lude hwloc or not. This,
however, assumes that hwloc will either fail in configure or the library
will be successfully built and might return an error during
initialization. But even in this case "make" should not fail.
Personally, I'd prefer the first approach, but can do with t
might want to check
the return value $ac_cv_prog_cc_c99 and do something with it (maybe abort).
2. I believe AM_CONDITIONAL is automake-1.11 specific. Can someone
verify this? If this is correct, then your AM_INIT_AUTOMAKE should
contain 1.11 as a prereq.
> make is supposed to always succee
On 10/21/2009 10:28 AM, Samuel Thibault wrote:
> Pavan Balaji, le Wed 21 Oct 2009 09:55:36 -0500, a écrit :
>> 1. I see a AC_PROG_CC_C99 in the configure.ac. Do you require the
>> compiler to be C99 capable always?
>
> No, we ended up using constructs which should pass c90 a
On 10/21/2009 10:38 AM, Samuel Thibault wrote:
> Pavan Balaji, le Wed 21 Oct 2009 10:36:33 -0500, a écrit :
>> On 10/21/2009 10:28 AM, Samuel Thibault wrote:
>>> Pavan Balaji, le Wed 21 Oct 2009 09:55:36 -0500, a écrit :
>>>> 1. I see a AC_PROG_CC_C99 in the co
> Pavan -- is it a problem to always compile hwloc with gcc?
Yeah, this will be a problem. For us to enable hwloc by default, it'll
need to build with all compilers and on all platforms without errors
(and hopefully without warnings either).
-- Pavan
--
Pavan Balaji
http://www.mcs
oring return value of 'fgets', declared with attribute
warn_unused_result
/home/balaji/projects/mpich2/hydra/hydra/tools/bind/hwloc/hwloc/src/topology-linux.c:931:
warning: ignoring return value of 'fgets', declared with attribute
warn_unused_result
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 10/21/2009 10:38 AM, Samuel Thibault wrote:
> Pavan Balaji, le Wed 21 Oct 2009 10:36:33 -0500, a écrit :
>> On 10/21/2009 10:28 AM, Samuel Thibault wrote:
>>> Pavan Balaji, le Wed 21 Oct 2009 09:55:36 -0500, a écrit :
>>>> 1. I see a AC_PROG_CC_C99 in the co
ought these over to the v0.9 branch.
>
> With all these changes, I'll cut an rc2 and post it shortly.
>
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
/autoconf-archive/ax_prefix_config_h.html).
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 10/21/2009 04:44 PM, Samuel Thibault wrote:
> Pavan Balaji, le Wed 21 Oct 2009 16:40:42 -0500, a écrit :
>>> Please read this thread:
>>>
>>> http://www.open-mpi.org/community/lists/hwloc-devel/2009/09/0054.php
>> Thanks. Is the only issue to not use AC_C_
lstopo-lstopo-fig.o lstopo-lstopo-cairo.o lstopo-lstopo-xml.o
-L/home/balaji/tmp/hwloc/hwloc-0.9.1rc2/src -lm ../src/.libs/libhwloc.so
-ltermcap
../src/.libs/libhwloc.so: undefined reference to `__CPU_SET'
../src/.libs/libhwloc.so: undefined reference to `__CPU_ZERO'
make[1]: *** [lstopo] Error 2
make[1]: Leaving directory
`/radix-homes/balaji/tmp/hwloc/hwloc-0.9.1rc2/utils'
make: *** [all-recursive] Error 1
==
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
m information:
% uname -a
Linux bblogin 2.6.24-24-generic #1 SMP Sat Aug 22 00:30:49 UTC 2009
x86_64 GNU/Linux
Attached /proc/cpuinfo as cpuinfo.log. Let me know if you need any more
information.
