Re: [Ecls-list] Frequent stackoverflow errors

2011-03-06 Thread Juan Jose Garcia-Ripoll
On Mon, Feb 28, 2011 at 11:52 AM, Marko Kocić marko.ko...@gmail.com wrote:

  I am a bit puzzled. Is this very recent?

 Not that recent. It happens from time to time. At first only in SLIME,
 but now it fails frequently on the command line too.
 Reliable way to trigger it is to try to quickload, for example
 closure-html, but it happens also on other libraries.


I tried to check whether closure-html really produces infinite loops in the
type propagator but I am unable to even start building it. It seems they
have some bogus pathnames there, because I get (with the new error rmessage
in CVS) something like

Condition of type: SIMPLE-ERROR
Pathname without a physical namestring:
:HOST NIL
:DEVICE NIL
:DIRECTORY (ABSOLUTE Users jjgarcia quicklisp dists quicklisp software
closure-html-20101006-git resources dtd)
:NAME Entities-Latin1
:TYPE UNSPECIFIC
:VERSION NEWEST

Could you verify? It is unsupported to have :unspecific components in
pathnames.

-- 
Instituto de Física Fundamental, CSIC
c/ Serrano, 113b, Madrid 28006 (Spain)
http://juanjose.garciaripoll.googlepages.com
--
What You Don't Know About Data Connectivity CAN Hurt You
This paper provides an overview of data connectivity, details
its effect on application quality, and explores various alternative
solutions. http://p.sf.net/sfu/progress-d2d___
Ecls-list mailing list
Ecls-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ecls-list


Re: [Ecls-list] Frequent stackoverflow errors

2011-03-02 Thread Juan Jose Garcia-Ripoll
On Mon, Feb 28, 2011 at 11:52 AM, Marko Kocić marko.ko...@gmail.com wrote:

  I am a bit puzzled. Is this very recent?

 Not that recent. It happens from time to time. At first only in SLIME,
 but now it fails frequently on the command line too.
 Reliable way to trigger it is to try to quickload, for example
 closure-html, but it happens also on other libraries.


Thanks for the hint. It seems the problem is in the type-propagation code --
I should set up a limit on recursion and abort the propagation when it
exceeds it. It is very useful that you managed to find a reproducible case.

Juanjo

-- 
Instituto de Física Fundamental, CSIC
c/ Serrano, 113b, Madrid 28006 (Spain)
http://juanjose.garciaripoll.googlepages.com
--
Free Software Download: Index, Search  Analyze Logs and other IT data in 
Real-Time with Splunk. Collect, index and harness all the fast moving IT data 
generated by your applications, servers and devices whether physical, virtual
or in the cloud. Deliver compliance at lower cost and gain new business 
insights. http://p.sf.net/sfu/splunk-dev2dev ___
Ecls-list mailing list
Ecls-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ecls-list


Re: [Ecls-list] Frequent stackoverflow errors

2011-02-28 Thread Marko Kocić
 I am a bit puzzled. Is this very recent?

Not that recent. It happens from time to time. At first only in SLIME,
but now it fails frequently on the command line too.
Reliable way to trigger it is to try to quickload, for example
closure-html, but it happens also on other libraries.

