6.4.20050215: panic: lookupVers1 MHashTable HT{d}

[Simon Peyton-Jones]

Ah, this one we fixed a few days ago.  Works for me with the head.

Thanks for your well-boiled-down bug reports; they are a lot faster to
fix.

Simon

| -----Original Message-----
| From: glasgow-haskell-bugs-bounces at haskell.org
[mailto:glasgow-haskell-bugs-
| bounces at haskell.org] On Behalf Of Remi Turk
| Sent: 27 February 2005 02:20
| To: glasgow-haskell-bugs at haskell.org
| Subject: 6.4.20050215: panic: lookupVers1 MHashTable HT{d}
| 
| Hi,
| 
| while trying to modify Data.HashTable to support both IO and ST
| without simply copying it, 6.4.20050215 again decided it doesn't
| like me:
| 
| /tmp/test% touch *.hs
| /tmp/test% /var/tmp/ghc/bin/ghc -O -c MHashTable.hs
| /tmp/test% /var/tmp/ghc/bin/ghc -O --make CompatHashTable.hs
| Chasing modules from: CompatHashTable.hs
| Skipping  MHashTable       ( ./MHashTable.hs, ./MHashTable.o )
| Compiling CompatHashTable  ( CompatHashTable.hs, CompatHashTable.o )
| ghc-6.4.20050215: panic! (the `impossible' happened, GHC version
6.4.20050215):
| 	lookupVers1 MHashTable HT{d}
| 
| To trigger it, compilation must actually be performed in two
| seperate steps, and the second one must be done using "--make".
| "-O" is also a vital ingredient for the panic.
| 
| The killing code is attached.
| 
| Happy Hacking,
| Remi
| 
| --
| Nobody can be exactly like me. Even I have trouble doing it.