On Sat, Jul 31, 2010 at 12:13:15PM -0300, Felipe Lessa wrote:

On Sat, Jul 31, 2010 at 11:14 AM, Lemmih wrote:

...

On Sat, Jul 31, 2010 at 1:16 PM, John Meacham wrote:

...

Announcing jhc 0.7.6! This release fixes a large number of bugs that cropped up when compiling hoskell out in the wild. In addition it has some major internal cleanups and better support for cross-compilation. This is a minor release in that I am just announcing it on jhc@haskell.org.

Benchmark updated with new results: http://mirror.seize.it/report.html

On "ansi" and on "digits-of-e1" the JHC-0.7.6-gc compiler is segfaulting gcc :).

Yes, I have noticed this a couple times, I have not taken the time to track down the root interaction that is crashing gcc, but it seems to be something brittle, as slightly changing the program or altering the optimization flags will cause it to go away. Hmm.. perhaps I need to do some exhaustive search for optimal gcc flags.. John -- John Meacham - ⑆repetae.net⑆john⑈ - http://notanumber.net/

On Sun, Aug 1, 2010 at 9:16 AM, John Meacham wrote:

On Sat, Jul 31, 2010 at 04:14:00PM +0200, Lemmih wrote:

...

On Sat, Jul 31, 2010 at 1:16 PM, John Meacham wrote:

...

Announcing jhc 0.7.6! This release fixes a large number of bugs that cropped up when compiling hoskell out in the wild. In addition it has some major internal cleanups and better support for cross-compilation. This is a minor release in that I am just announcing it on jhc@haskell.org.

Benchmark updated with new results: http://mirror.seize.it/report.html

Cool. I didn't expect too much change here, but it is good to see at least a little progress. Any luck on getting the RTS to limit ram use appropriately?

I've only been successful at substituting OOM errors with timeout errors so far.

I have a couple ideas for low hanging fruit to help the speed/resource usage of the compiler I'll try to work into an upcoming release. Along with some better diagnostics output of memory usage for you.

Excellent. -- Cheers,   Lemmih