Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-05 Thread Sebastian Kuzminsky
On Fri, Oct 5, 2012 at 5:01 PM, Samuel Thibault wrote:

> Sebastian Kuzminsky, le Sat 06 Oct 2012 00:55:57 +0200, a écrit :
> > binding to CPU0
> > could not bind to CPU0: Resource deadlock avoided
>
> Mmm, from what I read in the freebsd kernel:
>
> /*
>  * Create a set in the space provided in 'set' with the provided
> parameters.
>  * The set is returned with a single ref.  May return EDEADLK if the set
>  * will have no valid cpu based on restrictions from the parent.
>  */
>
> _cpuset_create(struct cpuset *set, struct cpuset *parent, const cpuset_t
> *mask,
> cpusetid_t id)
> {
>
> if (!CPU_OVERLAP(>cs_mask, mask))
> return (EDEADLK);
>
> Could it be that due to administration rules lstopo is not allowed to
> bind on cpu 0-9 ? In that case the x86 backend can not detect anything
> there.
>


Hm.  It may be that we're doing something funny and reserving those CPUs.
 I'll run some tests on Monday and get back to you.


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-05 Thread Samuel Thibault
Sebastian Kuzminsky, le Sat 06 Oct 2012 00:55:57 +0200, a écrit :
> binding to CPU0
> could not bind to CPU0: Resource deadlock avoided

Mmm, from what I read in the freebsd kernel:

/*
 * Create a set in the space provided in 'set' with the provided parameters.
 * The set is returned with a single ref.  May return EDEADLK if the set
 * will have no valid cpu based on restrictions from the parent.
 */

_cpuset_create(struct cpuset *set, struct cpuset *parent, const cpuset_t *mask,
cpusetid_t id)
{

if (!CPU_OVERLAP(>cs_mask, mask))
return (EDEADLK);

Could it be that due to administration rules lstopo is not allowed to
bind on cpu 0-9 ? In that case the x86 backend can not detect anything
there.

Samuel


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-05 Thread Sebastian Kuzminsky
On Thu, Oct 4, 2012 at 6:39 PM, Samuel Thibault wrote:

> Sebastian Kuzminsky, le Wed 03 Oct 2012 17:24:55 +0200, a écrit :
> > So that's an improvement over the svn trunk
> > yesterday, but it's not all the way fixed yet!
>
> Ok.  Apparemently hwloc can't bind itself to procs 0-9 for some reason.
> I have added debug to the trunk, could you try it again (no need for the
> config.log any more, but I still need --enable-debug).
>
>
Here's the output from r4884, configured with --enabled debug.  Thanks for
looking at this.


output.r4884
Description: Binary data


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-04 Thread Samuel Thibault
Sebastian Kuzminsky, le Wed 03 Oct 2012 17:24:55 +0200, a écrit :
> So that's an improvement over the svn trunk
> yesterday, but it's not all the way fixed yet!

Ok.  Apparemently hwloc can't bind itself to procs 0-9 for some reason.
I have added debug to the trunk, could you try it again (no need for the
config.log any more, but I still need --enable-debug).

Samuel


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-03 Thread Brice Goglin
Le 03/10/2012 17:23, Sebastian Kuzminsky a écrit :
> On Tue, Oct 2, 2012 at 5:14 PM, Samuel Thibault
> > wrote:
>
> There were two bugs which resulted into cpuid not being properly
> compiled. I have fixed them in the trunk, could you try again?
>
>
> I updated my checkout to r4882, reconfigured, rebuilt, and reran it,
> and it made the same output as 1.5.  So that's an improvement over the
> svn trunk yesterday, but it's not all the way fixed yet!
>
> I'll be around all day to run tests if you like ;-)
>

For what it's worth, I tested the x86 code on Linux on a dual E5-2650
machine and got the correct topology (exactly like your Linux on your
server). So the x86 detection code may be ok, but something else wouldn't.
There's still at least one bug in the freebsd code according to our
internal regression tool, stay tuned.

Brice



Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-03 Thread Sebastian Kuzminsky
On Tue, Oct 2, 2012 at 5:14 PM, Samuel Thibault wrote:

> There were two bugs which resulted into cpuid not being properly
> compiled. I have fixed them in the trunk, could you try again?
>

I updated my checkout to r4882, reconfigured, rebuilt, and reran it, and it
made the same output as 1.5.  So that's an improvement over the svn trunk
yesterday, but it's not all the way fixed yet!

I'll be around all day to run tests if you like ;-)


r4882.tar.bz2
Description: BZip2 compressed data


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-02 Thread Sebastian Kuzminsky
Wow, that was fast!

I'm heading out for the day, but i'll try it first thing in the morning.
Thanks :-)
On Oct 2, 2012 5:14 PM, "Samuel Thibault"  wrote:

> Hello,
>
> Sebastian Kuzminsky, le Wed 03 Oct 2012 01:08:46 +0200, a écrit :
> > Here you go (the list server rejected it because it was too big, but this
> > compressed version should make it through).
>
> Thanks!
>
> There were two bugs which resulted into cpuid not being properly
> compiled. I have fixed them in the trunk, could you try again?
>
> Samuel
> ___
> hwloc-users mailing list
> hwloc-us...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/hwloc-users
>


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-02 Thread Samuel Thibault
Hello,

Sebastian Kuzminsky, le Wed 03 Oct 2012 01:08:46 +0200, a écrit :
> Here you go (the list server rejected it because it was too big, but this
> compressed version should make it through).

