Hi all,
In spite of all the warnings, I tried to build ghc 5.02.3 from scratch.
The first step,
> ./distrib/hc-build
returned the error
>make[1]: *** No rule to make target `Addr.o', needed by `libHSlang.a'. Stop.
>make: *** [all] Error 1
I'm actually building on a standard Linux x86 platform
Hi Folks,
I've just updated the Building Guide to include a section on porting GHC
to a new platform. Please take a look:
http://www.haskell.org/ghc/docs/latest/building/sec-porting-ghc.html
Disclaimer: I haven't actually followed the steps in doing an
unregisterised bootstrap from beginnin
>> Do you provide in your interactive system a denotation for the last
>> result? (`:?' command does not show it). For example,
> It is called "it"
It is called $$ in Hugs and you can use :set -r flag to change it (you
can even use ':set -rit' if you want. It might be best if ghci and
Hugs we
> Dear GHC,
>
> Do you provide in your interactive system a denotation for
> the last result?
> (`:?' command does not show it). For example,
>
> Prelude> 1
> 1
> Prelude> % + 1
> 2
> ...
It is called "it"
Prelude> 1
1
Prelude> it + 1
2
Prelude> it + 1
3
This follows from other e
Dear GHC,
Do you provide in your interactive system a denotation for
the last result?
(`:?' command does not show it). For example,
Prelude> 1
1
Prelude> % + 1
2
...
-
Serge Mechveliani
[EMAIL PROTECTED]
___
Glasgow-haske
| Okay, I've only just toyed with profiling myself, but one
| observation I've made is that it's not a good idea to mix
| modules compiled with profiling and modules compiled without.
|
| That is, always 'make clean' (i.e. 'rm -f *.o *.hi') before
| you build with a new set of options.
Absol
Hal Daume wrote:
>I'm in a profiling kind of mood [...]
>[...] it seg faults.
>
>Here are some "unusual" things I'm doing; hopefully someone will be able
>to say "oh, there's your problem":
>
> - i'm only compiling the main program with profiling
"Oh, there's your problem"
Ketil Z. Malde wrote: