OpenSM/SA: In some SA records, send ERR_REQ_INVALID response on LID out
of range error
Signed-off-by: Hal Rosenstock [EMAIL PROTECTED]
Index: opensm/osm_sa_vlarb_record.c
===
--- opensm/osm_sa_vlarb_record.c(revision 8105)
Hi Hal,
Thanks for finding and fixing. Looks good to me.
Subject: [PATCH] OpenSM/SA: In some SA records, send
ERR_REQ_INVALIDresponse on LID out of range error
OpenSM/SA: In some SA records, send ERR_REQ_INVALID response on LID
out
of range error
Yes indeed we inserted one more critical bug
fix in SDP.
This bug is cause kernel oops in case
server and client do not open the same number of sockets. Thus it can easily happened
by any user level application using socket.
The reason we added it as a patch was to decrease
the risk, so
Hi,
After some testing of IBMgtSim I have a few questions:
1) If I try to build topologies using the MTS14400.ibnl as a building
block my simulation fails with a child process exited abnormally
message. I guess this is related to ibdmchk since the ibdmchk log
contains lots of errors like the
Hi Eitan,
On 14:46 Sun 18 Jun , Eitan Zahavi wrote:
This is a third take after incorporating Sasha's comments to the
partition manager patch I have previously provided.
Two small comments below.
/**
I've just noted an inconsistency with librdmacm of udapl calling
rdma_create_id without providing the PS param.
This is the trivial patch i was using to fix the compilation.
Or.
Index: dapl/openib_cma/dapl_ib_util.c
===
---
After fixing the ucma/port space issue with the calls to rdma_create_id i
am now trying to run
$ ./Target/dapltest -T S -D OpenIB-cma
and getting an immediate segfault with the below trace, any idea?
Or.
#0 0x2af6d3a97685 in ibv_open_device (device=0x537440) at device.c:128
128
Hi Sasha,
Thanks!
These two are real bugs.
I am sending PATCHv4...
Sasha Khapyorsky wrote:
Hi Eitan,
On 14:46 Sun 18 Jun , Eitan Zahavi wrote:
This is a third take after incorporating Sasha's comments to the
partition manager patch I have previously provided.
Two small comments
On 14:46 Sun 18 Jun , Eitan Zahavi wrote:
Another one is the handling of switch limited partition cap by
clearing the switch enforcement bit (on the specific port).
Some comment about this too. See below.
+ib_api_status_t
+osm_pkey_tbl_set_new_entry(
+ IN osm_pkey_tbl_t
Hi Hal
This is a 4th take after incorporating Sasha's new 2 bug reports
for the PATCHv3 for partition manager.
The difference from previous patch is very minor:
1. p_node is initialized in pkey_mgr_update_port
2. checking for a change in peer port pkey block first check for that block is
not
Running the openib stack from Redhat on a 2.6.9-34.ELsmp kernel, dual
Xeon. Running with openmpi v1.0.2 compiled w/gcc.
While we still have the problem with btl_openib_endpoint.c returning 0
byte(s) for max inline data, and realize that another IB stack addresses
this, another problem when
Hi Hal,
This fixes typo in the usage.
Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED]
---
osm/opensm/main.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/osm/opensm/main.c b/osm/opensm/main.c
index dfb2aec..4382fdb 100644
--- a/osm/opensm/main.c
+++
Hi Sasha,
Thanks. This is yet another bug.
The fix is trivial and is noted below.
Please let me know when you are done reviewing and I will post a new patch.
EZ
Sasha Khapyorsky wrote:
On 14:46 Sun 18 Jun , Eitan Zahavi wrote:
Another one is the handling of switch limited partition cap
On Mon, 2006-06-19 at 11:20, Sasha Khapyorsky wrote:
Hi Hal,
This fixes typo in the usage.
Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED]
Thanks. Applied.
-- Hal
___
openib-general mailing list
openib-general@openib.org
http://openib.org/bugzilla/show_bug.cgi?id=145
Summary: IB Core unable to communicate IPoIB on Fedora Core 4
Product: OpenFabrics Linux
Version: 1.0rc5
Platform: X86-64
OS/Version: Other
Status: NEW
Severity: major
On 18:24 Mon 19 Jun , Eitan Zahavi wrote:
Hi Sasha,
Thanks. This is yet another bug.
The fix is trivial and is noted below.
Please let me know when you are done reviewing and I will post a new patch.
I'm done. Did some running, enforcement works as expected now.
Sasha
EZ
Sasha
OpenSM/osm_sa_link_record.c: Fix LMC 0 handling
In osm_sa_link_record.c, properly handle non base LID requests
per C15-0.1.11: Query responses shall contain a port's base LID in
any LID component of a RID. So when LMC is non 0, the only records that
appear are those with the base LID and not
On Mon, 19 Jun 2006, Or Gerlitz wrote:
I've just noted an inconsistency with librdmacm of udapl calling
rdma_create_id without providing the PS param.
This is the trivial patch i was using to fix the compilation.
Yup. The RDMA CM update on Friday afternoon broke uDAPL. Fixed in
I don't see this.
The gdb sharedlibrary output looks suspicious. /usr/local/ib isn't a
standard path for our binaries.
Are you sure everything is up-to-date on your system? Is the provided
library that you have configured to handle IA OpenIB-cma the latest
and greatest?
On Mon, 19 Jun
Hello,
Anybody working on porting the MVAPICH code to use the RDMA CM for
connection setup? Just wondering how much work is needed to make
MVAPICH run on the iwarp devices.
