Walking User mode pages

!MAILaRCHIVE_VOTE_RePLACE
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
To: <linux-kernel@...>
Date: Thursday, February 28, 2008 - 10:41 am

hi,

I am a newbie to Linux kernel, so please forgive me if the question is
stupid.

To understand the swapping system, I decided to write a module to scan 
every task_struct using for_each_process() checking whether or not the 
page pointed by the pte is swapped out. To get the linear address 
address of the pte, I did something similar to "pgd_offset << 
PGDIR_SHIFT + pud_offset << PUD_SHIFT + pmd_offset << PMDSHIFT + 
pte_offset << PAGE_SHIFT". The offset for each level is simply the 
iteration of the walking of the table ranging from 0 to PTRS_PER_(PGD, 
PUD, PMD, PTE) depending on the table. However, the result of the 
calculation doesn't match anything in the /proc/<pid>/maps.

Is this to be expected? If that's the case, how can I find out what's in 
the swap file. Thanks in advance.

YM
--
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
Walking User mode pages, YM Yen, (Thu Feb 28, 10:41 am)