Posted: Sat Sep 20, 2003 9:31 am Post subject: Re: huge swap while printing
Max,
Max_Barel wrote:
On minor issues list I add this one: some of the shortcuts no longer work (they did previously). The command-A=select-all is an exemple.
I am thankful that we have testers throughout the world because I found that this bug appears on a French keyboard but not on a U.S. keyboard.
I bug was due to a NeoJ hack that strips off the key modifiers (e.g. Alt, Ctrl, etc.) if Mac OS X used the modifier to set the key pressed. For example, on a U.S. keyboard, Alt-C generates a C-cedilla. Apparently, the OOo code does not like the Alt key (probably because of the Alt+<number> feature in Windows?) and will ignore the key if NeoJ does not strip off the Alt modifier.
I changed the logic of this hack to only strip off the Alt modifier and leave the others unchanged. This fix works with both a French and U.S. keyboard but it will need to be tested.
Joined: May 31, 2003 Posts: 219 Location: French Alps
Posted: Mon Sep 22, 2003 4:13 am Post subject:
pluby wrote:
Of course, there might be bugs in my code. If a NeoJ print job uses up over 1 GB of swap and never ends and it is only a few pages long, then it is likely that one of my printing functions is recursively calling itself and never ending.
...
If you or JKT culd send me a document that has this memory problem, I would really appreciate it. It is much easier to run real documents while in the debugger to see what OOo and NeoJ are actually doing.
While writing this followup I'm uploading a tgz file of the sxd file and his linked images. This file is a press review of my glider project with many images embedded for a total of 14Mb. I'll send you the link on a private mail.
This post is to precise that the same file printed with OOo/X11 GM never use more than 200 Meg of RSS memory and no more than 6 swap files, where there were already 4 before launching OOo.
Max
Joined: Sep 18, 2003 Posts: 434 Location: London, UK
Posted: Mon Sep 22, 2003 4:21 am Post subject:
pluby wrote:
A gig of disk? That doesn't sound normal. I wonder if there is some sort of infinite recursion that is happening. Can you send the document to my private e-mail: patrick.luby@planamesa.com? Also, how are you measuring swap usage? I would like to see if I get similar swap measurements with your document on my machine.
As for the Command-H problem. That bug is in OOo X11 as well which is why you see if in NeoJ.
Patrick
Hi Patrick, sorry for the delay - the document is in the mail, together with the sample 15 I took at the time. As for swap files/paging:
1) I have the freeware Memory Monitor dockling which shows you how much active, wired, free memory you are using together with page-ins/outs. From this (along with the grinding noise from my disk ) I could see first hand that the paging was constant and very, very high.
2) I had the /var/vm directory open in the Finder to see how many swap files I had prior to starting printing and refreshed it every now and then throughout... it got to beyond 19 swapfiles prior to my force-quit (up from a starting position of 6). Each is 76.2MB in size, ergo about a GB of disk space was used up.
N.B. Perhaps it wasn't clear in my earlier posts, but the "Printing" display bar that appears at the bottom of the document window when you print something didn't indicate anything was printing suggesting that the print job hadn't even started spooling.
Thanks for the sample files. I am able to replicate each of the problems that both of you describe and it is clear that NeoJ is using up vast amounts of memory with these two documents.
I will do some debugging work and see if I can pinpoint where all of that memeory is being allocated. My guess is that I have a memory leak somewhere or memory is being retained too long.
In the meantime, I have posted an updated patch at http://www.planamesa.com/neojava/downloads/temp/printtest6.tar.gz. This patch contains fixes for Thomas' "black box" problem, a fix for the "missing bullets" problem, a fix for the "Mac view style" font problems, and all of the other fixes that I have mentioned in this thread.
Once I figure out a fix for the memory problem that Max and JKT have found, I will post a patch for that.
Posted: Mon Sep 22, 2003 10:44 pm Post subject: a couple more glitches
1. All Arrows seem to count double (e.g. left arrow = two lefts)
2. Copy randomly spontaneously reverts to Cmd-Insert from Cmd-C keybinding.
- Apricot
ps. This glitch appears in printtest5 and printtest6. I did some testing, and (1) only happens when one uses Cmd-arrow, and it goes away when one types a non-arrow key. A bit awkward when you use Cmd-arrow to navigate close to an insert point and then use arrows to finish the job...
ppps. I'm sure we are all ecstatic at pluby's work on this.
I found the problem that was causing the arrow key to move two characters. The fix will be in the next patch.
JKT,
This fix also fixes the bug for highlighting by holding the Shift keys while pressing the arrow or page up/down keys.
Also, while I was fixing the above problem, I realized that I could add support for Shift-mouse-click. Although I've talked about the conflict with Alt-mouse-click and Ctrl-mouse-click, I realized that Java has no conflict with te Shift-mouse-click.
This new Shift-mouse-click will be in the next patch.
All times are GMT - 7 Hours Goto page Previous1, 2
Page 2 of 2
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You cannot download files in this forum