All,
I have successfully ported coreboot to the relatively modern ASUS KGPE-D16 server board (dual AMD socket G34, 16 DDR3 DIMMs, https://www.asus.com/us/Commercial_Servers_Workstations/KGPED16/)! This port uses native Family 10h initialization (_not_ AGESA or CIMX).
The Libreboot folks will be interested to know that this board can run blob-free and still retain full functionality!
Port specifications: CPU: Dual AMD G34 Magny-Cours (Family 10h) RAM: 16 DDR3 DIMM slots with ECC support (tested with x4 4G DDR3-1333 unbuffered DIMMs)
Peripherals: PCIe slots: all functional PCI slot: functional RS-232: functional PS/2: expected to function, not tested (on SuperIO) ASpeed VGA device: functional (text mode, see below) IEEE1394: functional On-board USB: functional On-board NICs: functional ASUS PIKE SAS controller: functional PCIe ROMs: functional
Power management: DDR3 voltage set: functional ACPI/APIC: functional Suspend/resume: broken
Other: cbmem console: partial support (log truncated) cbmem timestamps: functional nvram: functional BIOS recovery jumper: functional
ASpeed VGA: The ASpeed VGA device initialises in text mode via its (new) coreboot driver, however this initialisation is incomplete, leading to distorted but quite usable VGA output. When Linux boots and engages the graphical framebuffer all distortion disappears.
This port was not trivial. Almost every device used was broken and required debugging/repair, with the notable exception of the SuperIO chip. The AMD DDR3 controller was severely broken to the point where large rewrites were needed in order to bring it in line with the BKDG. Even after the various component drivers were repaired
Due to the labor-intensive nature of the port and the extensive changes throughout the entire source tree, it is not economically feasible to merge this port upstream at this time (I estimate upward of 30 independent patches would be required just to get the board booting!). Raptor Engineering will, however, be continuing to maintain this port internally, and I am currently looking into adding native Family 15h support on top of this internal tree. Additionally, while it was not a priority for the initial port, I will be attempting to enable suspend/resume functionality as I have time.
If there is sufficient interest from the community in adding this board to coreboot I would consider merging the changes in exchange for a one-time contract payment in the vicinity of $35,000 USD. When considering this offer please bear in mind that this is a fully functional blobless board with a wide range of peripherals and expansion options available, and that once these large changes are merged I will continue to enhance coreboot functionality as before (e.g. with the KFSN4-DRE and the T400). I would also be willing to add this board to the test stand as the only fully supported 4-way Opteron board (socket G34 Magny-Cours CPUs contain two separate CPUs in one package, making this 2-socket board a 4-way system from a HyperTransport perspective).
Please let me know if you have any questions!
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
You should crowd-fund the $35,000 figure, there are lots of people who will be interested in this. I personally will chip in, and I'd ask others to as well.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 29/04/15 22:46, The Gluglug wrote:
You should crowd-fund the $35,000 figure, there are lots of people who will be interested in this. I personally will chip in, and I'd ask others to as well.
What about simply pushing the code as-is (make your non-upstream tree publicly available for people to git-clone), and let the community upstream it in their own time?
On 04/29/2015 04:56 PM, The Gluglug wrote:
What about simply pushing the code as-is (make your non-upstream tree publicly available for people to git-clone), and let the community upstream it in their own time?
There is no real incentive for Raptor to do this. There is no guarantee of code actually being up-streamed and maintained by the community, but after this release our competitors may use the code at will.
To be clear, we use coreboot internally on a large server cluster. This port is the result of an internal effort to upgrade said cluster, and while it is in our financial interest to upstream ports of older boards (e.g. the ASUS KFSN4-DRE) that did not require changes to the underlying support code, it is _not_ in our financial interest to simply release the code to modern boards that required major changes to underlying support code.
On Wed, Apr 29, 2015 at 10:46:29PM +0100, The Gluglug wrote:
You should crowd-fund the $35,000 figure, there are lots of people who will be interested in this. I personally will chip in, and I'd ask others to as well.
I would chip in too.
Thanks, Ward.
I'm curious, is there some Coreboot Foundation that would gather this money and purchase the copyright or would Raptor just somehow crowd-fund this money to license their work as GPLv2 while keeping the copyright?
--emi
On Thu, Apr 30, 2015 at 4:54 AM, Ward Vandewege ward@gnu.org wrote:
On Wed, Apr 29, 2015 at 10:46:29PM +0100, The Gluglug wrote:
You should crowd-fund the $35,000 figure, there are lots of people who will be interested in this. I personally will chip in, and I'd ask others to as well.
