[arm/tegra] Don't try to enable CoreSight on non-running CPUs

On exit from LP2, __enable_coresite_access was called to reset the
CoreSight interface and re-enable access on all CPUs. However, only
CPU0 would actually be running at the time (the other CPU would still
be held in reset). The attempt to unlock CoreSight on the non-running
CPU would cause a stall on the APB bus while the CoreSight access
timed out. The APB stall would cause SLINK DMA receiver overruns.

Rather than attempting to unlock CoreSight access up front for every
CPU, each CPU is now responsible for unlocking it's own access when
it starts up.

Bug 703311

Change-Id: Ie4611423ed72eb1cd0dbc8b7851f7a047bcffa14
Reviewed-on: http://git-master/r/4683
Reviewed-by: Gary King <gking@nvidia.com>
Reviewed-by: Scott Williams <scwilliams@nvidia.com>
Tested-by: Scott Williams <scwilliams@nvidia.com>
Reviewed-by: Michael Hsu <mhsu@nvidia.com>
Tested-by: Michael Hsu <mhsu@nvidia.com>

0 files changed, 0 insertions, 0 deletions