Thanks!

There were two bugs which resulted into cpuid not being properly
compiled. I have fixed them in the trunk, could you try again?

Samuel


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-02 Thread Samuel Thibault
Sebastian Kuzminsky, le Wed 03 Oct 2012 00:35:48 +0200, a écrit :
> [root@host-183 /home/nfshome/seb/tmp/hwloc-svn]# ./bin/lstopo
> 
> 
>  * CPU cpusets *
> 
> cpu 0 (os 0) has cpuset 0x0001
...
> cpu 31 (os 31) has cpuset 0x8000
> Machine#0(local=0KB total=33493800KB Backend=x86 Backend=FreeBSD 
> OSName=FreeBSD OSRelease=7.3-LROS-sawtooth.master.236 OSVersion="FreeBSD) 
> cpuset 0xf...f complete 0x online 0xf...f allowed 0xf...f nodeset 0x0 
> completeN 0x0 allowedN 0xf...f

Hum, so there's no actual trace of the x86 detection. Could you post
your config.log?

Samuel


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-02 Thread Sebastian Kuzminsky
On Tue, Oct 2, 2012 at 4:22 PM, Samuel Thibault wrote:

> Hello,
>
> Sebastian Kuzminsky, le Tue 02 Oct 2012 23:47:05 +0200, a écrit :
> > I've attached the output from both platforms.
>
> On freebsd, could you pass --enable-debug to ./configure and rerun
> lstopo, to get more debugging information?
>
>
Here's my configure line (and my output is attached):

 ./configure --disable-libxml2 --disable-cairo --disable-pci
--prefix=/home/seb/tmp/hwloc-svn --enable-debug

Thanks for looking at my problem :-)


debug-lstopo
Description: Binary data


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-02 Thread Sebastian Kuzminsky
On Tue, Oct 2, 2012 at 3:53 PM, Brice Goglin  wrote:

> Le 02/10/2012 23:45, Sebastian Kuzminsky a écrit :
> > Any ideas what I'm doing wrong?
>
> You're probably not doing anything wrong. The Linux output is indeed OK.
> The FreeBSD output is generated by reading cpuid information directly
> from the processor, we may need to update this code for recent
> processors like yours.
>
>
Thanks Brice!  I checked out r4874 of the svn trunk and built it and ran
it, and i think the output got less informative:

[testlab@host-183 /home/nfshome/seb/tmp/hwloc-svn]$ ./bin/lstopo
Machine (32GB)
  PU L#0 (P#0)
  PU L#1 (P#1)
  PU L#2 (P#2)
  PU L#3 (P#3)
  PU L#4 (P#4)
  PU L#5 (P#5)
  PU L#6 (P#6)
  PU L#7 (P#7)
  PU L#8 (P#8)
  PU L#9 (P#9)
  PU L#10 (P#10)
  PU L#11 (P#11)
  PU L#12 (P#12)
  PU L#13 (P#13)
  PU L#14 (P#14)
  PU L#15 (P#15)
  PU L#16 (P#16)
  PU L#17 (P#17)
  PU L#18 (P#18)
  PU L#19 (P#19)
  PU L#20 (P#20)
  PU L#21 (P#21)
  PU L#22 (P#22)
  PU L#23 (P#23)
  PU L#24 (P#24)
  PU L#25 (P#25)
  PU L#26 (P#26)
  PU L#27 (P#27)
  PU L#28 (P#28)
  PU L#29 (P#29)
  PU L#30 (P#30)
  PU L#31 (P#31)
[testlab@host-183 /home/nfshome/seb/tmp/hwloc-svn]$ ./bin/lstopo --version
lstopo 1.6a1


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-02 Thread Samuel Thibault
Hello,

Sebastian Kuzminsky, le Tue 02 Oct 2012 23:47:05 +0200, a écrit :
> I've attached the output from both platforms.

On freebsd, could you pass --enable-debug to ./configure and rerun
lstopo, to get more debugging information?

Samuel


Re: [hwloc-users] hwloc 1.5, freebsd and linux output on the same hardware

2012-10-02 Thread Brice Goglin
Le 02/10/2012 23:45, Sebastian Kuzminsky a écrit :
> Hi folks, I just discovered hwloc and it's really cool.  Very useful,
> so thanks!
>
> I'm trying to understand the hardware layout of a computer I'm working
> with, an HP Proliant DL360p G8 server with two Intel E5-2690 processors.
>
> I'm getting puzzling results from lstopo (from hwloc 1.5).  The
> results I get in Linux make good sense, but the results I get in
> FreeBSD (running on the same hardware) seem wrong.  Specifically,
> notice how the first 10 CPUs are not assigned to any socket or L1/L2
> cache.  Strange!
>
> I've attached the output from both platforms.
>
> Any ideas what I'm doing wrong?

Hello,

You're probably not doing anything wrong. The Linux output is indeed OK.
The FreeBSD output is generated by reading cpuid information directly
from the processor, we may need to update this code for recent
processors like yours.

Brice