Skip to content

ARM: OMAP2+: PRCM cleanups for 3.18 merge window

Tero Kristo linux-arm-kernel linux-omap 1 version New

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest0 of 26 patches 0 repliesNew

Patches in v1 No applies recorded yet

None of the 26 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 00/26] ARM: OMAP2+: PRCM cleanups for 3.18 merge window
Quoted 15 lines · Paul Walmsley
Quoted 13 lines · Tony Lindgren
Quoted 12 lines · Tero Kristo

Hi,

This set contains PRCM related cleanups meant for 3.18 merge window. These are based on top of 3.17-rc1 + the PRM set from Nishanth Menon
(http://article.gmane.org/gmane.linux.ports.arm.kernel/350305.) Nishanth's set is used as basis to avoid merge issues.

Purpose of this work is to eventually convert the PRCM code into a separate driver, but this is done in incremental parts as the amount of changes is substantial. Expected conclusion of this work is 3.19 if everything goes fine.

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

This part of the work mostly moves some of the SoC specific PRCM driver calls under generic version of the same, and adds SoC-ops to support these on the driver level.

Working branch posted here:

tree: https://github.com/t-kristo/linux-pm.git
branch: for-v3.18/prcm-cleanup

Hi,

I pushed v2 of this set as 'for-v3.18/prcm-cleanup-v2' to my tree. Basically to address the below comments, and additionally I changed a few public cm_* api names to omap_cm_* + add the tested-by/acked-by statuses. Tony/Paul, do you want me to repost the series, or are you ok with a local diff / pull only?

Quoted 6 lines
Quoted 2 lines

Paul, any comments on this series?

Patches 1, 3, 5, 8, 10, 13, 17, 18, 20, 21, 22, 23, 24 are:

Thanks, added acks to v2 patches.

Quoted 6 lines

Here are some specific comments on a few other patches:

Regarding patch 7: The kerneldoc-nano function comments are good, but should begin with "/**" rather than "/*". When that's fixed, it can be considered acked by me.

Yea, this was a typo, thanks for noticing.

Quoted 6 lines

Regarding patches 14, 15, 16: Non-static prm_* functions really should start with omap*_ to avoid potential naming conflicts with other drivers when these are moved to drivers/. (Obviously the same would apply for any CM function names in other, future patches.) When that's fixed, it can be considered acked by me.

Ok, changed. Additionally I changed some cm_* prototypes in patches #6, #7 and #9 to be of form omap_cm_*.

Quoted 2 lines

Regarding patch 25: What are "I/O wakeup gaes" -- gates? When that's fixed, an acked-by for me can be added.

Yes, gates. Fixed.

Quoted 6 lines

Regarding patch 26: It seems wise to ensure that omap_prm_reset_system() ends with a 'while(1) { cpu_relax(); }' or something similar, to ensure that execution flow does not proceed past that point. At that point, it should be possible to remove the "while(1) {}"s from omap44xx_restart(), omap2xxx_restart(), etc. When that's fixed, an acked-by for me can be added.

Ok, changed this also. Added the while(1) cpu_relax(); part to the public API, and removed the loops from everywhere else.

Quoted 6 lines

...

And some general comments: none of which should probably block this series, but seemed worth noting:

Quoted 7 lines

Regarding patches 6 and 19: Tero, could you please share the DT node data that you're planning to submit for the PRM, CM1, and CM2 on the OMAP4* platforms? According to the TRM, these are separate IP blocks, with separate OCP header register areas. So these should probably have separate DT nodes, regs, etc. if the DTS files are to match the hardware. The planned DTS data may impact the way these patches are written, at least, if more patches are to be avoided later.

We already have the nodes for these. Check out omap4.dtsi for example, we have nodes for prm/cm1/cm2/scrm there already. These were already defined when the clock DT data was introduced, and I don't foresee any changes to the nodes as of now. The nodes only contain register space info as of now, and Nishanth is adding the interrupt property for PRM interrupt purposes, but rest of the features are probably going to be hardcoded within the PRCM drivers themselves.

If you are interested, feel free to look at for-3.19/prcm-cleanup branch in my git tree, this contains the remaining patches I have for PRCM cleanup after this set available as of now.

-Tero

Quoted 7 lines

As far as patches 2, 4, 9, 11, and 12 go, I'll let those go without comment.

- Paul