GNU bug report logs -
#33548
25.1; GUD GDB doesn't decode file paths correctly
Previous Next
Reported by: ksqsf <i <at> ksqsf.moe>
Date: Thu, 29 Nov 2018 17:34:01 UTC
Severity: normal
Found in version 25.1
Done: Eli Zaretskii <eliz <at> gnu.org>
Bug is archived. No further changes may be made.
Full log
View this message in rfc822 format
[Message part 1 (text/plain, inline)]
Your message dated Fri, 30 Nov 2018 09:39:31 +0200
with message-id <83k1kvc8uk.fsf <at> gnu.org>
and subject line Re: bug#33548: 25.1; GUD GDB doesn't decode file paths correctly
has caused the debbugs.gnu.org bug report #33548,
regarding 25.1; GUD GDB doesn't decode file paths correctly
to be marked as done.
(If you believe you have received this mail in error, please contact
help-debbugs <at> gnu.org.)
--
33548: http://debbugs.gnu.org/cgi/bugreport.cgi?bug=33548
GNU Bug Tracking System
Contact help-debbugs <at> gnu.org with problems
[Message part 2 (message/rfc822, inline)]
Dear maintainers,
It seems that GDB MI doesn't decode file paths correctly, which can
result in some really weird unexpected behaviors.
For example, it attempts to find file
/home/ksqsf/我的坚果云/"/home/ksqsf/\346\210\221\347\232\204\345\235\232\346
\236\234\344\272\221/code.cpp"
I traced down to the source and found that gdb-get-source-file, gdb-
get-source-file-list, and gdb-get-location don't decode file paths
using a correct coding system. (In my case, UTF-8.) Instead, they
simply use 'read'.
Hope this can be fixed soon!
Regards,
_rika
[Message part 3 (message/rfc822, inline)]
> From: ksqsf <i <at> ksqsf.moe>
> Cc: 33548 <at> debbugs.gnu.org
> Date: Fri, 30 Nov 2018 10:03:19 +0800
> Feedback-ID: bizesmtp:ksqsf.moe:qybgforeign:qybgforeign4
>
> 在 2018-11-29四的 21:35 +0200,Eli Zaretskii写道:
> > Does it help to set gdb-mi-decode-strings to utf-8?
>
> Yes! This fixes this problem for me. Thanks. This report can be closed
> now.
Done.
> I think this variable deserves a mention in the doc anyway.
What documentation did you have in mind?
> Actually it should be made default.
No, I don't think so. It has some confusing side effects, and in most
cases is unnecessary.
> I noticed the warning, but can we do better? I mean, simply stops gdb
> from escaping strings (does gdb even support this?), as _inconsistent_
> escaping does nothing but introducing problems.
You can tell GDB to stop escaping in strings ("set print
sevenbit-strings", which is off by default), but GDB is under user
control, and the user is legitimately allowed to turn escaping on, we
cannot do anything about that. Plus, there are GDB settings to
control the encoding of text strings that GDB emits based on values of
the variables of the program being debugged.
> LLDB seems to do better job in this regard.
??? LLDB doesn't have a respectable implementation of the MI protocol
to begin with.
This bug report was last modified 6 years and 179 days ago.
Previous Next
GNU bug tracking system
Copyright (C) 1999 Darren O. Benham,
1997,2003 nCipher Corporation Ltd,
1994-97 Ian Jackson.