Valerio Mariani wrote: > I confirm, this seems to be what is causing the problem on my Fedora > Core 5. If I turn off the service > > /etc/init.d/cpuspeed stop > > and start the mythfrontend, I get no errors at all both in the frontend > output and in the output of dmesg (at least for the few minutes that I > tried. A few seconds were easily sufficient to see the errors before) > > Debian has no cpuspeed script in the /etc/init.d/ directory, but I don't > know if this feature is compiled in the kernel. Is there a way to tell > from the output of some command or from looking at the kernel > configuration file?
Hi Valerio, I don't know about the different ways of doing speed/power management - I believe there are both in-kernel managers and userspace manager (like in FC5, RHEL4 etc). The core functionality of this powermanagement feature is compiled into the kernel (or as a module). On my kernel (2.6.16.6), the feature is built as CONFIG_X86_POWERNOW_K8=m CONFIG_X86_POWERNOW_K8_ACPI=y ...for my Athlon64 cpu, in i386 mode - I assume it will be similar on your system. On my system (no errors for a week or so), I am not only not running the userspace manager, but I am also /not/ loading the module. I'm not sure if this has relevance - I will shortly resume testing with the module loaded but the manager stopped, and then if there are no errors, try to introduce errors by rapidly toggling the cpu frequency. It is possible that just locking CPU frequency when grabbing data is enough to avoid problems. Time will show. Yours, -S > The following information was asked in a previous post: > > The compiler that I used to compile ivtv on Debian is: > > gcc (GCC) 4.1.2 20060729 (prerelease) (Debian 4.1.1-10) > > On Fedora I got the same problem both with precompiled packages from > atrpms and compiling ivtv myself. In this latter case I used: > > gcc (GCC) 4.1.1 20060525 (Red Hat 4.1.1-1) > > Please let me know if you need me to run more test and provide more > information!! > > Thank you everyone for your help! > > Valerio > > > > > Simen Thoresen wrote: >> Hi Valerio, >> >> Altho I have not done any IVTV development work, I have been gathering >> a bit of data on this with the help of a few other list participants. >> If you still have both Debian and FC5 available, please compare >> powernow-support (cpu speed throttling) between them - either as >> kernel modules or built in. I've been running a longeish testseries >> here, and I'm pretty sure I never saw the DMA error before enabling >> the 'cpuspeed' service (it'll probably be called that on FC5, but I >> don't know how Debian implements this). Unfortunately, a simultanous >> CPU, kernel and ivtv -upgrade skewed my earlier impression a bit, but >> I'm pretty sure this is neither a kernel (including ivtv) -version nor >> an SMP issue. >> >> I've now been running close to a week with cpuspeed disabled and >> rather more grabbing than I normally do, and have yet to see a single >> error (with my previous schedule it seems to occur every two days or >> so with cpuspeed enabled). >> >> Yours, >> -S > -- Simen Thoresen, Dolphin ICS Systems Administration and Wulfkit Support _______________________________________________ ivtv-users mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-users
