You're right. This stuff is simply broken.
Don't use lockObject/unlockObject. Use lockMutex/unlockMutex instead.
If you can, that is. [ Alternatively, turn deferred-pop optimizations
off in your compiler (gcc can do that, look up the flag...) and hope
you get lucky. This would at least
I think that's better.
Btw, in code like this:
>
> To do that I used the following code snippet (patch attached):
>
> if( CLASS_IS_INTERFACE( clazz ) ) {
> Hjava_lang_Class* objClazz = (*env)->GetObjectClass(env, obj);
> errorInfo info;
>
Hi Enno,
thanks for your report.
>
> Attached is a patch, which to my opinion should fix this (using
> getInheritedMethodIndex() in classMethod.c), but I would appreciate
> if somebody more familiar with this code could check my explanation/fix.
>
I agree with your explanation, but I have
>
> Does this behavior sound like anything you saw bringing up Kaffe on the other
> LinuxVR boxes?
>
Never used LinuxVR - you may want to ask in the pocketlinux community.
- Godmar
No, we don't do that. There may be options, but they're not very portable.
There's two issues: blocking and async-signal safety.
As for blocking, we simply don't call any function that could block.
For instance, X11 first polls using a non-blocking select, and the
we get the event with whatev
139 = 128 + 11 -> SIGSEGV.
The newly built Kaffe crashes.
I advise you to install jikes, the latest version thereof.
Then reconfigure.
- Godmar
>
>
> In Kaffe cvs 2904, I can't seem to "make Klasses"
>
> I get the following errors,
>
> hazel:kool$ make Klasses
> make[1]: Ente
(I'm assuming you didn't declared gc_immortal_initialize as
static:)
Welcome to the undebuggable world of libtool and company.
I recommend you do a /bin/rm -rf * in your object tree,
and then try to reconfigure and rebuild again.
- Godmar
>
>
> This is probably more of a C program
Walter,
Are you saying that there's no access to any atomic instruction
(test-and-set, cmp-exchange, fetch-and-add, whatever) in usermode
on the MIPS R3000?
Hmmm...
http://www.ddci.dk/support/free/white_papers/monitor.html says the same:
(The reason semaphores and locks are relatively slow to
>
> With this fix I seem to have no more unexpected deadlocks.
> Just to be on the safe side, does anyone know of similar
> "supposed to be atomic" operations?
>
Not that, but we've repeatedly pointed out that the CMPEXCH
in locks.c in broken. By now, we've wasted more time discussing
this su
loader is null if the class is loaded by the system class or
bootstrap loader.
>
> Am I accessing it the wrong way OR Is it not implmented yet ? I have try
> to use a few different way to load the Testing class.
> - Class.forName() method
> - URLClassLoader object
> - new Testing()
>
>
Oops, sorry for sending that empty mail.
I was going to comment on Bart's mail.
The documentation his student wrote is definitely a good introduction;
there's also a link on that page to an earlier document written
by Kiyo Inaba.
I'd like to add that trampolines since take a third parameter,
>
>
> In message <20001219101041.I1069@marcus-lx> Marcus Smith wrote:
> > Is there a good document that explains what a "trampoline"
> > is? By studying the code, I think I know, but it would be
> > good to get a paragraph or two.
>
> One of my students wrote some documentation on this and
>
Oops, my fault.
I had even put that comment on top:
//
// Notes: 0xE1AF = mov ip, pc
//0xE59FF000 = ldr pc, [pc, #0]
// NB:0xE59FF004 = ldr pc, [pc, #4]
Btw, which CVSROOT do you use to get to "kaffe_all"?
I'm using ":pserver:[EMAIL PROTECTED]:/cvs/kaffe", but
I don't see a ka
yes. During startup.
