public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
From: Tom de Vries <tdevries@suse.de>
To: gdb-patches@sourceware.org
Cc: Tom Tromey <tom@tromey.com>
Subject: Re: [RFC 3/3] [gdb/dap] Ignore OSError on stream.flush in JSON writer
Date: Wed, 7 Feb 2024 11:29:02 +0100	[thread overview]
Message-ID: <6a588b22-2d3f-4a0d-8a84-b66b68ff9041@suse.de> (raw)
In-Reply-To: <20240207090224.27521-4-tdevries@suse.de>

On 2/7/24 10:02, Tom de Vries wrote:
> Due to the previous two fixes, I can now run test-case gdb.dap/eof.exp without
> coredumps happening.
> 
> However, I do run into:
> ...
> FAIL: gdb.dap/eof.exp: exceptions in log file
> ...
> due to:
> ...
> JSON writer: caught exception: err=OSError(5, 'Input/output error'),
>    type(err)=<class 'OSError'>
> Traceback (most recent call last):
>    File "python/gdb/dap/startup.py", line 72, in thread_wrapper
>      target(*args)
>    File "python/gdb/dap/io.py", line 80, in _json_writer
>      stream.flush()
> OSError: [Errno 5] Input/output error
> ...
> 
> I can image that the stream.flush would throw an OSError if it races with the
> gdb main thread handling a SIGHUP and finalizing python in the process.
> 
> Handle this is in the simplest way I can think of: ignore the OSError, and
> terminate the DAP thread.

I just found that the JSON reader thread does:
...
     except OSError:
         # Reading can also possibly throw an exception.  Treat this as 

         # EOF. 

         log_stack(LogLevel.FULL)
         return None
...

If we however threat this is in the JSON writer thread the same way, we 
get a FAIL from dap_check_log_file, because it checks for ^Traceback, 
which is generated by the log_stack.

Thanks,
- Tom


      reply	other threads:[~2024-02-07 10:29 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-07  9:02 [RFC 0/3] [gdb/dap] Fix issues triggered by gdb.dap/eof.exp Tom de Vries
2024-02-07  9:02 ` [RFC 1/3] [gdb/dap] Fix exit race Tom de Vries
2024-02-07 16:01   ` Tom Tromey
2024-02-13 15:04     ` Tom de Vries
2024-02-13 18:04       ` Tom Tromey
2024-02-13 18:11         ` Tom Tromey
2024-02-14 15:31         ` Tom de Vries
2024-02-14 15:34           ` Tom Tromey
2024-02-14 15:53             ` Tom de Vries
2024-02-14 16:18               ` Tom Tromey
2024-02-14 17:16                 ` Tom de Vries
2024-02-07  9:02 ` [RFC 2/3] [gdb/dap] Catch and log exceptions in dap threads Tom de Vries
2024-02-07 15:52   ` Tom Tromey
2024-02-12 15:15     ` Tom de Vries
2024-02-12 17:35       ` Tom Tromey
2024-02-07  9:02 ` [RFC 3/3] [gdb/dap] Ignore OSError on stream.flush in JSON writer Tom de Vries
2024-02-07 10:29   ` Tom de Vries [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=6a588b22-2d3f-4a0d-8a84-b66b68ff9041@suse.de \
    --to=tdevries@suse.de \
    --cc=gdb-patches@sourceware.org \
    --cc=tom@tromey.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).