On Sat, Apr 19, 2014 at 7:25 PM, Christoph Lameter c...@linux.com wrote:
On Fri, 18 Apr 2014, Nicolas Carlier wrote:
If the receiving QP does not have buffers available then the HCA will
silently drop UD packets. This is somethig that tripped us up initialy. So
its lossless only from HCA
c4iw_alloc() bails out without freeing the storage that 'devp' points to.
Picked up by Coverity - CID 1204241.
Fixes: fa658a98a2 (RDMA/cxgb4: Use the BAR2/WC path for kernel QPs and T5
devices)
Signed-off-by: Christoph Jaeger christophjae...@linux.com
---
drivers/infiniband/hw/cxgb4/device.c |
-Original Message-
From: Christoph Jaeger [mailto:christophjae...@linux.com]
Sent: Monday, April 21, 2014 10:03 AM
To: sw...@chelsio.com; rol...@kernel.org; sean.he...@intel.com;
hal.rosenst...@gmail.com
Cc: linux-rdma@vger.kernel.org; linux-ker...@vger.kernel.org; Christoph Jaeger
The effect of the check for -T resolve was reversed, so that -T with
any invalid value would result in the -T resolve behavior, and
-T resolve would result in an error.
---
examples/rstream.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/examples/rstream.c
Hi, Sean,
Add:
Signed-off-by: Patrick MacArthur pmaca...@iol.unh.edu
I can resend the patch if you would like.
Thanks,
Patrick
--
Patrick MacArthur pmaca...@iol.unh.edu
AIM: PmacarthAtIOL
Research and Development, High Performance Networking and Storage
UNH InterOperability Laboratory
On
Signed-off-by: Patrick MacArthur pmaca...@iol.unh.edu
I can resend the patch if you would like.
thanks - no need to resend
--
To unsubscribe from this list: send the line unsubscribe linux-rdma in
the body of a message to majord...@vger.kernel.org
More majordomo info at
On Mon, 21 Apr 2014, Nicolas Carlier wrote:
Interresting, I don't see this counter neither on fw/sw release notes nor on
the
systems. What the name of this counter ? On which HCA do you see this
counter ?
The mlx4 driver supports this counter and its in
/sys/class/infiniband/diag_counters/
On Wed, 2 Apr 2014, Or Gerlitz wrote:
From: Eli Cohen e...@dev.mellanox.co.il
Add support for the block multicast loopback QP creation flag along
the proper firmware API for that.
Could we get this merged? We need this on a lot of systems.
--
To unsubscribe from this list: send the line
We need this option for many hosts to avoid backflow of multicast
packages. Could we get that merged?
From 18ceae090b02b3055382e11c305dcb334d938122 Mon Sep 17 00:00:00 2001
From: Or Gerlitz ogerl...@mellanox.com
Date: Tue, 4 Mar 2014 17:20:00 +0200
Subject: [PATCH] IB/mlx4: Allow to always
commit f360d88a2efd upstream.
On Mon, Apr 21, 2014 at 10:56 AM, Christoph Lameter c...@linux.com wrote:
On Wed, 2 Apr 2014, Or Gerlitz wrote:
From: Eli Cohen e...@dev.mellanox.co.il
Add support for the block multicast loopback QP creation flag along
the proper firmware API for that.
On Mon, 21 Apr 2014, Roland Dreier wrote:
commit f360d88a2efd upstream.
Great. Is there a corresponding patch for mlx4? Dont see that in the
kernel and we have mostly mlx4.
--
To unsubscribe from this list: send the line unsubscribe linux-rdma in
the body of a message to
Is something more than commit 521e575b9a73 (from 2008) required?
On Mon, Apr 21, 2014 at 11:59 AM, Christoph Lameter c...@linux.com wrote:
On Mon, 21 Apr 2014, Roland Dreier wrote:
commit f360d88a2efd upstream.
Great. Is there a corresponding patch for mlx4? Dont see that in the
kernel and
On Mon, Apr 21, 2014 at 9:09 PM, Christoph Lameter c...@linux.com wrote:
We need this option for many hosts to avoid backflow of multicast
packages. Could we get that merged?
[...]
Currently, there's no way for user-space applications to specify
the IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK QP
On Mon, 21 Apr 2014, Roland Dreier wrote:
Is something more than commit 521e575b9a73 (from 2008) required?
Reviewing that right now. If it does what it seems to do then we are fine.
--
To unsubscribe from this list: send the line unsubscribe linux-rdma in
the body of a message to
On Mon, 21 Apr 2014, Or Gerlitz wrote:
Currently, there's no way for user-space applications to specify
the IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK QP creation flags
defined by commit 47ee1b9 IB/core: Add support for multicast loopback
blocking.
As a result, applications who send and
On Tue, Apr 22, 2014 at 12:06 AM, Christoph Lameter c...@linux.com wrote:
On Mon, 21 Apr 2014, Or Gerlitz wrote:
Currently, there's no way for user-space applications to specify
the IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK QP creation flags
defined by commit 47ee1b9 IB/core: Add support for
From: Allen Andrews allen.andr...@emulex.com
Two memory region leaks were found during testing:
1. rpcrdma_buffer_create: While allocating RPCRDMA_FRMR's
ib_alloc_fast_reg_mr is called and then ib_alloc_fast_reg_page_list is
called. If ib_alloc_fast_reg_page_list returns an error it bails out
From: Steve Wise sw...@opengridcomputing.com
Some rdma devices don't support a fast register page list depth of
at least RPCRDMA_MAX_DATA_SEGS. So xprtrdma needs to chunk its fast
register regions according to the minimum of the device max supported
depth or RPCRDMA_MAX_DATA_SEGS.
