-----Original Message----- From: Jean-Christophe PLAGNIOL-VILLARD [mailto:plagnioj@jcrosoft.com] Sent: Saturday, September 05, 2009 4:48 AM To: Wolfgang Denk Cc: u-boot@lists.denx.de; Tom Rix; Magnus Lilja; Prafulla Wadaskar; Dirk Behme; Minkyu Kang; Kyungmin Park; Harald Welte Subject: Re: RFC: split ARM repo and distribute workload

On 23:57 Fri 04 Sep , Wolfgang Denk wrote:

...

Hello everybody,

ARM has always been one of the architectures that generated a big number of different processors and SoCs, but recently the

activitiy in

...

this area is literally exploding. This is partially due to the fact that ARM is currently used in many designs, so many new ARM based processors and SoCs and even more new ARM boards show up,

but another

...

and at least as important change is that some silicon and board vendors have started to actively pushing their products into the U-Boot (and Linux) mainline source trees (and *welcome* they all are!).

It has become evident that this growing complexity has

become way too

...

massive to be shouldered by a single custodian, even a very

active one

...

like Jean-Christophe.

I think we have no other choice but to add more manpower to

this task,

...

i. e. split the ARM respository and distribute the workload across a few more custodians.

Unline with the Power architecture, where the split can be easily defined by processor lines, with ARM it seems more logical to me to differentiate by silicon vendors.

After much thinking I therefor suggest to implement the following change for the ARM architecture:

I think this will be better

master ARM repository : Jean-Christophe Plagniol-Villard

Freescale (i.MX) Magnus Lilja? Or are there any volunteers at Freescale?

Marvell Prafulla Wadaskar

This makes more sense for u-boot-mrvl.git (proposed) to go entire Marvell specific patches through it Regards.. Prafulla . .

Samsung (s3c, s5pc): Are there any volunteers at Samsung?

Texas Instruments (OMAP): Tom Rix

Best Regards, J.

Wolfgang, Dirk,

I can take responsibility for TI DaVinci SOCs.

Thanks, Sandeep ________________________________________ From: u-boot-bounces@lists.denx.de [u-boot-bounces@lists.denx.de] On Behalf Of Dirk Behme [dirk.behme@googlemail.com] Sent: Saturday, September 05, 2009 3:39 AM To: Wolfgang Denk Cc: Tom Rix; Kyungmin Park; u-boot@lists.denx.de; Magnus Lilja; Minkyu Kang Subject: Re: [U-Boot] RFC: split ARM repo and distribute workload

Dear Wolfgang,

Wolfgang Denk wrote:

Texas Instruments (OMAP): Dirk Behme? Or are there any volunteers at TI?

I really like to see someone from TI here, instead.

Looking at some past discussions, I think somebody from TI would be the best. E.g.:

- While discussing the board/vendor directory change, we were unsure who might be the vendor of which board. Somebody from TI would know best.

- While discussing the OMAP3 cache move, we were unsure what's about the ROM code calling and where/when this is necessary. Someone from TI would know best, too.

- etc, etc.

Best regards

Dirk _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot

Tom,

This is also what we have been discussing internally within TI to have an omap/davinci split.

BTW there are SOC's which have an OMAP part number but are actually very similar to DaVinci's.

Since i am quite familiar with the RBLs of all these devices and i have all the EVMs for the various SOC's with me i could help in maintaining the repository.

Lets hope this split is OK with the others as well.

Thanks, Sandeep ________________________________________ From: Tom [Tom.Rix@windriver.com] Sent: Sunday, September 06, 2009 9:54 AM To: Paulraj, Sandeep Cc: Dirk Behme; Wolfgang Denk; Kyungmin Park; u-boot@lists.denx.de; Magnus Lilja; Minkyu Kang Subject: Re: [U-Boot] RFC: split ARM repo and distribute workload

Paulraj, Sandeep wrote:

Wolfgang, Dirk,

I can take responsibility for TI DaVinci SOCs.

This means an omap/davinci split ? u-boot-arm-omap u-boot-arm-davinci

This is ok with me Tom

Thanks, Sandeep ________________________________________ From: u-boot-bounces@lists.denx.de [u-boot-bounces@lists.denx.de] On Behalf Of Dirk Behme [dirk.behme@googlemail.com] Sent: Saturday, September 05, 2009 3:39 AM To: Wolfgang Denk Cc: Tom Rix; Kyungmin Park; u-boot@lists.denx.de; Magnus Lilja; Minkyu Kang Subject: Re: [U-Boot] RFC: split ARM repo and distribute workload

Dear Wolfgang,

Wolfgang Denk wrote:

...

