Skip to content

ARM: OMAP: hwmod: remove runtime cpu_is checking

a0393909 linux-arm-kernel linux-omap 1 version New

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest0 of 6 patches 0 repliesNew

Patches in v1 No applies recorded yet

None of the 6 patches the cover letter announces are in this archive.

Reviews trailers per patch

No Reviewed-by, Acked-by or Tested-by trailers for v1 yet.

Read the v1 thread

Cover letter [PATCH 0/6] ARM: OMAP: hwmod: remove runtime cpu_is checking

Paul,

Quoted 31 lines · Paul Walmsley

Hi Kevin,

Quoted 8 lines · Kevin Hilman

This series attempts to remove all the runtime cpu_is* checking in omap_hwmod.c in favor of using function pointers initialized at init time.

This series was motivated by the addition of support for the AM335x series which was done by adding several more cpu_is* checks, and provided the proverbial straw that broke the camel's back.

Quoted 5 lines

In addition to the cleanup, this provides a much cleaner way of adding additional SoC support since it no longer requires adding additional runtime cpu_is* checks.

Boot tested on OMAP3530/Overo and OMAP4430/Panda.

Thanks Kevin, this series should bridge the gap nicely between the current code, and the point in time that we're able to move these functions into PRM/CM device driver code. And allow us to avoid adding more cpu_is_*() when we merge Vaibhav's hwmod changes.

As you've probably seen, the patches have been updated here. The function pointers have been made file-static rather than per-hwmod; this should conserve some memory and cache. Since they don't change on a per-hwmod basis, this should be safe. Also added quite a bit of kerneldoc -- basically I have an internal rule here now to ensure this exists on all functions and structures that are added, so it would be great to have this on future code :-)

Quoted 6 lines

The updated series is at git://git.pwsan.com/linux-2.6 in
the "hwmod_soc_conditional_cleanup_3.5" branch. Please let me know if you have any comments. It's only compile-tested here so far. I'll be rebasing my local copy of Vaibhav's series on it and will plan to send it upstream for 3.5.

I guess this series missed 3.5 merge window. I was re-basing the OMAP5 minimal support against the 3.5-rc2 and one of the dependency for that series is $subject series.

Tony,
How do we get the dependent series merged early enough so that OMAP5 support can be merged in linux-omap master to avoid any late build breaks.

Regards
santosh