Dear xen-devel members, My name is Diego Perez-Botero and I''m a Computer Science Master''s Degree student from Princeton University. I''m trying to analyze VM exits in the Xen Hypervisor, but I haven''t been able to find evidence of any VM exits up until now. I followed the "Fedora13Xen4Tutorial" tutorial from the Xen Wiki (http://wiki.xen.org/xenwiki/Fedora13Xen4Tutorial.html) and have the whole setup running smoothly on my laptop. Here''s what I have: - Fedora Xen 4.0 with Linux 2.6.32.50 pvops dom0 - A CentOS 5.5 DomU guest host running on that Xen installation I''m using the command "xentrace -e 0x81000 trace.out" on my Fedora Xen to capture TRC_HVM_ENTRYEXIT events. While the xentrace is listening, I''m restarting the CentOS DomU host. When CentOS boots up again, I''m running Apache HTTP Server on it as well as some assembly code that I wrote which includes a call to CPUID. The CPUID call returns without any problems and the HTTP Server works as it''s supposed to, so I then stop the xentrace command that has been running on my Fedora Xen only to find an empty trace file (0 bytes). If I do "xentrace -e all trace.out", a lot of hypercalls and other events are logged (I use "cat trace.out| xentrace_format formats | more" to read the output file), so my xentrace is indeed working. From what I''ve read, the CPUID call and the Apache server should generate at least 1 VMEXIT, so I don''t understand why no VMEXIT events are being logged. I have googled a lot on the topic and haven''t found any information regarding this sort of behavior. Any help is greatly appreciated. Am I missing something? Do I have to recompile the Xen source with some specific flags? Do you know any VMEXIT-intensive programs that I can use to generate VM exits? Thanks for your time. I apologize if this is a "newbie" question...I''m a Networking guy and this topic is completely new to me. Happy New Year! -Diego