Here'r the gdb session:
c:\opt\eclgdb ecl.exe
GNU gdb (GDB) 7.2
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later http://gnu.org/licenses/gpl.html
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type show copying
and show warranty for details.
This GDB was configured as mingw32.
For bug reporting instructions, please see:
http://www.gnu.org/software/gdb/bugs/...
Reading symbols from c:\opt\ecl/ecl.exe...(no debugging symbols found)...done.
(gdb) break ecl_cs_overflow
Function ecl_cs_overflow not defined.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 1 (ecl_cs_overflow) pending.
(gdb) run
Starting program: c:\opt\ecl/ecl.exe
[New Thread 220.0xc20]
;;; Loading #Pc:/opt/ecl/asdf.fas
;;; Loading #Pc:/opt/ecl/cmp.fas
;;; Loading C:/Users/mkocic/quicklisp/setup.lisp
ECL (Embeddable Common-Lisp) 11.1.1
Copyright (C) 1984 Taiichi Yuasa and Masami Hagiya
Copyright (C) 1993 Giuseppe Attardi
Copyright (C) 2000 Juan J. Garcia-Ripoll
ECL is free software, and you are welcome to redistribute it
under certain conditions; see file 'Copyright' for details.
Type :h for Help.
Top level in: #process TOP-LEVEL.
 (ql:quickload :closure-html)
To load closure-html:
  Load 1 ASDF system:
closure-html
; Loading closure-html
;;; Checking for wide character support... WARNING: Lisp
implementation doesn't use UTF-16, but accepts surrogate code p
oints.
 yes, using code points.
;;; Building Closure with CHARACTER RUNES
126 states..
; Table ready (total of forty-eight rules -- eighty-three states).
; Dumping: 
...
..
;;; Note:
;;;   Refusing to propagate #form C2PRINC 256e378
Breakpoint 1, 0x6e8e800b in ecl_cs_overflow () from c:\opt\ecl\ecl.dll
(gdb) bt
#0  0x6e8e800b in ecl_cs_overflow () from c:\opt\ecl\ecl.dll
#1  0x66ff0eaa in L2p1trivial () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#2  0x6e943893 in APPLY () from c:\opt\ecl\ecl.dll
#3  0x6e926887 in ecl_apply_from_stack_frame () from c:\opt\ecl\ecl.dll
#4  0x6e926d17 in cl_apply () from c:\opt\ecl\ecl.dll
#5  0x66ff0c78 in L1p1propagate () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#6  0x66ff19da in L11p1call_local () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#7  0x6e9438d0 in APPLY () from c:\opt\ecl\ecl.dll
#8  0x6e926887 in ecl_apply_from_stack_frame () from c:\opt\ecl\ecl.dll
#9  0x6e926d17 in cl_apply () from c:\opt\ecl\ecl.dll
#10 0x66ff0c78 in L1p1propagate () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#11 0x66ff12f9 in L5p1propagate_list () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#12 0x66ff2c89 in L24p1progn () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#13 0x6e94f9f2 in APPLY_fixed () from c:\opt\ecl\ecl.dll
#14 0x6e926866 in ecl_apply_from_stack_frame () from c:\opt\ecl\ecl.dll
#15 0x6e926d17 in cl_apply () from c:\opt\ecl\ecl.dll
#16 0x66ff0c78 in L1p1propagate () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#17 0x66ff1cc9 in L14p1if () from C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#18 0x6e94fa6a in APPLY_fixed () from c:\opt\ecl\ecl.dll
#19 0x6e926866 in ecl_apply_from_stack_frame () from c:\opt\ecl\ecl.dll
#20 0x6e926d17 in cl_apply () from c:\opt\ecl\ecl.dll
#21 0x66ff0c78 in L1p1propagate () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#22 0x66ff1d2c in L14p1if () from C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#23 0x6e94fa6a in APPLY_fixed () from c:\opt\ecl\ecl.dll
#24 0x6e926866 in ecl_apply_from_stack_frame () from c:\opt\ecl\ecl.dll
#25 0x6e926d17 in cl_apply () from c:\opt\ecl\ecl.dll
#26 0x66ff0c78 in L1p1propagate () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#27 0x66ff1d2c in L14p1if () from C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#28 0x6e94fa6a in APPLY_fixed () from c:\opt\ecl\ecl.dll
#29 0x6e926866 in ecl_apply_from_stack_frame () from c:\opt\ecl\ecl.dll
#30 0x6e926d17 in cl_apply () from c:\opt\ecl\ecl.dll
#31 0x66ff0c78 in L1p1propagate () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#32 0x66ff1d2c in L14p1if () from C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#33 0x6e94fa6a in APPLY_fixed () from c:\opt\ecl\ecl.dll
#34 0x6e926866 in ecl_apply_from_stack_frame () from c:\opt\ecl\ecl.dll
#35 0x6e926d17 in cl_apply () from c:\opt\ecl\ecl.dll
#36 0x66ff0c78 in L1p1propagate () from
C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#37 0x66ff1d2c in L14p1if () from C:\Users\mkocic\AppData\Local\Temp\eD7BA.tmp
#38 0x6e94fa6a in APPLY_fixed () from c:\opt\ecl\ecl.dll
#39 0x6e926866 in ecl_apply_from_stack_frame () from c:\opt\ecl\ecl.dll
#40 0x6e926d17 in cl_apply () from 

Re: [Ecls-list] Frequent stackoverflow errors

2011-02-25 Thread Anton Vodonosov
I had this error too! (I thought it's the result of the small changes in ECL 
I made. So it's a relief to hear it wasn't me).

I also don't know how to reproduce it reliably. I saw there is only two 
places in the ECL sources where this error message is printed.

If I have more info, I'll post it to the list.

Best regards,
- Anton

25.02.2011, 13:44, Marko Kocić marko.ko...@gmail.com:
 Hi all,

 I noticed that lately ECL frequently fails with stackoverflow errors,
 especially when compiling/loading packages.
 I can't reproduce it deterministically, but most often it happends
 when I try to quickload or asdfload some package from SLIME.

 However, it also happens from the command line. Right now I'm trying to do:
 (asdf:oos 'asdf:load-fasl-op :lispbuilder-sdl)
 and I get the following message:

 ;;;
 ;;; Stack overflow.
 ;;; Jumping to the outermost toplevel prompt
 ;;;

 However, using (asdf:oos 'asdf:load-fasl-op :lispbuilder-sdl) I was
 abe to compile/load it from commandline, but not from SLIME.

  *features*

 (:QUICKLISP :WSOCK :CLIM-GTKAIRO :DRAKMA-NO-SSL :HUNCHENTOOT-NO-SSL :ASDF2
  :ASDF :ASDF-WINDOWS :X86 :ECL-BYTECMP :FORMATTER :LONG-LONG :UINT64-T
  :UINT32-T :UINT16-T :RELATIVE-PACKAGE-NAMES :LONG-FLOAT :UNICODE :DFFI
  :CLOS-STREAMS :CMU-FORMAT :WINDOWS :WIN32 :MINGW32 :ECL-PDE :DLOPEN :CLOS
  :THREADS :BOEHM-GC :ANSI-CL :COMMON-LISP :IEEE-FLOATING-POINT :PREFIXED-API
  :FFI :PENTIUM3 :COMMON :ECL)

 ECL is built on Win7x64 using 32 bit mingw with unicode and threads enabled.

 Is there a way I can help to debug this stack overflow problem to find
 a root cause?

 Regards,
 Marko Kocić

 --
 Free Software Download: Index, Search  Analyze Logs and other IT data in
 Real-Time with Splunk. Collect, index and harness all the fast moving IT data
 generated by your applications, servers and devices whether physical, virtual
 or in the cloud. Deliver compliance at lower cost and gain new business
 insights. http://p.sf.net/sfu/splunk-dev2dev
 ___
 Ecls-list mailing list
 Ecls-list@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/ecls-list

--
Free Software Download: Index, Search  Analyze Logs and other IT data in 
Real-Time with Splunk. Collect, index and harness all the fast moving IT data 
generated by your applications, servers and devices whether physical, virtual
or in the cloud. Deliver compliance at lower cost and gain new business 
insights. http://p.sf.net/sfu/splunk-dev2dev 
___
Ecls-list mailing list
Ecls-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ecls-list