- Godmar
>
>
> Why is it not possible to find out about the current thread
> during a call to gcMalloc (from gc-incremental.c). I have
> written:
>
> Hjava_lang_Thread* tid;
> tid = getCurrentThread();
>
> but it raises a segmentation fault at runtime. Are
I'd say it's just a hodgepodge of ideas from various sources.
Don't read too much into it, especially don't read too much into the
use of doubly-linked lists in a conservative, non-incremental,
non-generational, non-moving collector...
- Godmar
>
>
> Is the garbage collection algo
Kaffe is way too big for the Palm in its current form.
Porting it would likely entail a complete redesign.
- Godmar
>
> Hello All
>
> Is anyone working on porting the Kaffe VM to PalmOS? What would this project
> entail? Would other Kaffe users be interested in a Kaffe VM for PalmOS
> /*
> * Lock a mutex - try to do this quickly but if we failed because
> * we can't determine if this is a multiple entry lock or we've got
> * contention then fall back on a slow lock.
> */
> void
> _lockMutex(iLock** lkp, void* where)
> {
> uintp val;
>
> val = (uintp)*lkp
> Let me try. But note - sometimes I had asserts
> related to synchronization, and they were "false".
> I mean, I have a dual processor machine, and,
> for example, and asseret
> assert(xxx==0 && yyy==0)
> was caught, but debugger shows, that both
> xxx and yyy are nulls. Probably, that's
> becau
yes.
>
>
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> Hello kaffe,
>
> Is timeout==0 for ksemGet mean infinite timeout?
>
> - --
> Best regards,
> Maxim mailto:[EMAIL PROTECTED]
>
> -BEGIN PGP SIGNATURE-
> Version: PGP 6.5i
>
> iQA/AwUBOiNqTu
>
> I have sometimes assertion erros
> from putHeavyLock method in
> assert(*lkp == LOCKINPROGRESS);
>
> So, the question is - how lkp is
> supposed to be protected in current thread
> from being modified from others threads?
>
Which COMPARE_AND_EXCHANGE macro does it use?
Do you h
>
> (a) Kaffe should keep its namespace clean
Yes, we should really prefix all our global symbols.
It's such a mess.
>
> (b) On ELF systems ISTM that you could link Kaffe with -Bsymbolic. This
> is not infinitely portable though. (Maybe something for libtool?) Also,
I don't know what -Bsy
If you add .java files, you must run developers/update-class-list
from the libraries/javalib directory to add those to the Makefile.am.
I'm not sure whether that will also add new directories automatically,
but studying that script should help you.
- Godmar
>
> Surely the Makefile.in
This could either be a problem with cache flushing (shouldn't be though)
or a problem with the compiler you're using.
You could try compiling with a different level of optimization
and you should doublecheck that the D-Cache is being flushed properly.
- Godmar
>
> [EMAIL PROTECTED]@
I don't remember the details, but the problem is essentially with some
of the installed libraries (png, jpeg, gif, ...).
I think we had to recompile them with -fPIC.
Unfortunately, with libtool still in the way, I have no idea how you'd
do that in 1.0.6.
- Godmar
>
>
> Hello,
>
>
> Concurrent incremental gc is in development and should be integrated
> in the near future. In kaffe 1.0.6, it is mem2.
>
mem2 is only in Transvirtual's still private tree. It's not part
of the public tree, and neither is it part of 1.0.6.
- Godmar
Thanks, but:
We don't use all that EXTRA_DIST crap anyway because we don't
create distributions with "make dist". Tim has also promised that
if he ever gets to merging the commercial and the public version
that automake will finally go where it belongs.
- Godmar
>
> Hi,
>
> Ther
>
>
> Hi,
>
> when kaffe resolves CONSTANT_Class in constant pool, it replaces it
> with CONSTANT_ResolvedClass, and records the pointer to class struct
> at constants.data[idx]. Why not treat field in the same way?
>
I don't think there's a particular reason.
I think you could do it, whic
>
>
> >>Godmar wrote
>
> >Which OS is this?
>
> Errr... See the header.
>
I noticed that after I sent the mail.
In any event, the sparc linux port uses its own context switch
routine, see sparc/linux/md.c
It depends on SP_OFFSET, but also on some other things.
>
> >Check that the SP_OF
I fixed this in the current CVS tree a while ago.
- Godmar
>
>
> In my environment where
>OS = Linux 2.0.38 #2 Sun Jan 2 11:44:53 JST 2000,
>Kaffe = kaffe-1.0.6,
> the following simple program does not run as expected.
> It prints nothing.
>
> $ cat TestTimeZone.java
> impor
Which OS is this?
Check that the SP_OFFSET value is correct (compile and run
developers/sp_offset.c and compare to config/sparc/youros/...h)
- Godmar
>
>
> >>> Patrick Tullmann <[EMAIL PROTECTED]> 10-Oct-00 10:00:54 PM >>>
>
> >When it seg faults, type 'bt' (for backtrace) and you'
There's no ppc jit support at this time.
You can write one, if you like.
- Godmar
>
>
> Hello Everybody
>
> Is JIT supported for Power PC ? If yes, then where could i
> find JIT executables to generate platform dependent code.
>
> Thanks
> Prashant Patel
>
Dali,
I was looking at this function in String.java:
private static StringBuffer decodeBytes(byte[] bytes, int offset,
int len, ByteToCharConverter encoding) {
StringBuffer sbuf = new StringBuffer(len);
char[] out = new char[512];
int outlen = encod
>
>
> Godmar Back wrote:
>
> [...] Every call results in a new
> > converter object being newinstanced, just to convert a bunch of bytes.
> > (The new converter was one of the changes done to make the
> > charset conversion thread-safe.)
> [...]
>
&g
>From what I understand, and someone correct me if I'm wrong,
there shouldn't be any reason not to include the change you suggest -
if someone implements it, of course.
If I understand your proposal right, you'd use an array for
the first 256 values and a hashtable or something like that
for t
>
>
> I am having some fun trying to work out kaffe's GC implementation.
>
> Before they lock me up for having no life can someone just confirm
> some stuff for me:
>
>
> 1. the kaffe gc interface defined in kaffevm/gc.h defines how the VM
> talks to the GC?
> ie: the VM interacts with the h
Call ExceptionInInitializerError.getException() first and then print the
stacktrace.
- Godmar
>
>
>
> How does Kaffe fill in the stack trace for the exception wrapped in an
> ExceptionInInitializerError? I had hoped that it would tell me the line
> number within the initializer whe
Note that there's nothing in Kaffe that prevents you from directly
accessing an OSS device in default mode. Similarly, you can easily
implement the equivalent of "cat file.au > /dev/audio" in Java.
If you write a small JNI library, you can also export the OSS ioctl()'s
to Java. It's really
> methods twice. It is safe to assume that two calls to getMethod() with
> the same arguments *in the same run of Kaffe* (and without the
> possibility of any classes being gc'd) will always return the same
> method, right?
Right.
- Godmar
Okay, try it again now.
We now also search superinterfaces if you invoke Class.getMethod() on
an interface class.
However, since there's no notion of superinterfaces at the bytecode level,
there's also no ordering - this means that I believe that the result of
such a getMethod call is not an
I know interfaces aren't handled right yet.
I'll look into it.
- Godmar
>
>
> Godmar Back wrote:
> >
> > Have you tried calling 1.2's Class.forName(,false,) and then
> > using getField().getValue() on a final static field?
> > Does doing this invoke ?
>
> I need to be able to run on 1.1...
In 1.1 w/ reflection, you&#
Have you tried calling 1.2's Class.forName(,false,) and then
using getField().getValue() on a final static field?
Does doing this invoke ?
If not, we should be able to fix Kaffe accordingly; I bet 10:1
it will call .
