This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
Re: SIGSEGV in jc1
- To: "Jeff Sturm" <jsturm at one-point dot com>,"Bryce McKinlay" <bryce at waitaki dot otago dot ac dot nz>
- Subject: Re: SIGSEGV in jc1
- From: "Tuomas Leikola" <tuomas dot leikola at digia dot com>
- Date: Wed, 9 May 2001 19:08:40 +0300
- Cc: <java at gcc dot gnu dot org>
- References: <Pine.LNX.4.10.10105091155130.4107-100000@mars.deadcafe.org>
would it not be a good idea to not compile a source file twice.. eg. reuse
the earlier parsing/compilation?
----- Original Message -----
From: "Jeff Sturm" <jsturm@one-point.com>
To: "Bryce McKinlay" <bryce@waitaki.otago.ac.nz>
Cc: <java@gcc.gnu.org>
Sent: Wednesday, May 09, 2001 7:17 PM
Subject: Re: SIGSEGV in jc1
> On Wed, 9 May 2001, Bryce McKinlay wrote:
> > In the past I have noticed that the problems with multifile compilation
> > were often dependent on the order in which the files were passed to jc1.
>
> I think you're right. Here's an example I was just looking at, again on
> the branch:
>
> ../src/jacl/tcl/lang/NamespaceCmd.java:32: Class `tcl.lang.NamespaceCmd'
> already defined in ./../src/jacl/tcl/lang/NamespaceCmd.java:32.
> public class NamespaceCmd extends InternalRep implements Command {
> ^
>
> It seems this class has already been parsed as a side effect of compiling
> some other class. One way to avoid the error is to compare the canonical
> path names, but I'm not really sure if there aren't other nasty side
> effects of parsing a source file twice.