Pike-devel June 2003

pike-devel@lists.lysator.liu.se

8 participants
29 discussions

Question about speed
by Peta, jo det �r jag ＠ Pike developers forum 12 Jun '03

12 Jun '03

As some of you probably know am I currently adding some ADTs to Pike, among those a sequence. I've been running some speed comparisons between the new sequence and the array regarding indexing and index assignment ( [] and []= ) and found that the array is about twice as fast as the sequence. I found this quite surprising since it use about the same code as the array, it is just a few rows extra, see the code below. Could the speed difference depends on the extra time it takes to find the '[]() lfun for the sequence? The CMOD-Code PIKEFUN mixed `[](mixed index) { simple_array_index_no_free(Pike_sp, THIS->a, index); Pike_sp++; } PIKEFUN mixed `[]=(mixed index, mixed value) { should_copy(); simple_set_index(THIS->a, index, value); } The final C-Code void f_Sequence_cq__backtick_5B_5D(INT32 args) { #line 67 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" struct svalue * index; #line 67 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" if(args != 1) wrong_number_of_args_error("`[]",args,1); #line 67 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" index=Pike_sp+0-1; dmalloc_touch_svalue(Pike_sp+0-1); #line 67 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" { simple_array_index_no_free(Pike_sp, THIS->a, index); Pike_sp++; } void f_Sequence_cq__backtick_5B_5D_eq(INT32 args) { #line 89 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" struct svalue * index; #line 89 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" struct svalue * value; #line 89 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" if(args != 2) wrong_number_of_args_error("`[]=",args,2); #line 89 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" index=Pike_sp+0-2; dmalloc_touch_svalue(Pike_sp+0-2); #line 89 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" value=Pike_sp+1-2; dmalloc_touch_svalue(Pike_sp+1-2); #line 89 "/export/spare/pike/home/peta/Pike/7.5/src/post_modules/_ADT/sequence.c mod" { should_copy(); simple_set_index(THIS->a, index, value); }

3 2

Debugging memory problems
by David Gourdelier 11 Jun '03

11 Jun '03

Hello, We have a memory leak inside Camas (and it exists most propably on IMHO as well). I've tried to debug in order to find where the leak is (I used Pike 7.2). We know the leak is inside the session object, it contains all sort of variables including other objects, mapping, string and so on (people interested in the source can get it here <http://cvsweb.caudiumforge.net/cgi-bin/cvsweb.cgi/camas/server/modules/cama…>). I've tried to force gc() every x minutes and when the user logout but it doesn't change the problem. I also added this in the destroy function of the session: void destroy_object(object o) { if(o && objectp(o)) destruct(o); } void destroy() { foreach(indices(this_object()), string var) { if(this_object()[var] && !intp(this_object()[var])) { int refs = _refs(this_object()[var]); if(refs > 2) write("variable %s is referenced %d times\n", var, refs); } } destroy_object(lock); destroy_object(imapclient); destroy_object(cmail); destroy_object(prefsmailobj); destroy_object(extendedabook); destroy_object(ebookmailobj); destruct(this_object()); } Which shows me lot of variables are referenced several times and for some of then, more than 7000 times. I also noted that the variables are usually referenced by similar number of other variables, for example I have this output: variable draftmail is referenced 1107 times variable misspelled is referenced 7067 times variable spelling is referenced 7067 times variable deletemails is referenced 7067 times variable deletemboxes is referenced 7067 times variable prefuids is referenced 7067 times variable searchstring is referenced 1107 times variable session is referenced 1107 times I also saw the destroy or dump_me functions are referenced 8 times while they are not used by other things in the program. So my first questions are: . How can I have so much objects referenced ? Is it because of some sort of reference loop ? Is _refs() working good on Pike 7.2 ? . What does a reference for a function mean ? . Is it possible to know which things reference which variables ? . What is the effect of adding a destroy() inside an object ? I also enable debug in gc.c to find the problem but the output looks quite cryptic to me. At the beginning of the destroy I have this: Garbage collecting ... | check: 33151 references checked, counted 489 weak refs | mark: 15799 markers referenced, 0 weak references freed, | 0 things really freed, got 0 tricky weak refs Then lot of enqueue 0x89223ec [(nil)]: back=(nil), prev=0x81d1990, next=(nil), data=0x850cd2 4, cycle=0, flags=0x01 gc_cycle_push, recurse 0x850cd24 [(nil)] marker at 0x89495ec: flags =0x6446, refs=1, weak=-1, xrefs=0, saved=0, frame=0x89223ec [back=(nil), prev=0x 81d1990, next=(nil), data=0x850cd24, cycle=0, flags=0x01] ... dequeue 0x89222b4 [0x89222cc]: back=0x89222cc, prev=0x89223ec, next=(nil), data= 0x88ade1c, cycle=0, flags=0x01 gc_cycle_pop, pop off 0x88ade1c [0x8606e0c] marker at 0x898bb2c: flags=0x040e, refs=1, weak=0, xrefs=0, saved=1, frame=0x89222b4 [back=0xfffffff f, prev=0x89223ec, next=(nil), data=0x88ade1c, cycle=0, flags=0x09] .... gc_cycle_push, no recurse 0x88ad8ac [(nil)] marker at 0x89986b4: flags =0x040e, refs=1, weak=0, xrefs=0, saved=1, frame=(nil) and finally: | cycle: 554 internal things visited, 14 cycle ids used, | 0 weak references freed, 0 things really freed, | space for 48 gc frames used | free: 26 really freed, 528 left with live references | Killing 0x850e4c4 with 25 refs .... | Killing 0x8606e0c with 25 refs | kill: 23 objects killed, 528 things really freed | destruct: 0 things really freed done (freed 554 of 14900 objects) . Can someone explained a little the output if it worth it ? Especially I would like to know what is a weak reference ? Why some references are weaker than others ? what does object killed mean ? I also want to know if _next() and _prev() can be of any help to me and if yes how can I use it ? Thank you for your time. -- David Gourdelier

