On 07/02/2013 05:22 PM, Stephen Boyd wrote: > On 07/02, Stephen Warren wrote: >> From: Stephen Warren <swarren@xxxxxxxxxx> >> >> Commit 281ecb7 "arm: delete __cpuinit/__CPUINIT usage from all ARM >> users" removed a __CPUINIT from head-common.S. However, the code >> immediately before the removed tag was marked __INIT, and was missing >> a match __FINIT. This caused code after the removed __CPUINIT to end >> up in the init section rather than the main text section as desired. >> This caused issues such as secondary CPU boot failures or crashes. >> Add the missing __FINIT to solve this. >> >> Signed-off-by: Stephen Warren <swarren@xxxxxxxxxx> >> --- >> arch/arm/kernel/head-common.S | 2 ++ >> 1 file changed, 2 insertions(+) >> >> diff --git a/arch/arm/kernel/head-common.S b/arch/arm/kernel/head-common.S >> index 529ce99..a9dc804 100644 >> --- a/arch/arm/kernel/head-common.S >> +++ b/arch/arm/kernel/head-common.S >> @@ -122,6 +122,8 @@ __mmap_switched_data: >> .long init_thread_union + THREAD_START_SP @ sp >> .size __mmap_switched_data, . - __mmap_switched_data >> >> + __FINIT >> + >> /* >> * This provides a C-API version of __lookup_processor_type >> */ > > Shouldn't this come after lookup_processor_type()? Otherwise you > just removed that function from the init section? I think this is correct. Prior to "arm: delete __cpuinit/__CPUINIT usage from all ARM users", lookup_processor_type() was in the cpuinit section. After that commit, it moved to the init section and hence got dumped soon after boot. The change above moves lookup_processor_type() back into the regular text section, which I believe was the intent of the __cpuinit removal patch. -- To unsubscribe from this list: send the line "unsubscribe linux-tegra" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
