LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Greg Thelen <gthelen@google.com>
To: Bart.VanAssche@wdc.com, hch@lst.de, dledford@redhat.com,
	jgg@ziepe.ca, axboe@fb.com, keith.busch@intel.com,
	sagi@grimberg.me, sfrench@samba.org, linux-rdma@vger.kernel.org,
	linux-nvme@lists.infradead.org, linux-cifs@vger.kernel.org,
	linux-kernel@vger.kernel.org, gthelen@google.com
Cc: linux-rdma@vger.kernel.org, linux-nvme@lists.infradead.org,
	linux-cifs@vger.kernel.org, linux-kernel@vger.kernel.org,
	Greg Thelen <gthelen@google.com>
Subject: [PATCH 0/6] IB: make INFINIBAND_ADDR_TRANS configurable
Date: Thu, 26 Apr 2018 11:19:29 -0700	[thread overview]
Message-ID: <20180426181935.32977-1-gthelen@google.com> (raw)

This series allows for CONFIG_INFINIBAND without
CONFIG_INFINIBAND_ADDR_TRANS (aka RDMA communication manager).
Fuzzing has been finding fair number of CM bugs.
So provide an option to disable it in systems which don't need it.

Changes since last posting (https://lkml.org/lkml/2018/4/25/1266):
- added ("ib_srp: depend on INFINIBAND_ADDR_TRANS") patch

This series passed the 0-day builder.

I perfomed manual static analysis of this series based on next-20180424,
looking for dependencies.

$ git grep -w CONFIG_INFINIBAND_ADDR_TRANS
drivers/infiniband/core/Makefile:infiniband-$(CONFIG_INFINIBAND_ADDR_TRANS)     := rdma_cm.o
drivers/infiniband/core/Makefile:user_access-$(CONFIG_INFINIBAND_ADDR_TRANS)    := rdma_ucm.o

$ grep -w -e rdma_cm -e rdma_ucm drivers/infiniband/core/Makefile
infiniband-$(CONFIG_INFINIBAND_ADDR_TRANS)      := rdma_cm.o
user_access-$(CONFIG_INFINIBAND_ADDR_TRANS)     := rdma_ucm.o
rdma_cm-y :=                    cma.o
rdma_cm-$(CONFIG_INFINIBAND_ADDR_TRANS_CONFIGFS) += cma_configfs.o
rdma_ucm-y :=                   ucma.o

$ grep EXPORT drivers/infiniband/core/{cma,ucma}.c
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_event_msg);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_reject_msg);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_is_consumer_reject);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_consumer_reject_data);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(__rdma_create_id);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_create_qp);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_destroy_qp);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_init_qp_attr);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_destroy_id);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_get_service_id);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_read_gids);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_set_service_type);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_set_ib_path);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_resolve_route);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_resolve_addr);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_set_reuseaddr);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_set_afonly);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_listen);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_bind_addr);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_connect);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(__rdma_accept);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_notify);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_reject);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_disconnect);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_join_multicast);
drivers/infiniband/core/cma.c:EXPORT_SYMBOL(rdma_leave_multicast);

$ git grep -lw \
  -e __rdma_accept \
  -e __rdma_create_id \
  -e rdma_bind_addr \
  -e rdma_connect \
  -e rdma_consumer_reject_data \
  -e rdma_create_qp \
  -e rdma_destroy_id \
  -e rdma_destroy_qp \
  -e rdma_disconnect \
  -e rdma_event_msg \
  -e rdma_get_service_id \
  -e rdma_init_qp_attr \
  -e rdma_is_consumer_reject \
  -e rdma_join_multicast \
  -e rdma_leave_multicast \
  -e rdma_listen \
  -e rdma_notify \
  -e rdma_read_gids \
  -e rdma_reject \
  -e rdma_reject_msg \
  -e rdma_resolve_addr \
  -e rdma_resolve_route \
  -e rdma_set_afonly \
  -e rdma_set_ib_path \
  -e rdma_set_reuseaddr \
  -e rdma_set_service_type
drivers/infiniband/core/cma.c
drivers/infiniband/core/ucma.c
drivers/infiniband/hw/qedr/verbs.c
drivers/infiniband/ulp/iser/iser_verbs.c
drivers/infiniband/ulp/isert/ib_isert.c
drivers/infiniband/ulp/srp/ib_srp.c
drivers/infiniband/ulp/srpt/ib_srpt.c
drivers/net/ethernet/qlogic/qed/qed_rdma.c
drivers/nvme/host/rdma.c
drivers/nvme/target/rdma.c
drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.c
drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c
fs/cifs/smbdirect.c
include/linux/qed/qed_rdma_if.h
include/rdma/rdma_cm.h
include/rdma/rdma_cm_ib.h
net/9p/trans_rdma.c
net/rds/ib.c
net/rds/ib_cm.c
net/rds/rdma_transport.c
net/sunrpc/xprtrdma/svc_rdma_transport.c
net/sunrpc/xprtrdma/transport.c
net/sunrpc/xprtrdma/verbs.c