I would chip in too.
Thanks, Ward.
-- Ward Vandewege GPG Key: 25F774AB
Do you use free software? Donate to join the FSF and support freedom at http://www.fsf.org/register_form?referrer=859
-- coreboot mailing list: coreboot@coreboot.org http://www.coreboot.org/mailman/listinfo/coreboot
Dear coreboot folks,
Timothy, congratulations again on making a coreboot port for the ASUS KGPE-D16 and therefore completing your third coreboot port! That’s really amazing!
Am Mittwoch, den 29.04.2015, 22:46 +0100 schrieb The Gluglug:
You should crowd-fund the $35,000 figure, there are lots of people who will be interested in this. I personally will chip in, and I'd ask others to as well.
I am thinking about organizing the crowdfunding campaign to raise the money.
If somebody else wants to do it, please speak up!
As this is more or less a donation by the backers, the process should be as open and transparent as possible. That’s why I am sharing the following information publicly.
1. Giant Monkey Software Engineering [1], the German company I work for, would be the organizing entity. As a company with five employees and a GmbH it might have enough credibility so that people would pledge/give their money to Giant Monkey compared to a private person or company run by a single person. Giant Monkey also has some PR/campaign knowledge, but most importantly knows a lot of people in the marketing sector.
2. After receiving the money, Giant Monkey would contract Raptor Engineering.
3. I’d like to have the domain campaign.coreboot.org redirect to the campaign page at the crowdfunding platform or set up a simple Web site there.
4. git-annex’ second funding was done by itself with PayPal. That saves the 4 % fee most other crowdfunding platforms charge.
Using a crowdfunding platform might be easier though, as they have experience and also provide a big community of possible backers.
Currently I’m thinking about Indiegogo, which Jolla also used to fund the Jolla Tablet. I heard, Kickstarter is also great with a big network.
5. I plan to raise 100.000 € (around $110.000) to upstream, that includes *paid review* and running the campaign, the whole port. (I’ll continue to use Euros.) More money would be used for stretch goals.
α) 4.000 € Indiegogo fees β) 35.000 € for Raptor Engineering for upstreaming for basic port γ) 15.000 € for Raptor Engineering for implementing support for S2R (S3) δ) 10.000 € for code review (inclusively hardware) (just an estimate) ε) 2.000 € for Gluglug to release images (I have not talked to Francis yet.) φ) 10.000 € campaign goodies (cf. 7.) ι) 4.000 € taxes (probably a lot more, depends if given money counts as donation) μ) 20.000 € Giant Monkey for running the campaign (Web site, press, marketing, videos, mile stone tasks (see below), …)
(Stretch goal) ν) 15.000 € for Raptor Engineering for upstreaming Family 15h support for the board
6. I’d start with a minimal Web site and campaign platform page and see how big the momentum alone through the coreboot community is. If we get 10.000 € in a week, I’d fully step in with a professional campaign. Otherwise I’d stop the campaign.
7. As a thank you for backers, I think of a payload included in the distributed coreboot based firmware image, reading a text file from CBFS with the names of the backers and displaying it. (Or a simple splash screen.) Of course just for those wanting it. Big backers (25.000 €) get a board with one CPU and RAM and coreboot preinstalled; medium backers (10.000 €) get some BLOB free laptop for example (Rockchip Chromebook or some Lenovo board). flash ROM chips are sent to backers donating 25 €.
8. Milestone tasks: At certain mile stones (probably each 10.000 €), I’d promise some more tasks to improve coreboot or the port (see the 5.000 € steps in the top). Possible are also SSL certificates for coreboot infrastructure, promising to run a 32-bit userspace build host, redesigning the Web site, implementing CBMEM time stamp support in SeaBIOS and GRUB, supporting Google’s verified boot, ….
9. Reasons for contributing
α) server, cluster companies; administrators Do hosting/server companies besides coreinfo [5] with AMD based offers exist? That means, is there a chance of getting big contributions?
What about the Free Software Foundation (FSF), FSF Europe (FSFE), Electronic Frontier Foundation (EFF)? What about governments?
β) free software enthusiasts I hope with the FSF, FSFE and EFF some big organizations will be able to motivate a lot of people to donate. γ) private “normal” people This is my main problem. Alexandru Gagniuc uses(?)/used(?) the board as a workstation, but the normal user will never use that expensive server board at home. Therefore they will never hold it in their hands. In the end the given money is a donation. Therefore, we need good arguments that Jane User will participate.
