A GitHub issue seems appropriate here (
https://github.com/pmodels/mpich/issues)
Jeff
On Tue, Aug 25, 2020 at 2:09 PM William Gropp via discuss <discuss(a)mpich.org>
wrote:
>
>
> (If there is a better way to submit these, let me know. The “bug report”
> link doesn’t work from mpich.org)
>
> —enable-fast in the top level configure accepts multiple, comma-separated
> items. The configure in modules/yaksa/configure.ac expects (without
> checking) an O level only. This causes the configure step to fail with an
> unrelated error (yaksa thinks that it can’t fine pthreads).
>
> Bill
>
>
>
> William Gropp
> Director and Chief Scientist, NCSA
> Thomas M. Siebel Chair in Computer Science
> University of Illinois Urbana-Champaign
>
>
>
>
>
>
>
>
>
>
> _______________________________________________
>
> discuss mailing list discuss(a)mpich.org
>
> To manage subscription options or unsubscribe:
>
> https://lists.mpich.org/mailman/listinfo/discuss
>
> --
Jeff Hammond
jeff.science(a)gmail.com
http://jeffhammond.github.io/
(If there is a better way to submit these, let me know. The “bug report” link doesn’t work from mpich.org)
—enable-fast in the top level configure accepts multiple, comma-separated items. The configure in modules/yaksa/configure.ac expects (without checking) an O level only. This causes the configure step to fail with an unrelated error (yaksa thinks that it can’t fine pthreads).
Bill
William Gropp
Director and Chief Scientist, NCSA
Thomas M. Siebel Chair in Computer Science
University of Illinois Urbana-Champaign
Thanks - I’m up to the runtime failure on FI_ADDR_NOTAVAIL. Is there an alternative path to build MPICH on OSX?
Bill
William Gropp
Director and Chief Scientist, NCSA
Thomas M. Siebel Chair in Computer Science
University of Illinois Urbana-Champaign
> On Aug 7, 2020, at 11:54 AM, Zhou, Hui via discuss <discuss(a)mpich.org> wrote:
>
> Follow up --
>
> If you hit this runtime error:
>
> Assertion failed in file src/mpid/ch4/netmod/ofi/ofi_init.c at line xxxx: mapped_table[i] != FI_ADDR_NOTAVAIL
>
> That is due to libfabric having trouble with ipv6 addresses and it ignores our hints on addr_format. I'll work out a patch later.
>
> --
> Hui Zhou
>
>
>
> On 8/7/20, 10:24 AM, "Raffenetti, Kenneth J." <raffenet(a)mcs.anl.gov> wrote:
>
> Just updating from git is not enough to fix it, though. Since it is a submodule update, you'll need to run
>
> git submodule update --init
>
> to pull in the latest fix.
>
> Ken
>
> On 8/6/20, 10:14 PM, "Zhou, Hui via discuss" <discuss(a)mpich.org> wrote:
>
> The particular link error should have been fixed since this Monday. Try make sure it is the correct source tree and clean build tree? That said, there will be additional run time issues building ch4 on mac. I’ll spend some time tomorrow and follow up if I can make progress.
>
> --
> Hui Zhou
> Principal Software Development Specialist
> Mathematics and Computer Science
> Argonne National Laboratory
>
>
>
> From: "Gropp, William D via discuss" <discuss(a)mpich.org>
> Reply-To: "discuss(a)mpich.org" <discuss(a)mpich.org>
> Date: Thursday, August 6, 2020 at 5:01 PM
> To: "discuss(a)mpich.org" <discuss(a)mpich.org>
> Cc: "Gropp, William D" <wgropp(a)illinois.edu>
> Subject: [mpich-discuss] Unable to build on OSX
>
>
>
> Using the current as of this morning from git, hash 1d340282 , and with CC=gcc-10 (needed to work with the corresponding version of gfortran, because gcc on OSX is really clang)
>
>
> ../../sources/mpich/configure --with-device=ch4:ofi --prefix=/Users/gropp/installs/mpich2-dbg
>
>
>
> works but make ends with
>
>
>
> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib: file: lib/.libs/libmpifort.a(mpi_c_interface_nobuf.o) has no symbols
>
> CXX src/binding/cxx/initcxx.lo
>
> CXXLD lib/libmpicxx.la <http://libmpicxx.la>
>
> CC src/env/mpichversion.o
>
> CCLD src/env/mpichversion
>
> Undefined symbols for architecture x86_64:
>
> "_ofi_process_vm_writev", referenced from:
>
> import-atom in libpmpi.dylib
>
> ld: symbol(s) not found for architecture x86_64
>
> collect2: error: ld returned 1 exit status
>
> make[2]: *** [src/env/mpichversion] Error 1
>
> make[1]: *** [all-recursive] Error 1
>
> make: *** [all] Error 2
>
> gropp@wdg-macbook hydra %
>
>
>
>
> Configure and make logs attached.
