I noticed last night that the nightly tests with Intel compilers on vanquish and octagon were not completing properly (no new output on the web page). The tests were also running the 1.4.x branch. I switched to the 1.4 branch and restarted the tests, but there were still some problems with Intel compilers. I have again started the tests. Let's hold off on releasing 1.4 until we figure out what is happening. I think the tests were fine until Monday. Rajeev
On 03/31/2011 08:01 AM, Rajeev Thakur wrote:
I noticed last night that the nightly tests with Intel compilers on vanquish and octagon were not completing properly (no new output on the web page). The tests were also running the 1.4.x branch. I switched to the 1.4 branch and restarted the tests, but there were still some problems with Intel compilers. I have again started the tests. Let's hold off on releasing 1.4 until we figure out what is happening. I think the tests were fine until Monday.
Also: 1. there seems to be some problem for the binary packagers with the MPL library ABI version. I'm trying to figure out what's going on there. 2. the NUDT guys reported that Hydra is having some problem with running on the full Tianhe machine. I'm looking into that as well. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji
On 03/31/2011 08:30 AM, Pavan Balaji wrote:
1. there seems to be some problem for the binary packagers with the MPL library ABI version. I'm trying to figure out what's going on there.
This didn't turn out to be a real problem.
2. the NUDT guys reported that Hydra is having some problem with running on the full Tianhe machine. I'm looking into that as well.
These guys are on vacation and will only try out my patches on Thursday or later. I don't know how long this will take, so we should not hold up the release for it. If there are no other pending issues, should we plan to push out 1.4 tomorrow? -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji
OK with me. Rajeev On Apr 3, 2011, at 8:54 AM, Pavan Balaji wrote:
On 03/31/2011 08:30 AM, Pavan Balaji wrote:
1. there seems to be some problem for the binary packagers with the MPL library ABI version. I'm trying to figure out what's going on there.
This didn't turn out to be a real problem.
2. the NUDT guys reported that Hydra is having some problem with running on the full Tianhe machine. I'm looking into that as well.
These guys are on vacation and will only try out my patches on Thursday or later. I don't know how long this will take, so we should not hold up the release for it.
If there are no other pending issues, should we plan to push out 1.4 tomorrow?
-- Pavan
-- Pavan Balaji http://www.mcs.anl.gov/~balaji
FYI, I'm working on the below ticket, so I didn't release 1.4 yet. Turns out that applications that spawn processes that span more than one node will fail because of this. https://trac.mcs.anl.gov/projects/mpich2/ticket/1434 -- Pavan On 04/03/2011 11:39 AM, Rajeev Thakur wrote:
OK with me.
Rajeev
On Apr 3, 2011, at 8:54 AM, Pavan Balaji wrote:
On 03/31/2011 08:30 AM, Pavan Balaji wrote:
1. there seems to be some problem for the binary packagers with the MPL library ABI version. I'm trying to figure out what's going on there.
This didn't turn out to be a real problem.
2. the NUDT guys reported that Hydra is having some problem with running on the full Tianhe machine. I'm looking into that as well.
These guys are on vacation and will only try out my patches on Thursday or later. I don't know how long this will take, so we should not hold up the release for it.
If there are no other pending issues, should we plan to push out 1.4 tomorrow?
-- Pavan
-- Pavan Balaji http://www.mcs.anl.gov/~balaji
-- Pavan Balaji http://www.mcs.anl.gov/~balaji
Darius, Dave and I discussed a hack-fix for this bug for now, which I'm planning to commit in shortly (this will disable shared-memory optimizations for dynamically spawned processes). I'm planning to release 1.4rc2 instead of 1.4 final as hwloc-1.1.2 got released with a few bug fixes that I care about. I've incorporated it into hydra, but I want it to be tested a bit more before we release 1.4. -- Pavan On 04/05/2011 11:26 AM, Pavan Balaji wrote:
FYI, I'm working on the below ticket, so I didn't release 1.4 yet.
Turns out that applications that spawn processes that span more than one node will fail because of this.
https://trac.mcs.anl.gov/projects/mpich2/ticket/1434
-- Pavan
On 04/03/2011 11:39 AM, Rajeev Thakur wrote:
OK with me.
Rajeev
On Apr 3, 2011, at 8:54 AM, Pavan Balaji wrote:
On 03/31/2011 08:30 AM, Pavan Balaji wrote:
1. there seems to be some problem for the binary packagers with the MPL library ABI version. I'm trying to figure out what's going on there.
This didn't turn out to be a real problem.
2. the NUDT guys reported that Hydra is having some problem with running on the full Tianhe machine. I'm looking into that as well.
These guys are on vacation and will only try out my patches on Thursday or later. I don't know how long this will take, so we should not hold up the release for it.
If there are no other pending issues, should we plan to push out 1.4 tomorrow?
-- Pavan
-- Pavan Balaji http://www.mcs.anl.gov/~balaji
-- Pavan Balaji http://www.mcs.anl.gov/~balaji
I've created the 1.4rc2 tarball and placed it here: http://www.mcs.anl.gov/research/projects/mpich2staging/balaji/mpich2/downloa... Please try it out and let me know if there are any problems. I'll release it in a few hours. -- Pavan On 04/06/2011 01:11 PM, Pavan Balaji wrote:
Darius, Dave and I discussed a hack-fix for this bug for now, which I'm planning to commit in shortly (this will disable shared-memory optimizations for dynamically spawned processes).
I'm planning to release 1.4rc2 instead of 1.4 final as hwloc-1.1.2 got released with a few bug fixes that I care about. I've incorporated it into hydra, but I want it to be tested a bit more before we release 1.4.
-- Pavan
On 04/05/2011 11:26 AM, Pavan Balaji wrote:
FYI, I'm working on the below ticket, so I didn't release 1.4 yet.
Turns out that applications that spawn processes that span more than one node will fail because of this.
https://trac.mcs.anl.gov/projects/mpich2/ticket/1434
-- Pavan
On 04/03/2011 11:39 AM, Rajeev Thakur wrote:
OK with me.
Rajeev
On Apr 3, 2011, at 8:54 AM, Pavan Balaji wrote:
On 03/31/2011 08:30 AM, Pavan Balaji wrote:
1. there seems to be some problem for the binary packagers with the MPL library ABI version. I'm trying to figure out what's going on there.
This didn't turn out to be a real problem.
2. the NUDT guys reported that Hydra is having some problem with running on the full Tianhe machine. I'm looking into that as well.
These guys are on vacation and will only try out my patches on Thursday or later. I don't know how long this will take, so we should not hold up the release for it.
If there are no other pending issues, should we plan to push out 1.4 tomorrow?
-- Pavan
-- Pavan Balaji http://www.mcs.anl.gov/~balaji
-- Pavan Balaji http://www.mcs.anl.gov/~balaji
participants (2)
-
Pavan Balaji -
Rajeev Thakur