> Though one can consider the fall-back in reverse order i.e. if the
> rdma connection fails continue with the already established connection (over
> the normal inet socket).
When I consider fallback, one of the issues is handling the case where one of
the two sides is not using rsockets. This
On Sat, 9 Jun 2012, Hefty, Sean wrote:
But to map standard networking applications to rsockets we will run into the
above problem i.e. fork() will not work. It would be very useful to
allow for the standard networking paradigm of: bind()->listen()->accept()-
->fork(), and then the server goes b
This patch is almost formatted perfectly, except in the future when
you send someone else's patch, please include your own Signed-off-by
line as well at the end of the chain. In other words, the email
should go
>From: Mahesh Vardhamanaiah
>
>Fixed code to read the max wqe and max rqe values fro
thanks, applied
--
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 http://vger.kernel.org/majordomo-info.html
On Mon, Jun 11, 2012 at 8:10 AM, Ursula Braun wrote:
> If zero based virtual addressing has never been implemented within
> Linux, does this mean an ib_post_send() with opcode IB_WR_RDMA_WRITE
> requires specification of both - rdma.rkey AND rdma.remote_addr?
> And thus means there is a need to ex
On Fri, 2012-05-25 at 09:01 -0700, Roland Dreier wrote:
> > Thanks for your good and quick help. May I ask another question about
> > zero based virtual addressing: The Infiniband Architecture Specification
> > talks about specifying the "type of VA" when registering a memory
> > region. Checking t
I was looking through old integer overflows, and I was still wondering
if this is buggy.
regards,
dan carpenter
On 12/1/11, Dan Carpenter wrote:
> This is a static checker fix, and I'm not super familiar with this code.
>
> My checker complains that user_wr->num_sge + sg_ind can have an integer
From: Oren Duer
MCG para-virtualization support includes:
- Creating multicast groups by VFs, and keeping accounting of them
- Leaving multicast groups by VFs
- SM will only be updated with real changes in the overall picture of MCGs
status
- Creation of MGID=0 groups (let SM choose MGID)
Note
The requires:
1. Replacing the paravirtualized pkey index (inserted
by the guest) with the real pkey index
2. For UD qp's, placing the guest's true source gid index
in the address path structure mgid field, and setting
the ud_force_mgid bit so that the mgid is taken from
the qp context
Allocate sriov paravirtualization resources and mad demuxing contexts
on the master.
This has two parts. The first part is to initialize the structures
to contain the contexts. This is done at master startup time
(mlx4_ib_init_sriov).
The second part is to actually create the tunneling resource
Special QPs are para-virtualized.
vHCAs are not given direct access to QP0/1. Rather, these QPs are operated by a
special context hosted by the PF, which mediates access to/from vHCAs.
This is done by opening a “tunnel” per vHCA port per QP0/1. A tunnel comprises
a pair of UD QPs: a “Tunnel QP” i
For an IB port, a slave should not show port active until that
slave has a valid alias-guid (provided by the subnet manager).
Therefore the port-up event should be passed to a slave only after
both the port is up, and the slave's alias-guid has been set.
Also, provide the infrastructure for propag
The MAD_IFC firmware command fulfills two functions.
In the first case, it is used in the QP0/QP1 MAD-handling flow to
obtain information from the FW (for answering queries), and for
setting variables in the HCA (MAD SET packets).
In this function, MAD_IFC should provide the FW (physical) view
of
To allow easy paravirtualization of pkey and gid table sizes,
keep paravirtualized sizes in mlx4_dev->caps, but save the
actual physical sizes in FW in struct: mlx4_dev->phys_cap.
In addition, in SRIOV mode, do the following:
1. Reduce reported pkey table size by 1.
This is done to reserve the
For IB ports, we paravirtualize the GUID at index 0 on slaves.
The GUID at index 0 seen by a slave is the actual GUID occupying
the GUID table at the slave-id index.
The driver, by default, request at startup time that subnet manager
populate its entire guid table with GUIDs. These guids are then
The guid cache introduced by the alias guid feature contains the all the
GUIDs for each port. GUIDs 1..127 are obtained from the subnet manager.
GUID 0, however, is obtained from the FW and cannot be modified by the
subnet manager. It must be introduced separately into the guid cache.
Initialize
This directory is added only for the master -- slaves do
not have it.
The sysfs iov directory is used to manage and examine the port
pkey and guid paravirtualization.
Under iov/ports, the administrator may examine the gid and pkey
tables as they are present in the device (and as are seen in the
"
Under SRIOV-IB, each slave has its own separate copy of the port-capabilities
flags. Thus, for example, the master can run a subnet manager
(which causes the IS_SM bit to be set in the master's port capabilities)
without affecting the port capabilities seen by the slaves
(i.e., the IS_SM bit will
Note: RoCE is still not supported on slaves. This
patch will be changed for V1 to reflect this.
Signed-off-by: Jack Morgenstein
---
drivers/infiniband/hw/mlx4/main.c |5 -
drivers/net/ethernet/mellanox/mlx4/fw.c |6 --
2 files changed, 0 insertions(+), 11 deletions(-)
diff
Normally, INIT_PORT and CLOSE_PORT are invoked when special QP0
transitions to RTR, or transitions to ERR/RESET respectively.
In SRIOV mode, however, the master is also paravirtualized.
