[U-Boot] [ANN] v2016.05-rc3
Hey all,
I know I did v2016.05-rc2 just last week, but that's because I was off schedule. We are two weeks out from scheduled release, so he's -rc3. Please test this everywhere you can. The diffstat here is large due to moving things from config.h to Kconfig but the binary delta is not. At this point forward, I expect things that come in to be bug fixes of a fairly important nature. Oh, and I'm going to re-generate the defconfigs and let people know about the surprises that spring up. Correcting those is also fine. Thanks all!
Dear Tom,
In message 20160426001803.GO29322@bill-the-cat you wrote:
I know I did v2016.05-rc2 just last week, but that's because I was off schedule. We are two weeks out from scheduled release, so he's -rc3.
rc2 and rc3 tarballs are now on FTP and ACD.
Best regards,
Wolfgang Denk
On 2016-04-25, Tom Rini wrote:
I know I did v2016.05-rc2 just last week, but that's because I was off schedule. We are two weeks out from scheduled release, so he's -rc3. Please test this everywhere you can.
I'm having issues with v2016.05-rc3 booting beaglebone black from eMMC and microSD, reverting ef5ebe951bec72631cdbc7cef9079e6c684e5d0b "ti_armv7_common.h: Fix U-Boot location on eMMC" seems to fix the issue for me.
It appears this patch changed where the raw offsets for loading u-boot.img are, though the "old" offsets are already in widespread use, so would be unfortunate to have to change them everywhere else.
live well, vagrant
On Tue, Apr 26, 2016 at 12:05:58PM -0700, Vagrant Cascadian wrote:
On 2016-04-25, Tom Rini wrote:
I know I did v2016.05-rc2 just last week, but that's because I was off schedule. We are two weeks out from scheduled release, so he's -rc3. Please test this everywhere you can.
I'm having issues with v2016.05-rc3 booting beaglebone black from eMMC and microSD, reverting ef5ebe951bec72631cdbc7cef9079e6c684e5d0b "ti_armv7_common.h: Fix U-Boot location on eMMC" seems to fix the issue for me.
It appears this patch changed where the raw offsets for loading u-boot.img are, though the "old" offsets are already in widespread use, so would be unfortunate to have to change them everywhere else.
Indeed this is a problem. I think for this release we might have to revert the change and then moving forward extend SPL to allow this location to come via the environment (which we do for example with bootcount) so that other use-cases can change things as needed. Sorry for the confusion, I had thought the patch was doing something slightly different than it is.
participants (3)
-
Tom Rini -
Vagrant Cascadian -
Wolfgang Denk