-- Pavan
On 10/22/2009 02:08 AM, Samuel Thibault wrote:
> Pavan Balaji, le Wed 21 Oct 2009 20:3
On 10/22/2009 10:10 AM, Jeff Squyres wrote:
> On Oct 21, 2009, at 5:01 PM, Pavan Balaji wrote:
>
>> The 0.9 branch doesn't build anymore -- missing openfabrics-verbs.h.
>
> Where are you looking? I see openfabrics-verbs.h in the rc2 tarball and
> on the svn v0.9 b
the return values and additional code this patch adds for
error cases; it is possible they might introduce a bug somewhere because
of my lack of knowledge on hwloc's internals.
Thanks,
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
Index: src/topology-li
all -Werror])
> +AM_INIT_AUTOMAKE([1.11 dist-bzip2 foreign tar-ustar silent-rules -Wall
> -Werror])
>
> # Make configure depend on the VERSION file, since it's used in AC_INIT
>
>
> ___
> hwloc-devel mailing list
> hwloc-
Log files attached.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
checking build system type... x86_64-unknown-linux-gnu
checking host system type... x86_64-unknown-linux-gnu
checking target system type... x86_64-unknown-linux-gnu
checking for a BSD-compatible install... /usr/bin
Attached.
On 10/29/2009 04:12 PM, Brice Goglin wrote:
> What's in
> /radix-homes/software/com/packages/pgi-9.0-4/linux86-64/9.0-4/include/sched.h
> ?
>
> Brice
>
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
/*
* Copyright 2009, STMicroelectronics, Incorpora
g, but the macros in this file look obsolete to me...
>
> Brice
>
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 10/29/2009 06:01 PM, Samuel Thibault wrote:
> In the meanwhile, I have commited detection code, could you check the
> svn version from
> https://svn.open-mpi.org/svn/hwloc/branches/v0.9
> ?
Yeah, this seems to build fine. Thanks.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
1
-- Pavan
On 11/04/2009 03:55 PM, Jeff Squyres wrote:
> Please beat it up and look for problems!
>
> http://www.open-mpi.org/software/hwloc/v0.9/
>
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
Apologies. This is my bad! My setup was messed up.
On 11/05/2009 01:18 PM, Pavan Balaji wrote:
> Am I supposed to do something special here for hwloc-0.9.2?
>
> % autoconf --version
> autoconf (GNU Autoconf) 2.63
> Copyright (C) 2008 Free Software Foundation, Inc.
> Licen
On 11/05/2009 01:21 PM, Jeff Squyres wrote:
> FWIW, you shouldn't need to run autogen on the tarball (or are you
> importing from svn?).
Importing from the svn.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
Sorry for going back-and-forth. I thought it was my error, but maybe it
isn't. A fresh svn checkout from
http://svn.open-mpi.org/svn/hwloc/tags/hwloc-v0.9.2 gives the error I
mentioned.
Am I missing something?
-- Pavan
On 11/05/2009 01:22 PM, Pavan Balaji wrote:
> On 11/05/2009 01:21
-
>
> ___
> hwloc-devel mailing list
> hwloc-de...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/hwloc-devel
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
pport C99, but whether to
drop C89-only support is a policy decision we are yet to make in MPICH2,
and will need the above patch to get hwloc integrated cleanly in all our
supported environments.
Thanks,
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
Darn. The patch is incorrect. Sorry, too sleepy. It should be checking
for $ac_prog_cc_c99 instead. But you get the idea.
-- Pavan
On 05/10/2010 08:56 PM, Pavan Balaji wrote:
I understand that hwloc requires C99 support. However, for compilers
that don't support C99, would you be wi
the delays folks; I was at the MPI Forum last week and that heavily
distracted me...)
On May 10, 2010, at 9:59 PM, Pavan Balaji wrote:
Darn. The patch is incorrect. Sorry, too sleepy. It should be checking
for $ac_prog_cc_c99 instead. But you get the idea.
-- Pavan
On 05/10/2010 08:56 PM
windows support on hwloc was.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
ever ask PGI why they do this?