But, actually, I don't really understand what that should be
good for anyway
>
> I worked around this by testing whether
> mth.equals(cls.getMethod(mth.getName(), mth.getParameterTypes())). This
> seems to expose another bug because on numerous occasions I get a
> NoSuchMethodException from this check, which shouldn't ever be
> possible... the imaginary conversation betw
>
>
> Kevin Gamiel wrote:
> > Absolutely. I just found one problem where I am using a Debug object
> > that writes debugging messages to a single file from a few hundred
> > different threads. I was using a PrintWriter but not explicitly closing
> > the stream after each message was written.
Just preload it once and for all and throw the same type in all threads,
even those that run in a classloader context.
This should be "close enough".
- Godmar
>
>
> hi,
>
> I've run into an odd problem that I don't know quite how to fix.
> Basically, a stack overflow isn't handl
>
> I've noticed that Kaffe gives some strange results in some situations;
> one particular case that I noticed is java.io.BufferedWriter. Running
> the attached class as "kaffe net.wuffies.japi.Japize -c
> java.io.BufferedWriter" I get duplicate lines for some methods. In
> particular, the meth
>
>
> > Didn't we have this discussion before?
>
> I did a short search in the archives, but it is a bit hard to find in
> the long lists.
>
Actually, my question wasn't rhetorical.
I thought we've discussed this before, but maybe it's something
different now? Why don't you produce a sho
> (2) What sort of "process model" are you after? Have you considered
> CSP?
>
Just a quick comment about that one:
When we say "process model" we mean "process" as in Unix process, i.e.,
an abstraction for a separate entity that include one or more threads of
execution. The emphasis is o
Didn't we have this discussion before?
Have you tried the java.awt.Defaults.AutoStop property?
- Godmar
>
>
> Hello all,
>
> When the last Frame is dispose()d, the prompt returns. I assume all
> threads die when the last Frame is dispose()d. That's nice, but I am
> trying to catch
Yes, Java does not have "monitors" as Dijkstra/Hoare/Hansen invented them,
that is, shared variables are not by design enclosed in monitors.
It is possible to access shared variables in an unsafe way, as anybody
who's done some Java programming knows.
To my knowledge, people have not demonstrat
Maybe Scott could help?
The problem is that on netwinders, some shared libraries are not
compiled with -fPIC, which results in this well-known error msg:
> BUG IN DYNAMIC LINKER ld.so: ../sysdeps/arm/dl-machine.h: 450:
> elf_machine_rel: Assertion `! "unexpected dynymic reloc type"' failed!
This is a problem with the dynamic libraries that come with your
netwinder. You need libraries compiled with the -fPIC option.
- Godmar
>
>
> Hi,
>
> I am trying to cross-compile kaffe on to arm-linux.
> I cross-compiled and executed "javac hello.java" on arm-linux.
> It gives the
Mikael,
I bet you did not download yesterday's snapshot, as they snapshot
mechanism is broken since Nov and Transvirtual still hasn't fixed it.
You probably downloaded a snapshot from Nov 2 that had this problem
which I since fixed. Check the beginning of the ChangeLog file.
Please try usin
>
> Based on the attached email from Alexandre Olivia on the Kaffe mailing list,
> I applied the following changes to the top level (/pkgsrc/lang/kaffe)
> Makefile:
Always keep in mind that changes to the Makefiles don't survive a new
configure since configure rebuilds Makefiles from Makefile.i
>
> I updated to latest CVS-sources and yes, the problem went away. The
> memory footprint is now almost identical to the case where I just
> create a new StringBuffer every time. The memory usage is still bit
> high (kaffe seems to use almost twice as much memory as jdk11.8, but
> the applicati
>
>
> Is there a way to tell the JIT to recompile a method? If not, would
> this be difficult to accomplish?
