From dalibor.topic at googlemail.com Tue Apr 1 07:16:15 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Tue, 1 Apr 2008 16:16:15 +0200 Subject: Call for vote In-Reply-To: <47EC698E.1060107@info9.net> References: <47EC698E.1060107@info9.net> Message-ID: <985bee770804010716u40a967c1kfb0ebce6988fe08b@mail.gmail.com> On Fri, Mar 28, 2008 at 5:44 AM, Tom Marble wrote: > Dalibor (et al): > > > * Should the porters group sponsor the project to port OpenJDK > > to the MIPS CPU architecture? > > Yes. > Thank you for your vote! cheers, dalibor topic From dalibor.topic at googlemail.com Tue Apr 1 07:17:22 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Tue, 1 Apr 2008 16:17:22 +0200 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: <47E94357.2090509@sun.com> References: <47E922D4.7060000@kaffe.org> <985bee770803250953g5ed1f77ctb5520ef9d62f52fd@mail.gmail.com> <47E94357.2090509@sun.com> Message-ID: <985bee770804010717s5497a4b6k2252272711043199@mail.gmail.com> On Tue, Mar 25, 2008 at 8:24 PM, David Herron wrote: > > > Dalibor Topic wrote: > On Tue, Mar 25, 2008 at 5:05 PM, Dalibor Topic wrote: > > > Dear members of the porters group, > > it is my pleasure to call you to vote in my role as the group's > moderator. The issue being voted on is > > * Should the porters group sponsor the project to port OpenJDK > to the MIPS CPU architecture? > > Please vote with yes or no. The voting period is two weeks. > > Yes. > Thank you for your vote! cheers, dalibor topic From dalibor.topic at googlemail.com Tue Apr 1 07:18:03 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Tue, 1 Apr 2008 16:18:03 +0200 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: <20080325174138.5DBAD5B656@eggemoggin.niobe.net> References: <47E922D4.7060000@kaffe.org> <20080325174138.5DBAD5B656@eggemoggin.niobe.net> Message-ID: <985bee770804010718s3425b119jffdec66d201c7941@mail.gmail.com> On Tue, Mar 25, 2008 at 7:41 PM, Mark Reinhold wrote: > Vote: yes > > - Mark (who has fond memories of hacking MIPS assembler code, way back when) > Thank you for your vote. cheers, dalibor topic From dalibor.topic at googlemail.com Tue Apr 1 07:20:13 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Tue, 1 Apr 2008 16:20:13 +0200 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: <47E922D4.7060000@kaffe.org> References: <47E922D4.7060000@kaffe.org> Message-ID: <985bee770804010720k62fead1dxbf1f962d8599e4b8@mail.gmail.com> On Tue, Mar 25, 2008 at 6:05 PM, Dalibor Topic wrote: > Dear members of the porters group, > > it is my pleasure to call you to vote in my role as the group's > moderator. The issue being voted on is > > * Should the porters group sponsor the project to port OpenJDK > to the MIPS CPU architecture? > > Please vote with yes or no. The voting period is two weeks. > > cheers, > dalibor topic The vote has passed. The members of the porters group have cast 4 votes for sponsoring the project, with no votes against it. The porters group therefore has decided to sponsor the project. I'd like to thank all the members of the group for their votes, and everyone for their contributions in the discussion of the proposal, and I'd like to congratulate the MIPS OpenJDK porters team, and welcome them into OpenJDK family of projects. cheers, dalibor topic From springer at reservoir.com Tue Apr 1 07:43:42 2008 From: springer at reservoir.com (Jonathan Springer) Date: Tue, 01 Apr 2008 09:43:42 -0500 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: <985bee770804010720k62fead1dxbf1f962d8599e4b8@mail.gmail.com> References: <47E922D4.7060000@kaffe.org> <985bee770804010720k62fead1dxbf1f962d8599e4b8@mail.gmail.com> Message-ID: <47F24A1E.9000609@reservoir.com> Dalibor Topic wrote: > On Tue, Mar 25, 2008 at 6:05 PM, Dalibor Topic wrote: >> Dear members of the porters group, >> >> it is my pleasure to call you to vote in my role as the group's >> moderator. The issue being voted on is >> >> * Should the porters group sponsor the project to port OpenJDK >> to the MIPS CPU architecture? >> >> Please vote with yes or no. The voting period is two weeks. >> >> cheers, >> dalibor topic > > The vote has passed. The members of the porters group have cast 4 > votes for sponsoring the project, with no votes against it. The porters > group therefore has decided to sponsor the project. > > I'd like to thank all the members of the group for their votes, and > everyone for their contributions in the discussion of the proposal, and > I'd like to congratulate the MIPS OpenJDK porters team, and > welcome them into OpenJDK family of projects. > > cheers, > dalibor topic Thanks to those who voted, and especially thanks to Dalibor for spearheading this vote. We will try to make the MIPS port as high-quality as we can and keep up with the pace of OpenJDK development. -Jonathan -- Jonathan Springer | Reservoir Labs, Inc. | http://www.reservoir.com/ From dalibor.topic at googlemail.com Tue Apr 1 11:25:37 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Tue, 1 Apr 2008 20:25:37 +0200 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: <47F24A1E.9000609@reservoir.com> References: <47E922D4.7060000@kaffe.org> <985bee770804010720k62fead1dxbf1f962d8599e4b8@mail.gmail.com> <47F24A1E.9000609@reservoir.com> Message-ID: <985bee770804011125t650e9a8bp5ee20036b237781d@mail.gmail.com> On Tue, Apr 1, 2008 at 4:43 PM, Jonathan Springer wrote: > > Thanks to those who voted, and especially thanks to Dalibor for > spearheading this vote. We will try to make the MIPS port as > high-quality as we can and keep up with the pace of OpenJDK > development. > Thanks, Jonathan! I CC:ed Mark Reinhold, who'll help you get set up with a mailing list and a web page for a start. If you'd like to have a hg forest, as well, and start migrating to the openjdk mercurial infrastructure, Mark can set that up for you, too. I'll update the porters page to add the mips port. Would you like to keep the UltraViolet name for the project? cheers, dalibor topic From springer at reservoir.com Tue Apr 1 16:32:40 2008 From: springer at reservoir.com (Jonathan Springer) Date: Tue, 01 Apr 2008 18:32:40 -0500 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: <985bee770804011125t650e9a8bp5ee20036b237781d@mail.gmail.com> References: <47E922D4.7060000@kaffe.org> <985bee770804010720k62fead1dxbf1f962d8599e4b8@mail.gmail.com> <47F24A1E.9000609@reservoir.com> <985bee770804011125t650e9a8bp5ee20036b237781d@mail.gmail.com> Message-ID: <47F2C618.9070209@reservoir.com> Dalibor Topic wrote: > On Tue, Apr 1, 2008 at 4:43 PM, Jonathan Springer > wrote: >> Thanks to those who voted, and especially thanks to Dalibor for >> spearheading this