Texas Instruments (OMAP): Dirk Behme? Or are there any volunteers at TI?

I really like to see someone from TI here, instead.

Looking at some past discussions, I think somebody from TI would be the best. E.g.:

• While discussing the board/vendor directory change, we were unsure

who might be the vendor of which board. Somebody from TI would know best.

• While discussing the OMAP3 cache move, we were unsure what's about

the ROM code calling and where/when this is necessary. Someone from TI would know best, too.

• etc, etc.

Best regards

Dirk _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot

Dear Sandeep,

In message 0554BEF07D437848AF01B9C9B5F0BC5D91E07BDB@dlee01.ent.ti.com you wrote:

This is also what we have been discussing internally within TI to have an omap/davinci split.

I'm not happy about the idea. If there really is need (or reason) to split these, I would prefer to do this in a seciond step, after the general ARM split has been completed and a stable work flow been established.

BTW there are SOC's which have an OMAP part number but are actually very similar to DaVinci's.

Since i am quite familiar with the RBLs of all these devices and i have all the EVMs for the various SOC's with me i could help in maintaining the repository.

Lets hope this split is OK with the others as well.

Maybe we can at least postpone it?

Would you also take responsibility as custodian for all TI, i. e. OMAP + DaVinci, at least for the time being?

Best regards,

Wolfgang Denk

Dear Kyungmin Park,

In message 9c9fda240909061806j5ecc26c3ie53215e8aa0175f0@mail.gmail.com you wrote:

...

Samsung (s3c, s5pc): Are there any volunteers at Samsung?

I recommended the Minky Kang. He's working for several years for u-boot from pxa, omap3, s3c6410 and s5pc1xx series.

Thanks.

But with our internal security policy it's hard to use ssh on u-boot git. so we want to use u-boot-arm as base.

Well, I think we can leave out the technical details here; we will find solutions to these.

Best regards,

Wolfgang Denk

Dear Wolfgang,

2009/9/7 Wolfgang Denk wd@denx.de

Hello everybody,

I wrote on

| Hello everybody, | | ARM has always been one of the architectures that generated a big | number of different processors and SoCs, but recently the activitiy in | this area is literally exploding. This is partially due to the fact | that ARM is currently used in many designs, so many new ARM based | processors and SoCs and even more new ARM boards show up, but another | and at least as important change is that some silicon and board | vendors have started to actively pushing their products into the | U-Boot (and Linux) mainline source trees (and *welcome* they all | are!). | | It has become evident that this growing complexity has become way too | massive to be shouldered by a single custodian, even a very active one | like Jean-Christophe. | | I think we have no other choice but to add more manpower to this task, | i. e. split the ARM respository and distribute the workload across a | few more custodians. | | Unline with the Power architecture, where the split can be easily | defined by processor lines, with ARM it seems more logical to me to | differentiate by silicon vendors.

Thanks for all the discussion so far. I really appreciate your help.

As we are in the middle of a merge window, I don't want to waste much time. Therefore changes become effective immediately.

This is the result we have so far:

master ARM repository: Tom Rix

    Tom has accepted his new responsibility. Thanks a lot for
   that. Tom succeeds Jean-Christophe Plagniol-Villard who did
   this work so far.

   Big thanks goes to Jean-Christophe for all his work so far. I
   hope this change helps to reduce your work load enough so you
   can focus n the many tasks you are running in parallel. Many
   of us are looking forward to seeing your results.

Atmel (AT91): Jean-Christophe Plagniol-Villard

    no changes here.

Freescale (i.MX)

   So far we didn't find a volunteer yet.

   Until we find one, we will not split out a separate repository
   either.

Marvell (PXA + IXP): Jean-Christophe Plagniol-Villard

    no changes here.

Marvell (all other): Prafulla Wadaskar

    Prafulla has accepted his new responsibility. Thanks a lot
   for that.

Samsung (s3c, s5pc): Minkyu Kang

   We don't have Minkyu Kang's word yet, but Kyungmin Park
   recommended him, and I'm optimistic to receiving his OK.

sorry to late response. I need all of your help :) Thank you for give me the opportunity.

Texas Instruments (OMAP,DaVinci): Sandeep Paulraj

   Sandeep has accepted responsibility for TI's DaVinci SOCs; at
   the moment it is not clear yet how we will handle OMAP -
   Sandeep suggested to split OMAP and DaVinci into separate
   repositories, but I would like to avoid this, at least until
   all the other changes have sunk in and stabilized. I hope that
   others (Dirk, Tom) can help out a bit here and support Sandeep
   where needed.

