| Message ID | 869aae2abe70587a0fe36dacfbb7d50b74f3b1c7.1249475102.git.mgoldish@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
I'd have one sugestion to make on this patch: On Wed, Aug 5, 2009 at 9:32 AM, Michael Goldish<mgoldish@redhat.com> wrote: > 1. As suggested by Yolkfull Chow, use a get_time() helper function that returns > Â the current host and guest times. > > 2. Use helper functions that set and restore the CPU affinity of a process. > > 3. In these helper functions, set the affinity of all threads of the process, > Â not just the main thread. > > Signed-off-by: Michael Goldish <mgoldish@redhat.com> > --- > Â client/tests/kvm/kvm_tests.py | Â 68 +++++++++++++++++++++++++++------------- > Â 1 files changed, 46 insertions(+), 22 deletions(-) > > diff --git a/client/tests/kvm/kvm_tests.py b/client/tests/kvm/kvm_tests.py > index 308db97..67cf713 100644 > --- a/client/tests/kvm/kvm_tests.py > +++ b/client/tests/kvm/kvm_tests.py > @@ -604,6 +604,45 @@ def run_timedrift(test, params, env): > Â Â @param params: Dictionary with test parameters. > Â Â @param env: Dictionary with the test environment. > Â Â """ > + Â Â # Helper functions > + Â Â def set_cpu_affinity(pid, mask): > + Â Â Â Â """ > + Â Â Â Â Set the CPU affinity of all threads of the process with PID pid. > + > + Â Â Â Â @param pid: The process ID. > + Â Â Â Â @param mask: The CPU affinity mask. > + Â Â Â Â @return: A dict containing the previous mask for each thread. > + Â Â Â Â """ > + Â Â Â Â tids = commands.getoutput("ps -L --pid=%s -o lwp=" % pid).split() > + Â Â Â Â prev_masks = {} > + Â Â Â Â for tid in tids: > + Â Â Â Â Â Â prev_mask = commands.getoutput("taskset -p %s" % tid).split()[-1] > + Â Â Â Â Â Â prev_masks[tid] = prev_mask > + Â Â Â Â Â Â commands.getoutput("taskset -p %s %s" % (mask, tid)) > + Â Â Â Â return prev_masks > + > + Â Â def restore_cpu_affinity(prev_masks): > + Â Â Â Â """ > + Â Â Â Â Restore the CPU affinity of several threads. > + > + Â Â Â Â @param prev_masks: A dict containing TIDs as keys and masks as values. > + Â Â Â Â """ > + Â Â Â Â for tid, mask in prev_masks.items(): > + Â Â Â Â Â Â commands.getoutput("taskset -p %s %s" % (mask, tid)) > + > + Â Â def get_time(): > + Â Â Â Â """ > + Â Â Â Â Returns the host time and guest time. > + > + Â Â Â Â @return: A tuple containing the host time and guest time. > + Â Â Â Â """ > + Â Â Â Â host_time = time.time() > + Â Â Â Â session.sendline(time_command) > + Â Â Â Â (match, s) = session.read_up_to_prompt() > + Â Â Â Â s = re.findall(time_filter_re, s)[0] > + Â Â Â Â guest_time = time.mktime(time.strptime(s, time_format)) > + Â Â Â Â return (host_time, guest_time) The above function. It would be slightly clearer to make it to take the time_command as a parameter rather than taking it from the global namespace 'implicitely'. > Â Â vm = kvm_utils.env_get_vm(env, params.get("main_vm")) > Â Â if not vm: > Â Â Â Â raise error.TestError("VM object not found in environment") > @@ -641,19 +680,12 @@ def run_timedrift(test, params, env): > Â Â guest_load_sessions = [] > Â Â host_load_sessions = [] > > - Â Â # Remember the VM's previous CPU affinity > - Â Â prev_cpu_mask = commands.getoutput("taskset -p %s" % vm.get_pid()) > - Â Â prev_cpu_mask = prev_cpu_mask.split()[-1] > Â Â # Set the VM's CPU affinity > - Â Â commands.getoutput("taskset -p %s %s" % (cpu_mask, vm.get_pid())) > + Â Â prev_affinity = set_cpu_affinity(vm.get_pid(), cpu_mask) > > Â Â try: > Â Â Â Â # Get time before load > - Â Â Â Â host_time_0 = time.time() > - Â Â Â Â session.sendline(time_command) > - Â Â Â Â (match, s) = session.read_up_to_prompt() > - Â Â Â Â s = re.findall(time_filter_re, s)[0] > - Â Â Â Â guest_time_0 = time.mktime(time.strptime(s, time_format)) > + Â Â Â Â (host_time_0, guest_time_0) = get_time() So the above call would be get_time(time_command). This way we can see clearly that the time_command is being passed to the function. However, not entirely necessary. Patch looks good, applied. http://autotest.kernel.org/changeset/3514 -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/client/tests/kvm/kvm_tests.py b/client/tests/kvm/kvm_tests.py index 308db97..67cf713 100644 --- a/client/tests/kvm/kvm_tests.py +++ b/client/tests/kvm/kvm_tests.py @@ -604,6 +604,45 @@ def run_timedrift(test, params, env): @param params: