Overview
Artifact ID: | 4ff704a9368d3ea503b608ca2afa2d7b51e5ec5a |
---|---|
Ticket: | f00009f7ce4b58c8b6caedf976e37ea78be6d3c0
file exists, mtime... procedures leaks |
User & Date: | sebres 2015-07-30 15:39:04 |
Changes
- closer changed to: "nobody"
- icomment:
Not so fast, because seems to be not so easy to fix... That fixes a leak only if I give this line direct in stdin inside tclsh: % time {file exists [string trim "C:/example.txt "]} 100000 If I do the same in a file (for example "test"), and execute it via `source test`, it leaks again. Possible another constellation regarding of objects, or refcounts (history involved) or literals, etc. Very strange also, that the tests have different execution times (constantly reproducible): - with leaks (% source ...): 16.03923 microseconds per iteration - w/o leaks (% time direct): 90.1751 microseconds per iteration But a 1 year older branch does the same without leaks within 15ms also!
- login: "sebres"
- mimetype: "text/plain"
- resolution changed to: "Wont Fix"
- status changed to: "Open"