[Patch] New throw test

[David Daney]

I preperation for hacking on MIPS signal handling changes, I wrote this 
test.

As indicated in the comments in the test, I am checking that unwinding 
is working properly if the very first instruction of a method faults. 
The test code does this for MIPS.  But as far as I can tell, it is 
impossible to generate code that falults on the first instruction for 
i686 so the test passes there as well.

In any event it was unclear if any existing test covered this case so I 
would like to add it.

Manually tested on mipsel-linux and i686-pc-linux-gnu,  Full make -k 
check in libjava running on i686-pc-linux-gnu.

OK to commit if it passes?

2006-11-02  David Daney  <ddaney@avtrex.com>

	* testsuite/libjava.lang/Throw_3.java: New Test.
	* testsuite/libjava.lang/Throw_3.out: Its expected output.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Throw_3.diff
Type: text/x-patch
Size: 1407 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/java-patches/attachments/20061102/a5817c5b/attachment.bin>