Emacs unicode 2 crash suddenly when do some *Replace String (M-%)*.

Emacs unicode 2 crash suddenly when do some *Replace String (M-%)*.

[Hongyi Zhao]

Hi all,

When I do some strings replacement with the latest unicode 2 CVS version 
compilled by me, some times it will crashed suddenly, it seems so 
stange.  Who can give me some hints on this thing?

Thanks in advance.

---
Hongyi Zhao (赵弘毅)
GnuPG DSA: 0xD108493B

老子道德经：第六十一章
大国者下流，天下之交，天下之牝，
牝常以静胜牡，以静为下。

Re: Emacs unicode 2 crash suddenly when do some *Replace String (M-%)*.

[Kenichi Handa]

In article <20070304143820.GA19858@debian-testing-hy.localdomain>, Hongyi Zhao <hongyi.zhao@gmail.com> writes:

> When I do some strings replacement with the latest unicode 2 CVS version 
> compilled by me, some times it will crashed suddenly, it seems so 
> stange.  Who can give me some hints on this thing?

Please find a reproducible test case for that bug.  If you
can't, please keep on using Emacs started from gdb as this:
% cd EMACS_BUILD_TREE/src
% gdb emacs
(gdb) run

and when it craches, show us the output of these gdb commands:

(gdb) bt full
(gdb) xbacktrace

---
Kenichi Handa
handa@m17n.org

Re: Emacs unicode 2 crash suddenly when do some *Replace String (M-%)*.

[Kenichi Handa]

In article <895f446b0703050401i1690839bh2d4dfc1f4bc790f4@mail.gmail.com>, "Hongyi Zhao" <hongyi.zhao@gmail.com> writes:
> I find that if I use the arg *--with-x-toolkit=lucid* or
> *--with-x-toolkit=athen* to compile emacs, the problem will dissapear.  But
> in this case, if the filename include chinese charaters beyond gb2312, the
> filename won't be displayed correctly tho the file's content  can be
> displayed  correctly be emacs.  Will you give me some hints on this thing?

Your report is too vague to give any hint.  Which arg did
you use to build Emacs before?  Exactly which character
can't be displayed?  Where it can't be displayed; in the
dired buffer? in modeline? or somewhere else?  When it can't
be displayed?  In which locale are you running Emacs?  First
of all, please use the command M-x report-emacs-bug to
report a bug.  It automatically inserts important
information in *mail* buffer.

> Furthmore, if I swith to ubuntu, the problem I sent to you in the last mail
> will dissapear.  So  I think that the reason is my Debian SID is screwed up
> by me and I will restore my Debian for retrying.

Don't guess anything, but just give us the currect and
concrete information.  Have you done the gdb session as I
wrote in the previous mail?  If not, why?  Are there any
problem in running Emacs under gdb?

---
Kenichi Handa
handa@m17n.org

Re: Emacs unicode 2 crash suddenly when do some *Replace String (M-%) *.

[Hongyi Zhao]

> Your report is too vague to give any hint.  Which arg did
> you use to build Emacs before? 

That arg is *--with-x-toolkit=gtk*

> Exactly which character
> can't be displayed?  Where it can't be displayed; in the
> dired buffer? in modeline? or somewhere else?  When it can't
> be displayed?  In which locale are you running Emacs?  First
> of all, please use the command M-x report-emacs-bug to
> report a bug.  It automatically inserts important
> information in *mail* buffer.
> 

My locale in UTF-8. That problem appears in the current file's buffer, and I want 
to replace the character *−* into *-* at that moment.

> > Furthmore, if I swith to ubuntu, the problem I sent to you in the last mail
> > will dissapear.  So  I think that the reason is my Debian SID is screwed up
> > by me and I will restore my Debian for retrying.
> 
> Don't guess anything, but just give us the currect and
> concrete information. 

Thank you for your winged words.

> Have you done the gdb session as I
> wrote in the previous mail?  If not, why?  Are there any
> problem in running Emacs under gdb?
> 

Now, I have restore my Debian sid, and recompiled the Emacs unicode 2.  
The problem I reported in the last is disapeared.  So I cann't give 
you any debug informations under gdb.

Best regards.
--- 
Hongsheng Zhao (赵红生)
GnuPG DSA: 0x2D767D70

老子道德经：第六十五章
知此两者亦稽式，常知稽式，是谓玄德。
玄德深矣、远矣，与物反矣，然后乃至大顺。