Andrei Dmitriev wrote:
> Hi folks,
>
> the problem still persists with soylatte as well as with
> openjdk6-b16-24_apr_2009-r1.
>
> $sw_vers
> ProductName:Mac OS X
> ProductVersion:10.5.7
> BuildVersion:9J61
>
> I feel that the javac of the newer version is required. How are you
> b
Hi!
This is just a request if someone else sees this problem.
When building a debug HotSpot and putting it into a bsd-port JDK build,
HotSpot eats memory up to 2GB. This does not happen with product or
fastdebug builds.
Still investigating...
-- Christian
Kurt Miller wrote:
> ALT_HOTSPOT_IMPORT_PATH, ALT_JDK_IMPORT_PATH should never be set
> when building the full bsd-port. This may be the source of the
> problem you are hitting.
I too think that this might be the problem. Try without setting them.
-- Christian
Andrei Dmitriev wrote:
> You've compiled it on BSD or Mac?
On Mac OS X 10.5
-- Christian
Andrei V. Dmitriev wrote:
>> 1) bsd-port will not build with gcc 4.2 due to the below compiler error
>>
> Correct.
The error is related to stuff I have commited recently. Was this a
clean build?
-- Christian
Christian Thalinger wrote:
> Andrei V. Dmitriev wrote:
>>> 1) bsd-port will not build with gcc 4.2 due to the below compiler error
>>>
>> Correct.
>
> The error is related to stuff I have commited recently. Was this a
> clean build?
A clean HotSpot de
Andrei V. Dmitriev wrote:
> Hi Christian,
>
> yes, it was a clean build. Actually I experienced two different things:
> 1) with ALT_HOTSPOT_IMPORT_PATH, ALT_JDK_IMPORT_PATH set *and* g++4.2
> I've seen finished build but binaries crashed.
> 2) with ALT_HOTSPOT_IMPORT_PATH, ALT_JDK_IMPORT_PATH UNs
Renju Mathew wrote:
> gmake[6]: Leaving directory
> `/teles/rema/jdk/1stMethod/bsd-port/build/bsd-i586/hotspot/outputdir/bsd_i486_compiler2/product'
>
> cd bsd_i486_compiler2/product && ./test_gamma
>
> java full version "1.6.0_14-b08"
>
> [1] Segmentation fault (core dumped) ./${gamma:-gamma.
Renju Mathew wrote:
> gdb gives the following
> #0 0x7ce25b12 in strrchr () from /usr/lib/libc.so.6
And a "backtrace"?
-- Christian
Renju Mathew wrote:
> Hi
>
> Here is the backtrace
>
> (gdb) bt
> #0 0x7ce25b12 in strrchr () from /usr/lib/libc.so.6
> #1 0xd550b87a in ?? ()
> #2 0x7cdd02fd in init () from /usr/lib/libc.so.6
I think it would be good to have a debug build. Please build again with:
$ make debug_build
--
Michael Franz wrote:
> Christian,
Hi!
Sorry for the long delay.
> I have been away from the OpenJDK for a while and am now running into
> this issue. Do you know what is the first version of OpenJDK 7 that can
> compile the JSR 292 features? I assume the ability to compile JSR 292
> features w
Hi (mostly Greg)!
I wanted to ask at what intervals a merge with upstream is done? I
guess it's just a matter of spare time and actually doing it. Maybe I
could help with that task, but then I'd need commit rights.
Furthermore it would be nice to integrate some patches to be able to
build on Li
Andrew John Hughes wrote:
> Hi Christian,
>
> Unless bsd-port is handled differently to the other OpenJDK forests,
> your existing OpenJDK account should just work. I've used mine across
> about 6-7 forests now.
Yeah, I'm not sure as I needed additional rights for the mlvm repository.
>
> I'm
Greg Lewis wrote:
> G'day Christian,
>
> On Fri, Aug 21, 2009 at 04:05:58PM +0200, Christian Thalinger wrote:
>> Hi (mostly Greg)!
>>
>> I wanted to ask at what intervals a merge with upstream is done? I
>> guess it's just a matter of spare time and act
Andrew John Hughes wrote:
>> Btw. what do you think of using Mercurials rebase extension[1]? I think
>> it would be more straight forward as the base repository would be
>> exactly the same (no changesets in between) and the BSD changes on top
>> of it.
>>
>
> This seems to have worked ok in movi
Andrew John Hughes wrote:
> True, but I don't see the advantage in destroying the history just to
> make the changes into a more convenient set of changesets.
