I know there hasnt been a lot of posts here, and that weve been posting over on G+ more (sorry Im going to try and do better about sharing things everywhere) but I want to give you guys a heads up. Weve got a rather serious bug in the firmware thats shipped and we know theres been a few people who have hit it and has the unfortunate side effect of causing the board to no longer boot. Weve been investigating the issue, and have been helping people who have hit the issue. We are fairly sure weve root caused the issue, and we have an early release of the firmware that should prevent that issue. We are getting this out there early because the issue is important enough that getting the fix out faster is important to us. Ive attached the firmware to the bug, but it can be directly downloaded at https://bugzilla.yoctoproject.org/attachment.cgi?id=2087 Instructions for flashing the firmware can be found here: https://bugzilla.yoctoproject.org/show_bug.cgi?id=6585#c26 the overall bug information can be found here: https://bugzilla.yoctoproject.org/show_bug.cgi?id=6585 The abbreviated version of the issue (as weve so far found) that theres enough data, during reboots, that gets written to the NVRAM to eventually invoke the NVRAM reclaim process. Unfortunately this can get triggered at any stage in the boot up sequence and there (until recently, I.E. this firmware), there was no indication anywhere (there is now an indication on the serial console when this occurs), and this was a fairly slow process (we fixed that too). *IF* you were to power cycle the board in the middle of the reclaim, it basically confused the next boot and it can no longer boot. Theres 4 things that have changed in this firmware to note: 1) *IF* you interrupt the NVRAM reclaim process, it can now recover on the next boot 2) The NVRAM reclaim process has been sped up by 9x (meaning you hopefully wont notice it and get concerned) 3) There is now a message on the serial console mentioning that this is happening. (Ive asked to try and have something on the screen, there are problems there mainly due to timing) 4) On a slightly more minor note, there is a fix for some HDMI and a lot of DVI monitors in now as well, so all displays should be working as expected now. Obviously I want to encourage everyone to upgrade their firmware, but keep in mind this is an early version, and hasnt necessarily gone through full validation, but we feel we need to get this to you sooner than later. Questions and issues, should get directed here, or taken to the mailing list, G+ or to the mailing list. - John Warthog9 Hawley
Posted on: Fri, 15 Aug 2014 01:34:12 +0000
Trending Topics
Recently Viewed Topics
© 2015