>
On the surface, all you'd have to do is tweak the Method struct to
mark the method as untranslated and invoke translate() again; you'll
also have to ensure you don't free the bytecod
A few of comments/statements/questions on my part:
+ The lookup in the system aka primordial classpath (as maintained in
findInJar.c) could be done in a direct hashtable instead of repeated
linear searches thru all files in all classpath entries. I'm thinking
of one hashtable for all,
>
>
> I have a modified sampNat.cc and sample.java from the GCJ home page.
> It compiles fine into 100% pure i586, but when I want to get a hybrid
> system, with the native libsampNat.so and jvm sample.class, I get nowhere.
The partially working gcj support does not support CNI, so
sampNat.cc
>
> I mean, you can run tests every day, but the logs will not change every
> day, because things don't get broken/fixed every day (Kaffe is not as big
> as Mozilla yet).
I think it would be great to have the option to only receive an
email when something breaks or gets fixed. Maybe Chris can
> Godmar Back wrote:
>
> > The .class file Brad sent does not verify because it violates a required
> > invariant about the stack size. I checked in a fix, and we now report the
> > verification error.
>
> Does Kaffe always require that code verify properly?
>
>
>
> Hi,
>
> I am facing the following problem while trying to compile Kaffe on a mips
> system. I simply used the "make" command at the top level directory and
> the gnu compiler, assembler, linker etc are all invoked appropriately.
>
How did you verify that the gnu assembler/linker is
The .class file Brad sent does not verify because it violates a required
invariant about the stack size. I checked in a fix, and we now report the
verification error.
>
> Theoretically, this could be a bug in Jasmin, but since this file works
> without fail in the Blackdown port, I think th
>
>
> Godmar Back wrote:
>
> > Run it under gdb ("setenv KAFFE_DEBUG gdb; kaffe Main")
> > and tell us what the backtrace says. Probably some assumption about the
> > way in which classfiles are constructed that's biting us.
>
> Here&#x
Run it under gdb ("setenv KAFFE_DEBUG gdb; kaffe Main")
and tell us what the backtrace says. Probably some assumption about the
way in which classfiles are constructed that's biting us.
- Godmar
>
>
>
> --qjSZefE6w0iW6XmQ
> Content-Type: text/plain; charset=us-ascii
>
> [ I have
>
>
>
> Thanks for the feedback. I have added the last-update date and version
> number of Kaffe to the document. This document was released today and
> was intended to describe Kaffe release 1.0.5. I would like to make this
> paper up-to-date, but I have not seen the anchor mentioned. Cou
That's great. I think Tim has someone redesign the kaffe.org website,
and I'm sure this could be included or at least linked to.
Only problem I have is that it describes the old, broken form of
trampolines. The fixed trampolines I put in last year take an additional
parameter for what I call
>
> lockObject(H_java_lang_Object*)
> unlockObject(H_java_lang_Object*)
>
> These should only be used for Java object monitors. They just take a
> pointer to the Java object. The intent is that if we ever create a
> facility to log the execution of all monitorenter/exit bytecodes,
I bet the problem is general, it's in all the fragility surrounding the use
of asm()'s in various versions of gcc/egcs/gcc on various architectures.
Unfortunately, the solution(s) is/are probably specific to a given
version (or range of versions) of the compiler and a given architecture. :-(
Stefan,
this is the desired behavior.
For instance, if there is a directory ./java with say
./java/lang/Object.class, then I *want* Kaffe to load this
file for java.lang.Object and not the java/lang/object.class from
Klasses.jar. Although it can be confusing (the whole CLASSPATH idea
is ill-
>
> assertion "!INTS_DISABLED()" failed: file "exception.c", line 308
>
> Tracing into the code - it would seem that interrupts are disabled in a =
> place that they should be enabled. If anyone has seen this error and =
> could give me some hint as to what I could do to remedy it, please let
There should be an #ifdef HAVE_LIBFFI in support.c now that activates
the libffi code.
- Godmar
>
>
> Pavel Roskin <[EMAIL PROTECTED]> writes:
>
> > Hello!
