[U-Boot] [STATUS] U-Boot v2010.09 released - Merge Window OPEN
Hello all,
U-Boot v2010.09 has been released and is available from the git repository and the FTP server.
The "next" branch has been pulled into mainline ("master" branch) and has been removed afterward.
******************** ******************** ** ** ** BIG FAT NOTE ** ** ** ******************** ********************
The commits we pulled from the next branch into mainline contain 3 major groups of changes:
1) New implementation for internal handling of environment variables.
In addition to the traditional commands "setenv", "printenv" and "saveenv" there is now a new "env" command with a number opf subcommands that offer many new features. Please see the commit message of commit ea882baf for details:
http://git.denx.de/?p=u-boot.git;a=commit;h=ea882baf
2) ARM cache support.
This greatly accelerates U-Boot on many ARM systems, especially functions like uncompressing images or loading images form external storage are now much faster, sometimes by a factor > 10. For details please see these commits:
7e4a9e6 ARM (ARM11): add data cache support, test on Qong board 95c6f6d ARM V7 (OMAP): add data cache support, test on Beagle board c3330e9 ARM (ARM926ejs): add data cache support, tested on magnesium and tx25 board
http://git.denx.de/?p=u-boot.git;a=commit;h=7e4a9e6 http://git.denx.de/?p=u-boot.git;a=commit;h=95c6f6d http://git.denx.de/?p=u-boot.git;a=commit;h=c3330e9
3) ARM relocation support.
This allows to efficiently use the same U-Bot binary image on systems with different RAM sizes, and brings the implementation much more in line with the code used for example on Power Architecture systems (eventually allowing to merge into common code). This seems especailly interesting now that ARM is getting Device Tree support as well.
Please note that *all* ARM boards need to be adapted to this noew code, which requires testing on the actual hardware, so this is a task for the respective board maintainers or other users.
***************************************************************** * Currently, most ARM boards are *broken* and *do not compile*. * *****************************************************************
Please see the commit message of commit f1d2b31 for details:
http://git.denx.de/?p=u-boot.git;a=commit;h=f1d2b31
Support for CONFIG_SYS_ARM_WITHOUT_RELOC will be removed after release v2010.03; all boards that have not been converted by then, i. e. that are still broken then, are considered unmaintained and without interest for the community and will be removed as well.
--- end of big fat note ---
The Merge Window for the next release (v2010.12) is open until Sat Oct 16, 2010, 23:59:59 CET = 17 days remaining.
The next release v2010.12 is scheduled for December 13, 2010.
See http://www.denx.de/wiki/U-Boot/ReleaseCycle for details, including guestimates for the next releases after that.
A little statistics [1] - changes since release v2020.06:
Processed 842 csets from 114 developers 32 employers found A total of 77775 lines added, 140638 removed (delta -62863)
Developers with the most changesets Mike Frysinger 77 (9.1%) Kumar Gala 69 (8.2%) Wolfgang Denk 55 (6.5%) Steve Sakoman 34 (4.0%) Minkyu Kang 33 (3.9%) Stefan Roese 28 (3.3%) Stefano Babic 19 (2.3%) Albert Aribaud 19 (2.3%) Graeme Russ 18 (2.1%) Marek Vasut 17 (2.0%) ...
Developers with the most changed lines Wolfgang Denk 111033 (56.0%) Minkyu Kang 6363 (3.2%) Kumar Gala 6342 (3.2%) Graeme Russ 5707 (2.9%) Vipin KUMAR 5698 (2.9%) Mike Frysinger 5461 (2.8%) Steve Sakoman 4111 (2.1%) Stefan Roese 3789 (1.9%) Matthias Kaehlcke 3008 (1.5%) kevin.morfitt@fearnside-systems.co.uk 2950 (1.5%) ...
Developers with the most lines removed Wolfgang Denk 105067 (74.7%) Graeme Russ 4244 (3.0%) Thomas Chou 979 (0.7%) kevin.morfitt@fearnside-systems.co.uk 867 (0.6%) Michal Simek 797 (0.6%) Michael Zaidman 612 (0.4%) Peter Tyser 383 (0.3%) Becky Bruce 206 (0.1%) Ladislav Michl 172 (0.1%) Nishanth Menon 104 (0.1%) ...