Brice
Message original
Sujet: Re: hwloc-devel post from bal...@mcs.anl.gov requires approval
Date : Sun, 01 Aug 2010 12:17:08 -0500
De :Pavan Balaji
Pour : Brice Goglin
D'oh! I forgot to forward stderr to the files. I've reatt
1 #0 (32KB) + Core #0 + PU #0 (phys=0)
L1 #1 (32KB) + Core #1 + PU #1 (phys=1)
Thanks,
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
helping in this case. So from
our perspective, if hwloc were to fail, we always want it to fail at
configure time, not make (or make install) time.
Thanks,
-- Pavan
On 08/03/2010 05:16 PM, Pavan Balaji wrote:
For the time being, can you add a test to the configure to error out in
this c
nterim, I'm
curious to know what other projects using hwloc are doing. Wouldn't they
run into this exact issue as well?
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
adds an odd logic, but it might just work around pgcc's
crazyness.
Samuel
___
hwloc-devel mailing list
hwloc-de...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/hwloc-devel
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
wloc/hwloc
Thanks,
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
y reserved keywords instead?
"depth" was first defined in the parent function (main()), and redefined
within a smaller scope.
"index" was defined in string.h.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
I'll discard this patch.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
ell. Using a sandbox configure allows us to catch hwloc's configure
errors and disable it for cases where it doesn't work.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
debug builds by default when
you find .svn), but I could probably be convinced to reverse it
pretty easily. So -- I'm ok with it.
Thanks. Let me know which way you guys decide on this, so I can keep my
patch-list current.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
Thanks for pointing it out.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
ted earlier.
https://svn.open-mpi.org/trac/hwloc/browser/branches/v1.1/config/hwloc_internal.m4?rev=3120#L162
Is this something that can be fixed in hwloc, or do you guys prefer
having it?
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 01/20/2011 03:33 PM, Brice Goglin wrote:
Le 20/01/2011 22:16, Pavan Balaji a écrit :
I'd rather keep it. Maybe Jeff can add a m4 macro that your m4 would
call to disable it? Something that would force the default value of
hwloc_want_picky to 0.
Ok, that's fine.
What's sur
ll tonight. It's generated
around 9pm US Eastern time.
http://www.open-mpi.org/software/hwloc/nightly/trunk/
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
accurate, IMO.
-- Pavan
On 11/06/2009 04:22 AM, Pavan Balaji wrote:
Thanks. That worked.
-- Pavan
On 11/05/2009 01:56 PM, Jeff Squyres wrote:
Ok, I see the issue. Apply the attached patch and see if that works for
you.
I'd prefer not to do another release today, if possible. ;-)
en
with the current version. It's just an additional -I path that's missing
in Makefile.am.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
mpi.org/mailman/listinfo.cgi/hwloc-devel
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
Hi all,
The hwloc code seems to shadow some previously defined variables. These
warnings show up when configured with CFLAGS=-Wshadow with gcc. Here's
a patch to fix these warnings:
https://trac.mcs.anl.gov/projects/mpich2/changeset/9656
-- Pavan
--
Pavan Balaji
http://www.mcs.an
On 03/31/2012 01:48 AM, Brice Goglin wrote:
Le 31/03/2012 04:06, Pavan Balaji a écrit :
The hwloc code seems to shadow some previously defined variables.
These warnings show up when configured with CFLAGS=-Wshadow with gcc.
Here's a patch to fix these warnings:
https://trac.mcs.an
o squash a
compiler warning, since state.tagname is a char *, while "topology" is a
const char *.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 05/11/2012 02:10 PM, Jeff Squyres wrote:
On May 11, 2012, at 2:35 PM, Pavan Balaji wrote:
1. The patch to Makefile.am is an ugly hack to get "make dist"
working in hydra. Specifically, autogen.sh was having trouble
running in the tarball created by "make dist", since
Folks,
Here's a minor compiler warning nit-pick in 1.5rc1:
https://trac.mcs.anl.gov/projects/mpich2/changeset/10012
Thanks,
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
ively just
typecast that one line, but this seems to be cleaner as you are not
really editing that string anywhere.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
In hwloc_bitmap_or(), is the resultant bitmap allowed to be the same as
one of the input bitmaps? It seems to work correctly in practice, but
the API doesn't seem to explicitly guarantee it.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
issue with automake-1.12 that Dave
noticed.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
at -Wall
would.
