Pike-devel

pike-devel@lists.lysator.liu.se

3044 discussions

Start a nNew thread

MIN_INT_TYPE
by Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum 27 Feb '03

27 Feb '03

Mast, please commit your changes to global.h so that Pike 7.4 compiles again...

2 1

Re: gcc/icc
by Johan Sundstr�m (a hugging punishment!) ＠ Pike (-) developers forum 27 Feb '03

27 Feb '03

I'm referring to the change to configure.in around lines 1367 - 1371: # Currently only supported with gcc. # # Further tests later on. - with_machine_code="$GCC" + with_machine_code="yes" ])

2 1

Re: gcc/icc
by Mirar ＠ Pike developers forum 27 Feb '03

27 Feb '03

"/home/mirar/pike/src/bignum.c:169: undefined reference to `eval_instruction' ipo_iccbiRPr9vu.o(.text+0x224b):/home/mirar/pike/src/bignum.c:175: undefined reference to `eval_instruction' ipo_iccbiRPr9vu.o(.text+0x2366):/home/mirar/pike/src/bignum.c:185: undefined reference to `eval_instruction' ipo_iccbiRPr9vu.o(.text+0x244e):/home/mirar/pike/src/bignum.c:192: undefined reference to `eval_instruction' ipo_iccbiRPr9vu.o(.text+0x2650):/home/mirar/pike/src/interpret.c:1861: more undefined references to `eval_instruction' follow" Any ideas? ^.^

2 2

Inf
by Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum 27 Feb '03

27 Feb '03

(If you do special handling of infinities, you should do it for NaN as well, so most of that code should be relevant.)

2 2

Inf
by Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum 27 Feb '03

27 Feb '03

Mirar, when you added the tests that infinity is always printed as "inf", it looks like you forgot to commit the fixes to sprintf that actually makes this happen... /pike/home/marcus/Pike/7.5/src/modules/sprintf/testsuite.in:121: Test 93 (shift 0) (CRNL) failed. 1: mixed a() { return sprintf(\"%f\",exp(65536)); } 2: mixed b() { return \"inf\"; } 3: o->a(): "Inf" o->b(): "inf" etc. There is a reason why the tests were just stringp() before.

3 7

Re: 64 bit ints
by David Hedbor ＠ Pike developers forum 26 Feb '03

26 Feb '03

But I'm doubtful that I think it'd be a really good idea anyway since it'd complicate things a lot.

1 0

Floating point (conversion) bug (affected: v7.4 & v7.5; may be Hilfe only)
by Alexander Demenshin 26 Feb '03

26 Feb '03

Hi there, Just playing around: Pike v7.5 release 3 running Hilfe v3.5 (Incremental Pike Frontend) > (float)"0.1e99"; Segmentation fault Same for 7.4 (CVS versions two days old). And another one: Pike v7.5 release 3 running Hilfe v3.5 (Incremental Pike Frontend) > float x; > x = 1e99; Segmentation fault Seems that the problem is related to Hilfe - from file it prints value as expected, no segfaults. Regards, /Al

5 5

Optimizer problem?
by Alexander Demenshin 26 Feb '03

26 Feb '03

Hi again, So one more problem: int main(int argc, array(string) argv) { int count = 0; int ts = 0; ts = time() + 1; while (time() <= ts) /* count++ */; write("%u %u\n", time(), ts); } Output: 1046227356 1046227357 The first number expected to be greater than second. If I insert "count++" into while loop (instead of empty body) everything is OK. This problem doesn't occur when time() is called indirectly. And, whatever it is, optimization of this kind is a bit weird, IMHO... :) Regards, /Al

1 0

Calling Pike functions with variable number of arguments from C
by Marek Habersack 25 Feb '03

25 Feb '03

Is it as simple as pushing all the required parameters to the stack and then calling the Pike function as usual? (i.e. with apply_svalue or apply_low) TIA, marek

2 1

Image.BMP sans RLE?
by Johan Sundstr�m (a hugging punishment!) ＠ Pike (-) developers forum 23 Feb '03

23 Feb '03

> Pike v7.4 release 13 running Hilfe v3.5 (Incremental Pike Frontend) > > Image.BMP.decode(Protocols.HTTP.get_url_data("http://bugzilla.mozilla.org/attachment.cgi?id=115302&action=view")); > Image.BMP.decode: advanced RLE decompression (literal run) is unimplemented (please send this image to mirar(a)roxen.com) Hmm? :-) / Johan Sundström (a hugging punishment!) Previous text: >2003-02-23 09:15: >Subject: Jahopp >-------------------------------------------------------------------- >Då verkar det som om vi ser slutet på Nilssons splash-bild i Mozilla. >En ny bild har plockats fram och den är markerad som "blocker" för >Mozilla 1.3 så den kommer att komma in: > >Bilden (BPM) >http://bugzilla.mozilla.org/attachment.cgi?id=115302&action=view > >Buggen >http://bugzilla.mozilla.org/show_bug.cgi?id=194291 > > / Daniel Bratell (Bara vara) >

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Pike-devel