Hubertus Franke wrote:
Here are the results of a simple "performance experiment".
I performed a series of kernel compiles "make -j3 bzImage" on my labtop for different configurations against 2.6.8.1. Times are reported "total (sys)" in seconds. "*" indicates after reboot.
vanilla CKRM-E16 + cpu-v8 CKRM-E16+cpu-v8+numtasks 336.75(32.01)* 337.92(31.96)* 332.29(31.87) 332.64(31.56) 332.71(31.48) 332.68(31.60) 332.99(31.53) 333.18(31.86) 332.44(31.83 332.53(31.56) 332.72(31.79) 332.81(31.65)
In the CKRM case, no only the CPU controller was compiled into the kernel (as required). numtasks was dynamically loaded. No other controllers were enabled.
Based on these result, I don't see in this workload any overhead for CKRM, the cpu scheduler or the numtasks controller.
The memory controller is next to test in this series. But I simply wanted to get a few potential culprits off the most-wanted list.
-- Hubertus
On my 1.8 GHz P4, 512KB mem machine, I'm able to successfully compile a 2.6.8.1 kernel with make -j4 with each controller added successively.
The times are given below.
To isolate the problem that Mark is seeing (I'm seeing a panic using his tree too), I'll try disabling the controllers in his tree one by one.
-- Shailabh
linux-2.6.8.1 ------------- stable, responsive
+ ckrm-e16 (CKRM disabled in config) ------------------------------------ stable, responsive
+ ckrm-e16 (numtasks, listenaq configured, NOT loaded) ------------------------------------------------------ stable, responsive
+ ckrm-e16 (numtasks, listenaq configured, loaded) ----------------------------------------------------
real 10m37.386s user 9m42.486s sys 0m51.418s
+ ckrm-e16 + cpu-v7 (loaded) ----------------------------
real 10m36.842s user 9m42.162s sys 0m51.232s
+ ckrm-e16 + cpu-v7 (loaded) + mem-v2 (loaded) -----------------------------------------------
real 10m46.066s user 9m46.602s sys 0m55.579s ---> somewhat of a jump here
+ ckrm-e16 + cpu-v7 (loaded) + mem-v2 (loaded) + io-v4 (loaded) ---------------------------------------------------------------
real 10m41.296s user 9m43.554s sys 0m54.936s
In the meantime, I've been able to successfully run the kernel compile on a Fedora 2.6.8-1.521 core with all controllers loaded.
I used a Fedora core from www.linuxant.com/driverloader/wlan/full/downloads-fc2-kernel-i686.php
which doesn't force the use of 4K stacks and without the 4G/4G split. and configured it to use the normal 8K stack (4G/4G still disabled).
Using your config file, Mark, the fedora core boots but hangs at a later time while trying to do "make modules_install" for another kernel. Using my config, both "Fedora core vanilla" and "Fedora core + all CKRM controllers" work ok.
I'll take up the config file issue w/you offline as we isolate the panics you're seeing.
-- Shailabh
------------------------------------------------------- This SF.net email is sponsored by: IT Product Guide on ITManagersJournal Use IT products in your business? Tell us what you think of them. Give us Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more http://productguide.itmanagersjournal.com/guidepromo.tmpl _______________________________________________ ckrm-tech mailing list https://lists.sourceforge.net/lists/listinfo/ckrm-tech