>
>
>
> (Some of the instructions on the MPICH web page for submitting bug reports sent me to unreachable github web pages)
>
>
>
>
>
>
>
> William Gropp
> Director and Chief Scientist, NCSA
> Thomas M. Siebel Chair in Computer Science
> University of Illinois Urbana-Champaign
>
>
>
>
>
>
>
>
>
>
>
>
>
> _______________________________________________
> discuss mailing list discuss(a)mpich.org
> To manage subscription options or unsubscribe:
> https://lists.mpich.org/mailman/listinfo/discuss
Follow up --
If you hit this runtime error:
Assertion failed in file src/mpid/ch4/netmod/ofi/ofi_init.c at line xxxx: mapped_table[i] != FI_ADDR_NOTAVAIL
That is due to libfabric having trouble with ipv6 addresses and it ignores our hints on addr_format. I'll work out a patch later.
--
Hui Zhou
On 8/7/20, 10:24 AM, "Raffenetti, Kenneth J." <raffenet(a)mcs.anl.gov> wrote:
Just updating from git is not enough to fix it, though. Since it is a submodule update, you'll need to run
git submodule update --init
to pull in the latest fix.
Ken
On 8/6/20, 10:14 PM, "Zhou, Hui via discuss" <discuss(a)mpich.org> wrote:
The particular link error should have been fixed since this Monday. Try make sure it is the correct source tree and clean build tree? That said, there will be additional run time issues building ch4 on mac. I’ll spend some time tomorrow and follow up if I can make progress.
--
Hui Zhou
Principal Software Development Specialist
Mathematics and Computer Science
Argonne National Laboratory
From: "Gropp, William D via discuss" <discuss(a)mpich.org>
Reply-To: "discuss(a)mpich.org" <discuss(a)mpich.org>
Date: Thursday, August 6, 2020 at 5:01 PM
To: "discuss(a)mpich.org" <discuss(a)mpich.org>
Cc: "Gropp, William D" <wgropp(a)illinois.edu>
Subject: [mpich-discuss] Unable to build on OSX
Using the current as of this morning from git, hash 1d340282 , and with CC=gcc-10 (needed to work with the corresponding version of gfortran, because gcc on OSX is really clang)
../../sources/mpich/configure --with-device=ch4:ofi --prefix=/Users/gropp/installs/mpich2-dbg
works but make ends with
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib: file: lib/.libs/libmpifort.a(mpi_c_interface_nobuf.o) has no symbols
CXX src/binding/cxx/initcxx.lo
CXXLD lib/libmpicxx.la <http://libmpicxx.la>
CC src/env/mpichversion.o
CCLD src/env/mpichversion
Undefined symbols for architecture x86_64:
"_ofi_process_vm_writev", referenced from:
import-atom in libpmpi.dylib
ld: symbol(s) not found for architecture x86_64
collect2: error: ld returned 1 exit status
make[2]: *** [src/env/mpichversion] Error 1
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2
gropp@wdg-macbook hydra %
Configure and make logs attached.
(Some of the instructions on the MPICH web page for submitting bug reports sent me to unreachable github web pages)
William Gropp
Director and Chief Scientist, NCSA
Thomas M. Siebel Chair in Computer Science
University of Illinois Urbana-Champaign
Just updating from git is not enough to fix it, though. Since it is a submodule update, you'll need to run
git submodule update --init
to pull in the latest fix.
Ken
On 8/6/20, 10:14 PM, "Zhou, Hui via discuss" <discuss(a)mpich.org> wrote:
The particular link error should have been fixed since this Monday. Try make sure it is the correct source tree and clean build tree? That said, there will be additional run time issues building ch4 on mac. I’ll spend some time tomorrow and follow up if I can make progress.
