Ticket #3903 (new bug)

Opened 3 years ago

Last modified 7 months ago

DPH bad sliceP causes RTS panic "allocGroup: requested zero blocks"

Reported by: benl Owned by: benl
Priority: low Milestone: 7.6.2
Component: Data Parallel Haskell Version: 6.13
Keywords: Cc:
Operating System: Unknown/Multiple Architecture: Unknown/Multiple
Type of failure: Compile-time crash Difficulty:
Test Case: Blocked By: #7330
Blocking: Related Tickets:

Description 

$ ghci -XPArr
...
Prelude> :m GHC.PArr
Prelude GHC.PArr> sliceP 10 10 [::]
<interactive>: internal error: allocGroup: requested zero blocks
    (GHC version 6.12.1 for i386_apple_darwin)
    Please report this as a GHC bug:  http://www.haskell.org/ghc/reportabug
Abort trap

sliceP 10 10 [::] is bogus. This should have been picked up in the libraries before hitting the RTS assertion.

Change History

Changed 3 years ago by igloo

  • priority changed from normal to high
  • milestone set to 6.12.3

Changed 3 years ago by igloo

  • milestone changed from 6.12.3 to 6.14.1

Changed 3 years ago by benl

So DPH doesn't do array bounds checks because they interfere with optimisations. I think in the long term we'll want to provide a Safe API as well as an Unsafe one. The idea being that the programmer develops and tests their program with the Safe API, before switching to the Unsafe (and faster) one.

Changed 3 years ago by igloo

  • priority changed from high to normal
  • milestone changed from 6.14.1 to 6.16.1

OK, sounds like there's nothing to do here for 6.14.1 then.

Changed 16 months ago by igloo

  • priority changed from normal to low
  • milestone changed from 7.4.1 to 7.6.1

Changed 8 months ago by igloo

  • milestone changed from 7.6.1 to 7.6.2

Changed 7 months ago by benl

  • blockedby 7330 added
Note: See TracTickets for help on using tickets.