Thanks,
Steve.
___
openib-general mailing list
http://openib.org/bugzilla/show_bug.cgi?id=145
--- Comment #1 from [EMAIL PROTECTED] 2006-06-19 10:32 ---
If I understand what you wrote correctly, IPoIB is running fine but ibping
reports some error. What is LID 0xC (and how was this determined) ? Is the
ibping kernel module running
Or Gerlitz wrote:
After fixing the ucma/port space issue with the calls to rdma_create_id i
am now trying to run
$ ./Target/dapltest -T S -D OpenIB-cma
and getting an immediate segfault with the below trace, any idea?
Hmm, no idea. I just updated to 8112 and everything runs fine for
Hi Hal,
This fixes umad send/recv retry counter in error report.
Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED]
---
libibmad/src/rpc.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/libibmad/src/rpc.c b/libibmad/src/rpc.c
index a3b29c9..e929ba4 100644
---
Hi Hal
This is a 5th take after incorporating Sasha's last reported bug
on bad assignment of the used_blocks.
This code was run again through my verification flow and also Sasha
had run some tests too.
Eitan
Signed-off-by: Eitan Zahavi [EMAIL PROTECTED]
Index: include/opensm/osm_port.h
Hi Hal
I have added InformInfo requests to the osmStress simulator flow.
Running it overnight exposed a bug as OpenSM segfaulted during
osm_report_notice. Some debug shows the following two flows were
missing a lock. Such that under stress the InformInfo DB was altered
while being accessed by the
Hi Sven,
Please see my response below:
Eitan
Sven-Arne Reinemo wrote:
Hi,
After some testing of IBMgtSim I have a few questions:
1) If I try to build topologies using the MTS14400.ibnl as a building
block my simulation fails with a child process exited abnormally
message. I guess this
Hi Sasha,
On Mon, 2006-06-19 at 14:30, Sasha Khapyorsky wrote:
Hi Hal,
This fixes umad send/recv retry counter in error report.
Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED]
---
Thanks. Applied.
-- Hal
___
openib-general mailing list
Hi Eitan,
On Mon, 2006-06-19 at 15:12, Eitan Zahavi wrote:
Hi Hal
I have added InformInfo requests to the osmStress simulator flow.
Running it overnight exposed a bug as OpenSM segfaulted during
osm_report_notice. Some debug shows the following two flows were
missing a lock. Such that
Here is a set of patches which adds mmapped completion queues and
receive queues for the InfiniPath HCA. This required changing
some of the core code in order to return HW specific data for
the ibv_resize_cq(), ibv_modify_qp(), and ibv_modify_srq().
I have included the minimal changes to mthca
This patch contains the mthca and ehca specific changes.
Signed-off-by: Ralph Campbell [EMAIL PROTECTED]
Index: src/userspace/libmthca/src/verbs.c
===
--- src/userspace/libmthca/src/verbs.c (revision 8021)
+++
This patch contains the libipathverbs specific changes.
Signed-off-by: Ralph Campbell [EMAIL PROTECTED]
Index: src/userspace/libipathverbs/src/verbs.c
===
--- src/userspace/libipathverbs/src/verbs.c (revision 8021)
+++
This patch contains the ib_ipath kernel driver specific changes.
Signed-off-by: Ralph Campbell [EMAIL PROTECTED]
Index: src/linux-kernel/infiniband/hw/ipath/ipath_qp.c
===
--- src/linux-kernel/infiniband/hw/ipath/ipath_qp.c
hello,
i'm trying to find whether i can do a lookup of ib_gid by either
node name or node's ip address. is this information available from
the subnet manager?
thanks,
Amit.
___
openib-general mailing list
openib-general@openib.org
On Fri, 2006-06-16 at 15:07 -0700, Roland Dreier wrote:
Robert, can you confirm that the new uverbs locking scheme helps the
performance problems you're having?
Yup - that was a big help. Thanks!
Regards,
Robert.
--
Robert Walsh Email: [EMAIL PROTECTED]
Robert, can you confirm that the new uverbs locking scheme helps the
performance problems you're having?
Yup - that was a big help. Thanks!
Good, because it's upstream now...
___
openib-general mailing list
openib-general@openib.org
Steve,
Anybody working on porting the MVAPICH code to use the RDMA CM for
connection setup? Just wondering how much work is needed to make
MVAPICH run on the iwarp devices.
Thanks for your note. We are taking a look at these issues. During the
next few weeks, we will have a complete
Linus pull my for-2.6.18 branch today, so the RDMA CM is upstream
now. He still has not pulled scsi-misc-2.6 so AFAIK there is still
more required before we can merge iSER.
However, I had to add the patch below to keep iSER compiling in my
tree, since scsi/scsi_request.h is now gone from the
Roland Dreier wrote:
Linus pull my for-2.6.18 branch today, so the RDMA CM is upstream
now. He still has not pulled scsi-misc-2.6 so AFAIK there is still
more required before we can merge iSER.
However, I had to add the patch below to keep iSER compiling in my
tree, since
This removes a compile warning : is_ud might be used uninitialized
in this function.
Signed-off-by: Krishna Kumar [EMAIL PROTECTED]
---
diff -ruNp 1/core/uverbs_cmd.c 2/core/uverbs_cmd.c
--- 1/core/uverbs_cmd.c 2006-06-20 10:14:46.0 +0530
+++ 2/core/uverbs_cmd.c 2006-06-20
39 matches
Mail list logo