public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
From: Pedro Alves <pedro@palves.net>
To: gdb-patches@sourceware.org
Subject: [PATCH] Make GDBserver abort on internal error in development mode
Date: Fri, 24 Jun 2022 14:26:45 +0100	[thread overview]
Message-ID: <20220624132645.1117504-1-pedro@palves.net> (raw)

Currently, if GDBserver hits some internal assertion, it exits with
error status, instead of aborting.  This makes it harder to debug
GDBserver, as you can't just debug a core file if GDBserver fails an
assertion.  I've had to hack the code to make GDBserver abort to debug
something several times before.

I believe the reason it exits instead of aborting, is to prevent
potentially littering the filesystem of smaller embedded targets with
core files.  I think I recall Daniel Jacobowitz once saying that many
years ago, but I can't be sure.  Anyhow, that seems reasonable to me.

Since we nowadays have a distinction between development and release
modes, I propose to make GDBserver abort on internal error is in
development mode, while keeping the status quo when in release mode.

Thus, after this patch, in development mode, you get:

 $ ../gdbserver/gdbserver
 ../../src/gdbserver/server.cc:3711: A problem internal to GDBserver has been detected.
 captured_main: Assertion `0' failed.
 Aborted (core dumped)
 $

while in release mode, you'll continue to get:

 $ ../gdbserver/gdbserver
 ../../src/gdbserver/server.cc:3711: A problem internal to GDBserver has been detected.
 captured_main: Assertion `0' failed.
 $ echo $?
 1

I do not think that this requires a separate configure switch.

A "--target_board=native-extended-gdbserver" run on Ubuntu 20.04 ends
up with:

		 === gdb Summary ===

 # of unexpected core files      29
 ...

for me, of which 8 are GDBserver core dumps, 7 more than without this
patch.

Change-Id: I6861e08ad71f65a0332c91ec95ca001d130b0e9d
---
 gdbserver/utils.cc      | 20 +++++++++++++++++---
 gdbsupport/config.in    |  3 +++
 gdbsupport/configure    | 13 +++++++++++++
 gdbsupport/configure.ac | 10 ++++++++++
 4 files changed, 43 insertions(+), 3 deletions(-)

diff --git a/gdbserver/utils.cc b/gdbserver/utils.cc
index 4f6516c9cf2..d24057c6012 100644
--- a/gdbserver/utils.cc
+++ b/gdbserver/utils.cc
@@ -28,13 +28,27 @@
 
 /* Generally useful subroutines used throughout the program.  */
 
+/* If in release mode, just exit.  This avoids potentially littering
+   the filesystem of small embedded targets with core files.  If in
+   development mode however, abort, producing core files to help with
+   debugging GDBserver.  */
+static void ATTRIBUTE_NORETURN
+abort_or_exit ()
+{
+#ifdef DEVELOPMENT
+  abort ();
+#else
+  exit (1);
+#endif
+}
+
 void
 malloc_failure (long size)
 {
   fprintf (stderr,
 	   PREFIX "ran out of memory while trying to allocate %lu bytes\n",
 	   (unsigned long) size);
-  exit (1);
+  abort_or_exit ();
 }
 
 /* Print the system error message for errno, and also mention STRING
@@ -82,7 +96,7 @@ vwarning (const char *string, va_list args)
   fprintf (stderr, "\n");
 }
 
-/* Report a problem internal to GDBserver, and exit.  */
+/* Report a problem internal to GDBserver, and abort/exit.  */
 
 void
 internal_verror (const char *file, int line, const char *fmt, va_list args)
@@ -91,7 +105,7 @@ internal_verror (const char *file, int line, const char *fmt, va_list args)
 %s:%d: A problem internal to " TOOLNAME " has been detected.\n", file, line);
   vfprintf (stderr, fmt, args);
   fprintf (stderr, "\n");
-  exit (1);
+  abort_or_exit ();
 }
 
 /* Report a problem internal to GDBserver.  */
diff --git a/gdbsupport/config.in b/gdbsupport/config.in
index a7ae23b4984..577866c97b3 100644
--- a/gdbsupport/config.in
+++ b/gdbsupport/config.in
@@ -11,6 +11,9 @@
 /* Define to 1 if using `alloca.c'. */
 #undef C_ALLOCA
 
+/* Define if development-mode features are enabled. */
+#undef DEVELOPMENT
+
 /* Define to 1 if translation of program messages to the user's native
    language is requested. */
 #undef ENABLE_NLS
diff --git a/gdbsupport/configure b/gdbsupport/configure
index 618f487749f..0b48521deb6 100755
--- a/gdbsupport/configure
+++ b/gdbsupport/configure
@@ -624,6 +624,7 @@ ac_subst_vars='am__EXEEXT_FALSE
 am__EXEEXT_TRUE
 LTLIBOBJS
 LIBOBJS
+CONFIG_STATUS_DEPENDENCIES
 WERROR_CFLAGS
 WARN_CFLAGS
 HAVE_PIPE_OR_PIPE2_FALSE
@@ -10452,6 +10453,15 @@ ac_compiler_gnu=$ac_cv_c_compiler_gnu
 
 
 
+# Set the 'development' global.
+. $srcdir/../bfd/development.sh
+
+if test "$development" = true ; then
+
+$as_echo "#define DEVELOPMENT 1" >>confdefs.h
+
+fi
+
 case ${host} in
   *mingw32*)
 
@@ -10460,6 +10470,9 @@ $as_echo "#define USE_WIN32API 1" >>confdefs.h
     ;;
 esac
 
+CONFIG_STATUS_DEPENDENCIES='$srcdir/../bfd/development.sh'
+
+
 ac_config_files="$ac_config_files Makefile"
 
 cat >confcache <<\_ACEOF
diff --git a/gdbsupport/configure.ac b/gdbsupport/configure.ac
index 1f794605f3c..ac2ade6a220 100644
--- a/gdbsupport/configure.ac
+++ b/gdbsupport/configure.ac
@@ -63,6 +63,14 @@ GDB_AC_PTRACE
 AM_GDB_COMPILER_TYPE
 AM_GDB_WARNINGS
 
+# Set the 'development' global.
+. $srcdir/../bfd/development.sh
+
+if test "$development" = true ; then
+   AC_DEFINE(DEVELOPMENT, 1,
+	     [Define if development-mode features are enabled.])
+fi
+
 case ${host} in
   *mingw32*)
     AC_DEFINE(USE_WIN32API, 1,
@@ -73,5 +81,7 @@ case ${host} in
     ;;
 esac
 
+AC_SUBST([CONFIG_STATUS_DEPENDENCIES], ['$srcdir/../bfd/development.sh'])
+
 AC_CONFIG_FILES([Makefile])
 AC_OUTPUT

base-commit: e83907ff5ffbac3d0224d31ee99e6dc056205f39
-- 
2.36.0


             reply	other threads:[~2022-06-24 13:26 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-24 13:26 Pedro Alves [this message]
2022-06-25 13:27 ` Simon Marchi
2022-06-27 12:06 ` Andrew Burgess

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=20220624132645.1117504-1-pedro@palves.net \
    --to=pedro@palves.net \
    --cc=gdb-patches@sourceware.org \
    /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).