4 18

Post-padding overwritten
by David Gourdelier 09 Jun '03

09 Jun '03

Hi, I've compiled Pike with --with-dmalloc --with-dmalloc-malloc-leaks --with-checker --with-rtldebug and when I load Caudium I get the following error when it loads one of the Camas modules: Loading camas_tags... Post-padding overwritten for block at 0xac7bda0 (size 20)! **Location: 0xac7bda0 Type: unknown Refs: 0 gc generation: 11/11 gc pass: 0/0 *** peep.c:187 (1 times) *** pike_memory.c:570 (1 times) **Cannot describe block of unknown type 247 ******************* Then the pike process exit. The previous modules loads without warning. Can you tell me what this mean ? -- David Gourdelier

5 6

dumpmaster
by Martin Nilsson (lambda) ＠ Pike (-) developers forum 06 Jun '03

06 Jun '03

Why is the program that dumps the master so bare bones?

2 1

configure problem with Freebsd
by David Gourdelier 03 Jun '03

03 Jun '03

Hi, I am having problems with Pike 7.4 on Freebsd 4.7. The problem is that it sources /etc/make.conf but make.conf is not really sourceable (at least on my system). It contains things like BDECFLAGS= -W -Wall or CPUTYPE=i386 and source /etc/make.conf reports errors in this case (morever i don't remember modifying make.conf). Another thing is: /configure --disable-make.conf configure: error: make.conf: invalid feature name -- David Gourdelier

3 3

New encoder
by Martin Nilsson (lambda) ＠ Pike (-) developers forum 03 Jun '03

03 Jun '03

It appears that the new encoder hangs most systems running "make verify" during precompilation of the test pike. That is bad, since there is no watchdog running during the installation phase.

3 4

Re: Pike Cookbook
by Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum 01 Jun '03

01 Jun '03

In case this wasn't 100% clear from the original post, I meant include the actual stat() call in the solution as well. It doesn't make it much longer, just much better. :-)

1 0

Pike Cookbook
by Martin Nilsson (har bott i google) ＠ Pike (-) developers forum 01 Jun '03

01 Jun '03

Johan Sundström and I are about to make some attempts to make a Pike Cookbook. Regardless if we have the strength to make it all the way into a book or not, I think this is a good project to spend some time on. While we can get some ideas what this cookbook should contain by looking at e.g. the Perl and Python cookbooks, much of what is in them deals with how to solve problems that are inherit in the langauges themselves. As a consequence we would like to know what problems you have with Pike and how you solve them (if you have). Johan and I will set up some sort of database or repository and collect the responses as well as inserting our own stuff. We will probably aim at getting the same basic layout as in the present cookbooks in the database, e.g.: ----------------------8<-------------------- Title: Hello World Contributor: Martin Nilsson Problem: Make a program that prints Hello World! and exits. Solution: int main() { write("Hello World!\n"); } Discussion: The above code is probably the smallest possible program, counting tokens, that solves the problem. Without unessecary white spaces this is 36 characters. A more "proper" solution would however be: int main(int argc, array(string) args, array(strings) env) { write("Hello World!\n"); return 0; } Here all the passed arguments are received in fairly typed variables, which you don't have to do. I say fairly typed, because there is always at least one argument passed to the program, the path of the program itself, so argc could be declared as int(1..). After the interesting code has been executed 0 is returned, as result code of the program. There is an implicit return 0; at the end of every function, which we relied on in the first code. Since the return value is always 0, the main function could be declared to return int(0..0). Since the Pike master only feeds positive return values from main into an exit call, return 0; can be replaced with exit(0);. While this doesn't really matter here it does matter if we chose another aproach for this program all together. When Pike executes a script it first loads the file and compiles it into a program. Then it instansiate an object from the program and call its main function. Thus we can put our code in the object constructor and when finished exit Pike, so it doesn't find out that the main function is missing. void create() { write("Hello World!\n"); exit(0); } ----------------------8<-------------------- So please give us your thoughts, problems and solutions, either here or as private messages.

8 26

'-g' is forbidden in paths...
by David Hedbor ＠ Pike developers forum 01 Jun '03

01 Jun '03

This brilliant code: CFLAGS="`echo $(CFLAGS) | sed -e 's/-g//g' -e 's/-W[^ ]*//g'`" \ In the GL and GTK Makefile.in's don't do what you'd want them to do when your directory is named, for example: /mnt/disk3/src/pike/7.4/build/linux-2.4.20-gaming-r3-i686 What are the reasons for the forced removal of -g and other stuff anyway?

3 3

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 June 2003