vote. We will try to make the MIPS port as >> high-quality as we can and keep up with the pace of OpenJDK >> development. > > Thanks, Jonathan! I CC:ed Mark Reinhold, who'll help you get set up with > a mailing list and a web page for a start. If you'd like to have a hg > forest, as well, > and start migrating to the openjdk mercurial infrastructure, Mark can set that > up for you, too. Thanks. I don't know if we're ready for a forest just yet, but a mailing list/page probably make sense. > I'll update the porters page to add the mips port. Would you like to > keep the UltraViolet > name for the project? It's a nice idea, but UltraViolet is actually the Reservoir trademark for its Java technology, so I'm not sure it's appropriate. I was just thinking to call it the OpenJDK MIPS Port Project (perhaps bland, but descriptive). I'm sure we could think of something more unique if that seems useful though. -- Jonathan Springer | Reservoir Labs, Inc. | http://www.reservoir.com/ From mr at sun.com Wed Apr 2 08:49:22 2008 From: mr at sun.com (Mark Reinhold) Date: Wed, 02 Apr 2008 08:49:22 -0700 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: springer@reservoir.com; Tue, 01 Apr 2008 18:32:40 CDT; <47F2C618.9070209@reservoir.com> Message-ID: <20080402154923.098C55B658@eggemoggin.niobe.net> > Date: Tue, 01 Apr 2008 18:32:40 -0500 > From: Jonathan Springer > Dalibor Topic wrote: >> I'll update the porters page to add the mips port. Would you like to >> keep the UltraViolet >> name for the project? > > It's a nice idea, but UltraViolet is actually the Reservoir trademark > for its Java technology, so I'm not sure it's appropriate. I was just > thinking to call it the OpenJDK MIPS Port Project (perhaps bland, but > descriptive). I'm sure we could think of something more unique if that > seems useful though. I agree that "MIPS Port" would be a good name for this project, following the "Haiku Port" that we already have. "MIPS" is, however, a trademark of MIPS Technologies, Inc., so I've asked Sun Legal for advice on whether and how to use that name. I expect an answer in the next day or two, at which point we can go ahead and set up your initial page and mailing list. - Mark (NAL, but often feels like one) From dalibor.topic at googlemail.com Wed Apr 2 10:07:18 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Wed, 2 Apr 2008 19:07:18 +0200 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: <20080402154923.098C55B658@eggemoggin.niobe.net> References: <47F2C618.9070209@reservoir.com> <20080402154923.098C55B658@eggemoggin.niobe.net> Message-ID: <985bee770804021007q419f6c17s9b8bebfd9902fef9@mail.gmail.com> On Wed, Apr 2, 2008 at 5:49 PM, Mark Reinhold wrote: > I agree that "MIPS Port" would be a good name for this project, following > the "Haiku Port" that we already have. > > "MIPS" is, however, a trademark of MIPS Technologies, Inc., so I've asked > Sun Legal for advice on whether and how to use that name. I expect an > answer in the next day or two, at which point we can go ahead and set up > your initial page and mailing list. Thank you very much of taking care of all the legal questions, Mark. cheers, dalibor topic From james.childers at gmail.com Wed Apr 2 17:10:59 2008 From: james.childers at gmail.com (James Childers) Date: Wed, 2 Apr 2008 19:10:59 -0500 Subject: Getting Eclipse to play nicely with SoyLatte (OS X, Java6) Message-ID: <759D0D4C-8B5B-4C4F-BA14-005B8CD140C7@gmail.com> I needed to get Java6 up and running within Eclipse on OS X and was able to successfully use SoyLatte as a JRE. There has been some interest expressed in this, but a bit of balking at having to recompile Eclipse. It's actually not all that much work; instructions available here: http://macjavadev.vox.com/library/post/eclipse-os-x-and-soylatte.html HTH. -= J From mr at sun.com Thu Apr 3 08:29:15 2008 From: mr at sun.com (Mark Reinhold) Date: Thu, 03 Apr 2008 08:29:15 -0700 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: mr@sun.com; Wed, 02 Apr 2008 08:49:22 PDT; <20080402154923.098C55B658@eggemoggin.niobe.net> Message-ID: <20080403152915.2A91D5B656@eggemoggin.niobe.net> > From: Mark Reinhold > Date: Wed, 02 Apr 2008 08:49:22 -0700 > ... > > I agree that "MIPS Port" would be a good name for this project, following > the "Haiku Port" that we already have. > > "MIPS" is, however, a trademark of MIPS Technologies, Inc., so I've asked > Sun Legal for advice on whether and how to use that name. I expect an > answer in the next day or two, at which point we can go ahead and set up > your initial page and mailing list. The advice of our trademark counsel is to call this the "Porting Project for MIPS". The appearance of the mark at the end of the phrase, rather than the beginning, constitutes fair use. To make this work in the left nav bar on the web pages we can shorten it to "Port: MIPS", and for consistency also change "Haiku Port" to "Port: Haiku". This will cause all the porting projects to be listed together in the nav bar, which is kind of nice too. The Project URL will be http://openjdk.java.net/projects/mips-port. Okay? - Mark From springer at reservoir.com Thu Apr 3 08:42:33 2008 From: springer at reservoir.com (Jonathan Springer) Date: Thu, 03 Apr 2008 10:42:33 -0500 Subject: Call for vote [Fwd: New Project Proposal: MIPS port of OpenJDK] In-Reply-To: <20080403152915.2A91D5B656@eggemoggin.niobe.net> References: <20080403152915.2A91D5B656@eggemoggin.niobe.net> Message-ID: <47F4FAE9.30109@reservoir.com> Mark Reinhold wrote: >> From: Mark Reinhold >> Date: Wed, 02 Apr 2008 08:49:22 -0700 > >> ... >> >> I agree that "MIPS Port" would be a good name for this project, following >> the "Haiku Port" that we already have. >> >> "MIPS" is, however, a trademark of MIPS Technologies, Inc., so I've asked >> Sun Legal for advice on whether and how to use that name. I expect an >> answer in the next day or two, at which point we can go ahead and set up >> your initial page and mailing list. > > The advice of our trademark counsel is to call this the "Porting Project > for MIPS". The appearance of the mark at the end of the phrase, rather > than the beginning, constitutes fair use. > > To make this work in the left nav bar on the web pages we can shorten it > to "Port: MIPS", and for consistency also change "Haiku Port" to "Port: > Haiku". This will cause all the porting projects to be listed together > in the nav bar, which is kind of nice too. > > The Project URL will be http://openjdk.java.net/projects/mips-port. > > Okay? Sounds great. Thanks for checking all this out! -Jonathan -- Jonathan Springer | Reservoir Labs, Inc. | http://www.reservoir.com/ From ebresie at gmail.com Thu Apr 3 19:34:47 2008 From: ebresie at gmail.com (Eric Bresie) Date: Thu, 3 Apr 2008 21:34:47 -0500 Subject: Porting Java ME to Open JDK Message-ID: <113b56700804031934y3db1b67fi754b9534eef08ecf@mail.gmail.com> Okay guys...this may seem silly or crazy with most of the comments I have been getting seeming to not be very hot on the idea, but I wanted to post it none the less... I would like to propose porting (maybe more appropriate, merge of) Java ME to OpenJDK. Since the idea would involve migrating the ME platform specific (at least the VM) with some additional build logic (a build task to build only ME specific elements) into the Open JDK side, I thought posting here would be appropriate. Some of the benefits in favor of this are: - Avoids two separate source baselines - Performance increases in Java SE make it into Java ME profiles - Helps minimize some of the problems with Java ME fragmentation - More up to date Java ME based on the newer revisions of Java SE / OpenJDK source - Recent java kernel activity in the Java 6 N could use loading of needed bundles. This fit well with the idea of only loading Java ME platform specific components. However, since these efforts are in the Java SE side, the Java ME folks do not really get the benefits of this at the moment. If they were merged, then they would. - In this context, this would be a benefit for Java SE, because the small memory footprints associated with ME platform may force some use case to optimize memory at the Java SE level. I compare this to the Linux Kernel which has platform support for just about every size device from mainframes and 64bit platforms to pda devices and Arm platforms. Optimizations occur in the perspective device specific area. - There are optional JSR that are focused on mobile devices (I assume item such as cell phone interfaces, PIMs, etc). But with VOIP, GPS, and bluetooth capabilities all inclusive on laptops, I still don't see anything that would not be appropriate in the Java SE realm as well. - Lots of refactoring opportunities. Some of the issue against seem to involve some of the following: - The target profile of the Java ME have optimizations - Optimizations for the mobile platform verses the full platform. However the above migration task would also be to identify the optimizations and modularize them so that an interface could be defined such that it works on both and then implement the interface depending upon the above mentioned build task. - Power could be a problems. I believe this concerns that have been mention revolves around the VM and threading technologies could result in more power usage, but I'm not sure. Maybe this needs to be a JSR suggestion. I may probably over simplifying things. Comments? Eric Bresie ebresie at gmail.com From carfield at carfield.com.hk Fri Apr 4 11:15:07 2008 From: carfield at carfield.com.hk (Carfield Yim) Date: Sat, 5 Apr 2008 02:15:07 +0800 Subject: Getting Eclipse to play nicely with SoyLatte (OS X, Java6) In-Reply-To: <759D0D4C-8B5B-4C4F-BA14-005B8CD140C7@gmail.com> References: <759D0D4C-8B5B-4C4F-BA14-005B8CD140C7@gmail.com> Message-ID: Great information , thanks a lot! On Thu, Apr 3, 2008 at 8:10 AM, James Childers wrote: > I needed to get Java6 up and running within Eclipse on OS X and was able > to successfully use SoyLatte as a JRE. There has been some interest > expressed in this, but a bit of balking at having to recompile Eclipse. It's > actually not all that much work; instructions available here: > > http://macjavadev.vox.com/library/post/eclipse-os-x-and-soylatte.html > > HTH. > > -= J > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/porters-dev/attachments/20080405/0daf1787/attachment.html From trevor at vocaro.com Mon Apr 14 13:17:27 2008 From: trevor at vocaro.com (Trevor Harmon) Date: Mon, 14 Apr 2008 13:17:27 -0700 Subject: JNI not working on SoyLatte Message-ID: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> I cannot get JNI working with SoyLatte 1.0.2. The same code and JNI library works fine with the default Java 1.5 installation in Mac OS X 10.5.2, but switching to SoyLatte causes an UnsatisfiedLinkError. Apparently the C library is not loading correctly. I tried both the 32- bit and 64-bit versions of SoyLatte but got the same result. Is this a known issue? I didn't see anything about JNI on the SoyLatte project page. Trevor From swingler at apple.com Mon Apr 14 14:32:00 2008 From: swingler at apple.com (Mike Swingler) Date: Mon, 14 Apr 2008 14:32:00 -0700 Subject: JNI not working on SoyLatte In-Reply-To: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> References: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> Message-ID: On Apr 14, 2008, at 1:17 PM, Trevor Harmon wrote: > I cannot get JNI working with SoyLatte 1.0.2. The same code and JNI > library works fine with the default Java 1.5 installation in Mac OS > X 10.5.2, but switching to SoyLatte causes an UnsatisfiedLinkError. > Apparently the C library is not loading correctly. I tried both the > 32-bit and 64-bit versions of SoyLatte but got the same result. > > Is this a known issue? I didn't see anything about JNI on the > SoyLatte project page. Have you tried renaming the native file .dylib instead of .jnilib? Cheers, Mike Swingler Java Runtime Engineer Apple Inc. From trevor at vocaro.com Mon Apr 14 16:30:52 2008 From: trevor at vocaro.com (Trevor Harmon) Date: Mon, 14 Apr 2008 16:30:52 -0700 Subject: JNI not working on SoyLatte In-Reply-To: References: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> Message-ID: <8E232C76-59D7-4EEE-803A-09B396C81FB1@vocaro.com> On Apr 14, 2008, at 2:32 PM, Mike Swingler wrote: > Have you tried renaming the native file .dylib instead of .jnilib? Thanks for the suggestion, but it did not change anything for me. Trevor From sajjan.jyothisv at gmail.com Mon Apr 14 14:16:33 2008 From: sajjan.jyothisv at gmail.com (sajjan jyothi) Date: Mon, 14 Apr 2008 23:16:33 +0200 Subject: porting of jdk to mips Message-ID: Hi All, I would like to port jdk to mips.Anybody can help me to give some tips on it.I am totally new to jdk source bundle. -- Thanks & Regards Sajjan Jyothi -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/porters-dev/attachments/20080414/5e2b0167/attachment.html From springer at reservoir.com Mon Apr 14 18:05:35 2008 From: springer at reservoir.com (Jonathan Springer) Date: Mon, 14 Apr 2008 20:05:35 -0500 Subject: porting of jdk to mips In-Reply-To: References: Message-ID: <4803FF5F.5050906@reservoir.com> sajjan jyothi wrote: > Hi All, > I would like to port jdk to mips.Anybody can help me to give some