Developers with the most signoffs (total 294) Sandeep Paulraj 54 (18.4%) Kumar Gala 54 (18.4%) Ben Warren 30 (10.2%) Stefan Roese 19 (6.5%) Minkyu Kang 19 (6.5%) Wolfgang Denk 13 (4.4%) Mike Frysinger 13 (4.4%) Kyungmin Park 11 (3.7%) Scott McNutt 8 (2.7%) Kim Phillips 6 (2.0%) ...
Developers with the most reviews (total 0)
Developers with the most test credits (total 20) Steve Sakoman 3 (15.0%) Thomas Chou 3 (15.0%) Stefan Roese 1 (5.0%) Minkyu Kang 1 (5.0%) Wolfgang Denk 1 (5.0%) Detlev Zundel 1 (5.0%) Peter Tyser 1 (5.0%) Ian Abbott 1 (5.0%) Philip Balister 1 (5.0%) Andreas BieÃmann 1 (5.0%) ...
Developers who gave the most tested-by credits (total 20) Mike Frysinger 5 (25.0%) Grazvydas Ignotas 2 (10.0%) Steve Sakoman 1 (5.0%) Thomas Chou 1 (5.0%) Stefan Roese 1 (5.0%) Wolfgang Denk 1 (5.0%) Reinhard Meyer 1 (5.0%) Aneesh V 1 (5.0%) Timur Tabi 1 (5.0%) kevin.morfitt@fearnside-systems.co.uk 1 (5.0%) ...
Developers with the most report credits (total 4) Wolfgang Denk 2 (50.0%) York Sun 1 (25.0%) Vivi Li 1 (25.0%)
Developers who gave the most report credits (total 4) Mike Frysinger 1 (25.0%) Kumar Gala 1 (25.0%) Kim Phillips 1 (25.0%) Anton Vorontsov 1 (25.0%)
Top changeset contributors by employer (Unknown) 168 (20.0%) Freescale 134 (15.9%) DENX Software Engineering 119 (14.1%) Analog Devices 85 (10.1%) Samsung 43 (5.1%) Sakoman Inc. 34 (4.0%) Texas Instruments 25 (3.0%) Wind River 20 (2.4%) Graeme Russ 18 (2.1%) Marvell 16 (1.9%) ...
Top lines changed by employer DENX Software Engineering 121675 (61.4%) (Unknown) 21198 (10.7%) Freescale 10652 (5.4%) Analog Devices 7846 (4.0%) Samsung 6709 (3.4%) Graeme Russ 5996 (3.0%) ST Microelectronics 5701 (2.9%) Sakoman Inc. 4121 (2.1%) Wind River 3766 (1.9%) Texas Instruments 2239 (1.1%) ...
Employers with the most signoffs (total 294) Freescale 90 (30.6%) Texas Instruments 61 (20.7%) (Unknown) 44 (15.0%) DENX Software Engineering 35 (11.9%) Samsung 34 (11.6%) Analog Devices 15 (5.1%) Psyent 8 (2.7%) Sakoman Inc. 4 (1.4%) Wind River 1 (0.3%) Marvell 1 (0.3%) Xilinx 1 (0.3%)
[1] See http://www.denx.de/wiki/U-Boot/UbootStat_2010_09 for full statistics, and http://www.denx.de/wiki/UBoot/ReleaseCycle for links to statistics for earlier releases.
Best regards,
Wolfgang Denk
Hi Wolfgang,
U-Boot v2010.09 has been released and is available from the git repository and the FTP server.
Cool, thanks!
[...]
- Currently, most ARM boards are *broken* and *do not compile*. *
Please see the commit message of commit f1d2b31 for details:
http://git.denx.de/?p=u-boot.git;a=commit;h=f1d2b31
Support for CONFIG_SYS_ARM_WITHOUT_RELOC will be removed after release v2010.03; all boards that have not been converted by then,
As many people won't have a time machine ready, I belive this should read "v2011.03", correct?
Cheers Detlev
Dear Detlev Zundel,
In message m24od8ki5t.fsf@ohwell.denx.de you wrote:
Support for CONFIG_SYS_ARM_WITHOUT_RELOC will be removed after release v2010.03; all boards that have not been converted by then,
As many people won't have a time machine ready, I belive this should read "v2011.03", correct?
