Re: [mpich-discuss] Problem configuring mpich-3.4a3 and 3.4a2 --with-device=ch4:ucx
Hello Zhou, Using simply –with-pmix=/usr as you suggested seems to work the same but is shorter and more elegant. I will use this formulation. Thanks for the suggestion, Martin Audet From: Zhou, Hui [mailto:[email protected]] Sent: Wednesday, September 23, 2020 19:25 To: Audet, Martin <[email protected]>; [email protected] Subject: Re: [mpich-discuss] Problem configuring mpich-3.4a3 and 3.4a2 --with-device=ch4:ucx Hi Martin, Try `--with-pmix=/usr` -- Hui Zhou From: "Audet, Martin" <[email protected]<mailto:[email protected]>> Date: Wednesday, September 23, 2020 at 4:38 PM To: "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Cc: "Zhou, Hui" <[email protected]<mailto:[email protected]>> Subject: RE: [mpich-discuss] Problem configuring mpich-3.4a3 and 3.4a2 --with-device=ch4:ucx Thanks Zhou for your answer. Now when I add the following options: --with-pmix-include=/usr/include --with-pmix-lib=/usr/lib64 to my configuration options (just after --with-pmix) it works as expected (i.e. I am able to use the pmix mechanism in Slurm). However if I don’t provide --with-pmix, even if --with-pmix-include=/usr/include --with-pmix-lib=/usr/lib64 are provided, pmix is not used (and I have to use the pmi2 mechanism in Slurm). Regards, Martin Audet From: Zhou, Hui [mailto:[email protected]] Sent: Wednesday, September 23, 2020 9:47 To: [email protected]<mailto:[email protected]> Cc: Audet, Martin <[email protected]<mailto:[email protected]>> Subject: Re: [mpich-discuss] Problem configuring mpich-3.4a3 and 3.4a2 --with-device=ch4:ucx
./configure --with-device=ch4:ucx --with-hcoll=/opt/mellanox/hcoll --with-pmix --prefix=/home/publique/depot/mpi/mpich-ch4_ucx-3.4a3 --enable-fast=all --enable-romio --with-file-system=ufs+nfs --enable-shared --enable-sharedlibs=gcc
I believe the offending option is `--with-pmix`, since that defaults to `--with-pmix=yes`. You are supposed to pass in the path to pmix installation, e.g. `--with-pmix=/usr/local`. I admit this is a bit not obvious. -- Hui Zhou From: "Audet, Martin via discuss" <[email protected]<mailto:[email protected]>> Reply-To: "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Date: Tuesday, September 22, 2020 at 9:41 PM To: "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Cc: "Audet, Martin" <[email protected]<mailto:[email protected]>> Subject: [mpich-discuss] Problem configuring mpich-3.4a3 and 3.4a2 --with-device=ch4:ucx Hello MPICH_Developers, I am having trouble running the configure script on two latest mpich release (3.4a2 and 3.4a3) when I use the flag --with-device=ch4:ucx (no such problems with ch3:sock or ch3:nemesis). It seems that the generated Makefile often contains two strange compilation and link flags: -Iyes/include and –Lyes/lib. To be able to compile the library I had to use the following command to remove those strange flags: find . -name Makefile -exec sed -i -e 's/-Iyes\/include\>//g' -e 's/-Lyes\/lib\>//g' {} \; After this step I am able to compile and the resulting library seems ok and appears to performs well (not extensive testing). For your information I call the configure script like this: ./configure --with-device=ch4:ucx --with-hcoll=/opt/mellanox/hcoll --with-pmix --prefix=/home/publique/depot/mpi/mpich-ch4_ucx-3.4a3 --enable-fast=all --enable-romio --with-file-system=ufs+nfs --enable-shared --enable-sharedlibs=gcc I had experienced this problem on the following three configurations: mpich-3.4a2 CentOS 7.6 MOFED 4.7 mpich-3.4a3 CentOS 7.6 MOFED 4.7 mpich-3.4a3 CentOS 7.8 MOFED 4.9 The architecture is x86_64. Could someone take a look at this problem ? Thanks, Martin Audet
participants (1)
-
Audet, Martin