tips > on it.I am totally new to jdk source bundle. What kind of MIPS, and on what OS? We developed an interpreter-only Linux MIPS64 port of OpenJDK (http://java.reservoir.com/). It runs Java code fine, though not yet quickly. Beyond a compiler, there are other tweaks that should be made, like more use of GOTs, etc. There is also IcedTea (http://fedoraproject.org/wiki/Features/IcedTea), which has a "zero assembler" patch that allows a C-interpreter configuration of OpenJDK to be built on other architectures with little/no extra work. -- Jonathan Springer | Reservoir Labs, Inc. | http://www.reservoir.com/ From glewis at eyesbeyond.com Mon Apr 14 18:56:08 2008 From: glewis at eyesbeyond.com (Greg Lewis) Date: Mon, 14 Apr 2008 18:56:08 -0700 Subject: JNI not working on SoyLatte In-Reply-To: <8E232C76-59D7-4EEE-803A-09B396C81FB1@vocaro.com> References: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> <8E232C76-59D7-4EEE-803A-09B396C81FB1@vocaro.com> Message-ID: <20080415015607.GA96433@misty.eyesbeyond.com> On Mon, Apr 14, 2008 at 04:30:52PM -0700, Trevor Harmon wrote: > On Apr 14, 2008, at 2:32 PM, Mike Swingler wrote: > > >Have you tried renaming the native file .dylib instead of .jnilib? > > Thanks for the suggestion, but it did not change anything for me. SoyLatte does require the library be .dylib instead of .jnilib, so thats at least part of the puzzle. If you want to find out what files its actually looking for in terms of the library you could run it under the MacOS X equivalent of ktrace and find out where its looking. You could also try setting LD_LIBRARY_PATH since IIRC it does search through there. -- Greg Lewis Email : glewis at eyesbeyond.com Eyes Beyond Web : http://www.eyesbeyond.com Information Technology FreeBSD : glewis at FreeBSD.org From trevor at vocaro.com Mon Apr 14 19:19:48 2008 From: trevor at vocaro.com (Trevor Harmon) Date: Mon, 14 Apr 2008 19:19:48 -0700 Subject: JNI not working on SoyLatte In-Reply-To: <20080415015607.GA96433@misty.eyesbeyond.com> References: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> <8E232C76-59D7-4EEE-803A-09B396C81FB1@vocaro.com> <20080415015607.GA96433@misty.eyesbeyond.com> Message-ID: <20D25C9F-9731-47E7-AC0B-E4B578E0137A@vocaro.com> On Apr 14, 2008, at 6:56 PM, Greg Lewis wrote: > SoyLatte does require the library be .dylib instead of .jnilib, so > thats > at least part of the puzzle. If you want to find out what files its > actually looking for in terms of the library you could run it under > the > MacOS X equivalent of ktrace and find out where its looking. Anybody know what the equivalent of ktrace is on Mac OS X? > You could also try setting LD_LIBRARY_PATH since IIRC it does search > through there. Yes, to get JNI to work on the bundled Apple port of Java 1.5, I have to set DYLD_LIBRARY_PATH to point to the directory containing the JNI library. However, neither DYLD_LIBRARY_PATH nor LD_LIBRARY_PATH seems to have any effect on SoyLatte. Trevor From landonf at bikemonkey.org Wed Apr 16 10:37:59 2008 From: landonf at bikemonkey.org (Landon Fuller) Date: Wed, 16 Apr 2008 10:37:59 -0700 Subject: JNI not working on SoyLatte In-Reply-To: <20D25C9F-9731-47E7-AC0B-E4B578E0137A@vocaro.com> References: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> <8E232C76-59D7-4EEE-803A-09B396C81FB1@vocaro.com> <20080415015607.GA96433@misty.eyesbeyond.com> <20D25C9F-9731-47E7-AC0B-E4B578E0137A@vocaro.com> Message-ID: <881E1CB2-4197-4A83-9C20-ED7829D76FE1@bikemonkey.org> On Apr 14, 2008, at 7:19 PM, Trevor Harmon wrote: > On Apr 14, 2008, at 6:56 PM, Greg Lewis wrote: > >> SoyLatte does require the library be .dylib instead of .jnilib, so >> thats >> at least part of the puzzle. If you want to find out what files its >> actually looking for in terms of the library you could run it under >> the >> MacOS X equivalent of ktrace and find out where its looking. > > Anybody know what the equivalent of ktrace is on Mac OS X? In Leopard, it would be dtrace (and dtruss). Tiger still has ktrace. >> You could also try setting LD_LIBRARY_PATH since IIRC it does >> search through there. > > Yes, to get JNI to work on the bundled Apple port of Java 1.5, I > have to set DYLD_LIBRARY_PATH to point to the directory containing > the JNI library. However, neither DYLD_LIBRARY_PATH nor > LD_LIBRARY_PATH seems to have any effect on SoyLatte. Do you have a stand-alone test case that works on Apple's JVM, but doesn't work on SoyLatte? Thanks, Landon -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.openjdk.java.net/pipermail/porters-dev/attachments/20080416/938cbaa2/attachment.bin From deb at freebsd.org Wed Apr 16 15:23:13 2008 From: deb at freebsd.org (Deb Goodkin) Date: Wed, 16 Apr 2008 16:23:13 -0600 Subject: BSD port In-Reply-To: <20080325170052.GB70511@misty.eyesbeyond.com> References: <20080325170052.GB70511@misty.eyesbeyond.com> Message-ID: <48067C51.7040309@freebsd.org> Hi Greg and Carla, I've read through the email and tried to understand what is needed for OpenJDK. First, I'd like to say that we fully support moving to OpenJDK and would like to do what is necessary to make this happen. The last major binary releases we did were for Java 1.5. We used the partner sources. The last two years, I believe Sun changed the license we are using to a JRL. I can't remember exactly why Sun requested us to go under that license, except for maybe because Sun was providing the license for free. We currently have a developer creating binaries using Java 1.6 from the partner sources. We are fine giving back the changes as long as Sun indemnifies us from any legal actions. We are under a license agreement to not make the source or changes available outside the foundation. So, we need to make sure we don't infringe on our license agreement. Please let me know what you need from the foundation to move forward. Sincerely, Deb Goodkin Director of Operations The FreeBSD Foundation Greg Lewis wrote: > G'day Carla, > > On Thu, Mar 20, 2008 at 03:33:14PM -0700, Carla Schroer wrote: >> Once again I find myself apologizing for the delay in responding. >> With licensing issues, we have to get legal advice, and it can be >> difficult to get time with our attorneys because they have a lot on >> their plates. I have a few more comments and questions below, and I >> appreciate you providing the information that you have here. > > No worries, I'm aware of how long it can take to get answers from legal > departments :). > >>> G'day Carla, >>> >>> On Thu, Feb 14, 2008 at 11:37:34AM -0800, Carla.Schroer at Sun.COM wrote: >>> >>>> Bringing the BSD port into the OpenJDK Community under the GPL >>>> license is