Package: g++-4.0
Version: 4.0.1-2
Severity: normal
Hi,
I've run into a build failure whilst transitioning one of my packages to
g++-4.0. The error occurs when HUGE_VAL is used with -pedantic in C++ code,
as in the following program:
#include
int main() {
double d = 1.0;
if (d
Hi,
> we cannot reasonably do anything about this bug without more info, and
> there's been no feedback for 3 month, so I close it.
Well, 1.5 months actually. But yeah, I probably wouldn't have made it
back to you after 3 months anyway. :/
Just rebuilt jython with gcj-3.3, and it seems to run
> well, I still could reenable building gij-3.2 / gcj-3.2, if that's
> necessary for the sarge release.
I don't have any reason to believe that this bug is new to gcj-3.3 - the
only reason it wasn't spotted until now was kaffe+jython used to be
broken for other reasons. So it's entirely possible
Hi.
> Ben, please could you update the status regarding newer gcj-3.3 /
> gcc-snapshot packages?
Still broken with both gcj 4:3.3.2-2 and gcc-snapshot 20040118-1.
Ben.
> Ben, please could you update the status regarding newer gcj-3.3 /
> gcc-snapshot packages?
Yep, sorry - got quite a bit busy these last few weeks. I will get to
it.
b.
Hi.
> please could you provide a reduced testcase and/or name the file
> beeing miscompiled?
This might or might not be an easy thing, and I'm about to leave for two
weeks where I will quite probably be down to email and nothing more. So
I can do this but it might be some time.
> what's the st
> Nevertheless it's the libstdc++ issue. Your code works fine with our
> gcc-3.2 build of 1.4.1.
Rightio, so it seems the debian j2se1.4 package does need a rebuild then.
Thanks - Ben. :)
> If you use a C++ dynamic_cast complied with g++-3.3 within a native Java
> method, the blackdown JVM (1.4.1) crashes.
Btw, further investigation suggests that this problem might be fixed by
rebuilding the blackdown JVM with gcc-3.3. Stephen, would this be possible?
Thanks - Ben. :)
Hi. I've just been converting an app to use C++-style casts instead of
C-style casts and I've come across a nasty problem.
If you use a C++ dynamic_cast complied with g++-3.3 within a native Java
method, the blackdown JVM (1.4.1) crashes.
I've attached three very short sample files that illus
Package: gij-3.2
Severity: normal
Hi. From what I can tell, zip file handling under gij-3.2 seems to be
broken. I have a transcript of a session below that shows a simple Java
program involving zip files that breaks under gij-3.2.
This interests me particularly, since Jython currently has a gr
Okay, sorry this has taken so long.
I've put a gij-wrapper-3.2 and a gcj-wrapper-3.2 up on
http://people.debian.org/~bab/java if you want to take a look.
I've done some basic testing and everything seems to be in order, but I
wouldn't mind doing a little more testing with some of the more comple
> > According to the Java policy, packages that provide java1-runtime must
> > support the the complete java runtime environment.
Well, AFAICT (looking at policy linked to www.debian.org), the Java policy
isn't particularly explicit. Section 2.1 states:
Java virtual machines must provide java
Hmm, okay, my bad. I'm reading through 12.8 now and it seems the
implicit copy constructor is always available, not just when no other
constructors are defined as I had thought.
If you agree with this then please go ahead and close this bug.
Thanks - Ben.
> > I would have expected a compile error, since the declaration
> > "Foo bar = foo;" tries to use the non-existant default constructor for
> > class Foo.
>
> No, it doesn't. This is not a initialization followed by an assignment
> operator; there is no assignment operator at all in this statement
Package: g++-3.2
Version: 1:3.2.1-1
Severity: normal
Hi. Both g++-3.2 and g++-2.95 seem to have problems with the following
code.
#include
struct Foo {
int val;
Foo(int newVal) {
val = newVal;
}
Foo& operator = (const Foo& other) {
val = 2;
ava/jnifix
It merely appends /usr/lib/jni to the JNI search path if it's not already
included and passes everything on to the real java runtime.
Thanks - Ben.
