linux.git/arch/arm/mach-hisi, branch master

ARM: hisi: consolidate the hisilicon machine entries

2016-07-07T08:37:59+00:00

The original patch was to add compatible string for Hi3519 soc and do some cleanup. Since the generic machine entry could meet most of the cases, so I did a further cleanup to reuse it and just keep one machine entry that needs map_io here. Signed-off-by: Jiancheng Xue Signed-off-by: Wei Xu

ARM: hisi: make unexported symbols static

2016-06-28T15:23:47+00:00

The two functions hix5hd2_set_scu_boot_addr() and hip01_set_boot_addr() are not declared or exported outside arch/arm/mach-hisi/platsmp.c file. Avoid the following warnings by making them static: arch/arm/mach-hisi/platsmp.c:142:6: warning: symbol 'hip01_set_boot_addr' was not declared. Should it be static? arch/arm/mach-hisi/platsmp.c:106:6: warning: symbol 'hix5hd2_set_scu_boot_addr' was not declared. Should it be static? Signed-off-by: Ben Dooks Signed-off-by: Wei Xu

ARM: use "depends on" for SoC configs instead of "if" after prompt

2015-12-01T21:44:51+00:00

Many ARM sub-architectures use prompts followed by "if" conditional, but it is wrong. Please notice the difference between config ARCH_FOO bool "Foo SoCs" if ARCH_MULTI_V7 and config ARCH_FOO bool "Foo SoCs" depends on ARCH_MULTI_V7 These two are *not* equivalent! In the former statement, it is not ARCH_FOO, but its prompt that depends on ARCH_MULTI_V7. So, it is completely valid that ARCH_FOO is selected by another, but ARCH_MULTI_V7 is still disabled. As it is not unmet dependency, Kconfig never warns. This is probably not what you want. The former should be used only when you need to do so, and you really understand what you are doing. (In most cases, it should be wrong!) For enabling/disabling sub-architectures, the latter is always correct. As a good side effect, this commit fixes some entries over 80 columns (mach-imx, mach-integrator, mach-mbevu). [Arnd: I note that there is not really a bug here, according to the discussion that followed, but I can see value in being consistent and in making the lines shorter] Signed-off-by: Masahiro Yamada Acked-by: Maxime Ripard Acked-by: Nicolas Ferre Acked-by: Heiko Stuebner Acked-by: Patrice Chotard Acked-by: Liviu Dudau Acked-by: Krzysztof Kozlowski Acked-by: Jun Nie Acked-by: Matthias Brugger Acked-by: Simon Horman Acked-by: Gregory CLEMENT Acked-by: Shawn Guo Acked-by: Sebastian Hesselbarth Acked-by: Thierry Reding Acked-by: Krzysztof Halasa Acked-by: Maxime Coquelin Signed-off-by: Arnd Bergmann

ARM: use const and __initconst for smp_operations

2015-12-01T21:17:45+00:00

These smp_operations structures are not over-written, so add "const" qualifier and replace __initdata with __initconst. Also, add "static" where it is possible. Signed-off-by: Masahiro Yamada Acked-by: Krzysztof Kozlowski Acked-by: Maxime Ripard Acked-by: Moritz Fischer Acked-by: Stephen Boyd # qcom part Acked-by: Viresh Kumar Acked-by: Patrice Chotard Acked-by: Heiko Stuebner Acked-by: Wei Xu Acked-by: Florian Fainelli Acked-by: Sebastian Hesselbarth Acked-by: Gregory CLEMENT Acked-by: Shawn Guo Acked-by: Matthias Brugger Acked-by: Thierry Reding Acked-by: Nicolas Pitre Acked-by: Liviu Dudau Acked-by: Linus Walleij Signed-off-by: Arnd Bergmann

ARM: hisi: do not export smp_operations structures

2015-12-01T21:17:13+00:00

These three structures are only defined and referenced in mach-hisi/platsmp.c. Drop the declarations from the header and add static qualifier to the definitions. Signed-off-by: Masahiro Yamada Signed-off-by: Arnd Bergmann Acked-by: Wei Xu

ARM: hisi: Remove clk-provider.h include

2015-07-20T17:52:29+00:00

This file doesn't use the clk provider APIs. Remove the include. Cc: Haojian Zhuang Cc: Wei Xu Signed-off-by: Stephen Boyd

Merge branch 'for-arm-soc' into for-next

2015-06-12T20:18:59+00:00

ARM: v7 setup function should invalidate L1 cache

2015-06-01T10:30:26+00:00

All ARMv5 and older CPUs invalidate their caches in the early assembly setup function, prior to enabling the MMU. This is because the L1 cache should not contain any data relevant to the execution of the kernel at this point; all data should have been flushed out to memory. This requirement should also be true for ARMv6 and ARMv7 CPUs - indeed, these typically do not search their caches when caching is disabled (as it needs to be when the MMU is disabled) so this change should be safe. ARMv7 allows there to be CPUs which search their caches while caching is disabled, and it's permitted that the cache is uninitialised at boot; for these, the architecture reference manual requires that an implementation specific code sequence is used immediately after reset to ensure that the cache is placed into a sane state. Such functionality is definitely outside the remit of the Linux kernel, and must be done by the SoC's firmware before _any_ CPU gets to the Linux kernel. Changing the data cache clean+invalidate to a mere invalidate allows us to get rid of a lot of platform specific hacks around this issue for their secondary CPU bringup paths - some of which were buggy. Reviewed-by: Florian Fainelli Tested-by: Florian Fainelli Tested-by: Heiko Stuebner Tested-by: Dinh Nguyen Acked-by: Sebastian Hesselbarth Tested-by: Sebastian Hesselbarth Acked-by: Shawn Guo Tested-by: Thierry Reding Acked-by: Thierry Reding Tested-by: Geert Uytterhoeven Tested-by: Michal Simek Tested-by: Wei Xu Signed-off-by: Russell King

ARM: 8370/1: hisi: fix hip04 build without HOTPLUG_CPU

2015-05-27T23:22:49+00:00

The hip04 smp implementation provides the hotplug operations (cpu_die and cpu_kill) unconditionally at the moment, which leads to a build error when HOTPLUG_CPU is disabled: mach-hisi/platmcpm.c:242:13: note: (near initialization for 'hip04_smp_ops') mach-hisi/platmcpm.c:242:2: error: unknown field 'cpu_die' specified in initializer mach-hisi/platmcpm.c:243:2: error: unknown field 'cpu_kill' specified in initializer This uses an #ifdef to remove the code from the build when that option is not set. Fixes: 905cdf9dda5d ("ARM: hisi/hip04: remove the MCPM overhead") Signed-off-by: Arnd Bergmann Acked-by: Nicolas Pitre Acked-by: Wei Xu Signed-off-by: Russell King

ARM: hisi/hip04: remove the MCPM overhead

2015-05-06T15:42:01+00:00

This platform is currently relying on the MCPM infrastructure for no apparent reason. The MCPM concurrency handling brings no benefits here as there is no asynchronous CPU wake-ups to be concerned about (this is used for CPU hotplug and secondary boot only, not for CPU idle). This platform is also different from the other MCPM users because a given CPU can't shut itself down completely without the assistance of another CPU. This is at odds with the on-going MCPM backend refactoring. To simplify things, this is converted to hook directly into the smp_operations callbacks, bypassing the MCPM infrastructure. Tested-by: Wei Xu Cc: Haojian Zhuang Signed-off-by: Nicolas Pitre