On Fri, Jul 1, 2016 at 6:39 PM, Hongbo Zhang macro.wave.z@gmail.com wrote:
On Sun, Jun 19, 2016 at 12:38 PM, Chen-Yu Tsai wens@csie.org wrote:
The original PSCI implementation assumed CONFIG_ARMV7_PSCI_NR_CPUS=4. Add this as a fallback value in case platforms have not defined it.
Signed-off-by: Chen-Yu Tsai wens@csie.org
arch/arm/include/asm/config.h | 5 +++++ 1 file changed, 5 insertions(+)
diff --git a/arch/arm/include/asm/config.h b/arch/arm/include/asm/config.h index 435fc4521c2e..f70302dfc4f1 100644 --- a/arch/arm/include/asm/config.h +++ b/arch/arm/include/asm/config.h @@ -23,4 +23,9 @@ #include <asm/arch/config.h> #endif
+/* Original code assumed 4 CPUs */ +#ifndef CONFIG_ARMV7_PSCI_NR_CPUS +#define CONFIG_ARMV7_PSCI_NR_CPUS 4
This makes platforms which have there own macro definition embarrassed somehow. we should add #define CONFIG_ARMV7_PSCI_NR_CPUS CONFIG_MAX_CPUS if this patch merged. some of our platform even has 16 cores.
I'm not sure how CONFIG_MAX_CPUS is used though.
Thinking about this, my first approach is probably wrong. I should add missing PSCI_NR_CPUS for the existing platforms. New platforms should define it themselves when enabling PSCI. Failure to do so will result in a compile error, rather than having a fallback value that subtly breaks later on.
Regards ChenYu
+#endif
#endif
2.8.1