I will thus create the following new git repositories:

   u-boot-marvell
   u-boot-samsung
   u-boot-ti

Tom, Prafulla, Minkyu and Sandeep: can you please send me your SSH public keys so I can enable your access to these repositories?

Minkyu - Kyungmin mentioned that you have problems to push files to an external git repository through SSH. Is there any other method we can set up to make access for you easier? Anything that can be done onour side (like providing a special configuration, say with non-standsard port numbers or so) is easy for us. Alternatively, if you could set up a local repository on your side and enable me to access it through ssh or rsync, I can easily use this and mirror it to the publicly accessable server. In short - if there is anything we can do to make work easier for you, please don't hesitate to tell me.

Best regards,

Wolfgang Denk

-- DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@denx.de WARNING: This Product Attracts Every Other Piece of Matter in the Universe, Including the Products of Other Manufacturers, with a Force Proportional to the Product of the Masses and Inversely Proportional to the Distance Between Them. _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot

Thanks. Minkyu Kang.

Dear Magnus,

In message 59b21cf20909070206t189fe96bpac3cce2770b6b752@mail.gmail.com you wrote:

...

Freescale (i.MX) Magnus Lilja? Or are there any volunteers at Freescale?

I don't have the time to handle this in a good way. It would be good if someone at Freescale could step in.

I see. Thank you very much anyway.

So until someone volunteers here, we leave this part untouched yet.

Best regards,

Wolfgang Denk

Dear Wolfgang,

On Tue, Sep 08, 2009 at 02:45:18PM +0200, Wolfgang Denk wrote:

in the context of splitting the ARM architecure into vendor-related sub-repositories an new "u-boot-samsung.git" repository has been created; it seems to make sense to merge the older u-boot-s3c24xx repository into this one.

As you're the custodian for the u-boot-s3c24xx repository I would like to hear your opinion about this.

I'm perfectly fine with this, as I have not really used my custodian status much anyway. I always assumed I would find more time to contribute patches from the Openmoko times, but never got around doing that, too many other projects.

Dear Jean-Christophe,

in message 20090921113701.GP29905@game.jcrosoft.org you wrote:

...

Atmel (AT91): Jean-Christophe Plagniol-Villard

Marvell (PXA + IXP): Jean-Christophe Plagniol-Villard

I decline it

Thank you for the clear statement. Thanks again for the work you did in the past.

May I inquire what your plans are for other development tasks which you said in the past you were working on (generic cache support for ARM, Kconfig, etc.)? Will you continue this work and submit any patches for these? If so, when might this happen?

Tom, I suggest we proceed as follows:

- IXP and PXA are really low traffic these days (EOL products); I think we can give up these repositories and just ran this through the generic ARM repo.

- AT91 needs a new custodian. Let's see if someone volunteers. Until then, please try to pick up related patches directly.

Best regards,

Wolfgang Denk

Dear Prafulla,

in message 73173D32E9439E4ABB5151606C3E19E202ED512BC1@SC-VEXCH1.marvell.com you wrote:

...

• IXP and PXA are really low traffic these days (EOL products); I think we can give up these repositories and just ran this through the generic ARM repo.

I would love to own this one too. But I will check if somebody more capable in Marvell can help to own this activity.

Given the low traffic for these processors I think we can stop having a separate repository. If you want to care about this, we can run this through the existing u-boot-marvell repository, which is only logical.

Thanks for your help.

Best regards,

Wolfgang Denk

Wolfgang Denk wrote:

Dear Jean-Christophe,

in message 20090921113701.GP29905@game.jcrosoft.org you wrote:

...

...

Atmel (AT91): Jean-Christophe Plagniol-Villard

Marvell (PXA + IXP): Jean-Christophe Plagniol-Villard

I decline it

Thank you for the clear statement. Thanks again for the work you did in the past.

May I inquire what your plans are for other development tasks which you said in the past you were working on (generic cache support for ARM, Kconfig, etc.)? Will you continue this work and submit any patches for these? If so, when might this happen?

Tom, I suggest we proceed as follows:

• IXP and PXA are really low traffic these days (EOL products); I think we can give up these repositories and just ran this through the generic ARM repo.

OK.

• AT91 needs a new custodian. Let's see if someone volunteers. Until then, please try to pick up related patches directly.

OK

Best regards,

Wolfgang Denk