something that Sun would very much like to see happen. >>>> >>> The BSD Java community would also very much like to see this happen, so >>> it seems like we're in complete agreement so far :). >>> >>> >>>> We believe that the work that's so far been done on the BSD port is >>>> based on code obtained under the Sun Community Source License >>>> (SCSL). We need to understand if any of the work was done on code >>>> obtained under a different license, such as the Java Research >>>> License (JRL). The SCSL license does provide for code modifications >>>> to be given back to Sun with sufficient rights and does not require >>>> a Sun Contributor Agreement (SCA) in order to do so. So we do need >>>> to make sure that we are getting the code back under SCSL. >>>> >>> Work on the part was done under the SCSL for the 1.2, 1.3, 1.4 and 1.5 >>> (initially) releases. I mention the previous versions since many >>> changes >>> would be forward ported from one version to the next. Once the JRL >>> update >>> releases commenced we started doing work based on them though, so those >>> changes would be under the JRL. >>> >>> However, everyone who did work under the JRL has agreed to release there >>> changes under the GPL as well. I need to check with the FreeBSD >>> Foundation >>> though on work done for them under contract, but I believe they would be >>> happy for those changes to be released under the GPL too. >>> >> Thanks for the clarification about who did the work and under what >> licenses. The JRL has different terms than SCSL. We think the >> simplest solution for getting all the BSD port code back to us with >> the rights we need, would be for us to draft a document that could be >> signed by the folks who worked on this under the JRL, and also FreeBSD >> Foundation, if they also have rights in the code you wish to donate. > > That would work. I already have email agreements that they are willing > to release changes under the GPL, so we should be able to obtain a signed > statement from everyone who has done work under the JRL (which is a much > smaller list than those that have done SCSL changes). > > One problem might be people who have contributed very small fixes on a > one time basis. Thankfully this is a small number for the JRL and we'll > deal with that when the time comes. > > The FreeBSD Foundation do have a stake in this as they have rights to some > of the changes Kurt made under contract with them. I've cc'ed Deb Goodkin > of the Foundation (I've already raised the matter with her previously and > I believe its under their consideration). > >>>> In order to minimize the engineering effort for Sun, we need to work >>>> out a way for you to provide diffs to us that correspond to the >>>> latest OpenJDK code rather than the SCSL'd code with which you >>>> started, since the OpenJDK code has already been cleared for >>>> distribution under the GPL. We would like to work with you on the >>>> best way to do this so that the code can be published in an OpenJDK >>>> project as soon as possible. >>>> >>> We can do that. I see OpenJDK 6 has been released very recently, so we >>> haven't ported to that yet. For OpenJDK 7 we already have a port based >>> on an earlier build and that is being updated to the current build. >>> >>> If OpenJDK 6 is part of this discussion (it might just be OpenJDK 7 >>> you're >>> talking about :) then, assuming its source base is similar to the JRL >>> source base, we should be able to port to it quickly given that we have>> a working port based on the JRL source. >>> >>> >> I think we could do 6 or 7. If you want a 6 port for BSD, then it >> seems like that's the way to go. We see this process as a one time >> thing to get the code back into Sun and out again as part of OpenJDK >> under the GPL. Then the group could work with the 6 open code and the >> 7 code to get a BSD port for 7. Then we wouldn't have to go through >> any of this again. We are going to need to work out the best way to >> get your diffs that minimize our engineering effort to put the code >> back out. I think Mark Reinhold will be driving that part of the discussion, once we get past the license issues. > > We'll look at whether to target 6 or 7 once the license issues are > resolved. But yes, we'd hope this would be a one time process that > would get the code back into OpenJDK and that the current porting > project would then shut up shop in favour of all future work being > based around OpenJDK. > > We already produce diffs and have endeavoured to add BSD support without > affect Solaris/Windows/Linux in any way although its been a while since > I compiled our source base on any of these (I have compiled it under > Linux once or twice). > >>>> It would also help us if someone that worked on this port could >>>> provide the name of the specific entity (or individuals) that signed >>>> the SCSL agreement where this work took place, as well as any other >>>> licenses the work was done under. >>> >>> I can get together a list of names of the people who have contributed. >>> However, the SCSL was never "signed" as such. One agreed to the license >>> by clicking on a button with something like "I Accept" on it when >>> downloading the SCSL based source releases. So while I can provide a >>> list >>> of names, I can't provide any paperwork. >>> >> You are correct that SCSL isn't signed if you want the code for >> internal use or research use. The SCSL is signed if someone wants >> the "commercial attachment" to distribute products based on the code. >> FreeBSD did sign a SCSL commercial attachment as well as a TCK >> license. So, we are going to need to get FreeBSD Foundation involved in giving the code back to us. > > Ok. I'll wait to hear from the Foundation on this. > From trevor at vocaro.com Thu Apr 17 18:29:27 2008 From: trevor at vocaro.com (Trevor Harmon) Date: Thu, 17 Apr 2008 18:29:27 -0700 Subject: JNI not working on SoyLatte In-Reply-To: <881E1CB2-4197-4A83-9C20-ED7829D76FE1@bikemonkey.org> References: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> <8E232C76-59D7-4EEE-803A-09B396C81FB1@vocaro.com> <20080415015607.GA96433@misty.eyesbeyond.com> <20D25C9F-9731-47E7-AC0B-E4B578E0137A@vocaro.com> <881E1CB2-4197-4A83-9C20-ED7829D76FE1@bikemonkey.org> Message-ID: On Apr 16, 2008, at 10:37 AM, Landon Fuller wrote: > Do you have a stand-alone test case that works on Apple's JVM, but > doesn't work on SoyLatte? Yes. Attached is a simple JNI test case. To run: javac HelloWorld.java javah -jni HelloWorld gcc -dynamiclib -I$JAVA_HOME/include HelloWorld.c -o libHelloWorld.dylib java HelloWorld This works fine under Mac OS X 10.5.2 with the built-in Java 1.5.0_13. However, if you then point JAVA_HOME to the location of SoyLatte 1.0.2 and make SoyLatte's bin directory the first element in the PATH, it doesn't work. I get "java.lang.UnsatisfiedLinkError: no HelloWorld in java.library.path" Trevor -------------- next part -------------- A non-text attachment was scrubbed... Name: HelloWorld.java Type: application/octet-stream Size: 177 bytes Desc: not available Url : http://mail.openjdk.java.net/pipermail/porters-dev/attachments/20080417/57e07c07/attachment.