Add collective member address and threading information
on a per transport basis. Call create/free service with
HCA open/close.
Signed-off-by: Arlin Davis
---
dapl/openib_ucm/dapl_ib_util.h | 17 +
dapl/openib_ucm/device.c | 13 +
2 files changed, 30 inserti
New collective support and the FCA provider will only be built with
configure option of "--enable-coll-type=fca". Dependencies include
fca devel package, fca library, and mverbs library. This will not
be included in default builds.
Signed-off-by: Arlin Davis
---
Makefile.am | 16 +++--
Add prototype and code to post extended events on dispatcher and
include collective definitions to dat_event_str function.
Signed-off-by: Arlin Davis
---
dapl/common/dapl_evd_util.c | 61 ++-
dapl/common/dapl_evd_util.h |7 +
2 files changed, 67
Add collective member address and threading information
on a per transport basis. Call create/free service with
HCA open/close.
Signed-off-by: Arlin Davis
---
dapl/openib_scm/dapl_ib_util.h | 18 +-
dapl/openib_scm/device.c | 13 +
2 files changed, 30 insert
Add collective member address and threading information
on a per transport basis. Call create/free service with
HCA open/close.
Signed-off-by: Arlin Davis
---
dapl/openib_cma/dapl_ib_util.h | 17 +
dapl/openib_cma/device.c | 13 +
2 files changed, 30 inserti
Signed-off-by: Arlin Davis
---
dapl/openib_common/util.c | 20
1 files changed, 20 insertions(+), 0 deletions(-)
diff --git a/dapl/openib_common/util.c b/dapl/openib_common/util.c
index 5076df8..29fc12f 100644
--- a/dapl/openib_common/util.c
+++ b/dapl/openib_common/util.
Signed-off-by: Arlin Davis
---
dapl/openib_common/ib_extensions.c | 190
1 files changed, 190 insertions(+), 0 deletions(-)
diff --git a/dapl/openib_common/ib_extensions.c
b/dapl/openib_common/ib_extensions.c
index 3c418e1..c85323c 100644
--- a/dapl/openib
Signed-off-by: Arlin Davis
---
dapl/include/dapl_debug.h |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/dapl/include/dapl_debug.h b/dapl/include/dapl_debug.h
index 6e332e6..ff473e3 100644
--- a/dapl/include/dapl_debug.h
+++ b/dapl/include/dapl_debug.h
@@ -69,7 +69,8 @
Support for bcast, barrier, reduce, allreduce, allgather, allgatherv
Signed-off-by: Arlin Davis
---
dapl/openib_common/collectives/fca_provider.c | 1400 +++
dapl/openib_common/collectives/fca_provider.h | 100 ++
dapl/openib_common/collectives/ib_collectives.h | 228 ++
New definitions for offloaded MPI collectives. Roll IB extensions to v2.0.6
and clean up some counter definitions for consistency.
Signed-off-by: Arlin Davis
---
dat/include/dat2/dat_ib_extensions.h | 685 +++---
1 files changed, 624 insertions(+), 61 deletions(-)
The collective extensions are designed to support MPI and general
multicast operations over IB fabrics that support offloaded collectives.
Where feasible, they come as close to MPI semantics as possible.
Unless otherwise stated, all members participating in a data collective
operation must call the
> Things only get complicated when the domain-allocator process allocates a
> single domain and simply
> uses that single domain for all jobs (i.e., the domain is never de-allocated
> for the lifetime of the
> allocating process, and the allocating process is the server for all jobs).
To help with
Some kernel components pin user space memory (infiniband and perf)
(by increasing the page count) and account that memory as "mlocked".
The difference between mlocking and pinning is:
A. mlocked pages are marked with PG_mlocked and are exempt from
swapping. Page migration may move them around
On Wed, Aug 10, 2011 at 11:18:33AM -0600, Rolf Manderscheid wrote:
> Hi Alex,
> >There are more places where lids defined as integer. Why did you choose to
> >change only these two?
>
> This structure defines what request packets look like on the wire to
> the simulator. The motivation for the ch
Bart,
Keep them coming. I have been using my available time for code to get the
crc32 patch upstream but that is about done. I've started to look at your
comments and really appreciate the time you are taking.
Bob
> -Original Message-
> From: linux-rdma-ow...@vger.kernel.org [mailto:lin
On Sun, Jul 24, 2011 at 9:43 PM, wrote:
> +int rxe_arbiter(void *arg)
> +{
> + int err;
> + unsigned long flags;
> + struct rxe_dev *rxe = (struct rxe_dev *)arg;
> + struct sk_buff *skb;
> + struct list_head *qpl;
> + struct rxe_qp *qp;
> +
> + /* get the next qp's sen
On Sun, Jul 24, 2011 at 9:43 PM, wrote:
> + /* heuristic sliding window algorithm to keep
> +sender from overrunning receiver queues */
> + if (qp_type(qp) == IB_QPT_RC) {
> + if (psn_compare(qp->req.psn, qp->comp.psn +
> + rxe_max_req_comp
Hi Alex,
There are more places where lids defined as integer. Why did you choose to
change only these two?
This structure defines what request packets look like on the wire to the
simulator.
The motivation for the change is a simulator client for python-rdma, and
this issue
shows up when cre
Hi Alex,
On 11-08-09 08:57 AM, Alex Netes wrote
This is what I get, when running the commands in the old way:
...
I get:
time: cannot run LD_PRELOAD=scripts/../umad2sim/libumad2sim.so: No such
file or directory
Command exited with non-zero status 127
0.00user 0.00system 0:00.00elapsed ?%CPU
19 matches
Mail list logo
