RTRS_SRV_CLOSING:
+ changed = true;
+ /* FALLTHRU */
+ default:
+ break;
+ }
+ break;
+ default:
+ break;
+ }
+ if (changed)
+ sess->state = new_state;
+
+ return changed;
+}
+
+static bool rtrs_srv_change_state_get_old(struct rtrs_srv_sess *sess,
+ enum rtrs_srv_state new_state,
+ enum rtrs_srv_state *old_state)
+{
+ bool changed;
+
+ spin_lock_irq(&sess->state_lock);
+ *old_state = sess->state;
+ changed = __rtrs_srv_change_state(sess, new_state);
+ spin_unlock_irq(&sess->state_lock);
+
+ return changed;
+}
+
+static bool rtrs_srv_change_state(struct rtrs_srv_sess *sess,
+ enum rtrs_srv_state new_state)
+{
+ enum rtrs_srv_state old_state;
+
+ return rtrs_srv_change_state_get_old(sess, new_state, &old_state);
+}
+
+static void free_id(struct rtrs_srv_op *id)
+{
+ if (!id)
+ return;
+ kfree(id->tx_wr);
+ kfree(id->tx_sg);
+ kfree(id);
+}
+
+static void rtrs_srv_free_ops_ids(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+ int i;
+
+ WARN_ON(atomic_read(&sess->ids_inflight));
+ if (sess->ops_ids) {
+ for (i = 0; i < srv->queue_depth; i++)
+ free_id(sess->ops_ids[i]);
+ kfree(sess->ops_ids);
+ sess->ops_ids = NULL;
+ }
+}
+
+static int rtrs_srv_alloc_ops_ids(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_srv_op *id;
+ int i;
+
+ sess->ops_ids = kcalloc(srv->queue_depth, sizeof(*sess->ops_ids),
+ GFP_KERNEL);
+ if (unlikely(!sess->ops_ids))
+ goto err;
+
+ for (i = 0; i < srv->queue_depth; ++i) {
+ id = kzalloc(sizeof(*id), GFP_KERNEL);
+ if (unlikely(!id))
+ goto err;
+
+ sess->ops_ids[i] = id;
+ id->tx_wr = kcalloc(MAX_SG_COUNT, sizeof(*id->tx_wr),
+ GFP_KERNEL);
+ if (unlikely(!id->tx_wr))
+ goto err;
+
+ id->tx_sg = kcalloc(MAX_SG_COUNT, sizeof(*id->tx_sg),
+ GFP_KERNEL);
+ if (unlikely(!id->tx_sg))
+ goto err;
+ }
+ init_waitqueue_head(&sess->ids_waitq);
+ atomic_set(&sess->ids_inflight, 0);
+
+ return 0;
+
+err:
+ rtrs_srv_free_ops_ids(sess);
+ return -ENOMEM;
+}
+
+static void rtrs_srv_get_ops_ids(struct rtrs_srv_sess *sess)
+{
+ atomic_inc(&sess->ids_inflight);
+}
+
+static void rtrs_srv_put_ops_ids(struct rtrs_srv_sess *sess)
+{
+ if (atomic_dec_and_test(&sess->ids_inflight))
+ wake_up(&sess->ids_waitq);
+}
+
+static void rtrs_srv_wait_ops_ids(struct rtrs_srv_sess *sess)
+{
+ wait_event(sess->ids_waitq, !atomic_read(&sess->ids_inflight));
+}
+
+static void rtrs_srv_rdma_done(struct ib_cq *cq, struct ib_wc *wc);
+
+static struct ib_cqe io_comp_cqe = {
+ .done = rtrs_srv_rdma_done
+};
+
+static void rtrs_srv_reg_mr_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_srv_con *con = cq->cq_context;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ rtrs_err(s, "REG MR failed: %s\n",
+ ib_wc_status_msg(wc->status));
+ close_sess(sess);
+ return;
+ }
+}
+
+static struct ib_cqe local_reg_cqe = {
+ .done = rtrs_srv_reg_mr_done
+};
+
+static int rdma_write_sg(struct rtrs_srv_op *id)
+{
+ struct rtrs_sess *s = id->con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ dma_addr_t dma_addr = sess->dma_addr[id->msg_id];
+ struct rtrs_srv_mr *srv_mr;
+ struct rtrs_srv *srv = sess->srv;
+ struct ib_send_wr inv_wr, imm_wr;
+ struct ib_rdma_wr *wr = NULL;
+ const struct ib_send_wr *bad_wr;
+ enum ib_send_flags flags;
+ size_t sg_cnt;
+ int err, i, offset;
+ bool need_inval;
+ u32 rkey = 0;
+ struct ib_reg_wr rwr;
+
+ sg_cnt = le16_to_cpu(id->rd_msg->sg_cnt);
+ need_inval = le16_to_cpu(id->rd_msg->flags) & RTRS_MSG_NEED_INVAL_F;
+ if (unlikely(!sg_cnt))
+ return -EINVAL;
+
+ offset = 0;
+ for (i = 0; i < sg_cnt; i++) {
+ struct ib_sge *list;
+
+ wr = &id->tx_wr[i];
+ list = &id->tx_sg[i];
+ list->addr = dma_addr + offset;
+ list->length = le32_to_cpu(id->rd_msg->desc[i].len);
+
+ /* WR will fail with length error
+ * if this is 0
+ */
+ if (unlikely(list->length == 0)) {
+ rtrs_err(s, "Invalid RDMA-Write sg list length 0\n");
+ return -EINVAL;
+ }
+
+ list->lkey = sess->s.dev->ib_pd->local_dma_lkey;
+ offset += list->length;
+
+ wr->wr.wr_cqe = &io_comp_cqe;
+ wr->wr.sg_list = list;
+ wr->wr.num_sge = 1;
+ wr->remote_addr = le64_to_cpu(id->rd_msg->desc[i].addr);
+ wr->rkey = le32_to_cpu(id->rd_msg->desc[i].key);
+ if (rkey == 0)
+ rkey = wr->rkey;
+ else
+ /* Only one key is actually used */
+ WARN_ON_ONCE(rkey != wr->rkey);
+
+ if (i < (sg_cnt - 1))
+ wr->wr.next = &id->tx_wr[i + 1].wr;
+
+ wr->wr.opcode = IB_WR_RDMA_WRITE;
+ wr->wr.ex.imm_data = 0;
+ wr->wr.send_flags = 0;
+ }
+
+ if (need_inval && always_invalidate) {
+ wr->wr.next = &rwr.wr;
+ rwr.wr.next = &inv_wr;
+ inv_wr.next = &imm_wr;
+ } else if (always_invalidate) {
+ wr->wr.next = &rwr.wr;
+ rwr.wr.next = &imm_wr;
+ } else if (need_inval) {
+ wr->wr.next = &inv_wr;
+ inv_wr.next = &imm_wr;
+ } else {
+ wr->wr.next = &imm_wr;
+ }
+ /*
+ * From time to time we have to post signalled sends,
+ * or send queue will fill up and only QP reset can help.
+ */
+ flags = atomic_inc_return(&id->con->wr_cnt) % srv->queue_depth ?
+ 0 : IB_SEND_SIGNALED;
+
+ if (need_inval) {
+ inv_wr.wr_cqe = &io_comp_cqe;
+ inv_wr.sg_list = NULL;
+ inv_wr.num_sge = 0;
+ inv_wr.opcode = IB_WR_SEND_WITH_INV;
+ inv_wr.send_flags = 0;
+ inv_wr.ex.invalidate_rkey = rkey;
+ }
+
+ imm_wr.next = NULL;
+ imm_wr.wr_cqe = &io_comp_cqe;
+ if (always_invalidate) {
+ struct ib_sge list;
+ struct rtrs_msg_rkey_rsp *msg;
+
+ srv_mr = &sess->mrs[id->msg_id];
+ rwr.wr.opcode = IB_WR_REG_MR;
+ rwr.wr.wr_cqe = &local_reg_cqe;
+ rwr.wr.num_sge = 0;
+ rwr.mr = srv_mr->mr;
+ rwr.wr.send_flags = 0;
+ rwr.key = srv_mr->mr->rkey;
+ rwr.access = (IB_ACCESS_LOCAL_WRITE |
+ IB_ACCESS_REMOTE_WRITE);
+ msg = srv_mr->iu->buf;
+ msg->buf_id = cpu_to_le16(id->msg_id);
+ msg->type = cpu_to_le16(RTRS_MSG_RKEY_RSP);
+ msg->rkey = cpu_to_le32(srv_mr->mr->rkey);
+
+ list.addr = srv_mr->iu->dma_addr;
+ list.length = sizeof(*msg);
+ list.lkey = sess->s.dev->ib_pd->local_dma_lkey;
+ imm_wr.sg_list = &list;
+ imm_wr.num_sge = 1;
+ imm_wr.opcode = IB_WR_SEND_WITH_IMM;
+ ib_dma_sync_single_for_device(sess->s.dev->ib_dev,
+ srv_mr->iu->dma_addr,
+ srv_mr->iu->size, DMA_TO_DEVICE);
+ } else {
+ imm_wr.sg_list = NULL;
+ imm_wr.num_sge = 0;
+ imm_wr.opcode = IB_WR_RDMA_WRITE_WITH_IMM;
+ }
+ imm_wr.send_flags = flags;
+ imm_wr.ex.imm_data = cpu_to_be32(rtrs_to_io_rsp_imm(id->msg_id,
+ 0, need_inval));
+
+ ib_dma_sync_single_for_device(sess->s.dev->ib_dev, dma_addr,
+ offset, DMA_BIDIRECTIONAL);
+
+ err = ib_post_send(id->con->c.qp, &id->tx_wr[0].wr, &bad_wr);
+ if (unlikely(err))
+ rtrs_err(s,
+ "Posting RDMA-Write-Request to QP failed, err: %d\n",
+ err);
+
+ return err;
+}
+
+/**
+ * send_io_resp_imm() - response with empty IMM on failed READ/WRITE requests or
+ * on successful WRITE request.
+ * @con the connection to send back result
+ * @id the id associated to io
+ * @errno the error number of the IO.
+ *
+ * Return 0 on success, errno otherwise.
+ */
+static int send_io_resp_imm(struct rtrs_srv_con *con, struct rtrs_srv_op *id,
+ int errno)
+{
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct ib_send_wr inv_wr, imm_wr, *wr = NULL;
+ struct ib_reg_wr rwr;
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_srv_mr *srv_mr;
+ bool need_inval = false;
+ enum ib_send_flags flags;
+ const struct ib_send_wr *bad_wr;
+ u32 imm;
+ int err;
+
+ if (id->dir == READ) {
+ struct rtrs_msg_rdma_read *rd_msg = id->rd_msg;
+ size_t sg_cnt;
+
+ need_inval = le16_to_cpu(rd_msg->flags) &
+ RTRS_MSG_NEED_INVAL_F;
+ sg_cnt = le16_to_cpu(rd_msg->sg_cnt);
+
+ if (need_inval) {
+ if (likely(sg_cnt)) {
+ inv_wr.wr_cqe = &io_comp_cqe;
+ inv_wr.sg_list = NULL;
+ inv_wr.num_sge = 0;
+ inv_wr.opcode = IB_WR_SEND_WITH_INV;
+ inv_wr.send_flags = 0;
+ /* Only one key is actually used */
+ inv_wr.ex.invalidate_rkey =
+ le32_to_cpu(rd_msg->desc[0].key);
+ } else {
+ WARN_ON_ONCE(1);
+ need_inval = false;
+ }
+ }
+ }
+
+ if (need_inval && always_invalidate) {
+ wr = &inv_wr;
+ inv_wr.next = &rwr.wr;
+ rwr.wr.next = &imm_wr;
+ } else if (always_invalidate) {
+ wr = &rwr.wr;
+ rwr.wr.next = &imm_wr;
+ } else if (need_inval) {
+ wr = &inv_wr;
+ inv_wr.next = &imm_wr;
+ } else {
+ wr = &imm_wr;
+ }
+ /*
+ * From time to time we have to post signalled sends,
+ * or send queue will fill up and only QP reset can help.
+ */
+ flags = atomic_inc_return(&con->wr_cnt) % srv->queue_depth ?
+ 0 : IB_SEND_SIGNALED;
+ imm = rtrs_to_io_rsp_imm(id->msg_id, errno, need_inval);
+ imm_wr.next = NULL;
+ imm_wr.wr_cqe = &io_comp_cqe;
+ if (always_invalidate) {
+ struct ib_sge list;
+ struct rtrs_msg_rkey_rsp *msg;
+
+ srv_mr = &sess->mrs[id->msg_id];
+ rwr.wr.next = &imm_wr;
+ rwr.wr.opcode = IB_WR_REG_MR;
+ rwr.wr.wr_cqe = &local_reg_cqe;
+ rwr.wr.num_sge = 0;
+ rwr.wr.send_flags = 0;
+ rwr.mr = srv_mr->mr;
+ rwr.key = srv_mr->mr->rkey;
+ rwr.access = (IB_ACCESS_LOCAL_WRITE |
+ IB_ACCESS_REMOTE_WRITE);
+ msg = srv_mr->iu->buf;
+ msg->buf_id = cpu_to_le16(id->msg_id);
+ msg->type = cpu_to_le16(RTRS_MSG_RKEY_RSP);
+ msg->rkey = cpu_to_le32(srv_mr->mr->rkey);
+
+ list.addr = srv_mr->iu->dma_addr;
+ list.length = sizeof(*msg);
+ list.lkey = sess->s.dev->ib_pd->local_dma_lkey;
+ imm_wr.sg_list = &list;
+ imm_wr.num_sge = 1;
+ imm_wr.opcode = IB_WR_SEND_WITH_IMM;
+ ib_dma_sync_single_for_device(sess->s.dev->ib_dev,
+ srv_mr->iu->dma_addr,
+ srv_mr->iu->size, DMA_TO_DEVICE);
+ } else {
+ imm_wr.sg_list = NULL;
+ imm_wr.num_sge = 0;
+ imm_wr.opcode = IB_WR_RDMA_WRITE_WITH_IMM;
+ }
+ imm_wr.send_flags = flags;
+ imm_wr.ex.imm_data = cpu_to_be32(imm);
+
+ err = ib_post_send(id->con->c.qp, wr, &bad_wr);
+ if (unlikely(err))
+ rtrs_err_rl(s, "Posting RDMA-Reply to QP failed, err: %d\n",
+ err);
+
+ return err;
+}
+
+void close_sess(struct rtrs_srv_sess *sess)
+{
+ enum rtrs_srv_state old_state;
+
+ if (rtrs_srv_change_state_get_old(sess, RTRS_SRV_CLOSING,
+ &old_state))
+ queue_work(rtrs_wq, &sess->close_work);
+ WARN_ON(sess->state != RTRS_SRV_CLOSING);
+}
+
+static inline const char *rtrs_srv_state_str(enum rtrs_srv_state state)
+{
+ switch (state) {
+ case RTRS_SRV_CONNECTING:
+ return "RTRS_SRV_CONNECTING";
+ case RTRS_SRV_CONNECTED:
+ return "RTRS_SRV_CONNECTED";
+ case RTRS_SRV_CLOSING:
+ return "RTRS_SRV_CLOSING";
+ case RTRS_SRV_CLOSED:
+ return "RTRS_SRV_CLOSED";
+ default:
+ return "UNKNOWN";
+ }
+}
+
+/*
+ * rtrs_srv_resp_rdma() - sends response to the client.
+ *
+ * Context: any
+ */
+void rtrs_srv_resp_rdma(struct rtrs_srv_op *id, int status)
+{
+ struct rtrs_srv_con *con = id->con;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ int err;
+
+ if (WARN_ON(!id))
+ return;
+
+ if (unlikely(sess->state != RTRS_SRV_CONNECTED)) {
+ rtrs_err_rl(s,
+ "Sending I/O response failed, session is disconnected, sess state %s\n",
+ rtrs_srv_state_str(sess->state));
+ goto out;
+ }
+ if (always_invalidate) {
+ struct rtrs_srv_mr *mr = &sess->mrs[id->msg_id];
+
+ ib_update_fast_reg_key(mr->mr, ib_inc_rkey(mr->mr->rkey));
+ }
+ if (status || id->dir == WRITE || !id->rd_msg->sg_cnt)
+ err = send_io_resp_imm(con, id, status);
+ else
+ err = rdma_write_sg(id);
+ if (unlikely(err)) {
+ rtrs_err_rl(s, "IO response failed: %d\n", err);
+ close_sess(sess);
+ }
+out:
+ rtrs_srv_put_ops_ids(sess);
+}
+EXPORT_SYMBOL(rtrs_srv_resp_rdma);
+
+void rtrs_srv_set_sess_priv(struct rtrs_srv *srv, void *priv)
+{
+ srv->priv = priv;
+}
+EXPORT_SYMBOL(rtrs_srv_set_sess_priv);
+
+static void unmap_cont_bufs(struct rtrs_srv_sess *sess)
+{
+ int i;
+
+ for (i = 0; i < sess->mrs_num; i++) {
+ struct rtrs_srv_mr *srv_mr;
+
+ srv_mr = &sess->mrs[i];
+ rtrs_iu_free(srv_mr->iu, DMA_TO_DEVICE,
+ sess->s.dev->ib_dev, 1);
+ ib_dereg_mr(srv_mr->mr);
+ ib_dma_unmap_sg(sess->s.dev->ib_dev, srv_mr->sgt.sgl,
+ srv_mr->sgt.nents, DMA_BIDIRECTIONAL);
+ sg_free_table(&srv_mr->sgt);
+ }
+ kfree(sess->mrs);
+}
+
+static int map_cont_bufs(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_sess *ss = &sess->s;
+ int i, mri, err, mrs_num;
+ unsigned int chunk_bits;
+ int chunks_per_mr = 1;
+
+ /*
+ * Here we map queue_depth chunks to MR. Firstly we have to
+ * figure out how many chunks can we map per MR.
+ */
+ if (always_invalidate) {
+ /*
+ * in order to do invalidate for each chunks of memory, we needs
+ * more memory regions.
+ */
+ mrs_num = srv->queue_depth;
+ } else {
+ chunks_per_mr =
+ sess->s.dev->ib_dev->attrs.max_fast_reg_page_list_len;
+ mrs_num = DIV_ROUND_UP(srv->queue_depth, chunks_per_mr);
+ chunks_per_mr = DIV_ROUND_UP(srv->queue_depth, mrs_num);
+ }
+
+ sess->mrs = kcalloc(mrs_num, sizeof(*sess->mrs), GFP_KERNEL);
+ if (unlikely(!sess->mrs))
+ return -ENOMEM;
+
+ sess->mrs_num = mrs_num;
+
+ for (mri = 0; mri < mrs_num; mri++) {
+ struct rtrs_srv_mr *srv_mr = &sess->mrs[mri];
+ struct sg_table *sgt = &srv_mr->sgt;
+ struct scatterlist *s;
+ struct ib_mr *mr;
+ int nr, chunks;
+ struct rtrs_msg_rkey_rsp *rsp;
+
+ chunks = chunks_per_mr * mri;
+ if (!always_invalidate)
+ chunks_per_mr = min_t(int, chunks_per_mr,
+ srv->queue_depth - chunks);
+
+ err = sg_alloc_table(sgt, chunks_per_mr, GFP_KERNEL);
+ if (unlikely(err))
+ goto err;
+
+ for_each_sg(sgt->sgl, s, chunks_per_mr, i)
+ sg_set_page(s, srv->chunks[chunks + i],
+ max_chunk_size, 0);
+
+ nr = ib_dma_map_sg(sess->s.dev->ib_dev, sgt->sgl,
+ sgt->nents, DMA_BIDIRECTIONAL);
+ if (unlikely(nr < sgt->nents)) {
+ err = nr < 0 ? nr : -EINVAL;
+ goto free_sg;
+ }
+ mr = ib_alloc_mr(sess->s.dev->ib_pd, IB_MR_TYPE_MEM_REG,
+ sgt->nents);
+ if (IS_ERR(mr)) {
+ err = PTR_ERR(mr);
+ goto unmap_sg;
+ }
+ nr = ib_map_mr_sg(mr, sgt->sgl, sgt->nents,
+ NULL, max_chunk_size);
+ if (unlikely(nr < sgt->nents)) {
+ err = nr < 0 ? nr : -EINVAL;
+ goto dereg_mr;
+ }
+
+ if (always_invalidate) {
+ srv_mr->iu = rtrs_iu_alloc(1, sizeof(*rsp), GFP_KERNEL,
+ sess->s.dev->ib_dev,
+ DMA_TO_DEVICE,
+ rtrs_srv_rdma_done);
+ if (unlikely(!srv_mr->iu)) {
+ rtrs_err(ss, "rtrs_iu_alloc(), err: %d\n",
+ -ENOMEM);
+ goto free_iu;
+ }
+ }
+ /* Eventually dma addr for each chunk can be cached */
+ for_each_sg(sgt->sgl, s, sgt->orig_nents, i)
+ sess->dma_addr[chunks + i] = sg_dma_address(s);
+
+ ib_update_fast_reg_key(mr, ib_inc_rkey(mr->rkey));
+ srv_mr->mr = mr;
+
+ continue;
+err:
+ while (mri--) {
+ srv_mr = &sess->mrs[mri];
+ sgt = &srv_mr->sgt;
+ mr = srv_mr->mr;
+free_iu:
+ rtrs_iu_free(srv_mr->iu, DMA_TO_DEVICE,
+ sess->s.dev->ib_dev, 1);
+dereg_mr:
+ ib_dereg_mr(mr);
+unmap_sg:
+ ib_dma_unmap_sg(sess->s.dev->ib_dev, sgt->sgl,
+ sgt->nents, DMA_BIDIRECTIONAL);
+free_sg:
+ sg_free_table(sgt);
+ }
+ kfree(sess->mrs);
+
+ return err;
+ }
+
+ chunk_bits = ilog2(srv->queue_depth - 1) + 1;
+ sess->mem_bits = (MAX_IMM_PAYL_BITS - chunk_bits);
+
+ return 0;
+}
+
+static void rtrs_srv_hb_err_handler(struct rtrs_con *c)
+{
+ close_sess(to_srv_sess(c->sess));
+}
+
+static void rtrs_srv_init_hb(struct rtrs_srv_sess *sess)
+{
+ rtrs_init_hb(&sess->s, &io_comp_cqe,
+ RTRS_HB_INTERVAL_MS,
+ RTRS_HB_MISSED_MAX,
+ rtrs_srv_hb_err_handler,
+ rtrs_wq);
+}
+
+static void rtrs_srv_start_hb(struct rtrs_srv_sess *sess)
+{
+ rtrs_start_hb(&sess->s);
+}
+
+static void rtrs_srv_stop_hb(struct rtrs_srv_sess *sess)
+{
+ rtrs_stop_hb(&sess->s);
+}
+
+static void rtrs_srv_info_rsp_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_srv_con *con = cq->cq_context;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct rtrs_iu *iu;
+
+ iu = container_of(wc->wr_cqe, struct rtrs_iu, cqe);
+ rtrs_iu_free(iu, DMA_TO_DEVICE, sess->s.dev->ib_dev, 1);
+
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ rtrs_err(s, "Sess info response send failed: %s\n",
+ ib_wc_status_msg(wc->status));
+ close_sess(sess);
+ return;
+ }
+ WARN_ON(wc->opcode != IB_WC_SEND);
+ rtrs_srv_update_wc_stats(&sess->stats);
+}
+
+static void rtrs_srv_sess_up(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_srv_ctx *ctx = srv->ctx;
+ int up;
+
+ mutex_lock(&srv->paths_ev_mutex);
+ up = ++srv->paths_up;
+ if (up == 1)
+ ctx->link_ev(srv, RTRS_SRV_LINK_EV_CONNECTED, NULL);
+ mutex_unlock(&srv->paths_ev_mutex);
+
+ /* Mark session as established */
+ sess->established = true;
+}
+
+static void rtrs_srv_sess_down(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_srv_ctx *ctx = srv->ctx;
+
+ if (!sess->established)
+ return;
+
+ sess->established = false;
+ mutex_lock(&srv->paths_ev_mutex);
+ WARN_ON(!srv->paths_up);
+ if (--srv->paths_up == 0)
+ ctx->link_ev(srv, RTRS_SRV_LINK_EV_DISCONNECTED, srv->priv);
+ mutex_unlock(&srv->paths_ev_mutex);
+}
+
+static int post_recv_sess(struct rtrs_srv_sess *sess);
+
+static int process_info_req(struct rtrs_srv_con *con,
+ struct rtrs_msg_info_req *msg)
+{
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct ib_send_wr *reg_wr = NULL;
+ struct rtrs_msg_info_rsp *rsp;
+ struct rtrs_iu *tx_iu;
+ struct ib_reg_wr *rwr;
+ int mri, err;
+ size_t tx_sz;
+
+ err = post_recv_sess(sess);
+ if (unlikely(err)) {
+ rtrs_err(s, "post_recv_sess(), err: %d\n", err);
+ return err;
+ }
+ rwr = kcalloc(sess->mrs_num, sizeof(*rwr), GFP_KERNEL);
+ if (unlikely(!rwr)) {
+ rtrs_err(s, "No memory\n");
+ return -ENOMEM;
+ }
+ memcpy(sess->s.sessname, msg->sessname, sizeof(sess->s.sessname));
+
+ tx_sz = sizeof(*rsp);
+ tx_sz += sizeof(rsp->desc[0]) * sess->mrs_num;
+ tx_iu = rtrs_iu_alloc(1, tx_sz, GFP_KERNEL, sess->s.dev->ib_dev,
+ DMA_TO_DEVICE, rtrs_srv_info_rsp_done);
+ if (unlikely(!tx_iu)) {
+ rtrs_err(s, "rtrs_iu_alloc(), err: %d\n", -ENOMEM);
+ err = -ENOMEM;
+ goto rwr_free;
+ }
+
+ rsp = tx_iu->buf;
+ rsp->type = cpu_to_le16(RTRS_MSG_INFO_RSP);
+ rsp->sg_cnt = cpu_to_le16(sess->mrs_num);
+
+ for (mri = 0; mri < sess->mrs_num; mri++) {
+ struct ib_mr *mr = sess->mrs[mri].mr;
+
+ rsp->desc[mri].addr = cpu_to_le64(mr->iova);
+ rsp->desc[mri].key = cpu_to_le32(mr->rkey);
+ rsp->desc[mri].len = cpu_to_le32(mr->length);
+
+ /*
+ * Fill in reg MR request and chain them *backwards*
+ */
+ rwr[mri].wr.next = mri ? &rwr[mri - 1].wr : NULL;
+ rwr[mri].wr.opcode = IB_WR_REG_MR;
+ rwr[mri].wr.wr_cqe = &local_reg_cqe;
+ rwr[mri].wr.num_sge = 0;
+ rwr[mri].wr.send_flags = mri ? 0 : IB_SEND_SIGNALED;
+ rwr[mri].mr = mr;
+ rwr[mri].key = mr->rkey;
+ rwr[mri].access = (IB_ACCESS_LOCAL_WRITE |
+ IB_ACCESS_REMOTE_WRITE);
+ reg_wr = &rwr[mri].wr;
+ }
+
+ err = rtrs_srv_create_sess_files(sess);
+ if (unlikely(err))
+ goto iu_free;
+ get_device(&sess->srv->dev);
+ rtrs_srv_change_state(sess, RTRS_SRV_CONNECTED);
+ rtrs_srv_start_hb(sess);
+
+ /*
+ * We do not account number of established connections at the current
+ * moment, we rely on the client, which should send info request when
+ * all connections are successfully established. Thus, simply notify
+ * listener with a proper event if we are the first path.
+ */
+ rtrs_srv_sess_up(sess);
+
+ ib_dma_sync_single_for_device(sess->s.dev->ib_dev, tx_iu->dma_addr,
+ tx_iu->size, DMA_TO_DEVICE);
+
+ /* Send info response */
+ err = rtrs_iu_post_send(&con->c, tx_iu, tx_sz, reg_wr);
+ if (unlikely(err)) {
+ rtrs_err(s, "rtrs_iu_post_send(), err: %d\n", err);
+iu_free:
+ rtrs_iu_free(tx_iu, DMA_TO_DEVICE, sess->s.dev->ib_dev, 1);
+ }
+rwr_free:
+ kfree(rwr);
+
+ return err;
+}
+
+static void rtrs_srv_info_req_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_srv_con *con = cq->cq_context;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct rtrs_msg_info_req *msg;
+ struct rtrs_iu *iu;
+ int err;
+
+ WARN_ON(con->c.cid);
+
+ iu = container_of(wc->wr_cqe, struct rtrs_iu, cqe);
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ rtrs_err(s, "Sess info request receive failed: %s\n",
+ ib_wc_status_msg(wc->status));
+ goto close;
+ }
+ WARN_ON(wc->opcode != IB_WC_RECV);
+
+ if (unlikely(wc->byte_len < sizeof(*msg))) {
+ rtrs_err(s, "Sess info request is malformed: size %d\n",
+ wc->byte_len);
+ goto close;
+ }
+ ib_dma_sync_single_for_cpu(sess->s.dev->ib_dev, iu->dma_addr,
+ iu->size, DMA_FROM_DEVICE);
+ msg = iu->buf;
+ if (unlikely(le16_to_cpu(msg->type) != RTRS_MSG_INFO_REQ)) {
+ rtrs_err(s, "Sess info request is malformed: type %d\n",
+ le16_to_cpu(msg->type));
+ goto close;
+ }
+ err = process_info_req(con, msg);
+ if (unlikely(err))
+ goto close;
+
+out:
+ rtrs_iu_free(iu, DMA_FROM_DEVICE, sess->s.dev->ib_dev, 1);
+ return;
+close:
+ close_sess(sess);
+ goto out;
+}
+
+static int post_recv_info_req(struct rtrs_srv_con *con)
+{
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct rtrs_iu *rx_iu;
+ int err;
+
+ rx_iu = rtrs_iu_alloc(1, sizeof(struct rtrs_msg_info_req),
+ GFP_KERNEL, sess->s.dev->ib_dev,
+ DMA_FROM_DEVICE, rtrs_srv_info_req_done);
+ if (unlikely(!rx_iu)) {
+ rtrs_err(s, "rtrs_iu_alloc(): no memory\n");
+ return -ENOMEM;
+ }
+ /* Prepare for getting info response */
+ err = rtrs_iu_post_recv(&con->c, rx_iu);
+ if (unlikely(err)) {
+ rtrs_err(s, "rtrs_iu_post_recv(), err: %d\n", err);
+ rtrs_iu_free(rx_iu, DMA_FROM_DEVICE, sess->s.dev->ib_dev, 1);
+ return err;
+ }
+
+ return 0;
+}
+
+static int post_recv_io(struct rtrs_srv_con *con, size_t q_size)
+{
+ int i, err;
+
+ for (i = 0; i < q_size; i++) {
+ err = rtrs_post_recv_empty(&con->c, &io_comp_cqe);
+ if (unlikely(err))
+ return err;
+ }
+
+ return 0;
+}
+
+static int post_recv_sess(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_sess *s = &sess->s;
+ size_t q_size;
+ int err, cid;
+
+ for (cid = 0; cid < sess->s.con_num; cid++) {
+ if (cid == 0)
+ q_size = SERVICE_CON_QUEUE_DEPTH;
+ else
+ q_size = srv->queue_depth;
+
+ err = post_recv_io(to_srv_con(sess->s.con[cid]), q_size);
+ if (unlikely(err)) {
+ rtrs_err(s, "post_recv_io(), err: %d\n", err);
+ return err;
+ }
+ }
+
+ return 0;
+}
+
+static void process_read(struct rtrs_srv_con *con,
+ struct rtrs_msg_rdma_read *msg,
+ u32 buf_id, u32 off)
+{
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_srv_ctx *ctx = srv->ctx;
+ struct rtrs_srv_op *id;
+
+ size_t usr_len, data_len;
+ void *data;
+ int ret;
+
+ if (unlikely(sess->state != RTRS_SRV_CONNECTED)) {
+ rtrs_err_rl(s,
+ "Processing read request failed, session is disconnected, sess state %s\n",
+ rtrs_srv_state_str(sess->state));
+ return;
+ }
+ rtrs_srv_get_ops_ids(sess);
+ rtrs_srv_update_rdma_stats(&sess->stats, off, READ);
+ id = sess->ops_ids[buf_id];
+ id->con = con;
+ id->dir = READ;
+ id->msg_id = buf_id;
+ id->rd_msg = msg;
+ usr_len = le16_to_cpu(msg->usr_len);
+ data_len = off - usr_len;
+ data = page_address(srv->chunks[buf_id]);
+ ret = ctx->rdma_ev(srv, srv->priv, id, READ, data, data_len,
+ data + data_len, usr_len);
+
+ if (unlikely(ret)) {
+ rtrs_err_rl(s,
+ "Processing read request failed, user module cb reported for msg_id %d, err: %d\n",
+ buf_id, ret);
+ goto send_err_msg;
+ }
+
+ return;
+
+send_err_msg:
+ ret = send_io_resp_imm(con, id, ret);
+ if (ret < 0) {
+ rtrs_err_rl(s,
+ "Sending err msg for failed RDMA-Write-Req failed, msg_id %d, err: %d\n",
+ buf_id, ret);
+ close_sess(sess);
+ }
+ rtrs_srv_put_ops_ids(sess);
+}
+
+static void process_write(struct rtrs_srv_con *con,
+ struct rtrs_msg_rdma_write *req,
+ u32 buf_id, u32 off)
+{
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_srv_ctx *ctx = srv->ctx;
+ struct rtrs_srv_op *id;
+
+ size_t data_len, usr_len;
+ void *data;
+ int ret;
+
+ if (unlikely(sess->state != RTRS_SRV_CONNECTED)) {
+ rtrs_err_rl(s,
+ "Processing write request failed, session is disconnected, sess state %s\n",
+ rtrs_srv_state_str(sess->state));
+ return;
+ }
+ rtrs_srv_get_ops_ids(sess);
+ rtrs_srv_update_rdma_stats(&sess->stats, off, WRITE);
+ id = sess->ops_ids[buf_id];
+ id->con = con;
+ id->dir = WRITE;
+ id->msg_id = buf_id;
+
+ usr_len = le16_to_cpu(req->usr_len);
+ data_len = off - usr_len;
+ data = page_address(srv->chunks[buf_id]);
+ ret = ctx->rdma_ev(srv, srv->priv, id, WRITE, data, data_len,
+ data + data_len, usr_len);
+ if (unlikely(ret)) {
+ rtrs_err_rl(s,
+ "Processing write request failed, user module callback reports err: %d\n",
+ ret);
+ goto send_err_msg;
+ }
+
+ return;
+
+send_err_msg:
+ ret = send_io_resp_imm(con, id, ret);
+ if (ret < 0) {
+ rtrs_err_rl(s,
+ "Processing write request failed, sending I/O response failed, msg_id %d, err: %d\n",
+ buf_id, ret);
+ close_sess(sess);
+ }
+ rtrs_srv_put_ops_ids(sess);
+}
+
+static void process_io_req(struct rtrs_srv_con *con, void *msg,
+ u32 id, u32 off)
+{
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct rtrs_msg_rdma_hdr *hdr;
+ unsigned int type;
+
+ ib_dma_sync_single_for_cpu(sess->s.dev->ib_dev, sess->dma_addr[id],
+ max_chunk_size, DMA_BIDIRECTIONAL);
+ hdr = msg;
+ type = le16_to_cpu(hdr->type);
+
+ switch (type) {
+ case RTRS_MSG_WRITE:
+ process_write(con, msg, id, off);
+ break;
+ case RTRS_MSG_READ:
+ process_read(con, msg, id, off);
+ break;
+ default:
+ rtrs_err(s,
+ "Processing I/O request failed, unknown message type received: 0x%02x\n",
+ type);
+ goto err;
+ }
+
+ return;
+
+err:
+ close_sess(sess);
+}
+
+static void rtrs_srv_inv_rkey_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_srv_mr *mr =
+ container_of(wc->wr_cqe, typeof(*mr), inv_cqe);
+ struct rtrs_srv_con *con = cq->cq_context;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct rtrs_srv *srv = sess->srv;
+ u32 msg_id, off;
+ void *data;
+
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ rtrs_err(s, "Failed IB_WR_LOCAL_INV: %s\n",
+ ib_wc_status_msg(wc->status));
+ close_sess(sess);
+ }
+ msg_id = mr->msg_id;
+ off = mr->msg_off;
+ data = page_address(srv->chunks[msg_id]) + off;
+ process_io_req(con, data, msg_id, off);
+}
+
+static int rtrs_srv_inv_rkey(struct rtrs_srv_con *con,
+ struct rtrs_srv_mr *mr)
+{
+ const struct ib_send_wr *bad_wr;
+ struct ib_send_wr wr = {
+ .opcode = IB_WR_LOCAL_INV,
+ .wr_cqe = &mr->inv_cqe,
+ .next = NULL,
+ .num_sge = 0,
+ .send_flags = IB_SEND_SIGNALED,
+ .ex.invalidate_rkey = mr->mr->rkey,
+ };
+ mr->inv_cqe.done = rtrs_srv_inv_rkey_done;
+
+ return ib_post_send(con->c.qp, &wr, &bad_wr);
+}
+
+static void rtrs_srv_rdma_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ struct rtrs_srv_con *con = cq->cq_context;
+ struct rtrs_sess *s = con->c.sess;
+ struct rtrs_srv_sess *sess = to_srv_sess(s);
+ struct rtrs_srv *srv = sess->srv;
+ u32 imm_type, imm_payload;
+ int err;
+
+ if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ if (wc->status != IB_WC_WR_FLUSH_ERR) {
+ rtrs_err(s,
+ "%s (wr_cqe: %p, type: %d, vendor_err: 0x%x, len: %u)\n",
+ ib_wc_status_msg(wc->status), wc->wr_cqe,
+ wc->opcode, wc->vendor_err, wc->byte_len);
+ close_sess(sess);
+ }
+ return;
+ }
+ rtrs_srv_update_wc_stats(&sess->stats);
+
+ switch (wc->opcode) {
+ case IB_WC_RECV_RDMA_WITH_IMM:
+ /*
+ * post_recv() RDMA write completions of IO reqs (read/write)
+ * and hb
+ */
+ if (WARN_ON(wc->wr_cqe != &io_comp_cqe))
+ return;
+ err = rtrs_post_recv_empty(&con->c, &io_comp_cqe);
+ if (unlikely(err)) {
+ rtrs_err(s, "rtrs_post_recv(), err: %d\n", err);
+ close_sess(sess);
+ break;
+ }
+ rtrs_from_imm(be32_to_cpu(wc->ex.imm_data),
+ &imm_type, &imm_payload);
+ if (likely(imm_type == RTRS_IO_REQ_IMM)) {
+ u32 msg_id, off;
+ void *data;
+
+ msg_id = imm_payload >> sess->mem_bits;
+ off = imm_payload & ((1 << sess->mem_bits) - 1);
+ if (unlikely(msg_id > srv->queue_depth ||
+ off > max_chunk_size)) {
+ rtrs_err(s, "Wrong msg_id %u, off %u\n",
+ msg_id, off);
+ close_sess(sess);
+ return;
+ }
+ if (always_invalidate) {
+ struct rtrs_srv_mr *mr = &sess->mrs[msg_id];
+
+ mr->msg_off = off;
+ mr->msg_id = msg_id;
+ err = rtrs_srv_inv_rkey(con, mr);
+ if (unlikely(err)) {
+ rtrs_err(s, "rtrs_post_recv(), err: %d\n",
+ err);
+ close_sess(sess);
+ break;
+ }
+ } else {
+ data = page_address(srv->chunks[msg_id]) + off;
+ process_io_req(con, data, msg_id, off);
+ }
+ } else if (imm_type == RTRS_HB_MSG_IMM) {
+ WARN_ON(con->c.cid);
+ rtrs_send_hb_ack(&sess->s);
+ } else if (imm_type == RTRS_HB_ACK_IMM) {
+ WARN_ON(con->c.cid);
+ sess->s.hb_missed_cnt = 0;
+ } else {
+ rtrs_wrn(s, "Unknown IMM type %u\n", imm_type);
+ }
+ break;
+ case IB_WC_RDMA_WRITE:
+ case IB_WC_SEND:
+ /*
+ * post_send() RDMA write completions of IO reqs (read/write)
+ * and hb
+ */
+ break;
+ default:
+ rtrs_wrn(s, "Unexpected WC type: %d\n", wc->opcode);
+ return;
+ }
+}
+
+int rtrs_srv_get_sess_name(struct rtrs_srv *srv, char *sessname, size_t len)
+{
+ struct rtrs_srv_sess *sess;
+ int err = -ENOTCONN;
+
+ mutex_lock(&srv->paths_mutex);
+ list_for_each_entry(sess, &srv->paths_list, s.entry) {
+ if (sess->state != RTRS_SRV_CONNECTED)
+ continue;
+ memcpy(sessname, sess->s.sessname,
+ min_t(size_t, sizeof(sess->s.sessname), len));
+ err = 0;
+ break;
+ }
+ mutex_unlock(&srv->paths_mutex);
+
+ return err;
+}
+EXPORT_SYMBOL(rtrs_srv_get_sess_name);
+
+int rtrs_srv_get_queue_depth(struct rtrs_srv *srv)
+{
+ return srv->queue_depth;
+}
+EXPORT_SYMBOL(rtrs_srv_get_queue_depth);
+
+static int find_next_bit_ring(struct rtrs_srv_sess *sess)
+{
+ struct ib_device *ib_dev = sess->s.dev->ib_dev;
+ int v;
+
+ v = cpumask_next(sess->cur_cq_vector, &cq_affinity_mask);
+ if (v >= nr_cpu_ids || v >= ib_dev->num_comp_vectors)
+ v = cpumask_first(&cq_affinity_mask);
+ return v;
+}
+
+static int rtrs_srv_get_next_cq_vector(struct rtrs_srv_sess *sess)
+{
+ sess->cur_cq_vector = find_next_bit_ring(sess);