> >
> > I'm sorry, I forgot to remember that config/config.h.in needs to be
> > regenerated after applying my patch for libff
Jason did that already.
- Godmar
>
>
> Hello!
>
> I'm sorry, I forgot to remember that config/config.h.in needs to be
> regenerated after applying my patch for libffi.
>
> Pavel Roskin
>
To complete the picture, and to evaluate the possible benefit
of the utf8 alternative, it would also be necessary to study
- how many of them are 7bit representable (for which charAt() would be
unaffected)
- how long the strings are and how often strings are copied during
construction or c
Two comments:
1. The configure option you're looking for is not --target, it's --host.
2. You may want to read the manual for size(1) and strip(1).
- Godmar
Just to make this clear. I don't plan on doing anything.
I'm just throwing out ideas here. But if I implemented it, I think
I would use such a flag to special-case regular strings.
Artur's numbers seem useful. However, as you say, a full investigation
would be needed to quantify the benefit
>
>
> Derek L Davies <[EMAIL PROTECTED]> writes:
>
> > Patrick Tullmann <[EMAIL PROTECTED]> writes:
> > > Its always easier to wrap a flexible interface an present it as a less
> > > flexible one, so I suggest going with the 'this' pointer approach. If
> > > it turns out that all the JVMDI
>
>
> Derek L Davies <[EMAIL PROTECTED]> writes:
>
> > I wonder why the JVMDI call isn't done like this:
> >
> > (*jvmdi)->SetEventHook(jvmdi, &evtFtn);
>
> They must have decided it was too annoying.
>
> > so that the benefits of having an interface pointer are realized for
> > JVMDI as th
Keep in mind that this C++ code
>
> jvmdi->SetEventHook(&evtFtn);
>
is roughly equivalent to this C code:
>
> (*jvmdi)->SetEventHook(jvmdi, &evtFtn);
>
> Also, does anyone know why the forth function table entry is reserved
> in the JNIEnv interface pointer? I understand the first thre
>
>
> Godmar Back wrote:
>
> > Thinking about it, maybe it would even be worth thinking about
> > not implementing Strings as char[] arrays internally, but as utf8strings.
> > This would save space for all ASCII strings, it would allows to directly
> >
>
>
> On Feb 4, 2000, Mo DeJong <[EMAIL PROTECTED]> wrote:
>
> > The current Kaffe implementation means that a \0 embedded in the
> > class name will screw up code that expects a \0 at the end of a
> > string.
>
> AFAIK, there can't be \0s embedded in class names.
Utf8strings can have \u0
http://www.research.ibm.com/journal/sj/391/suganuma.html
may provide for interesting reading.
- Godmar
>
>
> Hi,
>
> I have been looking into the kaffe sources for quite sometime,
> especially the implementation of JIT. I am not sure what all were the primary
> consideration
>
>
> >
> >
> > On Thu, 3 Feb 100, Godmar Back wrote:
> >
> > Ok, that sounds like a plan. The only other problem I noticed is that
> > Kaffe seems to use plain char* types for class names. These class names
> > really should be in unicode s
>
>
> On Thu, 3 Feb 100, Godmar Back wrote:
>
> Ok, that sounds like a plan. The only other problem I noticed is that
> Kaffe seems to use plain char* types for class names. These class names
> really should be in unicode strings or at least UTF8 strings. The
> curr
I'm working on it. We'll first try renaming the basic types internally
w/o breaking gcj (Jason sent me a patch for that), and see where that gets
us, and then the rest. I think we should be able to match Sun's output
exactly. Unfortunately, egcs broke on me right after I checked it out
toda
>
>
> >>>>> "GB" == Godmar Back <[EMAIL PROTECTED]> writes:
>
> GB> It's 17. So change 19 -> 17 in i386/jit.h and try that.
>
> Yep, my otherwise unmodified Kaffe is loading gcj-compiled code now. Neat.
That's grea
Give us more information, including but not limited to:
- the configure line you used.
