Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Leonardo Boquillon <leonardo.boquillon@tallertechnologies.com>
To: gdb-patches@sourceware.org,	tom@tromey.com
Subject: [PING][PATCH] Python: Fix Python error when "Quit"ting a paged info pretty-printers
Date: Tue, 15 Mar 2016 12:53:00 -0000	[thread overview]
Message-ID: <1458046365-21110-1-git-send-email-leonardo.boquillon@tallertechnologies.com> (raw)

Right now the "Quit" command used in the output paging is handled as an exception.
If we issue a "Quit" while outputting the registered pretty-printers list, the
Python handling layer will catch it and think it's a Python error.

The fix is to check if the error coming from Python is a Quit signal. If it is,
do not handle it as an error and resume the execution normally.

Is this ok to commit? I have a company-wide copyright assignment and a coworker
of mine is a write-after-approval maintainer.

gdb/ChangeLog:
2016-03-08  Leonardo Boquillon  <leonardo.boquillon@tallertechnologies.com>

	* python/py-cmd.c (cmdpy_function): Handle the "Quit" exception.

---
 gdb/python/py-cmd.c | 85 +++++++++++++++++++++++++++++------------------------
 1 file changed, 47 insertions(+), 38 deletions(-)

diff --git a/gdb/python/py-cmd.c b/gdb/python/py-cmd.c
index af6c5cf..edcaf79 100644
--- a/gdb/python/py-cmd.c
+++ b/gdb/python/py-cmd.c
@@ -165,46 +165,55 @@ cmdpy_function (struct cmd_list_element *command, char *args, int from_tty)
 	 we no longer own ptype, pvalue after the call to PyErr_Restore.  */
 
       msg = gdbpy_exception_to_string (ptype, pvalue);
-      make_cleanup (xfree, msg);
-
-      if (msg == NULL)
-	{
-	  /* An error occurred computing the string representation of the
-	     error message.  This is rare, but we should inform the user.  */
-	  printf_filtered (_("An error occurred in a Python command\n"
-			     "and then another occurred computing the "
-			     "error message.\n"));
-	  gdbpy_print_stack ();
-	}
-
-      /* Don't print the stack for gdb.GdbError exceptions.
-	 It is generally used to flag user errors.
-
-	 We also don't want to print "Error occurred in Python command"
-	 for user errors.  However, a missing message for gdb.GdbError
-	 exceptions is arguably a bug, so we flag it as such.  */
-
-      if (! PyErr_GivenExceptionMatches (ptype, gdbpy_gdberror_exc)
-	  || msg == NULL || *msg == '\0')
-	{
-	  PyErr_Restore (ptype, pvalue, ptraceback);
-	  gdbpy_print_stack ();
-	  if (msg != NULL && *msg != '\0')
-	    error (_("Error occurred in Python command: %s"), msg);
+      /* We need to check this because the "Quit" command used in the output 
+         paging is handled as an exception. If we issue a "Quit" while 
+         outputting the registered pretty-printers list, the Python handling 
+         layer will catch it and think it's a Python error                 */
+      if (strncmp(msg, "Quit", 5))
+	  {
+	    if (msg == NULL)
+	      {
+		/* An error occurred computing the string representation of the
+		   error message.  This is rare, but we should inform the user.*/
+		printf_filtered (_("An error occurred in a Python command\n"
+				   "and then another occurred computing the "
+				   "error message.\n"));
+		gdbpy_print_stack ();
+	      }
+
+	/* Don't print the stack for gdb.GdbError exceptions.
+	   It is generally used to flag user errors.
+
+	   We also don't want to print "Error occurred in Python command"
+	   for user errors.  However, a missing message for gdb.GdbError
+	   exceptions is arguably a bug, so we flag it as such.  */
+
+	  if (! PyErr_GivenExceptionMatches (ptype, gdbpy_gdberror_exc)
+	      || msg == NULL || *msg == '\0')
+	    {
+	      PyErr_Restore (ptype, pvalue, ptraceback);
+	      gdbpy_print_stack ();
+	      if (msg != NULL && *msg != '\0')
+	        error (_("Error occurred in Python command: %s"), msg);
+	      else
+	        error (_("Error occurred in Python command."));
+	    }
 	  else
-	    error (_("Error occurred in Python command."));
-	}
-      else
-	{
-	  Py_XDECREF (ptype);
-	  Py_XDECREF (pvalue);
-	  Py_XDECREF (ptraceback);
-	  error ("%s", msg);
-	}
-    }
+	    {
+	      Py_XDECREF (ptype);
+	      Py_XDECREF (pvalue);
+	      Py_XDECREF (ptraceback);
+	      error ("%s", msg);
+	    }
+	} // ELSE: This was not an error just quit signal. Do nothing.
 
-  Py_DECREF (result);
-  do_cleanups (cleanup);
+      make_cleanup (xfree, msg);
+    }
+  else
+    {
+      Py_DECREF (result);
+      do_cleanups (cleanup);
+    }
 }
 
 /* Helper function for the Python command completers (both "pure"
-- 
1.9.1


             reply	other threads:[~2016-03-15 12:53 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-03-15 12:53 Leonardo Boquillon [this message]
2016-03-15 15:45 ` Pedro Alves
2016-03-15 17:20 ` Tom Tromey
2016-03-15 17:51   ` Pedro Alves

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=1458046365-21110-1-git-send-email-leonardo.boquillon@tallertechnologies.com \
    --to=leonardo.boquillon@tallertechnologies.com \
    --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