+
+ return sess->cur_cq_vector;
+}
+
+static struct rtrs_srv *__alloc_srv(struct rtrs_srv_ctx *ctx,
+ const uuid_t *paths_uuid)
+{
+ struct rtrs_srv *srv;
+ int i;
+
+ srv = kzalloc(sizeof(*srv), GFP_KERNEL);
+ if (unlikely(!srv))
+ return NULL;
+
+ refcount_set(&srv->refcount, 1);
+ INIT_LIST_HEAD(&srv->paths_list);
+ mutex_init(&srv->paths_mutex);
+ mutex_init(&srv->paths_ev_mutex);
+ uuid_copy(&srv->paths_uuid, paths_uuid);
+ srv->queue_depth = sess_queue_depth;
+ srv->ctx = ctx;
+
+ srv->chunks = kcalloc(srv->queue_depth, sizeof(*srv->chunks),
+ GFP_KERNEL);
+ if (unlikely(!srv->chunks))
+ goto err_free_srv;
+
+ for (i = 0; i < srv->queue_depth; i++) {
+ srv->chunks[i] = mempool_alloc(chunk_pool, GFP_KERNEL);
+ if (unlikely(!srv->chunks[i])) {
+ pr_err("mempool_alloc() failed\n");
+ goto err_free_chunks;
+ }
+ }
+ list_add(&srv->ctx_list, &ctx->srv_list);
+
+ return srv;
+
+err_free_chunks:
+ while (i--)
+ mempool_free(srv->chunks[i], chunk_pool);
+ kfree(srv->chunks);
+
+err_free_srv:
+ kfree(srv);
+
+ return NULL;
+}
+
+static void free_srv(struct rtrs_srv *srv)
+{
+ int i;
+
+ WARN_ON(refcount_read(&srv->refcount));
+ for (i = 0; i < srv->queue_depth; i++)
+ mempool_free(srv->chunks[i], chunk_pool);
+ kfree(srv->chunks);
+ /* last put to release the srv structure */
+ put_device(&srv->dev);
+}
+
+static inline struct rtrs_srv *__find_srv_and_get(struct rtrs_srv_ctx *ctx,
+ const uuid_t *paths_uuid)
+{
+ struct rtrs_srv *srv;
+
+ list_for_each_entry(srv, &ctx->srv_list, ctx_list) {
+ if (uuid_equal(&srv->paths_uuid, paths_uuid) &&
+ refcount_inc_not_zero(&srv->refcount))
+ return srv;
+ }
+
+ return NULL;
+}
+
+static struct rtrs_srv *get_or_create_srv(struct rtrs_srv_ctx *ctx,
+ const uuid_t *paths_uuid)
+{
+ struct rtrs_srv *srv;
+
+ mutex_lock(&ctx->srv_mutex);
+ srv = __find_srv_and_get(ctx, paths_uuid);
+ if (!srv)
+ srv = __alloc_srv(ctx, paths_uuid);
+ mutex_unlock(&ctx->srv_mutex);
+
+ return srv;
+}
+
+static void put_srv(struct rtrs_srv *srv)
+{
+ if (refcount_dec_and_test(&srv->refcount)) {
+ struct rtrs_srv_ctx *ctx = srv->ctx;
+
+ WARN_ON(srv->dev.kobj.state_in_sysfs);
+ WARN_ON(srv->kobj_paths.state_in_sysfs);
+
+ mutex_lock(&ctx->srv_mutex);
+ list_del(&srv->ctx_list);
+ mutex_unlock(&ctx->srv_mutex);
+ free_srv(srv);
+ }
+}
+
+static void __add_path_to_srv(struct rtrs_srv *srv,
+ struct rtrs_srv_sess *sess)
+{
+ list_add_tail(&sess->s.entry, &srv->paths_list);
+ srv->paths_num++;
+ WARN_ON(srv->paths_num >= MAX_PATHS_NUM);
+}
+
+static void del_path_from_srv(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+
+ if (WARN_ON(!srv))
+ return;
+
+ mutex_lock(&srv->paths_mutex);
+ list_del(&sess->s.entry);
+ WARN_ON(!srv->paths_num);
+ srv->paths_num--;
+ mutex_unlock(&srv->paths_mutex);
+}
+
+static inline int sockaddr_cmp(const struct sockaddr *a,
+ const struct sockaddr *b)
+{
+ switch (a->sa_family) {
+ case AF_IB:
+ return memcmp(&((struct sockaddr_ib *)a)->sib_addr,
+ &((struct sockaddr_ib *)b)->sib_addr,
+ sizeof(struct ib_addr));
+ case AF_INET:
+ return memcmp(&((struct sockaddr_in *)a)->sin_addr,
+ &((struct sockaddr_in *)b)->sin_addr,
+ sizeof(struct in_addr));
+ case AF_INET6:
+ return memcmp(&((struct sockaddr_in6 *)a)->sin6_addr,
+ &((struct sockaddr_in6 *)b)->sin6_addr,
+ sizeof(struct in6_addr));
+ default:
+ return -ENOENT;
+ }
+}
+
+static inline bool __is_path_w_addr_exists(struct rtrs_srv *srv,
+ struct rdma_addr *addr)
+{
+ struct rtrs_srv_sess *sess;
+
+ list_for_each_entry(sess, &srv->paths_list, s.entry)
+ if (!sockaddr_cmp((struct sockaddr *)&sess->s.dst_addr,
+ (struct sockaddr *)&addr->dst_addr) &&
+ !sockaddr_cmp((struct sockaddr *)&sess->s.src_addr,
+ (struct sockaddr *)&addr->src_addr))
+ return true;
+
+ return false;
+}
+
+static void rtrs_srv_close_work(struct work_struct *work)
+{
+ struct rtrs_srv_sess *sess;
+ struct rtrs_srv_con *con;
+ int i;
+
+ sess = container_of(work, typeof(*sess), close_work);
+
+ rtrs_srv_destroy_sess_files(sess);
+ rtrs_srv_stop_hb(sess);
+
+ for (i = 0; i < sess->s.con_num; i++) {
+ if (!sess->s.con[i])
+ continue;
+ con = to_srv_con(sess->s.con[i]);
+ rdma_disconnect(con->c.cm_id);
+ ib_drain_qp(con->c.qp);
+ }
+ /* Wait for all inflights */
+ rtrs_srv_wait_ops_ids(sess);
+
+ /* Notify upper layer if we are the last path */
+ rtrs_srv_sess_down(sess);
+
+ unmap_cont_bufs(sess);
+ rtrs_srv_free_ops_ids(sess);
+
+ for (i = 0; i < sess->s.con_num; i++) {
+ if (!sess->s.con[i])
+ continue;
+ con = to_srv_con(sess->s.con[i]);
+ rtrs_cq_qp_destroy(&con->c);
+ rdma_destroy_id(con->c.cm_id);
+ kfree(con);
+ }
+ rtrs_ib_dev_put(sess->s.dev);
+
+ del_path_from_srv(sess);
+ put_srv(sess->srv);
+ sess->srv = NULL;
+ rtrs_srv_change_state(sess, RTRS_SRV_CLOSED);
+
+ kfree(sess->dma_addr);
+ kfree(sess->s.con);
+ kfree(sess);
+}
+
+static int rtrs_rdma_do_accept(struct rtrs_srv_sess *sess,
+ struct rdma_cm_id *cm_id)
+{
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_msg_conn_rsp msg;
+ struct rdma_conn_param param;
+ int err;
+
+ memset(¶m, 0, sizeof(param));
+ param.rnr_retry_count = 7;
+ param.private_data = &msg;
+ param.private_data_len = sizeof(msg);
+
+ memset(&msg, 0, sizeof(msg));
+ msg.magic = cpu_to_le16(RTRS_MAGIC);
+ msg.version = cpu_to_le16(RTRS_PROTO_VER);
+ msg.errno = 0;
+ msg.queue_depth = cpu_to_le16(srv->queue_depth);
+ msg.max_io_size = cpu_to_le32(max_chunk_size - MAX_HDR_SIZE);
+ msg.max_hdr_size = cpu_to_le32(MAX_HDR_SIZE);
+
+ if (always_invalidate)
+ msg.flags = cpu_to_le32(RTRS_MSG_NEW_RKEY_F);
+
+ err = rdma_accept(cm_id, ¶m);
+ if (err)
+ pr_err("rdma_accept(), err: %d\n", err);
+
+ return err;
+}
+
+static int rtrs_rdma_do_reject(struct rdma_cm_id *cm_id, int errno)
+{
+ struct rtrs_msg_conn_rsp msg;
+ int err;
+
+ memset(&msg, 0, sizeof(msg));
+ msg.magic = cpu_to_le16(RTRS_MAGIC);
+ msg.version = cpu_to_le16(RTRS_PROTO_VER);
+ msg.errno = cpu_to_le16(errno);
+
+ err = rdma_reject(cm_id, &msg, sizeof(msg));
+ if (err)
+ pr_err("rdma_reject(), err: %d\n", err);
+
+ /* Bounce errno back */
+ return errno;
+}
+
+static struct rtrs_srv_sess *
+__find_sess(struct rtrs_srv *srv, const uuid_t *sess_uuid)
+{
+ struct rtrs_srv_sess *sess;
+
+ list_for_each_entry(sess, &srv->paths_list, s.entry) {
+ if (uuid_equal(&sess->s.uuid, sess_uuid))
+ return sess;
+ }
+
+ return NULL;
+}
+
+static int create_con(struct rtrs_srv_sess *sess,
+ struct rdma_cm_id *cm_id,
+ unsigned int cid)
+{
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_sess *s = &sess->s;
+ struct rtrs_srv_con *con;
+
+ u16 cq_size, wr_queue_size;
+ int err, cq_vector;
+
+ con = kzalloc(sizeof(*con), GFP_KERNEL);
+ if (unlikely(!con)) {
+ rtrs_err(s, "kzalloc() failed\n");
+ err = -ENOMEM;
+ goto err;
+ }
+
+ con->c.cm_id = cm_id;
+ con->c.sess = &sess->s;
+ con->c.cid = cid;
+ atomic_set(&con->wr_cnt, 0);
+
+ if (con->c.cid == 0) {
+ /*
+ * All receive and all send (each requiring invalidate)
+ * + 2 for drain and heartbeat
+ */
+ wr_queue_size = SERVICE_CON_QUEUE_DEPTH * 3 + 2;
+ cq_size = wr_queue_size;
+ } else {
+ /*
+ * If we have all receive requests posted and
+ * all write requests posted and each read request
+ * requires an invalidate request + drain
+ * and qp gets into error state.
+ */
+ cq_size = srv->queue_depth * 3 + 1;
+ /*
+ * In theory we might have queue_depth * 32
+ * outstanding requests if an unsafe global key is used
+ * and we have queue_depth read requests each consisting
+ * of 32 different addresses. div 3 for mlx5.
+ */
+ wr_queue_size = sess->s.dev->ib_dev->attrs.max_qp_wr / 3;
+ }
+
+ cq_vector = rtrs_srv_get_next_cq_vector(sess);
+
+ /* TODO: SOFTIRQ can be faster, but be careful with softirq context */
+ err = rtrs_cq_qp_create(&sess->s, &con->c, 1, cq_vector, cq_size,
+ wr_queue_size, IB_POLL_WORKQUEUE);
+ if (unlikely(err)) {
+ rtrs_err(s, "rtrs_cq_qp_create(), err: %d\n", err);
+ goto free_con;
+ }
+ if (con->c.cid == 0) {
+ err = post_recv_info_req(con);
+ if (unlikely(err))
+ goto free_cqqp;
+ }
+ WARN_ON(sess->s.con[cid]);
+ sess->s.con[cid] = &con->c;
+
+ /*
+ * Change context from server to current connection. The other
+ * way is to use cm_id->qp->qp_context, which does not work on OFED.
+ */
+ cm_id->context = &con->c;
+
+ return 0;
+
+free_cqqp:
+ rtrs_cq_qp_destroy(&con->c);
+free_con:
+ kfree(con);
+
+err:
+ return err;
+}
+
+static struct rtrs_srv_sess *__alloc_sess(struct rtrs_srv *srv,
+ struct rdma_cm_id *cm_id,
+ unsigned int con_num,
+ unsigned int recon_cnt,
+ const uuid_t *uuid)
+{
+ struct rtrs_srv_sess *sess;
+ int err = -ENOMEM;
+
+ if (unlikely(srv->paths_num >= MAX_PATHS_NUM)) {
+ err = -ECONNRESET;
+ goto err;
+ }
+ if (unlikely(__is_path_w_addr_exists(srv, &cm_id->route.addr))) {
+ err = -EEXIST;
+ goto err;
+ }
+ sess = kzalloc(sizeof(*sess), GFP_KERNEL);
+ if (unlikely(!sess))
+ goto err;
+
+ sess->dma_addr = kcalloc(srv->queue_depth, sizeof(*sess->dma_addr),
+ GFP_KERNEL);
+ if (unlikely(!sess->dma_addr))
+ goto err_free_sess;
+
+ sess->s.con = kcalloc(con_num, sizeof(*sess->s.con), GFP_KERNEL);
+ if (unlikely(!sess->s.con))
+ goto err_free_dma_addr;
+
+ sess->state = RTRS_SRV_CONNECTING;
+ sess->srv = srv;
+ sess->cur_cq_vector = -1;
+ sess->s.dst_addr = cm_id->route.addr.dst_addr;
+ sess->s.src_addr = cm_id->route.addr.src_addr;
+ sess->s.con_num = con_num;
+ sess->s.recon_cnt = recon_cnt;
+ uuid_copy(&sess->s.uuid, uuid);
+ spin_lock_init(&sess->state_lock);
+ INIT_WORK(&sess->close_work, rtrs_srv_close_work);
+ rtrs_srv_init_hb(sess);
+
+ sess->s.dev = rtrs_ib_dev_find_or_add(cm_id->device, &dev_pool);
+ if (unlikely(!sess->s.dev)) {
+ err = -ENOMEM;
+ goto err_free_con;
+ }
+ err = map_cont_bufs(sess);
+ if (unlikely(err))
+ goto err_put_dev;
+
+ err = rtrs_srv_alloc_ops_ids(sess);
+ if (unlikely(err))
+ goto err_unmap_bufs;
+
+ __add_path_to_srv(srv, sess);
+
+ return sess;
+
+err_unmap_bufs:
+ unmap_cont_bufs(sess);
+err_put_dev:
+ rtrs_ib_dev_put(sess->s.dev);
+err_free_con:
+ kfree(sess->s.con);
+err_free_dma_addr:
+ kfree(sess->dma_addr);
+err_free_sess:
+ kfree(sess);
+
+err:
+ return ERR_PTR(err);
+}
+
+static int rtrs_rdma_connect(struct rdma_cm_id *cm_id,
+ const struct rtrs_msg_conn_req *msg,
+ size_t len)
+{
+ struct rtrs_srv_ctx *ctx = cm_id->context;
+ struct rtrs_srv_sess *sess;
+ struct rtrs_srv *srv;
+
+ u16 version, con_num, cid;
+ u16 recon_cnt;
+ int err;
+
+ if (unlikely(len < sizeof(*msg))) {
+ pr_err("Invalid RTRS connection request\n");
+ goto reject_w_econnreset;
+ }
+ if (unlikely(le16_to_cpu(msg->magic) != RTRS_MAGIC)) {
+ pr_err("Invalid RTRS magic\n");
+ goto reject_w_econnreset;
+ }
+ version = le16_to_cpu(msg->version);
+ if (unlikely(version >> 8 != RTRS_PROTO_VER_MAJOR)) {
+ pr_err("Unsupported major RTRS version: %d, expected %d\n",
+ version >> 8, RTRS_PROTO_VER_MAJOR);
+ goto reject_w_econnreset;
+ }
+ con_num = le16_to_cpu(msg->cid_num);
+ if (unlikely(con_num > 4096)) {
+ /* Sanity check */
+ pr_err("Too many connections requested: %d\n", con_num);
+ goto reject_w_econnreset;
+ }
+ cid = le16_to_cpu(msg->cid);
+ if (unlikely(cid >= con_num)) {
+ /* Sanity check */
+ pr_err("Incorrect cid: %d >= %d\n", cid, con_num);
+ goto reject_w_econnreset;
+ }
+ recon_cnt = le16_to_cpu(msg->recon_cnt);
+ srv = get_or_create_srv(ctx, &msg->paths_uuid);
+ if (unlikely(!srv)) {
+ err = -ENOMEM;
+ goto reject_w_err;
+ }
+ mutex_lock(&srv->paths_mutex);
+ sess = __find_sess(srv, &msg->sess_uuid);
+ if (sess) {
+ struct rtrs_sess *s = &sess->s;
+
+ /* Session already holds a reference */
+ put_srv(srv);
+
+ if (unlikely(sess->state != RTRS_SRV_CONNECTING)) {
+ rtrs_err(s, "Session in wrong state: %s\n",
+ rtrs_srv_state_str(sess->state));
+ mutex_unlock(&srv->paths_mutex);
+ goto reject_w_econnreset;
+ }
+ /*
+ * Sanity checks
+ */
+ if (unlikely(con_num != sess->s.con_num ||
+ cid >= sess->s.con_num)) {
+ rtrs_err(s, "Incorrect request: %d, %d\n",
+ cid, con_num);
+ mutex_unlock(&srv->paths_mutex);
+ goto reject_w_econnreset;
+ }
+ if (unlikely(sess->s.con[cid])) {
+ rtrs_err(s, "Connection already exists: %d\n",
+ cid);
+ mutex_unlock(&srv->paths_mutex);
+ goto reject_w_econnreset;
+ }
+ } else {
+ sess = __alloc_sess(srv, cm_id, con_num, recon_cnt,
+ &msg->sess_uuid);
+ if (IS_ERR(sess)) {
+ mutex_unlock(&srv->paths_mutex);
+ put_srv(srv);
+ err = PTR_ERR(sess);
+ goto reject_w_err;
+ }
+ }
+ err = create_con(sess, cm_id, cid);
+ if (unlikely(err)) {
+ (void)rtrs_rdma_do_reject(cm_id, err);
+ /*
+ * Since session has other connections we follow normal way
+ * through workqueue, but still return an error to tell cma.c
+ * to call rdma_destroy_id() for current connection.
+ */
+ goto close_and_return_err;
+ }
+ err = rtrs_rdma_do_accept(sess, cm_id);
+ if (unlikely(err)) {
+ (void)rtrs_rdma_do_reject(cm_id, err);
+ /*
+ * Since current connection was successfully added to the
+ * session we follow normal way through workqueue to close the
+ * session, thus return 0 to tell cma.c we call
+ * rdma_destroy_id() ourselves.
+ */
+ err = 0;
+ goto close_and_return_err;
+ }
+ mutex_unlock(&srv->paths_mutex);
+
+ return 0;
+
+reject_w_err:
+ return rtrs_rdma_do_reject(cm_id, err);
+
+reject_w_econnreset:
+ return rtrs_rdma_do_reject(cm_id, -ECONNRESET);
+
+close_and_return_err:
+ close_sess(sess);
+ mutex_unlock(&srv->paths_mutex);
+
+ return err;
+}
+
+static int rtrs_srv_rdma_cm_handler(struct rdma_cm_id *cm_id,
+ struct rdma_cm_event *ev)
+{
+ struct rtrs_srv_sess *sess = NULL;
+ struct rtrs_sess *s = NULL;
+
+ if (ev->event != RDMA_CM_EVENT_CONNECT_REQUEST) {
+ struct rtrs_con *c = cm_id->context;
+
+ s = c->sess;
+ sess = to_srv_sess(s);
+ }
+
+ switch (ev->event) {
+ case RDMA_CM_EVENT_CONNECT_REQUEST:
+ /*
+ * In case of error cma.c will destroy cm_id,
+ * see cma_process_remove()
+ */
+ return rtrs_rdma_connect(cm_id, ev->param.conn.private_data,
+ ev->param.conn.private_data_len);
+ case RDMA_CM_EVENT_ESTABLISHED:
+ /* Nothing here */
+ break;
+ case RDMA_CM_EVENT_REJECTED:
+ case RDMA_CM_EVENT_CONNECT_ERROR:
+ case RDMA_CM_EVENT_UNREACHABLE:
+ rtrs_err(s, "CM error (CM event: %s, err: %d)\n",
+ rdma_event_msg(ev->event), ev->status);
+ close_sess(sess);
+ break;
+ case RDMA_CM_EVENT_DISCONNECTED:
+ case RDMA_CM_EVENT_ADDR_CHANGE:
+ case RDMA_CM_EVENT_TIMEWAIT_EXIT:
+ close_sess(sess);
+ break;
+ case RDMA_CM_EVENT_DEVICE_REMOVAL:
+ close_sess(sess);
+ break;
+ default:
+ pr_err("Ignoring unexpected CM event %s, err %d\n",
+ rdma_event_msg(ev->event), ev->status);
+ break;
+ }
+
+ return 0;
+}
+
+static struct rdma_cm_id *rtrs_srv_cm_init(struct rtrs_srv_ctx *ctx,
+ struct sockaddr *addr,
+ enum rdma_ucm_port_space ps)
+{
+ struct rdma_cm_id *cm_id;
+ int ret;
+
+ cm_id = rdma_create_id(&init_net, rtrs_srv_rdma_cm_handler,
+ ctx, ps, IB_QPT_RC);
+ if (IS_ERR(cm_id)) {
+ ret = PTR_ERR(cm_id);
+ pr_err("Creating id for RDMA connection failed, err: %d\n",
+ ret);
+ goto err_out;
+ }
+ ret = rdma_bind_addr(cm_id, addr);
+ if (ret) {
+ pr_err("Binding RDMA address failed, err: %d\n", ret);
+ goto err_cm;
+ }
+ ret = rdma_listen(cm_id, 64);
+ if (ret) {
+ pr_err("Listening on RDMA connection failed, err: %d\n",
+ ret);
+ goto err_cm;
+ }
+
+ return cm_id;
+
+err_cm:
+ rdma_destroy_id(cm_id);
+err_out:
+
+ return ERR_PTR(ret);
+}
+
+static int rtrs_srv_rdma_init(struct rtrs_srv_ctx *ctx, unsigned int port)
+{
+ struct sockaddr_in6 sin = {
+ .sin6_family = AF_INET6,
+ .sin6_addr = IN6ADDR_ANY_INIT,
+ .sin6_port = htons(port),
+ };
+ struct sockaddr_ib sib = {
+ .sib_family = AF_IB,
+ .sib_addr.sib_subnet_prefix = 0ULL,
+ .sib_addr.sib_interface_id = 0ULL,
+ .sib_sid = cpu_to_be64(RDMA_IB_IP_PS_IB | port),
+ .sib_sid_mask = cpu_to_be64(0xffffffffffffffffULL),
+ .sib_pkey = cpu_to_be16(0xffff),
+ };
+ struct rdma_cm_id *cm_ip, *cm_ib;
+ int ret;
+
+ /*
+ * We accept both IPoIB and IB connections, so we need to keep
+ * two cm id's, one for each socket type and port space.
+ * If the cm initialization of one of the id's fails, we abort
+ * everything.
+ */
+ cm_ip = rtrs_srv_cm_init(ctx, (struct sockaddr *)&sin, RDMA_PS_TCP);
+ if (IS_ERR(cm_ip))
+ return PTR_ERR(cm_ip);
+
+ cm_ib = rtrs_srv_cm_init(ctx, (struct sockaddr *)&sib, RDMA_PS_IB);
+ if (IS_ERR(cm_ib)) {
+ ret = PTR_ERR(cm_ib);
+ goto free_cm_ip;
+ }
+
+ ctx->cm_id_ip = cm_ip;
+ ctx->cm_id_ib = cm_ib;
+
+ return 0;
+
+free_cm_ip:
+ rdma_destroy_id(cm_ip);
+
+ return ret;
+}
+
+static struct rtrs_srv_ctx *alloc_srv_ctx(rdma_ev_fn *rdma_ev,
+ link_ev_fn *link_ev)
+{
+ struct rtrs_srv_ctx *ctx;
+
+ ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+ if (!ctx)
+ return NULL;
+
+ ctx->rdma_ev = rdma_ev;
+ ctx->link_ev = link_ev;
+ mutex_init(&ctx->srv_mutex);
+ INIT_LIST_HEAD(&ctx->srv_list);
+
+ return ctx;
+}
+
+static void free_srv_ctx(struct rtrs_srv_ctx *ctx)
+{
+ WARN_ON(!list_empty(&ctx->srv_list));
+ kfree(ctx);
+}
+
+struct rtrs_srv_ctx *rtrs_srv_open(rdma_ev_fn *rdma_ev, link_ev_fn *link_ev,
+ unsigned int port)
+{
+ struct rtrs_srv_ctx *ctx;
+ int err;
+
+ ctx = alloc_srv_ctx(rdma_ev, link_ev);
+ if (unlikely(!ctx))
+ return ERR_PTR(-ENOMEM);
+
+ err = rtrs_srv_rdma_init(ctx, port);
+ if (unlikely(err)) {
+ free_srv_ctx(ctx);
+ return ERR_PTR(err);
+ }
+ /* Do not let module be unloaded if server context is alive */
+ __module_get(THIS_MODULE);
+
+ return ctx;
+}
+EXPORT_SYMBOL(rtrs_srv_open);
+
+static void close_sessions(struct rtrs_srv *srv)
+{
+ struct rtrs_srv_sess *sess;
+
+ mutex_lock(&srv->paths_mutex);
+ list_for_each_entry(sess, &srv->paths_list, s.entry)
+ close_sess(sess);
+ mutex_unlock(&srv->paths_mutex);
+}
+
+static void close_ctx(struct rtrs_srv_ctx *ctx)
+{
+ struct rtrs_srv *srv;
+
+ mutex_lock(&ctx->srv_mutex);
+ list_for_each_entry(srv, &ctx->srv_list, ctx_list)
+ close_sessions(srv);
+ mutex_unlock(&ctx->srv_mutex);
+ flush_workqueue(rtrs_wq);
+}
+
+void rtrs_srv_close(struct rtrs_srv_ctx *ctx)
+{
+ rdma_destroy_id(ctx->cm_id_ip);
+ rdma_destroy_id(ctx->cm_id_ib);
+ close_ctx(ctx);
+ free_srv_ctx(ctx);
+ module_put(THIS_MODULE);
+}
+EXPORT_SYMBOL(rtrs_srv_close);
+
+static int check_module_params(void)
+{
+ if (sess_queue_depth < 1 || sess_queue_depth > MAX_SESS_QUEUE_DEPTH) {
+ pr_err("Invalid sess_queue_depth value %d, has to be >= %d, <= %d.\n",
+ sess_queue_depth, 1, MAX_SESS_QUEUE_DEPTH);
+ return -EINVAL;
+ }
+ if (max_chunk_size < 4096 || !is_power_of_2(max_chunk_size)) {
+ pr_err("Invalid max_chunk_size value %d, has to be >= %d and should be power of two.\n",
+ max_chunk_size, 4096);
+ return -EINVAL;
+ }
+
+ /*
+ * Check if IB immediate data size is enough to hold the mem_id and the
+ * offset inside the memory chunk
+ */
+ if ((ilog2(sess_queue_depth - 1) + 1) +
+ (ilog2(max_chunk_size - 1) + 1) > MAX_IMM_PAYL_BITS) {
+ pr_err("RDMA immediate size (%db) not enough to encode %d buffers of size %dB. Reduce 'sess_queue_depth' or 'max_chunk_size' parameters.\n",
+ MAX_IMM_PAYL_BITS, sess_queue_depth, max_chunk_size);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int __init rtrs_server_init(void)
+{
+ int err;
+
+ init_cq_affinity();
+
+ pr_info("Loading module %s, proto %s: (cq_affinity_list: %s, max_chunk_size: %d (pure IO %ld, headers %ld) , sess_queue_depth: %d, always_invalidate: %d)\n",
+ KBUILD_MODNAME, RTRS_PROTO_VER_STRING,
+ cq_affinity_list, max_chunk_size,
+ max_chunk_size - MAX_HDR_SIZE, MAX_HDR_SIZE,
+ sess_queue_depth, always_invalidate);
+
+ rtrs_ib_dev_pool_init(0, &dev_pool);
+
+ err = check_module_params();
+ if (err) {
+ pr_err("Failed to load module, invalid module parameters, err: %d\n",
+ err);
+ return err;
+ }
+ chunk_pool = mempool_create_page_pool(sess_queue_depth * CHUNK_POOL_SZ,
+ get_order(max_chunk_size));
+ if (unlikely(!chunk_pool)) {
+ pr_err("Failed preallocate pool of chunks\n");
+ return -ENOMEM;
+ }
+ rtrs_dev_class = class_create(THIS_MODULE, "rtrs-server");
+ if (IS_ERR(rtrs_dev_class)) {
+ pr_err("Failed to create rtrs-server dev class\n");
+ err = PTR_ERR(rtrs_dev_class);
+ goto out_chunk_pool;
+ }
+ rtrs_wq = alloc_workqueue("rtrs_server_wq", WQ_MEM_RECLAIM, 0);
+ if (unlikely(!rtrs_wq)) {
+ pr_err("Failed to load module, alloc rtrs_server_wq failed\n");
+ goto out_dev_class;
+ }
+
+ return 0;
+
+out_dev_class:
+ class_destroy(rtrs_dev_class);
+out_chunk_pool:
+ mempool_destroy(chunk_pool);
+
+ return err;
+}
+
+static void __exit rtrs_server_exit(void)
+{
+ destroy_workqueue(rtrs_wq);
+ class_destroy(rtrs_dev_class);
+ mempool_destroy(chunk_pool);
+ rtrs_ib_dev_pool_deinit(&dev_pool);
+}
+
+module_init(rtrs_server_init);
+module_exit(rtrs_server_exit);
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 11/25] rtrs: server: statistics functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (9 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 10/25] rtrs: server: main functionality Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 12/25] rtrs: server: sysfs interface functions Jack Wang
` (15 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This introduces set of functions used on server side to account
statistics of RDMA data sent/received.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-srv-stats.c | 109 +++++++++++++++++++
1 file changed, 109 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-srv-stats.c
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-srv-stats.c b/drivers/infiniband/ulp/rtrs/rtrs-srv-stats.c
new file mode 100644
index 000000000000..c5632d27fefc
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-srv-stats.c
@@ -0,0 +1,109 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include "rtrs-srv.h"
+
+void rtrs_srv_update_rdma_stats(struct rtrs_srv_stats *s,
+ size_t size, int d)
+{
+ atomic64_inc(&s->rdma_stats.dir[d].cnt);
+ atomic64_add(size, &s->rdma_stats.dir[d].size_total);
+}
+
+void rtrs_srv_update_wc_stats(struct rtrs_srv_stats *s)
+{
+ atomic64_inc(&s->wc_comp.calls);
+ atomic64_inc(&s->wc_comp.total_wc_cnt);
+}
+
+int rtrs_srv_reset_rdma_stats(struct rtrs_srv_stats *stats, bool enable)
+{
+ if (enable) {
+ struct rtrs_srv_stats_rdma_stats *r = &stats->rdma_stats;
+
+ memset(r, 0, sizeof(*r));
+ return 0;
+ }
+
+ return -EINVAL;
+}
+
+ssize_t rtrs_srv_stats_rdma_to_str(struct rtrs_srv_stats *stats,
+ char *page, size_t len)
+{
+ struct rtrs_srv_stats_rdma_stats *r = &stats->rdma_stats;
+ struct rtrs_srv_sess *sess;
+
+ sess = container_of(stats, typeof(*sess), stats);
+
+ return scnprintf(page, len, "%lld %lld %lld %lld %u\n",
+ (s64)atomic64_read(&r->dir[READ].cnt),
+ (s64)atomic64_read(&r->dir[READ].size_total),
+ (s64)atomic64_read(&r->dir[WRITE].cnt),
+ (s64)atomic64_read(&r->dir[WRITE].size_total),
+ atomic_read(&sess->ids_inflight));
+}
+
+int rtrs_srv_reset_wc_completion_stats(struct rtrs_srv_stats *stats,
+ bool enable)
+{
+ if (enable) {
+ memset(&stats->wc_comp, 0, sizeof(stats->wc_comp));
+ return 0;
+ }
+
+ return -EINVAL;
+}
+
+int rtrs_srv_stats_wc_completion_to_str(struct rtrs_srv_stats *stats,
+ char *buf, size_t len)
+{
+ return snprintf(buf, len, "%lld %lld\n",
+ (s64)atomic64_read(&stats->wc_comp.total_wc_cnt),
+ (s64)atomic64_read(&stats->wc_comp.calls));
+}
+
+ssize_t rtrs_srv_reset_all_help(struct rtrs_srv_stats *stats,
+ char *page, size_t len)
+{
+ return scnprintf(page, PAGE_SIZE, "echo 1 to reset all statistics\n");
+}
+
+int rtrs_srv_reset_all_stats(struct rtrs_srv_stats *stats, bool enable)
+{
+ if (enable) {
+ rtrs_srv_reset_wc_completion_stats(stats, enable);
+ rtrs_srv_reset_rdma_stats(stats, enable);
+ return 0;
+ }
+
+ return -EINVAL;
+}
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 12/25] rtrs: server: sysfs interface functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (10 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 11/25] rtrs: server: statistics functions Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 13/25] rtrs: include client and server modules into kernel compilation Jack Wang
` (14 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is the sysfs interface to rtrs sessions on server side:
/sys/devices/virtual/rtrs-server/<SESS-NAME>/
*** rtrs session accepted from a client peer
|
|- paths/<SRC@DST>/
*** established paths from a client in a session
|
|- disconnect
| *** disconnect path
|
|- hca_name
| *** HCA name
|
|- hca_port
| *** HCA port
|
|- stats/
*** current path statistics
|
|- rdma
|- reset_all
|- wc_completions
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/ulp/rtrs/rtrs-srv-sysfs.c | 315 +++++++++++++++++++
1 file changed, 315 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/rtrs-srv-sysfs.c
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-srv-sysfs.c b/drivers/infiniband/ulp/rtrs/rtrs-srv-sysfs.c
new file mode 100644
index 000000000000..07f2caa13cdb
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/rtrs-srv-sysfs.c
@@ -0,0 +1,315 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Transport Layer
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jinpu Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include "rtrs-pri.h"
+#include "rtrs-srv.h"
+#include "rtrs-log.h"
+
+static struct kobj_type ktype = {
+ .sysfs_ops = &kobj_sysfs_ops,
+};
+
+static ssize_t rtrs_srv_disconnect_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ return scnprintf(page, PAGE_SIZE, "Usage: echo 1 > %s\n",
+ attr->attr.name);
+}
+
+static ssize_t rtrs_srv_disconnect_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct rtrs_srv_sess *sess;
+ struct rtrs_sess *s;
+ char str[MAXHOSTNAMELEN];
+
+ sess = container_of(kobj, struct rtrs_srv_sess, kobj);
+ s = &sess->s;
+ if (!sysfs_streq(buf, "1")) {
+ rtrs_err(s, "%s: invalid value: '%s'\n",
+ attr->attr.name, buf);
+ return -EINVAL;
+ }
+
+ sockaddr_to_str((struct sockaddr *)&sess->s.dst_addr, str, sizeof(str));
+
+ rtrs_info(s, "disconnect for path %s requested\n", str);
+ close_sess(sess);
+
+ return count;
+}
+
+static struct kobj_attribute rtrs_srv_disconnect_attr =
+ __ATTR(disconnect, 0644,
+ rtrs_srv_disconnect_show, rtrs_srv_disconnect_store);
+
+static ssize_t rtrs_srv_hca_port_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rtrs_srv_sess *sess;
+ struct rtrs_con *usr_con;
+
+ sess = container_of(kobj, typeof(*sess), kobj);
+ usr_con = sess->s.con[0];
+
+ return scnprintf(page, PAGE_SIZE, "%u\n",
+ usr_con->cm_id->port_num);
+}
+
+static struct kobj_attribute rtrs_srv_hca_port_attr =
+ __ATTR(hca_port, 0444, rtrs_srv_hca_port_show, NULL);
+
+static ssize_t rtrs_srv_hca_name_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rtrs_srv_sess *sess;
+
+ sess = container_of(kobj, struct rtrs_srv_sess, kobj);
+
+ return scnprintf(page, PAGE_SIZE, "%s\n",
+ sess->s.dev->ib_dev->name);
+}
+
+static struct kobj_attribute rtrs_srv_hca_name_attr =
+ __ATTR(hca_name, 0444, rtrs_srv_hca_name_show, NULL);
+
+static ssize_t rtrs_srv_src_addr_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rtrs_srv_sess *sess;
+ int cnt;
+
+ sess = container_of(kobj, struct rtrs_srv_sess, kobj);
+ cnt = sockaddr_to_str((struct sockaddr *)&sess->s.dst_addr,
+ page, PAGE_SIZE);
+ return cnt + scnprintf(page + cnt, PAGE_SIZE - cnt, "\n");
+}
+
+static struct kobj_attribute rtrs_srv_src_addr_attr =
+ __ATTR(src_addr, 0444, rtrs_srv_src_addr_show, NULL);
+
+static ssize_t rtrs_srv_dst_addr_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rtrs_srv_sess *sess;
+ int cnt;
+
+ sess = container_of(kobj, struct rtrs_srv_sess, kobj);
+ cnt = sockaddr_to_str((struct sockaddr *)&sess->s.src_addr,
+ page, PAGE_SIZE);
+ return cnt + scnprintf(page + cnt, PAGE_SIZE - cnt, "\n");
+}
+
+static struct kobj_attribute rtrs_srv_dst_addr_attr =
+ __ATTR(dst_addr, 0444, rtrs_srv_dst_addr_show, NULL);
+
+static struct attribute *rtrs_srv_sess_attrs[] = {
+ &rtrs_srv_hca_name_attr.attr,
+ &rtrs_srv_hca_port_attr.attr,
+ &rtrs_srv_src_addr_attr.attr,
+ &rtrs_srv_dst_addr_attr.attr,
+ &rtrs_srv_disconnect_attr.attr,
+ NULL,
+};
+
+static struct attribute_group rtrs_srv_sess_attr_group = {
+ .attrs = rtrs_srv_sess_attrs,
+};
+
+STAT_ATTR(struct rtrs_srv_sess, rdma,
+ rtrs_srv_stats_rdma_to_str,
+ rtrs_srv_reset_rdma_stats);
+
+STAT_ATTR(struct rtrs_srv_sess, wc_completion,
+ rtrs_srv_stats_wc_completion_to_str,
+ rtrs_srv_reset_wc_completion_stats);
+
+STAT_ATTR(struct rtrs_srv_sess, reset_all,
+ rtrs_srv_reset_all_help,
+ rtrs_srv_reset_all_stats);
+
+static struct attribute *rtrs_srv_stats_attrs[] = {
+ &rdma_attr.attr,
+ &wc_completion_attr.attr,
+ &reset_all_attr.attr,
+ NULL,
+};
+
+static struct attribute_group rtrs_srv_stats_attr_group = {
+ .attrs = rtrs_srv_stats_attrs,
+};
+
+static void rtrs_srv_dev_release(struct device *dev)
+{
+ struct rtrs_srv *srv = container_of(dev, struct rtrs_srv, dev);
+
+ kfree(srv);
+}
+
+static int rtrs_srv_create_once_sysfs_root_folders(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+ int err = 0;
+
+ mutex_lock(&srv->paths_mutex);
+ if (srv->dev_ref++) {
+ /*
+ * Just increase device reference. We can't use get_device()
+ * because we need to unregister device when ref goes to 0,
+ * not just to put it.
