f8l has quit [Remote host closed the connection]
f8l has joined #picolisp
freemint has quit [Ping timeout: 252 seconds]
freemint has joined #picolisp
stultulo has joined #picolisp
stultulo is now known as oldf8l
oldf8l is now known as stultulo
f8l has quit [Ping timeout: 240 seconds]
stultulo is now known as f8l
xkapastel has quit [Quit: Connection closed for inactivity]
karswell has quit [Ping timeout: 265 seconds]
orivej has joined #picolisp
rob_w has joined #picolisp
ubLIX has joined #picolisp
cil_z has joined #picolisp
ubLIX has quit [Quit: ubLIX]
<
Regenaxer>
Hi beneroth
cil_z has quit [Ping timeout: 252 seconds]
mtsd has joined #picolisp
mtsd has quit [Quit: Leaving]
lodsw_ has quit [Ping timeout: 240 seconds]
lodsw has joined #picolisp
ubLIX has joined #picolisp
ubLIX has quit [Quit: ubLIX]
rob_w has quit [Quit: Leaving]
<
tankf33der>
Regenaxer: when pil21 will be in public?
<
tankf33der>
pil21.tgz
<
Regenaxer>
Good question
<
Regenaxer>
Any time
<
Regenaxer>
You want to test? :)
<
Regenaxer>
I have only fragments so far
<
tankf33der>
of course i want, everybody wants
<
Regenaxer>
haha, everybody of 3 people ;)
<
tankf33der>
if ok i will ask mansur setup github repo for it
<
tankf33der>
yea, yoking
<
Regenaxer>
Repo is too early, everything changes all the time
<
Regenaxer>
ok, I will prepare later
<
Regenaxer>
at the moment I'm walking in the fields
<
tankf33der>
mansur’s script can easily handle add, del, changes inside tgz. Like all this year for picoLisp.tgz
<
tankf33der>
for git and mercurial
<
Regenaxer>
yes, perfect
<
tankf33der>
all you need rollout file with the same name.
<
Regenaxer>
This morning I found a bug, will fix first
<
beneroth>
what is it?
<
tankf33der>
beneroth: picolisp on llvm
<
Regenaxer>
(The bug I mentioned is not found yet - no time)
<
Regenaxer>
To build it, you must install make, llvm and clang
ubLIX has joined #picolisp
<
tankf33der>
compiled. playing.
<
Regenaxer>
only a handful of functions, car, cdr etc
<
Regenaxer>
see src/glob.l
<
Regenaxer>
in 'symTab'
<
tankf33der>
i definitely wants try this on SPARC tomorrow
<
Regenaxer>
I hope it is very portable
<
Regenaxer>
The 'native' issue is still unclear
<
tankf33der>
: (dec 0)
<
tankf33der>
-> 1152921504606846975
<
Regenaxer>
yes, zero check is missing
<
Regenaxer>
And after 'inc' or 'dec' the REPL behaves strange
<
tankf33der>
(cdr T)
<
Regenaxer>
Normally it should exit on NIL or Return
<
tankf33der>
my llvm 8.0.1, BTW
ubLIX has quit [Quit: ubLIX]
<
Regenaxer>
No bignums, no long symbol names
<
Regenaxer>
Nothing ..
<
tankf33der>
huge progress.
<
Regenaxer>
Evaluation is also only partially implemented
<
Regenaxer>
Only tests
<
Regenaxer>
What
*is* mainly finished is the compiler
<
tankf33der>
internal language picolisp, sweet.
<
Regenaxer>
ie 99% percent of work went into src/lib/llvm.l
<
Regenaxer>
A subset of pil
<
Regenaxer>
or, a very similar lang
<
tankf33der>
i read it like picolisp.
<
Regenaxer>
yeah, but lots of differences
<
Regenaxer>
It is a kind of assembler but with pil syntax
<
tankf33der>
you use (symbols)
<
Regenaxer>
redefined functions to compile themselves
<
tankf33der>
im thinking how i will compile on sparc then
<
Regenaxer>
yes, that will be interesting
<
Regenaxer>
if it is really portable at this stage
<
Regenaxer>
Should even work on 32 bit machines
<
Regenaxer>
Needs perhaps some more alignments then
<
Regenaxer>
Also no optimizations yet
<
Regenaxer>
found the bug
<
Regenaxer>
was in 'save' in llvm.l
<
Regenaxer>
released
<
tankf33der>
doing 32bit
<
tankf33der>
bootstraping
<
Regenaxer>
the release contains main.bc, a pre-built binary
<
Regenaxer>
So bootsrapping does no longer require a running picolisp
<
tankf33der>
segmentation fault on 32bit
<
tankf33der>
i manually did all steps.
<
Regenaxer>
Perhaps alignment missing
<
Regenaxer>
e.g. allocated memory is not aligned to 16 bits
<
tankf33der>
as i see Makefile doesnt respect main.bc and trying rewrite it
<
Regenaxer>
only if sources are newer
<
Regenaxer>
hmm, right with main.bc
<
Regenaxer>
Not sure what I should include into the release
<
Regenaxer>
picolisp.bc? But I think it is already system-dependent
<
Regenaxer>
yes, it has lib.bc
<
Regenaxer>
So Makefile must be changed to respect main.bc if it is there
<
Regenaxer>
Trying ...
andyjpb has joined #picolisp
<
Regenaxer>
Did it, and released again
<
Regenaxer>
Now Makefile should preserve main.bc
<
Regenaxer>
So build should also work without existing pil installation
<
tankf33der>
i will play tomorrow
<
Regenaxer>
Thanks!
ubLIX has joined #picolisp
<
tankf33der>
backtrace where crash on 32bit
<
Regenaxer>
hmm, pastebin hangs again
<
Regenaxer>
Perhaps I'm wrong that it works on 32 bit macxines
<
Regenaxer>
We still have the assumption that pointers are 64 bit
<
Regenaxer>
Not sure what LLVM does
orivej has quit [Ping timeout: 240 seconds]
<
Regenaxer>
Let's keep an eye on that
<
Regenaxer>
What is the contents of esi in movl -8(%esi) ... ?
<
tankf33der>
sparc failed also
<
tankf33der>
(gdb) p $esi
<
tankf33der>
$1 = 4747272
<
Regenaxer>
"value 0x400414 does not fit"
<
Regenaxer>
not symbol alignment
<
Regenaxer>
needs 8 at the end
<
tankf33der>
as is.
<
Regenaxer>
: (hex 4747272)
<
Regenaxer>
-> "487008"
<
Regenaxer>
is all right
<
Regenaxer>
May or may not work. Needs more debugging to find the real reason
<
Regenaxer>
cells must be 16 bytes, alignment etc.
<
tankf33der>
its ok, lets leave 32bit alone.
<
Regenaxer>
So the question is what exactly is wrong
<
Regenaxer>
pil21 uses i64 for such types
<
Regenaxer>
so it
*may* work in principle
<
tankf33der>
sleep.
<
tankf33der>
see you.
<
Regenaxer>
see you! :)
ubLIX has quit [Quit: ubLIX]
ubLIX has joined #picolisp
orivej has joined #picolisp