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: HelloWorld.c Type: application/octet-stream Size: 166 bytes Desc: not available Url : http://mail.openjdk.java.net/pipermail/porters-dev/attachments/20080417/57e07c07/attachment-0001.obj From landonf at bikemonkey.org Thu Apr 17 18:40:16 2008 From: landonf at bikemonkey.org (Landon Fuller) Date: Thu, 17 Apr 2008 18:40:16 -0700 Subject: JNI not working on SoyLatte In-Reply-To: References: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> <8E232C76-59D7-4EEE-803A-09B396C81FB1@vocaro.com> <20080415015607.GA96433@misty.eyesbeyond.com> <20D25C9F-9731-47E7-AC0B-E4B578E0137A@vocaro.com> <881E1CB2-4197-4A83-9C20-ED7829D76FE1@bikemonkey.org> Message-ID: <29E3C8A4-AA88-49BB-A420-D55B77ACE062@bikemonkey.org> On Apr 17, 2008, at 6:29 PM, Trevor Harmon wrote: > On Apr 16, 2008, at 10:37 AM, Landon Fuller wrote: > >> Do you have a stand-alone test case that works on Apple's JVM, but >> doesn't work on SoyLatte? > > Yes. Attached is a simple JNI test case. To run: > > javac HelloWorld.java > javah -jni HelloWorld > gcc -dynamiclib -I$JAVA_HOME/include HelloWorld.c -o > libHelloWorld.dylib > java HelloWorld > > This works fine under Mac OS X 10.5.2 with the built-in Java > 1.5.0_13. However, if you then point JAVA_HOME to the location of > SoyLatte 1.0.2 and make SoyLatte's bin directory the first element > in the PATH, it doesn't work. I get "java.lang.UnsatisfiedLinkError: > no HelloWorld in java.library.path" Thanks! landonf at max:~/Desktop/JNI> java -Djava.library.path=`pwd` HelloWorld Hello World! It looks like Apple's java includes '.' in java.library.path by default: scala> System.getProperty("java.library.path"); res0: java.lang.String = .:/Library/Java/Extensions:/System/Library/ Java/Extensions:/usr/lib/java Whereas SoyLatte: scala> System.getProperty("java.library.path"); res0: java.lang.String = /usr/local/soylatte16-i386-1.0.2/jre/lib/ i386/server:/usr/local/soylatte16-i386-1.0.2/jre/lib/i386:/usr/local/ soylatte16-i386-1.0.2/jre/../lib/i386:/usr/java/packages/lib/i386:/ lib:/usr/lib Additionally, SoyLatte doesn't seem to be respecting DYLD_LIBRARY_PATH. -landonf -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.openjdk.java.net/pipermail/porters-dev/attachments/20080417/0b88a9da/attachment.bin From trevor at vocaro.com Sun Apr 20 03:30:28 2008 From: trevor at vocaro.com (Trevor Harmon) Date: Sun, 20 Apr 2008 03:30:28 -0700 Subject: JNI not working on SoyLatte In-Reply-To: <29E3C8A4-AA88-49BB-A420-D55B77ACE062@bikemonkey.org> References: <04671351-F71F-472A-B7E7-7A04CDF6E971@vocaro.com> <8E232C76-59D7-4EEE-803A-09B396C81FB1@vocaro.com> <20080415015607.GA96433@misty.eyesbeyond.com> <20D25C9F-9731-47E7-AC0B-E4B578E0137A@vocaro.com> <881E1CB2-4197-4A83-9C20-ED7829D76FE1@bikemonkey.org> <29E3C8A4-AA88-49BB-A420-D55B77ACE062@bikemonkey.org> Message-ID: <94490E01-C7C0-47D6-95B6-02384ED249BB@vocaro.com> On Apr 17, 2008, at 6:40 PM, Landon Fuller wrote: > landonf at max:~/Desktop/JNI> java -Djava.library.path=`pwd` HelloWorld > Hello World! This bug has exposed another. My Ant script is, in fact, adding the appropriate directory to java.library.path, but it depends on the host OS, and somehow SoyLatte is reporting the wrong OS. This problem can be reproduced with the attached Ant script. With Apple's Java 1.5, it prints: You are on a Mac. true But with SoyLatte it prints: You are on a Mac. false > Additionally, SoyLatte doesn't seem to be respecting > DYLD_LIBRARY_PATH. This is also a problem for me. Would it be hard to fix? Trevor -------------- next part -------------- A non-text attachment was scrubbed... Name: build.xml Type: text/xml Size: 240 bytes Desc: not available Url : http://mail.openjdk.java.net/pipermail/porters-dev/attachments/20080420/d9b0b1a4/attachment.xml -------------- next part -------------- From dalibor.topic at googlemail.com Mon Apr 21 06:48:54 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Mon, 21 Apr 2008 15:48:54 +0200 Subject: Porting Java ME to Open JDK In-Reply-To: <113b56700804031934y3db1b67fi754b9534eef08ecf@mail.gmail.com> References: <113b56700804031934y3db1b67fi754b9534eef08ecf@mail.gmail.com> Message-ID: <985bee770804210648g2dc8c283v8cc70bb6b64312dc@mail.gmail.com> On Fri, Apr 4, 2008 at 4:34 AM, Eric Bresie wrote: > I would like to propose porting (maybe more appropriate, merge of) > Java ME to OpenJDK. > > Since the idea would involve migrating the ME platform specific (at > least the VM) with some additional build logic (a build task to build > only ME specific elements) into the Open JDK side, I thought posting > here would be appropriate. We looked at generating different 'profiles' of GNU Classpath every now and then, and it turned out to be non-trivial, since there are differences in core library behavior from SE release to SE release that would be quite hard to maintain in the same tree, without resorting to a lot of preprocessing. I would imagine the boundary between SE and ME to be much higher, so that would seem to me to require a significant amount of non-trivial build logic to accomplish, if you are after building certifiably compatible ME and SE releases from the same tree. A simpler project could be to try to make the OpenJDK class library work with the PhoneME VM. cheers, dalibor topic From glegris at thenesis.org Tue Apr 22 01:26:44 2008 From: glegris at thenesis.org (Guillaume Legris) Date: Tue, 22 Apr 2008 10:26:44 +0200 Subject: Porting Java ME to Open JDK Message-ID: <4a89b0860804220126g165956cdtaeae7b1ffc4d23ee@mail.gmail.com> Hi, I'm the maintainer of the MIDPath [1] project. MIDPath provides Java ME to Java SE environments and also works with the CLDC version of the CACAO VM [2] MIDPath is based extensively on the phoneME code. The KNI native code was replaced by JNI or pure Java code. MIDPath could be used as an implementation of the proposal in reusing totality or parts of its codebase. We would be glad to help in any way we can. Regards, Guillaume Legris [1] http://midpath.thenesis.org [2] http://www.cacaojvm.org From ebresie at gmail.com Tue Apr 22 14:33:33 2008 