+ */
+ goto unlock;
+ }
+ srv->dev.class = rtrs_dev_class;
+ srv->dev.release = rtrs_srv_dev_release;
+ dev_set_name(&srv->dev, "%s", sess->s.sessname);
+
+ err = device_register(&srv->dev);
+ if (unlikely(err)) {
+ pr_err("device_register(): %d\n", err);
+ goto unlock;
+ }
+ err = kobject_init_and_add(&srv->kobj_paths, &ktype,
+ &srv->dev.kobj, "paths");
+ if (unlikely(err)) {
+ pr_err("kobject_init_and_add(): %d\n", err);
+ device_unregister(&srv->dev);
+ goto unlock;
+ }
+unlock:
+ mutex_unlock(&srv->paths_mutex);
+
+ return err;
+}
+
+static void
+rtrs_srv_destroy_once_sysfs_root_folders(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+
+ mutex_lock(&srv->paths_mutex);
+ if (!--srv->dev_ref) {
+ kobject_del(&srv->kobj_paths);
+ kobject_put(&srv->kobj_paths);
+ device_unregister(&srv->dev);
+ }
+ mutex_unlock(&srv->paths_mutex);
+}
+
+static int rtrs_srv_create_stats_files(struct rtrs_srv_sess *sess)
+{
+ int err;
+ struct rtrs_sess *s = &sess->s;
+
+ err = kobject_init_and_add(&sess->kobj_stats, &ktype,
+ &sess->kobj, "stats");
+ if (unlikely(err)) {
+ rtrs_err(s, "kobject_init_and_add(): %d\n", err);
+ return err;
+ }
+ err = sysfs_create_group(&sess->kobj_stats,
+ &rtrs_srv_stats_attr_group);
+ if (unlikely(err)) {
+ rtrs_err(s, "sysfs_create_group(): %d\n", err);
+ goto err;
+ }
+
+ return 0;
+
+err:
+ kobject_del(&sess->kobj_stats);
+ kobject_put(&sess->kobj_stats);
+
+ return err;
+}
+
+int rtrs_srv_create_sess_files(struct rtrs_srv_sess *sess)
+{
+ struct rtrs_srv *srv = sess->srv;
+ struct rtrs_sess *s = &sess->s;
+ char str[NAME_MAX];
+ int err, cnt;
+
+ cnt = sockaddr_to_str((struct sockaddr *)&sess->s.dst_addr,
+ str, sizeof(str));
+ cnt += scnprintf(str + cnt, sizeof(str) - cnt, "@");
+ sockaddr_to_str((struct sockaddr *)&sess->s.src_addr,
+ str + cnt, sizeof(str) - cnt);
+
+ err = rtrs_srv_create_once_sysfs_root_folders(sess);
+ if (unlikely(err))
+ return err;
+
+ err = kobject_init_and_add(&sess->kobj, &ktype, &srv->kobj_paths,
+ "%s", str);
+ if (unlikely(err)) {
+ rtrs_err(s, "kobject_init_and_add(): %d\n", err);
+ goto destroy_root;
+ }
+ err = sysfs_create_group(&sess->kobj, &rtrs_srv_sess_attr_group);
+ if (unlikely(err)) {
+ rtrs_err(s, "sysfs_create_group(): %d\n", err);
+ goto put_kobj;
+ }
+ err = rtrs_srv_create_stats_files(sess);
+ if (unlikely(err))
+ goto remove_group;
+
+ return 0;
+
+remove_group:
+ sysfs_remove_group(&sess->kobj, &rtrs_srv_sess_attr_group);
+put_kobj:
+ kobject_del(&sess->kobj);
+ kobject_put(&sess->kobj);
+destroy_root:
+ rtrs_srv_destroy_once_sysfs_root_folders(sess);
+
+ return err;
+}
+
+void rtrs_srv_destroy_sess_files(struct rtrs_srv_sess *sess)
+{
+ if (sess->kobj.state_in_sysfs) {
+ kobject_del(&sess->kobj_stats);
+ kobject_put(&sess->kobj_stats);
+ kobject_del(&sess->kobj);
+ kobject_put(&sess->kobj);
+
+ rtrs_srv_destroy_once_sysfs_root_folders(sess);
+ }
+}
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 13/25] rtrs: include client and server modules into kernel compilation
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (11 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 12/25] rtrs: server: sysfs interface functions Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 14/25] rtrs: a bit of documentation Jack Wang
` (13 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
Add rtrs Makefile, Kconfig and also corresponding lines into upper
layer infiniband/ulp files.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/infiniband/Kconfig | 1 +
drivers/infiniband/ulp/Makefile | 1 +
drivers/infiniband/ulp/rtrs/Kconfig | 27 +++++++++++++++++++++++++++
drivers/infiniband/ulp/rtrs/Makefile | 17 +++++++++++++++++
4 files changed, 46 insertions(+)
create mode 100644 drivers/infiniband/ulp/rtrs/Kconfig
create mode 100644 drivers/infiniband/ulp/rtrs/Makefile
diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig
index ade86388434f..477418b37786 100644
--- a/drivers/infiniband/Kconfig
+++ b/drivers/infiniband/Kconfig
@@ -107,6 +107,7 @@ source "drivers/infiniband/ulp/srpt/Kconfig"
source "drivers/infiniband/ulp/iser/Kconfig"
source "drivers/infiniband/ulp/isert/Kconfig"
+source "drivers/infiniband/ulp/rtrs/Kconfig"
source "drivers/infiniband/ulp/opa_vnic/Kconfig"
diff --git a/drivers/infiniband/ulp/Makefile b/drivers/infiniband/ulp/Makefile
index 437813c7b481..4d0004b58377 100644
--- a/drivers/infiniband/ulp/Makefile
+++ b/drivers/infiniband/ulp/Makefile
@@ -5,3 +5,4 @@ obj-$(CONFIG_INFINIBAND_SRPT) += srpt/
obj-$(CONFIG_INFINIBAND_ISER) += iser/
obj-$(CONFIG_INFINIBAND_ISERT) += isert/
obj-$(CONFIG_INFINIBAND_OPA_VNIC) += opa_vnic/
+obj-$(CONFIG_INFINIBAND_RTRS) += rtrs/
diff --git a/drivers/infiniband/ulp/rtrs/Kconfig b/drivers/infiniband/ulp/rtrs/Kconfig
new file mode 100644
index 000000000000..1d6c670a4504
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/Kconfig
@@ -0,0 +1,27 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+
+config INFINIBAND_RTRS
+ tristate
+ depends on INFINIBAND_ADDR_TRANS
+
+config INFINIBAND_RTRS_CLIENT
+ tristate "RTRS client module"
+ depends on INFINIBAND_ADDR_TRANS
+ select INFINIBAND_RTRS
+ help
+ RDMA transport client module.
+
+ RTRS client allows for simplified data transfer and connection
+ establishment over RDMA (InfiniBand, RoCE, iWarp). Uses BIO-like
+ READ/WRITE semantics and provides multipath capabilities.
+
+config INFINIBAND_RTRS_SERVER
+ tristate "RTRS server module"
+ depends on INFINIBAND_ADDR_TRANS
+ select INFINIBAND_RTRS
+ help
+ RDMA transport server module.
+
+ RTRS server module processing connection and IO requests received
+ from the RTRS client module, it will pass the IO requests to its
+ user eg. RNBD_server.
diff --git a/drivers/infiniband/ulp/rtrs/Makefile b/drivers/infiniband/ulp/rtrs/Makefile
new file mode 100644
index 000000000000..89332be15c9e
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/Makefile
@@ -0,0 +1,17 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+
+rtrs-client-y := rtrs-clt.o \
+ rtrs-clt-stats.o \
+ rtrs-clt-sysfs.o
+
+rtrs-server-y := rtrs-srv.o \
+ rtrs-srv-stats.o \
+ rtrs-srv-sysfs.o
+
+rtrs-core-y := rtrs.o
+
+obj-$(CONFIG_INFINIBAND_RTRS) += rtrs-core.o
+obj-$(CONFIG_INFINIBAND_RTRS_CLIENT) += rtrs-client.o
+obj-$(CONFIG_INFINIBAND_RTRS_SERVER) += rtrs-server.o
+
+-include $(src)/compat/compat.mk
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 14/25] rtrs: a bit of documentation
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (12 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 13/25] rtrs: include client and server modules into kernel compilation Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:50 ` [PATCH v5 15/25] rnbd: private headers with rnbd protocol structs and helpers Jack Wang
` (12 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev, linux-kernel
From: Jack Wang <jinpu.wang@cloud.ionos.com>
README with description of major sysfs entries, sysfs documentation
has been moved to ABI dir as suggested by Bart.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
Cc: linux-kernel@vger.kernel.org
---
.../ABI/testing/sysfs-class-rtrs-client | 190 ++++++++++++++++++
.../ABI/testing/sysfs-class-rtrs-server | 81 ++++++++
drivers/infiniband/ulp/rtrs/README | 137 +++++++++++++
3 files changed, 408 insertions(+)
create mode 100644 Documentation/ABI/testing/sysfs-class-rtrs-client
create mode 100644 Documentation/ABI/testing/sysfs-class-rtrs-server
create mode 100644 drivers/infiniband/ulp/rtrs/README
diff --git a/Documentation/ABI/testing/sysfs-class-rtrs-client b/Documentation/ABI/testing/sysfs-class-rtrs-client
new file mode 100644
index 000000000000..8b219cf6c5c4
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-class-rtrs-client
@@ -0,0 +1,190 @@
+What: /sys/class/rtrs-client
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+When a user of RTRS API creates a new session, a directory entry with
+the name of that session is created under /sys/class/rtrs-client/<session-name>/
+
+What: /sys/class/rtrs-client/<session-name>/add_path
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RW, adds a new path (connection) to an existing session. Expected format is the
+following:
+
+ <[source addr,]destination addr>
+
+ *addr ::= [ ip:<ipv4|ipv6> | gid:<gid> ]
+
+What: /sys/class/rtrs-client/<session-name>/max_reconnect_attempts
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Maximum number reconnect attempts the client should make before giving up
+after connection breaks unexpectedly.
+
+What: /sys/class/rtrs-client/<session-name>/mp_policy
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Multipath policy specifies which path should be selected on each IO:
+
+ round-robin (0):
+ select path in per CPU round-robin manner.
+
+ min-inflight (1):
+ select path with minimum inflights.
+
+What: /sys/class/rtrs-client/<session-name>/paths/
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Each path belonging to a given session is listed here by its source and
+destination address. When a new path is added to a session by writing to
+the "add_path" entry, a directory <src@dst> is created.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/state
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains "connected" if the session is connected to the peer and fully
+functional. Otherwise the file contains "disconnected"
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/reconnect
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Write "1" to the file in order to reconnect the path.
+Operation is blocking and returns 0 if reconnect was successful.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/disconnect
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Write "1" to the file in order to disconnect the path.
+Operation blocks until RTRS path is disconnected.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/remove_path
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Write "1" to the file in order to disconnected and remove the path
+from the session. Operation blocks until the path is disconnected
+and removed from the session.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/hca_name
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains the the name of HCA the connection established on.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/hca_port
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains the port number of active port traffic is going through.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/src_addr
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains the source address of the path
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/dst_addr
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains the destination address of the path
+
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/stats/reset_all
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RW, Read will return usage help, write 0 will clear all the statistics.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/stats/sg_entries
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Data to be transferred via RDMA is passed to RTRS as scatter-gather
+list. A scatter-gather list can contain multiple entries.
+Scatter-gather list with less entries require less processing power
+and can therefore transferred faster. The file sg_entries outputs a
+per-CPU distribution table for the number of entries in the
+scatter-gather lists, that were passed to the RTRS API function
+rtrs_clt_request (READ or WRITE).
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/stats/cpu_migration
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RTRS expects that each HCA IRQ is pinned to a separate CPU. If it's
+not the case, the processing of an I/O response could be processed on a
+different CPU than where it was originally submitted. This file shows
+how many interrupts where generated on a non expected CPU.
+"from:" is the CPU on which the IRQ was expected, but not generated.
+"to:" is the CPU on which the IRQ was generated, but not expected.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/stats/reconnects
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains 2 unsigned int values, the first one records number of successful
+reconnects in the path lifetime, the second one records number of failed
+reconnects in the path lifetime.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/stats/rdma_lat
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Latency distribution of RTRS requests.
+The format is:
+ 1 ms: <CNT-LAT-READ> <CNT-LAT-WRITE>
+ 2 ms: <CNT-LAT-READ> <CNT-LAT-WRITE>
+ 4 ms: <CNT-LAT-READ> <CNT-LAT-WRITE>
+ 8 ms: <CNT-LAT-READ> <CNT-LAT-WRITE>
+ 16 ms: <CNT-LAT-READ> <CNT-LAT-WRITE>
+ ...
+ 65536 ms: <CNT-LAT-READ> <CNT-LAT-WRITE>
+ >= 65536 ms: <CNT-LAT-READ> <CNT-LAT-WRITE>
+ maximum ms: <CNT-LAT-READ> <CNT-LAT-WRITE>
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/stats/wc_completion
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains 2 unsigned int values, the first one records max number of work
+requests processed in work_completion in session lifetime, the second
+one records average number of work requests processed in work_completion
+in session lifetime.
+
+What: /sys/class/rtrs-client/<session-name>/paths/<src@dst>/stats/rdma
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains statistics regarding rdma operations and inflight operations.
+The output consists of 6 values:
+
+<read-count> <read-total-size> <write-count> <write-total-size> \
+<inflights> <failovered>
diff --git a/Documentation/ABI/testing/sysfs-class-rtrs-server b/Documentation/ABI/testing/sysfs-class-rtrs-server
new file mode 100644
index 000000000000..cac2a093d56f
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-class-rtrs-server
@@ -0,0 +1,81 @@
+What: /sys/class/rtrs-server
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+When a user of RTRS API creates a new session on a client side, a
+directory entry with the name of that session is created in here.
+
+What: /sys/class/rtrs-server/<session-name>/paths/
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+When new path is created by writing to "add_path" entry on client side,
+a directory entry named as <source address>@<destination address> is created
+on server.
+
+What: /sys/class/rtrs-server/<session-name>/paths/<src@dst>/disconnect
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+When "1" is written to the file, the RTRS session is being disconnected.
+Operations is non-blocking and returns control immediately to the caller.
+
+What: /sys/class/rtrs-server/<session-name>/paths/<src@dst>/hca_name
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains the the name of HCA the connection established on.
+
+What: /sys/class/rtrs-server/<session-name>/paths/<src@dst>/hca_port
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains the port number of active port traffic is going through.
+
+What: /sys/class/rtrs-server/<session-name>/paths/<src@dst>/src_addr
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains the source address of the path
+
+What: /sys/class/rtrs-server/<session-name>/paths/<src@dst>/dst_addr
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RO, Contains the destination address of the path
+
+What: /sys/class/rtrs-server/<session-name>/paths/<src@dst>/stats/reset_all
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RW, Read will return usage help, write 0 will clear all the statistics.
+
+What: /sys/class/rtrs-server/<session-name>/paths/<src@dst>/stats/rdma
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains statistics regarding rdma operations and inflight operations.
+The output consists of 5 values:
+<read-count> <read-total-size> <write-count> <write-total-size> <inflights>
+
+What: /sys/class/rtrs-server/<session-name>/paths/<src@dst>/stats/wc_completion
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains 3 values, the first one is int, records max number of work
+requests processed in work_completion in session lifetime, the second
+one long int records total number of work requests processed in
+work_completion in session lifetime and the 3rd one long int records
+total number of calls to the cq completion handler. Division of 2nd
+number through 3rd gives the average number of completions processed
+in completion handler.
diff --git a/drivers/infiniband/ulp/rtrs/README b/drivers/infiniband/ulp/rtrs/README
new file mode 100644
index 000000000000..e4985070e3c0
--- /dev/null
+++ b/drivers/infiniband/ulp/rtrs/README
@@ -0,0 +1,137 @@
+****************************
+InfiniBand Transport (RTRS)
+****************************
+
+RTRS (InfiniBand Transport) is a reliable high speed transport library
+which provides support to establish optimal number of connections
+between client and server machines using RDMA (InfiniBand, RoCE, iWarp)
+transport. It is optimized to transfer (read/write) IO blocks.
+
+In its core interface it follows the BIO semantics of providing the
+possibility to either write data from an sg list to the remote side
+or to request ("read") data transfer from the remote side into a given
+sg list.
+
+RTRS provides I/O fail-over and load-balancing capabilities by using
+multipath I/O (see "add_path" and "mp_policy" configuration entries).
+
+RTRS is used by the RNBD (Infiniband Network Block Device) modules.
+
+==================
+Transport protocol
+==================
+
+Overview
+--------
+An established connection between a client and a server is called rtrs
+session. A session is associated with a set of memory chunks reserved on the
+server side for a given client for rdma transfer. A session
+consists of multiple paths, each representing a separate physical link
+between client and server. Those are used for load balancing and failover.
+Each path consists of as many connections (QPs) as there are cpus on
+the client.
+
+When processing an incoming rdma write or read request rtrs client uses memory
+chunks reserved for him on the server side. Their number, size and addresses
+need to be exchanged between client and server during the connection
+establishment phase. Apart from the memory related information client needs to
+inform the server about the session name and identify each path and connection
+individually.
+
+On an established session client sends to server write or read messages.
+Server uses immediate field to tell the client which request is being
+acknowledged and for errno. Client uses immediate field to tell the server
+which of the memory chunks has been accessed and at which offset the message
+can be found.
+
+Connection establishment
+------------------------
+
+1. Client starts establishing connections belonging to a path of a session one
+by one via attaching RTRS_MSG_CON_REQ messages to the rdma_connect requests.
+Those include uuid of the session and uuid of the path to be
+established. They are used by the server to find a persisting session/path or
+to create a new one when necessary. The message also contains the protocol
+version and magic for compatibility, total number of connections per session
+(as many as cpus on the client), the id of the current connection and
+the reconnect counter, which is used to resolve the situations where
+client is trying to reconnect a path, while server is still destroying the old
+one.
+
+2. Server accepts the connection requests one by one and attaches
+RTRS_MSG_CONN_RSP messages to the rdma_accept. Apart from magic and
+protocol version, the messages include error code, queue depth supported by
+the server (number of memory chunks which are going to be allocated for that
+session) and the maximum size of one io.
+
+3. After all connections of a path are established client sends to server the
+RTRS_MSG_INFO_REQ message, containing the name of the session. This message
+requests the address information from the server.
+
+4. Server replies to the session info request message with RTRS_MSG_INFO_RSP,
+which contains the addresses and keys of the RDMA buffers allocated for that
+session.
+
+5. Session becomes connected after all paths to be established are connected
+(i.e. steps 1-4 finished for all paths requested for a session)
+
+6. Server and client exchange periodically heartbeat messages (empty rdma
+messages with an immediate field) which are used to detect a crash on remote
+side or network outage in an absence of IO.
+
+7. On any RDMA related error or in the case of a heartbeat timeout, the
+corresponding path is disconnected, all the inflight IO are failed over to a
+healthy path, if any, and the reconnect mechanism is triggered.
+
+CLT SRV
+*for each connection belonging to a path and for each path:
+RTRS_MSG_CON_REQ ------------------->
+ <------------------- RTRS_MSG_CON_RSP
+...
+*after all connections are established:
+RTRS_MSG_INFO_REQ ------------------->
+ <------------------- RTRS_MSG_INFO_RSP
+*heartbeat is started from both sides:
+ -------------------> [RTRS_HB_MSG_IMM]
+[RTRS_HB_MSG_ACK] <-------------------
+[RTRS_HB_MSG_IMM] <-------------------
+ -------------------> [RTRS_HB_MSG_ACK]
+
+IO path
+-------
+
+* Write *
+
+1. When processing a write request client selects one of the memory chunks
+on the server side and rdma writes there the user data, user header and the
+RTRS_MSG_RDMA_WRITE message. Apart from the type (write), the message only
+contains size of the user header. The client tells the server which chunk has
+been accessed and at what offset the RTRS_MSG_RDMA_WRITE can be found by
+using the IMM field.
+
+2. When confirming a write request server sends an "empty" rdma message with
+an immediate field. The 32 bit field is used to specify the outstanding
+inflight IO and for the error code.
+
+CLT SRV
+usr_data + usr_hdr + rtrs_msg_rdma_write -----------------> [RTRS_IO_REQ_IMM]
+[RTRS_IO_RSP_IMM] <----------------- (id + errno)
+
+* Read *
+
+1. When processing a read request client selects one of the memory chunks
+on the server side and rdma writes there the user header and the
+RTRS_MSG_RDMA_READ message. This message contains the type (read), size of
+the user header, flags (specifying if memory invalidation is necessary) and the
+list of addresses along with keys for the data to be read into.
+
+2. When confirming a read request server transfers the requested data first,
+attaches an invalidation message if requested and finally an "empty" rdma
+message with an immediate field. The 32 bit field is used to specify the
+outstanding inflight IO and the error code.
+
+CLT SRV
+usr_hdr + rtrs_msg_rdma_read --------------> [RTRS_IO_REQ_IMM]
+[RTRS_IO_RSP_IMM] <-------------- usr_data + (id + errno)
+or in case client requested invalidation:
+[RTRS_IO_RSP_IMM_W_INV] <-------------- usr_data + (INV) + (id + errno)
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 15/25] rnbd: private headers with rnbd protocol structs and helpers
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (13 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 14/25] rtrs: a bit of documentation Jack Wang
@ 2019-12-20 15:50 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 16/25] rnbd: client: private header with client structs and functions Jack Wang
` (11 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:50 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
These are common private headers with rnbd protocol structures,
logging, sysfs and other helper functions, which are used on
both client and server sides.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/rnbd/rnbd-common.c | 43 ++++
drivers/block/rnbd/rnbd-log.h | 61 ++++++
drivers/block/rnbd/rnbd-proto.h | 325 +++++++++++++++++++++++++++++++
3 files changed, 429 insertions(+)
create mode 100644 drivers/block/rnbd/rnbd-common.c
create mode 100644 drivers/block/rnbd/rnbd-log.h
create mode 100644 drivers/block/rnbd/rnbd-proto.h
diff --git a/drivers/block/rnbd/rnbd-common.c b/drivers/block/rnbd/rnbd-common.c
new file mode 100644
index 000000000000..713aad8e3571
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-common.c
@@ -0,0 +1,43 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#include "rnbd-proto.h"
+
+const char *rnbd_access_mode_str(enum rnbd_access_mode mode)
+{
+ switch (mode) {
+ case RNBD_ACCESS_RO:
+ return "ro";
+ case RNBD_ACCESS_RW:
+ return "rw";
+ case RNBD_ACCESS_MIGRATION:
+ return "migration";
+ default:
+ return "unknown";
+ }
+}
diff --git a/drivers/block/rnbd/rnbd-log.h b/drivers/block/rnbd/rnbd-log.h
new file mode 100644
index 000000000000..3a0eb6fa2346
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-log.h
@@ -0,0 +1,61 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#ifndef RNBD_LOG_H
+#define RNBD_LOG_H
+
+#include "rnbd-clt.h"
+#include "rnbd-srv.h"
+
+#define rnbd_clt_log(fn, dev, fmt, ...) ( \
+ fn("<%s@%s> " fmt, (dev)->pathname, \
+ (dev)->sess->sessname, \
+ ##__VA_ARGS__))
+#define rnbd_srv_log(fn, dev, fmt, ...) ( \
+ fn("<%s@%s>: " fmt, (dev)->pathname, \
+ (dev)->sess->sessname, ##__VA_ARGS__))
+
+#define rnbd_clt_err(dev, fmt, ...) \
+ rnbd_clt_log(pr_err, dev, fmt, ##__VA_ARGS__)
+#define rnbd_clt_err_rl(dev, fmt, ...) \
+ rnbd_clt_log(pr_err_ratelimited, dev, fmt, ##__VA_ARGS__)
+#define rnbd_clt_info(dev, fmt, ...) \
+ rnbd_clt_log(pr_info, dev, fmt, ##__VA_ARGS__)
+#define rnbd_clt_info_rl(dev, fmt, ...) \
+ rnbd_clt_log(pr_info_ratelimited, dev, fmt, ##__VA_ARGS__)
+
+#define rnbd_srv_err(dev, fmt, ...) \
+ rnbd_srv_log(pr_err, dev, fmt, ##__VA_ARGS__)
+#define rnbd_srv_err_rl(dev, fmt, ...) \
+ rnbd_srv_log(pr_err_ratelimited, dev, fmt, ##__VA_ARGS__)
+#define rnbd_srv_info(dev, fmt, ...) \
+ rnbd_srv_log(pr_info, dev, fmt, ##__VA_ARGS__)
+#define rnbd_srv_info_rl(dev, fmt, ...) \
+ rnbd_srv_log(pr_info_ratelimited, dev, fmt, ##__VA_ARGS__)
+
+#endif /* RNBD_LOG_H */
diff --git a/drivers/block/rnbd/rnbd-proto.h b/drivers/block/rnbd/rnbd-proto.h
new file mode 100644
index 000000000000..6369089d1134
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-proto.h
@@ -0,0 +1,325 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#ifndef RNBD_PROTO_H
+#define RNBD_PROTO_H
+
+#include <linux/types.h>
+#include <linux/blkdev.h>
+#include <linux/limits.h>
+#include <linux/inet.h>
+#include <linux/in.h>
+#include <linux/in6.h>
+#include <rdma/ib.h>
+
+#define RNBD_PROTO_VER_MAJOR 2
+#define RNBD_PROTO_VER_MINOR 0
+
+#define RNBD_PROTO_VER_STRING __stringify(RNBD_PROTO_VER_MAJOR) "." \
+ __stringify(RNBD_PROTO_VER_MINOR)
+
+#define RTRS_PORT 1234
+
+/**
+ * enum rnbd_msg_types - RNBD message types
+ * @RNBD_MSG_SESS_INFO: initial session info from client to server
+ * @RNBD_MSG_SESS_INFO_RSP: initial session info from server to client
+ * @RNBD_MSG_OPEN: open (map) device request
+ * @RNBD_MSG_OPEN_RSP: response to an @RNBD_MSG_OPEN
+ * @RNBD_MSG_IO: block IO request operation
+ * @RNBD_MSG_CLOSE: close (unmap) device request
+ */
+enum rnbd_msg_type {
+ RNBD_MSG_SESS_INFO,
+ RNBD_MSG_SESS_INFO_RSP,
+ RNBD_MSG_OPEN,
+ RNBD_MSG_OPEN_RSP,
+ RNBD_MSG_IO,
+ RNBD_MSG_CLOSE,
+};
+
+/**
+ * struct rnbd_msg_hdr - header of RNBD messages
+ * @type: Message type, valid values see: enum rnbd_msg_types
+ */
+struct rnbd_msg_hdr {
+ __le16 type;
+ __le16 __padding;
+};
+
+/**
+ * We allow to map RO many times and RW only once. We allow to map yet another
+ * time RW, if MIGRATION is provided (second RW export can be required for
+ * example for VM migration)
+ */
+enum rnbd_access_mode {
+ RNBD_ACCESS_RO,
+ RNBD_ACCESS_RW,
+ RNBD_ACCESS_MIGRATION,
+};
+
+/**
+ * struct rnbd_msg_sess_info - initial session info from client to server
+ * @hdr: message header
+ * @ver: RNBD protocol version
+ */
+struct rnbd_msg_sess_info {
+ struct rnbd_msg_hdr hdr;
+ u8 ver;
+ u8 reserved[31];
+};
+
+/**
+ * struct rnbd_msg_sess_info_rsp - initial session info from server to client
+ * @hdr: message header
+ * @ver: RNBD protocol version
+ */
+struct rnbd_msg_sess_info_rsp {
+ struct rnbd_msg_hdr hdr;
+ u8 ver;
+ u8 reserved[31];
+};
+
+/**
+ * struct rnbd_msg_open - request to open a remote device.
+ * @hdr: message header
+ * @access_mode: the mode to open remote device, valid values see:
+ * enum rnbd_access_mode
+ * @device_name: device path on remote side
+ */
+struct rnbd_msg_open {
+ struct rnbd_msg_hdr hdr;
+ u8 access_mode;
+ u8 resv1;
+ s8 dev_name[NAME_MAX];
+ u8 reserved[3];
+};
+
+/**
+ * struct rnbd_msg_close - request to close a remote device.
+ * @hdr: message header
+ * @device_id: device_id on server side to identify the device
+ */
+struct rnbd_msg_close {
+ struct rnbd_msg_hdr hdr;
+ __le32 device_id;
+};
+
+/**
+ * struct rnbd_msg_open_rsp - response message to RNBD_MSG_OPEN
+ * @hdr: message header
+ * @device_id: device_id on server side to identify the device
+ * @nsectors: number of sectors in the usual 512b unit
+ * @max_hw_sectors: max hardware sectors in the usual 512b unit
+ * @max_write_same_sectors: max sectors for WRITE SAME in the 512b unit
+ * @max_discard_sectors: max. sectors that can be discarded at once in 512b
+ * unit.
+ * @discard_granularity: size of the internal discard allocation unit in bytes
+ * @discard_alignment: offset from internal allocation assignment in bytes
+ * @physical_block_size: physical block size device supports in bytes
+ * @logical_block_size: logical block size device supports in bytes
+ * @max_segments: max segments hardware support in one transfer
+ * @secure_discard: supports secure discard
+ * @rotation: is a rotational disc?
+ */
+struct rnbd_msg_open_rsp {
+ struct rnbd_msg_hdr hdr;
+ __le32 device_id;
+ __le64 nsectors;
+ __le32 max_hw_sectors;
+ __le32 max_write_same_sectors;
+ __le32 max_discard_sectors;
+ __le32 discard_granularity;
+ __le32 discard_alignment;
+ __le16 physical_block_size;
+ __le16 logical_block_size;
+ __le16 max_segments;
+ __le16 secure_discard;
+ u8 rotational;
+ u8 reserved[11];
+};
+
+/**
+ * struct rnbd_msg_io - message for I/O read/write
+ * @hdr: message header
+ * @device_id: device_id on server side to find the right device
+ * @sector: bi_sector attribute from struct bio
+ * @rw: valid values are defined in enum rnbd_io_flags
+ * @bi_size: number of bytes for I/O read/write
+ * @prio: priority
+ */
+struct rnbd_msg_io {
+ struct rnbd_msg_hdr hdr;
+ __le32 device_id;
+ __le64 sector;
+ __le32 rw;
+ __le32 bi_size;
+ __le16 prio;
+};
+
+#define RNBD_OP_BITS 8
+#define RNBD_OP_MASK ((1 << RNBD_OP_BITS) - 1)
+
+/**
+ * enum rnbd_io_flags - RNBD request types from rq_flag_bits
+ * @RNBD_OP_READ: read sectors from the device
+ * @RNBD_OP_WRITE: write sectors to the device
+ * @RNBD_OP_FLUSH: flush the volatile write cache
+ * @RNBD_OP_DISCARD: discard sectors
+ * @RNBD_OP_SECURE_ERASE: securely erase sectors
+ * @RNBD_OP_WRITE_SAME: write the same sectors many times
+
+ * @RNBD_F_SYNC: request is sync (sync write or read)
+ * @RNBD_F_FUA: forced unit access
+ */
+enum rnbd_io_flags {
+
+ /* Operations */
+
+ RNBD_OP_READ = 0,
+ RNBD_OP_WRITE = 1,
+ RNBD_OP_FLUSH = 2,
+ RNBD_OP_DISCARD = 3,
+ RNBD_OP_SECURE_ERASE = 4,
+ RNBD_OP_WRITE_SAME = 5,
+
+ RNBD_OP_LAST,
+
+ /* Flags */
+
+ RNBD_F_SYNC = 1<<(RNBD_OP_BITS + 0),
+ RNBD_F_FUA = 1<<(RNBD_OP_BITS + 1),
+
+ RNBD_F_ALL = (RNBD_F_SYNC | RNBD_F_FUA)
+
+};
+
+static inline u32 rnbd_op(u32 flags)
+{
+ return (flags & RNBD_OP_MASK);
+}
+
+static inline u32 rnbd_flags(u32 flags)
+{
+ return (flags & ~RNBD_OP_MASK);
+}
+
+static inline bool rnbd_flags_supported(u32 flags)
+{
+ u32 op;
+
+ op = rnbd_op(flags);
+ flags = rnbd_flags(flags);
+
+ if (op >= RNBD_OP_LAST)
+ return false;
+ if (flags & ~RNBD_F_ALL)
+ return false;
+
+ return true;
+}
+
+static inline u32 rnbd_to_bio_flags(u32 rnbd_opf)
+{
+ u32 bio_opf;
+
+ switch (rnbd_op(rnbd_opf)) {
+ case RNBD_OP_READ:
+ bio_opf = REQ_OP_READ;
+ break;
+ case RNBD_OP_WRITE:
+ bio_opf = REQ_OP_WRITE;
+ break;
+ case RNBD_OP_FLUSH:
+ bio_opf = REQ_OP_FLUSH | REQ_PREFLUSH;
+ break;
+ case RNBD_OP_DISCARD:
+ bio_opf = REQ_OP_DISCARD;
+ break;
+ case RNBD_OP_SECURE_ERASE:
+ bio_opf = REQ_OP_SECURE_ERASE;
+ break;
+ case RNBD_OP_WRITE_SAME:
+ bio_opf = REQ_OP_WRITE_SAME;
+ break;
+ default:
+ WARN(1, "Unknown RNBD type: %d (flags %d)\n",
+ rnbd_op(rnbd_opf), rnbd_opf);
+ bio_opf = 0;
+ }
+
+ if (rnbd_opf & RNBD_F_SYNC)
+ bio_opf |= REQ_SYNC;
+
+ if (rnbd_opf & RNBD_F_FUA)
+ bio_opf |= REQ_FUA;
+
+ return bio_opf;
+}
+
+static inline u32 rq_to_rnbd_flags(struct request *rq)
+{
+ u32 rnbd_opf;
+
+ switch (req_op(rq)) {
+ case REQ_OP_READ:
+ rnbd_opf = RNBD_OP_READ;
+ break;
+ case REQ_OP_WRITE:
+ rnbd_opf = RNBD_OP_WRITE;
+ break;
+ case REQ_OP_DISCARD:
+ rnbd_opf = RNBD_OP_DISCARD;
+ break;
+ case REQ_OP_SECURE_ERASE:
+ rnbd_opf = RNBD_OP_SECURE_ERASE;
+ break;
+ case REQ_OP_WRITE_SAME:
+ rnbd_opf = RNBD_OP_WRITE_SAME;
+ break;
+ case REQ_OP_FLUSH:
+ rnbd_opf = RNBD_OP_FLUSH;
+ break;
+ default:
+ WARN(1, "Unknown request type %d (flags %llu)\n",
+ req_op(rq), (unsigned long long)rq->cmd_flags);
+ rnbd_opf = 0;
+ }
+
+ if (op_is_sync(rq->cmd_flags))
+ rnbd_opf |= RNBD_F_SYNC;
+
+ if (op_is_flush(rq->cmd_flags))
+ rnbd_opf |= RNBD_F_FUA;
+
+ return rnbd_opf;
+}
+
+const char *rnbd_access_mode_str(enum rnbd_access_mode mode);
+
+#endif /* RNBD_PROTO_H */
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 16/25] rnbd: client: private header with client structs and functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (14 preceding siblings ...)
