On Thu, Jul 19, 2018 at 07:41:03AM +0200, Duy Nguyen wrote: > On Thu, Jul 19, 2018 at 12:51 AM Elijah Newren <newren@xxxxxxxxx> wrote: > > > > I had a user report some poor behavior of 'git gc --aggressive' on a > > certain repo (which I sadly cannot share). Turns out that on this > > repo, this operation takes about 60% longer and produces a pack > > roughly twice the expected size. > > The intention was to make life better for weaker machines but > definitely should not slow down beefier ones, so yes this is > definitely a regression. > > Is it possible to share "verify-pack -v <pack file>" output of the > pack produced by 2.17.0 and 2.18.0? The only sensitive info there is > sha-1, which you can replace with just "SHA-1" if you want. I'm more > interested in delta sizes and distribution. Try this: -- >8 -- #!/bin/sh rm -rf repo git init repo cd repo dd if=/dev/urandom of=one.rand bs=1M count=2 dd if=/dev/urandom of=two.rand bs=1M count=2 dd if=/dev/urandom of=big.rand bs=1M count=20 cat one.rand big.rand >file git add file git commit -m one cat two.rand big.rand >file git add file git commit -m two git repack -ad git cat-file --batch-all-objects --batch-check='%(objectname) %(deltabase)' -- 8< -- Using git 2.17 for the repack results in a single delta (which should be about 2MB, because it will say "delete one.rand and add two.rand" or vice versa). Using git 2.18, I get no delta at all. -Peff