I have the bad feeling, that most of them won’t donate, as they do not see the benefit as coreboot is hard to explain. This is the biggest risk for the campaign. To get those a lot of money needs to be invested in marketing and promotion. But will the additional donations cover those costs and add to the original goal to support upstreaming the board?
In the end, coreboot developers should get paid for their work. The Varnish Cache developer has a good write-up about this issue [7].
But software is written by people, real people with kids, cars, mortgages, leaky roofs, sick pets, infirm parents and all other kinds of perfectly normal worries of an adult human being.
The best way to improve the quality of Free and Open Source Software, is to make it possible for these people to spend time on it.
They need time to review submissions carefully, time to write and run test-cases, time to respond and fix to bug-reports, time to code and most of all, time to think about the code.
But it would not even be close to morally defensible to ask these people to forego time to play with their kids, so that they instead develop and maintain the software that drives other peoples companies.
The right way to go -- the moral way to go -- and by far the most productive way to go, is to pay the developers so they can make the software they love their living.
Questions ========= 1. Recently Roundcube, the most(?) popular Webmail client, which is much more known than coreboot, started a crowdfunding campaign for $80.000 [6]. In the last two weeks they raised around $25.000. But experience shows that only in the beginning and in the end of a campaign a lot of money is collected.
So, what do you think? Is the coreboot crowdfunding campaign even realistic?
2. What crowdfunding platform should be used? Does Indiegogo [3] work for everyone? Are there platforms which will get boycotted by free software enthusiasts?
Thank you for reading up to this point. I am looking forward to your comments.
Thanks,
Paul
[1] https://www.giantmonkey.de [2] https://campaign.joeyh.name [3] https://www.indiegogo.com [4] https://www.indiegogo.com/projects/jolla-tablet-world-s-first-crowdsourced-t... [5] http://coreinfo.us [6] https://roundcu.be/next [7] https://www.varnish-cache.org/docs/4.0/phk/dough.html
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 17/05/15 15:11, Paul Menzel wrote:
Dear coreboot folks,
Timothy, congratulations again on making a coreboot port for the ASUS KGPE-D16 and therefore completing your third coreboot port! That’s really amazing!
Am Mittwoch, den 29.04.2015, 22:46 +0100 schrieb The Gluglug:
You should crowd-fund the $35,000 figure, there are lots of people who will be interested in this. I personally will chip in, and I'd ask others to as well.
I am thinking about organizing the crowdfunding campaign to raise the money.
If somebody else wants to do it, please speak up!
As this is more or less a donation by the backers, the process should be as open and transparent as possible. That’s why I am sharing the following information publicly.
- Giant Monkey Software Engineering [1], the German company I work
for, would be the organizing entity. As a company with five employees and a GmbH it might have enough credibility so that people would pledge/give their money to Giant Monkey compared to a private person or company run by a single person. Giant Monkey also has some PR/campaign knowledge, but most importantly knows a lot of people in the marketing sector.
- After receiving the money, Giant Monkey would contract Raptor
Engineering.
- I’d like to have the domain campaign.coreboot.org redirect to
the campaign page at the crowdfunding platform or set up a simple Web site there.
- git-annex’ second funding was done by itself with PayPal. That
saves the 4 % fee most other crowdfunding platforms charge.
Using a crowdfunding platform might be easier though, as they have experience and also provide a big community of possible backers.
Currently I’m thinking about Indiegogo, which Jolla also used to fund the Jolla Tablet. I heard, Kickstarter is also great with a big network.
- I plan to raise 100.000 € (around $110.000) to upstream, that
includes *paid review* and running the campaign, the whole port. (I’ll continue to use Euros.) More money would be used for stretch goals.
?) 4.000 € Indiegogo fees ?) 35.000 € for Raptor Engineering for upstreaming for basic port ?) 15.000 € for Raptor Engineering for implementing support for S2R (S3) ?) 10.000 € for code review (inclusively hardware) (just an estimate) ?) 2.000 € for Gluglug to release images (I have not talked to Francis yet.) ?) 10.000 € campaign goodies (cf. 7.) ?) 4.000 € taxes (probably a lot more, depends if given money counts as donation) ?) 20.000 € Giant Monkey for running the campaign (Web site, press, marketing, videos, mile stone tasks (see below), …)
(Stretch goal) ?) 15.000 € for Raptor Engineering for upstreaming Family 15h support for the board
- I’d start with a minimal Web site and campaign platform page and
see how big the momentum alone through the coreboot community is. If we get 10.000 € in a week, I’d fully step in with a professional campaign. Otherwise I’d stop the campaign.