2019-12-20 15:50 ` [PATCH v5 15/25] rnbd: private headers with rnbd protocol structs and helpers Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 17/25] rnbd: client: main functionality Jack Wang
` (10 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This header describes main structs and functions used by rnbd-client
module, mainly for managing RNBD sessions and mapped block devices,
creating and destroying sysfs entries.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/rnbd/rnbd-clt.h | 169 ++++++++++++++++++++++++++++++++++
1 file changed, 169 insertions(+)
create mode 100644 drivers/block/rnbd/rnbd-clt.h
diff --git a/drivers/block/rnbd/rnbd-clt.h b/drivers/block/rnbd/rnbd-clt.h
new file mode 100644
index 000000000000..b8b51a41af8e
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-clt.h
@@ -0,0 +1,169 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Swapnil Ingle <swapnil.ingle@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+
+#ifndef RNBD_CLT_H
+#define RNBD_CLT_H
+
+#include <linux/wait.h>
+#include <linux/in.h>
+#include <linux/inet.h>
+#include <linux/blk-mq.h>
+#include <linux/refcount.h>
+
+#include "rtrs.h"
+#include "rnbd-proto.h"
+#include "rnbd-log.h"
+
+#define BMAX_SEGMENTS 29
+#define RECONNECT_DELAY 30
+#define MAX_RECONNECTS -1
+
+enum rnbd_clt_dev_state {
+ DEV_STATE_INIT,
+ DEV_STATE_MAPPED,
+ DEV_STATE_MAPPED_DISCONNECTED,
+ DEV_STATE_UNMAPPED,
+};
+
+struct rnbd_iu_comp {
+ wait_queue_head_t wait;
+ int errno;
+};
+
+struct rnbd_iu {
+ union {
+ struct request *rq; /* for block io */
+ void *buf; /* for user messages */
+ };
+ struct rtrs_permit *permit;
+ union {
+ /* use to send msg associated with a dev */
+ struct rnbd_clt_dev *dev;
+ /* use to send msg associated with a sess */
+ struct rnbd_clt_session *sess;
+ };
+ blk_status_t status;
+ struct scatterlist sglist[BMAX_SEGMENTS];
+ struct work_struct work;
+ int errno;
+ struct rnbd_iu_comp comp;
+ atomic_t refcount;
+};
+
+struct rnbd_cpu_qlist {
+ struct list_head requeue_list;
+ spinlock_t requeue_lock;
+ unsigned int cpu;
+};
+
+struct rnbd_clt_session {
+ struct list_head list;
+ struct rtrs_clt *rtrs;
+ wait_queue_head_t rtrs_waitq;
+ bool rtrs_ready;
+ struct rnbd_cpu_qlist __percpu
+ *cpu_queues;
+ DECLARE_BITMAP(cpu_queues_bm, NR_CPUS);
+ int __percpu *cpu_rr; /* per-cpu var for CPU round-robin */
+ atomic_t busy;
+ int queue_depth;
+ u32 max_io_size;
+ struct blk_mq_tag_set tag_set;
+ struct mutex lock; /* protects state and devs_list */
+ struct list_head devs_list; /* list of struct rnbd_clt_dev */
+ refcount_t refcount;
+ char sessname[NAME_MAX];
+ u8 ver; /* protocol version */
+};
+
+/**
+ * Submission queues.
+ */
+struct rnbd_queue {
+ struct list_head requeue_list;
+ unsigned long in_list;
+ struct rnbd_clt_dev *dev;
+ struct blk_mq_hw_ctx *hctx;
+};
+
+struct rnbd_clt_dev {
+ struct rnbd_clt_session *sess;
+ struct request_queue *queue;
+ struct rnbd_queue *hw_queues;
+ u32 device_id;
+ /* local Idr index - used to track minor number allocations. */
+ u32 clt_device_id;
+ struct mutex lock;
+ enum rnbd_clt_dev_state dev_state;
+ char pathname[NAME_MAX];
+ enum rnbd_access_mode access_mode;
+ bool read_only;
+ bool rotational;
+ u32 max_hw_sectors;
+ u32 max_write_same_sectors;
+ u32 max_discard_sectors;
+ u32 discard_granularity;
+ u32 discard_alignment;
+ u16 secure_discard;
+ u16 physical_block_size;
+ u16 logical_block_size;
+ u16 max_segments;
+ size_t nsectors;
+ u64 size; /* device size in bytes */
+ struct list_head list;
+ struct gendisk *gd;
+ struct kobject kobj;
+ char blk_symlink_name[NAME_MAX];
+ refcount_t refcount;
+ struct work_struct unmap_on_rmmod_work;
+};
+
+/* rnbd-clt.c */
+
+struct rnbd_clt_dev *rnbd_clt_map_device(const char *sessname,
+ struct rtrs_addr *paths,
+ size_t path_cnt,
+ const char *pathname,
+ enum rnbd_access_mode access_mode);
+int rnbd_clt_unmap_device(struct rnbd_clt_dev *dev, bool force,
+ const struct attribute *sysfs_self);
+
+int rnbd_clt_remap_device(struct rnbd_clt_dev *dev);
+int rnbd_clt_resize_disk(struct rnbd_clt_dev *dev, size_t newsize);
+
+/* rnbd-clt-sysfs.c */
+
+int rnbd_clt_create_sysfs_files(void);
+
+void rnbd_clt_destroy_sysfs_files(void);
+void rnbd_clt_destroy_default_group(void);
+
+void rnbd_clt_remove_dev_symlink(struct rnbd_clt_dev *dev);
+
+#endif /* RNBD_CLT_H */
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 17/25] rnbd: client: main functionality
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (15 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 16/25] rnbd: client: private header with client structs and functions Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 18/25] rnbd: client: sysfs interface functions Jack Wang
` (9 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is main functionality of rnbd-client module, which provides
interface to map remote device as local block device /dev/rnbd<N>
and feeds RTRS with IO requests.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/rnbd/rnbd-clt.c | 1761 +++++++++++++++++++++++++++++++++
1 file changed, 1761 insertions(+)
create mode 100644 drivers/block/rnbd/rnbd-clt.c
diff --git a/drivers/block/rnbd/rnbd-clt.c b/drivers/block/rnbd/rnbd-clt.c
new file mode 100644
index 000000000000..cddaa6632286
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-clt.c
@@ -0,0 +1,1761 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Swapnil Ingle <swapnil.ingle@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include <linux/module.h>
+#include <linux/blkdev.h>
+#include <linux/hdreg.h>
+#include <linux/scatterlist.h>
+#include <linux/idr.h>
+
+#include "rnbd-clt.h"
+
+MODULE_DESCRIPTION("InfiniBand Network Block Device Client");
+MODULE_LICENSE("GPL");
+
+static int rnbd_client_major;
+static DEFINE_IDA(index_ida);
+static DEFINE_MUTEX(ida_lock);
+static DEFINE_MUTEX(sess_lock);
+static LIST_HEAD(sess_list);
+
+/*
+ * Maximum number of partitions an instance can have.
+ * 6 bits = 64 minors = 63 partitions (one minor is used for the device itself)
+ */
+#define RNBD_PART_BITS 6
+
+static inline bool rnbd_clt_get_sess(struct rnbd_clt_session *sess)
+{
+ return refcount_inc_not_zero(&sess->refcount);
+}
+
+static void free_sess(struct rnbd_clt_session *sess);
+
+static void rnbd_clt_put_sess(struct rnbd_clt_session *sess)
+{
+ might_sleep();
+
+ if (refcount_dec_and_test(&sess->refcount))
+ free_sess(sess);
+}
+
+static inline bool rnbd_clt_dev_is_mapped(struct rnbd_clt_dev *dev)
+{
+ return dev->dev_state == DEV_STATE_MAPPED;
+}
+
+static void rnbd_clt_put_dev(struct rnbd_clt_dev *dev)
+{
+ might_sleep();
+
+ if (refcount_dec_and_test(&dev->refcount)) {
+ mutex_lock(&ida_lock);
+ ida_simple_remove(&index_ida, dev->clt_device_id);
+ mutex_unlock(&ida_lock);
+ kfree(dev->hw_queues);
+ rnbd_clt_put_sess(dev->sess);
+ kfree(dev);
+ }
+}
+
+static inline bool rnbd_clt_get_dev(struct rnbd_clt_dev *dev)
+{
+ return refcount_inc_not_zero(&dev->refcount);
+}
+
+static int rnbd_clt_set_dev_attr(struct rnbd_clt_dev *dev,
+ const struct rnbd_msg_open_rsp *rsp)
+{
+ struct rnbd_clt_session *sess = dev->sess;
+
+ if (unlikely(!rsp->logical_block_size))
+ return -EINVAL;
+
+ dev->device_id = le32_to_cpu(rsp->device_id);
+ dev->nsectors = le64_to_cpu(rsp->nsectors);
+ dev->logical_block_size = le16_to_cpu(rsp->logical_block_size);
+ dev->physical_block_size = le16_to_cpu(rsp->physical_block_size);
+ dev->max_write_same_sectors = le32_to_cpu(rsp->max_write_same_sectors);
+ dev->max_discard_sectors = le32_to_cpu(rsp->max_discard_sectors);
+ dev->discard_granularity = le32_to_cpu(rsp->discard_granularity);
+ dev->discard_alignment = le32_to_cpu(rsp->discard_alignment);
+ dev->secure_discard = le16_to_cpu(rsp->secure_discard);
+ dev->rotational = rsp->rotational;
+
+ dev->max_hw_sectors = sess->max_io_size / dev->logical_block_size;
+ dev->max_segments = BMAX_SEGMENTS;
+
+ dev->max_hw_sectors = min_t(u32, dev->max_hw_sectors,
+ le32_to_cpu(rsp->max_hw_sectors));
+ dev->max_segments = min_t(u16, dev->max_segments,
+ le16_to_cpu(rsp->max_segments));
+
+ return 0;
+}
+
+static int rnbd_clt_change_capacity(struct rnbd_clt_dev *dev,
+ size_t new_nsectors)
+{
+ int err = 0;
+
+ rnbd_clt_info(dev, "Device size changed from %zu to %zu sectors\n",
+ dev->nsectors, new_nsectors);
+ dev->nsectors = new_nsectors;
+ set_capacity(dev->gd,
+ dev->nsectors * (dev->logical_block_size /
+ SECTOR_SIZE));
+ err = revalidate_disk(dev->gd);
+ if (err)
+ rnbd_clt_err(dev,
+ "Failed to change device size from %zu to %zu, err: %d\n",
+ dev->nsectors, new_nsectors, err);
+ return err;
+}
+
+static int process_msg_open_rsp(struct rnbd_clt_dev *dev,
+ struct rnbd_msg_open_rsp *rsp)
+{
+ int err = 0;
+
+ mutex_lock(&dev->lock);
+ if (dev->dev_state == DEV_STATE_UNMAPPED) {
+ rnbd_clt_info(dev,
+ "Ignoring Open-Response message from server for unmapped device\n");
+ err = -ENOENT;
+ goto out;
+ }
+ if (dev->dev_state == DEV_STATE_MAPPED_DISCONNECTED) {
+ u64 nsectors = le64_to_cpu(rsp->nsectors);
+
+ /*
+ * If the device was remapped and the size changed in the
+ * meantime we need to revalidate it
+ */
+ if (dev->nsectors != nsectors)
+ rnbd_clt_change_capacity(dev, nsectors);
+ rnbd_clt_info(dev, "Device online, device remapped successfully\n");
+ }
+ err = rnbd_clt_set_dev_attr(dev, rsp);
+ if (unlikely(err))
+ goto out;
+ dev->dev_state = DEV_STATE_MAPPED;
+
+out:
+ mutex_unlock(&dev->lock);
+
+ return err;
+}
+
+int rnbd_clt_resize_disk(struct rnbd_clt_dev *dev, size_t newsize)
+{
+ int ret = 0;
+
+ mutex_lock(&dev->lock);
+ if (dev->dev_state != DEV_STATE_MAPPED) {
+ pr_err("Failed to set new size of the device, device is not opened\n");
+ ret = -ENOENT;
+ goto out;
+ }
+ ret = rnbd_clt_change_capacity(dev, newsize);
+
+out:
+ mutex_unlock(&dev->lock);
+
+ return ret;
+}
+
+static inline void rnbd_clt_dev_requeue(struct rnbd_queue *q)
+{
+ if (WARN_ON(!q->hctx))
+ return;
+
+ /* We can come here from interrupt, thus async=true */
+ blk_mq_run_hw_queue(q->hctx, true);
+}
+
+enum {
+ RNBD_DELAY_10ms = 10,
+ RNBD_DELAY_IFBUSY = -1,
+};
+
+/**
+ * rnbd_get_cpu_qlist() - finds a list with HW queues to be rerun
+ * @sess: Session to find a queue for
+ * @cpu: Cpu to start the search from
+ *
+ * Description:
+ * Each CPU has a list of HW queues, which needs to be rerun. If a list
+ * is not empty - it is marked with a bit. This function finds first
+ * set bit in a bitmap and returns corresponding CPU list.
+ */
+static struct rnbd_cpu_qlist *
+rnbd_get_cpu_qlist(struct rnbd_clt_session *sess, int cpu)
+{
+ int bit;
+
+ /* First half */
+ bit = find_next_bit(sess->cpu_queues_bm, nr_cpu_ids, cpu);
+ if (bit < nr_cpu_ids) {
+ return per_cpu_ptr(sess->cpu_queues, bit);
+ } else if (cpu != 0) {
+ /* Second half */
+ bit = find_next_bit(sess->cpu_queues_bm, cpu, 0);
+ if (bit < cpu)
+ return per_cpu_ptr(sess->cpu_queues, bit);
+ }
+
+ return NULL;
+}
+
+static inline int nxt_cpu(int cpu)
+{
+ return (cpu + 1) % nr_cpu_ids;
+}
+
+/**
+ * rnbd_rerun_if_needed() - rerun next queue marked as stopped
+ * @sess: Session to rerun a queue on
+ *
+ * Description:
+ * Each CPU has it's own list of HW queues, which should be rerun.
+ * Function finds such list with HW queues, takes a list lock, picks up
+ * the first HW queue out of the list and requeues it.
+ *
+ * Return:
+ * True if the queue was requeued, false otherwise.
+ *
+ * Context:
+ * Does not matter.
+ */
+static inline bool rnbd_rerun_if_needed(struct rnbd_clt_session *sess)
+{
+ struct rnbd_queue *q = NULL;
+ struct rnbd_cpu_qlist *cpu_q;
+ unsigned long flags;
+ int *cpup;
+
+ /*
+ * To keep fairness and not to let other queues starve we always
+ * try to wake up someone else in round-robin manner. That of course
+ * increases latency but queues always have a chance to be executed.
+ */
+ cpup = get_cpu_ptr(sess->cpu_rr);
+ for (cpu_q = rnbd_get_cpu_qlist(sess, nxt_cpu(*cpup)); cpu_q;
+ cpu_q = rnbd_get_cpu_qlist(sess, nxt_cpu(cpu_q->cpu))) {
+ if (!spin_trylock_irqsave(&cpu_q->requeue_lock, flags))
+ continue;
+ if (likely(test_bit(cpu_q->cpu, sess->cpu_queues_bm))) {
+ q = list_first_entry_or_null(&cpu_q->requeue_list,
+ typeof(*q), requeue_list);
+ if (WARN_ON(!q))
+ goto clear_bit;
+ list_del_init(&q->requeue_list);
+ clear_bit_unlock(0, &q->in_list);
+
+ if (list_empty(&cpu_q->requeue_list)) {
+ /* Clear bit if nothing is left */
+clear_bit:
+ clear_bit(cpu_q->cpu, sess->cpu_queues_bm);
+ }
+ }
+ spin_unlock_irqrestore(&cpu_q->requeue_lock, flags);
+
+ if (q)
+ break;
+ }
+
+ /**
+ * Saves the CPU that is going to be requeued on the per-cpu var. Just
+ * incrementing it doesn't work because rnbd_get_cpu_qlist() will
+ * always return the first CPU with something on the queue list when the
+ * value stored on the var is greater than the last CPU with something
+ * on the list.
+ */
+ if (cpu_q)
+ *cpup = cpu_q->cpu;
+ put_cpu_var(sess->cpu_rr);
+
+ if (q)
+ rnbd_clt_dev_requeue(q);
+
+ return !!q;
+}
+
+/**
+ * rnbd_rerun_all_if_idle() - rerun all queues left in the list if
+ * session is idling (there are no requests
+ * in-flight).
+ * @sess: Session to rerun the queues on
+ *
+ * Description:
+ * This function tries to rerun all stopped queues if there are no
+ * requests in-flight anymore. This function tries to solve an obvious
+ * problem, when number of tags < than number of queues (hctx), which
+ * are stopped and put to sleep. If last permit, which has been just put,
+ * does not wake up all left queues (hctxs), IO requests hang forever.
+ *
+ * That can happen when all number of permits, say N, have been exhausted
+ * from one CPU, and we have many block devices per session, say M.
+ * Each block device has it's own queue (hctx) for each CPU, so eventually
+ * we can put that number of queues (hctxs) to sleep: M x nr_cpu_ids.
+ * If number of permits N < M x nr_cpu_ids finally we will get an IO hang.
+ *
+ * To avoid this hang last caller of rnbd_put_permit() (last caller is the
+ * one who observes sess->busy == 0) must wake up all remaining queues.
+ *
+ * Context:
+ * Does not matter.
+ */
+static inline void rnbd_rerun_all_if_idle(struct rnbd_clt_session *sess)
+{
+ bool requeued;
+
+ do {
+ requeued = rnbd_rerun_if_needed(sess);
+ } while (atomic_read(&sess->busy) == 0 && requeued);
+}
+
+static struct rtrs_permit *rnbd_get_permit(struct rnbd_clt_session *sess,
+ enum rtrs_clt_con_type con_type,
+ int wait)
+{
+ struct rtrs_permit *permit;
+
+ permit = rtrs_clt_get_permit(sess->rtrs, con_type,
+ wait ? RTRS_PERMIT_WAIT :
+ RTRS_PERMIT_NOWAIT);
+ if (likely(permit))
+ /* We have a subtle rare case here, when all permits can be
+ * consumed before busy counter increased. This is safe,
+ * because loser will get NULL as a permit, observe 0 busy
+ * counter and immediately restart the queue himself.
+ */
+ atomic_inc(&sess->busy);
+
+ return permit;
+}
+
+static void rnbd_put_permit(struct rnbd_clt_session *sess,
+ struct rtrs_permit *permit)
+{
+ rtrs_clt_put_permit(sess->rtrs, permit);
+ atomic_dec(&sess->busy);
+ /* Paired with rnbd_clt_dev_add_to_requeue(). Decrement first
+ * and then check queue bits.
+ */
+ smp_mb__after_atomic();
+ rnbd_rerun_all_if_idle(sess);
+}
+
+static struct rnbd_iu *rnbd_get_iu(struct rnbd_clt_session *sess,
+ enum rtrs_clt_con_type con_type,
+ int wait)
+{
+ struct rnbd_iu *iu;
+ struct rtrs_permit *permit;
+
+ permit = rnbd_get_permit(sess, con_type,
+ wait ? RTRS_PERMIT_WAIT :
+ RTRS_PERMIT_NOWAIT);
+ if (unlikely(!permit))
+ return NULL;
+ iu = rtrs_permit_to_pdu(permit);
+ iu->permit = permit;
+ /* yes, rtrs_permit_from_pdu() can be nice here,
+ * but also we have to think about MQ mode
+ */
+ /*
+ * 1st reference is dropped after finishing sending a "user" message,
+ * 2nd reference is dropped after confirmation with the response is
+ * returned.
+ * 1st and 2nd can happen in any order, so the rnbd_iu should be
+ * released (rtrs_permit returned to ibbtrs) only leased after both
+ * are finished.
+ */
+ atomic_set(&iu->refcount, 2);
+ init_waitqueue_head(&iu->comp.wait);
+ iu->comp.errno = INT_MAX;
+
+ return iu;
+}
+
+static void rnbd_put_iu(struct rnbd_clt_session *sess, struct rnbd_iu *iu)
+{
+ if (atomic_dec_and_test(&iu->refcount))
+ rnbd_put_permit(sess, iu->permit);
+}
+
+static void rnbd_softirq_done_fn(struct request *rq)
+{
+ struct rnbd_clt_dev *dev = rq->rq_disk->private_data;
+ struct rnbd_clt_session *sess = dev->sess;
+ struct rnbd_iu *iu;
+
+ iu = blk_mq_rq_to_pdu(rq);
+ rnbd_put_permit(sess, iu->permit);
+ blk_mq_end_request(rq, iu->status);
+}
+
+static void msg_io_conf(void *priv, int errno)
+{
+ struct rnbd_iu *iu = priv;
+ struct rnbd_clt_dev *dev = iu->dev;
+ struct request *rq = iu->rq;
+
+ iu->status = errno ? BLK_STS_IOERR : BLK_STS_OK;
+
+ blk_mq_complete_request(rq);
+
+ if (errno)
+ rnbd_clt_info_rl(dev, "%s I/O failed with err: %d\n",
+ rq_data_dir(rq) == READ ? "read" : "write",
+ errno);
+}
+
+static void wake_up_iu_comp(struct rnbd_iu *iu, int errno)
+{
+ iu->comp.errno = errno;
+ wake_up(&iu->comp.wait);
+}
+
+static void msg_conf(void *priv, int errno)
+{
+ struct rnbd_iu *iu = priv;
+
+ iu->errno = errno;
+ schedule_work(&iu->work);
+}
+
+enum {
+ NO_WAIT = 0,
+ WAIT = 1
+};
+
+static int send_usr_msg(struct rtrs_clt *rtrs, int dir,
+ struct rnbd_iu *iu, struct kvec *vec, size_t nr,
+ size_t len, struct scatterlist *sg, unsigned int sg_len,
+ void (*conf)(struct work_struct *work),
+ int *errno, bool wait)
+{
+ int err;
+
+ INIT_WORK(&iu->work, conf);
+ err = rtrs_clt_request(dir, msg_conf, rtrs, iu->permit,
+ iu, vec, nr, len, sg, sg_len);
+ if (!err && wait) {
+ wait_event(iu->comp.wait, iu->comp.errno != INT_MAX);
+ *errno = iu->comp.errno;
+ } else {
+ *errno = 0;
+ }
+
+ return err;
+}
+
+static void msg_close_conf(struct work_struct *work)
+{
+ struct rnbd_iu *iu = container_of(work, struct rnbd_iu, work);
+ struct rnbd_clt_dev *dev = iu->dev;
+
+ wake_up_iu_comp(iu, iu->errno);
+ rnbd_put_iu(dev->sess, iu);
+ rnbd_clt_put_dev(dev);
+}
+
+static int send_msg_close(struct rnbd_clt_dev *dev, u32 device_id, bool wait)
+{
+ struct rnbd_clt_session *sess = dev->sess;
+ struct rnbd_msg_close msg;
+ struct rnbd_iu *iu;
+ struct kvec vec = {
+ .iov_base = &msg,
+ .iov_len = sizeof(msg)
+ };
+ int err, errno;
+
+ iu = rnbd_get_iu(sess, RTRS_USR_CON, RTRS_PERMIT_WAIT);
+ if (unlikely(!iu))
+ return -ENOMEM;
+
+ iu->buf = NULL;
+ iu->dev = dev;
+
+ sg_mark_end(&iu->sglist[0]);
+
+ msg.hdr.type = cpu_to_le16(RNBD_MSG_CLOSE);
+ msg.device_id = cpu_to_le32(device_id);
+
+ WARN_ON(!rnbd_clt_get_dev(dev));
+ err = send_usr_msg(sess->rtrs, WRITE, iu, &vec, 1, 0, NULL, 0,
+ msg_close_conf, &errno, wait);
+ if (unlikely(err)) {
+ rnbd_clt_put_dev(dev);
+ rnbd_put_iu(sess, iu);
+ } else {
+ err = errno;
+ }
+
+ rnbd_put_iu(sess, iu);
+ return err;
+}
+
+static void msg_open_conf(struct work_struct *work)
+{
+ struct rnbd_iu *iu = container_of(work, struct rnbd_iu, work);
+ struct rnbd_msg_open_rsp *rsp = iu->buf;
+ struct rnbd_clt_dev *dev = iu->dev;
+ int errno = iu->errno;
+
+ if (errno) {
+ rnbd_clt_err(dev,
+ "Opening failed, server responded: %d\n",
+ errno);
+ } else {
+ errno = process_msg_open_rsp(dev, rsp);
+ if (unlikely(errno)) {
+ u32 device_id = le32_to_cpu(rsp->device_id);
+ /*
+ * If server thinks its fine, but we fail to process
+ * then be nice and send a close to server.
+ */
+ (void)send_msg_close(dev, device_id, NO_WAIT);
+ }
+ }
+ kfree(rsp);
+ wake_up_iu_comp(iu, errno);
+ rnbd_put_iu(dev->sess, iu);
+ rnbd_clt_put_dev(dev);
+}
+
+static void msg_sess_info_conf(struct work_struct *work)
+{
+ struct rnbd_iu *iu = container_of(work, struct rnbd_iu, work);
+ struct rnbd_msg_sess_info_rsp *rsp = iu->buf;
+ struct rnbd_clt_session *sess = iu->sess;
+
+ if (likely(!iu->errno))
+ sess->ver = min_t(u8, rsp->ver, RNBD_PROTO_VER_MAJOR);
+
+ kfree(rsp);
+ wake_up_iu_comp(iu, iu->errno);
+ rnbd_put_iu(sess, iu);
+ rnbd_clt_put_sess(sess);
+}
+
+static int send_msg_open(struct rnbd_clt_dev *dev, bool wait)
+{
+ struct rnbd_clt_session *sess = dev->sess;
+ struct rnbd_msg_open_rsp *rsp;
+ struct rnbd_msg_open msg;
+ struct rnbd_iu *iu;
+ struct kvec vec = {
+ .iov_base = &msg,
+ .iov_len = sizeof(msg)
+ };
+ int err, errno;
+
+ rsp = kzalloc(sizeof(*rsp), GFP_KERNEL);
+ if (unlikely(!rsp))
+ return -ENOMEM;
+
+ iu = rnbd_get_iu(sess, RTRS_USR_CON, RTRS_PERMIT_WAIT);
+ if (unlikely(!iu)) {
+ kfree(rsp);
+ return -ENOMEM;
+ }
+
+ iu->buf = rsp;
+ iu->dev = dev;
+
+ sg_init_one(iu->sglist, rsp, sizeof(*rsp));
+
+ msg.hdr.type = cpu_to_le16(RNBD_MSG_OPEN);
+ msg.access_mode = dev->access_mode;
+ strlcpy(msg.dev_name, dev->pathname, sizeof(msg.dev_name));
+
+ WARN_ON(!rnbd_clt_get_dev(dev));
+ err = send_usr_msg(sess->rtrs, READ, iu,
+ &vec, 1, sizeof(*rsp), iu->sglist, 1,
+ msg_open_conf, &errno, wait);
+ if (unlikely(err)) {
+ rnbd_clt_put_dev(dev);
+ rnbd_put_iu(sess, iu);
+ kfree(rsp);
+ } else {
+ err = errno;
+ }
+
+ rnbd_put_iu(sess, iu);
+ return err;
+}
+
+static int send_msg_sess_info(struct rnbd_clt_session *sess, bool wait)
+{
+ struct rnbd_msg_sess_info_rsp *rsp;
+ struct rnbd_msg_sess_info msg;
+ struct rnbd_iu *iu;
+ struct kvec vec = {
+ .iov_base = &msg,
+ .iov_len = sizeof(msg)
+ };
+ int err, errno;
+
+ rsp = kzalloc(sizeof(*rsp), GFP_KERNEL);
+ if (unlikely(!rsp))
+ return -ENOMEM;
+
+ iu = rnbd_get_iu(sess, RTRS_USR_CON, RTRS_PERMIT_WAIT);
+ if (unlikely(!iu)) {
+ kfree(rsp);
+ return -ENOMEM;
+ }
+
+ iu->buf = rsp;
+ iu->sess = sess;
+
+ sg_init_one(iu->sglist, rsp, sizeof(*rsp));
+
+ msg.hdr.type = cpu_to_le16(RNBD_MSG_SESS_INFO);
+ msg.ver = RNBD_PROTO_VER_MAJOR;
+
+ if (unlikely(!rnbd_clt_get_sess(sess))) {
+ /*
+ * That can happen only in one case, when RTRS has restablished
+ * the connection and link_ev() is called, but session is almost
+ * dead, last reference on session is put and caller is waiting
+ * for RTRS to close everything.
+ */
+ err = -ENODEV;
+ goto put_iu;
+ }
+ err = send_usr_msg(sess->rtrs, READ, iu,
+ &vec, 1, sizeof(*rsp), iu->sglist, 1,
+ msg_sess_info_conf, &errno, wait);
+ if (unlikely(err)) {
+ rnbd_clt_put_sess(sess);
+put_iu:
+ rnbd_put_iu(sess, iu);
+ kfree(rsp);
+ } else {
+ err = errno;
+ }
+
+ rnbd_put_iu(sess, iu);
+ return err;
+}
+
+static void set_dev_states_to_disconnected(struct rnbd_clt_session *sess)
+{
+ struct rnbd_clt_dev *dev;
+
+ mutex_lock(&sess->lock);
+ list_for_each_entry(dev, &sess->devs_list, list) {
+ rnbd_clt_err(dev, "Device disconnected.\n");
+
+ mutex_lock(&dev->lock);
+ if (dev->dev_state == DEV_STATE_MAPPED)
+ dev->dev_state = DEV_STATE_MAPPED_DISCONNECTED;
+ mutex_unlock(&dev->lock);
+ }
+ mutex_unlock(&sess->lock);
+}
+
+static void remap_devs(struct rnbd_clt_session *sess)
+{
+ struct rnbd_clt_dev *dev;
+ struct rtrs_attrs attrs;
+ int err;
+
+ /*
+ * Careful here: we are called from RTRS link event directly,
+ * thus we can't send any RTRS request and wait for response
+ * or RTRS will not be able to complete request with failure
+ * if something goes wrong (failing of outstanding requests
+ * happens exactly from the context where we are blocking now).
+ *
+ * So to avoid deadlocks each usr message sent from here must
+ * be asynchronous.
+ */
+
+ err = send_msg_sess_info(sess, NO_WAIT);
+ if (unlikely(err)) {
+ pr_err("send_msg_sess_info(\"%s\"): %d\n", sess->sessname, err);
+ return;
+ }
+
+ rtrs_clt_query(sess->rtrs, &attrs);
+ mutex_lock(&sess->lock);
+ sess->max_io_size = attrs.max_io_size;
+
+ list_for_each_entry(dev, &sess->devs_list, list) {
+ bool skip;
+
+ mutex_lock(&dev->lock);
+ skip = (dev->dev_state == DEV_STATE_INIT);
+ mutex_unlock(&dev->lock);
+ if (skip)
+ /*
+ * When device is establishing connection for the first
+ * time - do not remap, it will be closed soon.
+ */
+ continue;
+
+ rnbd_clt_info(dev, "session reconnected, remapping device\n");
+ err = send_msg_open(dev, NO_WAIT);
+ if (unlikely(err)) {
+ rnbd_clt_err(dev, "send_msg_open(): %d\n", err);
+ break;
+ }
+ }
+ mutex_unlock(&sess->lock);
+}
+
+static void rnbd_clt_link_ev(void *priv, enum rtrs_clt_link_ev ev)
+{
+ struct rnbd_clt_session *sess = priv;
+
+ switch (ev) {
+ case RTRS_CLT_LINK_EV_DISCONNECTED:
+ set_dev_states_to_disconnected(sess);
+ break;
+ case RTRS_CLT_LINK_EV_RECONNECTED:
+ remap_devs(sess);
+ break;
+ default:
+ pr_err("Unknown session event received (%d), session: %s\n",
+ ev, sess->sessname);
+ }
+}
+
+static void rnbd_init_cpu_qlists(struct rnbd_cpu_qlist __percpu *cpu_queues)
+{
+ unsigned int cpu;
+ struct rnbd_cpu_qlist *cpu_q;
+
+ for_each_possible_cpu(cpu) {
+ cpu_q = per_cpu_ptr(cpu_queues, cpu);
+
+ cpu_q->cpu = cpu;
+ INIT_LIST_HEAD(&cpu_q->requeue_list);
+ spin_lock_init(&cpu_q->requeue_lock);
+ }
+}
+
+static void destroy_mq_tags(struct rnbd_clt_session *sess)
+{
+ if (sess->tag_set.tags)
+ blk_mq_free_tag_set(&sess->tag_set);
+}
+
+static inline void wake_up_rtrs_waiters(struct rnbd_clt_session *sess)
+{
+ sess->rtrs_ready = true;
+ wake_up_all(&sess->rtrs_waitq);
+}
+
+static void close_rtrs(struct rnbd_clt_session *sess)
+{
+ might_sleep();
+
+ if (!IS_ERR_OR_NULL(sess->rtrs)) {
+ rtrs_clt_close(sess->rtrs);
+ sess->rtrs = NULL;
+ wake_up_rtrs_waiters(sess);
+ }
+}
+
+static void free_sess(struct rnbd_clt_session *sess)
+{
+ WARN_ON(!list_empty(&sess->devs_list));
+
+ might_sleep();
+
+ close_rtrs(sess);
+ destroy_mq_tags(sess);
+ if (!list_empty(&sess->list)) {
+ mutex_lock(&sess_lock);
+ list_del(&sess->list);
+ mutex_unlock(&sess_lock);
+ }
+ free_percpu(sess->cpu_queues);
+ free_percpu(sess->cpu_rr);
+ kfree(sess);
+}
+
+static struct rnbd_clt_session *alloc_sess(const char *sessname)
+{
+ struct rnbd_clt_session *sess;
+ int err, cpu;
+
+ sess = kzalloc_node(sizeof(*sess), GFP_KERNEL, NUMA_NO_NODE);
+ if (unlikely(!sess)) {
+ pr_err("Failed to create session %s, allocating session struct failed\n",
+ sessname);
+ return ERR_PTR(-ENOMEM);
+ }
+ strlcpy(sess->sessname, sessname, sizeof(sess->sessname));
+ atomic_set(&sess->busy, 0);
+ mutex_init(&sess->lock);
+ INIT_LIST_HEAD(&sess->devs_list);
+ INIT_LIST_HEAD(&sess->list);
+ bitmap_zero(sess->cpu_queues_bm, NR_CPUS);
+ init_waitqueue_head(&sess->rtrs_waitq);
+ refcount_set(&sess->refcount, 1);
+
+ sess->cpu_queues = alloc_percpu(struct rnbd_cpu_qlist);
+ if (unlikely(!sess->cpu_queues)) {
+ pr_err("Failed to create session to %s, alloc of percpu var (cpu_queues) failed\n",
+ sessname);
+ err = -ENOMEM;
+ goto err;
+ }
+ rnbd_init_cpu_qlists(sess->cpu_queues);
+
+ /**
+ * That is simple percpu variable which stores cpu indeces, which are
+ * incremented on each access. We need that for the sake of fairness
+ * to wake up queues in a round-robin manner.
+ */
+ sess->cpu_rr = alloc_percpu(int);
+ if (unlikely(!sess->cpu_rr)) {
+ pr_err("Failed to create session %s, alloc of percpu var (cpu_rr) failed\n",
+ sessname);
+ err = -ENOMEM;
+ goto err;
+ }
+ for_each_possible_cpu(cpu)
+ * per_cpu_ptr(sess->cpu_rr, cpu) = cpu;
+
+ return sess;
+
+err:
+ free_sess(sess);
+
+ return ERR_PTR(err);
+}
+
+static int wait_for_rtrs_connection(struct rnbd_clt_session *sess)
+{
+ wait_event(sess->rtrs_waitq, sess->rtrs_ready);
+ if (IS_ERR_OR_NULL(sess->rtrs))
+ return -ECONNRESET;
+
+ return 0;
+}
+
+static void wait_for_rtrs_disconnection(struct rnbd_clt_session *sess)
+__releases(&sess_lock)
+__acquires(&sess_lock)
+{
+ DEFINE_WAIT_FUNC(wait, autoremove_wake_function);
+
+ prepare_to_wait(&sess->rtrs_waitq, &wait, TASK_UNINTERRUPTIBLE);
+ if (IS_ERR_OR_NULL(sess->rtrs)) {
+ finish_wait(&sess->rtrs_waitq, &wait);
+ return;
+ }
+ mutex_unlock(&sess_lock);
+ /* After unlock session can be freed, so careful */
+ schedule();
+ mutex_lock(&sess_lock);
+}
+
+static struct rnbd_clt_session *__find_and_get_sess(const char *sessname)
+__releases(&sess_lock)
+__acquires(&sess_lock)
+{
+ struct rnbd_clt_session *sess;
+ int err;
+
+again:
+ list_for_each_entry(sess, &sess_list, list) {
+ if (strcmp(sessname, sess->sessname))
+ continue;
+
+ if (unlikely(sess->rtrs_ready && IS_ERR_OR_NULL(sess->rtrs)))
+ /*
+ * No RTRS connection, session is dying.
+ */
+ continue;
+
+ if (likely(rnbd_clt_get_sess(sess))) {
+ /*
+ * Alive session is found, wait for RTRS connection.
+ */
+ mutex_unlock(&sess_lock);
+ err = wait_for_rtrs_connection(sess);
+ if (unlikely(err))
+ rnbd_clt_put_sess(sess);
+ mutex_lock(&sess_lock);
+
+ if (unlikely(err))
+ /* Session is dying, repeat the loop */
+ goto again;
+
+ return sess;
+ }
+ /*
+ * Ref is 0, session is dying, wait for RTRS disconnect
+ * in order to avoid session names clashes.
+ */
+ wait_for_rtrs_disconnection(sess);
+ /*
+ * RTRS is disconnected and soon session will be freed,
+ * so repeat a loop.
+ */
+ goto again;
+ }
+
+ return NULL;
+}
+
+static struct rnbd_clt_session *find_and_get_sess(const char *sessname)
+{
+ struct rnbd_clt_session *sess;
+
+ mutex_lock(&sess_lock);
+ sess = __find_and_get_sess(sessname);
+ mutex_unlock(&sess_lock);
+
+ return sess;
+}
+
+static struct rnbd_clt_session *
+find_and_get_or_insert_sess(struct rnbd_clt_session *sess)
+{
+ struct rnbd_clt_session *found;
+
+ mutex_lock(&sess_lock);
+ found = __find_and_get_sess(sess->sessname);
+ if (!found)
+ list_add(&sess->list, &sess_list);
+ mutex_unlock(&sess_lock);
+
+ return found;
+}
+
+static int rnbd_client_open(struct block_device *block_device, fmode_t mode)
+{
+ struct rnbd_clt_dev *dev = block_device->bd_disk->private_data;
+
+ if (dev->read_only && (mode & FMODE_WRITE))
+ return -EPERM;
+
+ if (dev->dev_state == DEV_STATE_UNMAPPED ||
+ !rnbd_clt_get_dev(dev))
+ return -EIO;
+
+ return 0;
+}
+
+static void rnbd_client_release(struct gendisk *gen, fmode_t mode)
+{
+ struct rnbd_clt_dev *dev = gen->private_data;
+
+ rnbd_clt_put_dev(dev);
+}
+
+static int rnbd_client_getgeo(struct block_device *block_device,
+ struct hd_geometry *geo)
+{
+ u64 size;
+ struct rnbd_clt_dev *dev;
+
+ dev = block_device->bd_disk->private_data;
+ size = dev->size * (dev->logical_block_size / SECTOR_SIZE);
+ geo->cylinders = (size & ~0x3f) >> 6; /* size/64 */
+ geo->heads = 4;
+ geo->sectors = 16;
+ geo->start = 0;
+
+ return 0;
+}
+
+static const struct block_device_operations rnbd_client_ops = {
+ .owner = THIS_MODULE,
+ .open = rnbd_client_open,
+ .release = rnbd_client_release,
+ .getgeo = rnbd_client_getgeo
+};
+
+static size_t rnbd_clt_get_sg_size(struct scatterlist *sglist, u32 len)
+{
+ struct scatterlist *sg;
+ size_t tsize = 0;
+ int i;
+
+ for_each_sg(sglist, sg, len, i)
+ tsize += sg->length;
+ return tsize;
+}
+
+static int rnbd_client_xfer_request(struct rnbd_clt_dev *dev,
+ struct request *rq,
+ struct rnbd_iu *iu)
+{
+ struct rtrs_clt *rtrs = dev->sess->rtrs;
+ struct rtrs_permit *permit = iu->permit;
+ struct rnbd_msg_io msg;
+ unsigned int sg_cnt = 0;
+ struct kvec vec;
+ size_t size;
+ int err;
+
+ iu->rq = rq;
+ iu->dev = dev;
+ msg.sector = cpu_to_le64(blk_rq_pos(rq));
+ msg.bi_size = cpu_to_le32(blk_rq_bytes(rq));
+ msg.rw = cpu_to_le32(rq_to_rnbd_flags(rq));
+ msg.prio = cpu_to_le16(req_get_ioprio(rq));
+
+ /*
+ * We only support discards with single segment for now.
