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

[Mike Solomon]

On Aug 31, 2011, at 8:50 AM, address@hidden wrote:

> Status: Accepted
> Owner: ----
> Labels: Type-Critical Regression
> 
> New issue 1858 by address@hidden: 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 tests between 2.15.8 and 2.15.9.
> 
> "lyric-tie.profile" was in 2.15.8
> time: 0
> cells: 1125395
> 

I remember seeing a hike because of my stem work in certain regtests, but it 
was of a factor of about 1.15 and was due to the presence of beams and the 
caching of their pure heights.  I fixed this in part with my change to item.cc, 
but  this could still be problematic.

Is there any way to get a chart breaking down what grobs are doing the cell 
accessing?

Cheers,
MS