bug 18212 closed too early (by me) ?

Mon Mar 7 18:06:00 GMT 2005

Some months ago I reported bug 18212. In the recent days I discovered 
(after MANY hours of code stripping and testing) that it can be splitted 
to bugs 20351 and 20362. Because I thought that in the case when it 
produced the ICE that the reason is exactly like in 20362 (missing 
files), I marked it as a duplicate of 20362.
Now the problem: Bug 18212 was correctly reported. There's no simple 
trick to avoid the ICE. It could be that bug 20362 is completly 
independet of bug 18212 even if produces the same output! Or fixing bug 
20362 would fix bug 18212. Or many ways lead to ICE in GCJ ... (hoping 
for not!)
The question: Should I reopen bug 18212, or wait until bug 20362 (and 
bug 20351) are fixed to retest it and to reopen if necessary? Or both 
(reopen and retest later) ?
 Bojan