|
From: dilip d. <dil...@ya...> - 2008-06-21 00:07:43
|
Hello All
I am using Valgrind tool to measure the performance and memory leak on my fedora 8 machine . I have the following queries wrt Valgrind tool
1)Using massif tool of Valgrind I am able to see a massif.out.pid.ps file containing the total bytes allocated on heap as a graph with bytes on y-axis and time on x-axis.But I have an application which is dependent on a few shared object libraries. I want to see more details wrt the heap allocated for each individual function in all the libraries and the application.How can I do it
2)I had used the callgrind tool to find the call complexity and later used KCachegrind to display the data in a readable form .I am able to get details like number of times the function was called but am unable to find out what was the complexity in the function. How do I find details wrt how much time a function took , how many machine instructions it took .
3)
For example I saw this
calls: no_of_calls called_function_line
calling_function_line complexity_of_called_function
I am able to see numbers for complexity but am unable to understand what is meant by complexity.if an application executes for the first time and data is obtained from RAM , now if data is cached then for second time the complexity should be less isn' t it
3)
Lastly how do I find out the cache that is being used in my application.
Can my application explicitly cache data which is frequently used(although th underlying OS might be doing this) .If so how do I measure it .I am unable to gather uch infotmation from cachegrind
Regards
Dilip
Unlimited freedom, unlimited storage. Get it now, on http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html/ |