Bug 4223 - Source window slowness
Summary: Source window slowness
Alias: None
Product: frysk
Classification: Unclassified
Component: general (show other bugs)
Version: unspecified
: P2 normal
Target Milestone: ---
Assignee: Mike Cvet
Depends on:
Blocks: 1633 3392
  Show dependency treegraph
Reported: 2007-03-20 19:19 UTC by Rick Moseley
Modified: 2007-03-23 15:54 UTC (History)
0 users

See Also:
Last reconfirmed:


Note You need to log in before you can comment on or make changes to this bug.
Description Rick Moseley 2007-03-20 19:19:25 UTC
While fixing bz #4157 I downloaded the gnome-utils package so I could use the
"gcalctool" to use as an example of an rpm package that was downloaded, built
and then try to debug it.  In the process of debugging I noticed some slowness
in the loading of source files when stack frames are clicked on.  Especially
stack frame #7(gtk.c line #492).  (BTW, I was using gnome-utils-2.14.0-3.src.rpm
for the test so this could change with other versions.)

When clicking on the gtk.c stack frame it takes over 20 seconds on my i686/2GHz
P4 machine to bring up that window.  The source file for that has 3432 lines of
code in the current incarnation of that file.  I'm not sure right now where the
time is being spent.
Comment 1 Rick Moseley 2007-03-20 19:24:20 UTC
Added blockers.  It could be parser-related, but that would only cause it to be
slow the first time.  SOme investigation needs to occur.
Comment 2 Rick Moseley 2007-03-20 20:23:25 UTC
It appears the problem is NOT the parser being called too many times.  Debug
statements show that the parser is being called at the right times.
Comment 3 Mike Cvet 2007-03-23 12:55:18 UTC
	2007-03-23  Mike Cvet  <mcvet@redhat.com>
	* DOMSource.java (content): Added. String cache of the source code
	represented by the DOMSource. Addresses #4223.
	(setContent): Added.
	(getContent): Added.

	2007-03-23  Mike Cvet  <mcvet@redhat.com>
	* SourceBuffer.java (loadFile): Check to see if the source code has
	already been cached by the DOMSource in use. Addresses #4223.
Comment 4 Rick Moseley 2007-03-23 15:49:54 UTC
Those changes made significant progress Mike.  Changing between stack frames is
only 2-3 seconds on my x86 machine now.  Good work there sir!
Comment 5 Mike Cvet 2007-03-23 15:54:11 UTC
Closing via comment above. Thanks Rick!