- the version of gcc you used.
- the output of ldd /usr/local/kaffe/libexec/Kaffe
- a full stacktrace
- your CLASSPATH setting (if any)
- your LD_LIBRARY_PATH (if any)
It's impossible to guess from little you
>
> sh your_kaffe_tree/developers/geteh_from_libgcc2 your_gcc_cvs_tree/libgcc2.c | diff
>- your_kaffe_tree/kaffe/kaffevm/gcj/eh.c
>
> So, I just did that and noticed that they fixed a lot of warnings
> (imagine that libgcc2.c didn't even compile with -Wall two months ago),
> but there seems to
>
> So it seems to be something with exception handling. In the two cases
> below I had to take care to avoid class-lookup failures. Avoiding
> those failures, the gcj shared object loads and runs fine.
>
Yes, it sounds so familiar. Looking for these non-existent classes
is the first except
Another thing about the exception stuff.
Since I needed a hook into libgcc2.a's exception mechanism, I duplicated
a piece of libgcc2.c in eh.c. (Once/If Kaffe support works, this may be
included in the official libgcc2.c distribution since it's a hook that
could benefit other dcgs as well.)
T
Thanks for sending the info I requested.
> Ok, I did this blindly per FAQ/FAQ.gcj.
That's a cut-and-paste error in the FAQ. Thanks for pointing it out.
- Godmar
>
>
> >>>>> "GB" == Godmar Back <[EMAIL PROTECTED]> writes:
>
> GB> Or are you a "alpha test power user" who wants to help testing gcj
> GB> support?
>
> Yeah, I understand refinements are a long road.. But I must ad
Hi Marcus,
Could you describe what you were trying to do?
Are you a "normal user" who was just tripped up by this?
Or are you a "alpha test power user" who wants to help testing gcj support?
If so, did you remember to run the fixup program in developers?
Could you give the gcc version (gcc
>
> Jason Baker wrote:
> > ./developers/fixup.c:addSymbol("_Jv_intClass", CLASS, "int", "", "");
>
> That code is actually commented out already. Godmar will have to
> chime in on its intent.
>
At first, I thought I just fake the java::lang::Class _Jv_intClass structure
as a big bss symb
>
> I'm trying to compile kaffe 1.0.5 for RTEMS.
>
> I can not use alloca function so, this fails:
>
Then don't use it.
- Godmar
--
"Doctor, Doctor, it hurts when I do that."
"Well, then don't do it."
See the cross-compiling section in http://www.kaffe.org/porting.html
--target doesn't do you any good; it's for compilers only.
You must use --host (or give an argument) to specify the architecture that
will host the VM.
- Godmar
>
>
> 1) How does kaffe handle making object once again reachable from
> finalize method ? Are there two gc phases - first for detecting
> unreachable objects and second scannning everything again to actually
> free object both unreachable and already finalized ?
Take a look at gc-incremental.
That's great. I love to see (optional) support for libffi in kaffe.
> fails. The only difference is that following line doesn't appear in th
> output:
>
> at java.lang.reflect.Method.invoke(Method.java:native)
>
That fails with the interpreter and without libffi too.
Need to sort ou
>
>
> Edward Llewellyn writes:
> > Is it possible we could run multiple instances of Kaffe on the same Java
> > program? The reason I ask is because
> > we have an embedded app with two screens (for two users). We may need to
> > run multiple copies of
> > a browser, one for each user, or pos
>
> I'm not sure of the entomology of the phrase, but it refers
entomology is a branch of zoology that deals with insects.
I think the word you're looking for is etymology.
But your explanation of the technical aspects is correct.
I think it simply means the "thread package of the Green
I'd try "-verbose" first.
- Godmar
>
>
> If I wanted to see if my lacking a CLASSPATH setting is causing my
> problems, would I set it to every *.jar file that exist?
>
> Many thanks.
>
1 - 100 of 234 matches
Mail list logo
