On Thu, Sep 13, 2007 at 06:54:02PM -0700, Linus Torvalds wrote: > On Thu, 13 Sep 2007, Junio C Hamano wrote: > > Dmitry V. Levin writes: > > > Yes, the check looks almost redundant if run_status is enabled, but > > > - the bug fixed by commit v1.5.3.1-15-g4fb5fd5 was found due to > > > this check; > > > - the check itself is very cheap compared to run_status; > > > > That is worrisome. Maybe run_status is too expensive because it > > is doing something stupid? IOW, can you make run_status cheaper > > for you? > > It would be interesting to hear what the situation is that makes > run_status be too expensive. The situation when runstatus is too expensive is committing a large tree with massive changes. More or less real life example is committing several openoffice.org unpacked tarballs. First commit is relatively fast: $ mkdir 1 $ cd 1 $ git init $ tar xf ../OOo_2.0.4_src.tar.gz $ mv OO* tree $ git add tree (about 1.75 minutes) $ git commit -m 2.0.4 >/dev/null (about 30 seconds) $ git repack -a -d -q (about 8.5 minutes) $ du -h .git/objects/pack 316M .git/objects/pack $ cd - Second commit is slow: $ mkdir 2 $ cd 2 $ git init $ tar xf ../OOo_1.1.5_src.tar.gz $ mv OO* tree $ git add tree (about 50 seconds) $ git commit -m 1.1.5 >/dev/null (about 20 seconds) $ git repack -a -d -q (about 2 minutes) $ du -h .git/objects/pack 202M .git/objects/pack $ rm -rf tree $ rm -f .git/index $ tar xf ../OOo_2.0.4_src.tar.gz $ mv OO* tree $ git add tree (about 1.75 minutes) $ git commit -m 2.0.4 >/dev/null git-runstatus runs much longer (tens of minutes, I do not remember), and finally fails, either with OOM error (if no sufficient virtual memory available; on x86-64 64G was not enough) or segfaults. -- ldv
Attachment:
pgpouCAUdQ43J.pgp
Description: PGP signature
