I have added a get_cpu() library which should just do the right thing for you whether your system uses sched_getcpu() (the most common case), getcpu() or the vsyscall version of getcpu(). I have tested on Fedora, RHEL5 both 32 and 64-bit versions which cover all of the cases above. I would really appreciate more testing and reports on other distributions. Of course comments and patches are also welcome. While doing this I changed the Makefile quite extensively. I hope you will find the changes simply and unify the Make process considerably. This could be further improved too - for example by creating a build directory. In addition to the changes below, Clark has fixed the odd CRLF issue, and has a patch queued-up from Carsten that was missed last time. Clark, you can pull the following changes from: git://git.kernel.org/pub/scm/linux/kernel/git/jkacur/rt-tests.git branch: rt-tests-dev Thanks Carsten Emde (1): Completely remove VERSION_STRING until we have a better solution. John Kacur (3): rt-tests: Makefile - Changes to unify and simplify the Makefile rt-tests: Add a get_cpu() function to the library. rt-tests: Clean-up - protect rt-utils.h Makefile | 53 ++++++++++++++++++++------------ src/backfire/backfire.c | 2 +- src/backfire/sendme.c | 9 +++-- src/cyclictest/cyclictest.c | 1 - src/lib/rt-get_cpu.c | 25 +++++++++++++++ src/lib/rt-get_cpu.h | 46 ++++++++++++++++++++++++++++ src/lib/rt-utils.h | 5 +++ src/pi_tests/pi_stress.c | 7 +--- src/ptsematest/ptsematest.c | 12 +++---- src/rt-migrate-test/rt-migrate-test.c | 2 +- src/sigwaittest/sigwaittest.c | 8 +++-- src/svsematest/svsematest.c | 31 ++++--------------- 12 files changed, 135 insertions(+), 66 deletions(-) create mode 100644 src/lib/rt-get_cpu.c create mode 100644 src/lib/rt-get_cpu.h -- 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
