Mercurial > hg > mercurial-crew-with-dirclash
view tests/test-bundle @ 2474:1e32e2fe8a67
Fix cold cache diff performance
cold cache diff performance has regressed in two ways. localrepo.changes
has optimizations for diffing against the working dir parent that expect
node1 to be None. commands.revpair() usage means that commands.dodiff()
never sends node1 == None. This is fixed in localrepo.changes by checking
against the dirstate parents.
In the non-dirstate parents case, localrepo.changes does a loop comparing
files without first sorting the file names, leading to random access
across the disk.
| author | Chris Mason <mason@suse.com> |
|---|---|
| date | Wed, 21 Jun 2006 09:28:48 -0700 |
| parents | 89b15372df18 |
| children | 109a22f5434a |
line wrap: on
line source
#!/bin/sh hg init test cd test echo 0 > afile hg add afile hg commit -m "0.0" -d "1000000 0" echo 1 >> afile hg commit -m "0.1" -d "1000000 0" echo 2 >> afile hg commit -m "0.2" -d "1000000 0" echo 3 >> afile hg commit -m "0.3" -d "1000000 0" hg update -C 0 echo 1 >> afile hg commit -m "1.1" -d "1000000 0" echo 2 >> afile hg commit -m "1.2" -d "1000000 0" echo "a line" > fred echo 3 >> afile hg add fred hg commit -m "1.3" -d "1000000 0" hg mv afile adifferentfile hg commit -m "1.3m" -d "1000000 0" hg update -C 3 hg mv afile anotherfile hg commit -m "0.3m" -d "1000000 0" hg verify cd .. hg init empty hg -R test bundle full.hg empty hg -R test unbundle full.hg hg -R empty unbundle full.hg hg -R empty heads hg -R empty verify rm -rf empty hg init empty cd empty hg -R bundle://../full.hg log #doesn't work (yet ?) #hg -R bundle://../full.hg verify hg pull bundle://../full.hg cd .. rm -rf empty hg init empty hg clone -r 3 test partial hg clone partial partial2 cd partial hg -R bundle://../full.hg log hg incoming bundle://../full.hg hg -R bundle://../full.hg outgoing ../partial2 cd ..