Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-09-07 Thread lilypond
Updates: Status: Invalid Comment #8 on issue 1858 by percival.music.ca: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 using 73f365d8e1f6c7034f4106e6c06cfa95896e3639 (one commit before 52bea08ef73a55ee ), I get this: cells: 1945658

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-09-03 Thread Han-Wen Nienhuys
On Sat, Sep 3, 2011 at 12:05 PM, wrote: > > Comment #7 on issue 1858 by reinhold...@gmail.com: unexplained increase in > profile reports > http://code.google.com/p/lilypond/issues/detail?id=1858 > > That commit turned on debugging in guile again... > > So the question now is, what was the case be

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-09-03 Thread lilypond
Comment #7 on issue 1858 by reinhold...@gmail.com: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 That commit turned on debugging in guile again... So the question now is, what was the case before Ian turned debugging off in february with co

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-09-03 Thread lilypond
Comment #6 on issue 1858 by percival.music.ca: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 oops, that was looking at (apparently) the wrong problem; this specific issue was about a memory thing between 2.15.8 and 2.15.9. Git bisect point

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-09-03 Thread lilypond
Comment #5 on issue 1858 by percival.music.ca: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 out/bin/lilypond -ddump-profile ../input/regression/lyric-tie.ly more ../input/regression/lyric-tie.profile Git master gives me weird amounts: cells:

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-09-02 Thread lilypond
Comment #4 on issue 1858 by k-ohara5...@oco.net: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 `lilypond -ddump-profile a.ly` creates 'a.profile'. (It doesn't work on vanilla MS Windows, but will in a development setup, or in more unix-y sys

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-09-02 Thread lilypond
Comment #3 on issue 1858 by percival.music.ca: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 Is there any way to see the cells on the command-line? (i.e. from an individual run of lilypond, instead of the whole make test-compare) I don't

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-08-31 Thread lilypond
Comment #2 on issue 1858 by d...@gnu.org: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 A memory leak or significant increase in allocation need not noticeably affect performance. And it may even improve performance (some algorithms can be

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-08-31 Thread lilypond
Comment #1 on issue 1858 by philehol...@googlemail.com: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 I'm almost 100% certain it does not relate to performance. On my Windows Vista system, 2.15.9 is _just_ slower than 2.15.7 (sorry, don't h

Re: Issue 1858 in lilypond: unexplained increase in profile reports

2011-08-31 Thread Mike Solomon
On Aug 31, 2011, at 8:50 AM, lilyp...@googlecode.com wrote: > Status: Accepted > Owner: > Labels: Type-Critical Regression > > New issue 1858 by k-ohara5...@oco.net: unexplained increase in profile reports > http://code.google.com/p/lilypond/issues/detail?id=1858 > > The reported 'cells', i

Issue 1858 in lilypond: unexplained increase in profile reports

2011-08-30 Thread lilypond
Status: Accepted Owner: Labels: Type-Critical Regression New issue 1858 by k-ohara5...@oco.net: unexplained increase in profile reports http://code.google.com/p/lilypond/issues/detail?id=1858 The reported 'cells', in the profile output for the regression tests, doubled for several test