This in turn requires that we not do INIT_PORT until the
entire QP0 path (real QP0 and proxy QP0) is ready to re
1. Slave may not set the IS_SM capability for the port.
2. No DEV_MGR in multifunc mode.
Signed-off-by: Jack Morgenstein
---
drivers/net/ethernet/mellanox/mlx4/port.c | 10 ++
include/linux/mlx4/device.h |5 +
2 files changed, 15 insertions(+), 0 deletions(-)
dif
From: Amir Vadai
In CM para-virtualization:
1. Incoming requests are steered to the correct vHCA according to the embedded
GID
2. Communication IDs on outgoing requests are replaced by a globally unique ID,
generated
by the PPF, since there is no synchronization of ID generation between
gue
1. allow only master to change node description
2. prevent ah leakage in send mads
3. take device part number from PCI structure, so that
guests see the VF part number (and not the PF part number)
4. place the device revision ID into caps structure at startup
5. SET_PORT in update_gids_task need
When port pkey table potentially contains both full and partial
membership copies for the same pkey, we need a function to find
the exact (16-bit) pkey index.
This is particularly necessary when the master forwards QP1 MADS
sent by guests. If the guest has sent the MAD with a limited
membership p
pkey change and guid change events are not of interest to all slaves,
but only to those slaves which "see" the table slots whose contents
have change.
For example, if the guid at port 1, index 5 has changed in the
PPF, we wish to propagate the gid-change event only to the function
which has that g
Implement debug printouts to assist in supporting the driver.
The intent is to provide more detail in the log for errors which
would otherwise be returned as -EINVAL.
The facility is on-off switchable at run-time
(i.e., module-parameter controlled, off by default).
Error output is added here spe
Modify mlx4_dev_cap to allow IB support when SRIOV is active.
Modify mlx4_slave_cap to set the "rdma-supported" bit in its
flags area, and pass that to the guests (this is done in
QUERY_FUNC_CAP and its wrapper).
However, Do not yet activate IB support (i.e., do not yet remove the
error return at
In addition, pass the proxy and tunnel QP numbers to slaves so
the driver can perform sqp paravirtualization.
Signed-off-by: Jack Morgenstein
---
drivers/net/ethernet/mellanox/mlx4/fw.c| 21 +++
drivers/net/ethernet/mellanox/mlx4/fw.h|3 ++
drivers/net/e
The port management change event can replace smp_snoop. If the
capability bit for this event is set in dev-caps, the event is used
(by the driver setting the PORT_MNG_CHG_EVENT bit in the async event
mask in the MAP_EQ fw command). In this case, when the driver
passes incoming SMP PORT_INFO SET ma
The sriov special-qp tunneling mechanism uses proxy special-qp's
(instead of the real special qps) for MADs on guests. These proxy qp's
send their packets to a "tunnel" qp owned by the master. The master
then forwards the MAD (after any required paravirtualization) to the
real special QP, which s
From: Erez Shitrit
This query is needed for SRIOV alias GUID support.
The query is implemented per the IB Spec definition
in section 15.2.5.18 (GuidInfoRecord).
Signed-off-by: Erez Shitrit
Signed-off-by: Jack Morgenstein
Signed-off-by: Or Gerlitz
---
drivers/infiniband/core/sa_query.c | 133
This patch set adds SRIOV support for IB interfaces.
Patches 1-13 are "precondition" patches.
Patches 14-29 actually implement the feature.
This patch set introduces Infiniband SRIOV support for ConnectX2 and ConnectX3
devices. Each function presents itself as an independent vHCA (virtual HCA) t
1. Introduce the basic sriov parvirtualization context objects
for multiplexing and demultiplexing MADs.
2. Introduce support for the new proxy and tunnel QP types.
This patch introduces the objects required by the master
for managing QP paravirtualization for guests.
struct mlx4_ib_sriov{} is
The high-order bit of the port_num field is used for the forced-loopback flag.
Note: we just noticed that we made illegal use here of the MSB
of the port_num field in the ib core ah structure. We will correct
this error in the V1 patch set.
We did not want to delay the submission due to this.
S
Enhance the cached and non-cached pkey table lookups to enable limited and full
members of the same pkey to co-exist in the pkey table.
This is necessary for SRIOV to allow for a scheme where some guests would have
the full
membership pkey in their virtual pkey table, where other guests on the sa
Currently, VFs have 0 in their dev->caps.function field.
This is a valid pci id (usually of the PF).
Pass an invalid pci id to the VF via QUERY_FW, to make sure that
if the value gets accessed in the VF driver, it will not be valid.
Signed-off-by: Jack Morgenstein
---
drivers/net/ethernet/mella
These macros will be reused by the mlx4 SRIOV-IB CM
paravirtualization code, and there is no reason to have
them declared both in the IB core in the mlx4 IB driver.
Signed-off-by: Jack Morgenstein
Signed-off-by: Or Gerlitz
---
drivers/infiniband/core/cm_msgs.h | 12
include/rdma/
Fixed polling RQ/SRQ error CQE polling.
Returning error CQE to consumer for error case which was not returned
previously.
Signed-off-by: Parav Pandit
---
drivers/infiniband/hw/ocrdma/ocrdma_verbs.c |4 +++-
1 files changed, 3 insertions(+), 1 deletions(-)
diff --git a/drivers/infiniband/hw
38 matches
Mail list logo