--
Hui Zhou
Principal Software Development Specialist
Mathematics and Computer Science
Argonne National Laboratory
From: "Gropp, William D via discuss" <discuss(a)mpich.org>
Reply-To: "discuss(a)mpich.org" <discuss(a)mpich.org>
Date: Thursday, August 6, 2020 at 5:01 PM
To: "discuss(a)mpich.org" <discuss(a)mpich.org>
Cc: "Gropp, William D" <wgropp(a)illinois.edu>
Subject: [mpich-discuss] Unable to build on OSX
Using the current as of this morning from git, hash 1d340282 , and with CC=gcc-10 (needed to work with the corresponding version of gfortran, because gcc on OSX is really clang)
../../sources/mpich/configure --with-device=ch4:ofi --prefix=/Users/gropp/installs/mpich2-dbg
works but make ends with
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib: file: lib/.libs/libmpifort.a(mpi_c_interface_nobuf.o) has no symbols
CXX src/binding/cxx/initcxx.lo
CXXLD lib/libmpicxx.la <http://libmpicxx.la>
CC src/env/mpichversion.o
CCLD src/env/mpichversion
Undefined symbols for architecture x86_64:
"_ofi_process_vm_writev", referenced from:
import-atom in libpmpi.dylib
ld: symbol(s) not found for architecture x86_64
collect2: error: ld returned 1 exit status
make[2]: *** [src/env/mpichversion] Error 1
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2
gropp@wdg-macbook hydra %
Configure and make logs attached.
(Some of the instructions on the MPICH web page for submitting bug reports sent me to unreachable github web pages)
William Gropp
Director and Chief Scientist, NCSA
Thomas M. Siebel Chair in Computer Science
University of Illinois Urbana-Champaign
The particular link error should have been fixed since this Monday. Try make sure it is the correct source tree and clean build tree? That said, there will be additional run time issues building ch4 on mac. I’ll spend some time tomorrow and follow up if I can make progress.
--
Hui Zhou
Principal Software Development Specialist
Mathematics and Computer Science
Argonne National Laboratory
From: "Gropp, William D via discuss" <discuss(a)mpich.org>
Reply-To: "discuss(a)mpich.org" <discuss(a)mpich.org>
Date: Thursday, August 6, 2020 at 5:01 PM
To: "discuss(a)mpich.org" <discuss(a)mpich.org>
Cc: "Gropp, William D" <wgropp(a)illinois.edu>
Subject: [mpich-discuss] Unable to build on OSX
Using the current as of this morning from git, hash 1d340282 , and with CC=gcc-10 (needed to work with the corresponding version of gfortran, because gcc on OSX is really clang)
../../sources/mpich/configure --with-device=ch4:ofi --prefix=/Users/gropp/installs/mpich2-dbg
works but make ends with
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib: file: lib/.libs/libmpifort.a(mpi_c_interface_nobuf.o) has no symbols
CXX src/binding/cxx/initcxx.lo
CXXLD lib/libmpicxx.la<http://libmpicxx.la>
CC src/env/mpichversion.o
CCLD src/env/mpichversion
Undefined symbols for architecture x86_64:
"_ofi_process_vm_writev", referenced from:
import-atom in libpmpi.dylib
ld: symbol(s) not found for architecture x86_64
collect2: error: ld returned 1 exit status
make[2]: *** [src/env/mpichversion] Error 1
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2
gropp@wdg-macbook hydra %
Configure and make logs attached.
(Some of the instructions on the MPICH web page for submitting bug reports sent me to unreachable github web pages)
William Gropp
Director and Chief Scientist, NCSA
Thomas M. Siebel Chair in Computer Science
University of Illinois Urbana-Champaign
Using the current as of this morning from git, hash 1d340282 , and with CC=gcc-10 (needed to work with the corresponding version of gfortran, because gcc on OSX is really clang)
../../sources/mpich/configure --with-device=ch4:ofi --prefix=/Users/gropp/installs/mpich2-dbg
works but make ends with
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib: file: lib/.libs/libmpifort.a(mpi_c_interface_nobuf.o) has no symbols
CXX src/binding/cxx/initcxx.lo
CXXLD lib/libmpicxx.la<http://libmpicxx.la>
CC src/env/mpichversion.o
CCLD src/env/mpichversion
Undefined symbols for architecture x86_64:
"_ofi_process_vm_writev", referenced from:
import-atom in libpmpi.dylib
ld: symbol(s) not found for architecture x86_64
collect2: error: ld returned 1 exit status
make[2]: *** [src/env/mpichversion] Error 1
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2
gropp@wdg-macbook hydra %
Configure and make logs attached.
(Some of the instructions on the MPICH web page for submitting bug reports sent me to unreachable github web pages)
William Gropp
Director and Chief Scientist, NCSA
Thomas M. Siebel Chair in Computer Science
University of Illinois Urbana-Champaign
