On 23/09/2014, at 8:23 AM, john skaller wrote:
> The new compiler organisation is based on early monomorphisation.
Always some fun problem.
To monomorphise stuff, for Felix things we just replace
the type variables with concrete types.
But what about a primitive, eg a function?
To monomorphis
The new compiler organisation is based on early monomorphisation.
Initially I deliberately wrote the optimiser to work directly with
polymorphic entities. I did this to be absolutely certain that
Felix didn't make the same mistake as C++, and allow unprincipled
dependent name lookup, in other word
///
// fred.cpp
#include
int main()
{
::std::cout << "Hello World of Fred" << ::std::endl;
return 0;
}
///
// tcxx.flx
include "src/tools/toolchain_clang_config";
include "src/tools/toolchain_interface";
var toolchain-maker =
Dynlink::load-pl
I have just added (completely untested so don't use yet!):
// src/tools/toolchain_interface.flx
interface toolchain_t {
whatami : 1 -> string;
host_os : 1 -> string;
target_os : 1 -> string;
cxx_compiler_vendor : 1 -> string;
cxx_static_object_compiler : (dst:string,src: string) -> int
flx --test=build/release --expect --indir=test/regress/rt --time
--usage=prototype '--regex=.*flx'
provides some insight.
Note there's a bug... pkgconfig is returning a massive list of repeated include
files.
--
john skaller
skal...@users.sourceforge.net
http://felix-lang.org
-
Thought I'd have a look:
~/felix>time 'build/release/bin/flxg' -q --inline=500
'--output_dir=/Users/johnskaller/.felix/cache/text'
'--cache_dir=/Users/johnskaller/.felix/cache/binary' '-Ibuild/release/lib'
--syntax=@grammar/grammar.files --import=plat/flx.flxh std web
'/Users/johnskaller/fel
On 30/10/2010, at 2:03 AM, john skaller wrote:
>
> In any case the problem is simple: the function has to use the unbound
> symbol table to print the entry_kind.
and if you look at some of the commented out debug lines:
print_endline ("SUCCESS: overload chooses " ^
full_string_of_e
On 30/10/2010, at 1:17 AM, john skaller wrote:
>
> No, I don't know which bit is failing yet, but it has narrowed it down to
> something
> trivial (TGFT!)
Well now .. I have found the bug. It is here:
let full_string_of_entry_kind bsym_table {base_sym=i; spec_vs=vs; sub_ts=ts} =
let bsym =
On 29/10/2010, at 6:56 PM, john skaller wrote:
[]
Well I said the error had to be in the actual error handling and
indeed that is the case, I finally pinpointed the bug:
| None ->
print_endline "Overload resolution failed ..";
try
clierr sra
(
"[lo
Well I have found the error:
> private fun getcwd: &char * size -> &char = "getcwd($1,$2)";
> fun getcwd():string = {
>var b: array[char,1024];
>var p = getcwd(C_hack::cast[&char] (&b),size 1024);
>return if C_hack::isNULL p then "" else string p endif;
> }
saying:
> SYSTEM FAIL
10 matches
Mail list logo
