##// END OF EJS Templates
perf: perform a garbage collection before each iteration...
perf: perform a garbage collection before each iteration Currently, no explicit garbage collection is performed when running the microbenchmarks in `hg perf`. I think this is wrong because garbage collection can have a significant impact on execution times. And, if gc is triggered via the default heuristics, it will fire effectively randomly during subsequent benchmark iterations due to variable amount of garbage left over from previous runs. Running a gc before invoking the measured function will help ensure state is more consistent across all iterations.

File last commit:

r29641:13f90dde stable
r31397:8f5ed8fa default
Show More
packagelib.sh
35 lines | 1.1 KiB | application/x-sh | BashLexer
# Extract version number into 4 parts, some of which may be empty:
#
# version: the numeric part of the most recent tag. Will always look like 1.3.
#
# type: if an rc build, "rc", otherwise empty
#
# distance: the distance from the nearest tag, or empty if built from a tag
#
# node: the node|short hg was built from, or empty if built from a tag
gethgversion() {
make cleanbutpackages
make local || make local PURE=--pure
HG="$PWD/hg"
$HG version > /dev/null || { echo 'abort: hg version failed!'; exit 1 ; }
hgversion=`$HG version | sed -ne 's/.*(version \(.*\))$/\1/p'`
if echo $hgversion | grep + > /dev/null 2>&1 ; then
tmp=`echo $hgversion | cut -d+ -f 2`
hgversion=`echo $hgversion | cut -d+ -f 1`
distance=`echo $tmp | cut -d- -f 1`
node=`echo $tmp | cut -d- -f 2`
else
distance=''
node=''
fi
if echo $hgversion | grep -- '-' > /dev/null 2>&1; then
version=`echo $hgversion | cut -d- -f1`
type=`echo $hgversion | cut -d- -f2`
else
version=$hgversion
type=''
fi
}