From patchwork Fri Nov 11 17:44:05 2011
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: [testsuite] : Do not run simulate-thread on alpha*-*-linux*
Date: Fri, 11 Nov 2011 07:44:05 -0000
From: Uros Bizjak <ubizjak@gmail.com>
X-Patchwork-Id: 125237
Message-Id: <CAFULd4bb7QNNCBvgzDre9=AARKwwZcCNsN5PO0edez7fOgYs_Q@mail.gmail.com>
To: gcc-patches@gcc.gnu.org
Cc: Richard Henderson <rth@redhat.com>

Hello!

For some reason, single-stepping executable between ldl_l and stl_c
insns in gdb [1] breaks LL/SC chaining, so atomic operations never
finish. This calls for gdb bugreport.

Also taking into account that dejagnu timeout didn't trigger for
unattended testsuite run and considering huge amount of log
information in testsuite log logged during gdb session, this
all-together was quite devastating to disk space...

We can simply claim that gdb on alpha*-*-linux* is unusable for the
purpose of thread-simulate tests.

2011-11-11  Uros Bizjak  <ubizjak@gmail.com>

	* lib/gcc-simulate-thread.exp (simulate-thread): Do not run on
	alpha*-*-linux* targets.

Tested on alphaev68-pc-linux-gnu.

OK for mainline SVN?

[1] GNU gdb (Gentoo 7.3.1 p1) 7.3.1

Uros.


Index: lib/gcc-simulate-thread.exp
===================================================================
--- lib/gcc-simulate-thread.exp	(revision 181284)
+++ lib/gcc-simulate-thread.exp	(working copy)
@@ -22,6 +22,11 @@
 # Call 'fail' if a given test printed "FAIL:", otherwise call 'pass'.
 
 proc simulate-thread { args } {
+
+    # ??? Exit immediately if this is alpha*-*-linux* target, single-stepping
+    # executable between ldl_l and stl_c insns in gdb breaks LL/SC chaining.
+    if { [istarget alpha*-*-linux*] } { return }
+
     if { ![isnative] || [is_remote target] } { return }
 
     if { [llength $args] == 1 } {