Well, at least when the bsd-port is commited back to jdk7, the history
is destroyed anyway. I never used Mercurial's rebase, but maybe we
Greg Lewis wrote:
> On Tue, Sep 08, 2009 at 03:53:15PM +0800, Max (Weijun) Wang wrote:
>> On Sep 8, 2009, at 11:07 AM, gle...@eyesbeyond.com wrote:
>>
>>>
>>> Changeset: 6b3cf06c18e5
>>> Author:gle...@misty.eyesbeyond.com
>>> Date: 2009-09-07 19:40 -0700
>>> URL: http://hg.openj
Greg Lewis wrote:
> FWIW I can build the bsd-port base on Linux with these changes. I've done
> very little testing though.
I did a 32-bit build on Solaris and that was OK.
-- Christian
John Rose wrote:
> Here's a patch which addresses the printf-format problem. The root
> cause appears to be that GCC 4.2 has gotten a little smarter about
> handing out portability warnings. It objects to passing a "long int"
> to a "%d". The patch attempts to pacify the compiler by changin
Hi!
Here is a patch that enables hsdis compilation for Darwin.
-- Christian
diff --git a/src/share/tools/hsdis/Makefile b/src/share/tools/hsdis/Makefile
--- a/src/share/tools/hsdis/Makefile
+++ b/src/share/tools/hsdis/Makefile
@@ -22,7 +22,8 @@
#
#
-# Single gnu makefile for solaris, linu
On Fri, 2009-10-23 at 16:29 +0200, Oliver Schrenk wrote:
> // get John's patch, not hosted anywherem get it from attachment and
> unzip it
> $ patch -p1 /path/to/downlods>/snowleopard.patch
> $ cd ..
It's available here:
http://hg.openjdk.java.net/mlvm/mlvm/hotspot/raw-file/f9a6cda5b465/snowleo
On Mon, 2010-01-18 at 11:02 -0800, David Green wrote:
> I'm attempting to build OpenJDK from source according to these
> instructions and I'm running into the following error:
>
>
> cc1plus: warnings being treated as errors
>
> /Users/dgreen/Documents/packages/jdk7/bsd-port/hotsp
On 02/14/10 10:39 AM, Ugo Kanain wrote:
> Linking vm...
> Undefined symbols:
>"AbstractInterpreter::can_be_compiled(methodHandle)", referenced
> from:
>CompilationPolicy::canBeCompiled(methodHandle)in
> compilationPolicy.o
> ld: symbol(s) not found
That's just coincidence.
[mlvm-dev would be more appropriate. I'm CC'ing.]
On Tue, 2010-05-18 at 19:37 +0200, Eric Bodden wrote:
> Hello.
>
> I hope this is the right mailing list for this. If not, please correct
> me. I am trying to build JDK 7 on OSX Snow Leopard, to try out the new
> invokedynamic support.
Yeah, we
On Mon, 2010-07-19 at 20:39 -0500, Michael Hall wrote:
> Making adlc
> g++-4.0 -m32 -march=i586 -mstackrealign -m32 -march=i586 -
> mstackrealign -o ../generated/adfiles/adlc ../generated/adfiles/
> adlparse.o ../generated/adfiles/archDesc.o ../generated/adfiles/
> arena.o ../generated/adfiles/
On Tue, 2010-07-20 at 12:13 -0400, Stephen Bannasch wrote:
> At 1:19 PM +0200 7/20/10, Christian Thalinger wrote:
> >On Mon, 2010-07-19 at 20:39 -0500, Michael Hall wrote:
> >> Making adlc
> >> g++-4.0 -m32 -march=i586 -mstackrealign -m32 -march=i586 -
> >> m
On Wed, 2010-07-21 at 09:19 -0400, Stephen Bannasch wrote:
> At 9:57 AM +0200 7/21/10, Christian Thalinger wrote:
> >On Tue, 2010-07-20 at 12:13 -0400, Stephen Bannasch wrote:
> >> At 1:19 PM +0200 7/20/10, Christian Thalinger wrote:
> >> >On Mon, 2010-07-19 at 2
On Mon, 2010-07-19 at 20:39 -0500, Michael Hall wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> From googling I guess this is Java 7 on OS X Snow Leopard again. It
> seemed this has been discussed and resolved on this list before.
> I recently got curious about trying some Java 7 f
On Apr 15, 2011, at 3:12 PM, Stephen Bannasch wrote:
> After I successfully bsd-port build (and after I see if java can start and
> display it's version number) I first run the
> jdk/test/java/lang/String tests with jtreg as a second-level smoketest:
>
> $ jtreg -jdk:build/bsd-amd64/j2sdk-imag
29 matches
Mail list logo
