#2992: GHCi Memory Leak in Windows Vista
-----------------------+----------------------------------------------------
Reporter: Andir | Owner: igloo
Type: merge | Status: closed
Priority: high | Milestone: 6.10.2
Component: GHCi | Version: 6.10.1
Severity: normal | Resolution: fixed
Keywords: | Difficulty: Unknown
Testcase: | Os: Windows
Architecture: x86 |
-----------------------+----------------------------------------------------
Comment (by simonmar):
Compiling a stage 2 with `-debug` and then running it with `--interactive
+RTS -Ds` gave enough clues: every time I ran `ghc` in another window the
first `ghc` process would start running thread 2, which I happened to know
was the IO manager thread. The IO manager normally just sits in
`WaitForSingleObject` on the `Event` object, so I went looking for ways
that `WaitForSingleObject` could wake up, and then on a hunch I went to
look at the way the `Event` was created... and there it was. I verified
that after the fix I don't see the odd interaction between the two `ghc`
processes any more.
So I'm still not sure why it is that sometimes `WaitForSingleObject` would
get woken up many times in quick succession. I'm only guessing that it
was the `prodding` `IORef` that was leaking, but it certainly looks
suspicious, and there aren't any other obvious candidates. Anyway, I
can't repeat the problem any more.
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/2992#comment:9>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