+ * See queue limits.
+ */
+ if (req_op(rq) != REQ_OP_DISCARD)
+ sg_cnt = blk_rq_map_sg(dev->queue, rq, iu->sglist);
+
+ if (sg_cnt == 0)
+ /* Do not forget to mark the end */
+ sg_mark_end(&iu->sglist[0]);
+
+ msg.hdr.type = cpu_to_le16(RNBD_MSG_IO);
+ msg.device_id = cpu_to_le32(dev->device_id);
+
+ vec = (struct kvec) {
+ .iov_base = &msg,
+ .iov_len = sizeof(msg)
+ };
+ size = rnbd_clt_get_sg_size(iu->sglist, sg_cnt);
+ err = rtrs_clt_request(rq_data_dir(rq), msg_io_conf, rtrs, permit,
+ iu, &vec, 1, size, iu->sglist, sg_cnt);
+ if (unlikely(err)) {
+ rnbd_clt_err_rl(dev, "RTRS failed to transfer IO, err: %d\n",
+ err);
+ return err;
+ }
+
+ return 0;
+}
+
+/**
+ * rnbd_clt_dev_add_to_requeue() - add device to requeue if session is busy
+ * @dev: Device to be checked
+ * @q: Queue to be added to the requeue list if required
+ *
+ * Description:
+ * If session is busy, that means someone will requeue us when resources
+ * are freed. If session is not doing anything - device is not added to
+ * the list and @false is returned.
+ */
+static inline bool rnbd_clt_dev_add_to_requeue(struct rnbd_clt_dev *dev,
+ struct rnbd_queue *q)
+{
+ struct rnbd_clt_session *sess = dev->sess;
+ struct rnbd_cpu_qlist *cpu_q;
+ unsigned long flags;
+ bool added = true;
+ bool need_set;
+
+ cpu_q = get_cpu_ptr(sess->cpu_queues);
+ spin_lock_irqsave(&cpu_q->requeue_lock, flags);
+
+ if (likely(!test_and_set_bit_lock(0, &q->in_list))) {
+ if (WARN_ON(!list_empty(&q->requeue_list)))
+ goto unlock;
+
+ need_set = !test_bit(cpu_q->cpu, sess->cpu_queues_bm);
+ if (need_set) {
+ set_bit(cpu_q->cpu, sess->cpu_queues_bm);
+ /* Paired with rnbd_put_permit(). Set a bit first
+ * and then observe the busy counter.
+ */
+ smp_mb__before_atomic();
+ }
+ if (likely(atomic_read(&sess->busy))) {
+ list_add_tail(&q->requeue_list, &cpu_q->requeue_list);
+ } else {
+ /* Very unlikely, but possible: busy counter was
+ * observed as zero. Drop all bits and return
+ * false to restart the queue by ourselves.
+ */
+ if (need_set)
+ clear_bit(cpu_q->cpu, sess->cpu_queues_bm);
+ clear_bit_unlock(0, &q->in_list);
+ added = false;
+ }
+ }
+unlock:
+ spin_unlock_irqrestore(&cpu_q->requeue_lock, flags);
+ put_cpu_ptr(sess->cpu_queues);
+
+ return added;
+}
+
+static void rnbd_clt_dev_kick_mq_queue(struct rnbd_clt_dev *dev,
+ struct blk_mq_hw_ctx *hctx,
+ int delay)
+{
+ struct rnbd_queue *q = hctx->driver_data;
+
+ if (delay != RNBD_DELAY_IFBUSY)
+ blk_mq_delay_run_hw_queue(hctx, delay);
+ else if (unlikely(!rnbd_clt_dev_add_to_requeue(dev, q)))
+ /*
+ * If session is not busy we have to restart
+ * the queue ourselves.
+ */
+ blk_mq_delay_run_hw_queue(hctx, RNBD_DELAY_10ms);
+}
+
+static blk_status_t rnbd_queue_rq(struct blk_mq_hw_ctx *hctx,
+ const struct blk_mq_queue_data *bd)
+{
+ struct request *rq = bd->rq;
+ struct rnbd_clt_dev *dev = rq->rq_disk->private_data;
+ struct rnbd_iu *iu = blk_mq_rq_to_pdu(rq);
+ int err;
+
+ if (unlikely(!rnbd_clt_dev_is_mapped(dev)))
+ return BLK_STS_IOERR;
+
+ iu->permit = rnbd_get_permit(dev->sess, RTRS_IO_CON,
+ RTRS_PERMIT_NOWAIT);
+ if (unlikely(!iu->permit)) {
+ rnbd_clt_dev_kick_mq_queue(dev, hctx, RNBD_DELAY_IFBUSY);
+ return BLK_STS_RESOURCE;
+ }
+
+ blk_mq_start_request(rq);
+ err = rnbd_client_xfer_request(dev, rq, iu);
+ if (likely(err == 0))
+ return BLK_STS_OK;
+ if (unlikely(err == -EAGAIN || err == -ENOMEM)) {
+ rnbd_clt_dev_kick_mq_queue(dev, hctx, RNBD_DELAY_10ms);
+ rnbd_put_permit(dev->sess, iu->permit);
+ return BLK_STS_RESOURCE;
+ }
+
+ rnbd_put_permit(dev->sess, iu->permit);
+ return BLK_STS_IOERR;
+}
+
+static int rnbd_init_request(struct blk_mq_tag_set *set, struct request *rq,
+ unsigned int hctx_idx, unsigned int numa_node)
+{
+ struct rnbd_iu *iu = blk_mq_rq_to_pdu(rq);
+
+ sg_init_table(iu->sglist, BMAX_SEGMENTS);
+ return 0;
+}
+
+static struct blk_mq_ops rnbd_mq_ops = {
+ .queue_rq = rnbd_queue_rq,
+ .init_request = rnbd_init_request,
+ .complete = rnbd_softirq_done_fn,
+};
+
+static int setup_mq_tags(struct rnbd_clt_session *sess)
+{
+ struct blk_mq_tag_set *tags = &sess->tag_set;
+
+ memset(tags, 0, sizeof(*tags));
+ tags->ops = &rnbd_mq_ops;
+ tags->queue_depth = sess->queue_depth;
+ tags->numa_node = NUMA_NO_NODE;
+ tags->flags = BLK_MQ_F_SHOULD_MERGE |
+ BLK_MQ_F_TAG_SHARED;
+ tags->cmd_size = sizeof(struct rnbd_iu);
+ tags->nr_hw_queues = num_online_cpus();
+
+ return blk_mq_alloc_tag_set(tags);
+}
+
+static struct rnbd_clt_session *
+find_and_get_or_create_sess(const char *sessname,
+ const struct rtrs_addr *paths,
+ size_t path_cnt)
+{
+ struct rnbd_clt_session *sess, *found;
+ struct rtrs_attrs attrs;
+ int err;
+
+ sess = find_and_get_sess(sessname);
+ if (sess)
+ return sess;
+
+ sess = alloc_sess(sessname);
+ if (IS_ERR(sess))
+ return sess;
+
+ found = find_and_get_or_insert_sess(sess);
+ if (unlikely(found)) {
+ free_sess(sess);
+
+ return found;
+ }
+ /*
+ * Nothing was found, establish rtrs connection and proceed further.
+ */
+ sess->rtrs = rtrs_clt_open(sess, rnbd_clt_link_ev, sessname,
+ paths, path_cnt, RTRS_PORT,
+ sizeof(struct rnbd_iu),
+ RECONNECT_DELAY, BMAX_SEGMENTS,
+ MAX_RECONNECTS);
+ if (IS_ERR(sess->rtrs)) {
+ err = PTR_ERR(sess->rtrs);
+ goto wake_up_and_put;
+ }
+ rtrs_clt_query(sess->rtrs, &attrs);
+ sess->max_io_size = attrs.max_io_size;
+ sess->queue_depth = attrs.queue_depth;
+
+ err = setup_mq_tags(sess);
+ if (unlikely(err))
+ goto close_rtrs;
+
+ err = send_msg_sess_info(sess, WAIT);
+ if (unlikely(err))
+ goto close_rtrs;
+
+ wake_up_rtrs_waiters(sess);
+
+ return sess;
+
+close_rtrs:
+ close_rtrs(sess);
+put_sess:
+ rnbd_clt_put_sess(sess);
+
+ return ERR_PTR(err);
+
+wake_up_and_put:
+ wake_up_rtrs_waiters(sess);
+ goto put_sess;
+}
+
+static inline void rnbd_init_hw_queue(struct rnbd_clt_dev *dev,
+ struct rnbd_queue *q,
+ struct blk_mq_hw_ctx *hctx)
+{
+ INIT_LIST_HEAD(&q->requeue_list);
+ q->dev = dev;
+ q->hctx = hctx;
+}
+
+static void rnbd_init_mq_hw_queues(struct rnbd_clt_dev *dev)
+{
+ int i;
+ struct blk_mq_hw_ctx *hctx;
+ struct rnbd_queue *q;
+
+ queue_for_each_hw_ctx(dev->queue, hctx, i) {
+ q = &dev->hw_queues[i];
+ rnbd_init_hw_queue(dev, q, hctx);
+ hctx->driver_data = q;
+ }
+}
+
+static int index_to_minor(int index)
+{
+ return index << RNBD_PART_BITS;
+}
+
+static int minor_to_index(int minor)
+{
+ return minor >> RNBD_PART_BITS;
+}
+
+static int setup_mq_dev(struct rnbd_clt_dev *dev)
+{
+ dev->queue = blk_mq_init_queue(&dev->sess->tag_set);
+ if (IS_ERR(dev->queue)) {
+ rnbd_clt_err(dev, "Initializing multiqueue queue failed, err: %ld\n",
+ PTR_ERR(dev->queue));
+ return PTR_ERR(dev->queue);
+ }
+ rnbd_init_mq_hw_queues(dev);
+ return 0;
+}
+
+static void setup_request_queue(struct rnbd_clt_dev *dev)
+{
+ blk_queue_logical_block_size(dev->queue, dev->logical_block_size);
+ blk_queue_physical_block_size(dev->queue, dev->physical_block_size);
+ blk_queue_max_hw_sectors(dev->queue, dev->max_hw_sectors);
+ blk_queue_max_write_same_sectors(dev->queue,
+ dev->max_write_same_sectors);
+
+ /*
+ * we don't support discards to "discontiguous" segments
+ * in on request
+ */
+ blk_queue_max_discard_segments(dev->queue, 1);
+
+ blk_queue_max_discard_sectors(dev->queue, dev->max_discard_sectors);
+ dev->queue->limits.discard_granularity = dev->discard_granularity;
+ dev->queue->limits.discard_alignment = dev->discard_alignment;
+ if (dev->max_discard_sectors)
+ blk_queue_flag_set(QUEUE_FLAG_DISCARD, dev->queue);
+ if (dev->secure_discard)
+ blk_queue_flag_set(QUEUE_FLAG_SECERASE, dev->queue);
+
+ blk_queue_flag_set(QUEUE_FLAG_SAME_COMP, dev->queue);
+ blk_queue_flag_set(QUEUE_FLAG_SAME_FORCE, dev->queue);
+ blk_queue_max_segments(dev->queue, dev->max_segments);
+ blk_queue_io_opt(dev->queue, dev->sess->max_io_size);
+ blk_queue_virt_boundary(dev->queue, 4095);
+ blk_queue_write_cache(dev->queue, true, true);
+ dev->queue->queuedata = dev;
+}
+
+static void rnbd_clt_setup_gen_disk(struct rnbd_clt_dev *dev, int idx)
+{
+ dev->gd->major = rnbd_client_major;
+ dev->gd->first_minor = index_to_minor(idx);
+ dev->gd->fops = &rnbd_client_ops;
+ dev->gd->queue = dev->queue;
+ dev->gd->private_data = dev;
+ snprintf(dev->gd->disk_name, sizeof(dev->gd->disk_name), "rnbd%d",
+ idx);
+ pr_debug("disk_name=%s, capacity=%zu\n",
+ dev->gd->disk_name,
+ dev->nsectors * (dev->logical_block_size / SECTOR_SIZE)
+ );
+
+ set_capacity(dev->gd, dev->nsectors * (dev->logical_block_size /
+ SECTOR_SIZE));
+
+ if (dev->access_mode == RNBD_ACCESS_RO) {
+ dev->read_only = true;
+ set_disk_ro(dev->gd, true);
+ } else {
+ dev->read_only = false;
+ }
+
+ if (!dev->rotational)
+ blk_queue_flag_set(QUEUE_FLAG_NONROT, dev->queue);
+}
+
+static void rnbd_clt_add_gen_disk(struct rnbd_clt_dev *dev)
+{
+ add_disk(dev->gd);
+}
+
+static int rnbd_client_setup_device(struct rnbd_clt_session *sess,
+ struct rnbd_clt_dev *dev, int idx)
+{
+ int err;
+
+ dev->size = dev->nsectors * dev->logical_block_size;
+
+ err = setup_mq_dev(dev);
+ if (err)
+ return err;
+
+ setup_request_queue(dev);
+
+ dev->gd = alloc_disk_node(1 << RNBD_PART_BITS, NUMA_NO_NODE);
+ if (!dev->gd) {
+ rnbd_clt_err(dev, "Failed to allocate disk node\n");
+ blk_cleanup_queue(dev->queue);
+ return -ENOMEM;
+ }
+
+ rnbd_clt_setup_gen_disk(dev, idx);
+
+ return 0;
+}
+
+static struct rnbd_clt_dev *init_dev(struct rnbd_clt_session *sess,
+ enum rnbd_access_mode access_mode,
+ const char *pathname)
+{
+ struct rnbd_clt_dev *dev;
+ int ret;
+
+ dev = kzalloc_node(sizeof(*dev), GFP_KERNEL, NUMA_NO_NODE);
+ if (!dev)
+ return ERR_PTR(-ENOMEM);
+
+ dev->hw_queues = kcalloc(nr_cpu_ids, sizeof(*dev->hw_queues),
+ GFP_KERNEL);
+ if (unlikely(!dev->hw_queues)) {
+ pr_err("Failed to initialize device '%s' from session %s, allocating hw_queues failed.",
+ pathname, sess->sessname);
+ ret = -ENOMEM;
+ goto out_alloc;
+ }
+
+ mutex_lock(&ida_lock);
+ ret = ida_simple_get(&index_ida, 0, minor_to_index(1 << MINORBITS),
+ GFP_KERNEL);
+ mutex_unlock(&ida_lock);
+ if (ret < 0) {
+ pr_err("Failed to initialize device '%s' from session %s, allocating idr failed, err: %d\n",
+ pathname, sess->sessname, ret);
+ goto out_queues;
+ }
+ dev->clt_device_id = ret;
+ dev->sess = sess;
+ dev->access_mode = access_mode;
+ strlcpy(dev->pathname, pathname, sizeof(dev->pathname));
+ mutex_init(&dev->lock);
+ refcount_set(&dev->refcount, 1);
+ dev->dev_state = DEV_STATE_INIT;
+
+ /*
+ * Here we called from sysfs entry, thus clt-sysfs is
+ * responsible that session will not disappear.
+ */
+ WARN_ON(!rnbd_clt_get_sess(sess));
+
+ return dev;
+
+out_queues:
+ kfree(dev->hw_queues);
+out_alloc:
+ kfree(dev);
+ return ERR_PTR(ret);
+}
+
+static bool __exists_dev(const char *pathname)
+{
+ struct rnbd_clt_session *sess;
+ struct rnbd_clt_dev *dev;
+ bool found = false;
+
+ list_for_each_entry(sess, &sess_list, list) {
+ mutex_lock(&sess->lock);
+ list_for_each_entry(dev, &sess->devs_list, list) {
+ if (!strncmp(dev->pathname, pathname,
+ sizeof(dev->pathname))) {
+ found = true;
+ break;
+ }
+ }
+ mutex_unlock(&sess->lock);
+ if (found)
+ break;
+ }
+
+ return found;
+}
+
+static bool exists_devpath(const char *pathname)
+{
+ bool found;
+
+ mutex_lock(&sess_lock);
+ found = __exists_dev(pathname);
+ mutex_unlock(&sess_lock);
+
+ return found;
+}
+
+static bool insert_dev_if_not_exists_devpath(const char *pathname,
+ struct rnbd_clt_session *sess,
+ struct rnbd_clt_dev *dev)
+{
+ bool found;
+
+ mutex_lock(&sess_lock);
+ found = __exists_dev(pathname);
+ if (!found) {
+ mutex_lock(&sess->lock);
+ list_add_tail(&dev->list, &sess->devs_list);
+ mutex_unlock(&sess->lock);
+ }
+ mutex_unlock(&sess_lock);
+
+ return found;
+}
+
+static void delete_dev(struct rnbd_clt_dev *dev)
+{
+ struct rnbd_clt_session *sess = dev->sess;
+
+ mutex_lock(&sess->lock);
+ list_del(&dev->list);
+ mutex_unlock(&sess->lock);
+}
+
+struct rnbd_clt_dev *rnbd_clt_map_device(const char *sessname,
+ struct rtrs_addr *paths,
+ size_t path_cnt,
+ const char *pathname,
+ enum rnbd_access_mode access_mode)
+{
+ struct rnbd_clt_session *sess;
+ struct rnbd_clt_dev *dev;
+ int ret;
+
+ if (unlikely(exists_devpath(pathname)))
+ return ERR_PTR(-EEXIST);
+
+ sess = find_and_get_or_create_sess(sessname, paths, path_cnt);
+ if (IS_ERR(sess))
+ return ERR_CAST(sess);
+
+ dev = init_dev(sess, access_mode, pathname);
+ if (IS_ERR(dev)) {
+ pr_err("map_device: failed to map device '%s' from session %s, can't initialize device, err: %ld\n",
+ pathname, sess->sessname, PTR_ERR(dev));
+ ret = PTR_ERR(dev);
+ goto put_sess;
+ }
+ if (unlikely(insert_dev_if_not_exists_devpath(pathname, sess, dev))) {
+ ret = -EEXIST;
+ goto put_dev;
+ }
+ ret = send_msg_open(dev, WAIT);
+ if (unlikely(ret)) {
+ rnbd_clt_err(dev,
+ "map_device: failed, can't open remote device, err: %d\n",
+ ret);
+ goto del_dev;
+ }
+ mutex_lock(&dev->lock);
+ pr_debug("Opened remote device: session=%s, path='%s'\n",
+ sess->sessname, pathname);
+ ret = rnbd_client_setup_device(sess, dev, dev->clt_device_id);
+ if (ret) {
+ rnbd_clt_err(dev,
+ "map_device: Failed to configure device, err: %d\n",
+ ret);
+ mutex_unlock(&dev->lock);
+ goto del_dev;
+ }
+
+ rnbd_clt_info(dev,
+ "map_device: Device mapped as %s (nsectors: %zu, logical_block_size: %d, physical_block_size: %d, max_write_same_sectors: %d, max_discard_sectors: %d, discard_granularity: %d, discard_alignment: %d, secure_discard: %d, max_segments: %d, max_hw_sectors: %d, rotational: %d)\n",
+ dev->gd->disk_name, dev->nsectors,
+ dev->logical_block_size, dev->physical_block_size,
+ dev->max_write_same_sectors, dev->max_discard_sectors,
+ dev->discard_granularity, dev->discard_alignment,
+ dev->secure_discard, dev->max_segments,
+ dev->max_hw_sectors, dev->rotational);
+
+ mutex_unlock(&dev->lock);
+
+ rnbd_clt_add_gen_disk(dev);
+ rnbd_clt_put_sess(sess);
+
+ return dev;
+
+del_dev:
+ delete_dev(dev);
+put_dev:
+ rnbd_clt_put_dev(dev);
+put_sess:
+ rnbd_clt_put_sess(sess);
+
+ return ERR_PTR(ret);
+}
+
+static void destroy_gen_disk(struct rnbd_clt_dev *dev)
+{
+ del_gendisk(dev->gd);
+ blk_cleanup_queue(dev->queue);
+ put_disk(dev->gd);
+}
+
+static void destroy_sysfs(struct rnbd_clt_dev *dev,
+ const struct attribute *sysfs_self)
+{
+ rnbd_clt_remove_dev_symlink(dev);
+ if (dev->kobj.state_initialized) {
+ if (sysfs_self)
+ /* To avoid deadlock firstly remove itself */
+ sysfs_remove_file_self(&dev->kobj, sysfs_self);
+ kobject_del(&dev->kobj);
+ kobject_put(&dev->kobj);
+ }
+}
+
+int rnbd_clt_unmap_device(struct rnbd_clt_dev *dev, bool force,
+ const struct attribute *sysfs_self)
+{
+ struct rnbd_clt_session *sess = dev->sess;
+ int refcount, ret = 0;
+ bool was_mapped;
+
+ mutex_lock(&dev->lock);
+ if (dev->dev_state == DEV_STATE_UNMAPPED) {
+ rnbd_clt_info(dev, "Device is already being unmapped\n");
+ ret = -EALREADY;
+ goto err;
+ }
+ refcount = refcount_read(&dev->refcount);
+ if (!force && refcount > 1) {
+ rnbd_clt_err(dev,
+ "Closing device failed, device is in use, (%d device users)\n",
+ refcount - 1);
+ ret = -EBUSY;
+ goto err;
+ }
+ was_mapped = (dev->dev_state == DEV_STATE_MAPPED);
+ dev->dev_state = DEV_STATE_UNMAPPED;
+ mutex_unlock(&dev->lock);
+
+ delete_dev(dev);
+ destroy_sysfs(dev, sysfs_self);
+ destroy_gen_disk(dev);
+ if (was_mapped && sess->rtrs)
+ send_msg_close(dev, dev->device_id, WAIT);
+
+ rnbd_clt_info(dev, "Device is unmapped\n");
+
+ /* Likely last reference put */
+ rnbd_clt_put_dev(dev);
+
+ /*
+ * Here device and session can be vanished!
+ */
+
+ return 0;
+err:
+ mutex_unlock(&dev->lock);
+
+ return ret;
+}
+
+int rnbd_clt_remap_device(struct rnbd_clt_dev *dev)
+{
+ int err;
+
+ mutex_lock(&dev->lock);
+ if (likely(dev->dev_state == DEV_STATE_MAPPED_DISCONNECTED))
+ err = 0;
+ else if (dev->dev_state == DEV_STATE_UNMAPPED)
+ err = -ENODEV;
+ else if (dev->dev_state == DEV_STATE_MAPPED)
+ err = -EALREADY;
+ else
+ err = -EBUSY;
+ mutex_unlock(&dev->lock);
+ if (likely(!err)) {
+ rnbd_clt_info(dev, "Remapping device.\n");
+ err = send_msg_open(dev, WAIT);
+ if (unlikely(err))
+ rnbd_clt_err(dev, "remap_device: %d\n", err);
+ }
+
+ return err;
+}
+
+static void unmap_device_work(struct work_struct *work)
+{
+ struct rnbd_clt_dev *dev;
+
+ dev = container_of(work, typeof(*dev), unmap_on_rmmod_work);
+ rnbd_clt_unmap_device(dev, true, NULL);
+}
+
+static void rnbd_destroy_sessions(void)
+{
+ struct rnbd_clt_session *sess, *sn;
+ struct rnbd_clt_dev *dev, *tn;
+
+ /* Firstly forbid access through sysfs interface */
+ rnbd_clt_destroy_default_group();
+ rnbd_clt_destroy_sysfs_files();
+
+ /*
+ * Here at this point there is no any concurrent access to sessions
+ * list and devices list:
+ * 1. New session or device can'be be created - session sysfs files
+ * are removed.
+ * 2. Device or session can't be removed - module reference is taken
+ * into account in unmap device sysfs callback.
+ * 3. No IO requests inflight - each file open of block_dev increases
+ * module reference in get_disk().
+ *
+ * But still there can be user requests inflights, which are sent by
+ * asynchronous send_msg_*() functions, thus before unmapping devices
+ * RTRS session must be explicitly closed.
+ */
+
+ list_for_each_entry_safe(sess, sn, &sess_list, list) {
+ WARN_ON(!rnbd_clt_get_sess(sess));
+ close_rtrs(sess);
+ list_for_each_entry_safe(dev, tn, &sess->devs_list, list) {
+ /*
+ * Here unmap happens in parallel for only one reason:
+ * blk_cleanup_queue() takes around half a second, so
+ * on huge amount of devices the whole module unload
+ * procedure takes minutes.
+ */
+ INIT_WORK(&dev->unmap_on_rmmod_work, unmap_device_work);
+ queue_work(system_long_wq, &dev->unmap_on_rmmod_work);
+ }
+ rnbd_clt_put_sess(sess);
+ }
+ /* Wait for all scheduled unmap works */
+ flush_workqueue(system_long_wq);
+ WARN_ON(!list_empty(&sess_list));
+}
+
+static int __init rnbd_client_init(void)
+{
+ int err = 0;
+
+ pr_info("Loading module %s, proto %s:\n",
+ KBUILD_MODNAME, RNBD_PROTO_VER_STRING);
+
+ rnbd_client_major = register_blkdev(rnbd_client_major, "rnbd");
+ if (rnbd_client_major <= 0) {
+ pr_err("Failed to load module, block device registration failed\n");
+ return -EBUSY;
+ }
+
+ err = rnbd_clt_create_sysfs_files();
+ if (err) {
+ pr_err("Failed to load module, creating sysfs device files failed, err: %d\n",
+ err);
+ unregister_blkdev(rnbd_client_major, "rnbd");
+ }
+
+ return err;
+}
+
+static void __exit rnbd_client_exit(void)
+{
+ pr_info("Unloading module\n");
+ rnbd_destroy_sessions();
+ unregister_blkdev(rnbd_client_major, "rnbd");
+ ida_destroy(&index_ida);
+ pr_info("Module unloaded\n");
+}
+
+module_init(rnbd_client_init);
+module_exit(rnbd_client_exit);
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 18/25] rnbd: client: sysfs interface functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (16 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 17/25] rnbd: client: main functionality Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 19/25] rnbd: server: private header with server structs and functions Jack Wang
` (8 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is the sysfs interface to rnbd block devices on client side:
/sys/devices/virtual/rnbd-client/ctl/
|- map_device
| *** maps remote device
|
|- devices/
*** all mapped devices
/sys/block/rnbd<N>/rnbd/
|- unmap_device
| *** unmaps device
|
|- state
| *** device state
|
|- session
| *** session name
|
|- mapping_path
*** path of the dev that was mapped on server
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/rnbd/rnbd-clt-sysfs.c | 659 ++++++++++++++++++++++++++++
1 file changed, 659 insertions(+)
create mode 100644 drivers/block/rnbd/rnbd-clt-sysfs.c
diff --git a/drivers/block/rnbd/rnbd-clt-sysfs.c b/drivers/block/rnbd/rnbd-clt-sysfs.c
new file mode 100644
index 000000000000..91aecafdce61
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-clt-sysfs.c
@@ -0,0 +1,659 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Swapnil Ingle <swapnil.ingle@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include <linux/types.h>
+#include <linux/ctype.h>
+#include <linux/parser.h>
+#include <linux/module.h>
+#include <linux/in6.h>
+#include <linux/fs.h>
+#include <linux/uaccess.h>
+#include <linux/device.h>
+#include <rdma/ib.h>
+#include <rdma/rdma_cm.h>
+
+#include "rnbd-clt.h"
+
+static struct device *rnbd_dev;
+static struct class *rnbd_dev_class;
+static struct kobject *rnbd_devs_kobj;
+
+enum {
+ RNBD_OPT_ERR = 0,
+ RNBD_OPT_PATH = 1 << 0,
+ RNBD_OPT_DEV_PATH = 1 << 1,
+ RNBD_OPT_ACCESS_MODE = 1 << 3,
+ RNBD_OPT_SESSNAME = 1 << 6,
+};
+
+static const unsigned int rnbd_opt_mandatory[] = {
+ RNBD_OPT_PATH,
+ RNBD_OPT_DEV_PATH,
+ RNBD_OPT_SESSNAME,
+};
+
+static const match_table_t rnbd_opt_tokens = {
+ { RNBD_OPT_PATH, "path=%s" },
+ { RNBD_OPT_DEV_PATH, "device_path=%s" },
+ { RNBD_OPT_ACCESS_MODE, "access_mode=%s" },
+ { RNBD_OPT_SESSNAME, "sessname=%s" },
+ { RNBD_OPT_ERR, NULL },
+};
+
+/* remove new line from string */
+static void strip(char *s)
+{
+ char *p = s;
+
+ while (*s != '\0') {
+ if (*s != '\n')
+ *p++ = *s++;
+ else
+ ++s;
+ }
+ *p = '\0';
+}
+
+struct rnbd_map_options {
+ char *sessname;
+ struct rtrs_addr *paths;
+ size_t *path_cnt;
+ char *pathname;
+ enum rnbd_access_mode *access_mode;
+};
+
+static int rnbd_clt_parse_map_options(const char *buf, size_t max_path_cnt,
+ struct rnbd_map_options *opt)
+{
+ char *options, *sep_opt;
+ char *p;
+ substring_t args[MAX_OPT_ARGS];
+ int opt_mask = 0;
+ int token;
+ int ret = -EINVAL;
+ int i;
+ int p_cnt = 0;
+
+ options = kstrdup(buf, GFP_KERNEL);
+ if (!options)
+ return -ENOMEM;
+
+ sep_opt = strstrip(options);
+ strip(sep_opt);
+ while ((p = strsep(&sep_opt, " ")) != NULL) {
+ if (!*p)
+ continue;
+
+ token = match_token(p, rnbd_opt_tokens, args);
+ opt_mask |= token;
+
+ switch (token) {
+ case RNBD_OPT_SESSNAME:
+ p = match_strdup(args);
+ if (!p) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ if (strlen(p) > NAME_MAX) {
+ pr_err("map_device: sessname too long\n");
+ ret = -EINVAL;
+ kfree(p);
+ goto out;
+ }
+ strlcpy(opt->sessname, p, NAME_MAX);
+ kfree(p);
+ break;
+
+ case RNBD_OPT_PATH:
+ if (p_cnt >= max_path_cnt) {
+ pr_err("map_device: too many (> %zu) paths provided\n",
+ max_path_cnt);
+ ret = -ENOMEM;
+ goto out;
+ }
+ p = match_strdup(args);
+ if (!p) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ ret = rtrs_addr_to_sockaddr(p, strlen(p), RTRS_PORT,
+ &opt->paths[p_cnt]);
+ if (ret) {
+ pr_err("Can't parse path %s: %d\n", p, ret);
+ kfree(p);
+ goto out;
+ }
+
+ p_cnt++;
+
+ kfree(p);
+ break;
+
+ case RNBD_OPT_DEV_PATH:
+ p = match_strdup(args);
+ if (!p) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ if (strlen(p) > NAME_MAX) {
+ pr_err("map_device: Device path too long\n");
+ ret = -EINVAL;
+ kfree(p);
+ goto out;
+ }
+ strlcpy(opt->pathname, p, NAME_MAX);
+ kfree(p);
+ break;
+
+ case RNBD_OPT_ACCESS_MODE:
+ p = match_strdup(args);
+ if (!p) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ if (!strcmp(p, "ro")) {
+ *opt->access_mode = RNBD_ACCESS_RO;
+ } else if (!strcmp(p, "rw")) {
+ *opt->access_mode = RNBD_ACCESS_RW;
+ } else if (!strcmp(p, "migration")) {
+ *opt->access_mode = RNBD_ACCESS_MIGRATION;
+ } else {
+ pr_err("map_device: Invalid access_mode: '%s'\n",
+ p);
+ ret = -EINVAL;
+ kfree(p);
+ goto out;
+ }
+
+ kfree(p);
+ break;
+
+ default:
+ pr_err("map_device: Unknown parameter or missing value '%s'\n",
+ p);
+ ret = -EINVAL;
+ goto out;
+ }
+ }
+
+ for (i = 0; i < ARRAY_SIZE(rnbd_opt_mandatory); i++) {
+ if ((opt_mask & rnbd_opt_mandatory[i])) {
+ ret = 0;
+ } else {
+ pr_err("map_device: Parameters missing\n");
+ ret = -EINVAL;
+ break;
+ }
+ }
+
+out:
+ *opt->path_cnt = p_cnt;
+ kfree(options);
+ return ret;
+}
+
+static ssize_t state_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *page)
+{
+ struct rnbd_clt_dev *dev;
+
+ dev = container_of(kobj, struct rnbd_clt_dev, kobj);
+
+ switch (dev->dev_state) {
+ case DEV_STATE_INIT:
+ return snprintf(page, PAGE_SIZE, "init\n");
+ case DEV_STATE_MAPPED:
+ /* TODO fix cli tool before changing to proper state */
+ return snprintf(page, PAGE_SIZE, "open\n");
+ case DEV_STATE_MAPPED_DISCONNECTED:
+ /* TODO fix cli tool before changing to proper state */
+ return snprintf(page, PAGE_SIZE, "closed\n");
+ case DEV_STATE_UNMAPPED:
+ return snprintf(page, PAGE_SIZE, "unmapped\n");
+ default:
+ return snprintf(page, PAGE_SIZE, "unknown\n");
+ }
+}
+
+static struct kobj_attribute rnbd_clt_state_attr = __ATTR_RO(state);
+
+static ssize_t mapping_path_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *page)
+{
+ struct rnbd_clt_dev *dev;
+
+ dev = container_of(kobj, struct rnbd_clt_dev, kobj);
+
+ return scnprintf(page, PAGE_SIZE, "%s\n", dev->pathname);
+}
+
+static struct kobj_attribute rnbd_clt_mapping_path_attr =
+ __ATTR_RO(mapping_path);
+
+static ssize_t access_mode_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *page)
+{
+ struct rnbd_clt_dev *dev;
+
+ dev = container_of(kobj, struct rnbd_clt_dev, kobj);
+
+ return snprintf(page, PAGE_SIZE, "%s\n",
+ rnbd_access_mode_str(dev->access_mode));
+}
+
+static struct kobj_attribute rnbd_clt_access_mode =
+ __ATTR_RO(access_mode);
+
+static ssize_t rnbd_clt_unmap_dev_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *page)
+{
+ return scnprintf(page, PAGE_SIZE, "Usage: echo <normal|force> > %s\n",
+ attr->attr.name);
+}
+
+static ssize_t rnbd_clt_unmap_dev_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct rnbd_clt_dev *dev;
+ char *opt, *options;
+ bool force;
+ int err;
+
+ opt = kstrdup(buf, GFP_KERNEL);
+ if (!opt)
+ return -ENOMEM;
+
+ options = strstrip(opt);
+ strip(options);
+
+ dev = container_of(kobj, struct rnbd_clt_dev, kobj);
+
+ if (sysfs_streq(options, "normal")) {
+ force = false;
+ } else if (sysfs_streq(options, "force")) {
+ force = true;
+ } else {
+ rnbd_clt_err(dev,
+ "unmap_device: Invalid value: %s\n",
+ options);
+ err = -EINVAL;
+ goto out;
+ }
+
+ rnbd_clt_info(dev, "Unmapping device, option: %s.\n",
+ force ? "force" : "normal");
+
+ /*
+ * We take explicit module reference only for one reason: do not
+ * race with lockless rnbd_destroy_sessions().
+ */
+ if (!try_module_get(THIS_MODULE)) {
+ err = -ENODEV;
+ goto out;
+ }
+ err = rnbd_clt_unmap_device(dev, force, &attr->attr);
+ if (unlikely(err)) {
+ if (unlikely(err != -EALREADY))
+ rnbd_clt_err(dev, "unmap_device: %d\n", err);
+ goto module_put;
+ }
+
+ /*
+ * Here device can be vanished!
+ */
+
+ err = count;
+
+module_put:
+ module_put(THIS_MODULE);
+out:
+ kfree(opt);
+
+ return err;
+}
+
+static struct kobj_attribute rnbd_clt_unmap_device_attr =
+ __ATTR(unmap_device, 0644, rnbd_clt_unmap_dev_show,
+ rnbd_clt_unmap_dev_store);
+
+static ssize_t rnbd_clt_resize_dev_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ return scnprintf(page, PAGE_SIZE,
+ "Usage: echo <new size in sectors> > %s\n",
+ attr->attr.name);
+}
+
+static ssize_t rnbd_clt_resize_dev_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ int ret;
+ unsigned long sectors;
+ struct rnbd_clt_dev *dev;
+
+ dev = container_of(kobj, struct rnbd_clt_dev, kobj);
+
+ ret = kstrtoul(buf, 0, §ors);
+ if (ret)
+ return ret;
+
+ ret = rnbd_clt_resize_disk(dev, (size_t)sectors);
+ if (ret)
+ return ret;
+
+ return count;
+}
+
+static struct kobj_attribute rnbd_clt_resize_dev_attr =
+ __ATTR(resize, 0644, rnbd_clt_resize_dev_show,
+ rnbd_clt_resize_dev_store);
+
+static ssize_t rnbd_clt_remap_dev_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *page)
+{
+ return scnprintf(page, PAGE_SIZE, "Usage: echo <1> > %s\n",
+ attr->attr.name);
+}
+
+static ssize_t rnbd_clt_remap_dev_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct rnbd_clt_dev *dev;
+ char *opt, *options;
+ int err;
+
+ opt = kstrdup(buf, GFP_KERNEL);
+ if (!opt)
+ return -ENOMEM;
+
+ options = strstrip(opt);
+ strip(options);
+
+ dev = container_of(kobj, struct rnbd_clt_dev, kobj);
+ if (!sysfs_streq(options, "1")) {
+ rnbd_clt_err(dev,
+ "remap_device: Invalid value: %s\n",
+ options);
+ err = -EINVAL;
+ goto out;
+ }
+ err = rnbd_clt_remap_device(dev);
+ if (likely(!err))
+ err = count;
+
+out:
+ kfree(opt);
+
+ return err;
+}
+
+static struct kobj_attribute rnbd_clt_remap_device_attr =
+ __ATTR(remap_device, 0644, rnbd_clt_remap_dev_show,
+ rnbd_clt_remap_dev_store);
+
+static ssize_t session_show(struct kobject *kobj, struct kobj_attribute *attr,
+ char *page)
+{
+ struct rnbd_clt_dev *dev;
+
+ dev = container_of(kobj, struct rnbd_clt_dev, kobj);
+
+ return scnprintf(page, PAGE_SIZE, "%s\n", dev->sess->sessname);
+}
+
+static struct kobj_attribute rnbd_clt_session_attr =
+ __ATTR_RO(session);
+
+static struct attribute *rnbd_dev_attrs[] = {
+ &rnbd_clt_unmap_device_attr.attr,
+ &rnbd_clt_resize_dev_attr.attr,
+ &rnbd_clt_remap_device_attr.attr,
+ &rnbd_clt_mapping_path_attr.attr,
+ &rnbd_clt_state_attr.attr,
+ &rnbd_clt_session_attr.attr,
+ &rnbd_clt_access_mode.attr,
+ NULL,
+};
+
+void rnbd_clt_remove_dev_symlink(struct rnbd_clt_dev *dev)
+{
+ /*
+ * The module_is_live() check is crucial and helps to avoid annoying
+ * sysfs warning raised in sysfs_remove_link(), when the whole sysfs
+ * path was just removed, see rnbd_close_sessions().
