Hello CIP RT Folks! I'm pleased to announce the Linux 4.4.15-cip15-rt10 stable release. You can get this release via the git tree at: git://git.kernel.org/pub/scm/linux/kernel/git/wagi/linux-cip-rt.git https://git.kernel.org/pub/scm/linux/kernel/git/wagi/linux-cip-rt.git branch: linux-4.4.y-cip-rt Head SHA1: 6c97018d1d4626e21b5233eb28e607e412c79ebe Currently, I don't have uploaded the patches, because I need to figure out how this works on kernel.org. The release is based on the Linux 4.4.75-rt88 version. That is I applied the -rt88 patch on the 4.4.75-cip6 kernel and merged in all CIP releases. Note we got via the stable tree the new RT pull/push logic from upstream cb1831a83e54 ("sched/rt: Simplify the IPI based RT balancing logic"). I spend quite a while to get a baseline configuration working and the testing up and running. I have a BeagleBone Black and a MinnowBoard Turbot as test target. On the BBB, cyclictest reports reasonable numbers, though not the MinnowBoard Turbot. It seems the bios/uefi or so is eating CPU time. The exact same kernel works on different x86 hardware just fine. Furthermore, I let the kernel run on the Real-Time Linux Continuous Integration servers: https://ci-rt.linutronix.de/RT-Test/kbuild.jsp?id=112 Some of the builds are failing (allmodconfig), though it doesn't look cip-rt specific, eg: kernel/utilities/compiletest-runner@2/build/../include/linux/compiler.h:484 call to ‘__compiletime_assert_150’ declared with attribute error: BUILD_BUG_ON failed: hw_flag_names[NUM_IEEE80211_HW_FLAGS] != (void *)0x1 Not sure how to handle those. IIRC, we don't support allmodconfig. Next steps are improving my tests and identifying which 4.4.${x}-rt${y} changes should be added to the cip-rt kernel. Thanks, Daniel -- To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