4. The change to configure.ac is an issue with automake-1.12 that Dave
noticed.
These lines have been added in v1.5 already.
Thanks.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
Brice,
Sorry, I didn't get to try this patch out. Did it end up in any of the
hwloc releases?
-- Pavan
On 07/24/2012 02:45 PM, Brice Goglin wrote:
Le 24/07/2012 13:39, Brice Goglin a écrit :
Le 23/07/2012 18:40, Pavan Balaji a écrit :
3. The changes to config/hwloc.m4 and in
On 09/06/2012 12:22 AM, Brice Goglin wrote:
No, we didn't apply anything to v1.5. In trunk, we're using sysconf
instead of getpagesize when it supports SC_PAGESIZE or SC_PAGE_SIZE
(very often).
Is this going to go into 1.5.1?
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
er and Data Sciences Department Tel: +1-510-495-2352
> Lawrence Berkeley National Laboratory Fax: +1-510-486-6900
>
>
> ___
> hwloc-devel mailing list
> hwloc-de...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/hwloc-devel
>
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 02/05/2013 04:52 PM US Central Time, Pavan Balaji wrote:
> If libpci was disabled by default, would hwloc still come under the same
> GPL issue?
I realized that wasn't very clear. Let me rephrase -- if libpci was
disabled (either by default or a configure argument), would hwl
On 02/05/2013 05:29 PM US Central Time, Samuel Thibault wrote:
> Pavan Balaji, le Tue 05 Feb 2013 23:53:54 +0100, a écrit :
>> I checked libnuma, which seems to be LGPL (phew!), but didn't look at
>> the remaining libraries hwloc uses.
>
> Right. If hwloc was strictly
in the SETUP_CORE macro.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
tend
> to think you should remove _POSIX_SOURCE on this FreeBSD release.
Is this for any version of POSIX_SOURCE? Just giving _POSIX_SOURCE
falls back to POSIX.1-1990. Most times we (mpich) only care about
POSIX.1-2001 (given by -D_POSIX_C_SOURCE=200112L).
I can try it out later today.
-
's OK to not build hwloc on FreeBSD, but it should abort cleanly at
configure time. With the current model, the mpich build breaks on
FreeBSD because of this, even though the non-hwloc part of mpich builds
fine on FreeBSD.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 04/30/2013 02:22 AM US Central Time, Brice Goglin wrote:
> Le 29/04/2013 15:51, Pavan Balaji a écrit :
>> On 04/29/2013 08:36 AM US Central Time, Brice Goglin wrote:
>>>> I wonder if this CFLAGS=-D_POSIX_SOURCE is supported by FreeBSD headers.
>>> I could not fix
One more issue with hwloc-1.7 on the mac.
http://git.mpich.org/mpich.git/commitdiff/d9a67f40
This showed up when we did a strict build of mpich. I believe this can
be reproduced with "-Wall -Werror -O2", but I can find the exact set of
minimum required flags, if needed.
-- Pavan
e' to
non-matching type 'unsigned int'
lstopo-draw.c:444: warning: cast from function call of type 'double' to
non-matching type 'unsigned int'
lstopo-draw.c:444: warning: cast from function call of type 'double' to
non-matching type 'unsigned int'
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
grading to hwloc-1.7 for the next
major release of mpich (3.1). With the above fixes, it looks like
things are working well. Of course, we'll be going through a lot more
testing before the final release which would be later this year.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 05/05/2013 12:05 PM US Central Time, Brice Goglin wrote:
> Le 05/05/2013 18:18, Pavan Balaji a écrit :
>> - hwloc's check for whether an explicitly function declaration is
>> needed (using _HWLOC_CHECK_DECL) was relying on whether a dummy call to
>> the function t
don't really care about the hwloc library ABI.