- As a thank you for backers, I think of a payload included in
the distributed coreboot based firmware image, reading a text file from CBFS with the names of the backers and displaying it. (Or a simple splash screen.) Of course just for those wanting it. Big backers (25.000 €) get a board with one CPU and RAM and coreboot preinstalled; medium backers (10.000 €) get some BLOB free laptop for example (Rockchip Chromebook or some Lenovo board). flash ROM chips are sent to backers donating 25 €.
- Milestone tasks: At certain mile stones (probably each 10.000
€), I’d promise some more tasks to improve coreboot or the port (see the 5.000 € steps in the top). Possible are also SSL certificates for coreboot infrastructure, promising to run a 32-bit userspace build host, redesigning the Web site, implementing CBMEM time stamp support in SeaBIOS and GRUB, supporting Google’s verified boot, ….
- Reasons for contributing
?) server, cluster companies; administrators Do hosting/server companies besides coreinfo [5] with AMD based offers exist? That means, is there a chance of getting big contributions?
What about the Free Software Foundation (FSF), FSF Europe (FSFE), Electronic Frontier Foundation (EFF)? What about governments?
?) free software enthusiasts I hope with the FSF, FSFE and EFF some big organizations will be able to motivate a lot of people to donate. ?) private “normal” people This is my main problem. Alexandru Gagniuc uses(?)/used(?) the board as a workstation, but the normal user will never use that expensive server board at home. Therefore they will never hold it in their hands. In the end the given money is a donation. Therefore, we need good arguments that Jane User will participate.
I have the bad feeling, that most of them won’t donate, as they do not see the benefit as coreboot is hard to explain. This is the biggest risk for the campaign. To get those a lot of money needs to be invested in marketing and promotion. But will the additional donations cover those costs and add to the original goal to support upstreaming the board?
In the end, coreboot developers should get paid for their work. The Varnish Cache developer has a good write-up about this issue [7].
But software is written by people, real people with kids, cars, mortgages, leaky roofs, sick pets, infirm parents and all other kinds of perfectly normal worries of an adult human being.
The best way to improve the quality of Free and Open Source Software, is to make it possible for these people to spend time on it.
They need time to review submissions carefully, time to write and run test-cases, time to respond and fix to bug-reports, time to code and most of all, time to think about the code.
But it would not even be close to morally defensible to ask these people to forego time to play with their kids, so that they instead develop and maintain the software that drives other peoples companies.
The right way to go -- the moral way to go -- and by far the most productive way to go, is to pay the developers so they can make the software they love their living.
Questions ========= 1. Recently Roundcube, the most(?) popular Webmail client, which is much more known than coreboot, started a crowdfunding campaign for $80.000 [6]. In the last two weeks they raised around $25.000. But experience shows that only in the beginning and in the end of a campaign a lot of money is collected.
So, what do you think? Is the coreboot crowdfunding campaign even realistic?
- What crowdfunding platform should be used? Does Indiegogo [3]
work for everyone? Are there platforms which will get boycotted by free software enthusiasts?
Thank you for reading up to this point. I am looking forward to your comments.
Thanks,
Paul
[1] https://www.giantmonkey.de [2] https://campaign.joeyh.name [3] https://www.indiegogo.com [4] https://www.indiegogo.com/projects/jolla-tablet-world-s-first-crowdsourced-t...
[6] https://roundcu.be/next [7] https://www.varnish-cache.org/docs/4.0/phk/dough.html
I think we should get the FSF to run a campaign for this. They can keep the overheads low and keep the funding amount low. They are also based in the same country as Raptor Engineering.
On 04/29/2015 04:30 PM, Timothy Pearson wrote:
All,
I have successfully ported coreboot to the relatively modern ASUS KGPE-D16 server board (dual AMD socket G34, 16 DDR3 DIMMs, https://www.asus.com/us/Commercial_Servers_Workstations/KGPED16/)! This port uses native Family 10h initialization (_not_ AGESA or CIMX).
The Libreboot folks will be interested to know that this board can run blob-free and still retain full functionality!