From: ebresie at gmail.com (Eric Bresie) Date: Tue, 22 Apr 2008 16:33:33 -0500 Subject: Porting Java ME to Open JDK In-Reply-To: <985bee770804210648g2dc8c283v8cc70bb6b64312dc@mail.gmail.com> References: <113b56700804031934y3db1b67fi754b9534eef08ecf@mail.gmail.com> <985bee770804210648g2dc8c283v8cc70bb6b64312dc@mail.gmail.com> Message-ID: <113b56700804221433l6265a732rcf2a719834325df5@mail.gmail.com> I am sure I'm over simplifying this so I appreciate all your help to help me understand some of this. On 4/21/08, Dalibor Topic wrote: > On Fri, Apr 4, 2008 at 4:34 AM, Eric Bresie wrote: > > I would like to propose porting (maybe more appropriate, merge of) > > Java ME to OpenJDK. > > > > Since the idea would involve migrating the ME platform specific (at > > least the VM) with some additional build logic (a build task to build > > only ME specific elements) into the Open JDK side, I thought posting > > here would be appropriate. > > We looked at generating different 'profiles' of GNU Classpath every > now and then, > and it turned out to be non-trivial, since there are differences in core library > behavior from SE release to SE release that would be quite hard to maintain > in the same tree, without resorting to a lot of preprocessing. I would imagine What kind of differences and in what core libraries? Would they be good candidates for developing some wrapper interfaces to hide these differences? Stuff like graphic tool kits and audio toolkits could be abstracted and then implementation specific would make use of these? > the boundary between SE and ME to be much higher, so that would seem to This is probably a question I need to read up some more, but maybe you can answer in a few words. Isn't ME based on SE 1.4.x source? What kind of divergence (besides the VMs, hardware specific optimizations, and subsets or extensions for use with small footprints) occurred? I always thought ME was basically a subset of SE with a smaller footprint VM (for small resource devices). > me to require a significant amount of non-trivial build logic to accomplish, if > you are after building certifiably compatible ME and SE releases from the same > tree. That was basically my thought. Kind of like the linux kernel which has platform specific areas with the basic kernel abstracted on top of it with defined interfaces/apis. > A simpler project could be to try to make the OpenJDK class library work > with the PhoneME VM. > > cheers, > dalibor topic My thinking was more to write as much in java as low as possible and refactor out of the VM level to try and find common interfaces that are implemented as much into usable components as possible then isolate the platform specific in the VM as the location of the differences. So to integrate the phoneME VM into the OpenJDK with My thought here in the end was that the Java ME represents a whole new suite of use cases for Java that could help in optimizing Java as a whole to reduce memory usage and improve performance for lesser performing platforms and avoid having to maintain 2 separate source trees. It also allows the ME to leverage off the Java 5/6/7 a lot sooner. By the way congradulation on you new Java job by the way... -- Eric Bresie ebresie at gmail.com From twisti at complang.tuwien.ac.at Wed Apr 23 00:17:06 2008 From: twisti at complang.tuwien.ac.at (Christian Thalinger) Date: Wed, 23 Apr 2008 09:17:06 +0200 Subject: Porting Java ME to Open JDK In-Reply-To: <113b56700804221433l6265a732rcf2a719834325df5@mail.gmail.com> References: <113b56700804031934y3db1b67fi754b9534eef08ecf@mail.gmail.com> <985bee770804210648g2dc8c283v8cc70bb6b64312dc@mail.gmail.com> <113b56700804221433l6265a732rcf2a719834325df5@mail.gmail.com> Message-ID: <1208935026.31958.7.camel@imac523d.theobroma-systems.com> On Tue, 2008-04-22 at 16:33 -0500, Eric Bresie wrote: > This is probably a question I need to read up some more, but maybe you > can answer in a few words. Isn't ME based on SE 1.4.x source? What > kind of divergence (besides the VMs, hardware specific optimizations, > and subsets or extensions for use with small footprints) occurred? I > always thought ME was basically a subset of SE with a smaller > footprint VM (for small resource devices). I'm very sure it's not based on the SE source. I have implemented three different VM interfaces in CACAO[1]: 1. GNU Classpath 2. phoneME CLDC-1.1 3. OpenJDK And the OpenJDK one is very different to the others. There are a lot of smaller or bigger hacks in HotSpot do get things done properly and I'm not sure how easy it will be to get them out. But I think it could be done. The more important question is, is it worth to tailor HotSpot to be able to run phoneME classes with its powerful (and memory using) JIT compilers? [1] http://www.cacaovm.org/ - twisti From alvim at stud.ntnu.no Wed Apr 23 05:26:40 2008 From: alvim at stud.ntnu.no (alvim at stud.ntnu.no) Date: Wed, 23 Apr 2008 14:26:40 +0200 Subject: Porting OpenJDK to ARMEL Message-ID: <20080423142640.ev62nb5r8gg84ogs@webmail.ntnu.no> Hello! I'm trying do port OpenJDK to an ARMEL device, more known as Nokia N810 with Maemo os2008. While doing this I am running into a lot of dependencies. Have managed to install some of those manually, but I am not in goal with this project yet. As you see below, I am using the Scratchbox as an development environment. Anyone who knows how this can be done in an easy an convenient way? [sbox-CHINOOK_ARMEL: ~/pakka/openjdk/openjdk-6-6b09] > dpkg-buildpackage dpkg-buildpackage: source package is openjdk-6 dpkg-buildpackage: source version is 6b09-0ubuntu1 dpkg-buildpackage: source changed by Matthias Klose dpkg-buildpackage: host architecture armel dpkg-buildpackage: source version without epoch 6b09-0ubuntu1 : Using Scratchbox tools to satisfy builddeps : Dependency provided by Scratchbox: wget : Dependency provided by Scratchbox: zip : Dependency provided by Scratchbox: unzip : Dependency provided by Scratchbox: gawk : Dependency provided by Scratchbox: automake : Dependency provided by Scratchbox: automake dpkg-checkbuilddeps: Unmet build dependencies: ant gcj (>= 4:4.2.1) java-gcj-compat-dev (>= 1.0.76-2ubuntu3) libxp-dev libcupsys2-dev lesstif2-dev libxalan2-java liblcms1-dev mauve xvfb dpkg-buildpackage: Build dependencies/conflicts unsatisfied; aborting. dpkg-buildpackage: (Use -d flag to override.) From aph at redhat.com Wed Apr 23 05:41:30 2008 From: aph at redhat.com (Andrew Haley) Date: Wed, 23 Apr 2008 13:41:30 +0100 Subject: Porting OpenJDK to ARMEL In-Reply-To: <20080423142640.ev62nb5r8gg84ogs@webmail.ntnu.no> References: <20080423142640.ev62nb5r8gg84ogs@webmail.ntnu.no> Message-ID: <480F2E7A.7020506@redhat.com> alvim at stud.ntnu.no wrote: > I'm trying do port OpenJDK to an ARMEL device, more known as Nokia N810 > with Maemo os2008. > > While doing this I am running into a lot of dependencies. Have managed > to install some of those manually, but I am not in goal with this > project yet. > As you see below, I am using the Scratchbox as an development environment. > > Anyone who knows how this can be done in an easy an convenient way? You're going to have trouble. 