--
Ben Burton
[EMAIL PROTECTED] | [EMAIL PROTECTED]
Public Key: finger [EMAIL PROTECTED]
I go into a real vulnerable side of
Package: gcj-3.2
Version: 1:3.2.1-0pre4
Severity: normal
Hi. The gcj-wrapper-3.2 script currently cannot handle paths containing
spaces. As an example, create subdirectory "foo bar" containing Test.java.
[EMAIL PROTECTED]:/tmp> gcj-wrapper-3.2 foo\ bar/Test.java
gcj-3.2: cannot specify -o with
Package: gcj-3.2
Version: 1:3.2.1-0pre3
Severity: normal
Tags: patch
Hi. The wrapper /usr/bin/gcj-wrapper-3.2 fails if the first source it
is given contains path information beginning with a period.
Example:
gcj-wrapper-3.2 ../src/Test.java
This is because such a source is not recognised as
Package: gcj-3.2
Version: 1:3.2.1-0pre1
Severity: normal
Tags: patch
Hi. The script gcj-wrapper-3.2 will not compile java files whose names
begin with an underscore. I'm specifically encountering this problem in
trying to modify the jython build to use gcj.
The patch is simple (and has been tes
tags 150142 + patch
thanks
Hi. Below is a small patch that seems to fix the problem.
--- /usr/bin/gcj-wrapper-3.1Tue Jun 4 01:21:17 2002
+++ gcj-wrapper-3.1 Mon Jun 17 18:20:27 2002
@@ -11,7 +11,12 @@
fi
case "$arg" in
--classpath|-encoding|-bootclasspath|-extdirs)
+
Package: gcj-3.1
Version: 1:3.1.1-0pre2
Severity: normal
Hi. The script gcj-wrapper-3.1 claims to translate javac options to gcj
options (at least according to its manpage). However, the option -classpath
is passed to gcj unchanged at which point gcj-3.1 fails to recognise it
(although gcj-3.0 r
gt; 1.2 (for example Swing is missing and AWT is incomplete).
It's primarily the requirement on javax.swing that forces libbtools-java to
depend on java2-runtime.
Ben.
--
Ben Burton
[EMAIL PROTECTED] | [EMAIL PROTECTED]
http://baasil.humbug.org.au/bab/
Public Key: finger [EMAIL PROTECTE
Package: gcj-3.0
Version: 1:3.0.4-6
Severity: normal
Hi. I have the following piece of code that compiles correctly but then
gives an error when run. It involves calling a default constructor for
an inner class.
If gcj won't accept this default constructor, it should give a compile
error. If g
Package: gcj
Version: 2:3.0.4-5
Severity: normal
Hi. I'm getting a gcj compile error when calling a member function of a
newly created instance of an inner class.
Try compiling the following chunk of code:
class Test {
public static void main(String[] args) {
(new InnerTest())
Package: gcj-3.0
Version: 1:3.0.4-6
Severity: normal
Hi. I have a java file that causes gcj to lock up when it tries to
compile it to java bytecode. The specific example is in the jython
sources; I believe you can reproduce the problem as follows:
[EMAIL PROTECTED]:~/source/debian/made> apt-get
packages that depend on java-virtual-machine call
/usr/bin/java from their startup scripts.
Proposed java policy is at:
http://people.debian.org/~opal/java/policy.html
(and in incoming in package java-common, I believe).
Thanks - Ben.
--
Ben Burton
[EMAIL PROTECTED] | [EMAIL PROTECTED]
http
tags 103051 +patch
thanks
Hi.. there is a manpage for fastjar up on the missing manpages project at:
http://www.netmeister.org/misc/m2p2/man1/fastjar.1.gz
I realise this is right on the heels of the corresponding post to
debian-devel, but I'm mentioning this explicitly because I filed #103051
s
Package: fastjar
Version: 1:3.0-3
Severity: normal
Hi. Would it be possible to add a man page for fastjar?
Thanks - Ben.
-- System Information
Debian Release: testing/unstable
Architecture: i386
Kernel: Linux emerald 2.4.4-686 #1 Sun Apr 29 12:22:40 EST 2001 i686
Locale: LANG=C, LC_CTYPE=C
Ver
28 matches
Mail list logo