The second is a minor enhancement to autogen.sh, so we can make
autogen.sh less verbose based on user settings.
Neither patch changes the default behavior of hwloc.
Thanks,
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 05/05/2013 11:18 AM US Central Time, Pavan Balaji wrote:
> - For sysctl and sysctlbyname, I've updated hwloc/config.m4 to use a
> full link test instead of just using AC_CHECK_FUNCS, which only checks
> to see if the symbol exists or not. For example, the prototype of
>
vate/autogen/config.h stuff.
However, this also means that the configure.ac part of the versioning
patch has no effect for us. That part will need to sit in one of the
embeddable macros. Let me investigate it some more and get back to you.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
ode contribution instead of forwarding it along.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
(oh yeah, I hate that platform!) with
c89 builds. Can you pull in this patch:
http://git.mpich.org/mpich.git/commitdiff/939601e0
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 05/06/2013 04:08 AM US Central Time, Brice Goglin wrote:
> Le 05/05/2013 19:26, Pavan Balaji a écrit :
>> -autoreconf -ivf
>> +autoreconf ${autoreconf_args:"-ivf"}
>> -- 1.8.1.2
>
> This needs an additional dash:
>
> -autoreconf -ivf
> +autore
On 05/06/2013 06:07 AM US Central Time, Pavan Balaji wrote:
>
> On 05/06/2013 04:08 AM US Central Time, Brice Goglin wrote:
>> Le 05/05/2013 19:26, Pavan Balaji a écrit :
>>> -autoreconf -ivf
>>> +autoreconf ${autoreconf_args:"-ivf"}
>>&g
ese are all stdint.h types which is c99. It
looks like c89 support is not consistent in the code.
Can you check inside HWLOC_SETUP_CORE whether you have the necessary
level of c99 support and fail if you don't?
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
th
equivalent as far as mpich is concerned.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On 05/06/2013 07:37 AM US Central Time, Brice Goglin wrote:
> Le 06/05/2013 13:24, Pavan Balaji a écrit :
>> On 05/06/2013 01:25 AM US Central Time, Brice Goglin wrote:
>>> I think we removed C99 use (and configure check) from the code in 1.2
>>> because some compilers
nce regression testing is done.
> (5) should likely be OK too.
> (3) and more importantly (4) touch ugly configury, I am not sure I want
> to backport these. Or maybe later, in case we get more testing before
> v1.7.1 occurs.
>
> Brice
>
>
>
>
> Le 05/05/2013 18:18,
>
>
> Le 07/05/2013 16:59, Pavan Balaji a écrit :
>> Thanks Brice. What's the timeframe for 1.7.1? I want to see if we can
>> move to that for our mpich-3.1 release.
>>
>> -- Pavan
>>
>> On 05/07/2013 09:53 AM US Central Time, Brice Goglin wrote
YSCTL). The former provides the declaration for the function and
the latter provides the macros used within the definition of the
function (e.g., CTL_HW).
I hope this is a small enough patch to make it into 1.7.1 GA.
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
still supported, you guys might want to get that fixed
in there. Or at least put up a warning on the website about this.
Regards,
-- Pavan
--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
On Sep 6, 2013, at 10:30 AM, Pavan Balaji wrote:
>
>> The mpich-3.0.x series was released with hwloc-1.6.x. One our of partners
>> just brought it to our attention that this version of hwloc has LGPL-2.1
>> code in src/libltdl.
>
> Because GPL violations are quite
Hello,
Would you consider the following patch for hwloc-1.8 that we embed in the mpich
version of hwloc? The commit log has the description. Please ignore the extra
white-space piece of the commit.
8<
% git show 59cbc846 tests/embedded/configure.ac
commit 59cbc846d56eb259f6014b23114ee
84 matches
Mail list logo