1. OpenJDK is not set up to cross-compile. It builds its tools natively. Fixing this would be nice, but a huge task. 2. Assuming you're happy to use the C++ interpeter, it should already build on ARM if you're using the IcedTea build harness. 3. Although I can get it to build, and it does work, it still doesn't run stably. I don't know why this is. Did you want to do a real port, with a template interpreter, or a JIT, or did you just want the C++ interpreter? Andrew. From twisti at complang.tuwien.ac.at Wed Apr 23 05:48:10 2008 From: twisti at complang.tuwien.ac.at (Christian Thalinger) Date: Wed, 23 Apr 2008 14:48:10 +0200 Subject: Porting OpenJDK to ARMEL In-Reply-To: <480F2E7A.7020506@redhat.com> References: <20080423142640.ev62nb5r8gg84ogs@webmail.ntnu.no> <480F2E7A.7020506@redhat.com> Message-ID: <1208954890.31958.33.camel@imac523d.theobroma-systems.com> On Wed, 2008-04-23 at 13:41 +0100, Andrew Haley wrote: > 1. OpenJDK is not set up to cross-compile. > It builds its tools natively. Fixing this would be nice, but a > huge task. Well, it kind of works. I cross-compiled IcedTea for ARM[1], although it did not build the whole thing as I'm missing some libraries and headers. But yes, there are problems. [1] http://www.advogato.org/person/twisti/diary/11.html - twisti From aph at redhat.com Wed Apr 23 05:55:43 2008 From: aph at redhat.com (Andrew Haley) Date: Wed, 23 Apr 2008 13:55:43 +0100 Subject: Porting OpenJDK to ARMEL In-Reply-To: <1208954890.31958.33.camel@imac523d.theobroma-systems.com> References: <20080423142640.ev62nb5r8gg84ogs@webmail.ntnu.no> <480F2E7A.7020506@redhat.com> <1208954890.31958.33.camel@imac523d.theobroma-systems.com> Message-ID: <480F31CF.4070104@redhat.com> Christian Thalinger wrote: > On Wed, 2008-04-23 at 13:41 +0100, Andrew Haley wrote: >> 1. OpenJDK is not set up to cross-compile. >> It builds its tools natively. Fixing this would be nice, but a >> huge task. > > Well, it kind of works. I cross-compiled IcedTea for ARM[1], although > it did not build the whole thing as I'm missing some libraries and > headers. > > But yes, there are problems. Sure, but you didn't build it, at least not according to that blog entry -- you just took a j2re-image from an x86_64 build and cross-compiled CACAO. So what part of OpenJDK itself was cross-compiled? Andrew. From twisti at complang.tuwien.ac.at Wed Apr 23 05:59:56 2008 From: twisti at complang.tuwien.ac.at (Christian Thalinger) Date: Wed, 23 Apr 2008 14:59:56 +0200 Subject: Porting OpenJDK to ARMEL In-Reply-To: <480F31CF.4070104@redhat.com> References: <20080423142640.ev62nb5r8gg84ogs@webmail.ntnu.no> <480F2E7A.7020506@redhat.com> <1208954890.31958.33.camel@imac523d.theobroma-systems.com> <480F31CF.4070104@redhat.com> Message-ID: <1208955596.31958.41.camel@imac523d.theobroma-systems.com> On Wed, 2008-04-23 at 13:55 +0100, Andrew Haley wrote: > Christian Thalinger wrote: > > On Wed, 2008-04-23 at 13:41 +0100, Andrew Haley wrote: > >> 1. OpenJDK is not set up to cross-compile. > >> It builds its tools natively. Fixing this would be nice, but a > >> huge task. > > > > Well, it kind of works. I cross-compiled IcedTea for ARM[1], although > > it did not build the whole thing as I'm missing some libraries and > > headers. > > > > But yes, there are problems. > > Sure, but you didn't build it, at least not according to that blog > entry -- you just took a j2re-image from an x86_64 build and cross-compiled > CACAO. So what part of OpenJDK itself was cross-compiled? The native libraries. - twisti From tiansm at lemote.com Fri Apr 25 07:00:34 2008 From: tiansm at lemote.com (Songmao Tian) Date: Fri, 25 Apr 2008 22:00:34 +0800 Subject: about mips-port Message-ID: <4811E402.4010802@lemote.com> Hello all, I am a developer of Loongson[1] platform from Lemote Tech Inc. which based in China. Loongson is a mips64 variant. It's great that OpenJDK to start the mips-port, special thanks to Jonathan Springer, that's really great. And Lemote has pursued java port for a long time. There's jdk 1.5 working on Loongson, I have been using it for more than a year, but it's a pity that we don't have enough resource to push open jdk porting, so it's really great to hear that the work has been done! :-) Is there more plan about developing? I think some of our guys will likely join in:-). Songmao [1] http://www.linux-mips.org/wiki/Loongson From springer at reservoir.com Fri Apr 25 15:52:51 2008 From: springer at reservoir.com (Jonathan Springer) Date: Fri, 25 Apr 2008 17:52:51 -0500 Subject: about mips-port In-Reply-To: <4811E402.4010802@lemote.com> References: <4811E402.4010802@lemote.com> Message-ID: <481260C3.8050206@reservoir.com> Songmao Tian wrote: > Hello all, > I am a developer of Loongson[1] platform from Lemote Tech Inc. which > based in China. Loongson is a mips64 variant. It's great that OpenJDK to > start the mips-port, special thanks to Jonathan Springer, that's really > great. > And Lemote has pursued java port for a long time. There's jdk 1.5 > working on Loongson, I have been using it for more than a year, but it's > a pity that we don't have enough resource to push open jdk porting, so > it's really great to hear that the work has been done! :-) > Is there more plan about developing? I think some of our guys will > likely join in:-). Hi Songmao, Yes, I have been working on a MIPS64 port, and the basic interpreter functionality is there now. We are hoping to get going on the next stage, which is to look at performance, before too much longer. We have recently gotten an OpenJDK MIPS port project created[1], which should help, so it is good timing. It has a mailing list, mips-port (which hasn't been used yet since no one outside of Reservoir Labs has been working on this code so far). If you would like to help out, that sounds great. I need to figure out the best way to collaborate. I will work on setting something up that will work with our process here and hopefully take advantage of the support that OpenJDK offers. Thanks, -Jonathan [1] http://mail.openjdk.java.net/pipermail/announce/2008-April/000038.html -- Jonathan Springer | Reservoir Labs, Inc. | http://www.reservoir.com/