It is easier if you send actual darcs patches if possible, then they will automatically commute with my changes and merge without issue. darcs patching algorithms are much more advanced than gits in this regard. to do this easily from your git patch files just do: cd to a jhc repo pulled from my tree patch -p1 < yourfile.patch darcs record -a darcs send then it will email me and the list the patch, plus it will be properly attributed to you, and it will merge even if I modified the same files or even renamed them. John On Sat, Mar 16, 2013 at 5:26 AM, John Meacham wrote:

Thanks! I'll patch them into my repository.

John

On Sat, Mar 16, 2013 at 2:09 AM, Kiwamu Okabe wrote:

...

Hi John.

I think the patch is useful to run Haskell code on tiny CPU and small memory size. Github link is below.

https://github.com/ajhc/ajhc/commit/3167551530b0576cf1f42f928865868ce9aa0b50

The changes are...

1. Control BLOCK_SIZE and MEGABLOCK_SIZE from Makefile. 2. Do not allocate MEGABLOCK dynamically. 3. Set function pointer's bit0 for Cortex-M3. http://communities.mentor.com/community/cs/archives/arm-gnu/msg01904.html 4. Naive GC flag to run GC when no blocks. 5. Control stack_grow size.

Example of usage is found at https://github.com/ajhc/demo-cortex-m3.

Best regards, -- Kiwamu Okabe

_______________________________________________ jhc mailing list jhc@haskell.org http://www.haskell.org/mailman/listinfo/jhc