# analysis of each file:
drivers/infiniband/hw/qedr/verbs.c
drivers/net/ethernet/qlogic/qed/qed_rdma.c
include/linux/qed/qed_rdma_if.h
  - not a direct function call
include/rdma/rdma_cm.h
include/rdma/rdma_cm_ib.h
  - just prototypes
drivers/infiniband/core/cma.c
drivers/infiniband/core/ucma.c
  - depends on INFINIBAND_ADDR_TRANS
drivers/infiniband/ulp/iser/iser_verbs.c
  - depends on INFINIBAND_ISER => INFINIBAND_ADDR_TRANS
drivers/infiniband/ulp/isert/ib_isert.c
  - depends on INFINIBAND_ISERT => INFINIBAND_ADDR_TRANS
drivers/infiniband/ulp/srp/ib_srp.c
  - depends on INFINIBAND_SRP => INFINIBAND_ADDR_TRANS
    per this series' "ib_srp: depend on INFINIBAND_ADDR_TRANS"
drivers/infiniband/ulp/srpt/ib_srpt.c
  - depends on INFINIBAND_SRPT => INFINIBAND_ADDR_TRANS
    per this series' "ib_srpt: depend on INFINIBAND_ADDR_TRANS"
drivers/nvme/host/rdma.c
  - depends on NVME_RDMA => INFINIBAND_ADDR_TRANS
    per this series' "nvme: depend on INFINIBAND_ADDR_TRANS"
drivers/nvme/target/rdma.c
  - depends on NVME_TARGET_RDMA => INFINIBAND_ADDR_TRANS
    per this series' "nvmet-rdma: depend on INFINIBAND_ADDR_TRANS"
drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.c
drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h
drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c
  - depends on LNET_XPRT_IB => INFINIBAND_ADDR_TRANS
fs/cifs/smbdirect.c
  - depends on CIFS_SMB_DIRECT => INFINIBAND_ADDR_TRANS
    per this series' "cifs: smbd: depend on INFINIBAND_ADDR_TRANS"
net/9p/trans_rdma.c
  - depends on NET_9P_RDMA => INFINIBAND_ADDR_TRANS
net/rds/ib.c
net/rds/ib_cm.c
net/rds/rdma_transport.c
  - depends on RDS_RDMA => INFINIBAND_ADDR_TRANS
net/sunrpc/xprtrdma/svc_rdma_transport.c
net/sunrpc/xprtrdma/transport.c
net/sunrpc/xprtrdma/verbs.c
  - depends on SUNRPC_XPRT_RDMA => INFINIBAND_ADDR_TRANS

Greg Thelen (6):
  nvme: depend on INFINIBAND_ADDR_TRANS
  nvmet-rdma: depend on INFINIBAND_ADDR_TRANS
  ib_srpt: depend on INFINIBAND_ADDR_TRANS
  cifs: smbd: depend on INFINIBAND_ADDR_TRANS
  ib_srp: depend on INFINIBAND_ADDR_TRANS
  IB: make INFINIBAND_ADDR_TRANS configurable

 drivers/infiniband/Kconfig          | 5 ++++-
 drivers/infiniband/ulp/srp/Kconfig  | 2 +-
 drivers/infiniband/ulp/srpt/Kconfig | 2 +-
 drivers/nvme/host/Kconfig           | 2 +-
 drivers/nvme/target/Kconfig         | 2 +-
 fs/cifs/Kconfig                     | 2 +-
 6 files changed, 9 insertions(+), 6 deletions(-)

-- 
2.17.0.484.g0c8726318c-goog

             reply	other threads:[~2018-04-26 18:19 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-04-26 18:19 Greg Thelen [this message]
2018-04-26 18:19 ` [PATCH 1/6] nvme: depend on INFINIBAND_ADDR_TRANS Greg Thelen
2018-04-26 18:19 ` [PATCH 2/6] nvmet-rdma: " Greg Thelen
2018-04-26 18:19 ` [PATCH 3/6] ib_srpt: " Greg Thelen
2018-04-26 18:25   ` Bart Van Assche
2018-04-26 18:19 ` [PATCH 4/6] cifs: smbd: " Greg Thelen
2018-04-26 18:19 ` [PATCH 5/6] ib_srp: " Greg Thelen
2018-04-26 18:25   ` Bart Van Assche
2018-04-26 18:19 ` [PATCH 6/6] IB: make INFINIBAND_ADDR_TRANS configurable Greg Thelen
2018-04-26 18:26   ` Bart Van Assche
2018-04-27 16:07 ` [PATCH 0/6] " Doug Ledford

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180426181935.32977-1-gthelen@google.com \
    --to=gthelen@google.com \
    --cc=Bart.VanAssche@wdc.com \
    --cc=axboe@fb.com \
    --cc=dledford@redhat.com \
    --cc=hch@lst.de \
    --cc=jgg@ziepe.ca \
    --cc=keith.busch@intel.com \
    --cc=linux-cifs@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=sagi@grimberg.me \
    --cc=sfrench@samba.org \
    --subject='Re: [PATCH 0/6] IB: make INFINIBAND_ADDR_TRANS configurable' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).