Correct. Sorry, seems my fingers are wired for 2010...
Best regards,
Wolfgang Denk
On Wed, 29 Sep 2010 11:16:30 +0200, Wolfgang Denk wd@denx.de wrote:
This allows to efficiently use the same U-Bot
The U-Bot strikes again :-)
On Wed, Sep 29, 2010 at 02:51:49PM +0200, Bas Mevissen wrote:
On Wed, 29 Sep 2010 11:16:30 +0200, Wolfgang Denk wd@denx.de wrote:
This allows to efficiently use the same U-Bot
The U-Bot strikes again :-)
Striving for world domination.
SCNR, Wolfgang
On Wednesday, September 29, 2010 05:16:30 Wolfgang Denk wrote:
A little statistics [1] - changes since release v2020.06:
granted it's been years since i was in school taking math courses, but these totals dont seem to add up ...
Processed 842 csets from 114 developers
842 is the total # of commits ...
Developers with the most signoffs (total 294)
... but we only have 294 signoffs !? which means we have at least 548 commits merged that lack any signoff at all !?
Developers with the most changesets Mike Frysinger 77 (9.1%) Kumar Gala 69 (8.2%)
guessing this means Author ... i have 77 and Kumar has 69 ...
Developers with the most signoffs (total 294) Kumar Gala 54 (18.4%) Mike Frysinger 13 (4.4%)
yet here i only have 13 and Kumar has only 54 !? i'm fairly certain i've signed off on every single commit i either authored or pushed through my tree. -mike
Dear Mike Frysinger,
In message 201009290941.42366.vapier@gentoo.org you wrote:
On Wednesday, September 29, 2010 05:16:30 Wolfgang Denk wrote:
A little statistics [1] - changes since release v2020.06:
I guess you misses the time warp that I caused by used a version from 10 years in the future as refernce ;-)
[Man, that was a really buggy posting ;-) ]
granted it's been years since i was in school taking math courses, but these totals dont seem to add up ...
Please feel free to clone git://git.lwn.net/gitdm.git and check yourself.
[But make sure to replace the patch generated for b1f95b4438a with dummy data, or you will get even funnier results.]
Best regards,
Wolfgang Denk
On Sep 29, 2010, at 8:41 AM, Mike Frysinger wrote:
On Wednesday, September 29, 2010 05:16:30 Wolfgang Denk wrote:
A little statistics [1] - changes since release v2020.06:
granted it's been years since i was in school taking math courses, but these totals dont seem to add up ...
Processed 842 csets from 114 developers
842 is the total # of commits ...
Developers with the most signoffs (total 294)
... but we only have 294 signoffs !? which means we have at least 548 commits merged that lack any signoff at all !?
Developers with the most changesets Mike Frysinger 77 (9.1%) Kumar Gala 69 (8.2%)
guessing this means Author ... i have 77 and Kumar has 69 ...
Developers with the most signoffs (total 294) Kumar Gala 54 (18.4%) Mike Frysinger 13 (4.4%)
yet here i only have 13 and Kumar has only 54 !? i'm fairly certain i've signed off on every single commit i either authored or pushed through my tree. -mike
I'm guessing on this last bit the stat's dont count signoffs on which you are the author, but that's just a theory :)
- k
Kumar Gala wrote:
I'm guessing on this last bit the stat's dont count signoffs on which you are the author, but that's just a theory :)
More as a theory: in gitdm README:
-s Ignore Signed-off-by lines which match the author of each patch.
If Wolfgang has used this option, your theory is right !
Stefano
Dear Stefano Babic,
In message 4CA490D3.5030403@denx.de you wrote:
Kumar Gala wrote:
I'm guessing on this last bit the stat's dont count signoffs on which you are the author, but that's just a theory :)
More as a theory: in gitdm README:
-s Ignore Signed-off-by lines which match the author of each patch.
If Wolfgang has used this option, your theory is right !
The command was: ... | gitdm -u -s -a -o /tmp/results -h /tmp/results.html
Best regards,
Wolfgang Denk
participants (7)
-
Bas Mevissen -
Detlev Zundel -
Kumar Gala -
Mike Frysinger -
Stefano Babic -
Wolfgang Denk -
wolfgang＠leila.ping.de