Dictionary with test parameters. @param env: Dictionary with the test environment. """ + # Helper functions + def set_cpu_affinity(pid, mask): + """ + Set the CPU affinity of all threads of the process with PID pid. + + @param pid: The process ID. + @param mask: The CPU affinity mask. + @return: A dict containing the previous mask for each thread. + """ + tids = commands.getoutput("ps -L --pid=%s -o lwp=" % pid).split() + prev_masks = {} + for tid in tids: + prev_mask = commands.getoutput("taskset -p %s" % tid).split()[-1] + prev_masks[tid] = prev_mask + commands.getoutput("taskset -p %s %s" % (mask, tid)) + return prev_masks + + def restore_cpu_affinity(prev_masks): + """ + Restore the CPU affinity of several threads. + + @param prev_masks: A dict containing TIDs as keys and masks as values. + """ + for tid, mask in prev_masks.items(): + commands.getoutput("taskset -p %s %s" % (mask, tid)) + + def get_time(): + """ + Returns the host time and guest time. + + @return: A tuple containing the host time and guest time. + """ + host_time = time.time() + session.sendline(time_command) + (match, s) = session.read_up_to_prompt() + s = re.findall(time_filter_re, s)[0] + guest_time = time.mktime(time.strptime(s, time_format)) + return (host_time, guest_time) + vm = kvm_utils.env_get_vm(env, params.get("main_vm")) if not vm: raise error.TestError("VM object not found in environment") @@ -641,19 +680,12 @@ def run_timedrift(test, params, env): guest_load_sessions = [] host_load_sessions = [] - # Remember the VM's previous CPU affinity - prev_cpu_mask = commands.getoutput("taskset -p %s" % vm.get_pid()) - prev_cpu_mask = prev_cpu_mask.split()[-1] # Set the VM's CPU affinity - commands.getoutput("taskset -p %s %s" % (cpu_mask, vm.get_pid())) + prev_affinity = set_cpu_affinity(vm.get_pid(), cpu_mask) try: # Get time before load - host_time_0 = time.time() - session.sendline(time_command) - (match, s) = session.read_up_to_prompt() - s = re.findall(time_filter_re, s)[0] - guest_time_0 = time.mktime(time.strptime(s, time_format)) + (host_time_0, guest_time_0) = get_time() # Run some load on the guest logging.info("Starting load on guest...") @@ -676,22 +708,18 @@ def run_timedrift(test, params, env): timeout=0.5)) # Set the CPU affinity of the shell running the load process pid = host_load_sessions[-1].get_shell_pid() - commands.getoutput("taskset -p %s %s" % (cpu_mask, pid)) + set_cpu_affinity(pid, cpu_mask) # Try setting the CPU affinity of the load process itself pid = host_load_sessions[-1].get_pid() if pid: - commands.getoutput("taskset -p %s %s" % (cpu_mask, pid)) + set_cpu_affinity(pid, cpu_mask) # Sleep for a while (during load) logging.info("Sleeping for %s seconds..." % load_duration) time.sleep(load_duration) # Get time delta after load - host_time_1 = time.time() - session.sendline(time_command) - (match, s) = session.read_up_to_prompt() - s = re.findall(time_filter_re, s)[0] - guest_time_1 = time.mktime(time.strptime(s, time_format)) + (host_time_1, guest_time_1) = get_time() # Report results host_delta = host_time_1 - host_time_0 @@ -704,7 +732,7 @@ def run_timedrift(test, params, env): finally: logging.info("Cleaning up...") # Restore the VM's CPU affinity - commands.getoutput("taskset -p %s %s" % (prev_cpu_mask, vm.get_pid())) + restore_cpu_affinity(prev_affinity) # Stop the guest load if guest_load_stop_command: session.get_command_output(guest_load_stop_command) @@ -719,11 +747,7 @@ def run_timedrift(test, params, env): time.sleep(rest_duration) # Get time after rest - host_time_2 = time.time() - session.sendline(time_command) - (match, s) = session.read_up_to_prompt() - s = re.findall(time_filter_re, s)[0] - guest_time_2 = time.mktime(time.strptime(s, time_format)) + (host_time_2, guest_time_2) = get_time() # Report results host_delta_total = host_time_2 - host_time_0
1. As suggested by Yolkfull Chow, use a get_time() helper function that returns the current host and guest times. 2. Use helper functions that set and restore the CPU affinity of a process. 3. In these helper functions, set the affinity of all threads of the process, not just the main thread. Signed-off-by: Michael Goldish <mgoldish@redhat.com> --- client/tests/kvm/kvm_tests.py | 68 +++++++++++++++++++++++++++------------- 1 files changed, 46 insertions(+), 22 deletions(-)