Port specifications: CPU: Dual AMD G34 Magny-Cours (Family 10h) RAM: 16 DDR3 DIMM slots with ECC support (tested with x4 4G DDR3-1333 unbuffered DIMMs)
Just a quick update on this. Work has continued internally, and as of this message the KGPE-D16 port has two additional options available: full S3 support and native (non-AGESA) support for Family 15h processors. When all of the options are combined the KGPE-D16 is able to use the most modern Opterons available with no apparent issues; as a result, there should be very little development work remaining on this system.
The new Family 15h initialisation is integrated into the native Family 10h codebase; the two processors are similar enough overall (though still quite different on a more local scale) that this was feasible. The memory controllers are quite different between the two processor families; much of the additional work was related to updating the processor-independent mct_ddr3 code to handle the modified MCT design and training procedures. During this process the mct_ddr3 codebase has also been cleaned up significantly, and if the entire proposed patchset is merged the cleanup work will continue under Raptor's normal (free) open source development model.
Family 15h port specifications: Opteron 6200 processors: tested, fully functional Opteron 6300[P] processors: theoretically supported, not yet tested Dual packages: theoretically supported, not yet tested
Unbuffered DDR3 DIMMs: tested, fully functional Registered DDR3 DIMMs: theoretically supported, not yet tested Load-Reduced DDR3 DIMMs: incomplete support [1] Reduced DIMM voltage: tested, fully functional
ACPI object generation: tested, fully functional PowerNow!: tested, fully functional S3 suspend/resume: tested, fully functional
As mentioned before, this port was done for internal purposes to give Raptor Engineering a competitive advantage in certain markets. However, Raptor Engineering is willing to upstream the code if funded. The offer is currently broken up into three parts, with each part being dependent on the previous part:
1.) Full KGPE-D16 support as described in previous message, Family 10h only w/o S3: $35,000 USD
2.) Family 10h S3 support: $15,000 USD
3.) Native Family 15h support with S3 and MCT modifications: $40,000 USD
The port is broken up into these sections specifically so that the community can decide what is important enough to fund. Ideally the entire patch set would be funded, but if the community does not have sufficient interest in the most current Opteron processors I don't want to block the potential upstreaming of the other two sections.
Items marked as "theoretically supported" above should work but have not been tested; if they do not function Raptor Engineering would fix any observed failures as part of the quoted costs.
Thank you for your consideration!
1. These are not generally relevant to the KGPE-D16 on a technical basis due to the design of the G34 socket, and even less so given the extremely high cost of LRDIMM modules. However, if LRDIMM support is desired Raptor Engineering would be willing to add support as part of the quoted cost after the rest of the code is upstreamed.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
All,
Raptor Engineering is pleased to announce the public release of the ASUS KGPE-D16 support code, including support for Family 15h processors!
The current feature status matrix and link to the patchsets on Gerrit is available here: https://raptorengineeringinc.com/coreboot/kgpe-d16-status.php
Raptor Engineering would also like to thank Minifree Ltd. for sponsoring this release. I know earlier it looked like the crowdfunding push had died away, but we were working hard on a deal to get this code to the general public as quickly as possible. As such I couldn't say too much on the topic until now.
To get the funding requirements down to a manageable level, Raptor Engineering has split off several items that would normally have been included with this upstreaming push. Specifically, there is no blanket warranty provided with the patchset as was originally intended -- instead, if an functionality issue is found, you are encouraged to contract directly with Raptor Engineering to create a fix. Also, some of the features not directly related to our production operations here are incomplete; for example, we are currently looking for sponsors to sponsore the work needed to resolve certain RAM initialisation issues and to port the OpenBMC firmware to these boards.
Thank you all for your interest in this port; I hope to hear from you again in the future!
- -- Timothy Pearson Raptor Engineering +1 (415) 727-8645 (direct line) +1 (512) 690-0200 (switchboard) http://www.raptorengineeringinc.com
Hi all,
Timothy Pearson wrote:
Raptor Engineering is pleased to announce the public release of the ASUS KGPE-D16 support code, including support for Family 15h processors!
..
Raptor Engineering would also like to thank Minifree Ltd. for sponsoring this release.
I would like to take this opportunity to also express my gratitude and appreciation to Minifree Ltd. and Francis Rowe for enabling this most excellent contribution to coreboot possible!
Thank you very much.
I have been quite critical of the Libreboot project, but I do recognize that this contribution to coreboot would never have been possible if it were not for Francis starting out on the Libreboot adventure.
Well done! Keep them coming. :) Francis, now I owe you two beverages.
Have a great weekend
//Peter
-
Emilian Bold -
Paul Menzel -
Peter Stuge -
The Gluglug -
Timothy Pearson -
Ward Vandewege