Specifically, I am working on implementing a PMPI library (profiling interface) for a performance analysis tool that our lab is developing. Basically, I would like to report each instance of let's say mpi_send() separately if they originate from different part of the program. Since PMPI is done...
Actually, this is a 2 part question. Someone told me this has something to do with stackwalking but I am not entirely sure. Any help would be appreciated.
1. How do I get the line number for the operation that is currently executing.
Ex: presume getLine() returns the line number in the code...
Looks like the #1 concern for everyone is time related (i.e. involve timing of something and trying to get the fastest time). Anyone ever try analyzing the code that doesn't involve time?
Hung-Hsun
Hi,
Our research group is developing a performance analysis tool for parallel programming and I would like to see if I can get some feedback from you guys.
Basically, I would like to know the following when you, the programmer, goes through to optimize your code.
1. What is your approach in...
>> Also, casting the result of malloc in C is a bad thing.
How is that a bad thing? From what I can see, he/she is assigning the memory space using malloc to a char pointer. perfectly fine there.
Regardless, he's probably right on having to memorize the size somewhere else.
Our group is developing a performance analysis tool for SHMEM and UPC and I would like to see if I can find some users to contact with. Please let me know if you are an user. Thanks!
Hung-Hsun
I am using gcc and in my tests, everything is a power of 2 so yes they are all word aligned for large size. I presume that the Intel compiler give you better performance, but how much difference and why so? Can we use the same technique and write our own stuff?
First of all, thanx for all the replies.
I guess more information is in order. I am dealing with a high speed interconnect that require the data to be copy from a source location to a buffer location before it can be sent over the network (Think of it as MPI where you copy the source data into...
I wish I could do away with copying. There is a requirement in the setup currently that force me to do data copying and I pretty much just need to find a way to make it more efficient.
Hi,
I am wondering if anyone can tell me how to recompile the kernel for SuSe 8. The kernel that I have right now is running non-smp mode and i would like to make it so it runs smp mode so I would like to know how to configure and compile the source to accomplish this. thanks!
Hung-Hsun
Can anyone tell me if there is a file in SuSe 8 that is similar to module-info file in Redhat? I am trying to compile something and the system tells me at the end that there is a kernel-module version mistatch. It tells me that the thing I compiled for is version 2.4.19 while the running kernel...
so each time i change something in test.c I need to include hello.c as well? (i.e. rerun using gcc hello.c test.c)
How is that we don't need to include the standard header files when we include it in our program then? For example, why wouldn't I need to recompile stdio.c when I include stdio.h?
I am attempting to create header files for my functions but it doesn't work for me. Can someone tell me what I am doing wrong? Below is what I put in each file
hello.h
#include <stdio.h>
void hello(); /* prototype of function */
hello.c
void hello()
{
printf("Hello World!\n");
}...
I wish to create a static global array of something that could change each time the program is run base on some input (i.e. command line input size. Is it possible to do this?
Hi,
I have the following code that passes a pointer of a structure so that the function can print the content of the structure, but for some reason I am not able to get the correct value. I printed out the address that's passing in to the function (which points to the structure) and it is...
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.