From: SourceForge.net <no...@so...> - 2012-08-08 22:33:17
|
Bugs item #3555427, was opened at 2012-08-08 14:27 Message generated for change (Comment added) made by miesfeld You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=684730&aid=3555427&group_id=119701 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Interpreter Group: None >Status: Closed Resolution: None Priority: 5 Private: No Submitted By: Mark Miesfeld (miesfeld) Assigned to: Nobody/Anonymous (nobody) Summary: Memory problem on Windows. Initial Comment: Using the attached Rexx programs, I see the memory working set of rexx.exe steadily climbing until all usable memory is consumed. I've seen this on several systems, one of which is a Windows 64-bit system running Windows 7. Tested with trunk at r8139 or later. The base program is test4.rex which is a program supplied by a user for the SysFileTree crashing on Windows bug, #3405740. That program does a SysFileTree search for every file on the C: drive. My system has around 300,000 files. I used another Rexx program to run the base program in an endles loop. When the base program is run in an endless loop, I see memory climbing as noted above. I've created 2 base programs. One is the original test program that has this code in it: do i=1 to f.0 if \var("f.i") then do sw = 1 say "Stem variable f."i "has not been assigned a value." end end And another base program has that code removed. With the above code, memory keeps climbing. Without, it memory, on my system, reaches about 200 MB. The 4 attached programs: execTest4.rex -> calls test4.rex in an endless loop execTest4withVarTest.rex -> calls test4-withVarTest.rex in an endless loop Running execTest4withVarTest.rex shows rexx.exe working set continually climbing. Running execTest4.rex shows the working set increase and then level out and remain steady. I used task manager to monitor the working set. About 20 minutes of CPU time is enough to show the dramatic difference. ---------------------------------------------------------------------- >Comment By: Mark Miesfeld (miesfeld) Date: 2012-08-08 15:33 Message: Closing this. I didn't analyze it correctly. Rick explained it on the devel list. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=684730&aid=3555427&group_id=119701 |