Changes committed to git://oss.sgi.com/pcp/pcp.git
CHANGELOG | 5 +++
VERSION.pcp | 2 -
configure.in | 55 +++++++++++++++++++++++++++--------------
debian/changelog | 7 +++++
man/man1/pmval.1 | 27 +++++++++++++++++++-
src/libpcp_mmv/src/mmv_stats.c | 1
src/pmval/pmval.c | 48 +++++++++++++++++++++++++++++------
7 files changed, 117 insertions(+), 28 deletions(-)
commit 11549aca86588c4695d26829e02f2ec9e7b43c3d
Author: Nathan Scott <nathans@xxxxxxxxxx>
Date: Mon Sep 7 16:29:34 2009 +1000
Bump pcp stable version to 2.9.2.
commit 7997eea93ad9c240d2cbc7c2bfbd9ae21bb29e35
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date: Mon Sep 7 11:59:21 2009 +1000
Fix the number-of-samples value calculated in pmval once and for all.
For those who care, it turns out that the number of _fetches_ may be
known, the number of _reported_ samples cannot be precomputed correctly
in all cases ...
- only works for archives with -a (not -U or -h)
- is known for non-counter metrics
- most of the time is known for counter metrics but there are some
important corner cases where this is not the case, specifically there
will be one more reported sample than expected for a counter when the
report starting time is not at the start of the archive and the metric
has a defined value before the report starting time with no intervening
mark record (interpolate mode does return a value at the first report
sample in this case!).
With this pmval patch, the QA fallout is nil, except for 144 that is an
unrelated timezone problem. Also this patch avoids referencing the
context too soon, as there is no point in inspecting the metric's
semantics to try and better guess the value of smpls. As a bonus, this
patch also documents pmval's -U option that has remained a secret for
nearly a decade.
commit f4e6d55fd0466a706144ca61b4428436ecea7eb7
Author: Nathan Scott <nathans@xxxxxxxxxx>
Date: Mon Sep 7 11:51:58 2009 +1000
Attempt to resolve sys/stat related build issue on Alpha.
Not having access to an Alpha makes this difficult, but one thing
that is clearly not helping is that with Linux glibc on Alpha, we
end up defining *both* HAVE_STAT_TIMESPEC and HAVE_STAT_TIME_T, &
that breaks otherwise valid pmns.c coding assumptions.
This change means only one type for the stat time fields can be
defined, and it errs on the side of timespec (which should be the
one Alpha uses, AFAICT) - so perhaps this issue will finally be
resolved and PCP can progress in the Debian build/release system.
commit be8793c58935362184a1dd4351388ca6a0013f3b
Author: Nathan Scott <nathans@xxxxxxxxxx>
Date: Mon Sep 7 11:26:09 2009 +1000
Fix mmv library build on mac/win - missing a header.
|