+ */
+ if (strlen(dev->blk_symlink_name) && module_is_live(THIS_MODULE))
+ sysfs_remove_link(rnbd_devs_kobj, dev->blk_symlink_name);
+}
+
+static struct kobj_type rnbd_dev_ktype = {
+ .sysfs_ops = &kobj_sysfs_ops,
+ .default_attrs = rnbd_dev_attrs,
+};
+
+static int rnbd_clt_add_dev_kobj(struct rnbd_clt_dev *dev)
+{
+ int ret;
+ struct kobject *gd_kobj = &disk_to_dev(dev->gd)->kobj;
+
+ ret = kobject_init_and_add(&dev->kobj, &rnbd_dev_ktype, gd_kobj, "%s",
+ "rnbd");
+ if (ret)
+ rnbd_clt_err(dev, "Failed to create device sysfs dir, err: %d\n",
+ ret);
+
+ return ret;
+}
+
+static ssize_t rnbd_clt_map_device_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ return scnprintf(page, PAGE_SIZE,
+ "Usage: echo \"sessname=<name of the rtrs session> path=<[srcaddr@]dstaddr> [path=<[srcaddr@]dstaddr>] device_path=<full path on remote side> [access_mode=<ro|rw|migration>]\" > %s\n\naddr ::= [ ip:<ipv4> | ip:<ipv6> | gid:<gid> ]\n",
+ attr->attr.name);
+}
+
+static int rnbd_clt_get_path_name(struct rnbd_clt_dev *dev, char *buf,
+ size_t len)
+{
+ int ret;
+ char pathname[NAME_MAX], *s;
+
+ strlcpy(pathname, dev->pathname, sizeof(pathname));
+ while ((s = strchr(pathname, '/')))
+ s[0] = '!';
+
+ ret = snprintf(buf, len, "%s", pathname);
+ if (ret >= len)
+ return -ENAMETOOLONG;
+
+ return 0;
+}
+
+static int rnbd_clt_add_dev_symlink(struct rnbd_clt_dev *dev)
+{
+ struct kobject *gd_kobj = &disk_to_dev(dev->gd)->kobj;
+ int ret;
+
+ ret = rnbd_clt_get_path_name(dev, dev->blk_symlink_name,
+ sizeof(dev->blk_symlink_name));
+ if (ret) {
+ rnbd_clt_err(dev, "Failed to get /sys/block symlink path, err: %d\n",
+ ret);
+ goto out_err;
+ }
+
+ ret = sysfs_create_link(rnbd_devs_kobj, gd_kobj,
+ dev->blk_symlink_name);
+ if (ret) {
+ rnbd_clt_err(dev, "Creating /sys/block symlink failed, err: %d\n",
+ ret);
+ goto out_err;
+ }
+
+ return 0;
+
+out_err:
+ dev->blk_symlink_name[0] = '\0';
+ return ret;
+}
+
+static ssize_t rnbd_clt_map_device_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct rnbd_clt_dev *dev;
+ struct rnbd_map_options opt;
+ int ret;
+ char pathname[NAME_MAX];
+ char sessname[NAME_MAX];
+ enum rnbd_access_mode access_mode = RNBD_ACCESS_RW;
+
+ struct sockaddr_storage *addrs;
+ struct rtrs_addr paths[6];
+ size_t path_cnt;
+
+ opt.sessname = sessname;
+ opt.paths = paths;
+ opt.path_cnt = &path_cnt;
+ opt.pathname = pathname;
+ opt.access_mode = &access_mode;
+ addrs = kcalloc(ARRAY_SIZE(paths) * 2, sizeof(*addrs), GFP_KERNEL);
+ if (!addrs)
+ return -ENOMEM;
+
+ for (path_cnt = 0; path_cnt < ARRAY_SIZE(paths); path_cnt++) {
+ paths[path_cnt].src = &addrs[path_cnt * 2];
+ paths[path_cnt].dst = &addrs[path_cnt * 2 + 1];
+ }
+
+ ret = rnbd_clt_parse_map_options(buf, ARRAY_SIZE(paths), &opt);
+ if (ret)
+ goto out;
+
+ pr_info("Mapping device %s on session %s, (access_mode: %s)\n",
+ pathname, sessname,
+ rnbd_access_mode_str(access_mode));
+
+ dev = rnbd_clt_map_device(sessname, paths, path_cnt, pathname,
+ access_mode);
+ if (IS_ERR(dev)) {
+ ret = PTR_ERR(dev);
+ goto out;
+ }
+
+ ret = rnbd_clt_add_dev_kobj(dev);
+ if (unlikely(ret))
+ goto unmap_dev;
+
+ ret = rnbd_clt_add_dev_symlink(dev);
+ if (ret)
+ goto unmap_dev;
+
+ kfree(addrs);
+ return count;
+
+unmap_dev:
+ rnbd_clt_unmap_device(dev, true, NULL);
+out:
+ kfree(addrs);
+ return ret;
+}
+
+static struct kobj_attribute rnbd_clt_map_device_attr =
+ __ATTR(map_device, 0644,
+ rnbd_clt_map_device_show, rnbd_clt_map_device_store);
+
+static struct attribute *default_attrs[] = {
+ &rnbd_clt_map_device_attr.attr,
+ NULL,
+};
+
+static struct attribute_group default_attr_group = {
+ .attrs = default_attrs,
+};
+
+static const struct attribute_group *default_attr_groups[] = {
+ &default_attr_group,
+ NULL,
+};
+
+int rnbd_clt_create_sysfs_files(void)
+{
+ int err;
+
+ rnbd_dev_class = class_create(THIS_MODULE, "rnbd-client");
+ if (IS_ERR(rnbd_dev_class))
+ return PTR_ERR(rnbd_dev_class);
+
+ rnbd_dev = device_create_with_groups(rnbd_dev_class, NULL,
+ MKDEV(0, 0), NULL,
+ default_attr_groups, "ctl");
+ if (IS_ERR(rnbd_dev)) {
+ err = PTR_ERR(rnbd_dev);
+ goto cls_destroy;
+ }
+ rnbd_devs_kobj = kobject_create_and_add("devices", &rnbd_dev->kobj);
+ if (unlikely(!rnbd_devs_kobj)) {
+ err = -ENOMEM;
+ goto dev_destroy;
+ }
+
+ return 0;
+
+dev_destroy:
+ device_destroy(rnbd_dev_class, MKDEV(0, 0));
+cls_destroy:
+ class_destroy(rnbd_dev_class);
+
+ return err;
+}
+
+void rnbd_clt_destroy_default_group(void)
+{
+ sysfs_remove_group(&rnbd_dev->kobj, &default_attr_group);
+}
+
+void rnbd_clt_destroy_sysfs_files(void)
+{
+ kobject_del(rnbd_devs_kobj);
+ kobject_put(rnbd_devs_kobj);
+ device_destroy(rnbd_dev_class, MKDEV(0, 0));
+ class_destroy(rnbd_dev_class);
+}
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 19/25] rnbd: server: private header with server structs and functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (17 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 18/25] rnbd: client: sysfs interface functions Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 20/25] rnbd: server: main functionality Jack Wang
` (7 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This header describes main structs and functions used by rnbd-server
module, namely structs for managing sessions from different clients
and mapped (opened) devices.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/rnbd/rnbd-srv.h | 99 +++++++++++++++++++++++++++++++++++
1 file changed, 99 insertions(+)
create mode 100644 drivers/block/rnbd/rnbd-srv.h
diff --git a/drivers/block/rnbd/rnbd-srv.h b/drivers/block/rnbd/rnbd-srv.h
new file mode 100644
index 000000000000..770181e36f15
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-srv.h
@@ -0,0 +1,99 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#ifndef RNBD_SRV_H
+#define RNBD_SRV_H
+
+#include <linux/types.h>
+#include <linux/idr.h>
+#include <linux/kref.h>
+
+#include "rtrs.h"
+#include "rnbd-proto.h"
+#include "rnbd-log.h"
+
+struct rnbd_srv_session {
+ /* Entry inside global sess_list */
+ struct list_head list;
+ struct rtrs_srv *rtrs;
+ char sessname[NAME_MAX];
+ int queue_depth;
+ struct bio_set sess_bio_set;
+
+ rwlock_t index_lock ____cacheline_aligned;
+ struct idr index_idr;
+ /* List of struct rnbd_srv_sess_dev */
+ struct list_head sess_dev_list;
+ struct mutex lock;
+ u8 ver;
+};
+
+struct rnbd_srv_dev {
+ /* Entry inside global dev_list */
+ struct list_head list;
+ struct kobject dev_kobj;
+ struct kobject dev_sessions_kobj;
+ struct kref kref;
+ char id[NAME_MAX];
+ /* List of rnbd_srv_sess_dev structs */
+ struct list_head sess_dev_list;
+ struct mutex lock;
+ int open_write_cnt;
+};
+
+/* Structure which binds N devices and N sessions */
+struct rnbd_srv_sess_dev {
+ /* Entry inside rnbd_srv_dev struct */
+ struct list_head dev_list;
+ /* Entry inside rnbd_srv_session struct */
+ struct list_head sess_list;
+ struct rnbd_dev *rnbd_dev;
+ struct rnbd_srv_session *sess;
+ struct rnbd_srv_dev *dev;
+ struct kobject kobj;
+ struct completion *sysfs_release_compl;
+ u32 device_id;
+ fmode_t open_flags;
+ struct kref kref;
+ struct completion *destroy_comp;
+ char pathname[NAME_MAX];
+ enum rnbd_access_mode access_mode;
+};
+
+/* rnbd-srv-sysfs.c */
+
+int rnbd_srv_create_dev_sysfs(struct rnbd_srv_dev *dev,
+ struct block_device *bdev,
+ const char *dir_name);
+void rnbd_srv_destroy_dev_sysfs(struct rnbd_srv_dev *dev);
+int rnbd_srv_create_dev_session_sysfs(struct rnbd_srv_sess_dev *sess_dev);
+void rnbd_srv_destroy_dev_session_sysfs(struct rnbd_srv_sess_dev *sess_dev);
+int rnbd_srv_create_sysfs_files(void);
+void rnbd_srv_destroy_sysfs_files(void);
+
+#endif /* RNBD_SRV_H */
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 20/25] rnbd: server: main functionality
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (18 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 19/25] rnbd: server: private header with server structs and functions Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 21/25] rnbd: server: functionality for IO submission to file or block dev Jack Wang
` (6 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is main functionality of rnbd-server module, which handles IBTRS
events and rnbd protocol requests, like map (open) or unmap (close)
device. Also server side is responsible for processing incoming IBTRS
IO requests and forward them to local mapped devices.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/rnbd/rnbd-srv.c | 882 ++++++++++++++++++++++++++++++++++
1 file changed, 882 insertions(+)
create mode 100644 drivers/block/rnbd/rnbd-srv.c
diff --git a/drivers/block/rnbd/rnbd-srv.c b/drivers/block/rnbd/rnbd-srv.c
new file mode 100644
index 000000000000..706918848d95
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-srv.c
@@ -0,0 +1,882 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include <linux/module.h>
+#include <linux/blkdev.h>
+
+#include "rnbd-srv.h"
+#include "rnbd-srv-dev.h"
+
+MODULE_DESCRIPTION("InfiniBand Network Block Device Server");
+MODULE_LICENSE("GPL");
+
+static int __read_mostly port_nr = RTRS_PORT;
+
+module_param_named(port_nr, port_nr, int, 0444);
+MODULE_PARM_DESC(port_nr,
+ "The port number server is listening on (default: "
+ __stringify(RTRS_PORT)")");
+
+#define DEFAULT_DEV_SEARCH_PATH "/"
+
+static char dev_search_path[PATH_MAX] = DEFAULT_DEV_SEARCH_PATH;
+
+static int dev_search_path_set(const char *val, const struct kernel_param *kp)
+{
+ char *dup;
+
+ if (strlen(val) >= sizeof(dev_search_path))
+ return -EINVAL;
+
+ dup = kstrdup(val, GFP_KERNEL);
+
+ if (dup[strlen(dup) - 1] == '\n')
+ dup[strlen(dup) - 1] = '\0';
+
+ strlcpy(dev_search_path, dup, sizeof(dev_search_path));
+
+ kfree(dup);
+ pr_info("dev_search_path changed to '%s'\n", dev_search_path);
+
+ return 0;
+}
+
+static struct kparam_string dev_search_path_kparam_str = {
+ .maxlen = sizeof(dev_search_path),
+ .string = dev_search_path
+};
+
+static const struct kernel_param_ops dev_search_path_ops = {
+ .set = dev_search_path_set,
+ .get = param_get_string,
+};
+
+module_param_cb(dev_search_path, &dev_search_path_ops,
+ &dev_search_path_kparam_str, 0444);
+MODULE_PARM_DESC(dev_search_path,
+ "Sets the dev_search_path. When a device is mapped this path is prepended to the device path from the map device operation. If %SESSNAME% is specified in a path, then device will be searched in a session namespace. (default: "
+ DEFAULT_DEV_SEARCH_PATH ")");
+
+static DEFINE_MUTEX(sess_lock);
+static DEFINE_SPINLOCK(dev_lock);
+
+static LIST_HEAD(sess_list);
+static LIST_HEAD(dev_list);
+
+struct rnbd_io_private {
+ struct rtrs_srv_op *id;
+ struct rnbd_srv_sess_dev *sess_dev;
+};
+
+static void rnbd_sess_dev_release(struct kref *kref)
+{
+ struct rnbd_srv_sess_dev *sess_dev;
+
+ sess_dev = container_of(kref, struct rnbd_srv_sess_dev, kref);
+ complete(sess_dev->destroy_comp);
+}
+
+static inline void rnbd_put_sess_dev(struct rnbd_srv_sess_dev *sess_dev)
+{
+ kref_put(&sess_dev->kref, rnbd_sess_dev_release);
+}
+
+static void rnbd_endio(void *priv, int error)
+{
+ struct rnbd_io_private *rnbd_priv = priv;
+ struct rnbd_srv_sess_dev *sess_dev = rnbd_priv->sess_dev;
+
+ rnbd_put_sess_dev(sess_dev);
+
+ rtrs_srv_resp_rdma(rnbd_priv->id, error);
+
+ kfree(priv);
+}
+
+static struct rnbd_srv_sess_dev *
+rnbd_get_sess_dev(int dev_id, struct rnbd_srv_session *srv_sess)
+{
+ struct rnbd_srv_sess_dev *sess_dev;
+ int ret = 0;
+
+ read_lock(&srv_sess->index_lock);
+ sess_dev = idr_find(&srv_sess->index_idr, dev_id);
+ if (likely(sess_dev))
+ ret = kref_get_unless_zero(&sess_dev->kref);
+ read_unlock(&srv_sess->index_lock);
+
+ if (unlikely(!sess_dev || !ret))
+ return ERR_PTR(-ENXIO);
+
+ return sess_dev;
+}
+
+static int process_rdma(struct rtrs_srv *sess,
+ struct rnbd_srv_session *srv_sess,
+ struct rtrs_srv_op *id, void *data, u32 datalen,
+ const void *usr, size_t usrlen)
+{
+ const struct rnbd_msg_io *msg = usr;
+ struct rnbd_io_private *priv;
+ struct rnbd_srv_sess_dev *sess_dev;
+ u32 dev_id;
+ int err;
+
+ priv = kmalloc(sizeof(*priv), GFP_KERNEL);
+ if (unlikely(!priv))
+ return -ENOMEM;
+
+ dev_id = le32_to_cpu(msg->device_id);
+
+ sess_dev = rnbd_get_sess_dev(dev_id, srv_sess);
+ if (IS_ERR(sess_dev)) {
+ pr_err_ratelimited("Got I/O request on session %s for unknown device id %d\n",
+ srv_sess->sessname, dev_id);
+ err = -ENOTCONN;
+ goto err;
+ }
+
+ priv->sess_dev = sess_dev;
+ priv->id = id;
+
+ err = rnbd_dev_submit_io(sess_dev->rnbd_dev, le64_to_cpu(msg->sector),
+ data, datalen, le32_to_cpu(msg->bi_size),
+ le32_to_cpu(msg->rw),
+ srv_sess->ver < RNBD_PROTO_VER_MAJOR ||
+ usrlen < sizeof(*msg) ?
+ 0 : le16_to_cpu(msg->prio), priv);
+ if (unlikely(err)) {
+ rnbd_srv_err(sess_dev, "Submitting I/O to device failed, err: %d\n",
+ err);
+ goto sess_dev_put;
+ }
+
+ return 0;
+
+sess_dev_put:
+ rnbd_put_sess_dev(sess_dev);
+err:
+ kfree(priv);
+ return err;
+}
+
+static void destroy_device(struct rnbd_srv_dev *dev)
+{
+ WARN(!list_empty(&dev->sess_dev_list),
+ "Device %s is being destroyed but still in use!\n",
+ dev->id);
+
+ spin_lock(&dev_lock);
+ list_del(&dev->list);
+ spin_unlock(&dev_lock);
+
+ if (dev->dev_kobj.state_in_sysfs)
+ /*
+ * Destroy kobj only if it was really created.
+ * The following call should be sync, because
+ * we free the memory afterwards.
+ */
+ rnbd_srv_destroy_dev_sysfs(dev);
+
+ kfree(dev);
+}
+
+static void destroy_device_cb(struct kref *kref)
+{
+ struct rnbd_srv_dev *dev;
+
+ dev = container_of(kref, struct rnbd_srv_dev, kref);
+
+ destroy_device(dev);
+}
+
+static void rnbd_put_srv_dev(struct rnbd_srv_dev *dev)
+{
+ kref_put(&dev->kref, destroy_device_cb);
+}
+
+static void rnbd_destroy_sess_dev(struct rnbd_srv_sess_dev *sess_dev)
+{
+ DECLARE_COMPLETION_ONSTACK(dc);
+
+ write_lock(&sess_dev->sess->index_lock);
+ idr_remove(&sess_dev->sess->index_idr, sess_dev->device_id);
+ write_unlock(&sess_dev->sess->index_lock);
+
+ sess_dev->destroy_comp = &dc;
+ rnbd_put_sess_dev(sess_dev);
+ wait_for_completion(&dc);
+
+ rnbd_dev_close(sess_dev->rnbd_dev);
+ list_del(&sess_dev->sess_list);
+ mutex_lock(&sess_dev->dev->lock);
+ list_del(&sess_dev->dev_list);
+ if (sess_dev->open_flags & FMODE_WRITE)
+ sess_dev->dev->open_write_cnt--;
+ mutex_unlock(&sess_dev->dev->lock);
+
+ rnbd_put_srv_dev(sess_dev->dev);
+
+ rnbd_srv_info(sess_dev, "Device closed\n");
+ kfree(sess_dev);
+}
+
+static void destroy_sess(struct rnbd_srv_session *srv_sess)
+{
+ struct rnbd_srv_sess_dev *sess_dev, *tmp;
+
+ if (list_empty(&srv_sess->sess_dev_list))
+ goto out;
+
+ mutex_lock(&srv_sess->lock);
+ list_for_each_entry_safe(sess_dev, tmp, &srv_sess->sess_dev_list,
+ sess_list) {
+ rnbd_srv_destroy_dev_session_sysfs(sess_dev);
+ rnbd_destroy_sess_dev(sess_dev);
+ }
+ mutex_unlock(&srv_sess->lock);
+
+out:
+ idr_destroy(&srv_sess->index_idr);
+ bioset_exit(&srv_sess->sess_bio_set);
+
+ pr_info("RTRS Session %s disconnected\n", srv_sess->sessname);
+
+ mutex_lock(&sess_lock);
+ list_del(&srv_sess->list);
+ mutex_unlock(&sess_lock);
+
+ kfree(srv_sess);
+}
+
+static int create_sess(struct rtrs_srv *rtrs)
+{
+ struct rnbd_srv_session *srv_sess;
+ char sessname[NAME_MAX];
+ int err;
+
+ err = rtrs_srv_get_sess_name(rtrs, sessname, sizeof(sessname));
+ if (unlikely(err)) {
+ pr_err("rtrs_srv_get_sess_name(%s): %d\n", sessname, err);
+
+ return err;
+ }
+ srv_sess = kzalloc(sizeof(*srv_sess), GFP_KERNEL);
+ if (!srv_sess)
+ return -ENOMEM;
+ srv_sess->queue_depth = rtrs_srv_get_queue_depth(rtrs);
+
+ err = bioset_init(&srv_sess->sess_bio_set, srv_sess->queue_depth,
+ offsetof(struct rnbd_dev_blk_io, bio),
+ BIOSET_NEED_BVECS);
+ if (err) {
+ pr_err("Allocating srv_session for session %s failed\n",
+ sessname);
+ kfree(srv_sess);
+ return err;
+ }
+
+ idr_init(&srv_sess->index_idr);
+ rwlock_init(&srv_sess->index_lock);
+ INIT_LIST_HEAD(&srv_sess->sess_dev_list);
+ mutex_init(&srv_sess->lock);
+ mutex_lock(&sess_lock);
+ list_add(&srv_sess->list, &sess_list);
+ mutex_unlock(&sess_lock);
+
+ srv_sess->rtrs = rtrs;
+ strlcpy(srv_sess->sessname, sessname, sizeof(srv_sess->sessname));
+
+ rtrs_srv_set_sess_priv(rtrs, srv_sess);
+
+ return 0;
+}
+
+static int rnbd_srv_link_ev(struct rtrs_srv *rtrs,
+ enum rtrs_srv_link_ev ev, void *priv)
+{
+ struct rnbd_srv_session *srv_sess = priv;
+
+ switch (ev) {
+ case RTRS_SRV_LINK_EV_CONNECTED:
+ return create_sess(rtrs);
+
+ case RTRS_SRV_LINK_EV_DISCONNECTED:
+ if (WARN_ON(!srv_sess))
+ return -EINVAL;
+
+ destroy_sess(srv_sess);
+ return 0;
+
+ default:
+ pr_warn("Received unknown RTRS session event %d from session %s\n",
+ ev, srv_sess->sessname);
+ return -EINVAL;
+ }
+}
+
+static int process_msg_close(struct rtrs_srv *rtrs,
+ struct rnbd_srv_session *srv_sess,
+ void *data, size_t datalen, const void *usr,
+ size_t usrlen)
+{
+ const struct rnbd_msg_close *close_msg = usr;
+ struct rnbd_srv_sess_dev *sess_dev;
+
+ sess_dev = rnbd_get_sess_dev(le32_to_cpu(close_msg->device_id),
+ srv_sess);
+ if (IS_ERR(sess_dev))
+ return 0;
+
+ rnbd_srv_destroy_dev_session_sysfs(sess_dev);
+ rnbd_put_sess_dev(sess_dev);
+ mutex_lock(&srv_sess->lock);
+ rnbd_destroy_sess_dev(sess_dev);
+ mutex_unlock(&srv_sess->lock);
+ return 0;
+}
+
+static int process_msg_open(struct rtrs_srv *rtrs,
+ struct rnbd_srv_session *srv_sess,
+ const void *msg, size_t len,
+ void *data, size_t datalen);
+
+static int process_msg_sess_info(struct rtrs_srv *rtrs,
+ struct rnbd_srv_session *srv_sess,
+ const void *msg, size_t len,
+ void *data, size_t datalen);
+
+static int rnbd_srv_rdma_ev(struct rtrs_srv *rtrs, void *priv,
+ struct rtrs_srv_op *id, int dir,
+ void *data, size_t datalen, const void *usr,
+ size_t usrlen)
+{
+ struct rnbd_srv_session *srv_sess = priv;
+ const struct rnbd_msg_hdr *hdr = usr;
+ int ret = 0;
+ u16 type;
+
+ if (WARN_ON(!srv_sess))
+ return -ENODEV;
+
+ type = le16_to_cpu(hdr->type);
+
+ switch (type) {
+ case RNBD_MSG_IO:
+ return process_rdma(rtrs, srv_sess, id, data, datalen, usr,
+ usrlen);
+ case RNBD_MSG_CLOSE:
+ ret = process_msg_close(rtrs, srv_sess, data, datalen,
+ usr, usrlen);
+ break;
+ case RNBD_MSG_OPEN:
+ ret = process_msg_open(rtrs, srv_sess, usr, usrlen,
+ data, datalen);
+ break;
+ case RNBD_MSG_SESS_INFO:
+ ret = process_msg_sess_info(rtrs, srv_sess, usr, usrlen,
+ data, datalen);
+ break;
+ default:
+ pr_warn("Received unexpected message type %d with dir %d from session %s\n",
+ type, dir, srv_sess->sessname);
+ return -EINVAL;
+ }
+
+ rtrs_srv_resp_rdma(id, ret);
+ return 0;
+}
+
+static struct rnbd_srv_sess_dev
+*rnbd_sess_dev_alloc(struct rnbd_srv_session *srv_sess)
+{
+ struct rnbd_srv_sess_dev *sess_dev;
+ int error;
+
+ sess_dev = kzalloc(sizeof(*sess_dev), GFP_KERNEL);
+ if (!sess_dev)
+ return ERR_PTR(-ENOMEM);
+
+ idr_preload(GFP_KERNEL);
+ write_lock(&srv_sess->index_lock);
+
+ error = idr_alloc(&srv_sess->index_idr, sess_dev, 0, -1, GFP_NOWAIT);
+ if (error < 0) {
+ pr_warn("Allocating idr failed, err: %d\n", error);
+ goto out_unlock;
+ }
+
+ sess_dev->device_id = error;
+ error = 0;
+
+out_unlock:
+ write_unlock(&srv_sess->index_lock);
+ idr_preload_end();
+ if (error) {
+ kfree(sess_dev);
+ return ERR_PTR(error);
+ }
+
+ return sess_dev;
+}
+
+static struct rnbd_srv_dev *rnbd_srv_init_srv_dev(const char *id)
+{
+ struct rnbd_srv_dev *dev;
+
+ dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+ if (!dev)
+ return ERR_PTR(-ENOMEM);
+
+ strlcpy(dev->id, id, sizeof(dev->id));
+ kref_init(&dev->kref);
+ INIT_LIST_HEAD(&dev->sess_dev_list);
+ mutex_init(&dev->lock);
+
+ return dev;
+}
+
+static struct rnbd_srv_dev *
+rnbd_srv_find_or_add_srv_dev(struct rnbd_srv_dev *new_dev)
+{
+ struct rnbd_srv_dev *dev;
+
+ spin_lock(&dev_lock);
+ list_for_each_entry(dev, &dev_list, list) {
+ if (!strncmp(dev->id, new_dev->id, sizeof(dev->id))) {
+ if (!kref_get_unless_zero(&dev->kref))
+ /*
+ * We lost the race, device is almost dead.
+ * Continue traversing to find a valid one.
+ */
+ continue;
+ spin_unlock(&dev_lock);
+ return dev;
+ }
+ }
+ list_add(&new_dev->list, &dev_list);
+ spin_unlock(&dev_lock);
+
+ return new_dev;
+}
+
+static int rnbd_srv_check_update_open_perm(struct rnbd_srv_dev *srv_dev,
+ struct rnbd_srv_session *srv_sess,
+ enum rnbd_access_mode access_mode)
+{
+ int ret = -EPERM;
+
+ mutex_lock(&srv_dev->lock);
+
+ switch (access_mode) {
+ case RNBD_ACCESS_RO:
+ ret = 0;
+ break;
+ case RNBD_ACCESS_RW:
+ if (srv_dev->open_write_cnt == 0) {
+ srv_dev->open_write_cnt++;
+ ret = 0;
+ } else {
+ pr_err("Mapping device '%s' for session %s with RW permissions failed. Device already opened as 'RW' by %d client(s), access mode %s.\n",
+ srv_dev->id, srv_sess->sessname,
+ srv_dev->open_write_cnt,
+ rnbd_access_mode_str(access_mode));
+ }
+ break;
+ case RNBD_ACCESS_MIGRATION:
+ if (srv_dev->open_write_cnt < 2) {
+ srv_dev->open_write_cnt++;
+ ret = 0;
+ } else {
+ pr_err("Mapping device '%s' for session %s with migration permissions failed. Device already opened as 'RW' by %d client(s), access mode %s.\n",
+ srv_dev->id, srv_sess->sessname,
+ srv_dev->open_write_cnt,
+ rnbd_access_mode_str(access_mode));
+ }
+ break;
+ default:
+ pr_err("Received mapping request for device '%s' on session %s with invalid access mode: %d\n",
+ srv_dev->id, srv_sess->sessname, access_mode);
+ ret = -EINVAL;
+ }
+
+ mutex_unlock(&srv_dev->lock);
+
+ return ret;
+}
+
+static struct rnbd_srv_dev *
+rnbd_srv_get_or_create_srv_dev(struct rnbd_dev *rnbd_dev,
+ struct rnbd_srv_session *srv_sess,
+ enum rnbd_access_mode access_mode)
+{
+ int ret;
+ struct rnbd_srv_dev *new_dev, *dev;
+
+ new_dev = rnbd_srv_init_srv_dev(rnbd_dev->name);
+ if (IS_ERR(new_dev))
+ return new_dev;
+
+ dev = rnbd_srv_find_or_add_srv_dev(new_dev);
+ if (dev != new_dev)
+ kfree(new_dev);
+
+ ret = rnbd_srv_check_update_open_perm(dev, srv_sess, access_mode);
+ if (ret) {
+ rnbd_put_srv_dev(dev);
+ return ERR_PTR(ret);
+ }
+
+ return dev;
+}
+
+static void rnbd_srv_fill_msg_open_rsp(struct rnbd_msg_open_rsp *rsp,
+ struct rnbd_srv_sess_dev *sess_dev)
+{
+ struct rnbd_dev *rnbd_dev = sess_dev->rnbd_dev;
+
+ rsp->hdr.type = cpu_to_le16(RNBD_MSG_OPEN_RSP);
+ rsp->device_id =
+ cpu_to_le32(sess_dev->device_id);
+ rsp->nsectors =
+ cpu_to_le64(get_capacity(rnbd_dev->bdev->bd_disk));
+ rsp->logical_block_size =
+ cpu_to_le16(rnbd_dev_get_logical_bsize(rnbd_dev));
+ rsp->physical_block_size =
+ cpu_to_le16(rnbd_dev_get_phys_bsize(rnbd_dev));
+ rsp->max_segments =
+ cpu_to_le16(rnbd_dev_get_max_segs(rnbd_dev));
+ rsp->max_hw_sectors =
+ cpu_to_le32(rnbd_dev_get_max_hw_sects(rnbd_dev));
+ rsp->max_write_same_sectors =
+ cpu_to_le32(rnbd_dev_get_max_write_same_sects(rnbd_dev));
+ rsp->max_discard_sectors =
+ cpu_to_le32(rnbd_dev_get_max_discard_sects(rnbd_dev));
+ rsp->discard_granularity =
+ cpu_to_le32(rnbd_dev_get_discard_granularity(rnbd_dev));
+ rsp->discard_alignment =
+ cpu_to_le32(rnbd_dev_get_discard_alignment(rnbd_dev));
+ rsp->secure_discard =
+ cpu_to_le16(rnbd_dev_get_secure_discard(rnbd_dev));
+ rsp->rotational =
+ !blk_queue_nonrot(bdev_get_queue(rnbd_dev->bdev));
+}
+
+static struct rnbd_srv_sess_dev *
+rnbd_srv_create_set_sess_dev(struct rnbd_srv_session *srv_sess,
+ const struct rnbd_msg_open *open_msg,
+ struct rnbd_dev *rnbd_dev, fmode_t open_flags,
+ struct rnbd_srv_dev *srv_dev)
+{
+ struct rnbd_srv_sess_dev *sdev = rnbd_sess_dev_alloc(srv_sess);
+
+ if (IS_ERR(sdev))
+ return sdev;
+
+ kref_init(&sdev->kref);
+
+ strlcpy(sdev->pathname, open_msg->dev_name, sizeof(sdev->pathname));
+
+ sdev->rnbd_dev = rnbd_dev;
+ sdev->sess = srv_sess;
+ sdev->dev = srv_dev;
+ sdev->open_flags = open_flags;
+ sdev->access_mode = open_msg->access_mode;
+
+ return sdev;
+}
+
+static char *rnbd_srv_get_full_path(struct rnbd_srv_session *srv_sess,
+ const char *dev_name)
+{
+ char *full_path;
+ char *a, *b;
+
+ full_path = kmalloc(PATH_MAX, GFP_KERNEL);
+ if (!full_path)
+ return ERR_PTR(-ENOMEM);
+
+ /*
+ * Replace %SESSNAME% with a real session name in order to
+ * create device namespace.
+ */
+ a = strnstr(dev_search_path, "%SESSNAME%", sizeof(dev_search_path));
+ if (a) {
+ int len = a - dev_search_path;
+
+ len = snprintf(full_path, PATH_MAX, "%.*s/%s/%s", len,
+ dev_search_path, srv_sess->sessname, dev_name);
+ if (len >= PATH_MAX) {
+ pr_err("Tooooo looong path: %s, %s, %s\n",
+ dev_search_path, srv_sess->sessname, dev_name);
+ kfree(full_path);
+ return ERR_PTR(-EINVAL);
+ }
+ } else {
+ snprintf(full_path, PATH_MAX, "%s/%s",
+ dev_search_path, dev_name);
+ }
+
+ /* eliminitate duplicated slashes */
+ a = strchr(full_path, '/');
+ b = a;
+ while (*b != '\0') {
+ if (*b == '/' && *a == '/') {
+ b++;
+ } else {
+ a++;
+ *a = *b;
+ b++;
+ }
+ }
+ a++;
+ *a = '\0';
+
+ return full_path;
+}
+
+static int process_msg_sess_info(struct rtrs_srv *rtrs,
+ struct rnbd_srv_session *srv_sess,
+ const void *msg, size_t len,
+ void *data, size_t datalen)
+{
+ const struct rnbd_msg_sess_info *sess_info_msg = msg;
+ struct rnbd_msg_sess_info_rsp *rsp = data;
+
+ srv_sess->ver = min_t(u8, sess_info_msg->ver, RNBD_PROTO_VER_MAJOR);
+ pr_debug("Session %s using protocol version %d (client version: %d, server version: %d)\n",
+ srv_sess->sessname, srv_sess->ver,
+ sess_info_msg->ver, RNBD_PROTO_VER_MAJOR);
+
+ rsp->hdr.type = cpu_to_le16(RNBD_MSG_SESS_INFO_RSP);
+ rsp->ver = srv_sess->ver;
+
+ return 0;
+}
+
+/**
+ * find_srv_sess_dev() - a dev is already opened by this name
+ * @srv_sess the session to search.
+ * @dev_name string containing the name of the device.
+ *
+ * Return struct rnbd_srv_sess_dev if srv_sess already opened the dev_name
+ * NULL if the session didn't open the device yet.
+ */
+static struct rnbd_srv_sess_dev *
+find_srv_sess_dev(struct rnbd_srv_session *srv_sess, const char *dev_name)
+{
+ struct rnbd_srv_sess_dev *sess_dev;
+
+ if (list_empty(&srv_sess->sess_dev_list))
+ return NULL;
+
+ list_for_each_entry(sess_dev, &srv_sess->sess_dev_list, sess_list)
+ if (!strcmp(sess_dev->pathname, dev_name))
+ return sess_dev;
+
+ return NULL;
+}
+
+static int process_msg_open(struct rtrs_srv *rtrs,
+ struct rnbd_srv_session *srv_sess,
+ const void *msg, size_t len,
+ void *data, size_t datalen)
+{
+ int ret;
+ struct rnbd_srv_dev *srv_dev;
+ struct rnbd_srv_sess_dev *srv_sess_dev;
+ const struct rnbd_msg_open *open_msg = msg;
+ fmode_t open_flags;
+ char *full_path;
+ struct rnbd_dev *rnbd_dev;
+ struct rnbd_msg_open_rsp *rsp = data;
+
+ pr_debug("Open message received: session='%s' path='%s' access_mode=%d\n",
+ srv_sess->sessname, open_msg->dev_name,
+ open_msg->access_mode);
+ open_flags = FMODE_READ;
+ if (open_msg->access_mode != RNBD_ACCESS_RO)
+ open_flags |= FMODE_WRITE;
+
+ mutex_lock(&srv_sess->lock);
+
+ srv_sess_dev = find_srv_sess_dev(srv_sess, open_msg->dev_name);
+ if (srv_sess_dev)
+ goto fill_response;
+
+ if ((strlen(dev_search_path) + strlen(open_msg->dev_name))
+ >= PATH_MAX) {
+ pr_err("Opening device for session %s failed, device path too long. '%s/%s' is longer than PATH_MAX (%d)\n",
+ srv_sess->sessname, dev_search_path, open_msg->dev_name,
+ PATH_MAX);
+ ret = -EINVAL;
+ goto reject;
+ }
+ if (strstr(open_msg->dev_name, "..")) {
+ pr_err("Opening device for session %s failed, device path %s contains relative path ..\n",
+ srv_sess->sessname, open_msg->dev_name);
+ ret = -EINVAL;
+ goto reject;
+ }
+ full_path = rnbd_srv_get_full_path(srv_sess, open_msg->dev_name);
+ if (IS_ERR(full_path)) {
+ ret = PTR_ERR(full_path);
+ pr_err("Opening device '%s' for client %s failed, failed to get device full path, err: %d\n",
+ open_msg->dev_name, srv_sess->sessname, ret);
+ goto reject;
+ }
+
+ rnbd_dev = rnbd_dev_open(full_path, open_flags,
+ &srv_sess->sess_bio_set, rnbd_endio);
+ if (IS_ERR(rnbd_dev)) {
+ pr_err("Opening device '%s' on session %s failed, failed to open the block device, err: %ld\n",
+ full_path, srv_sess->sessname, PTR_ERR(rnbd_dev));
+ ret = PTR_ERR(rnbd_dev);
+ goto free_path;
+ }
+
+ srv_dev = rnbd_srv_get_or_create_srv_dev(rnbd_dev, srv_sess,
+ open_msg->access_mode);
+ if (IS_ERR(srv_dev)) {
+ pr_err("Opening device '%s' on session %s failed, creating srv_dev failed, err: %ld\n",
+ full_path, srv_sess->sessname, PTR_ERR(srv_dev));
+ ret = PTR_ERR(srv_dev);
+ goto rnbd_dev_close;
+ }
+
+ srv_sess_dev = rnbd_srv_create_set_sess_dev(srv_sess, open_msg,
+ rnbd_dev, open_flags,
+ srv_dev);
+ if (IS_ERR(srv_sess_dev)) {
+ pr_err("Opening device '%s' on session %s failed, creating sess_dev failed, err: %ld\n",
+ full_path, srv_sess->sessname, PTR_ERR(srv_sess_dev));
+ ret = PTR_ERR(srv_sess_dev);
+ goto srv_dev_put;
+ }
+
+ /* Create the srv_dev sysfs files if they haven't been created yet. The
+ * reason to delay the creation is not to create the sysfs files before
+ * we are sure the device can be opened.
