Hey Andrey,
While looking on publishing the kernel packages for Ubuntu based on llt, we noticed that it's unable to build in case CPU_IDLE is enabled.
This is basically because it seems one patch is applied twice, and both adding the same content at the cpuidle.h file: $ git blame drivers/cpuidle/cpuidle.h 78bb15d5 (Colin Cross 2012-05-03 10:23:24 +0800 35) #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED ...
b667844d (Colin Cross 2012-05-03 10:37:35 +0800 65) #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED
Looking further, it seems that the culprit is b667844d, as you can see at http://git.linaro.org/gitweb?p=kernel/linux-linaro-tracking.git%3Ba=commitdi...
Guess that we should not only revert it, but maybe fix on the tree that actually pulled the patch in.
Thanks,
Hey Ricardo,
On 06/26/2012 05:30 AM, Ricardo Salveti wrote:
Hey Andrey,
While looking on publishing the kernel packages for Ubuntu based on llt, we noticed that it's unable to build in case CPU_IDLE is enabled.
This is basically because it seems one patch is applied twice, and both adding the same content at the cpuidle.h file: $ git blame drivers/cpuidle/cpuidle.h 78bb15d5 (Colin Cross 2012-05-03 10:23:24 +0800 35) #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED ...
b667844d (Colin Cross 2012-05-03 10:37:35 +0800 65) #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED
Looking further, it seems that the culprit is b667844d, as you can see at http://git.linaro.org/gitweb?p=kernel/linux-linaro-tracking.git%3Ba=commitdi...
Guess that we should not only revert it
This is done. llt-20120626.0 should be ok (build tested with CPU_IDLE enabled). BTW, compared to the previous llt-20120613.0, llt-20120626.0 has vexpress support, and (as a side effect of adding vexpress) gator upgraded from 5.10.0 to 5.10.1.
but maybe fix on the tree that actually pulled the patch in.
This is in progress.
Thanks, Andrey
This is done. llt-20120626.0 should be ok (build tested with CPU_IDLE enabled). BTW, compared to the previous llt-20120613.0, llt-20120626.0 has vexpress support, and (as a side effect of adding vexpress) gator upgraded from 5.10.0 to 5.10.1.
I will add an llt-vexpress-3.4 kernel variant and respin all three llt kernels.
On 06/26/2012 06:51 PM, John Rigby wrote:
This is done. llt-20120626.0 should be ok (build tested with CPU_IDLE enabled). BTW, compared to the previous llt-20120613.0, llt-20120626.0 has vexpress support, and (as a side effect of adding vexpress) gator upgraded from 5.10.0 to 5.10.1.
I will add an llt-vexpress-3.4 kernel variant and respin all three llt kernels.
I've created llt-20120626.1 which is just one blank line different from llt-20120626.0 as regards to the sources. The real difference is that this build issue is now fixed in tilt-3.4 itself, and my tilt-3.4-fixes topic is no longer needed and is excluded from llt-20120626.1
Thanks, Andrey