[PATCH 0/7 (introduction)] cpu-timers and friends

From: Roland McGrath
Date: Mon Dec 13 2004 - 22:54:40 EST

This message is to introduce the coming seven messages. These relate to
process/thread CPU-time clocks for POSIX timers and some related fixes for
proper POSIX semantics in multithreaded programs (under NPTL).

These patches were made against a recent current tree, plus my stoprace-fix
signals patch. That patch follow-on and the group_exit cleanup are already
in 2.6.10-rc3-mm1, but these patches were made against a Linus tree with
just the stoprace-fix patch. (That fix is necessary for some of these
changes to be safe, while the group_exit cleanup is just a non-bug cleanup.)
If you need these patches remade against -mm or some other given tree
state, please let me know. I'm happy to do whatever makes it easiest to
get them into the tree ASAP.

Note that the first patch changes an ill-advised new user-kernel ABI that
has been introduced since 2.6.9; I really think we should not let the
existing clockid_t encoding change get out in 2.6.10. I hope this patch
(and ideally all the rest) can go into 2.6.10, but reverting the earlier
patch so that we have no new clockid_t values at all would be better than
letting the existing code for CPU clocks make it into 2.6.10, IMHO.

These can also be found in http://people.redhat.com/roland/kernel-patches/
Not all of the changes are strict dependencies, but these patches will need
to be applied in the order of posting, which is:


In http://people.redhat.com/roland/glibc/ you can find a glibc patch to
enable using the new clock_* and timer_* kernel support. This includes
some test programs. I've tested the kernel code using these programs, on a
2-CPU x86 machine. There shouldn't be anything machine-dependent about the
code, but architectures vary on how good the time they implement for
`sched_clock' is; on some, it's nothing but the same jiffies-based count
you see via getrusage.

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/