+ */
+ mutex_lock(&srv_dev->lock);
+ if (!srv_dev->dev_kobj.state_in_sysfs) {
+ ret = rnbd_srv_create_dev_sysfs(srv_dev, rnbd_dev->bdev,
+ rnbd_dev->name);
+ if (ret) {
+ mutex_unlock(&srv_dev->lock);
+ rnbd_srv_err(srv_sess_dev,
+ "Opening device failed, failed to create device sysfs files, err: %d\n",
+ ret);
+ goto free_srv_sess_dev;
+ }
+ }
+
+ ret = rnbd_srv_create_dev_session_sysfs(srv_sess_dev);
+ if (ret) {
+ mutex_unlock(&srv_dev->lock);
+ rnbd_srv_err(srv_sess_dev,
+ "Opening device failed, failed to create dev client sysfs files, err: %d\n",
+ ret);
+ goto free_srv_sess_dev;
+ }
+
+ list_add(&srv_sess_dev->dev_list, &srv_dev->sess_dev_list);
+ mutex_unlock(&srv_dev->lock);
+
+ list_add(&srv_sess_dev->sess_list, &srv_sess->sess_dev_list);
+
+ rnbd_srv_info(srv_sess_dev, "Opened device '%s'\n", srv_dev->id);
+
+ kfree(full_path);
+
+fill_response:
+ rnbd_srv_fill_msg_open_rsp(rsp, srv_sess_dev);
+ mutex_unlock(&srv_sess->lock);
+ return 0;
+
+free_srv_sess_dev:
+ write_lock(&srv_sess->index_lock);
+ idr_remove(&srv_sess->index_idr, srv_sess_dev->device_id);
+ write_unlock(&srv_sess->index_lock);
+ kfree(srv_sess_dev);
+srv_dev_put:
+ if (open_msg->access_mode != RNBD_ACCESS_RO) {
+ mutex_lock(&srv_dev->lock);
+ srv_dev->open_write_cnt--;
+ mutex_unlock(&srv_dev->lock);
+ }
+ rnbd_put_srv_dev(srv_dev);
+rnbd_dev_close:
+ rnbd_dev_close(rnbd_dev);
+free_path:
+ kfree(full_path);
+reject:
+ mutex_unlock(&srv_sess->lock);
+ return ret;
+}
+
+static struct rtrs_srv_ctx *rtrs_ctx;
+
+static int __init rnbd_srv_init_module(void)
+{
+ int err;
+
+ pr_info("Loading module %s, proto %s\n",
+ KBUILD_MODNAME, RNBD_PROTO_VER_STRING);
+
+ rtrs_ctx = rtrs_srv_open(rnbd_srv_rdma_ev, rnbd_srv_link_ev,
+ port_nr);
+ if (IS_ERR(rtrs_ctx)) {
+ err = PTR_ERR(rtrs_ctx);
+ pr_err("rtrs_srv_open(), err: %d\n", err);
+ return err;
+ }
+
+ err = rnbd_srv_create_sysfs_files();
+ if (err) {
+ pr_err("rnbd_srv_create_sysfs_files(), err: %d\n", err);
+ rtrs_srv_close(rtrs_ctx);
+ return err;
+ }
+
+ return 0;
+}
+
+static void __exit rnbd_srv_cleanup_module(void)
+{
+ rtrs_srv_close(rtrs_ctx);
+ WARN_ON(!list_empty(&sess_list));
+ rnbd_srv_destroy_sysfs_files();
+ pr_info("Module unloaded\n");
+}
+
+module_init(rnbd_srv_init_module);
+module_exit(rnbd_srv_cleanup_module);
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 21/25] rnbd: server: functionality for IO submission to file or block dev
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (19 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 20/25] rnbd: server: main functionality Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 22/25] rnbd: server: sysfs interface functions Jack Wang
` (5 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This provides helper functions for IO submission to file or block dev.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/rnbd/rnbd-srv-dev.c | 162 ++++++++++++++++++++++++++++++
drivers/block/rnbd/rnbd-srv-dev.h | 130 ++++++++++++++++++++++++
2 files changed, 292 insertions(+)
create mode 100644 drivers/block/rnbd/rnbd-srv-dev.c
create mode 100644 drivers/block/rnbd/rnbd-srv-dev.h
diff --git a/drivers/block/rnbd/rnbd-srv-dev.c b/drivers/block/rnbd/rnbd-srv-dev.c
new file mode 100644
index 000000000000..1f876d7fa685
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-srv-dev.c
@@ -0,0 +1,162 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include "rnbd-srv-dev.h"
+#include "rnbd-log.h"
+
+struct rnbd_dev *rnbd_dev_open(const char *path, fmode_t flags,
+ struct bio_set *bs, rnbd_dev_io_fn io_cb)
+{
+ struct rnbd_dev *dev;
+ int ret;
+
+ dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+ if (!dev)
+ return ERR_PTR(-ENOMEM);
+
+ dev->blk_open_flags = flags;
+ dev->bdev = blkdev_get_by_path(path, flags, THIS_MODULE);
+ ret = PTR_ERR_OR_ZERO(dev->bdev);
+ if (ret)
+ goto err;
+
+ dev->blk_open_flags = flags;
+ dev->io_cb = io_cb;
+ bdevname(dev->bdev, dev->name);
+ dev->ibd_bio_set = bs;
+
+ return dev;
+
+err:
+ kfree(dev);
+ return ERR_PTR(ret);
+}
+
+void rnbd_dev_close(struct rnbd_dev *dev)
+{
+ blkdev_put(dev->bdev, dev->blk_open_flags);
+ kfree(dev);
+}
+
+static void rnbd_dev_bi_end_io(struct bio *bio)
+{
+ struct rnbd_dev_blk_io *io = bio->bi_private;
+
+ io->dev->io_cb(io->priv, blk_status_to_errno(bio->bi_status));
+ bio_put(bio);
+}
+
+/**
+ * rnbd_bio_map_kern - map kernel address into bio
+ * @q: the struct request_queue for the bio
+ * @data: pointer to buffer to map
+ * @bs: bio_set to use.
+ * @len: length in bytes
+ * @gfp_mask: allocation flags for bio allocation
+ *
+ * Map the kernel address into a bio suitable for io to a block
+ * device. Returns an error pointer in case of error.
+ */
+static struct bio *rnbd_bio_map_kern(struct request_queue *q, void *data,
+ struct bio_set *bs,
+ unsigned int len, gfp_t gfp_mask)
+{
+ unsigned long kaddr = (unsigned long)data;
+ unsigned long end = (kaddr + len + PAGE_SIZE - 1) >> PAGE_SHIFT;
+ unsigned long start = kaddr >> PAGE_SHIFT;
+ const int nr_pages = end - start;
+ int offset, i;
+ struct bio *bio;
+
+ bio = bio_alloc_bioset(gfp_mask, nr_pages, bs);
+ if (!bio)
+ return ERR_PTR(-ENOMEM);
+
+ offset = offset_in_page(kaddr);
+ for (i = 0; i < nr_pages; i++) {
+ unsigned int bytes = PAGE_SIZE - offset;
+
+ if (len <= 0)
+ break;
+
+ if (bytes > len)
+ bytes = len;
+
+ if (bio_add_pc_page(q, bio, virt_to_page(data), bytes,
+ offset) < bytes) {
+ /* we don't support partial mappings */
+ bio_put(bio);
+ return ERR_PTR(-EINVAL);
+ }
+
+ data += bytes;
+ len -= bytes;
+ offset = 0;
+ }
+
+ bio->bi_end_io = bio_put;
+ return bio;
+}
+
+int rnbd_dev_submit_io(struct rnbd_dev *dev, sector_t sector, void *data,
+ size_t len, u32 bi_size, enum rnbd_io_flags flags,
+ short prio, void *priv)
+{
+ struct request_queue *q = bdev_get_queue(dev->bdev);
+ struct rnbd_dev_blk_io *io;
+ struct bio *bio;
+
+ /* check if the buffer is suitable for bdev */
+ if (WARN_ON(!blk_rq_aligned(q, (unsigned long)data, len)))
+ return -EINVAL;
+
+ /* Generate bio with pages pointing to the rdma buffer */
+ bio = rnbd_bio_map_kern(q, data, dev->ibd_bio_set, len, GFP_KERNEL);
+ if (IS_ERR(bio))
+ return PTR_ERR(bio);
+
+ io = container_of(bio, struct rnbd_dev_blk_io, bio);
+
+ io->dev = dev;
+ io->priv = priv;
+
+ bio->bi_end_io = rnbd_dev_bi_end_io;
+ bio->bi_private = io;
+ bio->bi_opf = rnbd_to_bio_flags(flags);
+ bio->bi_iter.bi_sector = sector;
+ bio->bi_iter.bi_size = bi_size;
+ bio_set_prio(bio, prio);
+ bio_set_dev(bio, dev->bdev);
+
+ submit_bio(bio);
+
+ return 0;
+}
diff --git a/drivers/block/rnbd/rnbd-srv-dev.h b/drivers/block/rnbd/rnbd-srv-dev.h
new file mode 100644
index 000000000000..724a2a650e2d
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-srv-dev.h
@@ -0,0 +1,130 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#ifndef RNBD_SRV_DEV_H
+#define RNBD_SRV_DEV_H
+
+#include <linux/fs.h>
+#include "rnbd-proto.h"
+
+typedef void rnbd_dev_io_fn(void *priv, int error);
+
+struct rnbd_dev {
+ struct block_device *bdev;
+ struct bio_set *ibd_bio_set;
+ fmode_t blk_open_flags;
+ char name[BDEVNAME_SIZE];
+ rnbd_dev_io_fn *io_cb;
+};
+
+struct rnbd_dev_blk_io {
+ struct rnbd_dev *dev;
+ void *priv;
+ /* have to be last member for front_pad usage of bioset_init */
+ struct bio bio;
+};
+
+/**
+ * rnbd_dev_open() - Open a device
+ * @flags: open flags
+ * @bs: bio_set to use during block io,
+ * @io_cb: is called when I/O finished
+ */
+struct rnbd_dev *rnbd_dev_open(const char *path, fmode_t flags,
+ struct bio_set *bs, rnbd_dev_io_fn io_cb);
+
+/**
+ * rnbd_dev_close() - Close a device
+ */
+void rnbd_dev_close(struct rnbd_dev *dev);
+
+static inline int rnbd_dev_get_logical_bsize(const struct rnbd_dev *dev)
+{
+ return bdev_logical_block_size(dev->bdev);
+}
+
+static inline int rnbd_dev_get_phys_bsize(const struct rnbd_dev *dev)
+{
+ return bdev_physical_block_size(dev->bdev);
+}
+
+static inline int rnbd_dev_get_max_segs(const struct rnbd_dev *dev)
+{
+ return queue_max_segments(bdev_get_queue(dev->bdev));
+}
+
+static inline int rnbd_dev_get_max_hw_sects(const struct rnbd_dev *dev)
+{
+ return queue_max_hw_sectors(bdev_get_queue(dev->bdev));
+}
+
+static inline int
+rnbd_dev_get_max_write_same_sects(const struct rnbd_dev *dev)
+{
+ return bdev_write_same(dev->bdev);
+}
+
+static inline int rnbd_dev_get_secure_discard(const struct rnbd_dev *dev)
+{
+ return blk_queue_secure_erase(bdev_get_queue(dev->bdev));
+}
+
+static inline int rnbd_dev_get_max_discard_sects(const struct rnbd_dev *dev)
+{
+ if (!blk_queue_discard(bdev_get_queue(dev->bdev)))
+ return 0;
+
+ return blk_queue_get_max_sectors(bdev_get_queue(dev->bdev),
+ REQ_OP_DISCARD);
+}
+
+static inline int rnbd_dev_get_discard_granularity(const struct rnbd_dev *dev)
+{
+ return bdev_get_queue(dev->bdev)->limits.discard_granularity;
+}
+
+static inline int rnbd_dev_get_discard_alignment(const struct rnbd_dev *dev)
+{
+ return bdev_get_queue(dev->bdev)->limits.discard_alignment;
+}
+
+/**
+ * rnbd_dev_submit_io() - Submit an I/O to the disk
+ * @dev: device to that the I/O is submitted
+ * @sector: address to read/write data to
+ * @data: I/O data to write or buffer to read I/O date into
+ * @len: length of @data
+ * @bi_size: Amount of data that will be read/written
+ * @prio: IO priority
+ * @priv: private data passed to @io_fn
+ */
+int rnbd_dev_submit_io(struct rnbd_dev *dev, sector_t sector, void *data,
+ size_t len, u32 bi_size, enum rnbd_io_flags flags,
+ short prio, void *priv);
+
+#endif /* RNBD_SRV_DEV_H */
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 22/25] rnbd: server: sysfs interface functions
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (20 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 21/25] rnbd: server: functionality for IO submission to file or block dev Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-23 8:14 ` Leon Romanovsky
2019-12-20 15:51 ` [PATCH v5 23/25] rnbd: include client and server modules into kernel compilation Jack Wang
` (4 subsequent siblings)
26 siblings, 1 reply; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
This is the sysfs interface to rnbd mapped devices on server side:
/sys/devices/virtual/rnbd-server/ctl/devices/<device_name>/
|- block_dev
| *** link pointing to the corresponding block device sysfs entry
|
|- sessions/<session-name>/
| *** sessions directory
|
|- read_only
| *** is devices mapped as read only
|
|- mapping_path
*** relative device path provided by the client during mapping
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/rnbd/rnbd-srv-sysfs.c | 234 ++++++++++++++++++++++++++++
1 file changed, 234 insertions(+)
create mode 100644 drivers/block/rnbd/rnbd-srv-sysfs.c
diff --git a/drivers/block/rnbd/rnbd-srv-sysfs.c b/drivers/block/rnbd/rnbd-srv-sysfs.c
new file mode 100644
index 000000000000..17258156cdf2
--- /dev/null
+++ b/drivers/block/rnbd/rnbd-srv-sysfs.c
@@ -0,0 +1,234 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * InfiniBand Network Block Driver
+ *
+ * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
+ * Authors: Fabian Holler <mail@fholler.de>
+ * Jack Wang <jinpu.wang@profitbricks.com>
+ * Kleber Souza <kleber.souza@profitbricks.com>
+ * Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Milind Dumbare <Milind.dumbare@gmail.com>
+ *
+ * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
+ * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
+ * Roman Penyaev <roman.penyaev@profitbricks.com>
+ *
+ * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
+ * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
+ * Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ */
+
+/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
+ * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
+ * Danil Kipnis <danil.kipnis@cloud.ionos.com>
+ * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
+ * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
+ */
+#undef pr_fmt
+#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
+
+#include <uapi/linux/limits.h>
+#include <linux/kobject.h>
+#include <linux/sysfs.h>
+#include <linux/stat.h>
+#include <linux/genhd.h>
+#include <linux/list.h>
+#include <linux/moduleparam.h>
+#include <linux/device.h>
+
+#include "rnbd-srv.h"
+
+static struct device *rnbd_dev;
+static struct class *rnbd_dev_class;
+static struct kobject *rnbd_devs_kobj;
+
+static struct kobj_type ktype = {
+ .sysfs_ops = &kobj_sysfs_ops,
+};
+
+int rnbd_srv_create_dev_sysfs(struct rnbd_srv_dev *dev,
+ struct block_device *bdev,
+ const char *dir_name)
+{
+ struct kobject *bdev_kobj;
+ int ret;
+
+ ret = kobject_init_and_add(&dev->dev_kobj, &ktype,
+ rnbd_devs_kobj, dir_name);
+ if (ret)
+ return ret;
+
+ ret = kobject_init_and_add(&dev->dev_sessions_kobj,
+ &ktype,
+ &dev->dev_kobj, "sessions");
+ if (ret)
+ goto err;
+
+ bdev_kobj = &disk_to_dev(bdev->bd_disk)->kobj;
+ ret = sysfs_create_link(&dev->dev_kobj, bdev_kobj, "block_dev");
+ if (ret)
+ goto err2;
+
+ return 0;
+
+err2:
+ kobject_del(&dev->dev_sessions_kobj);
+ kobject_put(&dev->dev_sessions_kobj);
+err:
+ kobject_del(&dev->dev_kobj);
+ kobject_put(&dev->dev_kobj);
+ return ret;
+}
+
+void rnbd_srv_destroy_dev_sysfs(struct rnbd_srv_dev *dev)
+{
+ sysfs_remove_link(&dev->dev_kobj, "block_dev");
+ kobject_del(&dev->dev_sessions_kobj);
+ kobject_put(&dev->dev_sessions_kobj);
+ kobject_del(&dev->dev_kobj);
+ kobject_put(&dev->dev_kobj);
+}
+
+static ssize_t read_only_show(struct kobject *kobj, struct kobj_attribute *attr,
+ char *page)
+{
+ struct rnbd_srv_sess_dev *sess_dev;
+
+ sess_dev = container_of(kobj, struct rnbd_srv_sess_dev, kobj);
+
+ return scnprintf(page, PAGE_SIZE, "%s\n",
+ (sess_dev->open_flags & FMODE_WRITE) ? "0" : "1");
+}
+
+static struct kobj_attribute rnbd_srv_dev_session_ro_attr =
+ __ATTR_RO(read_only);
+
+static ssize_t access_mode_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *page)
+{
+ struct rnbd_srv_sess_dev *sess_dev;
+
+ sess_dev = container_of(kobj, struct rnbd_srv_sess_dev, kobj);
+
+ return scnprintf(page, PAGE_SIZE, "%s\n",
+ rnbd_access_mode_str(sess_dev->access_mode));
+}
+
+static struct kobj_attribute rnbd_srv_dev_session_access_mode_attr =
+ __ATTR_RO(access_mode);
+
+static ssize_t mapping_path_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *page)
+{
+ struct rnbd_srv_sess_dev *sess_dev;
+
+ sess_dev = container_of(kobj, struct rnbd_srv_sess_dev, kobj);
+
+ return scnprintf(page, PAGE_SIZE, "%s\n", sess_dev->pathname);
+}
+
+static struct kobj_attribute rnbd_srv_dev_session_mapping_path_attr =
+ __ATTR_RO(mapping_path);
+
+static struct attribute *rnbd_srv_default_dev_sessions_attrs[] = {
+ &rnbd_srv_dev_session_access_mode_attr.attr,
+ &rnbd_srv_dev_session_ro_attr.attr,
+ &rnbd_srv_dev_session_mapping_path_attr.attr,
+ NULL,
+};
+
+static struct attribute_group rnbd_srv_default_dev_session_attr_group = {
+ .attrs = rnbd_srv_default_dev_sessions_attrs,
+};
+
+void rnbd_srv_destroy_dev_session_sysfs(struct rnbd_srv_sess_dev *sess_dev)
+{
+ DECLARE_COMPLETION_ONSTACK(sysfs_compl);
+
+ sysfs_remove_group(&sess_dev->kobj,
+ &rnbd_srv_default_dev_session_attr_group);
+
+ sess_dev->sysfs_release_compl = &sysfs_compl;
+ kobject_del(&sess_dev->kobj);
+ kobject_put(&sess_dev->kobj);
+ wait_for_completion(&sysfs_compl);
+}
+
+static void rnbd_srv_sess_dev_release(struct kobject *kobj)
+{
+ struct rnbd_srv_sess_dev *sess_dev;
+
+ sess_dev = container_of(kobj, struct rnbd_srv_sess_dev, kobj);
+ if (sess_dev->sysfs_release_compl)
+ complete_all(sess_dev->sysfs_release_compl);
+}
+
+static struct kobj_type rnbd_srv_sess_dev_ktype = {
+ .sysfs_ops = &kobj_sysfs_ops,
+ .release = rnbd_srv_sess_dev_release,
+};
+
+int rnbd_srv_create_dev_session_sysfs(struct rnbd_srv_sess_dev *sess_dev)
+{
+ int ret;
+
+ ret = kobject_init_and_add(&sess_dev->kobj, &rnbd_srv_sess_dev_ktype,
+ &sess_dev->dev->dev_sessions_kobj, "%s",
+ sess_dev->sess->sessname);
+ if (ret)
+ return ret;
+
+ ret = sysfs_create_group(&sess_dev->kobj,
+ &rnbd_srv_default_dev_session_attr_group);
+ if (ret)
+ goto err;
+
+ return 0;
+
+err:
+ kobject_del(&sess_dev->kobj);
+ kobject_put(&sess_dev->kobj);
+
+ return ret;
+}
+
+int rnbd_srv_create_sysfs_files(void)
+{
+ int err;
+
+ rnbd_dev_class = class_create(THIS_MODULE, "rnbd-server");
+ if (IS_ERR(rnbd_dev_class))
+ return PTR_ERR(rnbd_dev_class);
+
+ rnbd_dev = device_create(rnbd_dev_class, NULL,
+ MKDEV(0, 0), NULL, "ctl");
+ if (IS_ERR(rnbd_dev)) {
+ err = PTR_ERR(rnbd_dev);
+ goto cls_destroy;
+ }
+ rnbd_devs_kobj = kobject_create_and_add("devices", &rnbd_dev->kobj);
+ if (unlikely(!rnbd_devs_kobj)) {
+ err = -ENOMEM;
+ goto dev_destroy;
+ }
+
+ return 0;
+
+dev_destroy:
+ device_destroy(rnbd_dev_class, MKDEV(0, 0));
+cls_destroy:
+ class_destroy(rnbd_dev_class);
+
+ return err;
+}
+
+void rnbd_srv_destroy_sysfs_files(void)
+{
+ kobject_del(rnbd_devs_kobj);
+ kobject_put(rnbd_devs_kobj);
+ device_destroy(rnbd_dev_class, MKDEV(0, 0));
+ class_destroy(rnbd_dev_class);
+}
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* Re: [PATCH v5 22/25] rnbd: server: sysfs interface functions
2019-12-20 15:51 ` [PATCH v5 22/25] rnbd: server: sysfs interface functions Jack Wang
@ 2019-12-23 8:14 ` Leon Romanovsky
2019-12-23 8:33 ` Jinpu Wang
0 siblings, 1 reply; 45+ messages in thread
From: Leon Romanovsky @ 2019-12-23 8:14 UTC (permalink / raw)
To: Jack Wang
Cc: linux-block, linux-rdma, axboe, hch, sagi, bvanassche, dledford,
danil.kipnis, jinpu.wang, rpenyaev
On Fri, Dec 20, 2019 at 04:51:06PM +0100, Jack Wang wrote:
> From: Jack Wang <jinpu.wang@cloud.ionos.com>
>
> This is the sysfs interface to rnbd mapped devices on server side:
>
> /sys/devices/virtual/rnbd-server/ctl/devices/<device_name>/
> |- block_dev
> | *** link pointing to the corresponding block device sysfs entry
> |
> |- sessions/<session-name>/
> | *** sessions directory
> |
> |- read_only
> | *** is devices mapped as read only
> |
> |- mapping_path
> *** relative device path provided by the client during mapping
>
> Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
> ---
> drivers/block/rnbd/rnbd-srv-sysfs.c | 234 ++++++++++++++++++++++++++++
> 1 file changed, 234 insertions(+)
> create mode 100644 drivers/block/rnbd/rnbd-srv-sysfs.c
>
> diff --git a/drivers/block/rnbd/rnbd-srv-sysfs.c b/drivers/block/rnbd/rnbd-srv-sysfs.c
> new file mode 100644
> index 000000000000..17258156cdf2
> --- /dev/null
> +++ b/drivers/block/rnbd/rnbd-srv-sysfs.c
> @@ -0,0 +1,234 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +/*
> + * InfiniBand Network Block Driver
> + *
> + * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
> + * Authors: Fabian Holler <mail@fholler.de>
> + * Jack Wang <jinpu.wang@profitbricks.com>
> + * Kleber Souza <kleber.souza@profitbricks.com>
> + * Danil Kipnis <danil.kipnis@profitbricks.com>
> + * Roman Penyaev <roman.penyaev@profitbricks.com>
> + * Milind Dumbare <Milind.dumbare@gmail.com>
> + *
> + * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
> + * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
> + * Roman Penyaev <roman.penyaev@profitbricks.com>
> + *
> + * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
> + * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
> + * Jack Wang <jinpu.wang@cloud.ionos.com>
> + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> + */
> +
> +/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
> + * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
> + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> + * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
> + * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
> + */
> +#undef pr_fmt
> +#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
> +
> +#include <uapi/linux/limits.h>
> +#include <linux/kobject.h>
> +#include <linux/sysfs.h>
> +#include <linux/stat.h>
> +#include <linux/genhd.h>
> +#include <linux/list.h>
> +#include <linux/moduleparam.h>
> +#include <linux/device.h>
> +
> +#include "rnbd-srv.h"
> +
> +static struct device *rnbd_dev;
> +static struct class *rnbd_dev_class;
> +static struct kobject *rnbd_devs_kobj;
> +
> +static struct kobj_type ktype = {
> + .sysfs_ops = &kobj_sysfs_ops,
> +};
> +
> +int rnbd_srv_create_dev_sysfs(struct rnbd_srv_dev *dev,
> + struct block_device *bdev,
> + const char *dir_name)
> +{
> + struct kobject *bdev_kobj;
> + int ret;
> +
> + ret = kobject_init_and_add(&dev->dev_kobj, &ktype,
> + rnbd_devs_kobj, dir_name);
> + if (ret)
> + return ret;
> +
> + ret = kobject_init_and_add(&dev->dev_sessions_kobj,
> + &ktype,
> + &dev->dev_kobj, "sessions");
> + if (ret)
> + goto err;
> +
> + bdev_kobj = &disk_to_dev(bdev->bd_disk)->kobj;
> + ret = sysfs_create_link(&dev->dev_kobj, bdev_kobj, "block_dev");
> + if (ret)
> + goto err2;
> +
> + return 0;
> +
> +err2:
> + kobject_del(&dev->dev_sessions_kobj);
> + kobject_put(&dev->dev_sessions_kobj);
> +err:
> + kobject_del(&dev->dev_kobj);
> + kobject_put(&dev->dev_kobj);
You are using this _del/_put pattern a lot, from what I see
kobject_put() is the only is needed.
Thanks
^ permalink raw reply [flat|nested] 45+ messages in thread* Re: [PATCH v5 22/25] rnbd: server: sysfs interface functions
2019-12-23 8:14 ` Leon Romanovsky
@ 2019-12-23 8:33 ` Jinpu Wang
0 siblings, 0 replies; 45+ messages in thread
From: Jinpu Wang @ 2019-12-23 8:33 UTC (permalink / raw)
To: Leon Romanovsky
Cc: Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Bart Van Assche, Doug Ledford,
Danil Kipnis, rpenyaev
On Mon, Dec 23, 2019 at 9:14 AM Leon Romanovsky <leon@kernel.org> wrote:
>
> On Fri, Dec 20, 2019 at 04:51:06PM +0100, Jack Wang wrote:
> > From: Jack Wang <jinpu.wang@cloud.ionos.com>
> >
> > This is the sysfs interface to rnbd mapped devices on server side:
> >
> > /sys/devices/virtual/rnbd-server/ctl/devices/<device_name>/
> > |- block_dev
> > | *** link pointing to the corresponding block device sysfs entry
> > |
> > |- sessions/<session-name>/
> > | *** sessions directory
> > |
> > |- read_only
> > | *** is devices mapped as read only
> > |
> > |- mapping_path
> > *** relative device path provided by the client during mapping
> >
> > Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> > Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
> > ---
> > drivers/block/rnbd/rnbd-srv-sysfs.c | 234 ++++++++++++++++++++++++++++
> > 1 file changed, 234 insertions(+)
> > create mode 100644 drivers/block/rnbd/rnbd-srv-sysfs.c
> >
> > diff --git a/drivers/block/rnbd/rnbd-srv-sysfs.c b/drivers/block/rnbd/rnbd-srv-sysfs.c
> > new file mode 100644
> > index 000000000000..17258156cdf2
> > --- /dev/null
> > +++ b/drivers/block/rnbd/rnbd-srv-sysfs.c
> > @@ -0,0 +1,234 @@
> > +// SPDX-License-Identifier: GPL-2.0-or-later
> > +/*
> > + * InfiniBand Network Block Driver
> > + *
> > + * Copyright (c) 2014 - 2017 ProfitBricks GmbH. All rights reserved.
> > + * Authors: Fabian Holler <mail@fholler.de>
> > + * Jack Wang <jinpu.wang@profitbricks.com>
> > + * Kleber Souza <kleber.souza@profitbricks.com>
> > + * Danil Kipnis <danil.kipnis@profitbricks.com>
> > + * Roman Penyaev <roman.penyaev@profitbricks.com>
> > + * Milind Dumbare <Milind.dumbare@gmail.com>
> > + *
> > + * Copyright (c) 2017 - 2018 ProfitBricks GmbH. All rights reserved.
> > + * Authors: Danil Kipnis <danil.kipnis@profitbricks.com>
> > + * Roman Penyaev <roman.penyaev@profitbricks.com>
> > + *
> > + * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
> > + * Authors: Roman Penyaev <roman.penyaev@profitbricks.com>
> > + * Jack Wang <jinpu.wang@cloud.ionos.com>
> > + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> > + */
> > +
> > +/* Copyright (c) 2019 1&1 IONOS SE. All rights reserved.
> > + * Authors: Jack Wang <jinpu.wang@cloud.ionos.com>
> > + * Danil Kipnis <danil.kipnis@cloud.ionos.com>
> > + * Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
> > + * Lutz Pogrell <lutz.pogrell@cloud.ionos.com>
> > + */
> > +#undef pr_fmt
> > +#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
> > +
> > +#include <uapi/linux/limits.h>
> > +#include <linux/kobject.h>
> > +#include <linux/sysfs.h>
> > +#include <linux/stat.h>
> > +#include <linux/genhd.h>
> > +#include <linux/list.h>
> > +#include <linux/moduleparam.h>
> > +#include <linux/device.h>
> > +
> > +#include "rnbd-srv.h"
> > +
> > +static struct device *rnbd_dev;
> > +static struct class *rnbd_dev_class;
> > +static struct kobject *rnbd_devs_kobj;
> > +
> > +static struct kobj_type ktype = {
> > + .sysfs_ops = &kobj_sysfs_ops,
> > +};
> > +
> > +int rnbd_srv_create_dev_sysfs(struct rnbd_srv_dev *dev,
> > + struct block_device *bdev,
> > + const char *dir_name)
> > +{
> > + struct kobject *bdev_kobj;
> > + int ret;
> > +
> > + ret = kobject_init_and_add(&dev->dev_kobj, &ktype,
> > + rnbd_devs_kobj, dir_name);
> > + if (ret)
> > + return ret;
> > +
> > + ret = kobject_init_and_add(&dev->dev_sessions_kobj,
> > + &ktype,
> > + &dev->dev_kobj, "sessions");
> > + if (ret)
> > + goto err;
> > +
> > + bdev_kobj = &disk_to_dev(bdev->bd_disk)->kobj;
> > + ret = sysfs_create_link(&dev->dev_kobj, bdev_kobj, "block_dev");
> > + if (ret)
> > + goto err2;
> > +
> > + return 0;
> > +
> > +err2:
> > + kobject_del(&dev->dev_sessions_kobj);
> > + kobject_put(&dev->dev_sessions_kobj);
> > +err:
> > + kobject_del(&dev->dev_kobj);
> > + kobject_put(&dev->dev_kobj);
>
> You are using this _del/_put pattern a lot, from what I see
> kobject_put() is the only is needed.
>
> Thanks
Yes, you're right, for these 2 cases, kobject_put is enough, although
kobject_del is harmless here, which will unregister the kobject from
sysfs, and makes the kobject "invisible" to
user.
Will remove them next round.
Thanks
^ permalink raw reply [flat|nested] 45+ messages in thread
* [PATCH v5 23/25] rnbd: include client and server modules into kernel compilation
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (21 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 22/25] rnbd: server: sysfs interface functions Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 24/25] rnbd: a bit of documentation Jack Wang
` (3 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
Add rnbd Makefile, Kconfig and also corresponding lines into upper
block layer files.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
drivers/block/Kconfig | 2 ++
drivers/block/Makefile | 1 +
drivers/block/rnbd/Kconfig | 28 ++++++++++++++++++++++++++++
drivers/block/rnbd/Makefile | 17 +++++++++++++++++
4 files changed, 48 insertions(+)
create mode 100644 drivers/block/rnbd/Kconfig
create mode 100644 drivers/block/rnbd/Makefile
diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig
index 1bb8ec575352..1636a7d9e91e 100644
--- a/drivers/block/Kconfig
+++ b/drivers/block/Kconfig
@@ -468,4 +468,6 @@ config BLK_DEV_RSXX
To compile this driver as a module, choose M here: the
module will be called rsxx.
+source "drivers/block/rnbd/Kconfig"
+
endif # BLK_DEV
diff --git a/drivers/block/Makefile b/drivers/block/Makefile
index a53cc1e3a2d3..914f9d07835c 100644
--- a/drivers/block/Makefile
+++ b/drivers/block/Makefile
@@ -36,6 +36,7 @@ obj-$(CONFIG_BLK_DEV_PCIESSD_MTIP32XX) += mtip32xx/
obj-$(CONFIG_BLK_DEV_RSXX) += rsxx/
obj-$(CONFIG_ZRAM) += zram/
+obj-$(CONFIG_BLK_DEV_RNBD) += rnbd/
obj-$(CONFIG_BLK_DEV_NULL_BLK) += null_blk.o
null_blk-objs := null_blk_main.o
diff --git a/drivers/block/rnbd/Kconfig b/drivers/block/rnbd/Kconfig
new file mode 100644
index 000000000000..67882d2ff58f
--- /dev/null
+++ b/drivers/block/rnbd/Kconfig
@@ -0,0 +1,28 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+
+config BLK_DEV_RNBD
+ bool
+
+config BLK_DEV_RNBD_CLIENT
+ tristate "Network block device driver on top of RTRS transport"
+ depends on INFINIBAND_RTRS_CLIENT
+ select BLK_DEV_RNBD
+ help
+ RNBD client is a network block device driver using rdma transport.
+
+ RNBD client allows for mapping of a remote block devices over
+ RTRS protocol from a target system where RNBD server is running.
+
+ If unsure, say N.
+
+config BLK_DEV_RNBD_SERVER
+ tristate "Network block device over RDMA Infiniband server support"
+ depends on INFINIBAND_RTRS_SERVER
+ select BLK_DEV_RNBD
+ help
+ RNBD server is the server side of RNBD using rdma transport.
+
+ RNBD server allows for exporting local block devices to a remote client
+ over RTRS protocol.
+
+ If unsure, say N.
diff --git a/drivers/block/rnbd/Makefile b/drivers/block/rnbd/Makefile
new file mode 100644
index 000000000000..125c3576f221
--- /dev/null
+++ b/drivers/block/rnbd/Makefile
@@ -0,0 +1,17 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+
+ccflags-y := -Idrivers/infiniband/ulp/rtrs
+
+rnbd-client-y := rnbd-clt.o \
+ rnbd-common.o \
+ rnbd-clt-sysfs.o
+
+rnbd-server-y := rnbd-srv.o \
+ rnbd-common.o \
+ rnbd-srv-dev.o \
+ rnbd-srv-sysfs.o
+
+obj-$(CONFIG_BLK_DEV_RNBD_CLIENT) += rnbd-client.o
+obj-$(CONFIG_BLK_DEV_RNBD_SERVER) += rnbd-server.o
+
+-include $(src)/compat/compat.mk
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 24/25] rnbd: a bit of documentation
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (22 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 23/25] rnbd: include client and server modules into kernel compilation Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-20 15:51 ` [PATCH v5 25/25] MAINTAINERS: Add maintainers for RNBD/RTRS modules Jack Wang
` (2 subsequent siblings)
26 siblings, 0 replies; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev, linux-kernel
From: Jack Wang <jinpu.wang@cloud.ionos.com>
README with description of major sysfs entries, sysfs documentation
are moved to ABI dir as Bart suggested.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
Cc: linux-kernel@vger.kernel.org
---
Documentation/ABI/testing/sysfs-block-rnbd | 51 ++++++++
.../ABI/testing/sysfs-class-rnbd-client | 117 ++++++++++++++++++
.../ABI/testing/sysfs-class-rnbd-server | 57 +++++++++
drivers/block/rnbd/README | 80 ++++++++++++
4 files changed, 305 insertions(+)
create mode 100644 Documentation/ABI/testing/sysfs-block-rnbd
create mode 100644 Documentation/ABI/testing/sysfs-class-rnbd-client
create mode 100644 Documentation/ABI/testing/sysfs-class-rnbd-server
create mode 100644 drivers/block/rnbd/README
diff --git a/Documentation/ABI/testing/sysfs-block-rnbd b/Documentation/ABI/testing/sysfs-block-rnbd
new file mode 100644
index 000000000000..dea84280291a
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-block-rnbd
@@ -0,0 +1,51 @@
+What: /sys/block/rnbd<N>/rnbd/unmap_device
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+To unmap a volume, "normal" or "force" has to be written to:
+ /sys/block/rnbd<N>/rnbd/unmap_device
+
+When "normal" is used, the operation will fail with EBUSY if any process
+is using the device. When "force" is used, the device is also unmapped
+when device is in use. All I/Os that are in progress will fail.
+
+Example:
+
+ # echo "normal" > /sys/block/rnbd0/rnbd/unmap_device
+
+What: /sys/block/rnbd<N>/rnbd/state
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+The file contains the current state of the block device. The state file
+returns "open" when the device is successfully mapped from the server
+and accepting I/O requests. When the connection to the server gets
+disconnected in case of an error (e.g. link failure), the state file
+returns "closed" and all I/O requests submitted to it will fail with -EIO.
+
+What: /sys/block/rnbd<N>/rnbd/session
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+RNBD uses RTRS session to transport the data between client and
+server. The entry "session" contains the name of the session, that
+was used to establish the RTRS session. It's the same name that
+was passed as server parameter to the map_device entry.
+
+What: /sys/block/rnbd<N>/rnbd/mapping_path
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains the path that was passed as "device_path" to the map_device
+operation.
+
+What: /sys/block/rnbd<N>/rnbd/access_mode
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains the device access mode: ro, rw or migration.
diff --git a/Documentation/ABI/testing/sysfs-class-rnbd-client b/Documentation/ABI/testing/sysfs-class-rnbd-client
new file mode 100644
index 000000000000..7466a6aa2641
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-class-rnbd-client
@@ -0,0 +1,117 @@
+What: /sys/class/rnbd-client
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Provide information about RNBD-client.
+All sysfs files that are not read-only provide the usage information on read:
+
+Example:
+ # cat /sys/class/rnbd-client/ctl/map_device
+
+ > Usage: echo "sessname=<name of the rtrs session> path=<[srcaddr,]dstaddr>
+ > [path=<[srcaddr,]dstaddr>] device_path=<full path on remote side>
+ > [access_mode=<ro|rw|migration>] > map_device
+ >
+ > addr ::= [ ip:<ipv4> | ip:<ipv6> | gid:<gid> ]
+
+What: /sys/class/rnbd-client/ctl/map_device
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Expected format is the following:
+
+ sessname=<name of the rtrs session>
+ path=<[srcaddr,]dstaddr> [path=<[srcaddr,]dstaddr> ...]
+ device_path=<full path on remote side>
+ [access_mode=<ro|rw|migration>]
+
+Where:
+
+sessname: accepts a string not bigger than 256 chars, which identifies
+ a given session on the client and on the server.
+ I.e. "clt_hostname-srv_hostname" could be a natural choice.
+
+path: describes a connection between the client and the server by
+ specifying destination and, when required, the source address.
+ The addresses are to be provided in the following format:
+
+ ip:<IPv6>
+ ip:<IPv4>
+ gid:<GID>
+
+ for example:
+
+ path=ip:10.0.0.66
+ The single addr is treated as the destination.
+ The connection will be established to this
+ server from any client IP address.
+
+ path=ip:10.0.0.66,ip:10.0.1.66
+ First addr is the source address and the second
+ is the destination.
