w******g 发帖数: 67 | 1 I am working on transfer a MPI program from my laptop to one computing
cluster("Docker"). The system specifications are as following:
my laptop
Dell d810 Intel CPU
GCC 4.1.2
"Docker" cluster:
10 nodes , each node runs with 2 AMD64 processors and CentOS4.4
Linux system.
GCC 3.4.6
The problem is that I can compile the program successfully on the cluster.
But when running, it crashes with error(it crashes when I run it on one
processor)
*** glibc detected *** malloc(): memory corruption: 0x00000000 | l*****d 发帖数: 754 | 2 Could you provide more details, for example, the gcc compling options you us
ed? If you can give a simplified version of your source code that will be v
ery helpful.
【在 w******g 的大作中提到】 : I am working on transfer a MPI program from my laptop to one computing : cluster("Docker"). The system specifications are as following: : my laptop : Dell d810 Intel CPU : GCC 4.1.2 : "Docker" cluster: : 10 nodes , each node runs with 2 AMD64 processors and CentOS4.4 : Linux system. : GCC 3.4.6 : The problem is that I can compile the program successfully on the cluster.
| k****f 发帖数: 3794 | 3 跟AMD cluster没关系,
你自己程序出问题了。用valgrind查内存问题
【在 w******g 的大作中提到】 : I am working on transfer a MPI program from my laptop to one computing : cluster("Docker"). The system specifications are as following: : my laptop : Dell d810 Intel CPU : GCC 4.1.2 : "Docker" cluster: : 10 nodes , each node runs with 2 AMD64 processors and CentOS4.4 : Linux system. : GCC 3.4.6 : The problem is that I can compile the program successfully on the cluster.
|
|