Re: [Haskell-cafe] SIGSEGV in yieldCapability ()

On 23/05/2011 09:55, Erik Hesselink wrote:

On Sun, May 22, 2011 at 15:03, Johannes Waldmann wrote:

...

...

I think you should file a bug report with a test case on GHC.

I am willing to work on this, but I thought I'd go fishing for some advice first. My program uses: forkIO, STM, and FFI.

I've seen something like this, using only forkIO, so I'd start looking there. I was trying to create a reduced test case, but haven't reported a bug yet.

...

I think that "heap exhausted" sometimes gets reported as "evacuate: strange closure", (cf. http://hackage.haskell.org/trac/ghc/ticket/5085 ) and yieldCapability() might be another instance.

Thank you, we just had this message, and I had no idea what it was.

Please report all crashes like this as bugs: http://hackage.haskell.org/trac/ghc/wiki/ReportABug In particular, the RTS should never trigger a crash in a heap exhausted situation - if some documentation somewhere gives the wrong impression about that, please let me know. I will need a reproducible test case to track down the bug, though. Cheers, Simon