+
+ If multiple "path=" options are specified multiple connection
+ will be established and data will be sent according to
+ the selected multipath policy (see RTRS mp_policy sysfs entry
+ description).
+
+device_path: Path to the block device on the server side. Path is specified
+ relative to the directory on server side configured in the
+ 'dev_search_path' module parameter of the rnbd_server.
+ The rnbd_server prepends the <device_path> received from client
+ with <dev_search_path> and tries to open the
+ <dev_search_path>/<device_path> block device. On success,
+ a /dev/rnbd<N> device file, a /sys/block/rnbd_client/rnbd<N>/
+ directory and an entry in /sys/class/rnbd-client/ctl/devices
+ will be created.
+
+ If 'dev_search_path' contains '%SESSNAME%', then each session can
+ have different devices namespace, e.g. server was configured with
+ the following parameter "dev_search_path=/run/rnbd-devs/%SESSNAME%",
+ client has this string "sessname=blya device_path=sda", then server
+ will try to open: /run/rnbd-devs/blya/sda.
+
+access_mode: the access_mode parameter specifies if the device is to be
+ mapped as "ro" read-only or "rw" read-write. The server allows
+ a device to be exported in rw mode only once. The "migration"
+ access mode has to be specified if a second mapping in read-write
+ mode is desired.
+
+ By default "rw" is used.
+
+Exit Codes:
+
+If the device is already mapped it will fail with EEXIST. If the input
+has an invalid format it will return EINVAL. If the device path cannot
+be found on the server, it will fail with ENOENT.
+
+Finding device file after mapping
+---------------------------------
+
+After mapping, the device file can be found by:
+ o The symlink /sys/class/rnbd-client/ctl/devices/<device_id>
+ points to /sys/block/<dev-name>. The last part of the symlink destination
+ is the same as the device name. By extracting the last part of the
+ path the path to the device /dev/<dev-name> can be build.
+
+ o /dev/block/$(cat /sys/class/rnbd-client/ctl/devices/<device_id>/dev)
+
+How to find the <device_id> of the device is described on the next
+section.
+
+What: /sys/class/rnbd-client/ctl/devices/
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+For each device mapped on the client a new symbolic link is created as
+/sys/class/rnbd-client/ctl/devices/<device_id>, which points
+to the block device created by rnbd (/sys/block/rnbd<N>/).
+The <device_id> of each device is created as follows:
+
+- If the 'device_path' provided during mapping contains slashes ("/"),
+ they are replaced by exclamation mark ("!") and used as as the
+ <device_id>. Otherwise, the <device_id> will be the same as the
+ "device_path" provided.
diff --git a/Documentation/ABI/testing/sysfs-class-rnbd-server b/Documentation/ABI/testing/sysfs-class-rnbd-server
new file mode 100644
index 000000000000..1faa9faa8ca3
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-class-rnbd-server
@@ -0,0 +1,57 @@
+What: /sys/class/rnbd-server
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description: provide information about RNBD-server.
+
+What: /sys/class/rnbd-server/ctl/
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+When a client maps a device, a directory entry with the name of the
+block device is created under /sys/class/rnbd-server/ctl/devices/.
+
+What: /sys/class/rnbd-server/ctl/devices/<device_name>/block_dev
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Is a symlink to the sysfs entry of the exported device.
+
+Example:
+
+ block_dev -> ../../../../class/block/ram0
+
+What: /sys/class/rnbd-server/ctl/devices/<device_name>/sessions/
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+For each client a particular device is exported to, following directory will be
+created:
+
+/sys/class/rnbd-server/ctl/devices/<device_name>/sessions/<session-name>/
+
+When the device is unmapped by that client, the directory will be removed.
+
+What: /sys/class/rnbd-server/ctl/devices/<device_name>/sessions/<session-name>/read_only
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains '1' if device is mapped read-only, otherwise '0'.
+
+What: /sys/class/rnbd-server/ctl/devices/<device_name>/sessions/<session-name>/mapping_path
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains the relative device path provided by the user during mapping.
+
+What: /sys/class/rnbd-server/ctl/devices/<device_name>/sessions/<session-name>/access_mode
+Date: Jan 2020
+KernelVersion: 5.6
+Contact: Jack Wang <jinpu.wang@cloud.ionos.com> Danil Kipnis <danil.kipnis@cloud.ionos.com>
+Description:
+Contains the device access mode: ro, rw or migration.
diff --git a/drivers/block/rnbd/README b/drivers/block/rnbd/README
new file mode 100644
index 000000000000..2134d60ac3ae
--- /dev/null
+++ b/drivers/block/rnbd/README
@@ -0,0 +1,80 @@
+***************************************
+Infiniband Network Block Device (RNBD)
+***************************************
+
+Introduction
+------------
+
+RNBD (InfiniBand Network Block Device) is a pair of kernel modules
+(client and server) that allow for remote access of a block device on
+the server over RTRS protocol using the RDMA (InfiniBand, RoCE, iWarp)
+transport. After being mapped, the remote block devices can be accessed
+on the client side as local block devices.
+
+I/O is transferred between client and server by the RTRS transport
+modules. The administration of RNBD and RTRS modules is done via
+sysfs entries.
+
+Requirements
+------------
+
+ RTRS kernel modules
+
+Quick Start
+-----------
+
+Server side:
+ # modprobe rnbd_server
+
+Client side:
+ # modprobe rnbd_client
+ # echo "sessname=blya path=ip:10.50.100.66 device_path=/dev/ram0" > \
+ /sys/devices/virtual/rnbd-client/ctl/map_device
+
+ Where "sessname=" is a session name, a string to identify the session
+ on client and on server sides; "path=" is a destination IP address or
+ a pair of a source and a destination IPs, separated by comma. Multiple
+ "path=" options can be specified in order to use multipath (see RTRS
+ description for details); "device_path=" is the block device to be
+ mapped from the server side. After the session to the server machine is
+ established, the mapped device will appear on the client side under
+ /dev/rnbd<N>.
+
+
+RNBD-Server Module Parameters
+==============================
+
+dev_search_path
+---------------
+
+When a device is mapped from the client, the server generates the path
+to the block device on the server side by concatenating dev_search_path
+and the "device_path" that was specified in the map_device operation.
+
+The default dev_search_path is: "/".
+
+dev_search_path option can also contain %SESSNAME% in order to provide
+different deviec namespaces for different sessions. See "device_path"
+option for details.
+
+==============================
+Protocol (rnbd/rnbd-proto.h)
+==============================
+
+1. Before mapping first device from a given server, client sends an
+RNBD_MSG_SESS_INFO to the server. Server responds with
+RNBD_MSG_SESS_INFO_RSP. Currently the messages only contain the protocol
+version for backward compatibility.
+
+2. Client requests to open a device by sending RNBD_MSG_OPEN message. This
+contains the path to the device and access mode (read-only or writable).
+Server responds to the message with RNBD_MSG_OPEN_RSP. This contains
+a 32 bit device id to be used for IOs and device "geometry" related
+information: side, max_hw_sectors, etc.
+
+3. Client attaches RNBD_MSG_IO to each IO message send to a device. This
+message contains device id, provided by server in his rnbd_msg_open_rsp,
+sector to be accessed, read-write flags and bi_size.
+
+4. Client closes a device by sending RNBD_MSG_CLOSE which contains only the
+device id provided by the server.
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH v5 25/25] MAINTAINERS: Add maintainers for RNBD/RTRS modules
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (23 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 24/25] rnbd: a bit of documentation Jack Wang
@ 2019-12-20 15:51 ` Jack Wang
2019-12-22 9:55 ` Gal Pressman
2019-12-21 10:17 ` [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Leon Romanovsky
2020-01-02 18:18 ` Jason Gunthorpe
26 siblings, 1 reply; 45+ messages in thread
From: Jack Wang @ 2019-12-20 15:51 UTC (permalink / raw)
To: linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
From: Jack Wang <jinpu.wang@cloud.ionos.com>
Danil and me will maintain RNBD/RTRS modules.
Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
---
MAINTAINERS | 14 ++++++++++++++
1 file changed, 14 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index cc0a4a8ae06a..e0caeac43fc8 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -7955,6 +7955,20 @@ IBM ServeRAID RAID DRIVER
S: Orphan
F: drivers/scsi/ips.*
+RNBD BLOCK DRIVERS
+M: Danil Kipnis <danil.kipnis@cloud.ionos.com>
+M: Jack Wang <jinpu.wang@cloud.ionos.com>
+L: linux-block@vger.kernel.org
+S: Maintained
+F: drivers/block/rnbd/
+
+RTRS TRANSPORT DRIVERS
+M: Danil Kipnis <danil.kipnis@cloud.ionos.com>
+M: Jack Wang <jinpu.wang@cloud.ionos.com>
+L: linux-rdma@vger.kernel.org
+S: Maintained
+F: drivers/infiniband/ulp/rtrs/
+
ICH LPC AND GPIO DRIVER
M: Peter Tyser <ptyser@xes-inc.com>
S: Maintained
--
2.17.1
^ permalink raw reply related [flat|nested] 45+ messages in thread* Re: [PATCH v5 25/25] MAINTAINERS: Add maintainers for RNBD/RTRS modules
2019-12-20 15:51 ` [PATCH v5 25/25] MAINTAINERS: Add maintainers for RNBD/RTRS modules Jack Wang
@ 2019-12-22 9:55 ` Gal Pressman
2019-12-23 7:20 ` Jinpu Wang
0 siblings, 1 reply; 45+ messages in thread
From: Gal Pressman @ 2019-12-22 9:55 UTC (permalink / raw)
To: Jack Wang, linux-block, linux-rdma
Cc: axboe, hch, sagi, bvanassche, leon, dledford, danil.kipnis,
jinpu.wang, rpenyaev
On 20/12/2019 17:51, Jack Wang wrote:
> From: Jack Wang <jinpu.wang@cloud.ionos.com>
>
> Danil and me will maintain RNBD/RTRS modules.
>
> Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
> ---
> MAINTAINERS | 14 ++++++++++++++
> 1 file changed, 14 insertions(+)
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index cc0a4a8ae06a..e0caeac43fc8 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -7955,6 +7955,20 @@ IBM ServeRAID RAID DRIVER
> S: Orphan
> F: drivers/scsi/ips.*
>
> +RNBD BLOCK DRIVERS
> +M: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> +M: Jack Wang <jinpu.wang@cloud.ionos.com>
> +L: linux-block@vger.kernel.org
> +S: Maintained
> +F: drivers/block/rnbd/
> +
> +RTRS TRANSPORT DRIVERS
> +M: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> +M: Jack Wang <jinpu.wang@cloud.ionos.com>
> +L: linux-rdma@vger.kernel.org
> +S: Maintained
> +F: drivers/infiniband/ulp/rtrs/
> +
> ICH LPC AND GPIO DRIVER
> M: Peter Tyser <ptyser@xes-inc.com>
> S: Maintained
>
Entries should be added in alphabetical order.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 25/25] MAINTAINERS: Add maintainers for RNBD/RTRS modules
2019-12-22 9:55 ` Gal Pressman
@ 2019-12-23 7:20 ` Jinpu Wang
0 siblings, 0 replies; 45+ messages in thread
From: Jinpu Wang @ 2019-12-23 7:20 UTC (permalink / raw)
To: Gal Pressman
Cc: Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Bart Van Assche,
Leon Romanovsky, Doug Ledford, Danil Kipnis, rpenyaev
On Sun, Dec 22, 2019 at 10:55 AM Gal Pressman <galpress@amazon.com> wrote:
>
> On 20/12/2019 17:51, Jack Wang wrote:
> > From: Jack Wang <jinpu.wang@cloud.ionos.com>
> >
> > Danil and me will maintain RNBD/RTRS modules.
> >
> > Signed-off-by: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> > Signed-off-by: Jack Wang <jinpu.wang@cloud.ionos.com>
> > ---
> > MAINTAINERS | 14 ++++++++++++++
> > 1 file changed, 14 insertions(+)
> >
> > diff --git a/MAINTAINERS b/MAINTAINERS
> > index cc0a4a8ae06a..e0caeac43fc8 100644
> > --- a/MAINTAINERS
> > +++ b/MAINTAINERS
> > @@ -7955,6 +7955,20 @@ IBM ServeRAID RAID DRIVER
> > S: Orphan
> > F: drivers/scsi/ips.*
> >
> > +RNBD BLOCK DRIVERS
> > +M: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> > +M: Jack Wang <jinpu.wang@cloud.ionos.com>
> > +L: linux-block@vger.kernel.org
> > +S: Maintained
> > +F: drivers/block/rnbd/
> > +
> > +RTRS TRANSPORT DRIVERS
> > +M: Danil Kipnis <danil.kipnis@cloud.ionos.com>
> > +M: Jack Wang <jinpu.wang@cloud.ionos.com>
> > +L: linux-rdma@vger.kernel.org
> > +S: Maintained
> > +F: drivers/infiniband/ulp/rtrs/
> > +
> > ICH LPC AND GPIO DRIVER
> > M: Peter Tyser <ptyser@xes-inc.com>
> > S: Maintained
> >
>
> Entries should be added in alphabetical order.
Thanks, Gal,
Will fix the next round.
Regards,
Jack
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (24 preceding siblings ...)
2019-12-20 15:51 ` [PATCH v5 25/25] MAINTAINERS: Add maintainers for RNBD/RTRS modules Jack Wang
@ 2019-12-21 10:17 ` Leon Romanovsky
2020-01-02 18:18 ` Jason Gunthorpe
26 siblings, 0 replies; 45+ messages in thread
From: Leon Romanovsky @ 2019-12-21 10:17 UTC (permalink / raw)
To: Jack Wang
Cc: linux-block, linux-rdma, axboe, hch, sagi, bvanassche, dledford,
danil.kipnis, jinpu.wang, rpenyaev
On Fri, Dec 20, 2019 at 04:50:44PM +0100, Jack Wang wrote:
> Hi all,
>
> here is V5 of the RTRS (former IBTRS) rdma transport library and the
> corresponding RNBD (former IBNBD) rdma network block device.
>
> Main changes are the following:
> 1. Fix the security problem pointed out by Jason
> 2. Implement code-style/readability/API/etc suggestions by Bart van Assche
> 3. Rename IBTRS and IBNBD to RTRS and RNBD accordingly
> 4. Fileio mode support in rnbd-srv has been removed.
>
> The main functional change is a fix for the security problem pointed out by
> Jason and discussed both on the mailing list and during the last LPC RDMA MC 2019.
> On the server side we now invalidate in RTRS each rdma buffer before we hand it
> over to RNBD server and in turn to the block layer. A new rkey is generated and
> registered for the buffer after it returns back from the block layer and RNBD
> server. The new rkey is sent back to the client along with the IO result.
> The procedure is the default behaviour of the driver. This invalidation and
> registration on each IO causes performance drop of up to 20%. A user of the
> driver may choose to load the modules with this mechanism switched off
> (always_invalidate=N), if he understands and can take the risk of a malicious
> client being able to corrupt memory of a server it is connected to. This might
> be a reasonable option in a scenario where all the clients and all the servers
> are located within a secure datacenter.
>
> Huge thanks to Bart van Assche for the very detailed review of both RNBD and
> RTRS. These included suggestions for style fixes, better readability and
> documentation, code simplifications, eliminating usage of deprecated APIs,
> too many to name.
>
> The transport library and the network block device using it have been renamed to
> RTRS and RNBD accordingly in order to reflect the fact that they are based on
> the rdma subsystem and not bound to InfiniBand only.
>
> Fileio mode support in rnbd-server is not so efficent as pointed out by Bart,
> and we can use loop device in between if there is need, hence we just
> removed the fileio mode support.
Thanks for pushing the code forward.
^ permalink raw reply [flat|nested] 45+ messages in thread* Re: [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device
2019-12-20 15:50 [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Jack Wang
` (25 preceding siblings ...)
2019-12-21 10:17 ` [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device Leon Romanovsky
@ 2020-01-02 18:18 ` Jason Gunthorpe
2020-01-03 12:39 ` Jinpu Wang
26 siblings, 1 reply; 45+ messages in thread
From: Jason Gunthorpe @ 2020-01-02 18:18 UTC (permalink / raw)
To: Jack Wang
Cc: linux-block, linux-rdma, axboe, hch, sagi, bvanassche, leon,
dledford, danil.kipnis, jinpu.wang, rpenyaev
On Fri, Dec 20, 2019 at 04:50:44PM +0100, Jack Wang wrote:
> Hi all,
>
> here is V5 of the RTRS (former IBTRS) rdma transport library and the
> corresponding RNBD (former IBNBD) rdma network block device.
>
> Main changes are the following:
> 1. Fix the security problem pointed out by Jason
> 2. Implement code-style/readability/API/etc suggestions by Bart van Assche
> 3. Rename IBTRS and IBNBD to RTRS and RNBD accordingly
> 4. Fileio mode support in rnbd-srv has been removed.
>
> The main functional change is a fix for the security problem pointed out by
> Jason and discussed both on the mailing list and during the last LPC RDMA MC 2019.
> On the server side we now invalidate in RTRS each rdma buffer before we hand it
> over to RNBD server and in turn to the block layer. A new rkey is generated and
> registered for the buffer after it returns back from the block layer and RNBD
> server. The new rkey is sent back to the client along with the IO result.
> The procedure is the default behaviour of the driver. This invalidation and
> registration on each IO causes performance drop of up to 20%. A user of the
> driver may choose to load the modules with this mechanism switched
> off
So, how does this compare now to nvme over fabrics?
I recall there were questiosn why we needed yet another RDMA block
transport?
Jason
^ permalink raw reply [flat|nested] 45+ messages in thread* Re: [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device
2020-01-02 18:18 ` Jason Gunthorpe
@ 2020-01-03 12:39 ` Jinpu Wang
2020-01-03 16:28 ` Bart Van Assche
0 siblings, 1 reply; 45+ messages in thread
From: Jinpu Wang @ 2020-01-03 12:39 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Bart Van Assche,
Leon Romanovsky, Doug Ledford, Danil Kipnis, rpenyaev
On Thu, Jan 2, 2020 at 7:19 PM Jason Gunthorpe <jgg@ziepe.ca> wrote:
>
> On Fri, Dec 20, 2019 at 04:50:44PM +0100, Jack Wang wrote:
> > Hi all,
> >
> > here is V5 of the RTRS (former IBTRS) rdma transport library and the
> > corresponding RNBD (former IBNBD) rdma network block device.
> >
> > Main changes are the following:
> > 1. Fix the security problem pointed out by Jason
> > 2. Implement code-style/readability/API/etc suggestions by Bart van Assche
> > 3. Rename IBTRS and IBNBD to RTRS and RNBD accordingly
> > 4. Fileio mode support in rnbd-srv has been removed.
> >
> > The main functional change is a fix for the security problem pointed out by
> > Jason and discussed both on the mailing list and during the last LPC RDMA MC 2019.
> > On the server side we now invalidate in RTRS each rdma buffer before we hand it
> > over to RNBD server and in turn to the block layer. A new rkey is generated and
> > registered for the buffer after it returns back from the block layer and RNBD
> > server. The new rkey is sent back to the client along with the IO result.
> > The procedure is the default behaviour of the driver. This invalidation and
> > registration on each IO causes performance drop of up to 20%. A user of the
> > driver may choose to load the modules with this mechanism switched
> > off
>
> So, how does this compare now to nvme over fabrics?
>
> I recall there were questiosn why we needed yet another RDMA block
> transport?
>
> Jason
Performance results for the v5.5-rc1 kernel are here:
link: https://github.com/ionos-enterprise/ibnbd/tree/develop/performance/v5-v5.5-rc1
Some workloads RNBD are faster, some workloads NVMeoF are faster.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device
2020-01-03 12:39 ` Jinpu Wang
@ 2020-01-03 16:28 ` Bart Van Assche
2020-01-06 17:07 ` Jinpu Wang
0 siblings, 1 reply; 45+ messages in thread
From: Bart Van Assche @ 2020-01-03 16:28 UTC (permalink / raw)
To: Jinpu Wang, Jason Gunthorpe
Cc: Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Leon Romanovsky, Doug Ledford,
Danil Kipnis, rpenyaev
On 1/3/20 4:39 AM, Jinpu Wang wrote:
> Performance results for the v5.5-rc1 kernel are here:
> link: https://github.com/ionos-enterprise/ibnbd/tree/develop/performance/v5-v5.5-rc1
>
> Some workloads RNBD are faster, some workloads NVMeoF are faster.
Thank you for having shared these graphs.
Do the graphs in RNBD-SinglePath.pdf show that NVMeOF achieves similar
or higher IOPS, higher bandwidth and lower latency than RNBD for
workloads with a block size of 4 KB and also for mixed workloads with
less than 20 disks, whether or not invalidation is enabled for RNBD?
Is it already clear why NVMeOF performance drops if the number of disks
is above 25? Is that perhaps caused by contention on the block layer tag
allocator because multiple NVMe namespaces share a tag set? Can that
contention be avoided by increasing the NVMeOF queue depth further?
Thanks,
Bart.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device
2020-01-03 16:28 ` Bart Van Assche
@ 2020-01-06 17:07 ` Jinpu Wang
2020-01-07 10:56 ` Jinpu Wang
0 siblings, 1 reply; 45+ messages in thread
From: Jinpu Wang @ 2020-01-06 17:07 UTC (permalink / raw)
To: Bart Van Assche
Cc: Jason Gunthorpe, Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Leon Romanovsky, Doug Ledford,
Danil Kipnis, rpenyaev
On Fri, Jan 3, 2020 at 5:29 PM Bart Van Assche <bvanassche@acm.org> wrote:
>
> On 1/3/20 4:39 AM, Jinpu Wang wrote:
> > Performance results for the v5.5-rc1 kernel are here:
> > link: https://github.com/ionos-enterprise/ibnbd/tree/develop/performance/v5-v5.5-rc1
> >
> > Some workloads RNBD are faster, some workloads NVMeoF are faster.
>
> Thank you for having shared these graphs.
>
> Do the graphs in RNBD-SinglePath.pdf show that NVMeOF achieves similar
> or higher IOPS, higher bandwidth and lower latency than RNBD for
> workloads with a block size of 4 KB and also for mixed workloads with
> less than 20 disks, whether or not invalidation is enabled for RNBD?
Hi Bart,
Yes, that's the result on one pair of Server with Connect X4 HCA, I
did another test on another
2 servers with Connect X5 HCA, results are quite different, we will
double-check the
performance results also on old machines, will share new results later.
>
> Is it already clear why NVMeOF performance drops if the number of disks
> is above 25? Is that perhaps caused by contention on the block layer tag
> allocator because multiple NVMe namespaces share a tag set? Can that
> contention be avoided by increasing the NVMeOF queue depth further?
No yet, will check.
>
> Thanks,
>
> Bart.
>
>
Thanks
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device
2020-01-06 17:07 ` Jinpu Wang
@ 2020-01-07 10:56 ` Jinpu Wang
2020-01-16 16:41 ` Bart Van Assche
0 siblings, 1 reply; 45+ messages in thread
From: Jinpu Wang @ 2020-01-07 10:56 UTC (permalink / raw)
To: Bart Van Assche
Cc: Jason Gunthorpe, Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Leon Romanovsky, Doug Ledford,
Danil Kipnis, rpenyaev
On Mon, Jan 6, 2020 at 6:07 PM Jinpu Wang <jinpu.wang@cloud.ionos.com> wrote:
>
> On Fri, Jan 3, 2020 at 5:29 PM Bart Van Assche <bvanassche@acm.org> wrote:
> >
> > On 1/3/20 4:39 AM, Jinpu Wang wrote:
> > > Performance results for the v5.5-rc1 kernel are here:
> > > link: https://github.com/ionos-enterprise/ibnbd/tree/develop/performance/v5-v5.5-rc1
> > >
> > > Some workloads RNBD are faster, some workloads NVMeoF are faster.
> >
> > Thank you for having shared these graphs.
> >
> > Do the graphs in RNBD-SinglePath.pdf show that NVMeOF achieves similar
> > or higher IOPS, higher bandwidth and lower latency than RNBD for
> > workloads with a block size of 4 KB and also for mixed workloads with
> > less than 20 disks, whether or not invalidation is enabled for RNBD?
> Hi Bart,
>
> Yes, that's the result on one pair of Server with Connect X4 HCA, I
> did another test on another
> 2 servers with Connect X5 HCA, results are quite different, we will
> double-check the
> performance results also on old machines, will share new results later.
>
here are the results with ConnectX5 HCA MT4119 + Intel(R) Xeon(R) Gold
6130 CPU @ 2.10GHz, sorry no graph for now,
will prepare the next round.
disks 4k nvme dual 4k nvme single 4k rnbd dual 4k rnbd single
4k rnbd-inv dual 4k rnbd-inv single
x1 251637.436256 254312.068793 270886.311369 260934.306569
218632.336766 190800.519948
x2 460894.610539 463925.907409 496318.068193 466374.862514
418960.30397 372848.815118
x3 603263.673633 605004.49955 675073.892611 614552.144786
586223.077692 524221.977802
x4 731648.935106 733174.482552 850245.575442 743062.493751
744380.361964 656861.813819
x5 827732.326767 827444.855514 1026939.306069 840515.548445
897801.719828 762707.329267
x6 876705.329467 873963.0037 1142399.960004 876974.70253
1037773.522648 834073.892611
x7 893808.719128 893268.073193 1239282.471753 892728.027197
1135570.742926 871336.966303
x8 906589.741026 905938.006199 1287178.964207 906189.381062
1225040.9959 895292.070793
x9 912048.09519 912400.259974 1386211.878812 913885.311469
1302472.964884 910176.282372
x10 915566.243376 915602.739726 1442959.70403 916288.871113
1350296.325879 914966.40336
x11 917116.188381 916905.809419 1418574.942506 916264.373563
1370438.698083 915255.874413
x12 915852.814719 917710.128987 1423534.546545 916348.386452
1352357.364264 914966.656684
x13 919042.69573 918819.536093 1429697.830217 917631.036896
1378083.824558 916519.161677
x14 920000.49995 920031.59684 1443317.268273 917562.843716
1395023.56936 918935.706429
x15 920160.883912 920367.363264 1445306.425863 918278.472153
1440776.944611 916352.265921
x16 920652.869426 920673.832617 1454705.229477 917902.948526
1455708.2501 918198.001998
x17 916892.310769 916883.623275 1491071.841948 918936.706329
1436507.428457 917182.934132
x18 917247.775222 917762.523748 1612129.058036 918546.835949
1488716.583417 919521.095781
x19 920084.791521 920349.930014 1615690.87821 915371.496958
1406747.32954 918347.248577
x20 922339.232154 922208.058388 1591415.958404 917922.631526
1343806.744488 918903.393214
x21 923141.771646 923297.040592 1581547.169811 919371.025795
1342568.406347 919157.752944
x22 923063.787243 924072.385523 1574972.846162 919173.713143
1340318.639673 920577.995403
x23 923549.490102 924643.471306 1573597.440256 918705.060385
1333047.771337 917469.027431
x24 925584.483103 925224.955009 1578143.485651 921744.15117
1321494.708466 920001.498352
x25 926165.366927 926842.031594 1579288.271173 921845.392764
1319902.568202 920448.830702
x26 926852.729454 927382.123575 1585351.318945 922669.59912
1325670.791338 919137.796847
x27 928196.260748 928093.981204 1581427.557244 921379.155436
1325009.972078 919017.550858
x28 929330.433913 929606.778644 1581726.527347 924325.834833
1331721.074174 919557.761373
x29 929885.522895 929876.924615 1578317.436513 922977.240966
1333612.86783 921094.386736
x30 930635.972805 930599.520144 1583537.946205 922746.107784
1333446.651362 922821.171531
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device
2020-01-07 10:56 ` Jinpu Wang
@ 2020-01-16 16:41 ` Bart Van Assche
2020-01-16 16:46 ` Jinpu Wang
0 siblings, 1 reply; 45+ messages in thread
From: Bart Van Assche @ 2020-01-16 16:41 UTC (permalink / raw)
To: Jinpu Wang
Cc: Jason Gunthorpe, Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Leon Romanovsky, Doug Ledford,
Danil Kipnis, rpenyaev
On 1/7/20 2:56 AM, Jinpu Wang wrote:
> On Mon, Jan 6, 2020 at 6:07 PM Jinpu Wang <jinpu.wang@cloud.ionos.com> wrote:
>>
>> On Fri, Jan 3, 2020 at 5:29 PM Bart Van Assche <bvanassche@acm.org> wrote:
>>>
>>> On 1/3/20 4:39 AM, Jinpu Wang wrote:
>>>> Performance results for the v5.5-rc1 kernel are here:
>>>> link: https://github.com/ionos-enterprise/ibnbd/tree/develop/performance/v5-v5.5-rc1
>>>>
>>>> Some workloads RNBD are faster, some workloads NVMeoF are faster.
>>>
>>> Thank you for having shared these graphs.
>>>
>>> Do the graphs in RNBD-SinglePath.pdf show that NVMeOF achieves similar
>>> or higher IOPS, higher bandwidth and lower latency than RNBD for
>>> workloads with a block size of 4 KB and also for mixed workloads with
>>> less than 20 disks, whether or not invalidation is enabled for RNBD?
>> Hi Bart,
>>
>> Yes, that's the result on one pair of Server with Connect X4 HCA, I
>> did another test on another
>> 2 servers with Connect X5 HCA, results are quite different, we will
>> double-check the
>> performance results also on old machines, will share new results later.
>>
> here are the results with ConnectX5 HCA MT4119 + Intel(R) Xeon(R) Gold
> 6130 CPU @ 2.10GHz, sorry no graph for now,
> will prepare the next round.
>
> disks 4k nvme dual 4k nvme single 4k rnbd dual 4k rnbd single
> 4k rnbd-inv dual 4k rnbd-inv single
> x1 251637.436256 254312.068793 270886.311369 260934.306569
> 218632.336766 190800.519948
> x2 460894.610539 463925.907409 496318.068193 466374.862514
> 418960.30397 372848.815118
> x3 603263.673633 605004.49955 675073.892611 614552.144786
> 586223.077692 524221.977802
> x4 731648.935106 733174.482552 850245.575442 743062.493751
> 744380.361964 656861.813819
> x5 827732.326767 827444.855514 1026939.306069 840515.548445
> 897801.719828 762707.329267
> x6 876705.329467 873963.0037 1142399.960004 876974.70253
> 1037773.522648 834073.892611
> x7 893808.719128 893268.073193 1239282.471753 892728.027197
> 1135570.742926 871336.966303
> x8 906589.741026 905938.006199 1287178.964207 906189.381062
> 1225040.9959 895292.070793
> x9 912048.09519 912400.259974 1386211.878812 913885.311469
> 1302472.964884 910176.282372
> x10 915566.243376 915602.739726 1442959.70403 916288.871113
> 1350296.325879 914966.40336
> x11 917116.188381 916905.809419 1418574.942506 916264.373563
> 1370438.698083 915255.874413
> x12 915852.814719 917710.128987 1423534.546545 916348.386452
> 1352357.364264 914966.656684
> x13 919042.69573 918819.536093 1429697.830217 917631.036896
> 1378083.824558 916519.161677
> x14 920000.49995 920031.59684 1443317.268273 917562.843716
> 1395023.56936 918935.706429
> x15 920160.883912 920367.363264 1445306.425863 918278.472153
> 1440776.944611 916352.265921
> x16 920652.869426 920673.832617 1454705.229477 917902.948526
> 1455708.2501 918198.001998
> x17 916892.310769 916883.623275 1491071.841948 918936.706329
> 1436507.428457 917182.934132
> x18 917247.775222 917762.523748 1612129.058036 918546.835949
> 1488716.583417 919521.095781
> x19 920084.791521 920349.930014 1615690.87821 915371.496958
> 1406747.32954 918347.248577
> x20 922339.232154 922208.058388 1591415.958404 917922.631526
> 1343806.744488 918903.393214
> x21 923141.771646 923297.040592 1581547.169811 919371.025795
> 1342568.406347 919157.752944
> x22 923063.787243 924072.385523 1574972.846162 919173.713143
> 1340318.639673 920577.995403
> x23 923549.490102 924643.471306 1573597.440256 918705.060385
> 1333047.771337 917469.027431
> x24 925584.483103 925224.955009 1578143.485651 921744.15117
> 1321494.708466 920001.498352
> x25 926165.366927 926842.031594 1579288.271173 921845.392764
> 1319902.568202 920448.830702
> x26 926852.729454 927382.123575 1585351.318945 922669.59912
> 1325670.791338 919137.796847
> x27 928196.260748 928093.981204 1581427.557244 921379.155436
> 1325009.972078 919017.550858
> x28 929330.433913 929606.778644 1581726.527347 924325.834833
> 1331721.074174 919557.761373
> x29 929885.522895 929876.924615 1578317.436513 922977.240966
> 1333612.86783 921094.386736
> x30 930635.972805 930599.520144 1583537.946205 922746.107784
> 1333446.651362 922821.171531
Hi Jack,
What does "dual" mean? What explains the big difference between the NVMe
and RNBD results for the "dual" columns?
Thanks,
Bart.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v5 00/25] RTRS (former IBTRS) rdma transport library and the corresponding RNBD (former IBNBD) rdma network block device
2020-01-16 16:41 ` Bart Van Assche
@ 2020-01-16 16:46 ` Jinpu Wang
0 siblings, 0 replies; 45+ messages in thread
From: Jinpu Wang @ 2020-01-16 16:46 UTC (permalink / raw)
To: Bart Van Assche
Cc: Jason Gunthorpe, Jack Wang, linux-block, linux-rdma, Jens Axboe,
Christoph Hellwig, Sagi Grimberg, Leon Romanovsky, Doug Ledford,
Danil Kipnis, Roman Penyaev
On Thu, Jan 16, 2020 at 5:41 PM Bart Van Assche <bvanassche@acm.org> wrote:
>
> On 1/7/20 2:56 AM, Jinpu Wang wrote:
> > On Mon, Jan 6, 2020 at 6:07 PM Jinpu Wang <jinpu.wang@cloud.ionos.com> wrote:
> >>
> >> On Fri, Jan 3, 2020 at 5:29 PM Bart Van Assche <bvanassche@acm.org> wrote:
> >>>
> >>> On 1/3/20 4:39 AM, Jinpu Wang wrote:
> >>>> Performance results for the v5.5-rc1 kernel are here:
> >>>> link: https://github.com/ionos-enterprise/ibnbd/tree/develop/performance/v5-v5.5-rc1
> >>>>
> >>>> Some workloads RNBD are faster, some workloads NVMeoF are faster.
> >>>
> >>> Thank you for having shared these graphs.
> >>>
> >>> Do the graphs in RNBD-SinglePath.pdf show that NVMeOF achieves similar
> >>> or higher IOPS, higher bandwidth and lower latency than RNBD for
> >>> workloads with a block size of 4 KB and also for mixed workloads with
> >>> less than 20 disks, whether or not invalidation is enabled for RNBD?
> >> Hi Bart,
> >>
> >> Yes, that's the result on one pair of Server with Connect X4 HCA, I
> >> did another test on another
> >> 2 servers with Connect X5 HCA, results are quite different, we will
> >> double-check the
> >> performance results also on old machines, will share new results later.
> >>
> > here are the results with ConnectX5 HCA MT4119 + Intel(R) Xeon(R) Gold
> > 6130 CPU @ 2.10GHz, sorry no graph for now,
> > will prepare the next round.
> >
> > disks 4k nvme dual 4k nvme single 4k rnbd dual 4k rnbd single
> > 4k rnbd-inv dual 4k rnbd-inv single
> > x1 251637.436256 254312.068793 270886.311369 260934.306569
> > 218632.336766 190800.519948
> > x2 460894.610539 463925.907409 496318.068193 466374.862514
> > 418960.30397 372848.815118
> > x3 603263.673633 605004.49955 675073.892611 614552.144786
> > 586223.077692 524221.977802
> > x4 731648.935106 733174.482552 850245.575442 743062.493751
> > 744380.361964 656861.813819
> > x5 827732.326767 827444.855514 1026939.306069 840515.548445
> > 897801.719828 762707.329267
> > x6 876705.329467 873963.0037 1142399.960004 876974.70253
> > 1037773.522648 834073.892611
> > x7 893808.719128 893268.073193 1239282.471753 892728.027197
> > 1135570.742926 871336.966303
> > x8 906589.741026 905938.006199 1287178.964207 906189.381062
> > 1225040.9959 895292.070793
> > x9 912048.09519 912400.259974 1386211.878812 913885.311469
> > 1302472.964884 910176.282372
> > x10 915566.243376 915602.739726 1442959.70403 916288.871113
> > 1350296.325879 914966.40336
> > x11 917116.188381 916905.809419 1418574.942506 916264.373563
> > 1370438.698083 915255.874413
> > x12 915852.814719 917710.128987 1423534.546545 916348.386452
> > 1352357.364264 914966.656684
> > x13 919042.69573 918819.536093 1429697.830217 917631.036896
> > 1378083.824558 916519.161677
> > x14 920000.49995 920031.59684 1443317.268273 917562.843716
> > 1395023.56936 918935.706429
> > x15 920160.883912 920367.363264 1445306.425863 918278.472153
> > 1440776.944611 916352.265921
> > x16 920652.869426 920673.832617 1454705.229477 917902.948526
> > 1455708.2501 918198.001998
> > x17 916892.310769 916883.623275 1491071.841948 918936.706329
> > 1436507.428457 917182.934132
> > x18 917247.775222 917762.523748 1612129.058036 918546.835949
> > 1488716.583417 919521.095781
> > x19 920084.791521 920349.930014 1615690.87821 915371.496958
> > 1406747.32954 918347.248577
> > x20 922339.232154 922208.058388 1591415.958404 917922.631526
> > 1343806.744488 918903.393214
> > x21 923141.771646 923297.040592 1581547.169811 919371.025795
> > 1342568.406347 919157.752944
> > x22 923063.787243 924072.385523 1574972.846162 919173.713143
> > 1340318.639673 920577.995403
> > x23 923549.490102 924643.471306 1573597.440256 918705.060385
> > 1333047.771337 917469.027431
> > x24 925584.483103 925224.955009 1578143.485651 921744.15117
> > 1321494.708466 920001.498352
> > x25 926165.366927 926842.031594 1579288.271173 921845.392764
> > 1319902.568202 920448.830702
> > x26 926852.729454 927382.123575 1585351.318945 922669.59912
> > 1325670.791338 919137.796847
> > x27 928196.260748 928093.981204 1581427.557244 921379.155436
> > 1325009.972078 919017.550858
> > x28 929330.433913 929606.778644 1581726.527347 924325.834833
> > 1331721.074174 919557.761373
> > x29 929885.522895 929876.924615 1578317.436513 922977.240966
> > 1333612.86783 921094.386736
> > x30 930635.972805 930599.520144 1583537.946205 922746.107784
> > 1333446.651362 922821.171531
>
> Hi Jack,
>
> What does "dual" mean? What explains the big difference between the NVMe
> and RNBD results for the "dual" columns?
I just sent out v7 with also graphs, it's here:
https://github.com/ionos-enterprise/ibnbd/tree/develop/performance/v7-v5.5-rc2
Measurements are performed in single path (-single) and
multipath(-dual) setting. For multipath measurement two different
policies have been tested:
- RNBD - min-inflight (choose path with minimum inflights)
- NVMEoF - numa iopolicy
>
> Thanks,
>
> Bart.
>
>
Thanks
^ permalink raw reply [flat|nested] 45+ messages in thread