Hi all, In one of the postings a while ago, it was mentioned that kernel-space memory can also be paged in on demand (though it is not done for other reasons). The i386 do_page_fault() function has a comment to the same effect. But looking through the code, I am not able to figure out how this works. From what I understand, this is the flow when a page fault occurs for a kernel address: do_page_fault() gets the VMA containing the address by calling find_vma(). The mm struct passed is that of the current task. Since the kernel address is beyond the max. user-space address, find_vma() returns null. So, do_page_fault() jumps to the bad_area label, prints an error message and calls die(). If this is correct, the comment in do_page_fault() is erroneous. Or, more likely, I have missed something. Can someone clarify? -Thanks, Ravi. __________________________________________________ Do You Yahoo!? Find a job, post your resume. http://careers.yahoo.com -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ IRC Channel: irc.openprojects.net / #kernelnewbies Web Page: http://www.kernelnewbies.org/