After folks tried out RPCRDMA_REGISTER support as I requested in
the cover letter of the last version of this series, existing
problems were discovered already in the upstream kernel, starting
with the problem addressed by Steve's LOCAL_WRITE patch from last
week.
Rather than address them, this
An IB provider can invoke rpcrdma_conn_func() in an IRQ context,
thus rpcrdma_conn_func() cannot be allowed to directly invoke
generic RPC functions like xprt_wake_pending_tasks().
Signed-off-by: Chuck Lever chuck.le...@oracle.com
Tested-by: Steve Wise sw...@opengridcomputing.com
---
Section 4 of RFC 5667 (NFS/RDMA) says:
The server MUST ignore any Read list for other NFS procedures,
as well as additional Read list entries beyond the first in the
list.
Our XDR code adds a zero pad at the end of NFS WRITEs and SYMLINKs
whose content is not a multiple of 4 octets long.
Clean up: This memory registration mode is slow and was never
meant for use in production environments. Remove it to reduce
implementation complexity.
Signed-off-by: Chuck Lever chuck.le...@oracle.com
Tested-by: Steve Wise sw...@opengridcomputing.com
---
net/sunrpc/xprtrdma/rpc_rdma.c |8
The MEMWINDOWS and MEMWINDOES_ASYNC memory registration modes were
intended as stop-gap modes before the introduction of FRMR. They
are now considered obsolete.
MEMWINDOWS_ASYNC is also considered unsafe because it can leave
client memory registered and exposed for an indeterminant time after
ALLPHYSICAL is not a safe memory registration mode because it
permits NFS servers to write anywhere in a client's memory. NFS
server bugs could result in client memory being overwritten.
This can be useful for embedded systems which do not support more
surgical RDMA memory registration and
An audit of in-kernel RDMA providers that do not support the FRMR
memory registration shows that several of them support MTHCAFMR.
Prefer MTHCAFMR when FRMR is not supported.
If MTHCAFMR is not supported, only then choose ALLPHYSICAL.
Signed-off-by: Chuck Lever chuck.le...@oracle.com
---
Clean up: rpcrdma_ep_destroy() returns a value that is used
only to print a debugging message. rpcrdma_ep_destroy() already
prints debugging messages in all error cases.
Make rpcrdma_ep_destroy() return void instead.
Signed-off-by: Chuck Lever chuck.le...@oracle.com
Tested-by: Steve Wise
Clean up: All remaining callers of rpcrdma_deregister_external()
pass NULL as the last argument, so remove that argument.
Signed-off-by: Chuck Lever chuck.le...@oracle.com
Tested-by: Steve Wise sw...@opengridcomputing.com
---
net/sunrpc/xprtrdma/rpc_rdma.c |2 +-
The current CQ handler uses the ib_wc.opcode field to distinguish
between event types. However, the contents of that field are not
reliable if the completion status is not IB_WC_SUCCESS.
When an error completion occurs on a send event, the CQ handler
schedules a tasklet with something that is not
Skip the ib_poll_cq() after re-arming, if the provider knows there
are no additional items waiting. (Have a look at commit ed23a727 for
more details).
Signed-off-by: Chuck Lever chuck.le...@oracle.com
---
net/sunrpc/xprtrdma/verbs.c | 14 ++
1 files changed, 10 insertions(+), 4
While marshaling an RPC/RDMA request, the inline_{rsize,wsize}
settings determine whether an inline request is used, or whether
read or write chunks lists are built. The current default value of
these settings is 1024. Any RPC request smaller than 1024 bytes is
sent to the NFS server completely
Sagi Grimberg sa...@dev.mellanox.co.il points out that a steady
stream of CQ events could starve other work because of the boundless
loop pooling in rpcrdma_{send,recv}_poll().
Instead of a (potentially infinite) while loop, return after
collecting a budgeted number of completions.
Note that the
All kernel RDMA providers except amso1100 support either MTHCAFMR
or FRMR, both of which are faster than REGISTER. amso1100 can
continue to use ALLPHYSICAL.
The only other ULP consumer in the kernel that uses the reg_phys_mr
verb is Lustre.
Signed-off-by: Chuck Lever chuck.le...@oracle.com
---
If the selected memory registration mode is not supported by the
underlying provider/HCA, the NFS mount command reports that there was
an invalid mount option, and fails. This is misleading.
Reporting a problem allocating memory is a lot closer to the truth.
Signed-off-by: Chuck Lever
Change the completion handlers to grab up to 16 items per
ib_poll_cq() call. No extra ib_poll_cq() is needed if fewer than 16
items are returned.
Signed-off-by: Chuck Lever chuck.le...@oracle.com
---
net/sunrpc/xprtrdma/verbs.c | 56 ++-
On Tue, 22 Apr 2014, Or Gerlitz wrote:
Urgh. So we have all these flags and we cannot use them?
Christoph, the support (kernel IB core API and HW drivers e.g mlx4 and
now also mlx5) is there for **kernel** consumers (e.g IPoIB). and
missing for **user-space** consumers whose calls are
35 matches
Mail list logo