[committed] allow assert in inner class

Sat Feb 28 01:07:00 GMT 2004

Jeff Sturm wrote:
> build_assertion really should use build_incomplete_class_ref, not
> build_dot_class_method_invocation. The former already handles the case
> where GET_CPC is an inner class or interface.

I checked in a patch to do this, as attached.
I don't know whether this is appropriate for 3.4; I don't think
it counts as a regression, but it does fix some PRs.
This caused one regression running Jacks:
non-jls-jsr41.3-runtime-2 "Asserts do not cause class initialization"
(which is in jacks/tests/runtime/non-jls/assert/tests.tcl).
The test case does work work when compiling to native, but
fails when compiling to bytecode. However, JDK 1.4.2 fails
in the same manner, and there is something to be said for
compatibility. I haven't tried to see what JDK 1.5.0 does.
I think this patch is ok, but we should revisit this after
1.5.0 is released (i.e. not beta).
Presumably we should mark this as a Jacks XFAIL; could someone
with a Jacks account do that?
-- 
	--Per Bothner
per@bothner.com http://per.bothner.com/
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: assert-inner2.patch
URL: <http://gcc.gnu.org/pipermail/java/attachments/20040228/0e3ccec0/attachment.ksh>