FIX #1466 (partly), which was causing concprog001(ghci) to fail
authorSimon Marlow <simonmar@microsoft.com>
Tue, 11 Sep 2007 13:02:28 +0000 (13:02 +0000)
committerSimon Marlow <simonmar@microsoft.com>
Tue, 11 Sep 2007 13:02:28 +0000 (13:02 +0000)
commit066f10289f9711a0f6d0669aea97e134f1be2826
treef5b07282067007e3791602c1a471c9889b94e20e
parentdbbfdcaa31e71ec36566ce14640f297ecc699a3f
FIX #1466 (partly), which was causing concprog001(ghci) to fail
An AP_STACK now ensures that there is at least AP_STACK_SPLIM words of
stack headroom available after unpacking the payload.  Continuations
that require more than AP_STACK_SPLIM words of stack must do their own
stack checks instead of aggregating their stack usage into the parent
frame.  I have made this change for the interpreter, but not for
compiled code yet - we should do this in the glorious rewrite of the
code generator.
compiler/ghci/ByteCodeGen.lhs
compiler/main/Constants.lhs
includes/Constants.h
rts/Apply.cmm